opics LYMPHOMA · 7/15/2020 · Reem Karmali, Harry Miao, Igor Proscurshim, Shining Wang, Yujun...
129
LYMPHOMA Recent Articles Published in the AACR Journals Complimentary access to articles online: aacrjournals.org/hot-topics
Transcript of opics LYMPHOMA · 7/15/2020 · Reem Karmali, Harry Miao, Igor Proscurshim, Shining Wang, Yujun...
LYMPHOMA Recent Articles Published in the AACR Journals
Complimentary access to articles online:
aacrjournals.org/hot-topics
Cross-Journal Collection:Lymphoma
Table of Contents
Birth Order, Sibship Size, Childhood Environment and Immune-Related Disorders, and Risk of Lymphomain Lymphoid Cancer FamiliesSamantha J. Jones, Sumara Stroshein, Amy M. Williams, Dongmeng Liu, John J. Spinelli, Joseph M. Connors, andAngela R. Brooks-WilsonCancer Epidemiol Biomarkers Prev June 1, 2020 29:6 1168–78; doi: 10.1158/1055-9965.EPI-19-1204
Genomic Characterization of HIV-Associated Plasmablastic Lymphoma Identifies Pervasive Mutationsin the JAK–STAT PathwayZhaoqi Liu, Ioan Filip, Karen Gomez, Dewaldt Engelbrecht, Shabnum Meer, Pooja N. Lalloo, Pareen Patel, Yvonne Perner,Junfei Zhao, Jiguang Wang, Laura Pasqualucci, Raul Rabadan, and Pascale WillemBlood Cancer Discov Jul 1, 2020 1:1 112–25; doi: 10.1158/2643-3249.BCD-20-0051
Gene Expression Profiling of Mediastinal Gray Zone Lymphoma and Its Relationship to PrimaryMediastinal B-cell Lymphoma and Classical Hodgkin LymphomaStefania Pittaluga, Alina Nicolae, George W. Wright, Christopher Melani, Mark Roschewski, Seth Steinberg, DaWei Huang,Louis M. Staudt, Elaine S. Jaffe, and Wyndham H. WilsonBlood Cancer Discov Sep 1, 2020 1:2 1–7; doi: 10.1158/2643-3230.BCD-20-0009
Single-Cell Transcriptome Analysis Reveals Disease-Defining T-cell Subsets in the TumorMicroenvironment of Classic Hodgkin LymphomaTomohiro Aoki, Lauren C. Chong, Katsuyoshi Takata, Katy Milne, Monirath Hav, Anthony Colombo, Elizabeth A. Chavez,Michael Nissen, Xuehai Wang, Tomoko Miyata-Takata, Vivian Lam, Elena Viganò, Bruce W. Woolcock, Ad�ele Telenius,Michael Y. Li, Shannon Healy, Chanel Ghesquiere, Daniel Kos, Talia Goodyear, Johanna Veldman, Allen W. Zhang, Jubin Kim,Saeed Saberi, Jiarui Ding, Pedro Farinha, Andrew P. Weng, Kerry J. Savage, David W. Scott, Gerald Krystal, Brad H. Nelson,Anja Mottok, Akil Merchant, Sohrab P. Shah, and Christian SteidlCancer Discov Mar 1, 2020 10:3 406–21; doi: 10.1158/2159-8290.CD-19-0680
Editors of the AACR journals reviewed recently published content toidentify hot topics across the entire portfolio. This publication focuseson lymphoma and highlights articles based on a number of key metrics,such as usage and citations. We hope that you enjoy this complimentarycross-journal collection.
Lymphoma
The Th9 Axis Reduces the Oxidative Stress and Promotes the Survival of Malignant T Cells inCutaneous T-Cell Lymphoma PatientsSushant Kumar, Bhavuk Dhamija, Soumitra Marathe, Sarbari Ghosh, Alka Dwivedi, Atharva Karulkar, Neha Sharma, Manju Sengar,Epari Sridhar, Avinash Bonda, Jayashree Thorat, Prashant Tembhare, Tanuja Shet, Sumeet Gujral, Bhausaheb Bagal,Siddhartha Laskar, Hasmukh Jain, and Rahul PurwarMol Cancer Res Apr 1, 2020 18:4 657–68; doi: 10.1158/1541-7786.MCR-19-0894
IL1R8 Deficiency Drives Autoimmunity-Associated Lymphoma DevelopmentFederica Riva, Maurilio Ponzoni, Domenico Supino, Maria Teresa Sabrina Bertilaccio, Nadia Polentarutti, Matteo Massara,Fabio Pasqualini, Roberta Carriero, Anna Innocenzi, Achille Anselmo, Tania Veliz-Rodriguez, Giorgia Simonetti, Hans-Joachim Anders,Federico Caligaris-Cappio, Alberto Mantovani, Marta Muzio, and Cecilia GarlandaCancer Immunol Res June 1, 2019 7:6 874–85; doi: 10.1158/2326-6066.CIR-18-0698
Lymphoma Angiogenesis Is Orchestrated by Noncanonical Signaling PathwaysMarleen Gloger, Lutz Menzel, Michael Grau, Anne-Clemence Vion, Ioannis Anagnostopoulos, Myroslav Zapukhlyak, Kerstin Gerlach,Thomas Kammert€ons, Thomas Hehlgans, Maria Zschummel, Georg Lenz, Holger Gerhardt, Uta E. H€opken, and Armin RehmCancer Res Mar 15, 2020 80:6 1316–29; doi: 10.1158/0008-5472.CAN-19-1493
Selective Inhibition of HDAC3 Targets Synthetic Vulnerabilities and Activates Immune Surveillance inLymphomaPatrizia Mondello, Saber Tadros, Matt Teater, Lorena Fontan, Aaron Y. Chang, Neeraj Jain, Haopeng Yang, Shailbala Singh,Hsia-Yuan Ying, Chi-Shuen Chu, Man Chun John Ma, Eneda Toska, Stefan Alig, Matthew Durant, Elisa de Stanchina,Sreejoyee Ghosh, Anja Mottok, Loretta Nastoupil, Sattva S. Neelapu, Oliver Weigert, Giorgio Inghirami, Jos�e Baselga, Anas Younes,Cassian Yee, Ahmet Dogan, David A. Scheinberg, Robert G. Roeder, Ari M. Melnick, and Michael R. GreenCancer Discov Mar 1, 2020 10:3 440–59; doi: 10.1158/2159-8290.CD-19-0116
Phase I Study of TAK-659, an Investigational, Dual SYK/FLT3 Inhibitor, in Patients with B-Cell LymphomaLeo I. Gordon, Jason B. Kaplan, Rakesh Popat, Howard A. Burris III, Silvia Ferrari, Sumit Madan, Manish R. Patel, Giuseppe Gritti,Dima El-Sharkawi, Ian Chau, John A. Radford, Jaime P�erez de Oteyza, Pier Luigi Zinzani, Swaminathan Iyer, William Townsend,Reem Karmali, Harry Miao, Igor Proscurshim, Shining Wang, Yujun Wu, Kate Stumpo, Yaping Shou, Cecilia Carpio, and Francesc BoschClin Cancer Res Jul 15, 2020 26:14 3546–56; doi: 10.1158/1078-0432.CCR-19-3239
Tumor Microenvironment Composition and Severe Cytokine Release Syndrome (CRS) Influence Toxicity inPatients with Large B-Cell Lymphoma Treated with Axicabtagene CiloleucelRawan Faramand, Michael Jain, Verena Staedtke, Hiroshi Kotani, Renyuan Bai, Kayla Reid, Sae Bom Lee, Kristen Spitler, Xuefeng Wang,Biwei Cao, Javier Pinilla, Aleksander Lazaryan, Farhad Khimani, Bijal Shah, Julio C. Chavez, Taiga Nishihori, Asmita Mishra,John Mullinax, Ricardo Gonzalez, Mohammad Hussaini, Marian Dam, Brigett D. Brandjes, Christina A. Bachmeier, Claudio Anasetti,Frederick L. Locke, and Marco L. DavilaClin Cancer Res 2020; doi: 10.1158/1078-0432.CCR-20-1434
To read a full-text article within this pdf, please click on its title above. While viewing the full-text article, you may access it onlineby clicking its title on the article’s title page.
Lymphoma
CANCER EPIDEMIOLOGY, BIOMARKERS & PREVENTION | RESEARCH ARTICLE
Birth Order, Sibship Size, Childhood Environment andImmune-Related Disorders, and Risk of Lymphoma inLymphoid Cancer FamiliesSamantha J. Jones1,2, Sumara Stroshein1, Amy M. Williams1, Dongmeng Liu3, John J. Spinelli4,5,Joseph M. Connors6, and Angela R. Brooks-Wilson1,7
ABSTRACT◥
Background: Familial aggregation of lymphoid cancers andimmune-related disorders suggests a role for genetic susceptibility;however, few studies examine environmental factors. According tothe hygiene hypothesis, adult-onset immune-related diseases maybe a consequence of reduced childhood infectious exposures andaberrant immune development. In a cohort of 196 multiple-caselymphoid cancer families, we analyzed environmental factors relat-ed to the hygiene hypothesis.
Methods: Family structure, childhood environment, andimmune-related disorders were examined among 196 lymphoidcancer families, in relation to risk of lymphoid cancer. We reporton 450 lymphoid cancer cases and 1,018 unaffected siblingsusing logistic regression models with generalized estimatingequations to estimate ORs and 95% confidence intervals (CI)for association.
Results: The risk of lymphoma tended to decrease with later birthorder (OR¼ 0.83; 95% CI, 0.78–0.89) and larger sibship size (OR¼0.82; 95% CI, 0.79–0.85). High maternal education, above averagefamily income during childhood, allergies (OR¼ 2.25; 95% CI, 1.44–3.51), and tonsillectomy (OR ¼ 1.78; 95% CI, 1.14–2.78) wereindependent risk factors for lymphoma. Familial lymphoid cancercases were more likely to report environment (OR ¼ 1.90; 95% CI,1.21–2.98) and drug (OR ¼ 2.30; 95% CI, 1.41–3.73) allergies.
Conclusions: These associations underscore the complex etiologyof familial lymphoma. To our knowledge, this is the largest multiple-case family-based study that supports the hygiene hypothesis con-tributing to lymphoid cancer risk.
Impact: Understanding the mechanism by which environmentaland lifestyle factors affect lymphoid cancer risk may advance cancerprevention, even in the familial context.
IntroductionLymphoid cancers are a heterogeneous group of neoplasms that
arise from immune cells. Collectively, they represent the fifth highestglobal incidence of cancer. Established risk factors include older age,male sex, ethnicity, compromised immune function, and familyhistory of lymphoproliferative disorders (LPD; refs. 1, 2). Low-penetrance common genetic polymorphisms that affect pathwaysrelated to DNA integrity, B-cell growth, and survival and xenobioticmetabolism have also been implicated (3–5). Early-life environmentmay modulate risk of immune-related disorders, such as allergies andautoimmune conditions as well as some lymphoid cancers (6).
The hygiene hypothesis proposes that an early life environment thathas a relative lack of exposure to microorganisms and infectiousdisease inhibits a child's immune system frommaturing optimally (6).Consequently, such individuals are more susceptible to adult-onsetimmune-related disorders. Measures of family structure and crowdingrelate to the hygiene hypothesis as they may affect age and extent ofexposure to infectious diseases, with low birth order, and smallerfamilies correlating with higher risk.
Associations between early birth order and/or smaller sibship sizeand increased risk of lymphoma have been reported for lymphoidcancers collectively (7), and separately for non-Hodgkin lymphoma(NHL; refs. 7–11) and Hodgkin lymphoma (HL; refs. 7, 12–16).However, many other studies report no association between familystructure and risk of NHL (10, 16–19), HL (10–12, 16, 19, 20),chronic lymphocytic leukemia (CLL; refs. 16, 17, 21), or multiplemyeloma (MM; refs. 7, 16, 19). A few studies have observed apositive association between later birth order and NHL (17, 21, 22),and larger sibship size and risk of NHL (17, 18, 21, 22), andMM (16). The discordant findings among studies may be partlyexplained by variations in study design, study population, partic-ipant response rate, selection bias, hematologic subtypes assessed,and socioeconomic status (SES; ref. 23).
Few studies have examined family structure and environmentalfactors in the context of multiple-case lymphoid cancer families.Jønsson and colleagues observed a paternal birth order effect among24 parent–offspring pairs in 32 families enriched for CLL and B-cellmalignancies (24). Royer and colleagues (25) found that familialWaldenstr€om macroglobulinemia (WM) cases were more likely tohave immune-related disorders such as autoimmune diseases, aller-gies, and specific infections among 103 familial WM and related B-celldisorders (25).
1Canada's Michael Smith Genome Sciences Centre, British Columbia Cancer,Vancouver, British Columbia, Canada. 2Department of Medical Genetics, Uni-versity of British Columbia, Vancouver, British Columbia, Canada. 3Departmentof Statistics and Actuarial Science, Simon Fraser University, Burnaby, BritishColumbia, Canada. 4Population Oncology, British Columbia Cancer, Vancouver,British Columbia, Canada. 5School of Population and Public Health, University ofBritish Columbia, Vancouver, British Columbia, Canada. 6Centre for LymphoidCancer, British Columbia Cancer, Vancouver, British Columbia, Canada. 7Depart-ment of Biomedical Physiology and Kinesiology, Simon Fraser University,Burnaby, British Columbia, Canada.
Note: Supplementary data for this article are available at Cancer Epidemiology,Biomarkers & Prevention Online (http://cebp.aacrjournals.org/).
Corresponding Author: Angela R. Brooks-Wilson, BC Cancer, 675 West 10thAvenue, Vancouver, British Columbia V5Z 1L3, Canada. Phone: 604-675-8153;Fax: 604-675-8178; E-mail: [email protected]
Cancer Epidemiol Biomarkers Prev 2020;29:1168–78
doi: 10.1158/1055-9965.EPI-19-1204
�2020 American Association for Cancer Research.
AACRJournals.org | 1168
Currently, there remains a large gap in our understanding of theetiology of familial lymphoid cancers (24). We examined familystructure, childhood lifestyle, and immune-related disorders amonga large collection ofmultiple-case lymphoid cancer families, in relationto risk of lymphoid cancer.
Materials and MethodsStudy population
This study was approved by the BC Cancer-University of BritishColumbia Clinical Research Ethics Board. All participants providedwritten informed consent. Families were eligible for inclusion if theycontained a member diagnosed with lymphoma and at least oneadditional relative with a lymphoid cancer. Patients with the cancersof interest, all of their first-degree relatives, and additional relativesthat connect affected family members were invited to participate.Participants were recruited by physician-, self-, or genetic coun-sellor referral. Family participation was not limited to withinCanada, although most families were identified through a memberresiding in British Columbia, Canada. Families were ascertainedbetween 2006 and 2018.
Data collectionInformation about lymphoid malignancies, family structure, and
demographics was obtained systematically using a questionnaire andphone interviews with multiple family members. Family structure andearly lifestyle environment information, such as parental education,family income, farm residence, and urban/rural residential locationwas reported by sibship. Personal information regarding education,medical history (allergies, autoimmune diseases, surgical procedures)and early lifestyle data was obtained from a self-administered ques-tionnaire. Allergies were classified as drug, environmental, or food.Autoimmune diseases were categorized as systemic, organ-specific, orconditions without detectable autoantibodies (25, 26).
We report on 196 families with 524 lymphoid cancer cases among418 sibships. Of these 418 sibships, 52 lacking family structure (birthorder, sibship size) and 17 only-child cases were excluded. Theremaining sibships contained 453 cases and 1,112 siblings, fromwhich3 (0.7%) cases and 94 (8%) siblings were removed due tomissing age ofenrollment or sex. Analyses were conducted on 450 cases and 1,018siblings among 346 sibships.
Lymphoid cancer diagnoses were confirmed histologically orthrough medical records for 241 of the 450 (54%) cases. Cancers wereclassified according to the InterLymph hierarchical classification oflymphoid neoplasms for epidemiologic research (27).
Statistical analysisOur study examined multiple-case families with a history of hema-
tologic malignancies and does not represent a population-basedcollection.
A x2 goodness-of-fit test was performed to assess whether theobserved sex distribution of the families resembled that of the Cana-dian population (28). American population datawere used in instanceswhere distinct histologic subtype information was unavailable (29).
Standard logistic regression with generalized estimating equationThe relationship between lifestyle factors and risk of lymphoid
cancer was examined using logistic regression with a generalizedestimating equation (GEE) to accommodate correlated family data.ORs and 95% confidence intervals (CI) were clustered by family andadjusted for age (continuous) and sex. Potential confounding effects of
ethnicity did not change risk estimates ≥ 10% and were not retained inthe final analysis. Independent, exchangeable, and autoregressivecorrelation structures performed similarly; the autoregressive corre-lation structure was used in subsequent analyses. Covariates assessedinclude sex, age of enrollment (n¼ 1,468), highest level of participanteducation (n ¼ 494), maternal and paternal education (n ¼ 759 andn ¼ 770, respectively), family income during childhood (n ¼ 756),childhood farm residence (n ¼ 801), childhood residential location(n¼ 751), allergies (n¼ 354), asthma (n¼ 378), autoimmune diseases(n¼ 378), appendectomy (n¼ 353), and tonsillectomy (n¼ 353). Ageof death was used in replacement of age of enrollment for nonlivingparticipants. Individuals withmissing age, sex, or family structure datawere removed from the dataset. Because of their structural depen-dence, birth order and sibship size were investigated using separateGEEmodels. Covariates were independently assessedwithin eachGEEmodel. Statistical analysis was performed using R version 3.5.
Because of the variability between age-of-onset patterns, additionalanalyses were done with HL cases separated into childhood, young-adult, or adult onset, according to Cozen and colleagues (age-of-diagnosis ≤50, and <40; ref. 30) and Westergaard and colleagues(age-of-diagnosis <15, 15–42, and >42; ref. 15). Results from thesesensitivity analyses were essentially identical (SupplementaryTable S1) to those of all HL cases together; only outcomes based onall HL cases are presented. Additional analyses were stratified bysibship size (≤3 siblings and >3 siblings) according to Cozen andcolleagues (ref. 18; Supplementary Table S2).
Stepwise regressionThe availability of early life and disease information varied because
not all members of a given sibship completed the self-administeredquestionnaire. For example, if one member of a sibship reported themother's level of education, that datum was applied to all members ofthe sibship. In contrast, allergy information was only available if it wasself-reported. The variables had different amounts of missing data andso they were separated into three groups of comparable sample size toretain the most information. Three GEE models were built in astepwise manner to investigate the relationships between lymphoidcancer risk and lifestyle factors. Complete family structure (birth orderand sibship size), age of enrollment, and sex information was availablefor 1,468 participants, which constituted the base model. The middlemodel contained the basemodel variables and childhood environmentvariables:maternal education, paternal education, family income, farmresidence, and residential location. The full model was comprised ofthe middle model variables in addition to education, allergies, auto-immune diseases, asthma, appendectomy, and tonsillectomy.
Permutation testsThe association between family structure and lymphoid cancer risk
was also evaluated using standard logistic regression and x2 tests fortrend using permuted pairs of independent cases and controls. Eachfamily member was assigned a generation number relative to thefounding lymphoid cancer case or presumed carrier (31). One family(of 196) was excluded from permutations due to an unmatchedgeneration variable. The data set supported the random sampling of95 generation-matched case/control pairs of independent families,such that a maximum of one individual per family was selected(without resampling). Ninety-five pairs were permuted 10,000 timesin quadruplicate for use in: (i) logistic regression with birth order, (ii)logistic regression with sibship size, (iii) a x2 test for trend inproportion with birth order, and (iv) a x2 test for trend in proportionwith sibship size. The 10,000 P values and coefficient estimates from
Early Life Factors and Risk of Lymphoma in Families
AACRJournals.org Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 1169
each permuted analysis were compared with those observed with thefull family dataset. The logistic models contained base model variables(age of enrollment, sex, and birth order or sibship size).
ResultsWe report on 346 sibships with a lymphoid cancer–affected indi-
vidual within 196 multiple-case families (Table 1). The median age ofenrollment for cases and unaffected siblings was 62 and 63 years,respectively. 241 (54%) cases were confirmed histologically or bymedical records, all of which supported the reported diagnosis.Familial cases were 54% male; in comparison, Canadian populationNHL, HL, CLL, and MM cases are 55%, 57%, 61%, and 59% male,respectively (28). The sex distributions of familial MM (P ¼ 0.0454)and HL (P¼ 0.0126) cases were significantly less frequently male thanpopulation cases.
Family structureBirth order was inversely associated with lymphoid cancer (OR ¼
0.83; 95% CI, 0.78–0.89), such that earlier birth order has a higher riskof lymphoma (Tables 2 and 3). TheORswere 0.62 (95%CI: 0.41–0.82)for fourth born compared with first-born individuals, and 0.41 (95%CI: 0.30–0.57) for fifth or later born compared with first born. We alsoobserved a strong inverse relationship between sibship size andlymphoid cancer (OR ¼ 0.82; 95% CI, 0.79–0.85); smaller sibshipshad a higher risk of lymphoma. The OR was 0.58 (95% CI, 0.46–0.72)for sibships of 3, compared with sibships of 2. TheORs for sibships of 4and 5were 0.39 and 0.23, respectively. The adjusted effect estimates forbirth order changed slightly after stratifying by sibship size (Supple-mentary Table S2).
Table 3 shows the effects of family structure on the risk of lymphoidcancer types. Larger sibships were significantly associated with a lowerrisk of lymphoma and several histologic subtypes. Birth order wasinversely associated with risk of most major lymphoma entities [NHL,B-cell NHL, CLL, follicular lymphoma (FL), Mantle cell lymphoma(MCL), marginal zone lymphoma (MZL), and MM] but was notsignificant for HL, diffuse large B-cell lymphoma (DLBCL), or T-cellNHL. We observed no differences in the risk patterns associated withchildhood, young-adult, or older-adult onset HL (SupplementaryTable S1). Sibship size and birth order effects were similar amongfamilies with 2, 3, or 4 or more lymphoid cancer cases.
To estimate the probability of a chance association with birthorder position or sibship size, 40,000 independent permutation testswere performed. Despite a smaller sample size and lower power ofpermuted data (n ¼ 95 case/control pairs), both the regression andx2 tests for trend supported our findings, with approximately 51%and 93% of P values achieving statistical significance (P < 0.05) forbirth order and sibship size, respectively (Supplementary Fig. S1).Without the family dependence, the OR estimates remainedcomparable with those from GEE models for sibship size (medianOR ¼ 0.82) and birth order (median OR ¼ 0.82), validating ourobservations.
Early-life environment and immune-related diseasesHigher maternal education and an above average level of income
during childhood was associated with increasing risk of lymphoidcancer (Table 4). Childhood farm residents had a lower risk oflymphoma (OR¼ 0.65; 95% CI, 0.48–0.88), which was not significantafter adjusting for sibship size (OR ¼ 0.87; 95% CI, 0.70–1.08). Caseswere less likely than their unaffected siblings to have a post-secondaryeducation (OR ¼ 0.62; 95% CI, 0.38–0.99; Table 4), even when
adjusting for family structure. There was no relationship betweenpaternal education or childhood house location (urban vs. rural) andlymphoma or subtypes.
Allergies and tonsillectomy were independent risk factors for mostmajor lymphoma entities (Table 4). Lymphoid cancer risk wasincreased for individuals with environmental (e.g., hay fever) anddrug allergies for several lymphoma entities, whereas food allergieswere exclusively associated with risk of Nodular sclerosis classical HL(NSCHL; Table 4). History of appendectomy was significantly asso-ciated with a 9.7-fold increase in risk of DLBCL. Asthma was notsignificantly associated with risk of lymphoma with the exception ofMMwhere small sample sizemakes it unclear. Therewas no significantassociation between personal history of collective autoimmune dis-eases and lymphoma. However, familial lymphoid cancer cases weresignificantly less likely than their siblings to have had organ-specificautoimmune disease (OR¼ 0.44; 95%CI, 0.20–0.98) after adjusting forsibship size.
Stepwise model selectionThree GEE models encompassed selection of family structure,
early-life environment, and immune-related disorders. Because theavailability of lifestyle and disease information varied among par-ticipants, three GEE models were built in a stepwise manner(Table 5). The base model contained 1,468 individuals, in whichbirth order and sibship size were independent significant predictorsof lymphoid cancer status. The middle model (n ¼ 682) retainedfamily income during childhood in addition to birth order andsibship size as significant predictors of lymphoid cancer status,while the full model (n ¼ 321) included allergies, autoimmunediseases, tonsillectomy, and family structure, with maternal edu-cation included in the sibship size (but not birth order) model.
DiscussionWe assessed associations of family structure and childhood
environment with disease in families with multiple lymphoidcancers. We observed an inverse relationship between birth orderand cancer risk that was similar for lymphoid cancers collectivelyand most major subtypes (NHL, CLL, FL, and MM). Sibship sizewas also inversely associated with risk of lymphoma and all entities,with the exception of lymphoplasmacytic lymphoma. High mater-nal education, above average income during childhood, allergies,and tonsillectomy were independent risk factors for lymphoma. Toour knowledge, this is the largest multiple-case family study to datethat supports the hygiene hypothesis contributing to lymphoidcancers.
Familial lymphoid cancer casesweremore likely to bemale, which isconsistent among population studies (7–10, 12, 17, 18, 23, 32) but notalways true among multiple-case family studies (31, 33, 34). In thisstudy, NHL (and subtypes) resembled the population sex distribu-tion (28); however, HL andMMcaseswere significantly less likely to bemale. Familial cases may reflect a different and potentially moregenetic etiology in comparison with population cases (of which themajority are sporadic; ref. 31). Lower rates of B-cell malignancies inwomen may be attributed to nongenetic factors including body sizeand sex/reproductive hormones (35).
In this study, individuals earlier in birth order and/or of smallersibships had higher risk of lymphoma. With the exception of a fewstudies (16, 17, 21–23),many have shown inverse associations betweenbirth order and lymphoid cancers, including NHL (7, 8, 11),HL (7, 12, 14–16), and major subtypes (CLL, DLBCL, FL; refs. 7, 8).
Jones et al.
Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 CANCER EPIDEMIOLOGY, BIOMARKERS & PREVENTION1170
Opposite risk patterns for childhood- and adult-onset HL have beendocumented in population-based studies (15); however, we observedno difference in risk among young-adult or older-adult HL, whilechildhood-onset cases were limited in sample size. In our study, largersibships were protective of lymphoid cancers, suggesting thatmultiple-case families may have a different disease etiology than sporadic/nonfamilial lymphomas in the population. Our finding that MCL,MZL, and mucosa-associated lymphoid tissue (MALT) lymphomawere more frequent among earlier born siblings has not been previ-ously reported. Birth order and family size are inevitably correlatedand distinguishing between their effects is difficult. Generally, eldestsiblings receive more prenatal care and medical surveillance, and maybe better nourished than later born siblings (36). Children fromsmaller sibships are traditionally of higher SES and have an older ageat first bacterial or viral disease (36, 37).
Familial predisposition to lymphoma has been extensively inves-tigated, but few studies examine the effect of family size in multiple-case families (31). The effects of birth order were similar amongfamilies with 2, 3, and 4 or more lymphoid cancer cases. We wouldexpect families with more cases to have a more genetic etiology.Because the effects of birth order and sibship size do not vary with
Table 1. Demographic characteristics and family structure of participants, by lymphoid cancer status.a
Lymphoid affected, n (%)Subtypes
Characteristic Unaffected, n (%) All types NHL HL CLL MM Total, n
Total 1,018 (69.3) 450 (30.7) 221 (49.1) 70 (15.6) 133 (30.0) 26 (5.8) 1,468Sex
Male 510 (50.1) 242 (53.8) 124 (56.1) 29 (41.4) 79 (59.4) 10 (38.5) 752Female 508 (49.9) 208 (46.2) 97 (43.9) 41 (58.6) 54 (40.6) 16 (61.5) 716
Age of enrollment (y)b
Mean � SD 61.7 � 19.0 61.1 � 17.3 62.2 � 16.9 44.8 � 17.3 67.0 � 12.8 66.2 � 13.3 61.5 � 18.5Median 63 62 62 42 66 67 63Range 0.5–108 3–104 3–104 14–95 24–93 33–86 0.5–108<40 116 (11.4) 55 (12.2) 18 (8.1) 30 (42.9) 5 (3.8) <5 17140–49 109 (10.7) 42 (9.3) 23 (10.4) 14 (20.0) <5 <5 15150–59 207 (20.3) 90 (20.0) 51 (23.1) 14 (20.0) 24 (18.0) <5 29760–69 213 (20.9) 110 (24.4) 51 (23.1) 5 (3.8) 43 (32.3) 11 (42.3) 32370–79 200 (19.6) 94 (20.9) 44 (19.9) <5 38 (28.6) 8 (30.8) 249≥80 173 (17.0) 59 (13.1) 34 (15.4) <5 19 (14.3) <5 232
Birth orderFirst born 208 (20.4) 128 (28.4) 66 (29.4) 18 (25.7) 35 (26.3) 9 (34.6) 336Second born 226 (22.2) 106 (23.6) 54 (24.4) 18 (25.7) 26 (19.5) 8 (30.8) 332Third born 169 (16.6) 97 (21.6) 51 (23.1) 14 (20.0) 28 (21.1) <5 266Fourth born 138 (13.6) 51 (11.3) 22 (10.0) 8 (11.4) 19 (14.3) <5 189Fifth or later born 277 (27.2) 68 (15.1) 28 (12.7) 12 (17.1) 25 (18.8) <5 345
Sibship sizeTwo 58 (5.7) 72 (16.0) 36 (16.3) 17 (24.3) 14 (10.5) <5 130Three 131 (12.9) 97 (22.6) 47 (21.3) 19 (27.1) 21 (15.8) 10 (38.5) 228Four 180 (17.7) 90 (20.0) 51 (23.1) 10 (14.3) 27 (20.3) <5 270Five or more 649 (63.8) 191 (42.4) 87 (39.4) 24 (34.3) 71 (53.4) 9 (34.6) 840
Ethnicityc
White 967 (95.0) 431 (95.8) 209 (94.6) 68 (97.1) 132 (99.2) 22 (84.6) 1,398Other 51 (5.0) 19 (4.2) 12 (5.4) <5 <5 <5 70
No. affected per familyTwo 424 (41.7) 193 (42.9) 96 (43.4) 21 (30.0) 62 (46.6) 14 (53.8) 617Three 331 (32.5) 152 (33.8) 85 (38.5) 23 (32.9) 39 (29.3) 5 (19.2) 483Four or more 263 (25.8) 105 (23.3) 40 (18.1) 26 (37.1) 32 (24.1) 7 (26.9) 368
Abbreviations: CLL, chronic lymphocytic leukemia; HL, Hodgkin lymphoma; MM, multiple myeloma; NHL, non-Hodgkin lymphoma.aCells with <5 were suppressed for privacy.bAge at death was used for non-living participants. Family members missing age at enrollment (or age at death) were excluded.cEthnicity was classified according to SEER program race recode groups accessed through SEER�Stat (1).
Table 2. ORs for risk of lymphoma according to birth orderposition and sibship size.
Variable OR (95% CI)a,b
Birth orderFirst born 1.00 (Referent)Second born 0.76 (0.53–1.08)Third born 0.92 (0.65–1.27)Fourth born 0.62 (0.41–0.82)Fifth or later born 0.41 (0.30–0.57)
Sibship sizeTwo 1.00 (Referent)Three 0.58 (0.46–0.72)Four 0.39 (0.31–0.48)Five or more 0.23 (0.18–0.28)
aAdjusted for age at enrollment (continuous) and sex (male/female). Age atdeath was used for nonliving participants.bOR and 95% CI estimated by GEE logistic regression (clustered by family) withan autoregressive correlation structure. Bold type, 95% CI does not include 1.00,denoting a significant association.
Early Life Factors and Risk of Lymphoma in Families
AACRJournals.org Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 1171
the number of affected individuals in the family, we suggest that theycorrelate with an exposure that affects lymphoid cancer risk.
Indicators of infectious exposures that are correlated with child-hood SES were also supportive of the hygiene hypothesis, such thatindividuals with a high childhood SES were at an elevated risk oflymphoid cancer (38). Strong indicators of childhood SES includeparental education and income as they capture knowledge-relatedbehaviors that influence the age, extent, and response to infectiousagents (39). More protected or cleaner environments associated withhigher SES may delay infectious exposure and increase adult-onsetimmune-related disease risk, which is consistent with population-based associations (14, 17, 40), and our observations. In this study,childhood farm residents had a lower risk of lymphoma and FL, whichis consistent with the hygiene hypothesis and epidemiologic popula-tion-based studies (11, 41). Early and frequent farm visits and animalcontact (0–4 years of age) are thought to trigger an early immuneresponse and strong immune competence suggested to prevent child-hood lymphomas (11, 41, 42). Our study did not differentiate betweenage at childhood exposure, nor include farm-related exposures as anadult.
There is limited and contradictory information on the associationsbetween education and risk of lymphoma (43). In this study, familialcases were more likely to have lower educational attainment than theirunaffected siblings, which is consistent with sporadic DLBCL (43) andMM (43, 44) cases, but not all population-based studies (14, 38, 43).
The relationship between education and lymphoid cancer is complexand may be influenced by age of diagnosis, treatment regimens, andchildhood SES. Lower educational attainment may be attributed toneurocognitive impairments from chemotherapy as observed amongsurvivors of childhood HL, adult breast, and primary central nervoussystem lymphoma (45).
Familial cases were significantly more likely than their unaffectedsiblings to report a history of allergies and a tonsillectomy, which mayindicate defective immune regulation (46). A positive associationbetween lymphoid cancer and a tonsillectomy has been describedamong population NHL (47, 48), HL (47–49), and CLL (47) cases, butnot among multiple-case families (25). A tonsillectomy in youngerchildren may indicate severe recurrent tonsillitis (47, 49) caused by analtered or impaired immune response that affects lymphogenicmechanisms in adulthood (47). Lymphoid cancer risk may be morepronounced in tonsillectomized children because of the decliningimmunologic function of the tonsils from early childhood to adult-hood (48, 49). Viruses, such as EBV, have been implicated in this role,as it is associated with recurrent bouts of tonsillitis (40, 47, 48). Theassociation between tonsillectomy and lymphomamay be confoundedby SES, educational attainment, and family size although conflictingevidence has been reported (14, 38, 48).
An elevated risk of allergies has been observed among case–control (7, 14, 18) and cohort studies (46, 50–52) of nonfamilialoccurrences of lymphoma (50), includingNHL (18, 46, 51) andmature
Table 3. Associations between family structure and cancer risk by type and family size.a
Individuals within families, n (%) OR (95% CI)b,c
VariableFamilies,n (%)
Unaffectedsiblings
Lymphoidaffected Total Birth order Sibship size
Entityd
All types 196 (100) 1,018 (100) 450 (100) 1,468 (100) 0.83 (0.78–0.89) 0.82 (0.79–0.85)Lymphoid neoplasms 190 (96.9) 991 (97.3) 424 (94.2) 1,415 (96.4) 0.83 (0.77–0.89) 0.82 (0.80–0.85)
NHL 175 (89.3) 883 (86.7) 354 (78.7) 1,237 (84.3) 0.80 (0.75–0.87) 0.82 (0.79–0.84)B-cell NHL 162 (82.7) 753 (74.0) 307 (68.2) 1,060 (72.2) 0.80 (0.73–0.87) 0.81 (0.78–0.84)
CLL 81 (41.3) 357 (35.1) 133 (29.6) 490 (33.4) 0.88 (0.78–0.98) 0.84 (0.80–0.87)DLBCL 28 (14.3) 89 (8.7) 30 (6.7) 119 (8.1) 0.93 (0.74–1.17) 0.78 (0.70–0.87)FL 43 (21.9) 155 (15.2) 54 (12.0) 209 (14.2) 0.76 (0.62–0.93) 0.67 (0.61–0.74)Lymphoplasmacytic 13 (6.6) 34 (3.3) 17 (3.8) 51 (3.5) 0.88 (0.59–1.33) 0.75 (0.46–1.21)MCL 6 (3.1) 34 (3.3) 7 (1.6) 41 (2.8) 0.51 (0.39–0.68) 0.83 (0.81–0.86)MZL 9 (4.6) 41 (4.0) 10 (2.2) 51 (3.5) 0.47 (0.30–0.74) 0.82 (0.77–0.88)
MALT <5 25 (2.5) 5 (1.1) 30 (2.0) 0.56 (0.34–0.94) 0.81 (0.79–0.84)T-cell NHL 9 (4.6) 35 (3.4) 9 (2.0) 44 (3.0) 1.17 (0.90–1.51) 0.76 (0.67–0.85)
HL 46 (23.5) 185 (18.2) 70 (15.6) 255 (17.4) 0.93 (0.80–1.12) 0.83 (0.78–0.90)Classic HL 46 (23.5) 173 (17.0) 67 (14.9) 240 (16.3) 1.00 (0.86–1.12) 0.80 (0.73–0.89)
Nodular sclerosis 22 (11.2) 60 (5.9) 28 (6.2) 88 (6.0) 0.99 (0.76–1.30) 0.71 (0.61–0.82)MM 19 (9.7) 67 (6.6) 26 (5.8) 93 (6.3) 0.70 (0.52–0.94) 0.78 (0.65–0.95)
No. affected per familyTwo 107 (54.6) 426 (68.8) 193 (31.2) 619 (42.2) 0.83 (0.75–0.93) 0.79 (0.75–0.83)Three 61 (31.1) 332 (68.6) 152 (31.4) 484 (33.0) 0.73 (0.63–0.84) 0.81 (0.76–0.86)Four or more 28 (14.3) 260 (71.2) 105 (28.8) 365 (24.9) 0.94 (0.85–1.05) 0.85 (0.80–0.91)
Abbreviations: DLBCL, diffuse large B-cell lymphoma; CLL, chronic lymphocytic leukemia; FL, follicular lymphoma; HL, Hodgkin lymphoma; MALT, mucosa-associated lymphoid tissue; MCL, Mantle cell lymphoma; MM, multiple myeloma; MZL, marginal zone lymphoma; NHL, non-Hodgkin lymphoma.aCells with <5 were suppressed for privacy. Lymphomas of unknown lineage that are not otherwise specified (NOS), and entities with fewer than 5 cases were notanalyzed.bAdjusted for age at enrollment (continuous) and sex (male/female). Age at death was used for nonliving participants.cOR and 95% CI were estimated by GEE logistic regression (clustered by family) with an autoregressive correlation structure. Birth order referent group: first born;Sibship size referent group: two siblings. Bold type, 95% CI does not include 1.00, denoting significant association with risk of lymphoma.dGroupings are based on the InterLymph hierarchical classification of lymphoid neoplasm for epidemiologic research (27). Subtype family numbers (n) sums togreater than 196 because some families contain heterogeneous types of lymphoid cancers (e.g., NHL and CLL cases).
Jones et al.
Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 CANCER EPIDEMIOLOGY, BIOMARKERS & PREVENTION1172
Table
4.ORsforrisk
oflympho
maan
dhistologic
subtypes
forchild
hoodlifestyle
variab
lesan
dim
mun
edisordersin
GEEregressionan
alysis.a
Alltypes
InterLym
phclass
Lympho
idne
oplasm
s(LN)
Categ
ory
1Non-Hodgkinlympho
ma(N
HL)
Categ
ory
2B-cellNHL
Categ
ory
3Categ
ory
4or6
Chroniclympho
cytic
leuk
emia
(CLL
)Variable
cncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
Childho
odfarm
res.
No
201/37
01.0
0(Referen
t)188/360
1.00(Referen
t)157/33
61.0
0(Referen
t)139/293
1.00(Referen
t)65/160
1.00(Referen
t)Yes
65/165
0.65(0.48–0
.88)
63/167
0.70(0.52–
0.94)
58/152
0.76(0.55–
1.03)
53/146
0.70(0.51–0.96)
18/42
1.02(0.68–1.54)
Paterna
leducation
Less
than
HS
104/221
1.00(Referen
t)102/22
31.0
0(Referen
t)93/212
1.00(Referen
t)88/202
1.00(Referen
t)39
/101
1.00(Referen
t)HSgrad.
88/171
1.13(0.84–1.51)
81/166
1.09(0.81–1.4
6)
70/150
1.14(0.83–
1.57)
60/125
1.19(0.85–
1.68)
23/55
1.13(0.72–
1.77)
Post-sec.g
rad.
64/122
1.20(0.86–1.67)
61/118
1.19(0.86–1.66)
46/110
1.06(0.73–
1.52)
41/99
1.07(0.74–1.53)
19/46
1.26(0.74–2
.13)
Materna
leducation
Less
than
HS
94/218
1.00(Referen
t)92/22
01.0
0(Referen
t)81/20
31.0
0(Referen
t)77
/195
1.00(Referen
t)33
/96
1.00(Referen
t)HSgrad.
122/20
81.35
(1.01–1.7
9)
116/204
1.38
(1.04–1.83)
102/194
1.42(1.04–1.96)
85/164
1.42(1.03–
1.95)
36/76
1.55
(1.07–
2.26
)Post-sec.g
rad.
45/72
1.50
(1.09–2
.06)
41/72
1.42(1.03–
1.96)
28/59
1.23(0.88–1.73)
24/51
1.23(0.88–1.73)
13/30
1.26(0.76–2
.07)
Childho
odfamily
inco
me
Below
averag
e64/147
1.00(Referen
t)60/147
1.00(Referen
t)53
/136
1.00(Referen
t)49/128
1.00(Referen
t)22
/60
1.00(Referen
t)Ave
rage
138/287
1.13(0.84–1.51)
131/28
41.13(0.85–
1.50)
114/263
1.16(0.82–
1.65)
100/228
1.20(0.84–1.72)
50/118
1.35(0.91–1.9
9)
Above
averag
e51/69
1.75
(1.22–
2.50
)48/70
1.65(1.12
–2.43)
36/63
1.45(0.94–2.28)
31/57
1.45(0.96–2.17
)7/16
1.17(0.71–1.9
2)Childho
odresiden
ceRural
106/201
1.00(Referen
t)101/198
1.00(Referen
t)85/181
1.00(Referen
t)78
/174
1.00(Referen
t)36
/76
1.00(Referen
t)Urban
145/29
90.97(0.75–
1.25)
136/300
0.92(0.71–1.19)
116/278
0.93(0.71–1.2
2)102/23
90.98(0.75–
1.30)
45/121
0.81(0.55–
1.19)
Education
Less
than
HS
50/28
1.00(Referen
t)49/27
1.00(Referen
t)45/25
1.00(Referen
t)43/23
1.00(Referen
t)13/16
1.00(Referen
t)HSgrad.
105/83
0.79(0.48–1.30)
103/81
0.73(0.45–
1.19)
92/76
0.78(0.46–1.32)
77/77
0.68(0.40–1.15
)31/34
1.18(0.46–3
.06)
Post-sec.g
rad.
118/110
0.62(0.38–0
.99)
110/107
0.56(0.34–0
.90)
92/102
0.53(0.21–0.88)
83/94
0.50(0.30–0
.85)
40/39
1.21(0.50–2.96)
Asthm
aNo
149/170
1.00(Referen
t)143/165
1.00(Referen
t)123/153
1.00(Referen
t)112/144
1.00(Referen
t)52
/67
1.00(Referen
t)Yes
31/28
1.21(0.73–
2.03)
29/30
1.07(0.63–
1.83)
26/30
1.01(0.56–1.81)
22/26
1.08(0.57–
2.04)
13/13
1.22(0.56–2.65)
Autoim
mun
eNo
152/155
1.00(Referen
t)146/154
1.00(Referen
t)123/147
1.00(Referen
t)111/134
1.00(Referen
t)56
/65
1.00(Referen
t)Yes
28/43
0.68(0.39–1.17
)26
/41
0.69(0.40–1.20)
26/36
0.87(0.49–1.57)
23/36
0.80(0.44–1.45)
9/15
0.82(0.33–
2.02)
Organ
-specific,No
169/174
1.00(Referen
t)161/172
1.00(Referen
t)138/162
1.00(Referen
t)125/149
1.00(Referen
t)59
/71
1.00(Referen
t)Yes
11/24
0.49(0.35–
1.06)
11/23
0.54(0.25–
1.18)
11/21
0.66(0.29–1.50)
9/21
0.55(0.23–
1.30)
6/9
0.98(0.27–
3.49)
Systemic,N
o167/186
1.00(Referen
t)159/184
1.00(Referen
t)136/173
1.00(Referen
t)122/160
1.00(Referen
t)61/77
1.00(Referen
t)Yes
13/12
1.11(0.48–2
.57)
13/11
1.30(0.55–
3.08)
13/10
1.38(0.59–3
.21)
12/10
1.38(0.58–3
.29)
4/3
1.44(0.29–7.17
)Noau
toAb,N
o173/189
1.00(Referen
t)167/186
1.00(Referen
t)144/176
1.00(Referen
t)129/163
1.00(Referen
t)65/76
—
Yes
7/9
0.92(0.35–
2.43)
5/9
0.66(0.23–
1.87)
5/7
0.96(0.29–3
.29)
5/7
0.99(0.29–3
.31)
0/4
—
Allergies
No
58/99
1.00(Referen
t)144/152
1.00(Referen
t)51/85
1.00(Referen
t)48/79
1.00(Referen
t)22
/40
1.00(Referen
t)Yes
108/89
2.25
(1.44–3
.51)
28/43
2.25
(1.42–
3.56
)86/87
2.23
(1.35–
3.66)
76/81
2.06(1.23–
3.46)
34/35
2.52
(1.05–
6.07)
Drug,N
o104/144
1.00(Referen
t)99/140
1.00(Referen
t)124/88
1.00(Referen
t)79
/114
1.00(Referen
t)37
/57
1.00(Referen
t)Yes
62/44
2.30
(1.41–3.73
)60/44
2.30
(1.40–3
.79)
48/49
1.85(1.15
–3.07)
45/46
1.82(1.08–3
.07)
19/18
2.36
(1.01–5.55
)
(Con
tinu
edon
thefollowingpag
e)
Early Life Factors and Risk of Lymphoma in Families
AACRJournals.org Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 1173
Table
4.ORsforrisk
oflympho
maan
dhistologic
subtypes
forchild
hoodlifestyle
variab
lesan
dim
mun
edisordersin
GEEregressionan
alysis.a
(Cont'd)
Alltypes
InterLym
phclass
Lympho
idne
oplasm
s(LN)
Categ
ory
1Non-Hodgkinlympho
ma(N
HL)
Categ
ory
2B-cellNHL
Categ
ory
3Categ
ory
4or6
Chroniclympho
cytic
leuk
emia
(CLL
)Variable
cncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
Env
ironm
ent,No
94/132
1.00(Referen
t)91/129
1.00(Referen
t)79
/119
1.00(Referen
t)75
/111
1.00(Referen
t)35
/56
1.00(Referen
t)Yes
72/56
1.90(1.21–2.98)
68/55
1.83(1.13
–2.95)
58/53
1.96(1.15
–3.35)
49/49
1.67(0.96–2
.93)
21/19
1.77(0.71–4.41)
Food,N
o130/160
1.00(Referen
t)124/156
1.00(Referen
t)107/145
1.00(Referen
t)98/134
1.00(Referen
t)42/63
1.00(Referen
t)Yes
36/28
1.69(0.92–
3.11)
35/28
1.66(0.89–3
.07)
30/27
1.83(0.97–
3.45)
26/24
1.65(0.84–3
.22)
14/12
2.39
(0.86–6
.65)
Appen
dectomy
No
130/160
1.00(Referen
t)127/156
1.00(Referen
t)107/144
1.00(Referen
t)98/134
1.00(Referen
t)46/64
1.00(Referen
t)Yes
36/27
1.53(0.80–2
.96)
32/27
1.41(0.73–
2.72
)30
/27
1.29(0.64–2
.60)
26/25
1.24(0.60–2
.58)
10/11
1.30(0.39–4
.37)
Tonsillectomy
No
81/117
1.00(Referen
t)77
/111
1.00(Referen
t)61/100
1.00(Referen
t)53
/92
1.00(Referen
t)22
/37
1.00(Referen
t)Yes
85/70
1.78
(1.14
–2.78)
82/72
1.68(1.08–2.63)
76/71
1.53(0.97–
2.41)
71/67
1.65(1.02–
2.69)
34/38
1.38(0.71–2.67)
InterLym
phclass
Categ
ory
1Ly
mpho
idne
oplasm
s(LN)
Categ
ory
2Non-Hodgkinlympho
ma(N
HL)
Hodgkinlympho
ma(H
L)Categ
ory
3B-cellNHL
Classic
HL
Categ
ory
4or6
Diffuse
largeB-cell
Follicu
larlympho
ma
Nodular
sclerosing
Multiple
mye
loma(M
M)
Variable
cncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
Childho
odfarm
res.
No
20/50
1.00(Referen
t)28
/62
1.00(Referen
t)31/70
1.00(Referen
t)22
/41
1.00(Referen
t)13/38
1.00(Referen
t)Yes
7/29
0.71(0.43–
1.17)
11/42
0.57(0.36–0
.91)
5/19
1.08(0.47–
2.38
)3/13
0.45(0.15
–1.31)
2/7
0.90(0.42–
1.91)
Paterna
leducation
Less
than
HS
11/45
1.00(Referen
t)17/50
1.00(Referen
t)9/31
1.00(Referen
t)5/15
1.00(Referen
t)2/7
1.00(Referen
t)HSgrad.
9/20
1.40(0.63–
2.09)
14/35
1.19(0.67–
2.12)
11/20
1.38(0.67–
2.86)
6/12
1.46(0.73–
2.92)
7/16
1.53(0.75–
3.11)
Post-sec.g
rad.
7/14
1.96(0.80–4
.81)
5/11
1.34(0.82–
2.20
)15/31
1.10(0.50–2.43)
12/22
1.66(0.64–4
.31)
3/14
0.41(0.14
–1.16
)Materna
leducation
Less
than
HS
10/42
1.00(Referen
t)18/48
1.00(Referen
t)11/33
1.00(Referen
t)8/20
1.00(Referen
t)2/7
1.00(Referen
t)HSgrad.
14/31
1.61(0.78–3
.32)
16/42
1.02(0.60–1.73)
14/35
1.03(0.56–1.91)
7/18
0.94(0.33–
2.67)
6/16
1.35(0.62–
2.95)
Post-sec.g
rad.
3/5
2.09(0.44–9
.96)
2/5
1.03(0.61–1.7
4)
13/26
1.04(0.45–
2.42)
10/16
1.25(0.32–
4.91)
4/9
1.58(0.53–
4.70)
Childho
odfamily
inco
me
Below
averag
e7/37
1.00(Referen
t)8/32
1.00(Referen
t)7/23
1.00(Referen
t)4/8
1.00(Referen
t)4/7
1.00(Referen
t)Ave
rage
9/19
2.46(1.20–5
.05)
25/60
1.67(1.04–2
.69)
17/43
1.45(0.84–2.50)
11/30
0.84(0.39–1.81)
7/23
0.67(0.30–1.52)
Above
averag
e11/23
2.20
(1.05–
4.89)
2/5
1.75
(1.02–
3.01)
12/23
1.74(0.86–3
.51)
10/16
1.24(0.51–3.03)
3/9
0.65(0.32–
1.31)
Childho
odresiden
ceRural
12/37
1.00(Referen
t)12/46
1.00(Referen
t)16/36
1.00(Referen
t)10/21
1.00(Referen
t)5/9
1.00(Referen
t)Urban
13/38
0.76(0.46–1.26)
22/52
1.53(0.92–
2.55
)20
/53
0.82(0.47–
1.45)
15/33
1.07(0.50–2
.30)
9/30
0.66(0.30–1.45)
Education
Less
than
HS
4/11
1.00(Referen
t)13/5
1.00(Referen
t)4/4
1.00(Referen
t)4/1
1.00(Referen
t)1/7
1.00(Referen
t)HSgrad.
10/11
2.44(1.01–5.92)
19/17
0.47(0.15
–1.44)
11/18
0.57(0.10
–3.32)
6/10
0.26(0.04–1.62)
2/16
0.16
(0.01–6.23)
Post-sec.g
rad.
13/22
1.24(0.42–3.61)
12/19
0.30(0.09–0
.92)
18/28
0.76(0.17
–3.36)
13/17
0.39(0.10
–1.46)
8/9
0.80(0.01–46.1)
(Con
tinu
edon
thefollowingpag
e)
Jones et al.
Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 CANCER EPIDEMIOLOGY, BIOMARKERS & PREVENTION1174
Table
4.ORsforrisk
oflympho
maan
dhistologic
subtypes
forchild
hoodlifestyle
variab
lesan
dim
mun
edisordersin
GEEregressionan
alysis.a
(Cont'd)
InterLym
phclass
Categ
ory
1Ly
mpho
idne
oplasm
s(LN)
Categ
ory
2Non-Hodgkinlympho
ma(N
HL)
Hodgkinlympho
ma(H
L)Categ
ory
3B-cellNHL
Classic
HL
Categ
ory
4or6
Diffuse
largeB-cell
Follicu
larlympho
ma
Nodular
sclerosing
Multiple
mye
loma(M
M)
Variable
cncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
ncase/sib
OR(95%
CI)a,b
Asthm
aNo
17/33
1.00(Referen
t)26
/30
1.00(Referen
t)20
/39
1.00(Referen
t)4/14
1.00(Referen
t)6/13
1.00(Referen
t)Yes
3/5
1.62(0.37–
7.03)
4/5
1.01(0.39–2.58)
3/3
1.58(0.19
–13.2)
13/10
1.61(0.12
–20.8)
2/1
20.8
(1.04–4
17)
Autoim
mun
eNo
14/29
1.00(Referen
t)26
/26
1.00(Referen
t)23
/33
—18/18
—6/10
1.00(Referen
t)Yes
6/9
2.00(0.62–
6.43)
4/9
0.44(0.11–1.75)
0/9
—0/6
—2/4
1.33(0.13
–13.3)
Organ-specific,No
20/34
—30
/39
—23
/39
—18/22
—8/12
—Yes
0/4
—0/6
—0/3
—0/2
—0/2
—
Systemic,N
o17/34
1.00(Referen
t)27
/32
1.00(Referen
t)23
/40
—18/23
—8/12
—
Yes
3/4
1.89(0.55–
6.52)
3/3
1.18(0.26–5
.28)
0/2
—0/1
—0/2
—
Noau
toAb,N
o17/37
1.00(Referen
t)29
/34
1.00(Referen
t)23
/39
—18/21
—6/13
1.00(Referen
t)Yes
3/1
9.23(0.72–
118)
1/1
0.91(0.05–
17.8)
0/3
—0/3
—2/1
26.0
(2.67–
253)
Allergies
No
11/19
1.00(Referen
t)8/14
1.00(Referen
t)6/29
1.00(Referen
t)4/14
1.00(Referen
t)1/5
1.00(Referen
t)Yes
9/18
1.18(0.42–
3.29
)21/20
2.35
(0.74–7.41)
16/13
4.93(1.77–
13.7)
13/10
6.66(1.36–3
2.5)
6/9
1.90(0.05–
68.2)
Drug,N
o14/30
1.00(Referen
t)18/22
1.00(Referen
t)11/32
1.00(Referen
t)8/20
1.00(Referen
t)5/13
1.00(Referen
t)Yes
6/7
2.26
(0.62–
8.16
)11/12
1.35(0.52–
3.46)
11/10
5.93(1.37–
25.6)
9/4
4.68(0.80–2
7.4)
2/1
50.7
(5.72–
449)
Env
ironm
ent,No
13/27
1.00(Referen
t)17/20
1.00(Referen
t)12/32
1.00(Referen
t)9/15
1.00(Referen
t)3/8
1.00(Referen
t)Yes
7/10
2.78
(0.78–9
.96)
12/14
1.14(0.46–2
.78)
10/10
2.19
(0.69–6
.95)
8/9
2.37
(0.53–
10.5)
4/6
0.73(0.05–
10.9)
Food,N
o18/30
1.00(Referen
t)23
/29
1.00(Referen
t)17/38
1.00(Referen
t)12/22
1.00(Referen
t)6/10
1.00(Referen
t)Yes
2/7
0.57(0.07–
4.39)
6/5
2.03(0.56–7
.41)
5/4
2.15
(0.53–
8.67)
5/2
5.18
(1.25–
21.4)
1/4
0.73(0.03–
18.9)
Appen
dectomy
No
12/33
1.00(Referen
t)23
/24
1.00(Referen
t)20
/40
1.00(Referen
t)16/23
1.00(Referen
t)3/11
—
Yes
8/3
9.72(3.34–2
8.3)
6/9
0.80(0.27–
2.40)
2/2
8.38(0.59–118)
1/1
13.4
(0.27–
673
)4/3
—
Tonsillectomy
No
8/25
1.00(Referen
t)14/12
1.00(Referen
t)16/32
1.00(Referen
t)13/18
1.00(Referen
t)4/12
1.00(Referen
t)Yes
12/11
5.17
(1.74–15.3)
15/21
0.78(0.28–2
.21)
6/10
4.51(1.08–18.9)
4/6
4.61(1.07–
19.9)
3/2
7.78
(0.47–
129)
Note:R
esultswithfewer
than
5casesshouldbeview
edwithcaution.Ly
mpho
mas
ofu
nkno
wnlinea
gethat
areno
totherwisespecified
(NOS),an
den
tities
withfewer
than
5caseswereno
tana
lyzed.G
roup
ingsarebased
on
theInterLym
phhierarchical
classificationoflympho
idne
oplasm
forep
idem
iologic
research
(27).
Abbreviations:C
LL,chroniclympho
cyticleuk
emia;F
arm
res,farm
residen
ce;H
L,Hodgkinlympho
ma;HS,highscho
ol;Grad,g
radua
te;LN,lym
pho
idne
oplasm
s;MM,m
ultiplemye
loma;NHL,no
n-Hodgkinlympho
ma;No
autoAb,n
odetectable
autoan
tibodies;Post-sec,p
ost-second
ary.
aAdjusted
forag
eat
enrollm
ent(continuo
us)an
dsex(m
ale/female).Ageat
dea
thwas
used
forno
nlivingparticipan
ts.
bORan
d95%
CIw
ereestimated
byGEElogisticregression(clustered
byfamily)withan
autoregressiveco
rrelationstructure.Birth
order
referent
group
:firstb
orn;Sibshipsize
referent
group
:twosiblings.Boldtype,95%
CI
does
notinclud
e1.0
0,d
enoting
asignificant
association.
c Variablesareordered
bysample
size.
Early Life Factors and Risk of Lymphoma in Families
AACRJournals.org Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 1175
B-cell subtypes (18, 50, 53), HL (14, 52), MM (7, 50), and familial WM(103 cases; ref. 25); however, some case–control (but no cohort) studiesobserved the opposite effect (7, 18, 21, 22, 47, 54, 55). To date, mostepidemiologic studies examining immunologic factors and lymphomahave been case–control studies that may be biased from immune-altering effects of preclinical lymphoid cancer (52), which would bemitigated in cohort studies. Explicit correlations between lymphoidcancer subtypes and high molecular weight allergens (54), serum IgElevels (55), and type of allergy (e.g., food, environment;refs. 7, 18, 21, 46, 47, 50) complicate the elucidation of these relation-ships. Both disorders are polygenic multifactorial diseases with aheritable component and environmental modifiers of risk with var-iable disease outcomes among populations, exposed groups, andfamilies (56, 57). Various explanations have been proposed to accountfor the observed increase in incidence of lymphoma among individualswith immune dysregulation (50, 52). Families with heterogeneouslymphoid cancers may have genetic susceptibility factors that perturboptimal immune development in a way that increases risk of bothallergies and lymphoid cancers. These predisposition factors may bedifferent than those that contribute to sporadic lymphoid cancers.Nonfamilial lymphoid cancers may be more environmentally trig-gered, and less genetic, and therefore result from a different combi-nation of etiologic factors than multiple-case families. In our study,affected sibships weremultigenerational and had a higher frequency ofallergies (55.6%) relative to the expected population frequency of 30%–40% (58), suggesting that susceptibility factors may play a bigger rolethan environmental influences. In addition to genetic and environ-mental risk factors, chronic stimulation and proliferation of lympho-cytes may increase the chance of oncogenic mutations and subsequentcancer development as described by the antigenic stimulationhypothesis (50, 59).
We observed no association among collective autoimmune disor-ders and familial lymphoid cancer occurrence. Personal and familyhistory of autoimmune conditions are strong established risk factors
for lymphoid cancers (5, 25, 26, 60, 61), so this finding is unexpected.However, we were unable to examine subtype-specific associationsamong the biologically diverse autoimmune diseases, and a personalhistory of organ-specific autoimmune disease was associated withlower risk of lymphoma. Among individuals with an organ-specificautoimmune disease, unaffected siblings were on average 11 yearsyounger than lymphoid cases, suggesting they may not be trulylymphoid unaffected due to shorter duration of follow-up, and/orascertainment bias. In our study, an autoimmune condition wasobserved in 18.8% of individuals in a lymphoid-affected sibship, whichwas higher than the expected population frequency of 3%–6% (62),suggesting that these families may be enriched for genetic factors thatpredispose to immune dysregulation and associated conditions.
With the possible exception of MM, we observed no associationbetween familial lymphoid cancer and asthma, consistent with theliterature (9, 11, 13, 18, 21, 46). Some studies, including ours, wereunable to differentiate between allergic and non-allergic asthma whichmay explain inconsistency among associations/studies (5). In thisstudy, a higher risk of DLBCL (but no other subtype) was associatedwith an appendectomy, which is consistent with some (30, 63), but notall epidemiological population-based studies (9, 64). An appendecto-my/appendicitis may reflect susceptibility to infection/inflammation;however, this information was unavailable in our study cohort. Theremoval of the appendix and surrounding lymphoid tissue may alterthe natural immune response to pathogenic microorganisms (63).
Our observations support the antigen stimulation hypothe-sis (46, 59), wherein chronic immune stimulation progressively leadsto random oncogenic mutations and subsequent cancerdevelopment (7, 14, 18, 25, 46, 50–52). In contrast, the immunesurveillance hypothesis proposes that allergic conditions enhance theability of the immune system to detect and eliminate malignantcells (46, 59), and is also well supported (7, 18, 21, 22, 47, 54, 55).Inconsistencies among studies may be partially attributable to differ-ences in study designs, reverse causality, gender differences, selection
Table 5. ORs for risk of lymphoid cancer from stepwise GEE logistic regression models.
Adjusted for birth order Adjusted for sibship sizeModel OR (95% CI)a,b OR (95% CI)a,b
1. Base model, n ¼ 1,468Family structure 0.83 (0.78–0.89) 0.82 (0.79–0.85)
2. Middle model, n ¼ 682Family structure 0.83 (0.75–0.92) 0.82 (0.78–0.85)Childhood family income
Below average 1.00 (Referent) 1.00 (Referent)Average 1.00 (0.75–1.32) 0.78 (0.62–0.97)Above average 1.40 (0.97–1.97) 0.99 (0.73–1.34)
3. Full model, n ¼ 321Family structure 0.85 (0.75–0.98) 0.82 (0.74–0.89)Allergies 2.58 (1.59–4.20) 2.46 (1.52–3.98)Autoimmune 0.65 (0.35–1.22) 0.58 (0.31–1.09)Tonsillectomy 1.72 (1.06–2.81) 1.51 (0.91–2.56)Maternal education
Less than HS (not selected) 1.00 (Referent)HS graduate — 0.53 (0.31–0.93)Post-sec. graduate — 0.47 (0.23–0.96)
Abbreviations: HS, high school; Post-sec, post-secondary.aAdjusted for age at enrollment (continuous) and sex (male/female). Age at death was used for nonliving participants.bOR and 95% CI were estimated by GEE logistic regression (clustered by family) with an autoregressive correlation structure. Birth order referent group: first born;Sibship size referent group: two siblings. Bold type, 95% CI does not include 1.00, denoting a significant association.
Jones et al.
Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 CANCER EPIDEMIOLOGY, BIOMARKERS & PREVENTION1176
bias, diverse definition and measurement of allergy, hematologicsubtypes assessed, reliance on self-reported data/recall bias, andparticipant characteristics (e.g., families with a genetic etiology, spo-radic cases; refs. 7, 23, 31, 50).
Our study has several strengths, including extensive demographic,family structure, and exposure data, and inclusion of unaffected familymembers. Participation rates of cases and unaffected siblings were notdifferentiated by SES or education. Furthermore, case–control studiescan suffer from response bias due to education and SES disparitiesamong participants; our study design is able to differentiate risk amongsuch categories. Despite the rarity of familial hematologic malignan-cies, this study included a relatively large number of families (196).Wewere able to detect effects of birth order, sibship size, and childhoodenvironment among familial lymphoid cancer cases while controllingfor known lymphoma risk factors (age, sex, ethnicity). Limitationsinclude use of self-reported data which may be subject to recall/response biases. We did not have complete atopic disease data ordirect markers of infectious exposure, such as number and type ofinfections, age at infection, or serologic data. Shorter duration offollow-up may have biased some associations because insufficienttime elapsed for disease development among siblings and children.Families were not ascertained by means of a systematic population-based study, which may limit the generalizability of the findings tononfamilial lymphoma.However, this study represents the largest, andin terms of demographic and lifestyle information, the most exten-sively characterized cohort of lymphoid cancer families reported todate.
This investigation represents the first multiple-case family study toquantify the effects of family structure according to lymphoid cancertype. This is the first study to establish an inverse relationship betweenfamily structure (birth order and sibship size) and risk of CLL andMMin the context of families with heterogeneous lymphoid cancers. Theobserved inverse relationship between family structure and risk oflymphoma is supportive of the hygiene hypothesis, and that childhoodexposure to infectious agents may play a role in the risk of multiple
types of lymphoid cancers. Our observations indicate that lifestylefactors such as SES and education also correlatewith risk of lymphoma.The familial nature of these cancers implies a role of shared geneticand/or environmental factors. Such effectsmay bemodified by lifestylefactors that correlate with birth order and family structure, and couldlead to the identification of modifiable factors that protect againstlymphoid cancers, even in the context of multiple-case families.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors’ ContributionsConception and design: S.J. Jones, J.M. Connors, A.R. Brooks-WilsonDevelopment of methodology: S.J. Jones, J.J. SpinelliAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): J.M. Connors, A.R. Brooks-WilsonAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S.J. Jones, S. Stroshein, A.M. Williams, D. Liu,J.J. Spinelli, J.M. ConnorsWriting, review, and/or revision of the manuscript: S.J. Jones, S. Stroshein,A.M. Williams, J.J. Spinelli, J.M. Connors, A.R. Brooks-WilsonAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): J.M. ConnorsStudy supervision: A.R. Brooks-Wilson
AcknowledgmentsWe thank the families for their participation.We thank Susan Slager for her advice
on statistical analyses. S. Stroshein was supported by a Vice-President ResearchUndergraduate Student Research Award from Simon Fraser University. This study issupported by the research grant MOP-130311 (awarded to A.R. Brooks-Wilson) bythe Canadian Institutes of Health Research.
The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.
Received September 28, 2019; revised December 4, 2019; accepted March 9, 2020;published first March 13, 2020.
References1. National Cancer Institute, DCCPS, Surveillance Research Program. Surveillance,
Epidemiology, and EndResults (SEER) Program. SEER�StatDatabase: Incidence- SEER 18 Regs Research Data þ Hurricane Katrina Impacted Louisiana Cases,Nov 2018 Sub (1975–2016 varying); 2019. Available from: www.seer.cancer.gov.
2. American Cancer Society. Cancer facts & figures 2018. Atlanta (GA): AmericanCancer Society; 2018. Available from: https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/cancer-facts-figures-2018.html.
3. Skibola CF, Curry JD, Nieters A. Genetic susceptibility to lymphoma. Haema-tologica 2007;92:960–9.
4. Rothman N, Skibola CF, Wang SS, Morgan G, Lan Q, Smith MT, et al. Geneticvariation in TNF and IL10 and risk of non-Hodgkin lymphoma: a report fromthe InterLymph Consortium. Lancet Oncol 2005;7:27–38.
5. Alexander DD, Mink PJ, Adami H-O, Chang ET, Cole P, Mandel JS, et al. Thenon-Hodgkin lymphomas: a review of the epidemiologic literature. Int J Cancer2007;120:1–39.
6. Strachan DP. Hay fever, hygiene, and household size. BMJ 1989;299:1259–60.
7. Becker N, de Sanjose S, Nieters A,Maynadi�eM, Foretova L, Cocco PL, et al. Birthorder, allergies and lymphoma risk: results of the European collaborativeresearch project Epilymph. Leuk Res 2007;31:1365–72.
8. CrumpC, Sundquist K, SiehW,WinklebyMA, Sundquist J. Perinatal and familyrisk factors for non-Hodgkin lymphoma in early life: a Swedish national cohortstudy. J Natl Cancer Inst 2012;104:923–30.
9. Cartwright RA, McKinney PA, O'Brien C, Richards IDG, Roberts B, Lauder I,et al. Non-Hodgkin's lymphoma: case-control epidemiological study in York-shire. Leuk Res 1988;12:81–8.
10. Von Behren J, Spector LG,Mueller BA, Carozza SE, Chow EJ, Fox EE, et al. Birthorder and risk of childhood cancer: a pooled analysis from five U.S. states. Int JCancer 2012;128:2709–16.
11. Rudant J, Orsi L,MonnereauA, Patte C, Pacquement H, Landman-Parker J, et al.Childhood Hodgkin's lymphoma, non-Hodgkin's lymphoma and factors relatedto the immune system: The Escale Study (SFCE). Int J Cancer 2011;129:2236–47.
12. Chang ET,Montgomery SM, Richiardi L, Ehlin A, EkbomA, LambeM. Numberof siblings and risk of Hodgkin's lymphoma. Cancer Epidemiol Biomarkers Prev2004;13:1236–43.
13. Bernard SM, Cartwright RA, Darwin CM, Richards IDG, Roberts B, O'Brien C,et al. Hodgkin's disease: case-control epidemiological study in Yorkshire.Br J Cancer 1987;55:85–90.
14. GutensohnN, Cole P. Childhood social and environment andHodgkin's disease.N Engl J Med 1981;304:135–40.
15. Westergaard T, MelbyeM, Pedersen JB, FrischM, Olsen JH, Andersen PK. Birthorder, sibship size and risk of Hodgkin's disease in children and young adults: apopulation-based study of 31million person-years. Int J Cancer 1997;72:977–81.
16. Altieri A, Castro F, Bermejo JL, Hemminki K. Number of siblings and the risk oflymphoma, leukemia, and myeloma by histopathology. Cancer Epidemiol Bio-markers Prev 2006;15:1281–6.
17. Smedby KE,HjalgrimH, Chang ET, Rostgaard K, Glimelius B, AdamiH-O, et al.Childhood social environment and risk of non-Hodgkin lymphoma in adults.Cancer Res 2007;67:11074–82.
18. Cozen W, Cerhan JR, Martinez-Maza O, Ward MH, Linet M, Colt JS, et al. Theeffect of atopy, childhood crowding, and other immune-related factors on non-Hodgkin lymphoma risk. Cancer Causes Control 2007;18:821–31.
Early Life Factors and Risk of Lymphoma in Families
AACRJournals.org Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 1177
19. Bevier M,Weires M, Thomsen H, Sundquist J, Hemminki K. Influence of familysize and birth order on risk of cancer: a population-based study. BMC Cancer2011;11:163–73.
20. CrumpC, Sundquist K, SiehW,WinklebyMA, Sundquist J. Perinatal and familyrisk factors for Hodgkin lymphoma in childhood through young adulthood.Am J Epidemiol 2012;176:1147–58.
21. GrulichAE, Vajdic CM,Kaldor JM,Hughes AM,Kricker A, Fritschi L, et al. Birthorder, atopy, and risk of non-Hodgkin lymphoma. J Natl Cancer Inst 2005;97:587–94.
22. Bracci PM, Dalvi TB, Holly EA. Residential history, family characteristics andnon-Hodgkin lymphoma, a population-based case-control study in the SanFrancisco Bay Area. Cancer Epidemiol Biomarkers Prev 2006;15:1287–94.
23. Grulich AE, Vajdic CM, Falster MO, Kane E, Ekstrom Smedby K, Bracci PM,et al. Birth order and risk of non-Hodgkin lymphoma - true association or bias?Am J Epidemiol 2010;172:621–30.
24. Jønsson V, Tjønnfjord G, Samuelsen SO, Johannesen T, Olsen J, Sellick G, et al.Birth order pattern in the inheritance of chronic lymphocytic leukaemia andrelated lymphoproliferative disease. Leuk Lymphoma 2007;48:2387–96.
25. Royer RH, Koshiol J, Giambarresi TR, Vasquez LG, Pfeiffer RM, McMaster ML.Differential characteristics of Waldenstr€om macroglobulinemia according topatterns of familial aggregation. Blood 2010;115:4464–71.
26. Mellemkjaer L, Pfeiffer RM, Engels EA, Gridley G, Wheeler W, Hemminki K,et al. Autoimmune disease in individuals and close family members andsusceptibility to non-Hodgkin's lymphoma. Arthritis Rheum 2008;58:657–66.
27. Turner JJ, Morton LM, Linet MS, Clarke CA, Kadin ME, Vajdic CM, et al.InterLymph hierarchical classification of lymphoid neoplasms for epidemiologicresearch based on the WHO classification (2008): update and future directions.Blood 2010;116:90–9.
28. Canadian Cancer Society's Advisory Committee on Cancer Statistics.Canadian cancer statistics 2017 special topic: pancreatic cancer. CanadianCancer Society; 2017. Available from: https://www.cancer.ca/~/media/cancer.ca/CW/cancer%20information/cancer%20101/Canadian%20cancer%20statistics/Canadian-Cancer-Statistics-2017-EN.pdf?la=en.
29. Howlader N, Noone AM, Krapcho M, Garshell J, Miller D, Altekruse SF, et al.National Cancer Institute SEER Cancer Statistics Review 1975–2012. J NatlCancer Inst 2015;103:1975–2012.
30. CozenW, Hamilton AS, Zhao P, SalamMT, Deapen DM, Nathwani BN, et al. Aprotective role for early oral exposures in the etiology of young adult Hodgkinlymphoma. Blood 2009;114:4014–20.
31. Jones SJ, Voong J, Thomas R, English A, Schuetz J, Slack GW, et al. Nonrandomoccurrenceof lymphoid cancer types in 140 families. LeukLymphoma2017;58:1–10.
32. Altieri A, Bermejo JL, Hemminki K. Familial risk for non-Hodgkin lymphomaand other lymphoproliferative malignancies by histopathologic subtype: theSwedish Family-Cancer Database. Blood 2005;106:668–72.
33. Mauro FR,Giammartini E, GentileM, Sperduti I, Valle V, Pizzuti A, et al. Clinicalfeatures and outcome of familial chronic lymphocytic leukemia. Haematologica2006;91:1117–20.
34. Crowther-Swanepoel D, Wild R, Sellick G, Dyer MJS, Mauro FR, Cuthbert RJ,et al. Insight into the pathogenesis of chronic lymphocytic leukemia (CLL)through analysis of IgVH gene usage and mutation status in familial CLL.Neoplasia 2008;111:5691–3.
35. Li Q, Chang ET, Bassig BA, Dai M, Qin Q, Gao Y, et al. Body size and risk ofHodgkin's lymphoma by age and gender: a population-based case-control studyin Connecticut and Massachusetts. Cancer Causes Control 2013;24:287–95.
36. Horton S. Birth order and child nutritional status: evidence from the Philippines.Econ Dev Cult Change 1988;36:341–54.
37. Vineis P, Miligi L, Crosignani P, Fontana A, Masala G, Nanni O, et al. Delayedinfection, family size and malignant lymphomas. J Epidemiol CommunityHealth 2000;54:907–11.
38. Serraino D, Franceschi S, Talamini R, Barra S, Negri E, Carbone A, et al. Socio-economic indicators, infectious diseases and Hodgkin's disease. Int J Cancer1991;47:352–7.
39. Carozza SE, Puumala SE, Chow EJ, Fox EE, Horel S, Johnson KJ, et al. Parentaleducational attainment as an indicator of socioeconomic status and risk ofchildhood cancers. Br J Cancer 2010;103:136–42.
40. Hjalgrim H, Ekstrom Smedby K, Rostgaard K, Molin D, Hamilton-Dutoit S,Chang ET, et al. Infectious mononucleosis, childhood social environment, andrisk of Hodgkin lymphoma. Cancer Res 2007;67:2382–8.
41. Becker N, Deeg E, Nieters A. Population-based study of lymphoma in Germany:rationale, study design and first results. Leuk Res 2004;28:713–24.
42. Pearce N, Bethwaite P. Increasing incidence of non-Hodgkin's lymphoma:occupational and environmental factors. Cancer Res 1992;52:5496–501.
43. Hermann S, Rohrmann S, Linseisen J, Nieters A, Khan A, Gallo V, et al. Level ofeducation and the risk of lymphoma in the European prospective investigationinto cancer and nutrition. J Cancer Res Clin Oncol 2010;136:71–7.
44. Baris D, Brown LM, Silverman DT, Hayes R, Hoover RN, Swanson GM, et al.Socioeconomic status andmultiplemyeloma amongUSBlacks andWhites. Am JPublic Health 2000;90:1277–81.
45. Krull KR, Sabin ND, Reddick WE, Zhu L, Armstrong GT, Green DM, et al.Neurocognitive function and CNS integrity in adult survivors of childhoodHodgkin lymphoma. J Clin Oncol 2012;30:3618–24.
46. S€oderberg KC,Hagmar L, Schwartzbaum J, FeychtingM.Allergic conditions andrisk of hematological malignancies in adults: a cohort study. BMC Public Health2004;4:1–6.
47. Becker N, Deeg E, R€udiger T, Nieters A. Medical history and risk for lymphoma:results of a population-based case-control study in Germany. Eur J Cancer 2005;41:133–42.
48. Liaw K-L, Adami J, Gridley G, Nyren O, Linet MS. Risk of Hodgkin's diseasesubsequent to tonsillectomy: a population-based cohort study in Sweden. Int JCancer 1997;72:711–3.
49. Vestergaard H, Westergaard T, Wohlfahrt J, Hjalgrim H, Melbye M. Tonsillitis,tonsillectomy and Hodgkin's lymphoma. Int J Cancer 2010;127:633–7.
50. ShadmanM,White E,DeRoosAJ,Walter RB.Associations between allergies andrisk of hematologic malignancies: results from the VITamins and lifestyle cohortstudy. Am J Hematol 2013;88:1050–4.
51. Koshiol J, Lam TK, Gridley G, Check D, Brown LM, Landgren O. Racialdifferences in chronic immune stimulatory conditions and risk of non-Hodg-kin's lymphoma in veterans from theUnited States. J ClinOncol 2011;29:378–85.
52. Erber E, Lim U, Maskarinec G, Kolonel LN. Common immune-related riskfactors and incident non-Hodgkin lymphoma: the multiethnic cohort. Int JCancer 2009;125:1440–5.
53. Melbye M, Ekstrom Smedby K, Lehtinen T, Rostgaard K, Glimelius B, Munks-gaard L, et al. Atopy and risk of non-Hodgkin lymphoma. J Natl Cancer Inst2007;99:158–66.
54. Mirabelli MC, Zock J-P, D'Errico A, Kogevinas M, de Sanjos�e S, Miligi L, et al.Occupational exposure to high molecular weight allergens and lymphomarisk among Italian adults. Cancer Epidemiol Biomarkers Prev 2009;18:2650–4.
55. Ellison-Loschmann L, Benavente Y, Douwes J, Buendia E, Font R, Alvaro T, et al.Immunoglobulin E levels and risk of lymphoma in a case-control study in Spain.Cancer Epidemiol Biomarkers Prev 2007;16:1492–8.
56. Ober C, Yao T-C. The genetics of asthma and allergic disease: a 21st centuryperspective. Immunol Rev 2011;242:10–30.
57. Institute of Medicine (US) Roundtable on Environmental Health Sciences,Research, and Medicine. Cancer and the environment: gene-environmentinteractions. In: Wilson S, Jones L, Couseens C, editors. Gene–environmentinteraction. Washington (DC): National Academies Press; 2002.
58. Ryan D, Yusuf O, Ostergaard MS, Roman-Rodriguez M. World Allergy Orga-nization, white book on allergy: update 2013; 2013. Available from: http://www.worldallergy.org/UserFiles/file/WhiteBook2-2013-v8.pdf.
59. MusolinoC,AllegraA,Minciullo PL,Gangemi S. Allergy and risk of hematologicmalignancies: associations and mechanisms. Leuk Res 2014;38:1137–44.
60. Kleinstern G, Maurer MJ, Liebow M, Habermann TM, Koff JL, Allmer C, et al.History of autoimmune conditions and lymphoma prognosis. Blood Cancer J2018;8:1–10.
61. Hemminki K, F€orsti A, Sundquist K, Sundquist J, Li X. Familial associationsof lymphoma and myeloma with autoimmune diseases. Blood Cancer J 2017;7:1–5.
62. Hayter SM, CookMC. Updated assessment of the prevalence, spectrum and casedefinition of autoimmune disease. Autoimmun Rev 2012;11:754–65.
63. Mohammadi M, Song H, Cao Y, Glimelius I, Ekbom A, Ye W, et al. Risk oflymphoid neoplasms in a Swedish population-based cohort of 337,437 patientsundergoing appendectomy. Scand J Gastroenterol 2016;51:583–9.
64. Mellemkjær L, Johansen C, LinetMS, Gridley G, Olsen JH. Cancer risk followingappendectomy for acute appendicitis (Denmark). Cancer Causes Control 1998;9:183–7.
Cancer Epidemiol Biomarkers Prev; 29(6) June 2020 CANCER EPIDEMIOLOGY, BIOMARKERS & PREVENTION1178
Jones et al.
Genomic Characterization of HIV-Associated Plasmablastic Lymphoma Identifies Pervasive Mutations in the JAK–STAT PathwayZhaoqi Liu1,2, Ioan Filip1,2, Karen Gomez1,2, Dewaldt Engelbrecht3, Shabnum Meer4, Pooja N. Lalloo3, Pareen Patel3, Yvonne Perner5, Junfei Zhao1,2, Jiguang Wang6, Laura Pasqualucci7–9, Raul Rabadan1,2, and Pascale Willem3
RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 113
1Program for Mathematical Genomics, Columbia University, New York, New York. 2Departments of Systems Biology and Biomedical Informatics, Columbia University, New York, New York. 3Department of Haematology and Molecular Medicine, National Health Laboratory Service, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa. 4Department of Oral Pathology, Faculty of Health Sciences, Uni-versity of the Witwatersrand, Johannesburg, South Africa. 5Department of Anatomical Pathology, National Health Laboratory Service, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa. 6Division of Life Science, Department of Chemical and Biological Engineering, Center for Systems Biology and Human Health and State Key Laboratory of Molecular Neuroscience, The Hong Kong University of Sci-ence and Technology, Hong Kong SAR, China. 7Institute for Cancer Genetics. 8Herbert Irving Comprehensive Cancer Center, Columbia University, New York, New York. 9Department of Pathology and Cell Biology, Herbert Irving Comprehensive Cancer Center, Columbia University, New York, New York.Note: Supplementary data for this article are available at Blood Cancer Discovery Online (http://bloodcancerdiscov.aacrjournals.org/).L. Pasqualucci, R. Rabadan, and P. Willem contributed equally to this article.Corresponding Authors: Laura Pasqualucci, Institute for Cancer Genetics, Columbia University, 1130 St Nicholas Ave, Room 507B, New York, NY 10032. Phone: 212-851-5248; Fax: 212-851-5256; E-mail: [email protected]; Raul Rabadan, Columbia University Irving Medical Center, 622 West 168th Street, PH18-200, New York, NY 10032. Phone: 212-305-3896; Fax: 212-851-5149; Email: [email protected]; and Pascale Willem, University of the Witwatersrand and the National Health Laboratory Service, Wits Medi-cal School, 7 York Road Parktown, Johannesburg 2193, South Africa. Phone: 271-1489-8406; Fax: 271-1489-8480; E-mail: [email protected] Cancer Discov 2020;1:112–25doi: 10.1158/2643-3230.BCD-20-0051©2020 American Association for Cancer Research.
ABSTRACT Plasmablastic lymphoma (PBL) is an aggressive B-cell non-Hodgkin lymphoma asso-ciated with immunodeficiency in the context of human immunodeficiency virus (HIV)
infection or iatrogenic immunosuppression. While a rare disease in general, the incidence is dramatically increased in regions of the world with high HIV prevalence. The molecular pathogenesis of this disease is poorly characterized. Here, we defined the genomic features of PBL in a cohort of 110 patients from South Africa (15 by whole-exome sequencing and 95 by deep targeted sequencing). We identified recur-rent mutations in genes of the JAK–STAT signaling pathway, including STAT3 (42%), JAK1 (14%), and SOCS1 (10%), leading to its constitutive activation. Moreover, 24% of cases harbored gain-of-function mutations in RAS family members (NRAS and KRAS). Comparative analysis with other B-cell malignan-cies uncovered PBL-specific somatic mutations and transcriptional programs. We also found recurrent copy number gains encompassing the CD44 gene (37%), which encodes for a cell surface receptor involved in lymphocyte activation and homing, and was found expressed at high levels in all tested cases, independent of genetic alterations. These findings have implications for the understanding of the patho-genesis of this disease and the development of personalized medicine approaches.
SIGNIFICANCE: Plasmablastic lymphoma is a poorly studied and extremely aggressive tumor. Here we define the genomic landscape of this lymphoma in HIV-positive individuals from South Africa and iden-tify pervasive mutations in JAK–STAT3 and RAS–MAPK signaling pathways. These data offer a genomic framework for the design of improved treatment strategies targeting these circuits.
See related commentary by Küppers, p. 23.
INTRODUCTIONPlasmablastic lymphoma (PBL) is a highly aggressive lym-
phoma of preterminally differentiated B cells, which predomi-nantly occurs in patients with human immunodeficiency virus
(HIV)-related or iatrogenic immunodeficiency (1). As an AIDS-defining illness with a dismal prognosis, PBL is a particularly compelling problem in the Sub-Saharan African region, which accounts for approximately 54% of the estimated 37.9 million people living with HIV.
Despite the national rollout offering combination anti- retroviral therapy (cART) in South Africa since 2004, the prev-alence of HIV-associated mature B-cell lymphomas increases yearly (2). The burden of HIV-associated lymphomas on the country’s encumbered health are further compounded by the younger presentation age of HIV-infected patients, the challenging classification of these lymphomas, and an exceptionally aggressive clinical disease course that, although curable in a significant fraction of patients (3), often results in fatalities (4).
Previously being classified as diffuse large B-cell lymphoma (DLBCL), PBL was later recognized as a distinct entity (5) with well-defined histopathologic features including large plasmablastic or immunoblastic cell morphology, loss of B-cell lineage markers (CD20, PAX5), expression of plasma-cytic differentiation markers (CD38, CD138, IRF4/MUM1, BLIMP1), a high proliferation index, and frequent infection by the Epstein-Barr virus (EBV; ref. 6). While EBV infection and activating MYC translocations have been reported as the major features of these tumors in a number of studies (7), the molecular pathology of PBL remains elusive in both HIV-positive and -negative individuals. Numerous case stud-ies and some cohorts have been well described and reviewed (6–9). Only two studies have partially explored genetic aber-rations in this disease: a comparative genomic hybridiza-tion study showed a pattern of segmental gains in PBL that more closely resembled DLBCL than plasma cell mye-loma (10). The second study compared the transcriptional
Liu et al.RESEARCH ARTICLE
114 | BLOOD CANCER DISCOVERY�JULY 2020 AACRJournals.org
profiles of 15 PBL cases to DLBCL and extraosseous plasma-cytoma (11), and showed that B-cell receptor signaling genes including CD79A/B, BLK, LYN, and SWAP70 among others, were significantly downregulated in PBL compared with DLBCL; in contrast, targets of MYB and the oncogene MYC, as well as genes reflecting the known plasmacytic immu-nophenotype of PBL, were overexpressed. More recently, MYC and PRDM1 were investigated for mutations, struc-tural rearrangements, and copy number gains in 36 PBL cases (12). PRDM1 mutations were found in 8 of 16 cases, frequently in association with MYC overexpression, sug-gesting a coordinate role in the pathogenesis of the disease. While these studies have shed light on this rare disorder, a systematic characterization of protein-changing alterations in PBL has not been performed.
The elucidation of genes and pathways that drive the initiation and maintenance of PBL is essential to better understand the biology of this cancer and, critically, to imple-ment improved biomarkers and more effective treatment options. Here, we combined whole-exome and transcriptome sequencing followed by targeted resequencing of 110 HIV-associated PBL cases to elucidate the mutational, transcrip-tional, and copy number landscape of this disease. We show that mutations in various components of the JAK–STAT and MAPK–ERK pathway pervade this lymphoma, revealing this signaling cascade as a central oncogenic driver of the disease and a candidate for targeted therapy.
RESULTSThe Landscape of Somatic Mutations in PBL
To determine the mutational landscape of PBL, we per-formed whole-exome sequencing (WES) in a discovery panel of paired tumor and normal DNA collected from 15 HIV-positive patients, followed by deep targeted sequencing of the top 34 candidate genes in 95 additional cases (Supplemen-tary Table S1 and S2; Methods). Somatic mutations were identified from WES data using SAVI2 (13), an empirical Bayesian method. Overall, 2,149 nonsynonymous somatic variants were found in the 15 discovery cases, with a median of 45 per case and a total of 1,528 affected genes, of which 1,461 were mutated in the tumor-dominant clone (>15% cancer-specific allele frequency; Supplementary Table S3). Among the 15 WES cases, one was hypermutated (case PJ030), showing 845 sequence variants that were confirmed by RNA-sequencing (RNA-seq; 92% with a read depth ° 10; Supplementary Fig. S1A–S1C). Candidate genes were then selected for an extension screen based on the following cri-teria: (i) mutated in at least 2 discovery cases, (ii) expressed in normal and/or malignant B cells, (iii) known as a cancer driver gene, and/or (iv) with an established role in B-cell dif-ferentiation (Supplementary Tables S4 and S5; Methods). The 34 selected genes were all annotated in the Catalogue of Somatic Mutations in Cancer (COSMIC) database. The mean depth of coverage for WES was 54.0x. The mean depth of coverage for the targeted DNA sequencing was 121.7x and, on average, 99.7% of the target sequences were covered by at least 50 reads.
We found genes in the JAK–STAT, MAPK–ERK, and Notch signaling pathways were commonly mutated in our
PBL cohort (Fig. 1A–E). The most frequent genetic lesions affected the JAK–STAT signaling pathway with, in total, 62% of cases (68/110) harboring a mutation in at least one of five genes (STAT3, JAK1, SOCS1, JAK2, and PIM1, a direct transcriptional target of STAT3/5 that functions as part of a negative feedback loop; Fig. 1A). Among these, STAT3 was the predominant target of mutations, with 46 of 110 cases (42%) harboring clonal events (Supplementary Fig. S2A and S2B). With three exceptions, the identified variants were het-erozygous missense mutations clustering in exons 19 to 22, encoding part of the SH2 domain that is required for STAT3 molecular activation via receptor association and tyrosine phosphodimer formation (14). In particular, the majority of the mutations resulted in the amino acid changes Y640F (n = 11), D661V (n = 9), S614R (n = 5), and E616G/K (n = 4; Fig. 1B), which have been categorized as gain-of-function or likely gain-of-function mutations based on experimental validation (https://oncokb.org/gene/STAT3), and overlap with the STAT3 mutation pattern described in other aggres-sive B-cell lymphomas and T-cell and natural killer (NK) cell malignancies (15, 16). Three additional cases showed a >75% variant allele frequency in the absence of copy number changes, consistent with a copy neutral loss of heterozygo-sis. Sanger-based resequencing of the involved STAT3 exons, performed in 23 cases, validated all computationally identi-fied mutations on these samples (Fig. 1F; Supplementary Table S6).
In addition to STAT3, 15/110 PBL cases harbored heterozy-gous missense mutations of JAK1, encoding for a tyrosine kinase that phosphorylates STAT proteins. These events did not involve the canonical hotspot seen in many other can-cers (loss-of-function K860Nfs), but occurred at a highly conserved amino acid position in the JH1 kinase domain, G1097D/V (Fig. 1C). Mutations at the JAK1 G1097 codon were previously reported in intestinal T-cell lymphomas (17) and anaplastic large cell lymphoma (15), and the G1097D substitution has been documented as a gain-of-function event that triggers aberrant phosphorylation of STAT3, lead-ing to constitutive activation of the JAK–STAT signaling pathway (15). Of note, missense mutations in STAT3 and JAK1 were found to be mutually exclusive (P = 0.04, Fisher exact test), suggesting a convergent role in activating this cascade. In contrast to STAT3 and JAK1, mutations in SOCS1 (n = 11/110 cases) were more widely distributed, consistent with its previously established role as a target of AID-medi-ated aberrant somatic hypermutation (Fig. 1D). Although the consequences of most SOCS1 amino acid changes will require functional validation, the presence of a start-loss mutation in one case and a nonsense mutation in a second sample is expected to result in loss-of-function and inactiva-tion of this negative JAK/STAT regulator, consistent with a tumor suppressor role.
The second most commonly mutated program was the MAPK–ERK signaling pathway, affected in 28% of cases by mutations in the RAS gene family members NRAS (14%) and KRAS (9%), as well as in BRAF (5.5%) and MAP2K1 (3%). Nearly all (92.5%) RAS mutations were found at known func-tional hotspots that have been shown to affect the intrinsic RAS GTPase activity, namely G12, G13, and Q61 (ref. 18; Fig. 1E). BRAF, an integral component of the MAPK
Genomic Characterization of Plasmablastic Lymphoma RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 115
Figure 1. The landscape of putative driver gene mutations in PBL. A, Sample information, MYC translocation, and somatic mutation information are shown for 110 cases of PBL samples. The heatmap represents individual mutations in each sample, color-coded by type of mutation. B–E, Individual gene mutation maps for frequently mutated genes, showing mutation subtype, position, and evidence of mutational hotspots, based on COSMIC database information. Y-axis counts at the bottom of the maps reflect the number of identified mutations in the COSMIC database. F, Sanger validation of single nucleotide variants (SNV) in STAT3 mutants.
120
60
0
STAT int STAT alpha STAT bind SH21 770 aa
STAT3Chr17:42313324-42388568
Y640
F (1
1)D
661V
/N (9
)
S614
R (5
)
P715L (1)N647I (5)
E616
,G61
8 (7
)
N56
7K (2
)
C42
6R (2
)
R15
2W (1
)
Pkinase Tyr Pkinase Tyr 1154 aa1
JAK1Chr1:64833229-64966504
MissenseSplice G1097 (13)
X989 (1)S683
N (1
)K6
96I (
1)A6
99V
(1)
0
15
30
14
7
0
SH2 SOCS box1 211 aa
SOCS1Chr16:11254417-11256200
MissenseStart lostFrameshift
S125
N (2
)
G13
9V (1
)
R16
9G (1
)
T100
(2)
F79L
(2)
F58L
(2)
Y64*
(1)
M1T (1)
0
2,000
4,000
Ras1 189 aa
NRASChr1:114704469-114716894
Missense
Q61 (7)G13 (7)
G12D (2)
B C
CO
SMIC
CO
SMIC
D E
A
A T G G T T G TC C C A T T A AC CMet Asp Ala Pro Tyr Thr
A A
F PJ046 STAT3 D661V PJ042 STAT3 Y640F
Chr17:40474482Chr17:40474419
SART1UBR5
NFKBIEBLM
PTPRD
LTBNPHP4
B2MTNFRSF14
KLF2FOXP1PRDM1
TP53MYC
HDAC6TRRAPEP300
KMT2ATET2
NOTCH1SPEN
NCOR2MAP2K1
HSPA8
KRASBRAFNRASJAK2PIM1
SOCS1JAK1
STAT3MYC translocation
Sample sitePos Neg
DNAseq110 cases
JAK-STAT
MAPK
NOTCH
Others
Epigenetics
Other TF
Immuneevasion
Stop lostStart lost
Splice regionMissenseInframe indel Frameshift indel
WESTargeted DNAOral Not oral
# cases0 20 40
MissenseInframeFrameshift
signaling cascade, was found mutated in 6 cases, including 3 harboring mutations at or around the well-characterized V600 residue (namely: V600E, K601N, and T599TT) and 3 showing mutations at other common hotspots within the protein kinase domain (G464E, V471F, and M689V). The Valine at position 600 normally stabilizes the interaction between the BRAF glycine-rich loop and the activation
segments, and its glutamine substitution confers an over 500-fold increase in activity, leading to constitutive activa-tion of the MEK/ERK signaling cascade in the absence of extracellular stimuli. In line with their predicted gain of function, the mutant alleles in both RAS family members and BRAF were observed in heterozygosis (Supplementary Tables S3 and S5) and were actively expressed (median
Liu et al.RESEARCH ARTICLE
116 | BLOOD CANCER DISCOVERY�JULY 2020 AACRJournals.org
RPKM value of KRAS: 13.253, NRAS: 35.601, and BRAF: 9.906).
In addition, 24% of PBL cases carried mutations in genes implicated in the Notch signaling pathway, including those encoding for NOTCH1, its negative regulator SPEN, and the Notch pathway corepressor NCOR2. In particular, one sample displayed a frameshift variant in the NOTCH1 gene that is predicted to generate a truncated protein lacking the C-terminal PEST domain, and thus endowed with increased protein half-life, while 6 additional cases showed amino acid changes within the EGF repeats, the juxtamembrane heterodimerization domain (N-terminal portion), and the C-terminal PEST domain. SPEN mutations include monoal-lelic missense substitutions that were distributed along the protein coding exons with no particular clustering, and their functional effect remains to be determined.
Missense mutations, frequently multiple within the same allele, were also found in the MYC gene (10/110, 9%), with 4 additional cases displaying silent mutations that possibly reflect the aberrant activity of the physiologic somatic hypermutation mechanism (19) as well as the recruitment of the AID mutator enzyme by the juxtaposed immunoglobulin enhancer in cases harboring chromosomal MYC translocations. MYC rearrange-ments with the immunoglobulin genes are the most common cytogenetic feature in PBL, present in about half of reported cases (7, 12, 20), and were identified in 36% (31/76) of our samples using FISH (Fig. 1A). The functional consequences of MYC mutations will have to be experimentally tested; however, transcriptomic analysis revealed significantly higher expression of the MYC RNA in cases positive for the MYC translocation, consistent with MYC oncogenic activation (Supplementary Fig. S3A and S3B). The high proportion of cases showing evidence of MYC dysregulation reinforces the notion that MYC is a criti-cal contributor to this aggressive cancer.
Aside from the genes mentioned above, we observed recurrent mutations, including truncating loss-of-function events, in TET2 (10/110, 9%), TP53 (10/110, 9%), and NPHP4 (6/110, 5%). Finally, genes encoding epigenetic modifiers, transcription factors implicated in B-cell activation (FOXP1), positioning (KLF2), and terminal differentiation (PRDM1), and receptor molecules involved in tumor immune surveil-lance (e.g., B2M, TNFRSF14) were mutated to a lesser extent (range: 1%–5%; Fig. 1A).
Although the relatively small number of cases prevents robust statistical analyses, we did not find any significant difference in the rate of mutated genes between samples collected from the oral cavity and from other locations (Pearson correlation coef-ficient = 0.822, Supplementary Fig. S2B), suggesting a genetic homogeneity of the disease regardless of the tissue of origin.
Collectively, the data presented above uncover a pervasive role for mutations affecting the JAK–STAT and MAPK–ERK pathways in the genetic landscape of PBL, which may con-tribute to the pathogenesis of this lymphoma by enforcing constitutive signaling activation, in concert with dysregu-lated MYC activity.
Somatic Copy Number ChangesTo identify recurrent copy number alterations (CNA) asso-
ciated with PBL, we applied the SNP-FASST2 algorithm to WES data from the 15 discovery cases (Supplementary Fig.
S4A), followed by validation in an independent cohort of 31 additional PBL samples that had been processed with Affymetrix OncoScan microarrays (ref. 21; Supplementary Table S7). Comparison of CNA calls obtained in parallel from the WES and OncoScan approach in a subset of 9 samples confirmed the data were highly consistent with each other (Supplementary Fig. S4B), supporting the robustness of the analysis.
Frequent copy number gains involved large chromosomal regions (>10 Mb) including chromosome 1q (20/46 cases, 43%) and the whole or most of chromosome 7 (13/46 cases, 28%). When applying the GISTIC 2.0 algorithm to the com-bined cohort of 46 cases, we uncovered regions of highly recurrent amplification including 6p22.2 (the most signifi-cant), 6p22.1 and 1q21.3, all of which encompass histone gene clusters (Fig. 2A). In particular, the chromosome 6p gain covered 36 genes that encode canonical histones and has been previously reported as a common alteration in a variety of cancers, where histone gains have been linked to genetic instability (22). The significant region on chromo-some 1q21.3 also included the IL6R gene and the antiapop-totic MCL1 gene, which showed increased gene expression (Fig. 2C) analogous to what has been described in 26% of activated B-cell like (ABC) DLBCLs (23). Although further studies will be needed to determine the functional impact of these alterations in PBL, chromosome 1q gains have been associated with unfavorable prognosis in multiple myeloma, suggesting a role in the pathobiology of the disease.
The second most significant focal CNA was a chromo-some 11p13 regional gain targeting genes CD44 and PDHX, present in 17 of 46 cases (37%; Fig. 2A–C). CD44 is a nonki-nase transmembrane glycoprotein that is induced in B cells upon antigen-mediated activation and is critically involved in multiple lymphocyte functions, including migration, hom-ing, and the transmission of signals that regulate apoptosis. This CD44 protein is also thought to increase cancer cells’ adaptive plasticity in response to the microenvironment, thus giving them a survival and growth advantage (24). Analysis of 20 samples with available RNA-seq data revealed that CD44was consistently and highly expressed in cases harboring copy number gains (n = 2), whereas PDHX levels were undetect-able, indicating that CD44 is the specific target of the 11p13 amplicon. However, all samples displayed high CD44 mRNA expression, independent of the presence of genetic aberrations (Fig. 2C); consistently, IHC analysis with a specific CD44 anti-body showed very strong membranous staining in all cases tested (Fig. 2D) and confirmed this finding in an independent panel of 38 cases. Of note, although CD44 expression can be detected in normal plasma cells at both RNA and protein levels, the signal was markedly lower than in plasmablastic lymphoma cells, suggesting that elevated expression of CD44 does not simply reflect the cellular ontogeny of these tumors, and that alternative regulatory mechanisms may lead to CD44 upregu-lation in cases lacking CNAs (Supplementary Fig. S5A–S5C; Supplementary Table S8).
Plasmablastic Lymphoma Displays a Distinct Genetic and Transcriptional Program
To explore the genomic features of PBL in relation to other lymphoid neoplasms arising from the mature B-cell
Genomic Characterization of Plasmablastic Lymphoma RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 117
Figure 2. Recurrent copy number changes in PBL. A, GISTIC 2.0 results showing recurrent copy number changes in PBL samples. The green line indicates q-value = 1.0 × 10−6. B, A zoomed-in view of 11p13 on 17 cases of PBL, which shows consistent focal copy number gains of CD44. The figure was generated by the IGV browser using CNV segment files from SNP-FASST2 algorithm. C, Scatter plot representations of genes located in regions with recurrent copy number gains in PBL (q-value <1.0 × 10−6, GISTIC 2.0). The horizontal axis indicates the −log10(q-value) from the GISTIC report, and the ver-tical axis is the median gene expression level (normalized RPKM value) from PBL RNA-seq data (n = 20). D, IHC showing strong CD44 protein expression on tumor cells membrane in a representative sample (A10), which has a focal CD44 copy number gain.
1.10.8 1q21.31q23.11q31.31q32.11q432p24.12q132q22.12q32.3
6p22.26p22.16q23.2
7q34
11p13
1
2
34
56
78
910
11
1213
1415
1617
1819
202122
1
2
34
56
78
910
11
1213
1415
1617
1819
202122
0.40.320.2
10−1010−3 10−610−910−4 10−6
1p36.21
10q26.3
APRAMEF2HNRNPCL1PRAMEF10PRAMEF8PRAMEF4PRAMEF11PRAMEF7
TCERG1LMIR378C
HIST2H2AA3HIST2H2ACHIST2H2BEHIST2H4AHIST2H3CHIST2H2AB
HIST2H3AHIST2H2BCHIST2H2BFHIST2H4BHIST2H3DHIST2H2AA4
HIST1H1DHIST1H1EHIST1H1THIST1H2AEHIST1H2ADHIST1H2BDHIST1H2ACHIST1H2BG
HIST1H2BFHIST1H2BEHIST1H2BHHIST1H2BIHIST1H2BCHIST1H3DHIST1H3EHIST1H3G
HIST1H4DHIST1H2BKHIST1H4FHIST1H4CHIST1H4HHIST1H4EHIST1H4G...
PDHXCD44
chr11:32,620,498-44,306,998B
Chr 11
33 Mb 35 Mb 37 Mb 39 Mb 41 Mb 43 Mb
37%
Gai
n
CD44 (chr11:35,160,417-35,253,949)
Sam
ple
PJ021PJ026PJ030
A05A08D03A10B03B05B07B08B09B11C03C05C06C11
D
Sample: A10
C
−Log10 (q-value) from GISTIC 2.0
Gen
e ex
pres
sion
leve
l in
PBL
15q11.2
16p13.3
22q13.32
12p11.23
18q11.1
7q11.21
6p22.211p131q21.3
6 8 10 12 14 16
150
100
50
0
HIST1H1E
HIST1H4CHIST1H2BG
HIST1H2BI
HIST1H2BCHIST1H3G
MCL1HIST1H1B
HIST1H2BM
CD44
PDHX
HIST1H2AJHIST2H2AC
HIST2H2BE
HIST1H2AI
BUB1
HIST1H2BLADAR
HIST1H2AE
HIST1H2BHHIST1H4D
HIST1H2BDHIST1H1D
6p22.1
Liu et al.RESEARCH ARTICLE
118 | BLOOD CANCER DISCOVERY�JULY 2020 AACRJournals.org
lineage, we performed unsupervised clustering analysis based on mutation frequency of the top mutated genes from three cancer types obtained from public repositories (Fig. 3A). The analysis included chronic lymphocytic leukemia (CLL; ref. 25), diffuse large B-cell lymphoma (DLBCL) transcription-ally defined as ABC and germinal center B-cell-like (GCB) subtypes (26), and multiple myeloma (27). As expected on the basis of their presumed derivation from B cells committed to plasma cell differentiation, the mutational landscape of PBL was overall closer to multiple myeloma than to other mature B-cell malignancies, with mutations in RAS family members being detected in as many as 20% of cases in both diseases, while rare in DLBCL (Fig. 3A). Conversely, the mutational landscape of PBL was highly distinct from that of both GCB- and ABC-DLBCL (Fig. 3A). In particular, mutations affecting the methyltransferase KMT2D and acetyltransferase CREBBP, two among the most commonly mutated genes in DLBCL (28), were absent in plasmablastic lymphoma. ABC-DLBCL–specific mutations such as CD79A/B and MYD88 were also lacking in PBL. Of note, STAT3 was the top mutated gene in our cohort at significantly different frequencies compared with other B-cell malignancies, making it a hallmark of PBL (Fig. 3A). The differential genetic landscape of PBL is consist-ent with its status as a distinct entity among mature B-cell neoplasms.
To define the transcriptional profile of HIV-positive PBL and to identify unique signatures that may distinguish it from other lymphoma types, we performed RNA-seq analysis of 20 PBL samples (including 12 of the 15 discovery cases), and compared their transcriptome to that of normal B-cell subsets and other B-cell lymphoma types previously characterized in our laboratories and/or obtained from pub-lic repositories, including germinal center centroblasts (CB), naïve B cells (NB), and memory B cells (MB; ref. 29), as well as CLL (30), DLBCL (26), and multiple myeloma cell lines (31). As expected, hierarchical clustering of the top 1,000 most aberrantly expressed genes revealed that PBL and multiple myeloma were closer to each other compared with other B-cell malignancies and to normal B cells, reflecting their presumed cell of origin (Fig. 3B). Consistently, plasma-blastic lymphoma and multiple myeloma lacked expression of common B-cell markers (CD19, CD20, CD40, and PAX5) and transcription factors involved in the germinal center reaction (BCL6, BCL7A, BCL11A, and SPIB), whereas expres-sion of the master regulator of plasma cell differentiation PRDM1 and other plasma cell markers (CD138, XBP1, and IRF4) were increased (Supplementary Fig. S6). MYC expres-sion was also higher in PBL and multiple myeloma. Other notable differences included the upregulation of IL6R, a known STAT3 target, and the downregulation of SWAP70, previously suggested as a potential biomarker of PBL (Sup-plementary Fig. S6; ref. 11). Finally, recent studies have sug-gested that EBV positive PBLs evade immune recognition
by expressing the programmed cell death protein 1 (PD-1) and its PD-L1 ligand (32). We confirmed high expression of PD-L1 in our PBL cohort, although PD-1 did not follow this pattern (Supplementary Fig. S6).
We then performed functional enrichment analysis (g:profiler) to identify biological pathways that are pref-erentially enriched in PBL as compared with other B-cell malignancies. Whereas DLBCL and CLL were enriched for B-cell related biological programs (e.g., B-cell activation and proliferation, lymphocyte activation and differentiation, and adaptive immune response), multiple myeloma was enriched for mitotic cell-cycle processes, with highly expressed genes including MCM10, BIRC5, CENPE, BUB1, and AURKA (Fig. 3B). The most significantly enriched pathways in PBL were related to chromatin/nucleosome assembly (Fig. 3B); par-ticularly, histone genes encoding basic nuclear proteins were highly expressed in PBL, possibly related to the recurrent copy number gains/amplification encompassing this gene cluster on chromosomes 6p22.2 and 1q21.3 (Supplemen-tary Fig. S7A–S7D). However, the biological significance of this observation in relation to tumorigenesis remains unknown.
Activation of the JAK–STAT Pathway in Plasmablastic Lymphoma
Given the elevated frequency of mutations targeting the JAK–STAT pathway, we used computational and IHC approaches to assess the extent of activation of this signal-ing cascade in PBL (Methods). To this end, we first inter-rogated the genome-wide transcriptional profile of PBL for enrichment in the JAK–STAT signaling pathway using previously identified signatures available in the MSigDB database. This analysis revealed a significant positive enrichment for genes implicated in this pathway, consist-ent with constitutive activation (Fig. 4A). This signature was expressed at higher levels in PBL when compared with multiple myeloma and CLL (Fig. 4B), with the most signifi-cant difference being observed between PBL and multiple myeloma (t test, P = 4.7 × 10−11), which is congruent with the fact that the latter had few mutations in this pathway (Fig. 3A). We then performed IHC staining with anti-STAT3 and anti-phospho-STAT3 antibodies to assess the STAT3 subcellular localization and phosphorylation, used as a read-out for signaling activation. We observed strong positive nuclear signal in 61% of tested cases (19/31), which also stained positive for the phospho-STAT3 antibody. Of note, evidence of constitutively active STAT3 was detected even in the absence of genetic alterations affecting this pathway, suggesting alternative (epigenetic) mechanisms of activation and indicating a prominent role for this cascade in the pathogenesis of the disease (Fig. 4C and D; Supplementary Fig. S8; Supplementary Table S9).
Figure 3. Comparative analysis of PBL and other B-cell malignancies. A, Unsupervised clustering based on frequencies of the most recurrently mutated genes from PBL, multiple myeloma, CLL, and two main subtypes of DLBCL (Methods). B, Hierarchical clustering of mRNA expression profiles across plasmablastic lymphoma, multiple myeloma, CLL, DLBCL and normal B cells, including centroblast (CB), naïve B (NB), and memory B cells (MB). Note that the JAK–STAT signaling pathway does not appear in this figure because only the top 1,000 most aberrantly expressed genes were selected for this analysis. Functional enrichment analysis was performed using g:profiler.
Genomic Characterization of Plasmablastic Lymphoma RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 119
0.2 0.4 0.2 0.40.2 0.4MM ABC DLBCL GCB DLBCL
A
4
MMPBL
CBNBMBCLL
ABCGCB
B
Leukocyte activationImmune responseLymphocyte activationLymphocyte differentiation
B-cell activationB-cell proliferationAdaptive immune responseHemopoiesis
Nucleosome assemblyChromatin assemblyDNA packagingChromatin organization
Chromatin silencingNucleosome positioningHematopoietic developmentSensory perception of taste
Mitotic cell cycleNuclear divisionRegulation of cell cycleOrganelle fission
Chromosome segregationCell-cycle checkpointDNA metabolic processSpindle checkpoint
Immune system processCell activationDefense responseSecretion by cell
Cell adhesionInflammatory responseCell migrationNeutrophil degranulation
Z-score
0.2 0.4PBL
0.2 0.4CLL
0−4
STAT3NCOR2
TET2JAK1
NPHP4MYCLTB
KRASNRASDIS3
TENT5CBRAFFAT3
SF3B1ATM
TP53NOTCH1
POT1XPO1BIRC3
PIM1MYD88CD79BPRDM1
DTX1IRF2BP2
TNFRSF14CREBBP
SOCS1EZH2BCL2SGK1
GNA13B2M
BTG2BTG1
TNFAIP3CARD11
HIST1H1EMEF2BKLHL6CD58SPEN
Liu et al.RESEARCH ARTICLE
120 | BLOOD CANCER DISCOVERY�JULY 2020 AACRJournals.org
Figure 4. Activation of JAK–STAT pathway in PBL. A, Preranked gene set enrichment analysis indicating the significant positive enrichment of KEGG JAK–STAT signaling pathway in PBL. The preranked gene list was generated on the basis of the median expression level on PBL samples. B, Enrichment comparisons of JAK–STAT pathway between PBL and other B-cell malignancies by performing single-sample GSEA (ssGSEA). Pairwise P values were derived from t test. C and D, IHC plots showing pSTAT3 protein expression in >75% of tumor cells, confirming STAT3 activation in STAT3 mutated cases. Results using anti-pSTAT3 antibody are photographed at ×100 and ×400 magnification.
A
0.00
0.25
0.50
0.75
PBL MM CLL ABC GCB
Enric
hmen
t sco
re o
f ssG
SEA
KEGG JAK–STAT pathwayB
C
PBL vs. MM P = 4.7e-11PBL vs. CLL P = 2.3e-07PBL vs. DLBCL P = 0.94
0 10,000 20,000 30,000 40,000 50,000Rank in ordered dataset
0.0
5.0
10.0
Ran
ked
list m
etric
Pos (positively correlated)
Neg (negatively correlated)
Zero score at 19,158
0.0
0.2
0.4
Enric
hmen
t sco
re (E
S)
P: 0.000NES: 2.088
KEGG JAK–STAT signaling pathway
PJ122 N567K
N567K
×100 ×400
×100 ×400PJ154
PJ159 D661V×100 ×400
×100 ×400PJ225 Y640F
D
Virus Detection in Plasmablastic Lymphoma
To analyze the viral and bacterial make-up of PBL tumors, we used the Pandora pipeline, which extracts and aligns nonhost genetic material from tumor RNA-seq data (Meth-ods). Potentially pathogenic species were then identified by applying the BLAST algorithm against the NCBI database of viruses and bacteria reference genomes. In our PBL cohort, 12 of 20 samples contained HIV-1 transcripts, with only three samples exceeding 100 mapped reads (Supplementary
Fig. S9). In addition, EBV transcripts were detected in 18 of 20 samples, HCMV (human cytomegalovirus, or human betaherpesvirus 5) in 3 of 20 samples, and KSHV (Kaposi sar-coma-associated herpes virus, human herpes virus 8) in one sample (Supplementary Fig. S9). These three herpes viruses have a broad tropism, naturally infect B cells, and are known to be associated with many tumor types.
EBV reactivation is considered a major driver of PBL, predominantly with a latency I infection program although low levels of LMP1 gene expression can be detected (33).
Genomic Characterization of Plasmablastic Lymphoma RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 121
We thus performed a detailed RNA profiling of the EBV genome. Recapitulating previous reports, all PBL cases tested by in situ hybridization were positive for EBV-encoded RNA (EBER), while only 4 of 13 samples showed expression of the LMP1 protein on IHC (Supplementary Table S1). When using the levels of BLLF1 and EBNA2, two genes that are invariably not expressed in PBL, as threshold for posi-tive calls, we noted that nearly the entire viral genome was transcribed at background levels in the majority of sam-ples (Fig. 5). This is analogous to what has been observed in nonreplicating infected B cells, where a background of transcripts can be detected in the absence of protein expression, due to regulation at the ribosomal level (34). However, several genes in the BamHI-A region of the virus were abundantly transcribed in at least 50% of cases across the cohort, including those encoding for components of the viral replication machinery (namely, the DNA polymerase catalytic subunit BALF5, the single-stranded DNA-binding protein BALF2, and the lytic origin of DNA replication oriLyt), the BART encoded protein RPMS1, the viral envelope glycoprotein BALF4, and the G-protein–coupled receptor BILF1, which is expressed predominantly during the imme-diate early phases of infection in vitro (35, 36). LF1, LF2, and LF3 were also highly expressed in PBL, but their functional roles are less understood. Expression of the LMP-1 mRNA was detected in 17 of 20 samples, but at much lower levels (Fig. 5), consistent with the known EBV latency programs of PBL (33).
DISCUSSIONOur study provides a comprehensive snapshot of the
genetic landscape of HIV-associated PBL and reveals highly recurrent somatic mutations affecting the JAK–STAT3 and RAS–MAPK signaling pathway as a genetic hallmark of this disease, with STAT3 representing the most prominent target.
These findings underscore a central role for this transcription factor in the pathogenesis of PBL and have implications for the diagnosis and treatment of these diseases.
The STAT3 protein is an important player in multiple immune cells where it modulates a variety of physiological processes. Within the B-cell lineage, a selective role has been recognized for this factor in the differentiation of B cells into plasma cells upon antigen stimulation, as documented by both in vitro and in vivo studies (14). Briefly, in response to CD40L- and IL21-mediated signaling by T follicular heper cells, STAT3 is phosphorylated by activated JAK kinases, translocates into the nucleus as homo- or hetero-dimers, and activates the transcription of multiple targets, including the plasma cell master regulator BLIMP1. In turn, BLIMP1 down-regulates BCL6 expression, an absolute prerequisite to GC exit and plasma cell differentiation (14). The STAT3 amino acid changes identified in our study include experimentally documented gain-of-function events that are predicted to have oncogenic effects by enhancing its phosphorylation and transactivation potential (37). In addition, other well- documented genetic mechanisms were found that can acti-vate the JAK–STAT signaling pathway, including mutually exclusive upstream mutations of JAK1 or JAK2 (18/110 cases) and the loss of the STAT3 negative regulator SOCS1 (mutated in 11/110 cases). Notably, a targeted sequencing study pub-lished while this manuscript was under revision identified recurrent STAT3 mutations in 5 of 42 cases, all of which were EBV positive (38). Thus, STAT3 dysregulation may contribute to the pathogenesis of PBL by rendering these cells signaling-independent while providing proliferation and survival signals.
Constitutive activation of the JAK–STAT signaling path-way has been reported in a number of solid and hematologic malignancies and plays a central role in two lymphomas that are immunophenotypically closely related to plasma-blastic lymphoma: primary effusion lymphoma (PEL) and
Figure 5. EBV transcription programs in PBL. Heatmap illustrating the full genome expression of EBV in the RNA-seq data of PBL samples (virus gene counts are normalized per million host reads). 16/17 EBV-positive samples show significantly higher expression of lytic genes such as BALF4 (encoding the envelope glycoprotein B) and BALF5 (encoding the DNA polymerase catalytic subunit) compared with the expression of canonical latency programs.
PJ008PJ117PJ024PJ046PJ042PJ123PJ049PJ121PJ030PJ021PJ020PJ007PJ039PJ120PJ119PJ026PJ129PJ122PJ028PJ125
Cp
TRBD
LF3.
5BG
LF3.
5BX
RF1
BVLF
1BB
RF2
BTR
F1oriP
BGLF
3BD
LF4
BcR
F1BF
RF1
ABF
LF1
BGR
F1/B
DR
F1EB
NA−
3BEB
NA−
3CBB
LF2/
BBLF
3Ba
RF1
BcLF
1BF
RF2
BFLF
2or
iLyt
BSLF
2/BM
LF1
LF3
RPM
S1LF1
LF2
BBLF
1BC
RF1
BLR
F1BF
RF3
BKR
F2BS
RF1
BVR
F2BV
RF1
BDLF
1BB
RF1
BLR
F2BZ
LF2
BBR
F3BG
LF1
BDLF
2BG
LF2
BRR
F2BD
LF3
BALF
3BI
LF2
BLLF
1BX
LF2
BOR
F1BN
RF1
BOLF
1BK
RF4
BMR
F2BP
LF1
BILF
1BA
LF4
BBLF
4BG
LF4
BLLF
3BS
LF1
BXLF
1BR
RF1
BGLF
5BF
RF1
BKR
F3BR
LF1
BALF
1BA
RF1
BMR
F1BO
RF2
BHR
F1BH
LF1
BALF
2BA
LF5
LMP−
2BEB
NA−
2EB
NA−
1EB
NA−
3ALM
P−2A
LMP−
1
0 60 120
Normalized virus read counts
Latent Early lytic Late lyticLatency II,III Latency III Latency I,II,III
Liu et al.RESEARCH ARTICLE
122 | BLOOD CANCER DISCOVERY�JULY 2020 AACRJournals.org
ALK-positive large B-cell lymphoma (LBCL). PEL is a rare and aggressive AIDS-defining disease, which is associated with infection by HHV8 and is clinically distinguishable from PBL by the presence of lymphomatous effusion in body cavities. In these cells, constitutive STAT3 activity is achieved via expression of the HHV8 viral protein IL6, which contributes to the disease in an autocrine fashion by promoting proliferation and survival (39). In ALK-positive LBCL, STAT3 activation is sustained by the ALK kinase mediated by chromosomal translocations with the CLTCL gene or the NPM1 gene (40, 41). However, direct genetic alterations of STAT3 are rare in mature B-cell lympho-mas. In particular, while multiple genomic hits leading to potentiation of the JAK–STAT oncogenic pathway have been detected in 87% of Hodgkin lymphomas as well as in primary mediastinal B-cell lymphomas, the most commonly affected STAT member in these tumors is STAT6 (42–45). The high incidence of STAT3 mutational activation in HIV-associated PBL points to STAT or JAK inhibitors as promising treatment options in this lymphoma type. While anti-STAT3 therapeutic attempts are still in development, JAK inhibitor therapy, currently used in the clinical setting, was shown to be an effective antagonist to STAT3 activa-tion, inducing apoptosis in both anaplastic large T-cell lymphomas and ovarian cancer (46).
The second important finding of this study is the iden-tification of frequent hotspot mutations in RAS–MAPK family members. Functional RAS activation is a common molecular feature of multiple myeloma, particularly in the relapsed/refractory setting, while it is rarely observed in de novo DLBCL, suggesting a specific role in the pathogen-esis of plasma cell dyscrasias. Interestingly, and different from multiple myeloma, NRAS, and KRAS mutations in PBL were never concurrently observed in the same case and were often well represented in the dominant tumor clone, consistent with early events. These data have direct implica-tions for the clinical exploration of treatments inhibiting this pathway.
Our study also showed overexpression of the transmem-brane glycoprotein CD44, frequently associated with copy number gains/amplifications at this locus. CD44 is an adhe-sion molecule that mediates cellular interaction with the microenvironment and participates in the trafficking of neoplastic cells in multiple myeloma, CLL, and ALL (47); moreover, CD44 was shown to increase cell resistance to apoptosis and to enhance cancer cell invasiveness. Whereas the role of CD44 in PBL requires functional dissection, its high expression is likely to compound the aggressiveness of the disease, as previously described in DLBCL of the ABC subtype (48). In light of this, successful anti-CD44– targeted therapy in a mice xenograft model of human multi-ple myeloma may in future represent an attractive therapeutic option for PBL (49).
The finding of increased histones mRNA abundance in PBL, together with recurrent copy number gains encom-passing this gene cluster on chromosome 6, is of interest because it emerged as a distinctive feature of this disease compared with other lymphoid malignancies. Histones rep-resent a basic component of the chromatin structure and their involvement may suggest a selective role for nucleoso-
mal plasticity in the pathogenesis of PBL, which warrants further investigations.
Due to the small number of EBV-negative cases in our cohort (n = 2) and the overall rarity of this subset, we could not assess whether EBV status is associated with differing genetic features, as recently reported for Burkitt lymphoma (50) and suggested by the evidence of distinct transcrip-tomic profiles between EBV-positive and –negative PBL (32). Moreover, it remains to be determined whether regional dif-ferences exist with PBL occurring outside the context of HIV immunodeficiency, or within HIV-associated PBL, given the high homogeneity of our cohort from the Gauteng region in South Africa. Thus, additional work will be required to specifically address these questions.
We found that most of the EBV genome was transcribed at very low levels in the majority of PBL cases, while prominent expression was detected for several genes in the BamHI-A region of the virus, including some early lytic genes. However, transcription of the key early lytic gene BZLF1 was noticeably absent, suggesting an incomplete lytic program. Supporting this notion, a recent study showed that increased STAT3 expression, as observed in the majority of PBL cases, decreases the susceptibility of latently infected cells to EBV lytic activa-tion signals via an RNA-binding protein PCBP2 (51). Thus, further protein expression studies, in particular for BRLF1 and for the BLLF1-encoded viral envelope glycoprotein gp350, are required to assess whether the observed transcriptome program translates in lytic infection. Indeed, recent transla-tion ribosome profiling studies have clearly demonstrated a marked heterogeneity of lytic genes translation and complex levels of intracellular translation repression mechanisms at work in infected B cells (34).
In conclusion, the results of our study characterize HIV-associated PBL as a distinct subset of aggressive B-cell lym-phoma and significantly contribute to our knowledge about the molecular pathogenesis of this disease through the iden-tification of recurrently mutated genes, uncovering a major role for the dysregulation of JAK–STAT3 and RAS–MAPK signaling pathways. These results reveal new points of poten-tial therapeutic intervention in these patients.
METHODSFor complete experimental details and computational analyses, see
also Supplementary Methods.
Patient CohortsProspective samples (n = 15 with paired tumor-normal tissue;
discovery cohort) were collected from patients with suspected PBL, upon informed written consent in line with the Declara-tion of Helsinki, and diagnosis was further confirmed by two independent pathologists. Matched normal DNA for these 15 patients was extracted from peripheral blood samples that were documented to be tumor-free. For targeted sequencing (n = 95 samples; extension cohort), formalin-fixed paraffin-embedded (FFPE) material was retrieved from the archives of the Department of Oral Pathology and Anatomical Pathology, National Health Laboratory Service and the University of the Witwatersrand. This panel included 36 nonoral PBL samples that were part of a pub-lished cohort (52). The study protocol was approved by the local Human Research Ethics Committee (IRB Reference M150390). A
Genomic Characterization of Plasmablastic Lymphoma RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 123
summary of demographics and phenotypic markers of the discovery cohort are displayed in Supplementary Table S1, whereas demo-graphic information for the extension cohort is included in Supple-mentary Table S3. For downstream nucleic acid extraction, the tumor area (>70% tumor cells) was ringed for microdissection in all samples. A third cohort of 31 well-characterized PBL samples obtained as archived FFPE material (local Human Research Ethics Committee IRB reference M101171 and 96/2011; Supplementary Table S7) was used to validate and refine copy number aberration results observed in the WES data by using Microarray OncoScan (21).
WES and RNA-SeqBoth exome and RNA-seq library preparations and sequencing
were outsourced to Centrillion Genomics Services and BGI (Americas Corporation). Whole-exome libraries were prepared using the Agi-lent SureSelect Human All Exon V6 Kit (Agilent Technologies) and sequenced on HiSeq 2500 using TruSeq SBS v2 Reagent Kit (Illumina) at 2 × 100 bp paired-end reads with on-target coverage of 100X per sample. RNA-seq libraries were prepared using the Illumina TruSeq Stranded Total RNA Sample Prep Kit (Illumina). Flow cells with multiplexed samples were run on the HiSeq 2500 using an Illu-mina TruSeq SBS v2 Reagent Kit at 2 × 100 bp paired-end reads and a coverage of 50M reads per sample.
Mutation CallingFastq files were aligned to the human genome assembly (hg19) using
the Burrows–Wheeler Aligner (version 0.6.2). Before further analysis, the initially aligned BAM files were subjected to preprocessing that sorted, indexed, and marked duplicated reads using SAMtools (ver-sion 1.7) and Picard (version 1). To identify somatic mutations from WES data for tumor samples with matched blood control, we applied the variant-calling software SAVI2, based on an empirical Bayesian method as published (13). Somatic mutations were identified on the basis of the final report of SAVI2, and following five additional criteria: (i) not annotated as a synonymous variant, intragenic variant, or intron variant; (ii) not annotated as a common SNP (dbSnp138); (iii) a variant allele frequency of >5% in the tumor sample; (iv) a variant allele read depth of <2 in the matched normal control; and (v) variant associated reads with overall mismatch rate ≤ 0.02 as estimated by bam-readcount (version 0.8.0, https://github.com/genome/bam-readcount). All muta-tions described throughout this manuscript refer to nonsynonymous mutations, unless otherwise specified.
Design of Targeted Sequencing PanelThe full coding exons of 34 genes found recurrently mutated in
the discovery cohort and/or previously implicated in lymphoma were analyzed in an extension panel of 95 cases (Supplementary Table S3) by targeted capture and next generation sequencing. Genes were selected according to the following criteria: (i) allele frequency > 15%; (ii) mutated in at least 2 of 15 cases; (iii) expressed in normal or transformed B cells; and (iv) functionally annotated. In addition, we included 16 genes that were only mutated in one sample but have known roles in the pathogenesis of lymphoma and 4 genes that were not found mutated in the discovery cohort but have been previously implicated in PBL. Genes lacking clear functional annotation and/or known to represent common nonspecific muta-tional targets in sequencing studies (e.g., TTN, PCLO) were excluded. The complete gene list is reported in Supplementary Table S4.
Targeted Next-Generation SequencingThe entire coding region of the 34 selected genes was isolated
using the IDT xGen Predesigned Gene Capture Custom Target Enrichment Technology (Integrated DNA Technologies) and sub-jected to library preparation and next-generation sequencing on the Illumina HiSeq platform with 2 × 150 bp configuration. Targeted
sequencing was performed at GENEWIZ. Read alignments and con-ventional preprocessing were conducted as described for the WES analysis. For samples lacking matched normal control, variants were filtered out if found in dbSNP database (dbSNP138) as well as in any normal sample of the WES cohort.
Copy Number AnalysisFor copy number analysis from WES data (n = 15), the Biodiscov-
ery Multiscale BAM Reference Builder (53) was used to construct a multiscale reference (MSR) file from 14 paired normal samples alignments (BAM). The MSR file was used as reference for copy-number variation (CNV) calling of all tumor alignments with the SNP-FASST2 algorithm (54), using Nexus Copy Number, v10.0 (Bio-Discovery, Inc.; ref. 54). Gains and losses were defined as at least +0.3 and −0.3 log2 ratio changes, respectively, in the tumor alignment.
To validate CNV calling from WES data, 31 additional samples were processed on Oncoscan FFPE Express Arrays (Affymetrix, Thermo Fisher Scientific) (Supplementary Table S7) according to the manu-facturer’s instructions, followed by scanning on a GeneChip Scanner 3000 7G, with the Affymetrix GeneChip Command Console (AGCC). Paired A+T and G+C CEL files were combined and analyzed with Chromosome Analysis Suite v3.3.0.139 (ChAS, Applied Biosystems, Thermo Fisher Scientific), using the OncoScan CNV workflow for FFPE, without manual recentering. We confirmed that copy number calls from OncoScan and WES data were consistent with each other by performing OncoScan on 7 cases that had been sequenced by WES and comparing the calls obtained from the two methods. Probe genomic coordinates were aligned to hg19 and the resulting OSCHP files were analyzed by Nexus Copy Number v10.0 using the SNP-FASST2 algorithm (54) with default parameters.
To detect recurrent copy number aberrations, we applied the GISTIC 2.0 algorithm using GenePattern (https://www.genepattern.org/) to the copy number segmentations of the combined cohort of 46 patients (15 cases analyzed by WES, 31 cases analyzed by OncoScan). Recurrent regions of copy number aberrations with q-value < 1.0 × 10−6 were considered significant. Furthermore, we assessed the expression level of each gene within the significant GISTIC peaks using the median value of quantile normalized RPKM from RNA-seq data (n = 20).
Pandora: A High-Performing Pipeline for Quantifying the Bacterial and Viral Microenvironment of Bulk RNA-Seq Samples
Pandora is an open-source pipeline (https://github.com/RabadanLab/Pandora) that takes as input the total RNA sequence reads from a single sample and outputs the spectrum of detected microbial transcripts, focusing on bacteria and viruses. The Pan-dora workflow is divided into the following four main modules: (i) mapping to the human host genome using STAR and Bowtie2 to filter out the host reads from downstream analysis; (ii) de novo assembly of host-subtracted short reads using Trinity (55) to create contiguous full-length transcripts (contigs), which help increase the accuracy of alignment to the correct species of origin in the next step; (iii) identification of the most likely species of origin for each assem-bled contig with BLAST; and (iv) filtering and parsing of the BLAST results into a final report on the detected microbial abundances. We also performed gene expression profiling of all the reads that mapped to the EBV genome (56) using FeatureCounts (57) to fully characterize lytic and latent EBV programs.
Data AvailabilityThe data that support the findings of this study are available upon
request. The sequencing data have been deposited in NCBI Sequence Read Archive (SRA) under accession number PRJNA598849.
Liu et al.RESEARCH ARTICLE
124 | BLOOD CANCER DISCOVERY�JULY 2020 AACRJournals.org
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors’ ContributionsZ. Liu: Conceptualization, investigation, visualization, method-
ology, writing-original draft, writing-review, and editing. I. Filip: Investigation, methodology, and writing-original draft. K. Gomez: Investigation, methodology, writing-review, and editing. D. Engel-brecht: Resources, validation, investigation, and visualization. S. Meer: Resources and investigation. P. Lalloo: Resources, validation, and investigation. P. Patel: Resources, validation and investigation. Y. Perner: Resources and investigation. J. Zhao: Investigation, visuali-zation, and methodology. J. Wang: Investigation, writing-review, and editing. L. Pasqualucci: Conceptualization, resources, supervision, funding acquisition, investigation, methodology, writing-original draft, project administration, writing-review, and editing. R. Rabadan: Conceptualization, supervision, funding acquisition, investigation, methodology, project administration, writing-review, and editing. P. Willem: conceptualization, resources, supervision, funding acqui-sition, validation, investigation, visualization, writing-original draft, project administration, writing-review, and editing.
AcknowledgmentsThis work has been funded by NIH grants R21 CA192854 (to
P. Willem, L. Pasqualucci, and R. Rabadan), R01GM117591 and U54 CA193313 (to R. Rabadan), and was initiated under the Columbia- South Africa Training Program for Research on HIV-associated Malignancies D43 CA153715, with the support of Judith Jacobson. We thank Stephen P. Goff and Henri-Jacques Delecluse for their helpful suggestions. We also thank Sonja Boy for providing addi-tional samples for the independent copy number validation cohort, Nicole Crawford for assistance in samples collection and data cura-tion, and Jacky Brown for help with Sanger sequencing. Whole exome capture and sequencing, and RNA sequencing were completed at Centrillion Biosciences, Inc and BGI Tech Solutions (Hong Kong). Targeted DNA sequencing was performed at Genewiz Inc.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received April 2, 2020; revised May 11, 2020; accepted May 11, 2020; published first June 10, 2020.
REFERENCES1. Campo E SH, Harris NL. WHO classification of tumors of the hemat-
opoietic and lymphoid tissues. In: Swerdlow SH, Campo E, Harris NL, editors. Lyon, France: IARC Press; 2017. p. 321–2.
2. Carbone A, Vaccher E, Gloghini A, Pantanowitz L, Abayomi A, de Paoli P, et al. Diagnosis and management of lymphomas and other cancers in HIV-infected patients. Nat Rev Clin Oncol 2014;11: 223–38.
3. Noy A, Lensing SY, Moore PC, Gupta N, Aboulafia D, Ambinder R, et al. Plasmablastic lymphoma is treatable in the HAART era. A 10 year retrospective by the AIDS Malignancy Consortium. Leuk Lymphoma 2016;57:1731–4.
4. Pather S, Mohamed Z, McLeod H, Pillay K. Large cell lymphoma: correlation of hiv status and prognosis with differentiation profiles assessed by immunophenotyping. Pathol Oncol Res 2013;19:695–705.
5. Delecluse HJ, Anagnostopoulos I, Dallenbach F, Hummel M, Mara-fioti T, Schneider U, et al. Plasmablastic lymphomas of the oral cav-ity: a new entity associated with the human immunodeficiency virus infection. Blood 1997;89:1413–20.
6. Castillo JJ, Bibas M, Miranda RN. The biology and treatment of plas-mablastic lymphoma. Blood 2015;125:2323–30.
7. Valera A, Balague O, Colomo L, Martinez A, Delabie J, Taddesse-Heath L, et al. IG/MYC rearrangements are the main cytogenetic alteration in plasmablastic lymphomas. Am J Surg Pathol 2010;34:1686–94.
8. Tchernonog E, Faurie P, Coppo P, Monjanel H, Bonnet A, Algarte Genin M, et al. Clinical characteristics and prognostic factors of plas-mablastic lymphoma patients: analysis of 135 patients from the LYSA group. Ann Oncol 2017;28:843–8.
9. Teruya-Feldstein J, Chiao E, Filippa D, Lin O, Comenzo R, Coleman M, et al. CD20-negative large-cell lymphoma with plasmablastic features: a clinically heterogenous spectrum in both HIV-positive and-negative patients. Ann Oncol 2004;15:1673–9.
10. Chang CC, Zhou X, Taylor JJ, Huang WT, Ren X, Monzon F, et al. Genomic profiling of plasmablastic lymphoma using array compara-tive genomic hybridization (aCGH): revealing significant overlapping genomic lesions with diffuse large B-cell lymphoma. J Hematol Oncol 2009;2:47.
11. Chapman J, Gentles AJ, Sujoy V, Vega F, Dumur CI, Blevins TL, et al. Gene expression analysis of plasmablastic lymphoma identifies downregulation of B-cell receptor signaling and additional unique transcriptional programs. Leukemia 2015;29:2270–3.
12. Montes-Moreno S, Martinez-Magunacelaya N, Zecchini-Barrese T, Villambrosia SG, Linares E, Ranchal T, et al. Plasmablastic lymphoma phenotype is determined by genetic alterations in MYC and PRDM1. Mod Pathol 2017;30:85–94.
13. Trifonov V, Pasqualucci L, Tiacci E, Falini B, Rabadan R. SAVI: a sta-tistical algorithm for variant frequency identification. BMC Syst Biol 2013;7:S2.
14. Hillmer EJ, Zhang H, Li HS, Watowich SS. STAT3 signaling in immu-nity. Cytokine Growth Factor Rev 2016;31:1–15.
15. Crescenzo R, Abate F, Lasorsa E, Tabbo F, Gaudiano M, Chiesa N, et al. Convergent mutations and kinase fusions lead to oncogenic STAT3 activation in anaplastic large cell lymphoma. Cancer Cell 2015; 27:516–32.
16. Song TL, Nairismagi ML, Laurensia Y, Lim JQ, Tan J, Li ZM, et al. Oncogenic activation of the STAT3 pathway drives PD-L1 expression in natural killer/T-cell lymphoma. Blood 2018;132:1146–58.
17. Nicolae A, Xi L, Pham TH, Pham TA, Navarro W, Meeker HG, et al. Mutations in the JAK/STAT and RAS signaling pathways are com-mon in intestinal T-cell lymphomas. Leukemia 2016;30:2245.
18. Prior IA, Lewis PD, Mattos C. A comprehensive survey of Ras muta-tions in cancer. Cancer Res 2012;72:2457–67.
19. Pasqualucci L, Neumeister P, Goossens T, Nanjangud G, Chaganti R, Küppers R, et al. Hypermutation of multiple proto-oncogenes in B-cell diffuse large-cell lymphomas. Nature 2001;412:341–6.
20. Taddesse-Heath L, Meloni-Ehrig A, Scheerle J, Kelly JC, Jaffe ES. Plas-mablastic lymphoma with MYC translocation: evidence for a com-mon pathway in the generation of plasmablastic features. Mod Pathol 2010;23:991–9.
21. Boy SC, van Heerden MB, Babb C, van Heerden WF, Willem P. Domi-nant genetic aberrations and coexistent EBV infection in HIV-related oral plasmablastic lymphomas. Oral Oncol 2011;47:883–7.
22. Maya Miles D, Penate X, Sanmartin Olmo T, Jourquin F, Munoz Centeno MC, Mendoza M, et al. High levels of histones promote whole-genome-duplications and trigger a Swe1(WEE1)-dependent phosphorylation of Cdc28(CDK1). Elife 2018;7:e35337.
23. Wenzel S, Grau M, Mavis C, Hailfinger S, Wolf A, Madle H, et al. MCL1 is deregulated in subgroups of diffuse large B-cell lymphoma. Leukemia 2013;27:1381–90.
24. Chen C, Zhao S, Karnad A, Freeman JW. The biology and role of CD44 in cancer progression: therapeutic implications. J Hematol Oncol 2018;11:64.
25. Landau DA, Tausch E, Taylor-Weiner AN, Stewart C, Reiter JG, Bahlo J, et al. Mutations driving CLL and their evolution in progression and relapse. Nature 2015;526:525.
26. Schmitz R, Wright GW, Huang DW, Johnson CA, Phelan JD, Wang JQ, et al. Genetics and pathogenesis of diffuse large B-cell lymphoma. N Engl J Med 2018;378:1396–407.
Genomic Characterization of Plasmablastic Lymphoma RESEARCH ARTICLE
JULY 2020�BLOOD CANCER DISCOVERY | 125
27. Lohr JG, Stojanov P, Carter SL, Cruz-Gordillo P, Lawrence MS, Auclair D, et al. Widespread genetic heterogeneity in multiple myeloma: implications for targeted therapy. Cancer Cell 2014;25: 91–101.
28. Pasqualucci L, Dalla-Favera R. Genetics of diffuse large B-cell lym-phoma. Blood 2018;131:2307–19.
29. Brescia P, Schneider C, Holmes AB, Shen Q, Hussein S, Pasqualucci L, et al. MEF2B instructs germinal center development and acts as an oncogene in B cell lymphomagenesis. Cancer Cell 2018;34:453–65.
30. Ramsay AJ, Martinez-Trillos A, Jares P, Rodriguez D, Kwarciak A, Quesada V. Next-generation sequencing reveals the secrets of the chronic lymphocytic leukemia genome. Clin Transl Oncol 2013;15: 3–8.
31. Chapman MA, Lawrence MS, Keats JJ, Cibulskis K, Sougnez C, Schinzel AC, et al. Initial genome sequencing and analysis of multiple myeloma. Nature 2011;471:467–72.
32. Gravelle P, Pericart S, Tosolini M, Fabiani B, Coppo P, Amara N, et al. EBV infection determines the immune hallmarks of plasmablastic lymphoma. Oncoimmunology 2018;7:e1486950.
33. Shannon-Lowe C, Rickinson A. The global landscape of EBV-associated tumours. Front Oncol 2019;9:713.
34. Bencun M, Klinke O, Hotz-Wagenblatt A, Klaus S, Tsai MH, Poirey R, et al. Translational profiling of B cells infected with the Epstein-Barr virus reveals 5′ leader ribosome recruitment through upstream open reading frames. Nucleic Acids Res 2018;46:2802–19.
35. Hammerschmidt W, Sugden B. Replication of Epstein–Barr viral DNA. Cold Spring Harb Perspect Biol 2013;5:a013029.
36. Sugimoto A, Sato Y, Kanda T, Murata T, Narita Y, Kawashima D, et al. Different distributions of Epstein-Barr virus early and late gene transcripts within viral replication compartments. J Virol 2013;87: 6693–9.
37. Yu H, Lee H, Herrmann A, Buettner R, Jove R. Revisiting STAT3 sig-nalling in cancer: new and unexpected biological functions. Nat Rev Cancer 2014;14:736–46.
38. Garcia-Reyero J, Martinez Magunacelaya N, Gonzalez de Villambrosia S, Loghavi S, Gomez Mediavilla A, Tonda R, et al. Genetic lesions in MYC and STAT3 drive oncogenic transcription factor overexpression in plasmablastic lymphoma. Haematologica 2020 Apr 9 [Epub ahead of print].
39. Cousins E, Gao Y, Sandford G, Nicholas J. Human herpesvirus 8 viral interleukin-6 signaling through gp130 promotes virus replica-tion in primary effusion lymphoma and endothelial cells. J Virol 2014;88:12167–72.
40. Chiarle R, Simmons WJ, Cai H, Dhall G, Zamo A, Raz R, et al. Stat3 is required for ALK-mediated lymphomagenesis and provides a possible therapeutic target. Nat Med 2005;11:623–9.
41. Valera A, Colomo L, Martinez A, de Jong D, Balague O, Matheu G, et al. ALK-positive large B-cell lymphomas express a terminal B-cell differentiation program and activated STAT3 but lack MYC rear-rangements. Mod Pathol 2013;26:1329–37.
42. Tiacci E, Ladewig E, Schiavoni G, Penson A, Fortini E, Pettirossi V, et al. Pervasive mutations of JAK-STAT pathway genes in classical Hodgkin lymphoma. Blood 2018;131:2454–65.
43. Steidl C, Gascoyne RD. The molecular pathogenesis of primary medi-astinal large B-cell lymphoma. Blood 2011;118:2659–69.
44. Joos S, Kupper M, Ohl S, von Bonin F, Mechtersheimer G, Bentz M, et al. Genomic imbalances including amplification of the tyrosine kinase gene JAK2 in CD30+ Hodgkin cells. Cancer Res 2000;60:549–52.
45. Ritz O, Guiter C, Castellano F, Dorsch K, Melzner J, Jais JP, et al. Recurrent mutations of the STAT6 DNA binding domain in primary mediastinal B-cell lymphoma. Blood 2009;114:1236–42.
46. Johnson DE, O’Keefe RA, Grandis JR. Targeting the IL-6/JAK/STAT3 signalling axis in cancer. Nat Rev Clin Oncol 2018;15:234.
47. Redondo-Muñoz J, García-Pardo A, Teixidó J. Molecular players in hematologic tumor cell trafficking. Front Immunol 2019;10:156.
48. Tzankov A, Pehrs AC, Zimpfer A, Ascani S, Lugli A, Pileri S, et al. Prognostic significance of CD44 expression in diffuse large B cell lymphoma of activated and germinal centre B cell-like types: a tissue microarray analysis of 90 cases. J Clin Pathol 2003;56:747–52.
49. Zhong Y, Meng F, Zhang W, Li B, van Hest JC, Zhong Z. CD44-targeted vesicles encapsulating granzyme B as artificial killer cells for potent inhibition of human multiple myeloma in mice. J Control Release 2020;320:421–30.
50. Grande BM, Gerhard DS, Jiang A, Griner NB, Abramson JS, Alexander TB, et al. Genome-wide discovery of somatic coding and noncoding mutations in pediatric endemic and sporadic Burkitt lymphoma. Blood 2019;133:1313–24.
51. Koganti S, Clark C, Zhi J, Li X, Chen EI, Chakrabortty S, et al. Cellular STAT3 functions via PCBP2 to restrain Epstein-Barr virus lytic activa-tion in B lymphocytes. J Virol 2015;89:5002–11.
52. Meer S PY, McAlpine ED, Willem P. Extraoral plasmablastic lympho-mas in a high human immunodeficiency virus endemic area. Histopa-thology 2019;76:212–21.
53. Villela D, Costa SS, Vianna-Morgante AM, Krepischi AC, Rosenberg C. Efficient detection of chromosome imbalances and single nucleo-tide variants using targeted sequencing in the clinical setting. Eur J Med Genet 2017;60:667–74.
54. Ostrup O, Ahlborn LB, Lassen U, Mau-Sorensen M, Nielsen FC. Detection of copy number alterations in cell-free tumor DNA from plasma. BBA Clin 2017;7:120–6.
55. Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, et al. Full-length transcriptome assembly from RNA-Seq data with-out a reference genome. Nat Biotechnol 2011;29:644.
56. Arvey A, Tempera I, Tsai K, Chen HS, Tikhmyanova N, Klichinsky M, et al. An atlas of the Epstein-Barr virus transcriptome and epig-enome reveals host-virus regulatory interactions. Cell Host Microbe 2012;12:233–45.
57. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinfor-matics 2013;30:923–30.
SEPTEMBER 2020�BLOOD CANCER DISCOVERY | 1
Gene Expression Profi ling of Mediastinal Gray Zone Lymphoma and Its Relationship to Primary Mediastinal B-cell Lymphoma and Classical Hodgkin Lymphoma Stefania Pittaluga 1 , Alina Nicolae 1 , George W. Wright 2 , Christopher Melani 3 , Mark Roschewski 3 , Seth Steinberg 4 , DaWei Huang 3 , Louis M. Staudt 3 , Elaine S. Jaffe 1 , and Wyndham H. Wilson 3
RESEARCH BRIEF
ABSTRACT Mediastinal gray zone lymphoma (MGZL) has immunopathologic features between classical Hodgkin lymphoma (cHL) and primary mediastinal thymic B-cell lymphoma
(PMBL), leading to uncertainty regarding its biological relationship to these entities. We performed gene expression profi ling from patients with MGZL (20), cHL (18), and PMBL (17) and show MGZL clus-ters between cHL and PMBL. Expression signatures reveal germinal B-cell and IFN regulatory factor 4 (IRF4) signatures were relatively low in MGZL and cHL compared with PMBL, indicating downregula-tion of the B-cell program in MGZL, a hallmark of cHL. T-cell and macrophage signatures were higher in MGZL and cHL compared with PMBL, consistent with infi ltrating immune cells, which are found in cHL. The NF κ B signature was higher in MGZL than PMBL, and like cHL, MGZL and PMBL express NF κ B inducing kinase (NIK), indicating noncanonical signaling. These fi ndings indicate that while MGZL has distinctive clustering, it is biologically closer to cHL.
SIGNIFICANCE: We performed comparative gene expression analysis of MGZL, cHL, and PMBL and show most MGZL cases are biologically closer to cHL. MGZL has signifi cantly higher tumor cell density than cHL and greater NF κ B activation compared with PMBL, which may explain its greater treatment resistance compared with cHL and PMBL.
1 Laboratory of Pathology, NCI, Bethesda, Maryland. 2 Biometrics Research Program, NCI, Bethesda, Maryland. 3 Lymphoid Malignancies Branch, NCI, Bethesda, Maryland. 4 Center for Cancer Research, NCI, Bethesda, Maryland. S. Pittaluga and W.H. Wilson contributed equally as the co-senior authors of this article. Corresponding Author: Wyndham H. Wilson, National Cancer Institute, 10 Center Dr, Bethesda, MD 20892. Phone: 301-312-5484; E-mail: [email protected]
Blood Cancer Discov 2020;1:1–7 doi: 10.1158/2643-3230.BCD-20-0009 ©2020 American Association for Cancer Research.
INTRODUCTION
Mediastinal gray zone lymphoma (MGZL) is a pathologic entity with morphologic and/or IHC features intermediate
between primary mediastinal B-cell lymphoma (PMBL) and mediastinal classical Hodgkin lymphoma (cHL; refs. 1–4 ). Historically, these cases were sometimes termed “Hodgkin-like anaplastic large cell lymphoma”, but a connection to what
Pittaluga et al.RESEARCH BRIEF
2 | BLOOD CANCER DISCOVERY�SEPTEMBER 2020 AACRJournals.org
Table 1. Clinical and pathologic characteristics
Characteristics MGZL cHL PMBLTotal patients 20 18 17
Male gender 15 (75%) 8 (44%) 10 (59%)
Median age (years) 34 28 33¡Age range 17–77 19–72 17–52
Mediastinal tumor present 16/17 a (94%) 18 (100%) 17 (100%)
Pretreatment biopsy 18/18 b (100%) 15/18 (83%) 17/17 (100%)MGZL Tumor morphology¡PMBL-Like 5/20 (25%) NA NA¡cHL-Like 5/20 (25%) NA NA¡Composite PMBL/cHL-like 1/20 (5%) NA NA¡Intermediate 9/20 (45%) NA NA
Immunophenotype¡CD20 19/20 (95%) 3/16 (19%) 17 (100%)¡CD30 20/20 (100%) 18/18 (100%) 12/14 (85%)¡CD15 16/20 (80%) 17/18 (95%) 1/11 (9%)¡EBER in situ hybridization 0/11 (0%) NA NA
Percent tumor cell quartiles¡<25 3/17 (18%) 8/18 (44%) 3/17 (18%)¡25–50 4/17 (23.5%) 7/18 (39%) 3/17 (18%)¡50–75 6/17 (35%) 2/18 (11%) 2/17 (12%)¡75–100 4/17 (23.5%) 1/18 (6%) 9/17 (53%)
Abbreviation: NA, not applicable. a Three cases unknown. b Two cases unknown.
we recognize today as anaplastic large-cell lymphoma was not confi rmed ( 5, 6 ). As a group, PMBL and cHL are hypothesized to derive from a thymic B cell and share molecular features including gains as well as amplifi cation of the REL/BCL11A locus (2p16.1), JAK2/PDL2 locus (9p24.1), and CIITA locus (16p13.13), which also occur in MGZL ( 7–10 ). PMBL and cHL also have overlapping gene expression profi les, suggest-ing they lie along a pathobiological continuum with MGZL described as the “missing link” between these entities ( 2, 8 ). Methylation profi ling of MGZL showed a unique signature, but also shared components with cHL and PMBL, further supporting its association with these entities ( 9 ). MGZL also has overlapping clinical features with PMBL and cHL, raising the question of where it lies within the biological and clini-cal spectrum of mediastinal B-cell lymphomas. To address these questions, we performed gene expression profi ling of MGZL, cHL, and PMBL and compared expression signatures of immune cells, NF κ B survival pathway, and B-cell differen-tiation.
RESULTS Patient Characteristics and Pathobiology
Fifty-fi ve tumor samples comprising 20 MGZL, 18 cHL, and 17 PMBL were analyzed ( Table 1 ). Of these biopsies, 50 were pretreatment, 2 MGZL had unknown treatment status, and 3 cHL were at relapse. The median (range) patient age of the
respective groups was 34 (17–77), 28 (19–72) and 33 (17–52) years, and 15 (75%), 8 (44%), and 10 (59%) were male. Patients with cHL and PMBL had features consistent with these diagno-ses, whereas patients with MGZL had histologic and/or pheno-typic features intermediate between cHL and PMBL according to the World Health Organization Classifi cation ( 11 ).
The diagnosis of MGZL is dependent on a discordance between morphology and phenotype, which is inconsistent with classifi cation as either cHL or PMBL. In general, MGZL is relatively rich in tumor cells compared with cHL, and although Hodgkin Reed-Sternberg (HRS) cells may be present, they lack a typical infl ammatory background and phenotypically have a stronger B-cell phenotype with expression of CD20 and CD79 compared with cHL. MGZL tumors may have a predominant morphology, which allows further categorization of these tumors as PMBL-like, cHL-like, or rarely composite with two separate components; however, cases with an intermediate morphology are more typically seen. MGZL cHL-like tumors show Reed-Sternberg cells in an infl ammatory background with relatively robust CD20 staining ( Fig. 1A ), whereas cases predominantly composed of large cells as seen in PMBL may lack or only partially express B-cell markers such as CD20, but often strongly express CD30 and CD15 ( Fig. 1B ). In the current series of 20 MGZL cases, the categories of cHL-like, PMBL-like, and composite were present in 25%, 25%, and 5%, respectively, with 45% of cases showing an intermediate mor-phology ( Table 1 ). Almost all MGZL cases showed CD20 and
Gene Expression of Mediastinal Gray Zone Lymphomas RESEARCH BRIEF
SEPTEMBER 2020�BLOOD CANCER DISCOVERY | 3
CD15 expression and all were positive for CD30 (Table 1). We also assessed EBV encoded RNA in situ (EBER) in 11 MGZL cases and found no positive cases.
We determined the percent distribution of tumor cells in our MGZL, cHL, and PMBL cohorts as a control for tumor cell density (Table 1). As expected, PMBL had signifi-cantly more tumor cells compared with cHL with the median quartile being 75%–100% and 25%–50% cells in each group, respectively (P = 0.003). MGZL also showed significantly higher tumor cells density compared with cHL with a median quartile of 50% to 75% and 25% to 50% cells, respectively (P = 0.015). However, there was no significant difference in tumor cell density between MGZL and PMBL (P = 0.455).
Gene ExpressionGene expression profiling was performed on all 55 biopsy
samples from patients with MGZL, cHL, and PMBL. On the basis of the hypothesis that MGZL is biologically intermedi-ate between cHL and PMBL, we first analyzed expression differences between cHL and PMBL and identified 103 genes that were significant at P < 10−6 (FDR = 0.0001). The result-ant clustering generally placed the MGZL cases between core clusters for cHL and PMBL (Fig. 2). The cHL core, termed cluster group 1, includes all cHL samples and one MGZL, and the PMBL core, termed cluster group 4, includes all but one PMBL and two MGZL cases. The MGZL cases broke down into two cluster groups. One abutted the cHL core, termed cluster group 2, and includes most MGZL samples and one outlier PMBL sample that had an extremely low tumor con-tent, and one abutted the PMBL core, termed cluster group 3, and includes 4 MGZL samples.
This division represents the most stable grouping of sam-ples and recognizes two MGZL clusters distinguished by proximity to cHL or PMBL cases. On the basis of the patho-logic recognition of MGZL as cHL-like, PMBL-like, or a composite of the two, we looked at their distribution in the array (Fig. 2). Among 5 cHL-like cases, one was at the edge of cluster group 1 (cHL core), 3 were in cluster group 2 (cHL abutting), and one was in cluster group 3 (PMBL abutting). Of the five cases with PMBL-like features, 3 were in cluster group 3 (PMBL abutting), one was in cluster group 4 (PMBL core), and one in cluster group 2 (cHL abutting). The single case that showed composite features clustered within cluster group 4 (PMBL core), raising the suggestion that the PMBL-like component was overrepresented in the biopsy specimen. The majority of cases with intermediate features were distrib-uted in cluster groups 2 and 3.
The gene expression signature revealed 43 genes more highly expressed in cHL compared with PMBL (Fig. 2). We explored signatures of cellular differentiation and found that like cHL, MGZL had lower expression of germinal center B-cell genes (GCB) and IFN regulatory factor 4 (IRF4) genes, indicating these tumors down regulate B-cell differentia-tion relative to PMBL (Fig. 3A and B). Genes associated with infiltrating T cells, including IL6ST, CTLA4, CD28, and ICOS, and immune regulation, including IL1R2, IL32, IL7R, and TNIP3, were of interest because of their association with the inflammatory background found in cHL (11). MGZL tumors showed variable expression of these genes with cluster group 2 abutting cHL showing relatively high levels and cluster group 3 abutting PMBL showing low expression. We also looked at two immune-associated signatures defined as a
Figure 1. MGZL IHC. A, MGZL cHL-like tumors show Reed-Sternberg cells in an inflammatory background on hematoxylin and eosin (H&E) stain along with robust CD20 and low CD15- and CD30-positive cells. B, MGZL PMBL-like tumors show diffuse large B cells on H&E stain but with robust CD30- and CD15-positive cells and low CD20.
AH&E stain
MGZL cHL-like MGZL PMBL-like
CD20 stain
CD30 stain CD15 stain
H&E stain CD20 stain
CD30 stain CD15 stain
B
20 µm
20 µm 20 µm
Pittaluga et al.RESEARCH BRIEF
4 | BLOOD CANCER DISCOVERY�SEPTEMBER 2020 AACRJournals.org
HistologyClassical hodgkin lymphoma
1. cHL core
Mediastinal gray zone lymphoma
2. MGZL abut cHL
Primary mediastinal lymphoma
4. PMBL core
KIF16BCLUAP1CLUCCND2IL6STCOL29A1COL4A4ACVR1ADAMTS10ARHGEF10NFIAZMYND11IL1R2IL32GGT5VWA5AICOSLRRC8CIL7RGIMAP7GMFGAMICA1MAFMAN1C1NPDC1CD28CTLA4MAP3K5PTPN13SATB1TRIB2RASGRP2TNIP3SLC12A6S1PR1IL18R1SULT1B1TRPS1PRKCANFASCPLA2G4ASERPINB6TMEM59LOC100288765KCNMA1GNAZKANK1CSMAP4K2CD70DNAJC5BC17orf45ATP2A1AP1S3FANCAMGAT3DEPDC5GRHPRCHDHIL17RBCTPS2TMEM57LAT2BTKBCAR3P0U2F2P0U2AF1C17orf99LRCH1
3. MGZL abut PMBLCore groups
ADATLE4GCNT2KBTBD8CCDC138DENND4AMSH6MTA3ANKRD13ARFTN1SLC9A7DNAH8ELL3BLNKGGA2LRMPKIAA0922MS4A1PAX5RASGRP3PRPSAP2RAB3GAP2PLCG2SWAP70RGS13PRDM2WEE1P0U2F1ZCCHC7SPATS2PCDHGA6MYBL1SIPA1L1SNX11UVRAGWDR66KIAA1432PDCD1LG2NFX1UBAP2TP63PCDHGA10PCDHGB5MAST2MRPL39RFC3SNX22P0F1BSMS
Figure 2. Gene expression cluster-ing of 103 genes in tumor samples. MGZL clustered between cHL and PMBL. Four cluster groups were identified and included a core cHL group (cluster 1), a core PMBL group (cluster 4), and two MGZL groups divided by their abutment to the cHL (cluster 2) and PMBL (cluster 3) core groups. MGZL cluster groups 2 and 3 had gene expression closer to cHL and PMBL, respectively.
Gene Expression of Mediastinal Gray Zone Lymphomas RESEARCH BRIEF
SEPTEMBER 2020�BLOOD CANCER DISCOVERY | 5
thymic CD4 T-cell signature and a macrophage cell signature (Fig. 3C and D). In both instances, the median gene expres-sion of MGZL tumors was intermediate between cHL and PMBL, but in many cases closer to cHL. These results show gene expression associated with infiltrating immune cells in MGZL is distributed among cases abutting PMBL and cHL and include cases that are cHL-like and PMBL-like.
Finally, we explored the nuclear factor κB (NFκB) signaling pathway that regulates cell survival and is expressed in both cHL and PMBL. NFκB expression was highest in cHL with most MGZL cases showing greater expression than PMBL, including those in cluster group 3 that abutted the PMBL core (Fig. 3E). A recent study showed increased copies of the NFκB inducing kinase (NIK) activates the noncanonical NFκB pathway in cHL and accounts for much of the NFκB pathway activation in cHL (12). We also found increased NIK
expression by IHC in our MGZL (5/6 positive) and PMBL (5/6 positive) cases, indicating they share a similar mechanism of NFκB activation (Fig. 4).
DISCUSSIONThe variable immunopathologic features and distinguish-
ing clinical outcome of MGZL necessitates a greater biological understanding. Our cases show a distribution of tumor mor-phology that reflects the reported pathologic diversity, with a quarter each showing a PMBL-like and cHL-like morphology and the balance mostly showing an intermediate morphol-ogy. The variable morphology of MGZL is highlighted by a recent study of 139 cases where 62% and 38% of cases were identified as cHL-like and large B-cell lymphoma-like, respec-tively, with only two cases interpreted as having an intermediate
8.1
67
8
9
10
11
12
8
10
12
14
8.5 8.1
9.1
10.1
11.1
9.0
9.5
10.0
10.5
11.0
9.1
10.1
11.1
GCB signature IRF4 signature
cHL MGZL PMBL
cHL MGZL PMBL cHL MGZL PMBL
cHL MGZL PMBL cHL MGZL
Legend
cHL coreMGZL abut cHLMGZL abut PMBLPMBL core
PMBL
A B
Macrophage signature NFκB signatureD E
C CD4 T-cell signature
Figure 3. MGZL, cHL, and PMBL gene expression signatures are shown for GCB cells (A), IRF4 (B), CD4 T cells (C), macrophages (D), and NFκB (E). The distribution of the cluster groups within each signature and cell type is color coded as shown in the legend.
cHL PMBL MGZL
2 µm2 µm 2 µm
Figure 4. Representative IHC of NIK staining in cHL, PMBL, and MGZL. Analysis of NIK staining in MGZL and PMBL showed 5 of 6 positives in both cohorts. Published analysis of NIK staining in cHL showed 30 of 31 positive cases (12).
Pittaluga et al.RESEARCH BRIEF
6 | BLOOD CANCER DISCOVERY�SEPTEMBER 2020 AACRJournals.org
morphology (7). This stands in contrast to our own series in which we found an intermediate morphology in 45% of cases, raising the likelihood that the discrepancy between the series reflects the difficulty in such assignments rather than a differ-ence in biology (13). Furthermore, MGZL rarely express EBV, which was reported in 24% of cases in the above series, and not found in any of 11 cases analyzed in our series. Indeed, the presence of EBV raises the likelihood that such cases are EBV+ diffuse large B-cell lymphomas, where Hodgkin-like fea-tures are often present, and we believe they should be largely excluded from series of MGZL (11).
The frequency of CD20 and CD15 expression in 95% and 80% of our MGZL cases, respectively, points to a shared biol-ogy with PMBL and cHL. Gene expression clustering of our cases showed two stable groupings with most MGZL cases (13) abutting the cHL core and a few cases (4) abutting the PMBL core cases. Although limited in numbers, all but one MGZL case with PMBL-like features either abutted or clustered in the PMBL core, and all but one cHL-like case abutted or clustered in the cHL-core, suggesting an association between gene expres-sion and morphology and phenotype. Of course, most MGZL cases were interpreted as intermediate and abutted the cHL core. Overall, these findings suggest that most MGZL cases have gene expression features more closely aligned with cHL than with PMBL. Notably, this gene expression distribution does not appear to be driven by tumor cell number per se based on our finding that the MGZL and PMBL cohorts have similar tumor cell densities, whereas the cHL cohort is significantly lower.
Analysis of gene expression signatures provides a more gran-ular view of MGZL. Downregulation of the B-cell program as reflected in the GCB signature, which is a feature of cHL, was prominent in most MGZL cases with the cases abutting the PMBL core showing the least down regulation. The IRF4 signature, which is associated with B-cell differentiation and germinal center formation, was also lower in MGZL compared with PMBL and more similar to cHL. Most MGZL cases exhibit somewhat higher expression of NFκB genes compared with PMBL, and like cHL, this appears to be driven by the noncanon-ical pathway (12). Features of the microenvironment as deter-mined by the T-cell and macrophage signatures were increased in MGZL, indicating another biological similarity to cHL (4).
Because of the rarity of MGZL and its biological diver-sity, our results are limited by relatively low sample sizes. Nonetheless, the totality of data confirms the diversity of these tumors, but also indicates that most cases are biologi-cally closer to cHL than to PMBL. While this suggests these cases should be treated as cHL, historical data indicates such treatment produces poorer outcomes compared with those observed in cHL (5, 6). The alternative therapeutic strategy to treat these patients with regimens designed for PMBL has also been tested and shows an inferior outcome compared with PMBL, albeit still with a significant rate of cure (4, 14, 15). These data suggest that the poorer outcome of MGZL, with regimens designed for cHL or PMBL, likely reflects the presence of a higher degree of chemotherapy resistance as opposed to an issue with the therapeutic approach per se, possibly related to increased activation of NFκB relative to PMBL and higher tumor cell density relative to cHL. Thus, the totality of data suggests that most MGZL are biologically closer to cHL, but harbor biological features of PMBL and
future therapy should integrate active components used in both diseases including anti-CD20 antibodies and immune checkpoint inhibitors.
METHODSTissue Specimens
Deidentified tissue samples obtained from the Laboratory of Pathol-ogy, Center for Cancer Research, National Cancer Institute included MGZL (20 cases), cHL (18 cases), and PMBL (17 cases). Eight MGZL cases in the current series were included in a previously published study ClinicalTrials.gov (NCT00001337). The study samples were from a protocol approved by the Institutional Review Board, NCI or had approved waiver of written informed consent. Histologic review was conducted by the authors (S. Pittaluga, A. Nicolae, and E.S. Jaffe) in accordance with the World Health Organization Classification.
IHC AnalysisIHC was performed on formalin-fixed, paraffin-embedded (FFPE)
tissue sections and included CD20, CD3, CD15, CD30, CD79, PAX-5, and OCT-2, and performed as described previously. Positive controls were run with each set of slides and showed appropriate staining pat-terns. The percentage of tumor cells was assessed by morphology and CD20, PAX5, or CD30 depending on immunoreactivity of the tumor cells. CD30 and CD15 immunoreactivity were not quantitated but scored as positive if there was any staining on the malignant cells. IHC for NIK and EBV-encoded RNA (EBER) in situ hybridization was performed on a subset of cases. All slides were independently reviewed and the scores agreed upon by joint rereview (S. Pittaluga, A. Nicolae, and E.S. Jaffe).
Gene Expression AnalysisThe methods used to extract, sequence, and analyze RNA from
FFPE samples to generate digital gene expression values were fol-lowed as described previously (16). An RNA library employing TruSeq (Illumina Prep Kit v2) was prepared and genes with zero signal on more than 20% of the samples were excluded from analysis.
Statistical AnalysisTwo-sided t tests were used to identify genes that were differen-
tially expressed between histologically defined cHL and PMBL sam-ples. Genes were centered such that zero represented the midpoint between the cHL and PMBL averages. Hierarchical clustering was performed using centroid linkage with an uncentered correlation distance metric. Summarized gene expression measure for signa-tures were generated as follows: For each signature and each sample calculated an initial average of all genes that had 10% or fewer zero values. We then refined this list by excluding those genes that were not correlated (Pearson r > 0.25) with this signature average across all samples. The average of the remaining correlated genes for each sample was then reported as the signature gene expression for that sample. P values for signature gene expression were calculated on the basis of an F-test. An exact Cochran–Armitage test for trend was used to test the difference in the ordered categorical percent tumor cells in three sets of two pathologic groups each; P values are two-tailed.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors’ ContributionsS. Pittaluga: Conceptualization, resources, data curation, for-
mal analysis, supervision, funding acquisition, validation, inves-tigation, visualization, methodology, project administration,
Gene Expression of Mediastinal Gray Zone Lymphomas RESEARCH BRIEF
SEPTEMBER 2020�BLOOD CANCER DISCOVERY | 7
writing−review, and editing. A. Nicolae: Resources, validation, inves-tigation, visualization, writing−review, and editing. G.W. Wright: Conceptualization, data curation, software, formal analysis, vali-dation, investigation, methodology, writing−review, and editing. C. Melani: Resources, investigation, writing−review, and editing. M. Rochewski: Resources, investigation, writing−review, and edit-ing. S. Steinberg: Data curation, software, formal analysis, writing−review, and editing. D. Huang: Resources, data curation, software, formal analysis, validation, methodology, writing−review, and edit-ing. L.M. Staudt: Conceptualization, resources, data curation, for-mal analysis, supervision, validation, investigation, methodology, writing−review, and editing. E.S. Jaffe: Conceptualization, resources, formal analysis, investigation, methodology, writing−review, and editing. W.H. Wilson: Conceptualization, resources, data curation, formal analysis, supervision, validation, investigation, methodology, writing−original draft, project administration, writing−review, and editing.
AcknowledgmentsThis work was supported by grants from the Center for Cancer
Research, NCI.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received January 28, 2020; revised February 27, 2020; accepted February 28, 2020; published first March 30, 2020.
REFERENCES1. World Health Organisation. WHO Classification of tumours of hae-
matopoietic and lymphoid tissues (revised 4th edition). Lyon, France: International Agency for Research on Cancer; 2017.
2. Traverse-Glehen A, Pittaluga S, Gaulard P, Sorbara L, Alonso MA, Raffeld M, et al. Mediastinal gray zone lymphoma: the missing link between classic Hodgkin’s lymphoma and mediastinal large B-cell lymphoma. Am J Surg Pathol 2005;29:1411–21.
3. Quintanilla-Martinez L, de Jong D, de Mascarel A, Hsi ED, Kluin P, Natkunam Y, et al. Gray zones around diffuse large B cell lym-phoma. Conclusions based on the workshop of the XIV meeting of the European Association for Hematopathology and the Soci-
ety of Hematopathology in Bordeaux, France. J Hematop 2009;2:211–36.
4. Wilson WH, Pittaluga S, Nicolae A, Camphausen K, Shovlin M, Steinberg SM, et al. A prospective study of mediastinal gray-zone lymphoma. Blood 2014;124:1563–9.
5. Pileri S, Bocchia M, Baroni CD, Martelli M, Falini B, Sabattini E, et al. Anaplastic large cell lymphoma (CD30+/Ki-1+): results of a prospective clinico-pathological study of 69 cases. Br J Haematol 1994;86:513–23.
6. Zinzani PL, Martelli M, Magagnoli M, Zaccaria A, Ronconi F, Cantonetti M, et al. Anaplastic large cell lymphoma Hodgkin’s-like: a randomized trial of ABVD versus MACOP-B with and without radia-tion therapy. Blood 1998;92:790–4.
7. Sarkozy C, Copie-Bergman C, Damotte D, Ben-Neriah S, Burroni B, Cornillon J, et al. Gray-zone lymphoma between cHL and large B-cell lymphoma: a histopathologic series from the LYSA. Am J Surg Pathol 2019;43:341–51.
8. Rosenwald A, Wright G, Leroy K, Yu X, Gaulard P, Gascoyne RD, et al. Molecular diagnosis of primary mediastinal B cell lymphoma identifies a clinically favorable subgroup of diffuse large B cell lym-phoma related to Hodgkin lymphoma. J Exp Med 2003;198:851–62.
9. Eberle FC, Rodriguez-Canales J, Wei L, Hanson JC, Killian JK, Sun HW, et al. Methylation profiling of mediastinal gray zone lymphoma reveals a distinctive signature with elements shared by classical Hodgkin’s lymphoma and primary mediastinal large B-cell lymphoma. Haemato-logica 2011;96:558–66.
10. Eberle FC, Salaverria I, Steidl C, Summers TA Jr, Pittaluga S, Neriah SB, et al. Gray zone lymphoma: chromosomal aberrations with immu-nophenotypic and clinical correlations. Mod Pathol 2011;24:1586–97.
11. Swerdlow SH, Campo E, Pileri SA, Harris NL, Stein H, Siebert R, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood 2016;127:2375–90.
12. Ranuncolo SM, Pittaluga S, Evbuomwan MO, Jaffe ES, Lewis BA. Hodgkin lymphoma requires stabilized NIK and constitutive RelB expression for survival. Blood 2012;120:3756–63.
13. Parker K, Venkataraman G. Challenges in the diagnosis of gray zone lymphomas. Surg Pathol Clin 2019;12:709–18.
14. Dunleavy K, Grant C, Eberle FC, Pittaluga S, Jaffe ES, Wilson WH. Gray zone lymphoma: better treated like Hodgkin lymphoma or medi-astinal large B-cell lymphoma? Curr Hematol Malig Rep 2012;7:241–7.
15. Dunleavy K, Pittaluga S, Maeda LS, Advani R, Chen CC, Hessler J, et al. Dose-adjusted EPOCH-rituximab therapy in primary mediastinal B-cell lymphoma. N Engl J Med 2013;368:1408–16.
16. Schmitz R, Wright GW, Huang DW, Johnson CA, Phelan JD, Wang JQ, et al. Genetics and pathogenesis of diffuse large B-cell lymphoma. N Engl J Med 2018;378:1396–407.
RESEARCH ARTICLE
Single-Cell Transcriptome Analysis Reveals Disease-Defining T-cell Subsets in the Tumor Microenvironment of Classic Hodgkin Lymphoma Tomohiro Aoki1,2, Lauren C. Chong1, Katsuyoshi Takata1, Katy Milne3,4, Monirath Hav5, Anthony Colombo5, Elizabeth A. Chavez1, Michael Nissen6, Xuehai Wang6, Tomoko Miyata-Takata1, Vivian Lam6, Elena Viganò1,2, Bruce W. Woolcock1, Adèle Telenius1, Michael Y. Li1,2, Shannon Healy1, Chanel Ghesquiere3,4, Daniel Kos3,4, Talia Goodyear3,4, Johanna Veldman7, Allen W. Zhang8,9, Jubin Kim6, Saeed Saberi8, Jiarui Ding8,10, Pedro Farinha1, Andrew P. Weng6, Kerry J. Savage1, David W. Scott1, Gerald Krystal6, Brad H. Nelson3,4,11, Anja Mottok1,12, Akil Merchant5, Sohrab P. Shah2,8,9, and Christian Steidl1,2
MARCH 2020�CANCER DISCOVERY | 407
ABSTRACT Hodgkin lymphoma is characterized by an extensively dominant tumor microenviron-ment (TME) composed of different types of noncancerous immune cells with rare
malignant cells. Characterization of the cellular components and their spatial relationship is crucial to understanding cross-talk and therapeutic targeting in the TME. We performed single-cell RNA sequenc-ing of more than 127,000 cells from 22 Hodgkin lymphoma tissue specimens and 5 reactive lymph nodes, profi ling for the fi rst time the phenotype of the Hodgkin lymphoma–specifi c immune microenvironment at single-cell resolution. Single-cell expression profi ling identifi ed a novel Hodgkin lymphoma–associ-ated subset of T cells with prominent expression of the inhibitory receptor LAG3, and functional analy-ses established this LAG3 + T-cell population as a mediator of immunosuppression. Multiplexed spatial assessment of immune cells in the microenvironment also revealed increased LAG3 + T cells in the direct vicinity of MHC class II–defi cient tumor cells. Our fi ndings provide novel insights into TME biology and suggest new approaches to immune-checkpoint targeting in Hodgkin lymphoma.
SIGNIFICANCE: We provide detailed functional and spatial characteristics of immune cells in classic Hodgkin lymphoma at single-cell resolution. Specifi cally, we identifi ed a regulatory T-cell–like immuno-suppressive subset of LAG3 + T cells contributing to the immune-escape phenotype. Our insights aid in the development of novel biomarkers and combination treatment strategies targeting immune checkpoints.
See related commentary by Fisher and Oh, p. 342.
1 Centre for Lymphoid Cancer, British Columbia Cancer, Vancouver, British Columbia, Canada. 2 Department of Pathology and Laboratory Medicine, Uni-versity of British Columbia, Vancouver, British Columbia, Canada. 3 Deeley Research Centre, British Columbia Cancer, Vancouver, British Columbia, Canada. 4 Department of Biochemistry and Microbiology, University of Vic-toria, Victoria, British Columbia, Canada. 5 Cedars-Sinai Medical Center, Los Angeles, California. 6 Terry Fox Laboratory, British Columbia Cancer, Vancouver, British Columbia, Canada. 7 Department of Pathology and Medi-cal Biology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands. 8 Department of Molecular Oncology, British Columbia Cancer, Vancouver, British Columbia, Canada. 9 Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York. 10 Broad Institute of MIT and Harvard, Cambridge, Mas-sachusetts. 11 Department of Medical Genetics, University of British Colum-bia, Vancouver, British Columbia, Canada. 12 Institute of Human Genetics, Ulm University and Ulm University Medical Center, Ulm, Germany. Note: Supplementary data for this article are available at Cancer Discovery Online (http://cancerdiscovery.aacrjournals.org/). T. Aoki and L.C. Chong contributed equally to this work. Corresponding Author: Christian Steidl, British Columbia Cancer, 675 West 10th Avenue, Room 12-110, Vancouver, BC V5Z 1L3, Canada. Phone: 604-675-8046; Fax: 604-675-8183; E-mail: [email protected] Cancer Discov 2020;10:406–21 doi: 10.1158/2159-8290.CD-19-0680 ©2019 American Association for Cancer Research.
INTRODUCTION Classic Hodgkin lymphoma (cHL) is the most common lym-
phoma subtype among adolescents and young adults ( 1 ). cHL is characterized by an extensive microenvironment composed of different types of noncancerous normal immune cells, such as several types of T cells, B cells, eosinophils and macrophages, and a rare population (∼1%) of clonal malignant Hodgkin and Reed–Sternberg (HRS) cells ( 1–3 ). Although some fi ndings support the concept that the HRS cells recruit these immune cells to form a tumor-supporting, regulatory tumor microen-vironment (TME) with limited antitumor activity in cHL ( 4–6 ),
the complex interactions between HRS cells and their TME remain only partially understood. A deeper understanding of this symbiotic cellular cross-talk (“ecosystem”) may lead to the development of novel biomarkers and therapeutic approaches.
Immune-checkpoint inhibitors, such as the programmed death 1 (PD-1) inhibitors nivolumab and pembrolizumab, have shown dramatic effi cacy in relapsed or refractory cHL, with an overall response rate of 65% to 87% ( 7, 8 ) and durable remissions of approximately 1.5 years ( 8 ), which compares very favorably with other agents in this setting ( 9 ). Although the emergence of novel drugs emphasizes the need for the identifi cation of predictive biomarkers that can provide a rationale for treatment selection, it remains unclear which cells are the most important targets of immune-checkpoint inhibitors and which components are most relevant for the immune-escape phenotype in cHL. Thus, further comprehen-sive investigations of this interaction are needed.
Previous studies have applied IHC, microarray, cytometry by time-of-fl ight, and NanoString assays to characterize the immune phenotype of the TME in cHL and have identifi ed some important associations between the presence of certain immune cell types and clinical outcome ( 4, 6, 10 ). Although previous reports have described enrichment of CD4 + T cells in the TME of cHL ( 10–12 ), their study scale has been limited, and detailed coexpression patterns of important markers such as inhibitory receptors have not been examined.
Recently, the landscape of tumor-infi ltrating T cells has been assessed using single-cell transcriptome sequencing in several solid tumors, mostly of epithelial origin ( 13, 14 ). These single-cell RNA sequencing (scRNA-seq) studies have revealed diverse immune phenotypes, such as cells exhibiting an exhaustion signature, as well as clonal expansion patterns of T-cell lineages ( 14 ). However, such analyses are currently lacking in lympho-mas, which differ from most solid cancers in that they are clon-ally derived from lymphocytes that professionally interact with other immune cells in the ecosystem of the microenvironment.
Aoki et al.RESEARCH ARTICLE
408 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
In this study, we performed high-dimensional and spatial profiling of immune cells in cHL using scRNA-seq of 127,786 cells, multicolor IHC, and imaging mass cytometry (IMC). We identified a unique regulatory T cell–like subset that expressed lymphocyte activation gene 3 (LAG3+ T cells) in cHL and was mostly absent in normal reactive lymph nodes. LAG3+ T cells were characterized by expression of IL10 and TGFβ, and we demonstrated an immunosuppressive function of these cells. Further topologic analysis revealed that HRS cells were closely surrounded by frequent LAG3+ T cells in the subset of patients with cHL with loss of major histocompatibility class II (MHC-II) expression on tumor cells. Our data provide an unprec-edented number of single-cell transcriptomes in combination with multiplexed spatial assessment, allowing us to decipher the unique immune cell architecture of the TME in cHL with implications for novel therapies, including rational combina-tions and predictive biomarker development.
RESULTSThe cHL-Specific Immune Microenvironment at Single-Cell Resolution
To characterize the transcriptional profile of immune cells in the TME of cHL, we performed scRNA-seq on single-cell suspensions collected from lymph nodes of 22 patients with cHL, including 12 of the nodular sclerosis (NS) subtype, 9 of the mixed cellularity (MC) subtype, and 1 of the lymphocyte-rich subtype. We also sequenced reactive lymph nodes (RLN; n = 5) from healthy donors as normal controls (Supplementary Tables S1 and S2). Transcriptome data were obtained for a total of 127,786 sorted live cells, with a median of 1,203 genes detected per cell (Supplementary Table S3). To perform a systematic comparative analysis of the cHL TME and RLN, we merged the expression data from all cells (cHL and RLN) and performed batch correction and normalization. Removal of batch effects (caused by single-cell isolation and library preparation in dif-ferent experimental runs) resulted in improved mixing of cells across samples, as demonstrated by a significant increase in cell entropy (Wilcoxon–Mann–Whitney P < 0.001; Supplementary Fig. S1A and S1B).
Unsupervised clustering using PhenoGraph followed by visualization in t-SNE space (15, 16) identified 22 expression- based cell clusters that were annotated and assigned to a cell type based on the expression of genes described in published transcriptome data of sorted immune cells (17) and known canonical markers (Fig. 1A; Supplementary Figs. S2A–S2E and S3). These included four naïve T-cell clusters, two CD8+ T-cell clusters, six CD4+ T-cell clusters, seven B-cell clusters, one mac-rophage cluster, one plasmacytoid dendritic cell cluster, and one progenitor cell cluster. We could not observe an HRS cell cluster, which may be due to limitations of the microfluidics approach. Although most immune cell phenotypes exhibited overlap between cHL and RLN as demonstrated by clusters con-taining a mixture of cell types, we observed an enrichment of cells from cHL in some specific cell clusters (Fig. 1B). Of inter-est, we found that all three regulatory T-cell (Treg) clusters were quantitatively dominated by cells derived from the cHL samples with only a minor proportion originating from RLNs (Fig. 1C), and that the proportion of cells assigned to Treg clusters was significantly higher in cHL samples compared with RLN
(P = 0.0001; t test; Fig. 1D). The cluster containing the highest proportion of immune cells from cHL samples (CD4-C5-Treg) also exhibited relatively high expression of LAG3 and CTLA4 (Fig. 1A). Conversely, clusters enriched in RLN were mostly B-cell and CD8+ T-cell clusters (Fig. 1C). Further examination of the non-Treg CD4+ T-cell clusters revealed that they were primarily composed of type 2 T helper (Th2) cells, and that Th1 and type 17 T helper (Th17) cells were also enriched in cHL samples compared with RLN (Fig. 1E). We also performed dif-ferential expression analysis between cHL and RLN cells within each cluster, and identified IL32 as consistently upregulated in cHL T cells compared with RLN T cells (Supplementary Fig. S4). IL32 is a known proinflammatory cytokine that can induce the production of other cytokines such as IL6 (18).
EBV Status Affects the Immune Cell Subset Composition in cHL
Thirty to forty percent of cHLs are associated with latent Epstein–Barr virus (EBV) infection of the malignant HRS cells (19), and several reports indicate that EBV infection can recruit specific Treg populations to the TME in cHL (20, 21). To more precisely define immune cell composition according to EBV status, we compared the RNA-seq data of 5 EBV+ with 17 EBV− cases (Supplementary Fig. S5A). The proportion of CD4+
T cells with a Th17 profile was significantly decreased in EBV+
cHL (P = 0.004; t test; Fig. 1F and G). However, there was no sig-nificant difference between EBV+ and EBV− cases with respect to CD8+ T-cell or Treg proportions (Fig. 1F; Supplementary Fig. S5B). Similarly, the cHL MC subtype, which is more com-monly associated with EBV-related cHL, was associated with a lower proportion of Th17 polarized immune cells as compared with the NS subtype (Fig. 1H; Supplementary Fig. S5C).
Single-Cell Expression Patterns of Novel cHL-Specific Immune Subsets
Our data demonstrated the preferential enrichment of Tregs in cHL as compared with RLN (Fig. 1B and D). Considering the importance of an immunosuppressive microenvironment as a cancer hallmark, and its implications for biomarker develop-ment and targeted immunotherapy, we focused our analyses on the detailed characterization of Treg subsets. The most cHL-enriched Treg cluster, CD4-C5-Treg (Fig. 1A), was character-ized by high expression of LAG3, in addition to common Treg markers such as IL2RA (CD25) and TNFRSF18 (GITR; Fig. 2A). However, other canonical Treg markers such as FOXP3 were not coexpressed in this cluster, suggesting these cells may exhibit a type 1 regulatory (Tr1) T-cell phenotype (refs. 20, 22; Fig. 2B; Supplementary Fig. S6A). To confirm the expression pattern of immune cells in cHL, we also assessed the expression of surface and intracellular markers in all cHL cases using multicolor IHC and IMC. The orthogonal data confirmed the inversely corre-lated expression pattern of LAG3 and FOXP3 on CD4+ T cells at the protein level (Supplementary Fig. S6B–S6C).
Inhibitory receptor–mediated immune tolerance that can be hijacked by tumors has been a major target of cancer immuno-therapy (23, 24). To gain more insight into the characteristics of inhibitory receptor expression in the TME of cHL, we explored expression patterns among individual T cells. Although LAG3-expressing cells were mostly assigned to Treg clusters, PD-1–expressing cells were primarily assigned to non-Treg
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 409
Figure 1. Immune cell atlas of the Hodgkin lymphoma (HL) microenvironment at single-cell resolution. Cells from 22 cHL and 5 RLN cases were clustered using the PhenoGraph algorithm to identify groups of cells with similar expression patterns. A, Heat map summarizing mean expression (normalized and log-transformed) of selected canonical markers in each cluster. Data have been scaled row-wise for visualization. The covariate bar on the left side indicates the component associated with each gene, and black boxes highlight prominent expression of known subtype genes. DC, dendritic cell; NK, natural killer. B, Single-cell expression of all cells from cHL and RLN in t-SNE space (first two dimensions). Cells are colored according to the PhenoGraph cluster. Subsets of cells from cHL and RLN samples are shown on the same coordinates below, respectively. C, Proportion of cells in each cluster originating from cHL and RLN sam-ples. Clusters labeled in red highlight Treg clusters. Dashed white line represents the proportion of RLN cells in the total population. D, The proportion of cells assigned to a given immune cell type (as determined by cluster) was calculated for each sample. Box plots summarize the distribution of the proportions for all samples, grouped by tissue type (cHL or RLN). P values calculated using a t test are shown above, and demonstrate a significant expansion in the proportion of Tregs present in cHL compared with RLN. E, Proportion of CD4+ T cells (non-Treg) assigned to various subsets, calculated per sample and summarized with box plots (see Methods for definition of subtypes). TFH, follicular helper T cells. F–G, Proportion of immune cell types as in D–E, with samples separated according to EBV status (RLN not included). H, Proportion of immune cell types as in E, with samples separated according to histologic subtype (RLN not included).
Cluster_name
Progenitor
B-C1-Naïve
T-C4-Naïve
CD8-C1-Cytotoxic
CD8-C2-Memory
CD4-C1-Helper
CD4-C2-Helper
CD4-C3-Helper
CD4-C4-Treg
CD4-C5-Treg
CD4-C6-Treg
Plasmacytoid-DC
Macrophage
B-C2-Naïve
B-C3-Mature
B-C4-Mature
B-C5-Mature
B-C6-GCB
B-C7-Plasma
T-C1-Naïve
T-C2-Naïve
T-C3-Naïve cHL only RLN only
Naïve T cells
Naïve
NaïveTh2 Th17
ExhaustedT cells
NK cellsMature
B cells
Macrophages
PlasmacytoidDCs
B
CytotoxicT cells
Th1
TregsT cellsCD3D Component
B cell
CD4-C5-T
reg
CD4-C1-H
elper
CD4-C3-H
elper
Macrop
hage
CD4-C4-T
reg
CD4-C6-T
reg
Progen
itor
T-C1-N
aïve
CD8-C1-C
ytotox
ic
B-C3-M
ature
B-C7-P
lasma
B-C2-N
aïve
Co-stimulatory moleculeCyokine/chemokineEffector moleculeInhibitory receptorMacrophageMemory T cellNaïve T cellNK cellPlasma cellPlasmacytoid DCProgenitorT cellT helperTranscription factorTreg
CD8BCD4CD19MS4A1(CD20)IGHDSDC1 (CD138)NCAM1CLEC4C (CD303)NRP1 (CD304)CD68IL3RAIDO1CD34CCR7IL7RLEF1SELL (CD62L)CD44EOMESID2TIGITHAVCR2CTLA4LAG3CD274 (PD-L1)PDCD1 (PD-1)GZMAGZMKICOS (CD278)ICOSLGCD28TNFRSF18 (GITR)TNFRSF8 (CD30)CD40LGFOXP3IL2RA (CD25)IKZF2CXCR5BCL6KLRB1 (CD161)CCR4TBX21GATA3IL2IL4IFNGM
acrophagePlasm
acytoid-DC
CD
4-C6-Treg
CD
4-C5-Treg
CD
4-C4-Treg
CD
4-C3-H
elperC
D4-C
2-Helper
CD
4-C1-H
elperC
D8-C
2-Mem
oryC
D8-C
1-Cytotoxic
T-C4-N
aïveT-C
3-Naïve
T-C2-N
aïveT-C
1-Naïve
B-C7-Plasm
aB-C
6-GC
BB-C
5-Mature
B-C4-M
atureB-C
3-Mature
B-C2-N
aïveB-C
1-Naïve
Progenitor
1.00
0.75
0.50
Type
HLRLN
TypeHLRLN
PathologyEBV statusNEGPOS
EBV statusNEGPOS
E
HGF
0.5
0.4
0.3
0.2
0.1
0.0
0.3
0.2
0.1
0.0
0.3
0.8
0.6
0.4
0.2
0.0
0.2
0.1
0.0
P = 0.0159 P = 0.364 P = 0.00585 P = 0.0534
Th1 Th2 Th17 TFH
Th1 Th2 Th17 TFHTh1 Th2 Th17 TFH
0.25Prop
ortio
n of
cel
ls in
clu
ster
Prop
ortio
n of
CD
4+
T ce
lls (n
on-T
reg)
Prop
ortio
n of
CD
4+ T c
ells
(non
-Tre
g)
Prop
ortio
n of
CD
4+ T c
ells
(non
-Tre
g)
Prop
ortio
n of
all
cells
0.00
CD4+ T cells(non-Treg)
CD8+ T cells Tregs
P = 0.098 P = 0.302 P = 0.00422 P = 0.695
HDMCHDNS
P = 0.00422 P = 0.498 P = 0.00487 P = 0.44
B-C1-N
aïve
T-C3-N
aïve
B-C4-M
ature
T-C2-N
aïve
CD4-C2-H
elper
T-C4-N
aïve
Plasmac
ytoid-
DC
B-C5-M
ature
B-C6-G
CB
CD8-C2-M
emory
−4
−2
0
2
4
P = 0.00415 P = 0.962 P = 0.166
TypeHLRLN
0.8DC
A
P = 0.264
CD4+ T cells(non-Treg)
CD8+ T cells Tregs
P = 0.231 P = 0.000136
0.6
0.4
0.2
0.0Pr
opor
tion
of a
ll ce
lls
Typ
Aoki et al.RESEARCH ARTICLE
410 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Figure 2. Detailed characterization and coexpression patterns of regulatory T cells in the TME of cHL. A, Violin plots showing distribution of expres-sion values (normalized log-transformed) for genes associated with Treg function. Cells from three cluster types are included: CD4+ T cells (non-Treg; CD4-C1-Helper, CD4-C2-Helper, and CD4-C3-Helper), LAG3+ Tregs (CD4-C5-Treg), and other Tregs (CD4-C4-Treg and CD4-C6-Treg). B, The number of individual cells coexpressing the Treg markers LAG3 and FOXP3 in all Treg clusters. C, Proportion of LAG3 and PDCD1 (PD-1)–positive cells in each cluster. D, Proportion of LAG3 and PD-1–positive cells in all Tregs, CD4+ T cells (non-Tregs), and all CD8+ T cells. E, Heat map showing mean expression of inhibitory receptors for cluster subsets. Expression values have been scaled row-wise for visualization. F, UpSet plot showing coexpression patterns of inhibitory receptors (LAG3, PD-1, TIGIT, TIM3, and CTLA4) for individual cells in the LAG3+ Treg cluster. G, Cellular trajectories were inferred using diffusion map analysis of cells in all CD4+ T-cell clusters (cHL cells only). Individual cells are shown in the first two resulting dimensions, and are colored according to cluster (LAG3+ Treg cluster is shown in bold). Axis labels indicate the signature most correlated with each dimension (see Methods).
PDCD1 (PD-1) TIGITLAG3 PIM3 BATF
LAG3+
Tregs (cell-level expression)
FOXP3+
3,884 376 1,134
CTLA4
Nor
mal
ized
log
coun
ts
0
0.0 0.0
0.1
0.2
0.3 B cellsCD4+ T cells (non-Treg)CD8+ T cellsMacrophagesNaïve T cellsPlasmacytoid DCsProgenitorsTregs
0.1
0.2
0.3
0.4
Prop
ortio
n of
LAG
3+ cel
ls
Prop
ortio
n of
PD
-1+ c
ells
0.5
B cellsCD4+ T cells (non-Treg)CD8+ T cellsMacrophagesNaïve T cellsPlasmacytoid DCsProgenitorsTregs
LAG3 PDCD1 (PD-1)1
0.5
0
−0.5
−1
TIGIT
FOXP3
IL2RA (CD25)
CTLA4
LAG3
CD
4+ T cells (non-Treg)
LAG3
+ Tregs
Other Tregs
CTLs
HAVCR2 (TIM3)
PD-1
CD8+ T ce
lls
CD8+ T ce
lls (no
n-Treg
)Tre
gs
LAG3+ Tregs(cell-level expression)
0.00
0
−0.02
0.00
−0.01 0.00 0.01 0.02
Cluster_nameCD4-C1-HelperCD4-C2-HelperCD4-C3-HelperCD4-C4-TregCD4-C5-TregCD4-C6-Treg
0.02
Dim
ensi
on 2
G2−
M
Dimension 1T-cell terminal differentiation01,0002,000
Num cells3,000
PD-1+
TIM3+
TIGIT+
LAG3+
CTLA4+
600
300
Ove
rlap
900
100198
0
388
304
24922
319
217
1
12011
711
094 85 80 59 48 46 45 36 34 34 34 29 29 23 17 14 5 4 1
0.050.100.150.20
0.00
Prop
ortio
n po
sitiv
e ce
lls
0.1
0.2
0.3
2
4
0
0
CD
4+ T
cel
ls (n
on-T
reg)
CD
4-C
5-Tr
egC
D8-
C1-
Cyt
otox
icC
D4-
C3-
Hel
per
CD
4-C
4-Tr
egPr
ogen
itor
CD
4-C
6-Tr
egM
acro
phag
eC
D4-
C1-
Hel
per
CD
4-C
2-H
elpe
rT-
C3-
NaÏ
vePl
asm
acyt
oid-
DC
T-C
1-N
aÏve
T-C
2-N
aÏve
CD
8-C
2-M
emor
yT-
C4-
NaÏ
veB-
C7-
Plas
ma
B-C
6-G
CB
B-C
2-N
aÏve
B-C
1-N
aÏve
B-C
3-M
atur
eB-
C4-
Mat
ure
B-C
5-M
atur
e
CD
4-C
3-H
elpe
rC
D4-
C2-
Hel
per
CD
8-C
1-C
ytot
oxic
CD
4-C
5-Tr
egPr
ogen
itor
CD
4-C
1-H
elpe
r
B-C
6-G
CB
CD
4-C
4-Tr
egC
D4-
C6-
Treg
T-C
3-N
aÏve
Plas
mac
ytoi
d-D
C
T-C
1-N
aÏve
CD
8-C
2-M
emor
yT-
C4-
NaÏ
veB-
C7-
Plas
ma
T-C
2-N
aÏve
B-C
2-N
aÏve
B-C
1-N
aÏve
B-C
5-M
atur
e
B-C
4-M
atur
eB-
C3-
Mat
ure
Mac
roph
age
CD
4+ T
cel
ls (n
on-T
reg)
CD
4+ T
cel
ls (n
on-T
reg)
CD
4+ T
cel
ls (n
on-T
reg)
LAG
3+Tr
egs
LAG
3+Tr
egs
LAG
3+Tr
egs
LAG
3+Tr
egs
LAG
3+Tr
egs
CD
4+ T
cel
ls (n
on-T
reg)
Oth
erTr
egs
Oth
erTr
egs
Oth
erTr
egs
Oth
erTr
egs
Oth
erTr
egs
0
1
2
3
1234
0 0
0
1
2
3
4
0
1
2
3
4
0
1
2
3
01234
12345
1
2
3
2
4
IL2RA FOXP3 TNFRSF18 (GITR) ITGB2
C
A
D E
F G
B
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 411
CD4+ T-cell clusters (Fig. 2C). Interestingly, CD8+ T cells, including CTLs, are not the dominant population expressing PD-1 and LAG3 (Fig. 2C and D), indicating the importance of the CD4+ T-cell population for immune-checkpoint regulation in cHL. Notably, the expression pattern of inhibitory receptors was variable among T-cell subsets (Fig. 2E), suggesting a specific role of each inhibitory receptor in each T-cell subset in cHL. Analyzing coexpression patterns on the single-cell level revealed that the majority of LAG3+ T cells coexpressed CTLA4, which is known as a more universal Treg marker, but not PD-1 (Fig. 2F). Similarly, most PD-1+ T cells did not coexpress LAG3. CTLA4 was also coexpressed by FOXP3+ T cells (Supplementary Fig. S6A). These coexpression patterns were validated using flow cytometry (Supplementary Fig. S7A and S7B). Interestingly, LAG3, TIGIT, and PD-1 were not coexpressed by the majority of CD8+ T cells. Furthermore, although we observed a trend toward higher proportions of non–T follicular helper PD-1+ CD4+ T cells in RLN samples, the proportion of LAG3+ cells was signifi-cantly higher in cHL, suggesting a unique role of LAG3+ CD4+ T cells in cHL pathogenesis (Supplementary Fig. S7C).
To explore the functional role of LAG3+ T cells, we next applied the diffusion map algorithm (25, 26) with the aim of characterizing differentiation states among CD4+ T cells (Fig. 2G). Most T cells were grouped by PhenoGraph cluster, and the first dimension showed a trajectory beginning with naïve T cells and ending with Tregs. LAG3+ T cells were enriched at the far end of this dimension, which was correlated with genes repre-sentative of a terminal differentiation signature (Methods; Sup-plementary Fig. S8A). Consistent with a previous report that showed LAG3+ T cells confer suppressive activity through their significantly reduced proliferation activity (27), LAG3+ T cells were also located in the middle-to-negative end of the second dimension, which correlated with G2–M cell-cycle and glycolysis signature genes (Supplementary Fig. S8B). Furthermore, the most positively correlated genes with dimension 1 were LAG3, LGMN, and CTLA4, which are known markers of suppressive function in Tregs, indicating the suppressive signature of LAG3 in these T cells (Supplementary Fig. S8C and S8D).
cHL Cell Line Supernatant Can Induce LAG3+ T CellsTo characterize the immunosuppressive signature of Tregs
in cHL, we investigated the cytokine expression of LAG3+ T cells. Among the CD4+ cluster T cells, LAG3+ T cells had higher expression of the immune-suppressive cytokines IL10, TGFβ, and IFNγ compared with LAG3− T cells (Fig. 3A). These characteristics are consistent with the profile of type 1 regula-tory T cells (28, 29).
Taken together, our data consistently demonstrate a sup-pressive phenotype of LAG3+ T cells in cHL. We hypothesized that cytokines or chemokines produced by HRS cells might influence the TME in cHL. Thus, we next assessed the effect of supernatant transfer of various lymphoma cell lines on the expansion of T cells in vitro. After 14 days of activation of T cells, flow cytometry analysis confirmed that CD4+ CD25+ T cells cocultured with cHL cell line supernatant expressed significantly higher levels of LAG3 as compared with those cocultured with diffuse large B-cell lymphoma (DLBCL) cell line supernatant or medium only (Fig. 3B and C). Luminex analysis revealed that the presence of cHL cell line supernatant resulted in enrichment of multiple cytokines and chemokines
as compared with DLBCL cell lines, including TARC/CCL17, TGFβ, and IL6, which are known enhancers of Treg migra-tion and differentiation (refs. 30–38; Fig. 3D). Consistent with scRNA-seq results, CD4+ LAG3+ T cells isolated by FACS secreted significantly higher amounts of IL10 and TGFβ com-pared with CD4+ LAG3− T cells (Fig. 3E). Notably, CD4+ LAG3+ T cells suppressed the proliferation of responder CD4+ T cells when cocultured in vitro, confirming an immunosup-pressive function of the LAG3+ T cells (Fig. 3F).
Spatial Assessment of LAG3+ T Cells and HRS CellsWe next sought to understand the spatial relationship between
LAG3+ T cells and malignant HRS cells. IHC of all cases revealed that LAG3+ T cells were enriched in the cHL TME compared with RLN, and in a subset of cHL cases, HRS cells were closely surrounded by LAG3+ T cells (Fig. 4A). Of note, our single-cell analysis revealed that LAG3 expression was significantly higher in cases with MHC class II–negative HRS cells (n = 6) as compared with those with MHC class II–positive HRS cells (n = 16), but was not correlated with EBV status or histologic subtype (Fig. 4B; Supplementary Fig. S9A–S9C). Strikingly, when examining cells within the CD4-C5-Treg cluster, LAG3 was identified as the most upregulated gene in MHC class II–negative cells compared with MHC class II–positive cells (Fig. 4C). Characterization of immune markers using IHC showed not only a marked increase in LAG3+ T cells but also a decrease in FOXP3+ T cells in MHC-II–negative cases when compared with MHC-II–positive cases (Fig. 4D). There was no difference in the proportion of CTLA4+ CD4+ T cells by MHC-II status, suggesting the LAG3+ cells represent a distinct subpopulation of the Hodgkin lymphoma–specific CTLA4+ cells previously reported (ref. 12; Supplementary Fig. S9D). To validate these findings, we assessed the spatial relation-ship between HRS cells and LAG3+ CD4+ T cells using multicolor IHC (Fig. 4E–G). We confirmed that the density of LAG3+ T cells in HRS-surrounding regions was significantly increased in MHC-II–negative cases, but not correlated with either MHC-I status, pathologic subtype, or EBV status (Fig. 4E; Supplemen-tary Fig. S10A). Similarly, the average nearest-neighbor distance between CD30+ cells (HRS cells) and their closest LAG3+ T cell was significantly shorter in MHC-II–negative cHL cases (Fig. 4F). In contrast, the density of HRS-surrounding FOXP3+ T cells was higher in cases with MHC-II–positive HRS cells (Fig. 4E; Sup-plementary Fig. S10B), and the nearest-neighbor distance from HRS cells to FOXP3+ cells was also shorter in these cases (Fig. 4F; Supplementary Fig. S11A and S11B).
To further investigate the spatial relationship between HRS cells and their surrounding cells, we next assessed the expres-sion of surface and intracellular markers in all cHL study cases using IMC, which allows for simultaneous interrogation and visualization of 35 protein markers in the spatial context of the TME. Consistent with IHC analysis, IMC revealed that MHC-II–negative cHL cases showed numerous LAG3+ CD4+ cells, with rare FOXP3+ CD4+ cells (Fig. 5A; Supplementary Fig. S12A). In contrast, MHC-II–positive cases showed rare LAG3+ CD4+ T cells and abundant FOXP3+ CD4+ T cells rosetting the HRS cells. We also confirmed the observed sig-nificantly shorter nearest-neighbor distances between HRS cells and their closest LAG3+ T cell in MHC-II–negative cHL cases when compared with MHC-II–positive cHL cases using IMC data (Supplementary Fig. S12B and S12C).
Aoki et al.RESEARCH ARTICLE
412 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Figure 3. An immune-suppressive microenvironment is characteristic of cHL and is associated with LAG3 positivity. A, Density plots showing the expression of suppressive cytokines for cells in the LAG3+ Treg cluster (CD4-C5-Treg). Cells are grouped by LAG3 positivity, and P values were calculated using t tests. B, Representative flow-cytometric analysis of CD25 and LAG3 expression on T cells isolated from PBMCs cultured with supernatant of the cHL cell line L-1236 or medium, respectively. C, The proportion of LAG3+ cells among CD4+ T cells cultured with supernatant of cHL cell lines (KM-H2, L-428, and L-1236), DLBCL cell lines (OCI-Ly1 and Karpas-422), or medium only. Data, mean ± SEM (n = 3). *, P ≤ 0.05; **, P ≤ 0.01. D, The amount of cytokines and chemokines in the supernatant of cHL cell lines and DLBCL cell lines by Luminex analysis. Data, mean ± SEM (n = 3). E, The amount of cytokines and chemokines in the supernatant of FACS-sorted CD4+ LAG3+ cells and CD4+ LAG3− cells by Luminex analysis. Data, mean ± SEM (n = 4). ****, P ≤ 0.0001. F, Left, a representative experiment showing proliferation of CD4+ responder T cells alone (bottom), cocultured with FACS-sorted CD4+ LAG− T cells (middle), or cocultured with FACS-sorted CD4+ LAG3+ T cells (top). Right, the percentage of proliferating CD4+ responder T cells in each coculture condition, relative to the normal proliferation rate (alone). Data, mean ± SEM (n = 4). *, P ≤ 0.05.
25IL10
IL2
L-1236
CD
4-BV
786
LAG
3-PE
CD25-BB515
CD4-BV786
10
8
8
6
4
2
0
**
*
HL cell
line
DLBCL c
ell line
HL cell
line
Medium
6
4
Percentage of CD25+LAG3+ T cells
Percentage of CD25+LAG3+ T cells
Percentage of CD25+LAG3+ T cells
Perc
enta
ge a
mon
g C
D4+ T
cel
l
8
6
4
2
0
Perc
enta
ge a
mon
g C
D4+ T
cel
l
Perc
enta
ge a
mon
g C
D4+ T
cel
ls
2
25,000
200,000
150,000
800 400 2,000
1,500
1,000
500
0
300
200
100
0
600
400
200
0
100,000
50,000
0
300E
D
B
A
C
F1,500
150
CD4+LAG3+
CD4+LAG3−
ResponderT-cell only
Proliferation dye
*
100
50
100 101 102 103 104 105
0
1,000
500
0
**** ****
200
100
0
LAG3-p
ositiv
e cells
LAG3-n
egati
ve ce
lls
LAG3-p
ositiv
e cells
LAG3-n
egati
ve ce
lls
LAG3-p
ositiv
e cells
LAG3-n
egati
ve ce
lls
1,500 4,000 1,000
800
600
400
200
0
3,000
2,000
1,000
0
1,000
500
0
TAR
C/C
CL1
7 (p
g/m
L)R
ANTE
S (p
g/m
L)
IL10
(pg/
mL)
TGFβ
1 (p
g/m
L)
Rel
ativ
e %
of C
D4
prol
ifera
tion
TNF°
(pg/
mL)
IL16
(pg/
mL)
MD
C/C
CL2
2 (p
g/m
L)
TGFβ
1 (p
g/m
L)
IL6
(pg/
mL)
SDF-
1a+b
/CXC
L12
(pg/
mL)
20,000
15,000
10,000
5,000
0
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
0
KM-H2
L-428
L-123
6
OCI-Ly1
Karpas
-422
Medium
Medium
L-1236 Medium
TGFB1
IFNG
15
10
5
0
201510
5
−50
50
10
5
0
40302010
−100
105
104
103
102
101
100
100 101 102 103
34.2% 11.9%
104 105
105
104
103
102
101
100
100 101 102 103
18.2%
104 105
105
104
103
102
101
100
100 101 102 103
2.83%
104 105
105
104
103
102
101
100
100 101 102 103 104 105
Normalized log counts
Normalized log counts
LAG3 statusLAG3−
LAG3+
LAG3 statusLAG3−
LAG3+
LAG3 statusLAG3−
LAG3+
LAG3 statusLAG3−
LAG3+
Den
sity
Den
sity
Den
sity
Den
sity
P = 8.8e–85
P = 0.18 P = 1.1e–35
P = 6.8e–12
0.00
0.00
0.01
0.02
0.03
0.04
0.05
0.05
0.10
0.15
0.20
0.25
Normalized log counts
0.00
0.05
0.10
0.15
0.20
0.25
Normalized log counts
0.00
0.05
0.10
0.15
0.20
0.25
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 413
Figure 4. Spatial distribution of HRS cells and LAG3+ T cells in cHL tumors. A, Representative LAG3 IHC of cHL tumor biopsies and a reactive lymph node (× 400, CHL03 and CHL05). B, Box plot showing mean LAG3 expression of cells in the LAG3+ Treg cluster separated by MHC class II (MHC-II) status on HRS cells as determined by IHC (P = 0.0186; t test). C, Volcano plot showing differentially expressed genes between cells in the LAG3+ Treg cluster originating from MHC-II–positive or MHC-II–negative cases. The top five genes by absolute log fold change in each direction are highlighted in red. The y-axis summarizes P values corrected for multiple testing using the Benjamini–Hochberg method. D, IHC staining for major immune cell markers in repre-sentative cases with either MHC-II–positive HRS cells (left) or MHC-II–negative HRS cells (right; × 400). (continued on next page)
125
3
2
1
NEG POSMHC class II status (HRS cells)
P = 0.0186
Mea
n LA
G3
expr
essi
on (n
orm
aliz
ed lo
g co
unts
)
CD4-C5-Treg (by MHC class II status (HRS cells))
100
SELL
RPS4Y1
RPS2
IFITM1
LAG3
CD81ETS1
DDX5PTMS XIST
MALAT1 JUNBFOSSRGN
GZMKH3F3BTSHZ2
XBPDUSP1
RGS1
Up in NEG >< Up in POS
MT–ND3RPL23A
RPS24GAPDH
RPSAACTB- MT-CO1
COTL1PKM
S100A4LTB
TM B10INC00152
RPLP0RPL5
RPL7IL2RA
YBX1LDHB
RPS3A
75
50
25
0
−1.0 −0.5 0.0 0.5LogFC
−Log
10 (q
valu
e)
1.0
H&E
CD30
CD3
CD4
MHC-II
LAG3
FOXP3
RLN HL surrounded HL nonsurrounded
A
BD
C
MHC-IIpositive
MHC-IInegative
The Number of LAG3+ T Cells in the TME Is Correlated with Loss of MHC-II Expression in a Large Validation Cohort
We next validated our findings using IHC of an independ-ent cohort of 166 patients uniformly treated with first-line ABVD [adriamycin (aka doxorubicin), bleomycin, vinblastine, and dacarbazine] as described in Steidl and colleagues (6), and investigated the potential prognostic value of the presence of
LAG3+ T cells. Consistent with the results from scRNA-seq, we found that the proportion of LAG3+ T cells present in tumor tissue was significantly higher in cases with MHC-II– negative HRS cells as compared with those with MHC-II–positive HRS cells, but was not associated with EBV status (Fig. 5B and C). In addition, we observed a trend toward shortened disease-specific survival (DSS; P = 0.072) and over-all survival (OS; P = 0.12) in patients with an increased number of LAG3+ T cells (Supplementary Fig. S13A
Aoki et al.RESEARCH ARTICLE
414 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
and S13B). Of note, a high proportion of LAG3+ T cells (>15%) and CD68+ tumor-associated macrophages (≥5%; ref. 6) were identified as independent prognostic factors for DSS by multivariate Cox regression analysis (also considering MHC-II expression and International Prognostic Score as variables; Supplementary Fig. S13C). In the absence of statisti-cally significant outcome correlates in the present cohorts of pretreatment Hodgkin lymphoma samples, we examined an independent cohort of patients with relapsed cHL uniformly treated with high-dose chemotherapy followed by autologous stem cell transplantation (ASCT; ref. 4). We similarly found that abundant LAG3+ T cells were associated with unfavorable post-ASCT survival, although statistical significance was not reached, likely due to sample size (Supplementary Fig. S13D).
Cross-talk between HRS Cells and LAG3+ T Cells in cHL
To investigate the role of HRS cells in their interaction with the cHL microenvironment, we next explored Affymetrix gene-expression data generated from microdissected HRS cells of primary Hodgkin lymphoma samples (ref. 39; see Supplementary Methods for details). We validated the high expression level of the cytokines and chemokines that we observed in the in vitro Luminex assay (Fig. 6A). Notably, IL6, encoding a known promoter of Tr1 cell differentiation (38), was the only cytokine gene that showed significantly higher expression in MHC-II–negative HRS cells compared with
MHC-II–positive HRS cells. CD4+LAG3+ T cells were also induced by IL6 in vitro (Fig. 6B), indicating that IL6 might play a role in inducing CD4+LAG3+ T cells in cHL.
MHC-II is also a known LAG3 ligand (40, 41). To investigate the interaction between LAG3+ T cells and MHC-II on HRS cells, we generated CIITA knockouts in the L-428 cHL cell line, as CIITA is the master regulator of MHC-II expression, and confirmed the MHC-II–negative status of these CIITA knock-out cells (Supplementary Fig. S14A). Next, we isolated LAG3+ T cells induced from peripheral blood mononuclear cells (PBMC) using L-428 supernatant transfer. In coculture of these LAG3+ T cells with either CIITA wild-type or knockout L-428 cells, we observed that LAG3 expression was significantly decreased with MHC-II–positive L-428, suggesting negative regulation of LAG3+ T-cell function through a direct MHC-II–LAG3 interaction (Fig. 6C). We also evaluated expression of cytokines, including IL6 and TARC, from both CIITA wild-type and knockout L-428 cells, and observed no significant differ-ence (Supplementary Fig. S14B). Taken together, these find-ings suggest that although IL6 induces LAG3+ T cells, MHC-II positivity actively depletes them; thus, a mechanism for induc-tion and persistence is present only in MHC-II–negative tumors. We also investigated the expression of other LAG3 ligands on HRS cells according to MHC-II status in the Affym-etrix data set, and found that their expression was not signifi-cantly increased relative to normal GCB cells (Supplementary Fig. S14C). In addition, there was no correlation between the
Figure 4. (Continued) E, Box plot showing the density of CD4+ LAG3+ T cells (left) or CD4+ FOXP3+ (right) in the region surrounding CD30+ cells (HRS) for each sample, separated by tissue type and MHC-II status on HRS cells (t test; ns: P > 0.05; *, P ≤ 0.05; ***, P ≤ 0.001; ****, P ≤ 0.0001). F, Average nearest-neighbor (NN) distance from an HRS cell (CD30+) to the closest CD4+ LAG3+ cell (left) or CD4+ FOXP3+ cell (right) was calculated per sample, and separated by MHC-II status on HRS cells. P values were calculated using t tests. G, Multicolor immunofluorescence staining (CHL03 and CHL05) for CD30 (yellow), MHC-II (green), and LAG3 (magenta) shows localization of LAG3+ CD4+ T cells to the region surrounding HRS cells in cases with MHC-II–negative HRS cells.
4,000
E F
G
10,000
7,500
5,000
2,500
0
2,000
0
80
60
60
40
20
40
20NEG
P = 0.00144 P = 0.038HRS to LAG3+ HRS to FOXP3+
POSMHC class II status (HRS cells)
MHC-IIpositive
DAPI CD30 MHC-II LAG3 Merge
MHC-IInegative
NEG POSMHC class II status (HRS cells)
Aver
age
NN
dis
tanc
e (µ
m)
Aver
age
NN
dis
tanc
e (µ
m)
MHC-IINEG
MHC-IIPOS
RLN MHC-IINEG
MHC-IIPOS
RLN
LAG
3+ den
sity
(HR
S re
gion
)(c
ells
/mm
2 )
FOXP
3+ den
sity
(HR
S re
gion
)(c
ells
/mm
2 )
****
*******
*ns
ns
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 415
Figure 5. Coexpression patterns and localization of immune cells according to HRS MHC-II status, using IMC. A, A representative case with MHC-II–negative cHL case (CHL5) shows numerous LAG3+ CD4+ T cells (i) and few FOXP3+ CD4+ T cells (ii), with the LAG3+ cells rosetting the HRS cells (iii–iv). In contrast, a representative MHC-II–positive cHL case (CHL3) shows rare LAG3+ CD4+ T cells (v) and abundant FOXP3+ CD4+ T cells (vi), the latter surround-ing HRS cells (vii–viii). B, Comparison of the proportion of LAG3+ cells by MHC-II status in a validation cohort (6). P values were calculated using t tests. C, Comparison of the proportion of LAG3+ cells by EBV status in a validation cohort (6). P values were calculated using t tests.
MHC-IInegative
i
A
B C
ii iii iv
viiiviiviv
MHC-IIpositive
40
30
Prop
ortio
n of
pos
itive
cel
ls
20
P < 0.001
LAG3+ cells
10
0
MHC II+
MHC II−
CD30/LAG3LAG3+ CD4 cellsFOXP3+ CD4 cellsCD8 cells
CD30/FOXP3 CD30/CD4/CD8/FOXP3/LAG3 LAG3+CD4/FOXP3+CD4/CD8
RSRS
RS
RS RS
RS
25 lm
25 lm25 lm
25 lm
RSRS
CD30 LAG3 CD30 FOXP3 CD30 CD4 CD8 FOXP3 LAG3
LAG3+ CD4 cellsFOXP3+ CD4 cellsCD8 cells
CD30 LAG3 CD30 FOXP3 CD30 CD4 CD8 FOXP3 LAG3
40
30
Prop
ortio
n of
pos
itive
cel
ls
20
P = 0.47
LAG3+ cells
10
0
EBV+
EBV−
expression level of LAG3 ligands according to MHC-II status, suggesting no direct interaction with these ligands in cHL.
T Cells from cHL Clinical Samples Are Activated after Removal of LAG3+ T Cells
To confirm the pathogenic role of LAG3+ T cells in cHL clin-ical samples, we sorted both CD4+ LAG3+ CD25+ T cells and the remaining T cells from cell suspensions of 4 patients. We then cocultured T cells with or without CD4+ LAG3+ CD25+ T cells in vitro, and observed that proliferation was suppressed in the T cells cocultured with the LAG3+ population, whereas proliferation and expression of the intracellular cytokine, TNFα, were significantly increased in the population cultured without LAG3+ cells (Fig. 6D and E; Supplementary Fig. S15).
These results support an immunosuppressive function of CD4+ LAG3+ T cells in cHL clinical samples, providing a pre-clinical rationale for targeting LAG3+ T cells and their interac-tions to promote reactivation of T cells in a subset of patients.
Our results suggest a model in which the immunosuppres-sive microenvironment of MHC-II–negative HRS cells (type 1) is highly organized and in part induced by CD4+ LAG3+ T cells, which in turn are induced by cytokines and chemokines produced by HRS cells (Fig. 7). Aggregating all of these results, we reason that cross-talk between LAG3+ T cells and HRS cells may be an essential mechanism of immune escape in cHL, with potential implications for outcome prediction of differential checkpoint inhibitor therapy, including response durability and overcoming resistance.
Aoki et al.RESEARCH ARTICLE
416 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Figure 6. Interactions of HRS cells and CD4+ LAG3+ T cells. A, The expression of cytokines and chemokines on microdissected HRS cells from primary Hodgkin lymphoma samples (separated by MHC class II status) and germinal center cells from reactive tonsil (GCB; t test; ns: P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ****, P ≤ 0.0001). B, The proportion of LAG3+ cells among CD4+ T cells after coculture with supernatant of cHL cell lines (L-1236), medium with IL6, or medium only. Data, mean ± SEM (n = 4; **, P ≤ 0.01). C, Left, a representative experiment showing LAG3 expression of CD4+LAG3+ T cells (HLA-matched with L-428) after coculture with either CIITA wild-type (WT; red) or CIITA knockout (KO) L-428 (blue). LAG3 expression on the T cells was significantly decreased after coculture with MHC-II–positive (CIITA KO) cells. Right, the percentage of highly expressing LAG3+ T cells after coculture with L-428 CIITA variants (wild-type or knockout). Data, mean ± SEM (n = 3). *, P ≤ 0.05. D, Left, a representative experiment showing proliferation of CD4+ T cells sorted from cHL clinical samples (red), and the same cells cocultured with CD4+LAG3+ CD25+T cells from cHL clinical samples (blue). Right, the percentage of proliferating cells in each condition is shown as the mean ± SEM (n = 4). *, P ≤ 0.05 (t test). E, The expression of TNFα in the populations described in D is shown as the mean ± SEM (n = 3). *, P ≤ 0.05 (t test).
12
15
12
9
6
IL6A
C
D E
B
RANTES/CCL5 TNFα
TARC/CCL17
********
**** ****
ns****
ns
* ns****
****
10
8
6
8
10
10
Percentage of CD25+ LAG3+ T cells
NS
**8
6
4
2
0
60
40
**
*
20
0
100
80
60
40
20
Nor
mal
ized
to m
ode
Nor
mal
ized
exp
ress
ion
Nor
mal
ized
to m
ode
% o
f div
ided
cel
ls
% o
f TN
Fα+
cells
0
100
80
60
40T cells
Proliferation dye
T cells + LAG3+ T cells
CIITA KO L428
WT L428
T cells (no LAG3+ T cells)
20
0
100
80
60
40
20
* *
0
100
80
60
40
20
0
T cells
cultu
red
with LA
G3+ cells T ce
lls
(no LA
G3+ cells)
T cells
cultu
red
with LA
G3+ T cells T ce
lls
(No L
AG3+ cells)
101 102 103
LAG3
104
WT
CIITA KO 1
CIITA KO 2
Perc
enta
ge a
mon
g C
D4+ T
cel
ls
LAG
3-hi
gh e
xpre
ssed
cel
ls (%
)
L-123
6 IL6
Medium
only
8
66
4
GCB
MHC-II NEG
MHC-II POS
GCB
MHC-II NEG
MHC-II POS
GCB
MHC-II NEG
MHC-II POS
GCB
MHC-II NEG
MHC-II POS
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 417
DISCUSSION
Using scRNA-seq and IMC at an unprecedented scale, we comprehensively characterized immune cell populations to generate an immune cell atlas of the TME in cHL at both the RNA and protein levels. In addition to reproducing known TME characteristics in cHL at single-cell resolution, such as a Treg/Th2-rich environment (10, 11), a Th17-predominant profile in EBV+ cHL (42), and a CTLA4+ PD-1− T-cell popu-lation (12), we also identified and characterized in detail novel cellular subpopulations, including immunosuppres-sive LAG3+ T cells (40) that are linked to unique pathologic and clinical parameters. Strikingly, Treg populations and the LAG3+ T-cell population in particular emerged as the most highly enriched and cHL-characteristic cellular component.
LAG3 is a selective marker of Tr1 cells, which in contrast to natural Tregs derived from the thymus, are known as induced Tregs that exhibit strong immunosuppressive activ-ity (20–22, 27). Consistent with characteristics of Tr1 cells, the expression of the suppressive cytokines IL10 and TGFβ (22, 27) was very high in LAG3+ T cells, whereas FOXP3 was not coexpressed in LAG3+ T cells in our scRNA-seq and IMC data. Furthermore, LAG3+ T cells demonstrated substantial suppressive activity in vitro, indicating an immunosuppressive role of these cells in the TME of cHL.
Unlike previous reports that found that EBV infection increased Tr1-related gene expression including LAG3 in cHL (20), we identified a significant LAG3+ Treg population regardless of EBV status by scRNA-seq, multicolor IHC, IMC, and single-color IHC analyses in independent cohorts. However, our study revealed that LAG3+ CD4+ T cells were enriched in cases with MHC-II–negative HRS cells. Interest-ingly, MHC-II deficiency was reported as a predictor of unfa-vorable outcome after PD-1 blockade (43). Our scRNA-seq data revealed that each T-cell subset had a specific expression pattern of inhibitory receptors including PD-1 and LAG3. Most notably, the majority of LAG3+ CD4+ T cells did not coexpress PD-1, and the absence of PD-1 has been reported to represent functionally active Tregs in solid cancer (44), indicating the potential of LAG3 as a separate and comple-mentary immunotherapeutic target in cHL. The FOXP3+
Tregs that are enriched in MHC-II–positive HRS cells in this study might be similar to the PD-1–negative FOXP3+ Tregs previously reported (10).
MHC-II is one of the major ligands of LAG3 (40, 41), and we showed negative regulation of LAG3+ T-cell expres-sion through MHC-II and LAG3 interaction using Hodgkin lymphoma cell lines in vitro. These results are consistent with the patient data showing that LAG3+ CD4+ T cells were preferentially observed surrounding MHC-II–negative HRS cells. Additionally, our in vitro coculture findings suggest that cytokines and chemokines produced by HRS cells may be an important inducer of LAG3+ CD4+ T cells in the TME. In par-ticular, reanalysis of expression on laser microdissected HRS cells revealed that MHC-II–negative HRS cells had higher levels of IL6, a cytokine known to induce Tr1 cells (38). Alter-native ligands of LAG3 that mediate the immune-suppressive function (45, 46) might contribute to these interactions, although we did not observe any differences in their expres-sion on HRS cells according to MHC-II status.
Our findings suggest that LAG3+ T cells induced by cytokines and chemokines from HRS cells play an important role in substantial immunosuppressive activity in the TME of cHL. Importantly, LAG3 is a cancer immunotherapeutic target in ongoing clinical trials in malignant lymphoma, including cHL (NCI trial ID 02061761), and we showed the potential of removing the LAG3+ population as a means of reactivating T-cell activity. Although currently our data do not demonstrate the value of LAG3+ T cells as a prognostic biomarker, and further studies in additional cohorts are pending, it will be critical to evaluate the potential of LAG3+
T cells as a predictive biomarker in the context of treatments targeting LAG3+ T cells and their cellular interactions. In particular, ongoing trials of LAG3-targeting antibodies and antibody–drug conjugates against CTLA4 or CD25 (which would target LAG3+ cells among others) will allow this evalu-ation. Moreover, additional investigations into the biology of immune cell interactions, including LAG3+ T cells and other immune cell types, may be beneficial for the future therapeu-tic development of alternative checkpoint inhibitors.
In conclusion, our comprehensive analysis provides, for the first time, detailed functional and spatial characteristics
Figure 7. Hypothetical model of LAG3+ T-cell and HRS cell interactions in cHL. MHC-II–negative HRS cells (type 1) secrete cytokines that induce LAG3 in CD4+ T cells. CD4+ LAG3+ T cells surround HRS cells and secrete suppressive cytokines. MHC-II–positive cells (type 2) secrete a distinct set of cytokines that attract FOXP3+ and Th17 cells.
Type 1: LAG3 high
IL10
IL6
TARC
MDC/CCL22
MHC class II
LAG3+ Treg
FOXP3+ Treg
Th17 T cellTGF° TNFα
Type 2: LAG3 low
Aoki et al.RESEARCH ARTICLE
418 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
of immune cells in the cHL microenvironment at single-cell resolution. We identified unique expression signatures of TME cells, including LAG3+ T cells, and our findings provide novel insights and texture to the central hypothesis of CD4+ T cell–mediated immune-suppressive activity in the patho-genesis of cHL. Importantly, our findings will facilitate a deeper understanding of the mechanisms underlying the immune-escape phenotype in cHL, and aid in the develop-ment of novel biomarkers and treatment strategies.
METHODSDetailed materials and methods are available in the Supplemen-
tary Data file.
Tissue SamplesFor scRNA-seq, a total of 22 patients with histologically con-
firmed diagnostic (n = 21) or relapsed (n = 1) cHL and reactive lymphoid hyperplasia (but no evidence of malignant disease or systemic autoimmune disease; n = 5) were included in this study. Patients were selected based on the availability of tissue that had been mechanically dissociated and cryopreserved as cell sus-pensions following diagnostic lymph node biopsy from British Columbia (BC) Cancer. Patient characteristics are summarized in Supplementary Tables S1 and S2.
The independent validation cohort consisted of 166 patients with cHL uniformly treated with ABVD at BC Cancer between 1994 and 2007 from the cohort described in Steidl and colleagues (6). This cohort was derived from a population-based registry (Centre for Lymphoid Cancer database, BC Cancer Agency) enriched for treatment failure. The median follow-up time for living patients was 4.1 years (range, 0.5–24.4 years). The relapse cohort consisted of 55 relapsed or refractory cHL patients uniformly treated with high-dose chemotherapy and ASCT at BC Cancer, from the cohort described in Chan and colleagues (4).
This study was reviewed and approved by the University of British Columbia–BC Cancer Agency Research Ethics Board (H14-02304), in accordance with the Declaration of Helsinki. We obtained writ-ten informed consent from the patients or the need for consent was waived in the retrospective study.
scRNA-seq Sample PreparationTo identify live cells, we used DAPI (Sigma-Aldrich) for live–dead
discrimination. Cell suspensions from cHL tumors or reactive lymph node were rapidly defrosted at 37°C, washed in 10 mL of RPMI-1640/10% fetal bovine serum (FBS) solution or RPMI1640/20% FBS solution containing DNase I (Millipore Sigma) and washed in PBS. Cells were resuspended in PBC containing 3% FBS and stained with DAPI for 15 minutes at 4°C in the dark. Viable cells (DAPI-negative) were sorted on a FACS ARIAIII or FACS Fusion (BD Biosciences) using an 85-μm nozzle (Supplementary Fig. S16). Sorted cells were collected in 0.5 mL of medium, centrifuged and diluted in 1× PBS with 0.04% bovine serum albumin (BSA). Cell number was determined using a Countess II Automated Cell Counter whenever possible.
Library Preparation and scRNA-seqIn total, 8,600 cells per sample were loaded into a Chromium
Single-Cell 3′ Chip Kit v2 (PN-120236) and processed according to the Chromium Single-Cell 3′ Reagent Kit v2 User Guide. Libraries were constructed using the Single 3′ Library and Gel Bead Kit v2 (PN-120237) and Chromium i7 Mulitiplex Kit v2 (PN-120236). Single-cell libraries from two samples were pooled and sequenced on one HiSeq 2500 125 base PET lane. CellRanger software (v2.1.0; 10X Genomics) was used to demultiplex the raw data, generate quality metrics, and generate per-gene count data for each cell.
Normalization and Batch CorrectionAnalysis and visualization of scRNA-seq data were performed in
the R statistical environment (v3.5.0). CellRanger count data from all cells (n = 131,151) were read into a single “SingleCellExperiment” object. Cells were filtered if they had ≥ 20% reads aligning to mito-chondrial genes, or if their total number of detected genes was ≥ 3 median absolute deviations from the sample median. This yielded a total of 127,786 cells for analysis. The scran package (v1.9.11) was used to quick-cluster the cells and compute cell-specific sum factors with the method described by Lun and colleagues (ref. 47; see Supple-mentary Methods for details). The scater package (v1.8.0) was used to log-normalize the count data using the cell-specific sum factors.
To remove batch effects resulting from different chips and library preparation, the fast mutual nearest neighbors (MNN) batch correc-tion technique in the scran package was utilized, grouping cells by their chip and using the expression of genes with positive biological components (see Supplementary Methods for details). This produced a matrix of corrected low-dimensional component coordinates (d = 50) for each cell, which was used as input for downstream analyses. Entropy of cell expression before and after batch correction was assessed in R using the method described by Azizi and colleagues (ref. 13; Supplementary Fig. S1B; Supplementary Methods).
Clustering and AnnotationUnsupervised clustering was performed with the PhenoGraph algo-
rithm (48), using the first 10 MNN-corrected components as input. Clusters from PhenoGraph were manually assigned to a cell type by comparing the mean expression of known markers across cells in a cluster (see Supplementary Methods for details). For visualization pur-poses, t-SNE transformation was performed with the scater package using the first 10 MNN-corrected components as input. All differential expression results were generated using the findMarkers function of the scran package, which performs gene-wise t tests between pairs of clusters, and adjusts for multiple testing with the Benjamini–Hochberg method. Diffusion map analysis (25) was performed using the algo-rithm implemented by the scater package (Supplementary Methods).
Multicolor IHC on TMA, Scanning, and Image AnalysisTissue microarray (TMA) slides were deparaffinized and incubated
with each marker of interest (MHC class II, FOXP3, CD8, LAG3, CD4, and CD30), followed by detection using Mach2 horseradish peroxi-dase and visualization using Opal fluorophores (Supplementary Table S4; see Supplementary Methods for details). Nuclei were visualized with DAPI staining. TMA slides were scanned using the Vectra multi-spectral imaging system (PerkinElmer) following the manufacturer’s instructions to generate .im3 image cubes for downstream analysis. To analyze the spectra for all fluorophores included, inForm image analysis software (v2.4.4; PerkinElmer) was used. Cells were first clas-sified into tissue categories using DAPI and CD30 to identify CD30+
DAPI+, CD30− DAPI+, and CD30−DAPI− areas via manual circling and training (Supplementary Fig. S17). The CD30+ DAPI+ regions were considered to be HRS-surrounding regions. Cells were then pheno-typed as positive or negative for each of the six markers (MHC class II, FOXP3, CD8, LAG3, CD4, and CD30). Data were merged in R by X–Y coordinates so that each cell could be assessed for all markers simul-taneously. Nearest-neighbor analysis was performed with the spatstat R package (v1.58-2).
IMCIMC was performed on a 5-μm section of the same TMA described
above. The section was baked at 60°C for 90 minutes on a hot plate, dewaxed for 20 minutes in xylene and rehydrated in a graded series of alcohol (100%, 95%, 80%, and 70%) for 5 minutes each. Heat-induced antigen retrieval was conducted on a hot plate at 95°C in Tris-EDTA buffer at pH 9 for 30 minutes. After blocking with 3% BSA in PBS
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 419
for 45 minutes, the section was incubated overnight at 4°C with a cocktail of 35 antibodies tagged with rare lanthanide isotopes (Supplementary Table S5). The section was counterstained the next day for 40 minutes with iridium and 3 minutes with ruthenium tetroxide (RuO4) as described in Catena and colleagues (49). Slides were imaged using the Fluidigm Hyperion IMC system with a 1-µm laser ablation spot size and frequency of 100 to 200 Hz. A tissue area of 1,000 μm2 per sample was ablated and imaged. Duplicate cores of the same samples were ablated when morphologic heterogene-ity was identified a priori on H&E. IMCTools (https://github.com/ BodenmillerGroup/imctools) was used in conjunction with CellPro-filer (v2.2.0) to segment images and identify cell objects (see Supple-mentary Methods for details).
Cell LinesThe cHL cell lines KMH2, L428, and L-1236 were obtained from
the German Collection of Microorganisms and Cell Cultures (DSMZ; http://www.dsmz.de/) between 2007 and 2010, and were used for experiments within 20 passages. Cultures were grown according to the standard conditions. Human DLBCL cell lines (Karpas-422) were purchased from DSMZ, and maintained in RPMI1640 (Life Tech-nologies) containing 20% FBS. The cell line OCI-Ly1 was obtained from Dr. L. Staudt (NIH) in 2009 and maintained in RPMI1640 (Life Technologies) containing 10% FBS. All cell lines were confirmed negative for Mycoplasma prior to culture using the Venor GeM Myco-plasma Detection Kit, PCR-based (Sigma-Aldrich, MP0025). All cell lines were authenticated using short tandem repeat profiling.
Cell Isolation and Purification of Human T CellsWe purified CD4+ and CD8+ T lymphocytes from PBMCs (see
Supplementary Methods for details). Isolated CD4+ and CD8+ T cells were incubated in either supernatants from cHL cell lines (L-1236, L-428, KM-H2) or DLBCL cell lines (OCI-Ly1 and Karpas-422) or culture medium. At the end of day 14, we washed and analyzed the T cells using flow cytometry for characterization. We purified CD4+ LAG3+ T cells and CD4+ LAG3− T cells by flow sorting on a FACS Fusion (BD Biosciences) using a 85-μm nozzle.
Flow CytometryTo characterize T cells in vitro, we stained cells with a panel of
antibodies including CD3, CD4, CD8, and LAG3 (see Supplementary Methods for details) and assessed them using flow cytometry (LSR-Fortessa or FACSymphony, BD). Flow cytometry data were analyzed using FlowJo software (v10.2; TreeStar; Supplementary Fig. S18). Statistical analyses were performed using GraphPad Prism Version 7 (GraphPad Software Inc.).
In Vitro Suppression AssayTo evaluate the suppressive activity of LAG3+ T cells, we stained
CD4+ T cells (responder cells) with proliferation dye (VPD450; BD Biosciences or CellTrace Violet Cell Proliferation Kit; Thermo Fisher) and activated them using soluble monoclonal antibodies to CD3 and CD28 in PRIME XV T-cell CDM medium or CD3/CD28 Beads (Thermo Fisher). We added purified CD4+ LAG3+ T cells induced by cHL cell line supernatant transfer, or purified from cell suspensions of cHL clinical samples (suppressor cells) at a ratio of 1:1. We calculated the percentage of divided responder T cells by gating on CD4+ cells, and T-cell proliferation was determined based on proliferation dye dilution using flow cytometry (LSRFortessa and FACSymphony, BD).
Cytokine and Chemokine DetectionCytokines and chemokines were measured by ELISA and custom
Bio-Plex assays (see Supplementary Methods for details).
Generation of CIITA Knockout CellsL-428 cell lines were transduced with lentivirus expressing guide
sequence against CIITA to generate CIITA knockout cells that abro-gate the expression of MHC class II (Supplementary Fig. S19A–S19B; see Supplementary Methods for details). MHC class II expression was evaluated by staining the cells with FITC-HLA DR/DP/DQ antibodies (1:100, BD Biosciences #555558) and analyzed using the BD LSRFortessa. Subsequently, CIITA knockout cells were sorted by mCherry+, HLA DR−/DP−/DQ−, and DAPI− using the BD FACSAria Fusion sorter.
In Vitro HRS Cell and T-cell Coculture AssayWe purified CD4+ LAG3+ T cells from HLA class II–matched (to
L-428) PBMCs as described above. CD4+ LAG3+ T cells were cocul-tured with either CIITA wild-type or CIITA knockout L-428 at 2:1 ratio in a 96-well plate.
Survival AnalysisOS (death from any cause), DSS (the time from initial diagnosis
to death from lymphoma or its treatment, with data for patients who died of unrelated causes censored at the time of death), and post–autologous stem-cell transplant failure-free survival (time from ASCT treatment to cHL progression, or death from cHL) were analyzed using the Kaplan–Meier method and results were compared using the log-rank test. Univariate and multivariate Cox regression analyses were performed to assess the effects of prog-nostic factors. Survival analyses were performed in the R statistical environment (v3.5.2).
Statistical Results and VisualizationAll t tests reported are two-sided Student t tests, and P values
< 0.05 were considered to be statistically significant. In all box plots, boxes represent the interquartile range with a horizontal line indi-cating the median value. Whiskers extend to the farthest data point within a maximum of 1.5 × the interquartile range, and colored dots represent outliers.
Data AvailabilityscRNA-seq BAM files (generated with CellRanger v2.1.0) are depos-
ited in the European Genome-phenome Archive (EGAS00001004085) and are available by request. The figures associated with the above raw data sets are Figs. 1–4 and Supplementary Figs. S1–S10.
Code AvailabilityScripts used for data analysis are available upon request.
Disclosure of Potential Conflicts of InterestK.J. Savage is a consultant for Seattle Genetics, BMS, Merck,
Verastem, AbbVie, Servier, and AstraZeneca, and reports receiving commercial research support from Roche. S.P. Shah is a consultant for Contextual Genomics Inc. C. Steidl is an advisory board member for Curis Inc., AbbVie, Seattle Genetics, and Roche, reports receiving commercial research grants from Bristol-Myers Squibb and Trillium Therapeutics, and has received other remuneration from Bayer and Juno Therapeutics. No potential conflicts of interest were disclosed by the other authors.
Authors’ ContributionsConception and design: T. Aoki, X. Wang, A.P. Weng, B.H. Nelson, A. Mottok, S.P. Shah, C. SteidlDevelopment of methodology: T. Aoki, L.C. Chong, K. Takata, K. Milne, E.A. Chavez, M. Nissen, X. Wang, A. Telenius, C. Ghesquiere, J. Veldman, J. Kim, A. Mottok, A. Merchant, C. Steidl
Aoki et al.RESEARCH ARTICLE
420 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Aoki, K. Takata, K. Milne, M. Hav, M. Nissen, T. Miyata-Takata, V. Lam, E. Viganò, M.Y. Li, S. Healy, C. Ghesquiere, D. Kos, T. Goodyear, J. Kim, P. Farinha, K.J. Savage, D.W. Scott, G. Krystal, B.H. Nelson, A. Mottok, C. SteidlAnalysis and interpretation of data (e.g., statistical analysis, bio-statistics, computational analysis): T. Aoki, L.C. Chong, K. Takata, K. Milne, M. Hav, A. Colombo, E.A. Chavez, M. Nissen, X. Wang, T. Miyata-Takata, E. Viganò, M.Y. Li, A.W. Zhang, S. Saberi, J. Ding, A.P. Weng, B.H. Nelson, A. Mottok, A. Merchant, C. SteidlWriting, review, and/or revision of the manuscript: T. Aoki, L.C. Chong, K. Takata, M. Hav, A. Colombo, T. Miyata-Takata, V. Lam, J. Veldman, A.W. Zhang, K.J. Savage, D.W. Scott, B.H. Nelson, A. Mot-tok, A. Merchant, C. SteidlAdministrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L.C. Chong, K. Takata, B.W. Woolcock, A. Telenius, P. Farinha, D.W. Scott, C. SteidlStudy supervision: D.W. Scott, S.P. Shah, C. Steidl
AcknowledgmentsThis study was funded by a research grant from the Canadian Cancer
Society Research Institute and a Paul Allen Distinguished Investigator award (Frontiers Group; C. Steidl). This study was also supported by a Program Project Grant from the Terry Fox Research Institute (C. Steidl, grant no. 1061), Genome Canada, Genome British Columbia, Canadian Institutes of Health Research, and the British Columbia Cancer Foun-dation. T. Aoki was supported by fellowships from the Japanese Society for the Promotion of Science and the Uehara Memorial Foundation. T. Aoki received research funding support from The Kanae Founda-tion for the Promotion of Medical Science. E. Viganò is supported by a Michael Smith Foundation for Health Research trainee award.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received June 12, 2019; revised November 13, 2019; accepted December 13, 2019; published first December 19, 2019.
REFERENCES1. Swerdlow SH, Campo E, Harris NL, Jaffe ES, Pileri SA, Stein H, et al.
WHO classification of tumours of haematopoietic and lymphoid tis-sues. Revised 4th ed. Lyon, France: International Agency for Research on Cancer (IARC); 2017.
2. Mottok A, Steidl C. Biology of classical Hodgkin lymphoma: implica-tions for prognosis and novel therapies. Blood 2018;131:1654–65.
3. Aoki T, Steidl C. Novel biomarker approaches in classic Hodgkin lymphoma. Cancer J 2018;24:206–14.
4. Chan FC, Mottok A, Gerrie AS, Power M, Nijland M, Diepstra A, et al. Prognostic model to predict post-autologous stem-cell trans-plantation outcomes in classical Hodgkin lymphoma. J Clin Oncol 2017;35:3722–33.
5. Steidl C, Shah SP, Woolcock BW, Rui L, Kawahara M, Farinha P, et al. MHC class II transactivator CIITA is a recurrent gene fusion partner in lymphoid cancers. Nature 2011;471:377–81.
6. Steidl C, Lee T, Shah SP, Farinha P, Han G, Nayar T, et al. Tumor-associated macrophages and survival in classic Hodgkin’s lymphoma. N Engl J Med 2010;362:875–85.
7. Chen R, Zinzani PL, Fanale MA, Armand P, Johnson NA, Brice P, et al. Phase II study of the efficacy and safety of pembrolizumab for relapsed/refractory classic Hodgkin lymphoma. J Clin Oncol 2017;35:2125–32.
8. Armand P, Engert A, Younes A, Fanale M, Santoro A, Zinzani PL, et al. Nivolumab for relapsed/refractory classic Hodgkin lymphoma after failure of autologous hematopoietic cell transplantation: extended
follow-up of the multicohort single-arm phase II CheckMate 205 trial. J Clin Oncol 2018;36:1428–39.
9. Younes A, Gopal AK, Smith SE, Ansell SM, Rosenblatt JD, Savage KJ, et al. Results of a pivotal phase II study of brentuximab vedotin for patients with relapsed or refractory Hodgkin’s lymphoma. J Clin Oncol 2012;30:2183–9.
10. Cader FZ, Schackmann RCJ, Hu X, Wienand K, Redd R, Chapuy B, et al. Mass cytometry of Hodgkin lymphoma reveals a CD4(+) regula-tory T-cell-rich and exhausted T-effector microenvironment. Blood 2018;132:825–36.
11. Greaves P, Clear A, Owen A, Iqbal S, Lee A, Matthews J, et al. Defining characteristics of classical Hodgkin lymphoma microenvironment T-helper cells. Blood 2013;122:2856–63.
12. Patel SS, Weirather JL, Lipschitz M, Lako A, Chen PH, Griffin GK, et al. The microenvironmental niche in classic Hodgkin lymphoma is enriched for CTLA-4-positive T-cells that are PD-1-negative. Blood 2019;134:2059–69.
13. Azizi E, Carr AJ, Plitas G, Cornish AE, Konopacki C, Prabhakaran S, et al. Single-cell map of diverse immune phenotypes in the breast tumor microenvironment. Cell 2018;174:1293–308.
14. Han A, Glanville J, Hansmann L, Davis MM. Linking T-cell receptor sequence to functional phenotype at the single-cell level. Nat Biotech-nol 2014;32:684–92.
15. Zilionis R, Nainys J, Veres A, Savova V, Zemmour D, Klein AM, et al. Single-cell barcoding and sequencing using droplet microfluidics. Nat Protoc 2017;12:44–73.
16. Klein AM, Mazutis L, Akartuna I, Tallapragada N, Veres A, Li V, et al. Droplet barcoding for single-cell transcriptomics applied to embry-onic stem cells. Cell 2015;161:1187–201.
17. Novershtern N, Subramanian A, Lawton LN, Mak RH, Haining WN, McConkey ME, et al. Densely interconnected transcriptional circuits control cell states in human hematopoiesis. Cell 2011;144:296–309.
18. Netea MG, Azam T, Ferwerda G, Girardin SE, Walsh M, Park JS, et al. IL-32 synergizes with nucleotide oligomerization domain (NOD) 1 and NOD2 ligands for IL-1beta and IL-6 production through a caspase 1-dependent mechanism. Proc Natl Acad Sci U S A 2005;102:16309–14.
19. Schmitz R, Stanelle J, Hansmann ML, Kuppers R. Pathogenesis of classical and lymphocyte-predominant Hodgkin lymphoma. Annu Rev Pathol 2009;4:151–74.
20. Morales O, Mrizak D, Francois V, Mustapha R, Miroux C, Depil S, et al. Epstein-Barr virus infection induces an increase of T regulatory type 1 cells in Hodgkin lymphoma patients. Br J Haematol 2014;166:875–90.
21. Gandhi MK, Lambley E, Duraiswamy J, Dua U, Smith C, Elliott S, et al. Expression of LAG-3 by tumor-infiltrating lymphocytes is coincident with the suppression of latent membrane antigen-spe-cific CD8+ T-cell function in Hodgkin lymphoma patients. Blood 2006;108:2280–9.
22. Gagliani N, Magnani CF, Huber S, Gianolini ME, Pala M, Licona-Limon P, et al. Coexpression of CD49b and LAG-3 identifies human and mouse T regulatory type 1 cells. Nat Med 2013;19:739–46.
23. Andrews LP, Marciscano AE, Drake CG, Vignali DA. LAG3 (CD223) as a cancer immunotherapy target. Immunol Rev 2017;276:80–96.
24. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 2014;515:568–71.
25. Haghverdi L, Buettner F, Theis FJ. Diffusion maps for high-dimensional single-cell analysis of differentiation data. Bioinformatics 2015;31: 2989–98.
26. Coifman RR, Lafon S, Lee AB, Maggioni M, Nadler B, Warner F, et al. Geometric diffusions as a tool for harmonic analysis and structure definition of data: diffusion maps. Proc Natl Acad Sci U S A 2005;102: 7426–31.
27. Huang CT, Workman CJ, Flies D, Pan X, Marson AL, Zhou G, et al. Role of LAG-3 in regulatory T cells. Immunity 2004;21:503–13.
28. Bacchetta R, Sartirana C, Levings MK, Bordignon C, Narula S, Ron-carolo MG. Growth and expansion of human T regulatory type 1 cells are independent from TCR activation but require exogenous cytokines. Eur J Immunol 2002;32:2237–45.
Single-Cell Characterization of Hodgkin Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 421
29. Groux H, O’Garra A, Bigler M, Rouleau M, Antonenko S, de Vries JE, et al. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 1997;389:737–42.
30. Skinnider BF, Mak TW. The role of cytokines in classical Hodgkin lymphoma. Blood 2002;99:4283–97.
31. Lin Y, Xu L, Jin H, Zhong Y, Di J, Lin QD. CXCL12 enhances exog-enous CD4+CD25+ T cell migration and prevents embryo loss in non-obese diabetic mice. Fertil Steril 2009;91:2687–96.
32. McFadden C, Morgan R, Rahangdale S, Green D, Yamasaki H, Center D, et al. Preferential migration of T regulatory cells induced by IL-16. J Immunol 2007;179:6439–45.
33. Wang X, Lang M, Zhao T, Feng X, Zheng C, Huang C, et al. Cancer-FOXP3 directly activated CCL5 to recruit FOXP3(+)Treg cells in pancreatic ductal adenocarcinoma. Oncogene 2017;36: 3048–58.
34. Pierini A, Strober W, Moffett C, Baker J, Nishikii H, Alvarez M, et al. TNF-alpha priming enhances CD4+FoxP3+ regulatory T-cell sup-pressive function in murine GVHD prevention and treatment. Blood 2016;128:866–71.
35. Tran DQ. TGF-beta: the sword, the wand, and the shield of FOXP3(+) regulatory T cells. J Mol Cell Biol 2012;4:29–37.
36. Gobert M, Treilleux I, Bendriss-Vermare N, Bachelot T, Goddard-Leon S, Arfi V, et al. Regulatory T cells recruited through CCL22/CCR4 are selectively activated in lymphoid infiltrates surrounding primary breast tumors and lead to an adverse clinical outcome. Cancer Res 2009;69:2000–9.
37. Mizukami Y, Kono K, Kawaguchi Y, Akaike H, Kamimura K, Sugai H, et al. CCL17 and CCL22 chemokines within tumor microenvironment are related to accumulation of Foxp3+ regulatory T cells in gastric cancer. Int J Cancer 2008;122:2286–93.
38. Jin JO, Han X, Yu Q. Interleukin-6 induces the generation of IL-10-producing Tr1 cells and suppresses autoimmune tissue inflamma-tion. J Autoimmun 2013;40:28–44.
39. Steidl C, Diepstra A, Lee T, Chan FC, Farinha P, Tan K, et al. Gene expression profiling of microdissected Hodgkin Reed-Sternberg cells
correlates with treatment outcome in classical Hodgkin lymphoma. Blood 2012;120:3530–40.
40. Huard B, Prigent P, Pages F, Bruniquel D, Triebel F. T cell major histo-compatibility complex class II molecules down-regulate CD4+ T cell clone responses following LAG-3 binding. Eur J Immunol 1996;26:1180–6.
41. Baixeras E, Huard B, Miossec C, Jitsukawa S, Martin M, Hercend T, et al. Characterization of the lymphocyte activation gene 3-encoded protein. A new ligand for human leukocyte antigen class II antigens. J Exp Med 1992;176:327–37.
42. Duffield AS, Ascierto ML, Anders RA, Taube JM, Meeker AK, Chen S, et al. Th17 immune microenvironment in Epstein-Barr virus-negative Hodgkin lymphoma: implications for immunotherapy. Blood Adv 2017;1:1324–34.
43. Roemer MGM, Redd RA, Cader FZ, Pak CJ, Abdelrahman S, Ouyang J, et al. Major histocompatibility complex class II and programmed death ligand 1 expression predict outcome after programmed death 1 blockade in classic Hodgkin lymphoma. J Clin Oncol 2018;36:942–50.
44. Zhang B, Chikuma S, Hori S, Fagarasan S, Honjo T. Nonoverlapping roles of PD-1 and FoxP3 in maintaining immune tolerance in a novel autoimmune pancreatitis mouse model. Proc Natl Acad Sci U S A 2016;113:8490–5.
45. Wang J, Sanmamed MF, Datar I, Su TT, Ji L, Sun J, et al. Fibrinogen-like protein 1 is a major immune inhibitory ligand of LAG-3. Cell 2019;176:334–47.
46. Xu F, Liu J, Liu D, Liu B, Wang M, Hu Z, et al. LSECtin expressed on melanoma cells promotes tumor progression by inhibiting antitumor T-cell responses. Cancer Res 2014;74:3418–28.
47. Lun AT, Bach K, Marioni JC. Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol 2016;17:75.
48. Levine JH, Simonds EF, Bendall SC, Davis KL, Amir el AD, Tadmor MD, et al. Data-driven phenotypic dissection of AML reveals progen-itor-like cells that correlate with prognosis. Cell 2015;162:184–97.
49. Catena R, Montuenga LM, Bodenmiller B. Ruthenium counterstain-ing for imaging mass cytometry. J Pathol 2018;244:479–84.
MOLECULAR CANCER RESEARCH | TUMOR MICROENVIRONMENTAND IMMUNOBIOLOGY
The Th9 Axis Reduces the Oxidative Stress and Promotesthe Survival of Malignant T Cells in Cutaneous T-CellLymphoma PatientsSushant Kumar1, Bhavuk Dhamija1, Soumitra Marathe1, Sarbari Ghosh1, Alka Dwivedi1, Atharva Karulkar1,Neha Sharma2, Manju Sengar2, Epari Sridhar3, Avinash Bonda2, Jayashree Thorat2, Prashant Tembhare3,Tanuja Shet3, Sumeet Gujral3, Bhausaheb Bagal2, Siddhartha Laskar4, Hasmukh Jain2, and Rahul Purwar1
ABSTRACT◥
Immune dysfunction is critical in pathogenesis of cutaneousT-cell lymphoma (CTCL). Few studies have reported abnormalcytokine profile and dysregulated T-cell functions during theonset and progression of certain types of lymphoma. However,the presence of IL9-producing Th9 cells and their role in tumorcell metabolism and survival remain unexplored. With thisclinical study, we performed multidimensional blood endotyp-ing of CTCL patients before and after standard photo/chemo-therapy and revealed distinct immune hallmarks of the disease.Importantly, there was a higher frequency of “skin homing”Th9 cells in CTCL patients with early (T1 and T2) andadvanced-stage disease (T3 and T4). However, advanced-stage CTCL patients had severely impaired frequencyof skin-homing Th1 and Th17 cells, indicating attenuatedimmunity. Treatment of CTCL patients with standard photo/
chemotherapy decreased the skin-homing Th9 cells andincreased the Th1 and Th17 cells. Interestingly, T cells ofCTCL patients express IL9 receptor (IL9R), and there wasnegligible IL9R expression on T cells of healthy donors. Mech-anistically, IL9/IL9R interaction on CD3þ T cells of CTCLpatients and Jurkat cells reduced oxidative stress, lactic acido-sis, and apoptosis and ultimately increased their survival. Inconclusion, coexpression of IL9 and IL9R on T cells in CTCLpatients indicates the autocrine-positive feedback loop of Th9axis in promoting the survival of malignant T cells by reducingthe oxidative stress.
Implications: The critical role of Th9 axis in CTCL pathogenesisindicates that strategies targeting Th9 cells might harbor signif-icant potential in developing robust CTCL therapy.
IntroductionCutaneous T-cell lymphoma (CTCL) represents a diverse group of
non-Hodgkin lymphomas derived frommature malignant T cells thattraffic to the human skin (1, 2). On the basis of Surveillance Epide-miology and EndResults registry data, the incidence of CTCL is 6.4 permillion individual in the United States and the highest incidence hasbeen reported among African-Americans (3). According to theWorldHealth Organization/European Organization for Research and Treat-ment of Cancer (WHO-EORTC) classification, primary CTCL includemultiple variants with distinct clinical manifestations (4). The path-ogenesis of CTCL is not clear, however, the immune dysregulation isbelieved to have a vital role in the pathogenesis of lymphoma (5–7).Few studies reported the abnormal production of various cytokinesduring the onset of certain lymphomas (8).
IL9 is a pleiotropic cytokine and IL9/IL9 receptor (IL9R) interactionpromotes T-cell growth (9) and is responsible for diverse functions ininflammatory and immune responses (10, 11). Multiple cell typesincluding Th2, Th17, regulatory T cells, mast cells, dendritic cells, andnatural killer (NK)/T cell have been reported to secrete IL9 (12–16).Later, IL9-producing Th9 cells, a distinct Th subset was described. TheTh9 cells are reported to be “skin tropic” as a majority of Th9 cellsexpress cutaneous lymphocyte antigen (CLA; ref. 17) and their pres-ence was demonstrated in the patients with various skin diseasesincluding psoriasis and atopic dermatitis (17, 18). Increased levels ofIL9 were demonstrated in biopsies or sera in multiple types ofperipheral T-cell lymphomas including adult T-cell leukemia, ana-plastic large-cell lymphoma (ALCL), Hodgkin lymphoma, and nasalNK/T-cell lymphoma patients (19–24). A recent study reported thepresence of CD3þ IL9-producing T cells in the dermis and dermo-epidermal junction in mycosis fungoides skin lesions (19). However,the presence of Th9 cells as well as IL9 production bymalignant T cellsin patients with early- and advanced-stage CTCL have not beenreported so far.
The role of IL9 in tumor development depends on multiple factorsincluding the nature of tumor and tumor microenvironment. We andothers have reported the antitumor role of Th9 cells in amurine modelof melanoma, which were mediated by promoting CD8þ cytotoxicT cells and mast cell functions (25, 26). Unlike melanoma, theprotumor activity of IL9 has been reported in hematologic cancersusing murine models or cell lines. For example, transgenic miceoverexpressing IL9 have been shown to develop thymic lymphomaat the age of 3–9 months (27). Moreover, IL9 stimulation of mousethymic lymphoma cells induced by N-methyl-N-nitrosourea or byX-ray irradiation showed increased proliferative responses (28). IL9promoted the proliferation of human Hodgkin lymphoma cell lines in
1Department of Biosciences & Bioengineering, Indian Institute of TechnologyBombay, Mumbai, Maharashtra, India. 2Medical oncology, Tata Memorial Hos-pital, Mumbai, Maharashtra, India. 3Pathology, Tata Memorial Hospital, Mumbai,Maharashtra, India. 4Radiation Oncology, Tata Memorial Hospital, Mumbai,Maharashtra, India.
Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).
Corresponding Author: Rahul Purwar, Indian Institute of Technology Bombay,Lab302, BSBE, IIT Bombay, Powai, Mumbai, Maharashtra 400076, India.Phone: 9122-2576-7737; Fax: 9122-2572-7760; E-mail: [email protected]
Mol Cancer Res 2020;18:657–68
doi: 10.1158/1541-7786.MCR-19-0894
�2020 American Association for Cancer Research.
AACRJournals.org | 657
a dose-dependent manner, and the ablation of IL9 by IL9-specificantisense oligomer inhibited the proliferation of Hodgkin lymphomacells (29). In primary T-cell lymphoma mouse models, IL9/IL9Rinteraction activated STAT proteins and contributed to in vivo growthof tumor (30–32). However, the roles of IL9 in malignant T cells’survival in CTCL patients have not been probed.
We performed blood endotyping of early- and advanced-stage CTCLpatients [mycosis fungoides, CD30þ lymphoproliferative disease(CD30þLPD), subcutaneouspanniculitis-like T-cell lymphoma (SPTL),and CTCL, not otherwise specified (NOS)] focusing on three majorcomponents of “skin-homing” and “systemic”T-cell–mediated immuneresponses: (i) T-cell cytokine profile, (ii) T-cell activation, and (iii) T-cellsubsets. We demonstrated the increased frequency of “skin-homing”Th9 cells in early as well as advanced-stage CTCL. Interestingly, weobserved the coexpressionof IL9 and IL9RonmalignantT cells of CTCLpatients. Functionally, IL9/IL9R interaction reduced the toxic reactiveoxygen species (ROS) production, lactic acidosis, and apoptosis, andultimately increased the survival ofmalignant T cells. In conclusion, Th9axis promotes malignant T-cell survival by reducing the oxidative stressof T cells in CTCL patients.
Materials and MethodsStudy protocol and blood samples
The study protocol was approved by the Institute Ethics Committeeof Tata Memorial Hospital and IIT Bombay and is in compliance withthe Declaration of Helsinki. The clinical trial (CTRI/2017/04/008356)is registered in the Clinical Trial Registry of India. In this study werecruited primary CTCL patients, mainly mycosis fungoides, CD30þ
LPD, SPTL, and CTCL-NOS. Table 1 describes the patient's details,risk status, symptoms, clinical examination, staging and treatmenthistory of CTCL patients at the baseline. In brief, de novo andpreviously treated adult patients (≥18 years) suffering from CTCL(n ¼ 17) were recruited, provided they were not on any activeimmunosuppression at least 4 weeks prior to the sampling. Amongthese, 6 patients of CTCL were enrolled as a follow-up study afterstandard photo/chemotherapy. The treatment details and patientinformation of follow-up patients are described in SupplementaryTable S1. CTCL diagnoses were established in accordance with theWHO-EORTC classification. Good clinical practice guidelines werefollowed and written consent was obtained from all patients for bloodsample collection. Whole blood was collected from 17 CTCL patients(10 early stage and 7 advanced-stage patients) and 19 healthy donors inBD Vacutainer EDTA tubes.
Surface marker and intracellular cytokine analysis by flowcytometry
The peripheral blood mononuclear cell (PBMC) from peripheralblood were isolated by ficoll-histopaque (Sigma-Aldrich; 10771) den-sity gradient centrifugation method. Cells (1 � 106 cells/tube) werewashed in FACS staining buffer (2% FBS in PBS) and resuspended in50 mL staining buffer. The fluorochrome-conjugated surface antibo-dies (CD3, CD4, CD8, CLA, CCR7, L-selectin, CD45RA, CD45RO,CD25, and CD69) were added in different combinations in tubes andmixed gently by tapping. Similarly, fluorochrome-conjugated IL9Rantibody was added to Jurkat cells (Source: National Centre for CellScience, Pune, India) and sorted T cells. All tubes were incubated for40 minutes in dark on ice and then washed twice with staining bufferand then resuspended in 250 mL staining buffer for acquisition by BDFACSVerse or BDFACSAria Fusion III (BDBiosciences) and analyseswere done using BD FACSuite software.
For intracellular staining, the PBMCs were cultured in T-cell media(Iscove's modified Dulbecco's mediumwith 10% heat-inactivated FBS,1% penicillin/streptomycin, and 1% L-glutamine). PBMCs were stim-ulated in 1 mL of T-cell media in 24-well plate with PMA (Merck;P8139, 10 ng/mL) and Ionomycin (Invitrogen; I24222, 500 ng/mL) inthe presence of Brefeldin-A (Golgi Plug- BD Biosciences; 555028,0.75 mL/mL) for 5 hours at 37�C. The PBMCs were first stained withfluorochrome-conjugated surface antibodies (CD4 andCLA) and thensurface-stained cells were washed twice with staining buffer and cellswere resuspended in 250 mL 1� Cytofix/Cytoperm Buffer (BD Bios-ciences; 555028), mixed gently, and incubated at room temperature for20 minutes. After incubation, cells were washed twice with 500 mLperm/wash buffer. Fluorochrome-conjugated antibodies for intracel-lular marker (IL9, IL4, IL17, and IFNg) were added in differentcombinations in the respective tubes and mixed gently by tapping inperm/wash buffer (total volume not exceeding 100 mL). All tubes wereincubated for 30 minutes in dark at room temperature and thenwashed twice with perm/wash buffer and resuspended in 250 mLstaining buffer for acquisition by BD FACSVerse or BD FACS AriaFusion III (BD Biosciences) and the analyses were done using BDFACSuite software.
All antibodies were procured from BD Biosciences, BioLegend, andThermo Fisher Scientific.
Cytokine analysis by ELISAIn T-cell receptor (TCR) stimulation (recall assay), the PBMCswere
cultured in T-cell media and were stimulated using anti-CD3/CD28Dynabeads Human T-cell activator (Bead: cell ratio ¼ 1:1, Gibco;11131D) and incubated for 48 hours. Post 48 hours, IL9 productionwas quantified in the 100 mL of cell-free supernatant using the humanIL9 ELISA Kit (Invitrogen; 88-7958-88) as per the manufacturer'sinstructions.
T-cell sortingThe PBMCs (1 � 106 cells/tube) were washed using FACS staining
buffer (2% FBS in PBS) and resuspended in 50 mL staining buffer. Thefluorochrome-conjugated CD3 antibodies were added in tubes andmixed gently by tapping. The tubes were incubated for 40 minutes indark on ice and then washed twice with staining buffer and thenresuspended in 250 mL staining buffer for sorting by BD FACS AriaFusion III (BD Biosciences) and purity analyses were done using BDFACSDiva software.
Cell survival analysisJurkat cells were procured from National Centre for Cell Science,
Pune, India and Mycoplasma testing was performed using PCRMycoplasma Detection Kit (Applied Biological Materials Inc.;G238) as per the manufacturer's instruction. No additional cell lineauthentication was performed. Jurkat cells (0.1 million/mL) weresuspended in RPMI media with 10% FBS and 1% penicillin/strepto-mycin and seeded in a 24-well plate. Similarly, sorted T cells of CTCLpatients and healthy donors (0.1 million/mL) were seeded in T-cellmedia in a 24-well plate. Cells were cultured for 72 hours in thepresence and absence of cytokines (IL9: 20 ng/mL and IFNg : 2 ng/mL).Post incubation, cells were harvested and stained for trypan blue, andlive cells were counted.
Flow cytometric apoptosis assayA total of 0.1 million/mL Jurkat cells were seeded in 12-well plates
and cultured in 5%CO2 incubator at 37�C for 72 hours in the presenceand absence of cytokines (IL9, 20 ng/mL and IFNg , 2 ng/mL). At the
Kumar et al.
Mol Cancer Res; 18(4) April 2020 MOLECULAR CANCER RESEARCH658
Table
1.Patientsinform
ation.
S.No
Age
Sex
Diagno
sis
Stag
eRiskstatus
Symptoms
Clin
ical
exam
ination
Stag
ing
Trea
tmen
thistory
127
MMF
CTCLea
rly
stag
eT2A
N0M0,
stag
eIB
Lymphno
desw
ellingan
drash
PS-1,no
lympha
den
opathy
,no
organ
omeg
aly
PETCTscan
,nono
dal
disea
se;
bone
marrow,notdone
;CSF,
notdone
PUVA
223
FMF
T2N0M0,stage
IB
Pruritus
PS-1,no
lympha
den
opathy
,no
organ
omeg
aly,
skin;p
atches
ove
rtrun
k,thighs,u
pper
limbs
CTscan
,multiple
subcm
retroperitone
alpositive
;bone
marrow,notdone
;CSF,
notdone
PUVA
350
MMF
T2N0M0,stage
IB
Excessive
dryne
ss,o
ccasiona
litching
PS-1,no
lympha
den
opathy
,no
organ
omeg
aly.
Nostag
ingdone
.Biopsy
showed
noev
iden
ceof
lympho
ma
Offprotoco
l.Referred
todermatologistan
dcardiologistopinion.
453
MMF
T2N0M0B0,
stag
eIB
Erythem
atous
plaque
sonchest
abdomen
andthighs
present.
PS-1,n
olympha
den
opathy
,no
organ
omeg
aly.
PET/C
T,n
otdone
;bone
marrow,uninv
olved
;CSF,not
done
Onobservation
561
MMF
T2N0M0B0,
stag
eIB
Itching,m
acules
inbilateralh
ands,
napeofne
ck,a
nteriorchestwall,
andforehe
ad.
PS-1,n
olympha
den
opathy
,no
organ
omeg
aly.
USGab
domen
,nodisea
se;
bone
marrow,u
ninv
olved
;CSF,n
otdone
TSET
631
FMF
T1N0M0,stage
IA
Hyp
opigmen
tedpatches
ontrun
k,lim
b,a
ndab
domen
PS-1,n
olympha
den
opathy
,no
organ
omeg
aly.
CTscan
,norm
al,b
one
marrow,
notdone
;CSF,u
ninv
olved
Referredto
dermaun
it
755
MMF
T1A
N0M0stag
eIA
Swellingin
right
forearm
PS-1,n
olympha
den
opathy
,no
organ
omeg
aly
CT,right
paratrachea
lan
dsubcarina
llymphno
des.B
M,
likelyto
beun
invo
lved
.CSF,
notdone
Observation
840
FMF
T2A
N0M0stag
eIB
Hyp
erpigmen
tedpatch
seen
inthe
leftbreast,erythe
matous
patch
seen
ontheleftchestwall,
dep
igmen
tedpatch
seen
onthe
lower
abdominal
wall,legs
(poplitea
lfossa)
PS1,no
periphe
rallym
pha
den
opathy
,no
hepatospleno
meg
alychest
clea
r.Macular
lesions
>10%
BSA
PET/C
T,n
otdone
;bone
marrow,n
otdone
;CSF,n
ot
done
;USGab
domen
,no
nodes.
Patient
istaking
PUVA
therap
y
932
FMF
HPRno
tsuggestive
of
MF
White
patch,n
onitchy
and
nonp
ainful
Exten
sive
macular,scaly
lesionall
ove
rthe
bodyinvo
lvingchest,back,
abdomen
,and
B/L
upper
limban
dlower
limb.N
olympha
den
opathy
.Noorgan
omeg
aly.
PECT/C
T,n
otdone
;USG
abdomen
pelvis,no
abno
rmalitydetected;bone
marrow,n
otdone
;HPRno
tsuggestive
ofMF.
Patient
isun
der
observation
1055
FMF
T2A
N0M0,
stag
eIB
—PS-1,n
olympha
den
opathy
,no
organ
omeg
aly
PETscan
,B/L
cervicalno
des,B/
Laxillaryno
des,externa
liliac
nodes,B
/Linguina
lnodes
Che
motherap
y,gem
citabine.
Cha
nged
toCEOP
aftercycleof
gem
citabine.
(Con
tinu
edon
thefollowingpag
e)
Th9 Axis Promotes T-Cell Survival in CTCL
AACRJournals.org Mol Cancer Res; 18(4) April 2020 659
Table
1.Patientsinform
ation.
(Cont'd)
S.No
Age
Sex
Diagno
sis
Stag
eRiskstatus
Symptoms
Clin
ical
exam
ination
Stag
ing
Trea
tmen
thistory
1168
FMF
CTCL
advanced
stag
e
T3NxM0stag
eIIB
Red
patches
inB/L
upper
limb,
abdomen
,che
st,lower
limbs,an
dno
dular
swellingin
lower
abdomen
andright
inguina
lregion
PS-3,n
olympha
den
opathy
,no
organ
omeg
aly
Notdone
TSET
1255
MMF
T3N0M0B0,
stag
eIIB
Multiple
maculopap
ular
lesions,
lesiononleftsideofthelip
with
desqua
mationan
dserous
discharge.
PS-1,no
lympha
den
opathy
,no
organ
omeg
aly
PET,n
otavailable;m
arrow,
uninvo
lved
;CSF,n
otdone
Gem
citabinepallinten
t
1357
FMF
T4NxM0B0,
extensive
cutane
ous
disea
se
Pruritus,fissuringwithdischargeall
ove
rthebody,
associated
with
redne
ss,itching
,and
flaking.
PS-2,a
xillary
lympha
den
opathy
,diffuse
erythroderma,
andscaling
PET,n
otavailable;b
one
marrow,n
otavailable;C
SF,
notavailable
INFplusMTXan
dRT
opiniononTSET
1431
FSPTL
T3B
N0M0
Fev
er,sub
cutane
ous
edem
aove
rB/
Lup
per
andlower
limbs,multiple
nodules
palpab
leove
rleft
cervical,B
/L15
upper
limbs,B/L
lower
limb,b
ack,
andan
orexia
PS-2,n
olympha
den
opathy
,spleno
meg
aly
PET-B/L
axillaryno
des,B
/Lexternal
iliac
nodes,B
/Linguina
lnodes,
retroperitone
alno
des,b
one
marrow,uninv
olved
;CSF,not
done
6cycles
CHOPfrom
Jul7,
2018
1563
MCTCL:NOS
T3N0M0B0,
stag
eIIB
Ulcerativelesiononright
axillaan
dreddishno
dular
lesionove
rleft
lower
limb
PS-1,no
lympha
den
opathy
,no
organ
omeg
aly
CTscan
,right
axillaryregion;
distalleftthigh,
leftinferior
thyroid,m
esen
tericno
des
OralM
TX
1649
FCD30
þLP
DT4N1M0,stage
IIIA
Fev
er,w
eight
loss,p
ruritus,lymph
nodesw
elling
PS-1,lympha
den
opathy
B/L
axillary
nodes;n
oorgan
omeg
aly,
skin;
diffuse
erythroderma,
extensive
scaling,a
ndplaque
s
PETCTscan
,B/L
axillaryno
des:
2.4�
1.4cm
,ing
uina
lno
de,
bone
marrow,notdone
;CSF,
notdone
Che
motherap
y-metho
trexate
1745
MCutan
eous
ALC
LT3N0M0,stage
IIB
Skinlesions
onchestwall
PS-0,n
olympha
den
opathy
,no
organ
omeg
aly.
CT/PET,n
otdone
;bone
marrow,n
otdone
;CSF,n
ot
done
Observation
Abbreviations:ALC
L,an
aplastic
largecelllympho
ma;
B/L,bilateral;CSF,cerebrospinal
fluid;CD30
þLP
D,CD30
þlympho
proliferativedisea
se;HPR,histopatho
logyreport;INF,interferon;
MF,mycosisfung
oides;
MTX,m
etho
trexate;
NOS,n
ototherwisespecified
;PUVA,p
soralenan
dultravioletA;SPTL,
subcutane
ous
pan
niculitis-likeT-celllym
pho
ma;
TSET,totalskin
electrontherap
y.
Kumar et al.
Mol Cancer Res; 18(4) April 2020 MOLECULAR CANCER RESEARCH660
end of the incubation period, cells were harvested and washed withAnnexin V binding buffer (1�). The apoptotic cells were stained usingthe FITC Annexin V Apoptosis Detection Kit (BD Pharmingen;556547) as per the manufacturer's instructions and acquisition wasperformed by BD FACSVerse (BD Biosciences). The analyses of FACSdata were done using BD FACSuite software. 7ADD was used insteadof PE provided in the Apoptosis Detection Kit.
Measurement of oxidative stressMeasurement of intracellular ROS was carried out using 2,
7-dichlorodihydrofluorescein diacetate (H2DCF-DA), a dye that afterhydrolysis by intracellular esterases reacts with superoxide, hydroxyl,or oxygen radicals and forms fluorescent green product, dichloro-fluorescein (DCF). After 24-hour treatment of cytokines (IL9, 20 ng/mL and IFNg , 2 ng/mL), cells were collected, washed with PBS, andstained with 1 mmol/L H2DCFDA for 30 minutes at 37�C and ROSlevels were measured by BD FACSVerse or BD FACS Aria Fusion III(BD Biosciences).
Lactate measurement in supernatantJurkat cells and sorted T cells from patients and healthy individuals
were cultured in the presence and absence of cytokines (IL9, 20 ng/mLand IFNg , 2 ng/mL). Post 24 hours, the supernatant was collected andlactate was measured through the spectrofluorimetric method. Lactatedehydrogenase (1 U/mL) was used to convert lactate to pyruvate,further generating NADH, which was measured at 340 nm usingELISA plate reader. Incubation was carried out in 96-well ELISA platesfor 1 hour at 37�C. The pyruvate generated was trapped usinghydrazine present in glycine-hydrazine buffer (pH, 9.0), to preventreverse production of lactate.
Statistical analysis of individual immune featuresFor studies described in Figs. 1–6, comparative data were analyzed
by using Student t test. The specific statistical test chosen for eachexperiment is mentioned in the figure legend. Prism 7.02 software wasused to perform the statistical analysis. In the figures, �, P < 0.05; ��, P <0.01; ���, P < 0.001 and ns for nonsignificant.
Gate onCD4+CLA+ IL17− IL4−
IL17
A
Gate onIL4
CD4+CLA− IL17− IL4−
Skin homing
SystemicIL
9
IFNγ
5Th9cells
1.37
1.08Th1 cells
0.13Th9cells
3.48
0.27Th1 cells
IL9
IFNγ
IL17
A
IL4
Gatting strategyA
CD4CLA
CLA+ Th1 cells CLA+ Th2 cells CLA+ Th9 cells CLA+ Th17 cellsB
F G H I
Freq
uenc
y (%
)
50
40
30
20
10
0
Freq
uenc
y (%
)
50
40
30
20
10
0
CLA− Th1 cells CLA− Th2 cells CLA− Th9 cells CLA− Th17 cells
C D E
Healthy Total Early
CTCL
Advanced
Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced
Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced
Freq
uenc
y (%
)
10
8
6
4
2
0
Freq
uenc
y (%
)
8
6
4
2
0
Freq
uenc
y (%
)
4
3
2
1
0
Freq
uenc
y (%
)
25
20
15
10
5
0
Freq
uenc
y (%
)
15
10
5
0
Freq
uenc
y (%
)
50
40
30
20
10
0
Size
Gra
nula
rity
Figure 1.
Increased skin-homingTh9 cells inCTCLpatientswith early- andadvanced-stagedisease, humanPBMCsofCTCLpatients [early (T1 andT2): n¼ 10, advanceddisease(T3 and T4): n ¼ 7] and healthy individuals (n ¼ 19) were stimulated with PMA/ionomycin in presence of brefeldin-A and stained as described in Materials andMethods. A, Representative dot plots and gating strategy for identification of Th9 and Th1 cells are depicted. Cumulative data of skin homing (CLAþ, B–E) andsystemic (CLA�, F–I) Th9 cells, Th1 cells, Th2 cells, and Th17 cells are shown. Bar plots represent means� SEM. Statistical significance was determined as comparedwith healthy donors by Student unpaired t test (P values are designated as ��� , <0.001; ��, <0.01; � , <0.05).
Th9 Axis Promotes T-Cell Survival in CTCL
AACRJournals.org Mol Cancer Res; 18(4) April 2020 661
ResultsIncreased skin-homing Th9 cells in blood of CTCL patients ascompared with healthy controls
A large proportion of CLAþ skin-resident effector T cells are knownto secrete IL9 with distinct Th9 phenotype in healthy and inflamedskin (17). In this study, we quantified skin-homing (CLAþ) andsystemic (CLA�) Th1 cells (CD4þ IFNgþ IL9� IL4� IL17�), Th2cells (CD4þ IL4þ IFNg� IL9� IL17�), Th9 cells (CD4þ IL9þ IFNg�
IL4� IL17�), and Th17 cells (CD4þ IL17þ IL4� IFNg� IL-9�) inprimary CTCL patients (early and advanced). Figure 1A representsthe gating strategy to quantify the skin-homing (CLAþ) and systemic(CLA�) Th9 cells and Th1 cells. Similar strategies were used for gatingTh2 and Th17 cells. In early- as well as advanced-stage CTCL patients,there was an increased frequency of skin-homing (CLAþ) Th9 cells ascompared with healthy donors (Fig. 1D). However, higher numbers ofsystemic (CLA�) Th9 cells were found in advanced-stage CTCLpatients as compared with healthy donors (Fig. 1H).
There was a significantly lower frequency of skin-homing Th1cells (Fig. 1B) and skin-homing Th17 cells (Fig. 1E) in advanced-stage CTCL patients as compared with healthy donors. There wasno difference in the frequency of systemic Th1 cells (Fig. 1F), skin-homing and systemic Th2 cells (Fig. 1C and G), and systemic Th17cells (Fig. 1I) among healthy donors and CTCL patients. Collec-tively, this data suggest that advanced-stage CTCL patients haveattenuated skin immunity (Th1 and Th17 responses), and Th9 cellsare increased in early- and advanced-stage CTCL patients.
CTCL patient–derived T cells secrete increased levels of IL9 andexpress high levels of IL9R on their surface
Th9 cells are the major source of IL9, which signals through a gCfamily receptor (IL9R) on target cells. We examined whether patient-derived T cells secrete increased levels of IL9 upon T-cell activation.The IL9 production was quantified by ELISA in the cell-free culturesupernatant of PBMCs isolated from CTCL patients and healthysubjects upon TCR stimulation using anti-CD3/CD28–coated Dyna-beads. Similar to flow cytometry data, IL9 production was higher inCTCL patients as compared with healthy donors (Fig. 2A).
Next, we examined the surface expression of IL9R on malignantT cells. CD3þT cells were sorted fromCTCL patients as well as healthyindividuals.We observed a higher expression of IL9R onCTCL patientT cells and negligible expression of IL9R on healthy T cells (Fig. 2B).Similarly, there was increased IL9R expression on Jurkat cells (a T-celllymphoma cell line). This data collectively suggest the coexpression ofIL9 and IL9R expression on malignant T cells.
IL9 promotes the survival of T cells of CTCL patients and Jurkatcells
We further examined the functional relevance of the coexpression ofIL9 and IL9R on malignant T cells. IL9 promoted T-cell survival ofCTCL patients, Jurkat cells, and had no impact on healthy T cells(Fig. 3A). Because IFNg is known to be proapoptotic and antiproli-ferative (33) and has the ability to induce oxidative stress (34) invarious cancer cells, we also examined the role of IL9 in IFNg milieu tofurther strengthen our observation. IFNg inhibited the cell survival
Healthy Total Early
CTCL
Advanced
200
150
100
50
0
IL9
(pg/
mL)
/10
6 cel
ls
A
CTCL patients Healthy donors Jurkat cellsB
Coun
t
Coun
t
Coun
t
IsotypeIL9R
IsotypeIL9R
IsotypeIL9R
IL9R IL9R IL9R
IL9RIsotype IL9RIsotype IL9RIsotype
1,000
800
600
400
200
0Fluo
resc
ence
inte
nsity 120
115
110
105
100Fluo
resc
ence
inte
nsity 400
300
200
100
0Fluo
resc
ence
inte
nsityns
Figure 2.
T cells of patients with CTCL produce increased levels ofIL9 and express high levels of IL9R. A, Isolated PBMCsfrom healthy donors (n¼ 23) and patientswith CTCL (n¼11; 6 early and 5 advanced disease) were stimulated usinganti-CD3/CD28–coated Dynabeads (Bead to cell ratio,1:1). Post 48-hour incubation, IL9 production was quan-tified by ELISA in the cell-free supernatant as per themanufacturer's instruction. Bar plots represent means �SEM. Statistical significance was determined by compar-ing with healthy donors using Student unpaired t test. B,Sorted CD3þ T cells from CTCL patients (n ¼ 3), healthydonors (n ¼ 3), and Jurkat cells (n ¼ 3) were stained forIL9R and analyzed by flow cytometry. Representativehistogram (top) for CTCL, healthy, and Jurkat cells areshown and cumulative data are depicted in the bottompanel. Statistical significance was determined by pairedt test (P values are designated as � , <0.05; ns, notsignificant).
Kumar et al.
Mol Cancer Res; 18(4) April 2020 MOLECULAR CANCER RESEARCH662
and IL9 reversed the IFNg-mediated inhibition of cell survival inCTCL patients and Jurkat cells (Fig. 3A). Next, we examined themechanism of IL9-mediated increase in tumor cell survival. IL9reduced the Jurkat cell apoptosis as quantified by Annexin-V/7-AADstaining through flow cytometry (Fig. 3B).
IL9 reduces oxidative stress and lactic acidosis of T cells of CTCLpatients and Jurkat cells
Because reduced toxic ROS levels are described to promote healthyT-cell survival (35), we examined whether IL9/IL9R interactionimpacts metabolic alterations, which would be beneficial for thesurvival of malignant T cells. Sorted CD3þ T cells fromCTCL patientsand healthy donors were cultured in the presence and absence of IL9and/or IFNg and ROS levels were quantified byDCFDA staining usingflow cytometry. Interestingly, IL9 alone and/or in presence of IFNgsignificantly reduced the ROS levels inside the T cells of CTCL patientsand Jurkat cells; however there was no significant reduction in ROSlevels in healthy T cells (Fig. 4A).
Because lactic acidosis promotes ROS release in cancer cells andincreases oxidative stress inside the cells (36), we quantified extracel-
lular lactate in cell-free supernatant of CTCL patients, Jurkat cells, andhealthy donors. IL9 decreased the extracellular lactate concentration inT cells of CTCL patients and Jurkat cells, however, had no impact onlactate production in healthy donors (Fig. 4B). Collectively this datasuggest that IL9 promotes cell survival by reducing the intracellulartoxic ROS levels, lactic acidosis, and apoptosis.
T-cell subsets and immune activation profile of patients withCTCL and healthy donors
An earlier report shows increased migratory memory T cells (TMM)in fewCTCLpatients and depletion of TMM cells in both the circulationand the skin of CTCL patients treated with alemtuzumab (37). Veryrecently, TMM cells were demonstrated to be the connecting linkbetween skin and lymph nodes and critical to the pathogenesis ofL-CTCL, a malignancy of central memory T cells (TCM; ref. 38). Here,we quantified the relative percentage of TMM and TCM in CTCLpatients as compared with healthy donors. There was an increase inpercentage of skin-homing (CLAþ) as well as systemic (CLA�) TMM
(CCR7þ L-Selectin�) in CTCL patients as compared with healthydonors. However, we observed increased frequency of skin-homing
CTCL patients Healthy donors Jurkat cells
Live
cel
ls (
×10
5 )
4
3
2
1
2.5
2
1.5
1
4
3
2
1
A
IFNγ IFNγ+IL9
Control IL9 Control IL9 IFNγ IFNγ+IL9
Control IL9 IFNγ IFNγ+IL9
Control IL9
7-A
DD
FITC Annexin V
Live
cel
ls (
×10
5 )
2.5
2
1.5
1
Live
cel
ls (
×10
5 )
10
8
6
4
2
10
8
6
4
2
ns ns ns
B
10
8
6
4
2
0
Dea
d ce
lls (
%)
10
8
6
4
2
0
Dea
d ce
lls (
%)ns
IFNγ IFNγ+IL9Control IL9
IFNγ+IL9IFNγ
Figure 3.
IL9 promotes survival of T cells of CTCL patient and Jurkat cells and reduces the apoptosis in Jurkat cells, CD3þ-sorted T cells from CTCL patients and healthy donorsand Jurkat cells were cultured in presence of cytokines (IL9 and/or IFNg) or left untreated (control).A,Numbers of live cells were quantified to assess the cell survival(n ¼ 3). B, Representative images of Jurkat cells apoptosis after IFNg and IL9 treatment for 72 hours. Apoptotic cells were stained for Annexin and 7-AAD andanalyzed by flow cytometry. Live cells are shown in the bottom left quadrant (7-AAD�/Annexin�); early apoptotic cells are shown in the bottom right quadrant(7-AAD�/Annexinþ); apoptotic cells are shown in the top right quadrant (7-AADþ/Annexinþ) and top left quadrant (7-AADþ/Annexin�). Percentage of dead cells(7-AADþ/Annexinþand7-AADþ/Annexin�)wereplotted and cumulative dataof four individual experiments are depicted. Statistical significancewasdeterminedbycomparing with control using Student paired t test (P values are designated as �� , <0.01; � , <0.05; ns, not significant).
Th9 Axis Promotes T-Cell Survival in CTCL
AACRJournals.org Mol Cancer Res; 18(4) April 2020 663
TCM (CCR7þL-Selectinþ) in advanced-stage CTCL patients as com-pared with healthy donors (Fig. 5A–D). In addition, decreasednumbers of CD4þ na€�ve T cells (CD4þ Tna€�ve) and increased CD4þ
effector T cells (CD4þ Teff) were found in CTCL patients as comparedwith the healthy donors (Fig. 5E and F). However, there was com-parable frequency of CD8þ na€�ve T cells (CD8þ Tna€�ve) and CD8þ
effector T cells (CD8þ Teff) in CTCL patients and healthy donors(Fig. 5G and H).
To understand the activation state of T cells (CD4þ/CD8þ) inCTCLpatients as compared with healthy donors, the expression and fre-quency of CD69þ and CD25þ T cells were quantified. CD69 is anactivation marker which appears very early, while CD25 appears a bitlate on the surface of lymphocyte's plasma membrane. Interestingly,the frequency of CD8þCD69þCD25� T cells was higher in CTCLpatients as compared with healthy donors (Fig. 5K) and no differencewas observed in the frequency of CD4þCD69þCD25� T cells in CTCLpatients as compared with healthy donors (Fig. 5I). Furthermore,
CD69� CD25þ T cells (CD4þ and CD8þ) frequency was higher inCTCL patients as compared with healthy donors (Fig. 5J and L). Thisdata indicate that CTCL showed increased early activation in CD8þ Tcells.
Patients under standard photo/chemotherapy exhibitattenuated skin-homing Th9 cells and reestablished Th1 andTh17 immunity
Finally, we examined the cytokine profile in 6 follow-up CTCLpatients [mycosis fungoides early stage (n ¼ 2), CD30þ LPDadvanced stage (n ¼ 1), and SPTL advanced stage (n ¼ 3)] whowere under standard photo/chemotherapy. The treatment andpatients’ clinical details are provided in Supplementary Table S1.Interestingly, in these patients there was a significant reduction inthe frequency of skin-homing Th9 cells. However, the frequency ofTh1 and Th17 cells significantly increased upon treatment infollow-up patients (Fig. 6A).
UntreatedIL9
IFNγIFNγ+IL9
UntreatedIL9
IFNγIFNγ+IL9
IFNγIFNγ+IL9
UntreatedIL9
CTCL patients Healthy donors Jurkat cells
ROS level (fluorescent intensity)
CTCL patients Healthy donors Jurkat cells
2,000
1,500
1,000
500
0
Fluo
resc
ent i
nten
sity 12,000
10,000
8,000
6,000
4,000
2,000
0
12,000
10,000
8,000
6,000
4,000
2,000
0
Fluo
resc
ent i
nten
sity
2,000
1,500
1,000
500
0
4,000
3,000
2,000
1,000
0
4,000
3,000
2,000
1,000
0
Fluo
resc
ent i
nten
sity
Control IL9 IFNγ IFNγ+IL9
Control IL9 IFNγ IFNγ+IL9
Control IL9 IFNγ IFNγ+IL9
CTCL patients Healthy donors Jurkat cells
A
B
800
600
400
200
0
Lact
ate
conc
entr
atio
n (m
mol
/L)
600
400
200
0
400
300
200
100
0
400
300
200
100
0
400
300
200
100
0
Lact
ate
conc
entr
atio
n (m
mol
/L)
Lact
ate
conc
entr
atio
n (m
mol
/L)
400
300
200
100
0Control IL9 IFNγ IFNγ
+IL9IFNγ IFNγ
+IL9IFNγ IFNγ
+IL9
ns
nsns
Cell
coun
t
Control IL9 Control IL9
ns ns ns
Figure 4.
IL9 reduces toxic ROS levels and lactic acidosis of T cells of CTCL patient and Jurkat cells, CD3þ-sorted T cells fromCTCL patients and healthy donors and Jurkat cellswere cultured in presence of cytokines (IL9 and/or IFNg) or left untreated (control). A, ROS levels were quantified by H2DCFDA staining using flow cytometry. Toppanel depicts a representative histogram for each condition. Bottom panel demonstrates the cumulative data of CTCL patients (n¼ 4), healthy donors (n¼ 3), andJurkat cells (n¼ 8).B, Lactate concentration wasmeasured in cell-free supernatant upon appropriate cytokine treatment as shown in figure in CTCL (n¼ 4), healthydonors (n¼ 4), and Jurkat cells (n¼ 5). Bar plots represent means� SEM. Statistical significance was determined by comparing with control using Student pairedt test (P values are designated as �� , <0.01; � , <0.05; ns, not significant).
Kumar et al.
Mol Cancer Res; 18(4) April 2020 MOLECULAR CANCER RESEARCH664
DiscussionWith this study, we have observed a distinct cytokine profile in
early- and advanced-stage CTCL patients. Early-stage CTCLpatients had an increased frequency of skin-homing Th9 cells andcomparable numbers of other effector Th cells as compared with thehealthy individuals. In contrast, patients with advanced disease hadincreased frequency of Th9 cells (skin-homing and systemic) andimpaired frequency of skin-homing Th1 and Th17 cells. Thisindicates that skin-homing Th9 cells may play a role in initiationas well as in maintenance of the disease in CTCL. The diseaseprogression (advanced-stage CTCL: T3 and T4) led to an impairedantitumor immunity, which might have contributed to the infectionand persistence of lymphoma in these patients. In addition, therewas elevated IL9 production in CTCL patients as compared withhealthy donors. Coelevation of Th9 frequency and IL9 productionindicates that Th9 cells are the major source of IL9 in CTCLpatients. A recent study has reported the presence of CD3þ IL9-producing T cells in skin biopsy of mycosis fungoides lesion, one ofthe major subtypes of CTCL (19). In this study, we have demon-strated the presence of skin-homing Th9 and elevated IL9 produc-tion across different types of CTCL including mycosis fungoides,CD30þ LPD, SPTL, and CTCL-NOS. Importantly, we observed
attenuated Th1 immunity in a large cohort of patients withadvanced disease. Furthermore, there was an increase in Th1 cellfrequency upon photo/chemotherapy treatment. A recent study hasreported that increase in Th1 response reduces the malignant T-cellburden (39). Our observation of impaired Th17 in advanced CTCLdirectly correlates with a previous study where deficiency of RORg ,a transcription factor required for differentiation of IL17, wasshown to promote rapid development of T-cell lymphoma (40).Finally, we observed the reversal of cytokine profile, increase infrequency of skin-homing Th1 and Th17 cells, as well as a signif-icant decrease in Th9 cells upon standard photo/chemotherapy in asubset of follow-up CTCL patients. Collectively, this reveals theinterplay of cytokines in pathogenesis of CTCL and posits thatstrategies promoting the Th1 and Th17 responses and inhibitingTh9 cells might be beneficial and effective in treating CTCL.
Next, we demonstrated the role of IL9 in tumorigenesis anddelineated the underlying mechanisms. Our study revealed the coex-pression of IL9 and IL9R on patient-derived T cells, indicating thepresence of autocrine role of IL9 on tumor cells. Various studies, usingcell lines and murine models, have reported the roles of IL9 in thepathogenesis of different cancers including lung cancer, breast cancer,thyroid cancer, and leukemia (41–44).
CLA+ TMM CLA− TMM CLA− TCMCLA+ TCMA B C D
CL4+ Tnaïve CL4+ Teff CD8+ TeffCD8+ TnaïveE F G H
Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced Healthy Total Early
CTCL
Advanced
Healthy Total EarlyAdvanced Healthy Total EarlyAdvanced Healthy Total EarlyAdvanced Healthy Total EarlyAdvanced
CTCLCTCLCTCL CTCL
Healthy Total EarlyAdvanced
CTCL
Healthy Total Early Advanced
CTCL
Healthy Total Early Advanced
CTCL
Healthy Total Early Advanced
CTCL
CL4+CL69− CL25+ CL8+CL69+ CL25− CL8+CL69− CL25+J K L
Freq
uenc
y (%
) 10080604020
0
Freq
uenc
y (%
)
80
60
40
20
0
Freq
uenc
y (%
)
50
40
30
20
10
0
Freq
uenc
y (%
)
50
40
30
20
10
0
Freq
uenc
y (%
) 80
60
40
20
0
Freq
uenc
y (%
)
100
80
60
40
20
0
Freq
uenc
y (%
)
100
80
60
40
20
0
Freq
uenc
y (%
) 80
60
40
20
0
CL4+CL69+ CL25−I
Freq
uenc
y (%
) 30
20
10
0
Freq
uenc
y (%
)
50
40
30
20
10
0
Freq
uenc
y (%
)50
40
30
20
10
0
Freq
uenc
y (%
)
2520151050
Figure 5.
T-cell subsets and immune activation profile of CTCL patients and healthy donors, human PBMCs isolated from CTCL patients [early (T1 and T2): n ¼ 9, advanceddisease (T3, T4): n¼ 6] and healthy donors (n¼ 19) were stained for surfacemarkers (CD3, CD4, CD8, CD45RA, CD45RO, CLA, CCR7, L-Selectin, CD69, and CD25) asdescribed inMaterials andMethods. Cumulative data of skin-homing TMM (CD4þCLAþCCR7þ L-Selectin�;A), systemic TMM (CD4þCLA�CCR7þ L-Selectin�;B), skin-homing TCM (CD4þ CLAþ CCR7þ L-Selectinþ; C), systemic TCM (CD4þ CLA� CCR7þ L-Selectinþ; D), CD4þ Tna€�ve cells (CD4
þ CD45RAþ CD45RO�; E), CD4þ Teff cells(CD4þ CD45ROþ CD45RA�; F), CD8þ Tna€�ve cells (CD8
þ CD45RAþ CD45RO�; G), CD8þ Teff cells (CD8þ CD45ROþ CD45RA�; H), CD4þ CD69þ CD25� T cells (I),
CD4þCD69� CD25þ T cells (J), CD8þ CD69þ CD25� T cells (K), and CD4þCD69� CD25þ T cells (L) are shown for healthy donors and CTCL patients. Bar plotsrepresent means � SEM. Statistical significance was determined by Student unpaired t test (P values are designated as ��� , <0.001; �� , <0.01; � , <0.05).
Th9 Axis Promotes T-Cell Survival in CTCL
AACRJournals.org Mol Cancer Res; 18(4) April 2020 665
We observed that IL9 significantly decreased the rate of apoptosisand ROS levels (oxidative stress) inside the T cells of CTCL patients,thus indicating the role of IL9 in maintaining a steady-state ROS.Concurrently, IL9 increased the malignant T-cell survival, whichultimately resulted in high cell numbers. The state of redox balancein CTCL and the effect of IL9 in regulating ROS were not knownpreviously. The ROS are chemically reactive oxygen-containing spe-cies that are generated directly or indirectly from free oxygen. In cancercells, ROS is relatively higher as compared with the normal cells, whichhelp them in inducing tumorigenesis (45). However, ROS generationhigher than the steady-state level can be toxic even for the cancercells (46). Once the balance is disturbed, it can lead to oxidative stressinside the cells. Our data indicate the role of IL9 in maintaining ROSlevels in CTCL to promote the malignant T-cell survival. Rapidincreases in intracellular ROS may lead to cellular transformation,DNA damage, and activation of p53, which can ultimately lead toapoptosis (46). ROS levels have been shown to be elevated in variouscell types when the extracellular environment is rich in lactate (36).The decrease in ROS thus could be due to the less acidic environment(reduced lactate levels) generated by IL9. We also observed that IL9
stimulation reduced the extracellular lactate levels in the T-cell culturesupernatant of CTCL patients, indicating less acidosis. This couldbe due to the fact that Warburg effect is energetically inefficient ascompared with oxidative phosphorylation, and could limit the tumorgrowth in glucose-depleted environment, where the cells would thenstart utilizing lactate as an alternative nutrient source (47, 48). Thus,the alteration in redox balance caused by elevated IL9 helps in theoverall survival and maintenance of malignant T cells in CTCL,generating a protumor response. These findings suggest a possiblerole of IL9 in reducing the oxidative stress, which otherwise could bedetrimental for the cancer cells and can lead to apoptosis instead ofhelping in tumorigenesis. Another interesting observation was iden-tifying the unique immune hallmarks of CTCL (summarizedin Fig. 6B). In addition to cytokine profile, we examined the skin-homing and systemic immune–phenotyping focusing on T-cell sub-sets (Tna€�ve, Teff, TMM, and TCM) and T-cell (CD4/CD8) activationprofile (CD69/CD25). Surprisingly, in our study, we have found anincrease in skin homing as well as systemic TMM cells in CTCL ascompared with healthy donors. In addition, there was decreasedfrequency of CD4þ Tna€�ve and increased frequency of CD4þTeff in
CLA+ (skin homing) CLA- (systemic)
Freq
uenc
y (%
)
30
20
10
0
Freq
uenc
y (%
)
40
30
20
10
0
Th2
cells
Freq
uenc
y (%
)
50
40
30
20
10
0
Th1 c
ells
Freq
uenc
y (%
)40
30
20
10
0
Th9
cells
Freq
uenc
y (%
)
15
10
5
0
Th17
cel
lsFr
eque
ncy
(%)
50
40
30
20
10
0
Freq
uenc
y (%
)
1.5
1.0
0.5
0.0
Freq
uenc
y (%
) 5
4
3
2
1
0
A B
C
Baseline Posttherapy
Baseline Posttherapy
ns
ns
nsns
Hall marks of immunefeatures in CTCL
Increased Th9 cells in CTCL
Attenuated Th1 and Th17cells in advanced CTCL
Higher TMM in CTCLpatients
Increased Effector/Naïveratio
Increased early and lateactivation in CD8compartment of advancedCTCL patients
CTCL
Early Advanced
Th1
Th9
Th17
Naïve
Effector
TMM
TCM
Cytokineprofile
Skin
hom
ing
Th cellsubsets
Activationstatus
Healthy CTCL
CD69+
CD25+
CD8+
CD4+
CD8+
CD4+
No IL9R on normal T cellsNo IL9/IL9R interaction
Th9, IL9, IL9RROS Lactic acidosis
Apoptosis Cell survival
IL9
Der
mis
CLA−
Th9 cellsCLA−
Th9 cells
E-selectinCLA
E-selectinCLA
T cell
Th9
Th9Th9
Th9 Th9Th9 Th9
T cell
T cellT cell
T cell
Th9
T cell
Th9Der
mis
IL9
IL9RIL9R
Figure 6.
Cytokine profile after standard photo/chemotherapy, immune hallmarks of disease, and graphical summary of the study. CTCL patients were treated withstandard photo/chemotherapy as described in Supplementary Table S1. A, Skin homing (CLAþ) and systemic (CLA�) Th1, Th2, Th9, and Th17 cells werequantified. Cumulative data of 6 follow-up patients are depicted. Statistical significance was determined by Student paired t test; P values are shown. B,Summary of blood endotyping represents hallmarks of immune features in CTCL. C, Graphical summary depict the role of Th9 cells in patient with CTCL andhealthy individual. Statistical significance was determined by comparing with control using Student paired t test (P values are designated as �� , <0.01; �, <0.05;ns, not significant).
Kumar et al.
Mol Cancer Res; 18(4) April 2020 MOLECULAR CANCER RESEARCH666
CTCL patients as compared with healthy donors. Interestingly, thefrequency of CD8þCD69þCD25� T cells was higher in CTCL patientand percentage of CD69� CD25þ T cells (CD4þ and CD8þ) was higherin CTCL patient as compared with healthy donors.
In conclusion, this study provides first evidence of the presence ofincreased skin-homing Th9 cells and elevated IL9 in CTCL.Moreover,it demonstrates the autocrine-positive feedback loop of Th9 axis inCTCL anddelineates the role of IL9 in reducing the oxidative stress andrate of apoptosis to promote the survival of malignant T cells (graph-ical summary: Fig. 6C). Strategies targeting Th9 cells and inhibition ofIL9 may harbor significant potential in the development of noveleffective therapeutics for these difficult-to-treat malignancies.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors’ ContributionsConception and design: H. Jain, R. PurwarDevelopment of methodology: S. Kumar, B. DhamijaAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): S. Kumar, B. Dhamija, S. Ghosh, A. Dwivedi, N. Sharma, M. Sengar,E. Sridhar, A. Bonda, J. Thorat, P. Tembhare, B. Bagal, S. Laskar, H. JainAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S. Kumar, B. Dhamija, R. PurwarWriting, review, and/or revision of the manuscript: S. Kumar, B. Dhamija,S. Marathe, H. Jain, R. Purwar
Administrative, technical, or material support (i.e., reporting or organizing data,constructing databases): S. Kumar, N. Sharma, R. PurwarStudy supervision: H. Jain, R. PurwarOther (helped in performing experiments: procuring blood samples fromTMH. For assays: T-cell isolation, setting up assays surface and intracellularimmunostaining, and setting up assays for ELISA): S. MaratheOther (helped in performing experiments mainly in flow cytometry): A. DwivediOther (sample collection and performing experiments): A. Karulkar
AcknowledgmentsThis work was supported by Department of Science and Technology
(DST) grant (RD/0119-DST0000-10), Indian Council of Medical Research(ICMR) grant (RD/0119-ICMR000-001), Tata Trust Fund (RD/0117TATAE00-001), Bristol-Myers Squibb (15BMS001), and intramural fund of IIT Bombay to R.Purwar. This work was partially supported by a grant from Department ofBiotechnology, Government of India awarded to Wadhwani Research Centre forBioengineering, and IIT Bombay (BT/INF/22/SP23026/2017). We would like tothank the individuals (patients and healthy volunteers) who donated the blood forthis study. We also thank family members of the patients, the medical staff of TataMemorial Hospital, and Central FACS facility of BSBE, IIT Bombay for theirsupport.
The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.
Received September 4, 2019; revised December 9, 2019; accepted January 24, 2020;published first January 29, 2020.
References1. Girardi M, Heald PW, Wilson LD. The pathogenesis of mycosis fungoides.
N Engl J Med 2004;350:1978–88.2. Kim EJ, Hess S, Richardson SK, Newton S, Showe LC, Benoit BM, et al.
Immunopathogenesis and therapy of cutaneous T cell lymphoma. J Clin Invest2005;115:798–812.
3. Criscione VD, Weinstock MA. Incidence of cutaneous T-cell lymphoma in theUnited States, 1973-2002. Arch Dermatol 2007;143:854–9.
4. Willemze R, Jaffe ES, Burg G, Cerroni L, Berti E, Swerdlow SH, et al.WHO-EORTC classification for cutaneous lymphomas. Blood 2005;105:3768–85.
5. Chen Y, Zheng T, Lan Q, Foss F, Kim C, Chen X, et al. Cytokine polymorphismsin Th1/Th2 pathway genes, body mass index, and risk of non-Hodgkin lym-phoma. Blood 2011;117:585–90.
6. Skinnider BF, Mak TW. The role of cytokines in classical Hodgkin lymphoma.Blood 2002;99:4283–97.
7. Purdue MP, Lan Q, Bagni R, Hocking WG, Baris D, Reding DJ, et al. Pre-diagnostic serum levels of cytokines and other immunemarkers and risk of non-Hodgkin lymphoma. Cancer Res 2011;71:4898–907.
8. Izraeli S, Shochat C, Tal N, Geron I. Towards precision medicine in childhoodleukemia–insights from mutationally activated cytokine receptor pathways inacute lymphoblastic leukemia. Cancer Lett 2014;352:15–20.
9. Hultner L, Druez C, Moeller J, Uyttenhove C, Schmitt E, Rude E, et al. Mast cellgrowth- enhancing activity (MEA) is structurally related and functionallyidentical to the novel mouse T cell growth factor P40/TCGFIII (interleukin9). Eur J Immunol 1990;20:1413–6.
10. Renauld JC, Druez C, Kermouni A,Houssiau F, UyttenhoveC, VanRoost E, et al.Expression cloning of the murine and human interleukin 9 receptor cDNAs.Proc Natl Acad Sci U S A 1992;89:5690–4.
11. Liu J, Harberts E, Tammaro A, Girardi N, Filler RB, Fishelevich R, et al. IL-9regulates allergen-specific Th1 responses in allergic contact dermatitis. J InvestDermatol 2014;134:1903–11.
12. Nowak EC,Weaver CT, Turner H, Begum-Haque S, Becher B, Schreiner B, et al.IL-9 as a mediator of Th17-driven inflammatory disease. J Exp Med 2009;206:1653–60.
13. Jones TG, Hallgren J, Humbles A, Burwell T, Finkelman FD, Alcaide P, et al.Antigen- induced increases in pulmonary mast cell progenitor numbers dependon IL-9 and CD1d- restricted NKT cells. J Immunol 2009;183:5251–60.
14. Leech MD, Grencis RK. Induction of enhanced immunity to intestinal nema-todes using IL-9-producing dendritic cells. J Immunol 2006;176:2505–11.
15. Lauwerys BR,GarotN, Renauld JC,Houssiau FA. Cytokine production and killeractivity of NK/T-NK cells derived with IL-2, IL-15, or the combination of IL-12and IL- 18. J Immunol 2000;165:1847–53.
16. Lu LF, Lind EF, GondekDC, Bennett KA, GleesonMW, Pino-LagosK, et al.Mastcells are essential intermediaries in regulatory T-cell tolerance. Nature 2006;442:997–1002.
17. Schlapbach C, Gehad A, Yang C, Watanabe R, Guenova E, Teague JE, et al.Human TH9 cells are skin-tropic and have autocrine and paracrine proin-flammatory capacity. Sci Transl Med 2014;6:219ra8.
18. Czarnowicki T, He H, Leonard A, Kim HJ, Kameyama N, Pavel AB, et al. Bloodendotyping distinguishes the profile of vitiligo from that of other inflammatoryand autoimmune skin diseases. J Allergy Clin Immunol 2019;143:2095–107.
19. Vieyra-Garcia PA, Wei T, Naym DG, Fredholm S, Fink-Puches R, Cerroni L,et al. STAT3/5-dependent IL9 overexpression contributes to neoplastic cellsurvival in mycosis fungoides. Clin Cancer Res 2016;22:3328–39.
20. Lv X, Feng L, Ge X, Lu K, Wang X. Interleukin-9 promotes cell survival anddrug resistance in diffuse large B-cell lymphoma. J Exp Clin Cancer Res 2016;35:106.
21. Fischer M, Bijman M, Molin D, Cormont F, Uyttenhove C, van Snick J, et al.Increased serum levels of interleukin-9 correlate to negative prognostic factors inHodgkin's lymphoma. Leukemia 2003;17:2513–6.
22. Chen J, Petrus M, Bryant BR, Phuc Nguyen V, Stamer M, Goldman CK, et al.Induction of the IL-9 gene by HTLV-I Tax stimulates the spontaneous prolif-eration of primary adult T-cell leukemia cells by a paracrine mechanism. Blood2008;111:5163–72.
23. Nagato T, Kobayashi H, Kishibe K, Takahara M, Ogino T, Ishii H, et al.Expression of interleukin-9 in nasal natural killer/T-cell lymphoma cell linesand patients. Clin Cancer Res 2005;11:8250–7.
24. Merz H, Houssiau FA, Orscheschek K, Renauld JC, Fliedner A, Herin M, et al.Interleukin-9 expression in human malignant lymphomas: unique associationwith Hodgkin's disease and large cell anaplastic lymphoma. Blood 1991;78:1311–7.
25. Purwar R, Schlapbach C, Xiao S, Kang HS, Elyaman W, Jiang X, et al. Robusttumor immunity to melanoma mediated by interleukin-9-producing T cells.Nat Med 2012;18:1248–53.
AACRJournals.org Mol Cancer Res; 18(4) April 2020 667
Th9 Axis Promotes T-Cell Survival in CTCL
26. Lu Y, Hong S, Li H, Park J, Hong B, Wang L, et al. Th9 cells promote antitumorimmune responses in vivo. J Clin Invest 2012;122:4160–71.
27. Renauld JC, van der Lugt N, Vink A, van Roon M, Godfraind C, Warnier G,et al. Thymic lymphomas in interleukin 9 transgenic mice. Oncogene 1994;9:1327–32.
28. Vink A, Renauld JC, Warnier G, Van Snick J. Interleukin-9 stimulates in vitrogrowth of mouse thymic lymphomas. Eur J Immunol 1993;23:1134–8.
29. Gruss HJ, Brach MA, Drexler HG, Bross KJ, Herrmann F. Interleukin 9 isexpressed by primary and cultured Hodgkin and Reed-Sternberg cells.Cancer Res 1992;52:1026–31.
30. Knoops L, Renauld JC. IL-9 and its receptor: from signal transduction totumorigenesis. Growth Factors 2004;22:207–15.
31. Chen J, Petrus M, Bryant BR, Nguyen VP, Goldman CK, Bamford R,et al. Autocrine/paracrine cytokine stimulation of leukemic cell prolif-eration in smoldering and chronic adult T-cell leukemia. Blood 2010;116:5948–56.
32. Shang Y, Kakinuma S, Amasaki Y, Nishimura M, Kobayashi Y, Shimada Y.Aberrant activation of interleukin-9 receptor and downstream Stat3/5 in pri-mary T-cell lymphomas in vivo in susceptible B6 and resistant C3Hmice. In vivo2008;22:713–20.
33. Zaidi MR,Merlino G. The two faces of interferon-gamma in cancer. Clin CancerRes 2011;17:6118–24.
34. Lin CF, Chen CL, Chien SY, Tseng PC, Wang YC, Tsai TT. Oxidative stressfacilitates IFN-gamma-induced mimic extracellular trap cell death in A549 lungepithelial cancer cells. PLoS One 2016;11:e0162157.
35. Kong H, Chandel NS. Regulation of redox balance in cancer and T cells. J BiolChem 2018;293:7499–507.
36. El Sayed SM, Mahmoud AA, El Sawy SA, Abdelaal EA, Fouad AM, Yousif RS,et al. Warburg effect increases steady-state ROS condition in cancer cellsthrough decreasing their antioxidant capacities (anticancer effects of 3-bro-mopyruvate through antagonizing Warburg effect). Med Hypotheses 2013;81:866–70.
37. Watanabe R, Gehad A, Yang C, Scott LL, Teague JE, Schlapbach C, et al.Human skin is protected by four functionally and phenotypically discrete
populations of resident and recirculating memory T cells. Sci Transl Med2015;7:279ra39.
38. Campbell JJ, Clark RA, Watanabe R, Kupper TS. Sezary syndrome and mycosisfungoides arise from distinct T-cell subsets: a biologic rationale for their distinctclinical behaviors. Blood 2010;116:767–71.
39. Guenova E,Watanabe R, Teague JE, Desimone JA, Jiang Y, DowlatshahiM, et al.TH2 cytokines frommalignant cells suppress TH1 responses and enforce a globalTH2 bias in leukemic cutaneous T-cell lymphoma. Clin Cancer Res 2013;19:3755–63.
40. Ueda E, Kurebayashi S, Sakaue M, Backlund M, Koller B, Jetten AM. Highincidence of T-cell lymphomas in mice deficient in the retinoid-related orphanreceptor RORgamma. Cancer Res 2002;62:901–9.
41. Ye ZJ, Zhou Q, YinW, Yuan ML, YangWB, Xiong XZ, et al. Differentiation andimmune regulation of IL-9-producing CD4þ T cells in malignant pleuraleffusion. Am J Respir Crit Care Med 2012;186:1168–79.
42. Hsieh TH, Hsu CY, Tsai CF, Chiu CC, Liang SS, Wang TN, et al. A novel cell-penetrating peptide suppresses breast tumorigenesis by inhibiting beta-catenin/LEF-1 signaling. Sci Rep 2016;6:19156.
43. Zivancevic-Simonovic S, Mihaljevic O, Majstorovic I, Popovic S, Markovic S,Milosevic-Djordjevic O, et al. Cytokine production in patients with papillarythyroid cancer and associated autoimmune Hashimoto thyroiditis.Cancer Immunol Immunother 2015;64:1011–9.
44. Lavorgna A, Matsuoka M, Harhaj EW. A critical role for IL-17RB signaling inHTLV-1 tax-induced NF-kappaB activation and T-cell transformation.PLoS Pathog 2014;10:e1004418.
45. Fruehauf JP, Meyskens FL Jr. Reactive oxygen species: a breath of life or death?Clin Cancer Res 2007;13:789–94.
46. Trachootham D, Alexandre J, Huang P. Targeting cancer cells by ROS-mediated mechanisms: a radical therapeutic approach? Nat Rev Drug Discov2009;8:579–91.
47. Hui S, Ghergurovich JM,Morscher RJ, Jang C, Teng X, LuW, et al. Glucose feedsthe TCA cycle via circulating lactate. Nature 2017;551:115–8.
48. Faubert B, Li KY, Cai L, Hensley CT, Kim J, Zacharias LG, et al. Lactatemetabolism in human lung tumors. Cell 2017;171:358–71.
Mol Cancer Res; 18(4) April 2020 MOLECULAR CANCER RESEARCH668
Kumar et al.
Research Article
IL1R8 Deficiency Drives Autoimmunity-Associated Lymphoma DevelopmentFederica Riva1,2, Maurilio Ponzoni3, Domenico Supino2,Maria Teresa Sabrina Bertilaccio4, Nadia Polentarutti2, Matteo Massara2,Fabio Pasqualini2, Roberta Carriero2, Anna Innocenzi3, Achille Anselmo2,Tania Veliz-Rodriguez4, Giorgia Simonetti4, Hans-Joachim Anders5,Federico Caligaris-Cappio4, Alberto Mantovani2,6,7, Marta Muzio4, andCecilia Garlanda2,6
Abstract
Chronic inflammation, including that driven by autoim-munity, is associated with the development of B-cell lym-phomas. IL1R8 is a regulatory receptor belonging to theIL1R family, which negatively regulates NF-kB activationfollowing stimulation of IL1R or Toll-like receptor familymembers. IL1R8 deficiency is associated with the develop-ment of severe autoimmune lupus-like disease in lpr mice.We herein investigated whether concomitant exacerbatedinflammation and autoimmunity caused by the deficiencyof IL1R8 could recapitulate autoimmunity-associated lym-phomagenesis. We thus monitored B-cell lymphoma devel-
opment during the aging of IL1R8-deficient lprmice, observ-ing an increased lymphoid cell expansion that evolved todiffuse large B-cell lymphoma (DLBCL). Molecular andgene-expression analyses showed that the NF-kB pathwaywas constitutively activated in Il1r8�/�/lpr B splenocytes.In human DLBCL, IL1R8 had reduced expression comparedwith normal B cells, and higher IL1R8 expression wasassociated with a better outcome. Thus, IL1R8 silencingis associated with increased lymphoproliferation and trans-formation in the pathogenesis of B-cell lymphomas associ-ated with autoimmunity.
IntroductionThe association between chronic inflammation and promotion
of malignancy was first described in the nineteenth century (1)and is supported by epidemiologic andmechanistic data (2, 3). Inparticular, patients suffering from certain autoimmune or inflam-
matory conditions, such as systemic lupus erythematosus (SLE),rheumatoid arthritis, and Sjogren syndrome, are prone to developlymphomas, namely, B-cell non-Hodgkin lymphomas (B-NHL;refs. 4–7). The mechanisms triggering the transition from benignB-cell proliferation to malignancy are still only partially defined.Chronic inflammation, antigen stimulation, and B-cell receptorsignaling, associated with the inherent genetic instability of lym-phocytes, are known to play a central role in lymphoma devel-opment (8, 9). More specifically, gain-of-function mutations ofMYD88 and constitutive activation of NF-kB have emergedamong the most frequently recurring mutations in B-cell lym-phoproliferative diseases (10).
Mice homozygous for the lymphoproliferation spontaneousinactivatingmutation (Faslpr) show systemic autoimmunity, mas-sive lymphadenopathy associated with proliferation of aberrantT cells, arthritis, and immune complex glomerulonephrosis (11).In humans, germline mutations in the FAS gene have beenassociated with autoimmune lymphoproliferative syndrome(ALPS; ref. 12), and somatic FAS mutations have been found inmultiple myeloma and B-NHL (4).
IL1R8 (also known as TIR8 or single Ig IL1-related receptor,SIGIRR) is a member of the interleukin-1 receptor (IL1R)family acting as a negative regulatory receptor (13). IL1R8 inhibitsNF-kB and JNK activation following stimulation of IL1R orTLR family members by interfering with the recruitment of TIRdomain-containing adaptor molecules (14–17). In combinationwith IL18Ra, IL1R8 also serves as one of the receptor chains forthe anti-inflammatory cytokine IL37, thereby activating anti-inflammatory responses (18).
IL1R8deficiency leads touncontrolled activationof IL1Ror TLRfamilymembers and is associated with exacerbated inflammatory
1Department of Veterinary Medicine, University of Milan, Milan, Italy. 2HumanitasResearch Hospital, Rozzano, Italy. 3Ateneo Vita-Salute and Unit of LymphoidMalignancies, IRCCS San Raffaele Scientific Institute; Pathology Unit, SanRaffaele Scientific Institute, Milano, Italy. 4Division of Experimental Oncology,IRCCS San Raffaele Scientific Institute, Milano, Italy. 5Medizinische Klinik andPoliklinik IV, Klinikum der Universit€at M€unchen, LMU M€unchen, Germany.6Humanitas University, Pieve Emanuele, Italy. 7The William Harvey ResearchInstitute, Queen Mary University of London, London, United Kingdom.
Note: Supplementary data for this article are available at Cancer ImmunologyResearch Online (http://cancerimmunolres.aacrjournals.org/).
Current address for M.T.S. Bertilaccio: Department of Experimental Therapeu-tics, The University of Texas MD Anderson Cancer Center, Houston, Texas;current address for G. Simonetti, Biosciences Laboratory, Istituto ScientificoRomagnolo per lo Studio e la Cura dei Tumori (IRST) IRCCS, Meldola, Italy; andcurent address for F. Caligaris-Cappio, Associazione Italiana per la Ricerca sulCancro (AIRC), Milano, Italy.
Corresponding Authors: Cecilia Garlanda, Istituto Clinico Humanitas, ViaManzoni 113, 20089 Rozzano, Italy. Phone: 39-028-224-5115; Fax: 39-028-224-5101; E-mail: [email protected]; and Marta Muzio,IRCCS San Raffaele Hospital, Milano, Italy, Via Olgettina 60, 20132 Milano, Italy.Phone: 39-02-26437104; Fax: 39-02-26434575; E-mail: [email protected]
Cancer Immunol Res 2019;7:874–85
doi: 10.1158/2326-6066.CIR-18-0698
�2019 American Association for Cancer Research.
CancerImmunologyResearch
Cancer Immunol Res; 7(6) June 2019874
responses (14, 19) and autoimmunity (16, 20–22). Accordingly,downregulationof IL1R8 is observed inpsoriasis (23).Dependingon the context, IL1R8 is involved in modulating either inflam-mation-associated tumorigenesis and tumor progression, includ-ing colorectal cancer (19, 24, 25) and chronic lymphocytic leu-kemia (CLL; ref. 26), or NK cell–mediated antitumor immuneresponses in mouse models (27, 28).
IL1R8 deficiency in lpr mice is associated with severe lympho-proliferation and autoimmune lupus-like disease (16), due toincreased dendritic cell (DC) activation and B-cell proliferation inresponse to TLR7- and TLR9-activating autoantigens or nucleo-somes (29, 30).
The involvement of IL1R8 in autoimmunity, and the criticalrole of constitutive activation of MyD88-dependent NF-kBactivation in B-cell transformation, raised the hypothesis thatIL1R8 might be involved in the autoimmunity-associated riskof developing lymphoma. Here, we show that IL1R8 deficiencywas associated with significantly earlier death and increasedsusceptibility to lymphoproliferation, which evolved in trans-plantable diffuse large B-cell lymphoma (DLBCL). Analysis ofclonality showed that multiple independent transformationevents occurred in the same host. In humans, IL1R8 was poorlyexpressed in DLBCL cell lines and primary lesions whencompared with peripheral blood or germinal center B cellsand was associated with better outcome in terms of overallsurvival, suggesting that IL-1R8 downregulation is a driver oflymphomagenesis.
Materials and MethodsAnimals and samples
IL1R8-deficient (Il1r8�/�) mice were generated asdescribed (14) and backcrossed to the C57BL/6J background(Charles River Laboratories) up to the F11 generation. Il1r8�/�
andB6lpr/lpr (Charles River Laboratories)were crossed to generateIl1r8�/�/lpr mice. Mice were housed in the SPF animal facility ofHumanitas Research Hospital in individually ventilated cages.Micewere sacrificed at 12 to 18months of age, unless they reachedthe established endpoints andorganswere collected for histologicand molecular analysis. Procedures involving animals have beenconducted in accordance with, and with the approval of theInstitutional Animal Care and Use Committee of HumanitasResearch Hospital and Italian Health Ministry (authorizations43/2012-B released on August 2, 2012 and 828/2015-PR releasedon July 8, 2015), in compliance with national (D.L. n.116, G.U.,suppl. 40, February 18, 1992; D.L. n.26, March 4, 2014) andinternational law and policies (EEC Council Directive 86/609, OJL 358,1,12-12-1987; EEC Council Directive 2010/63/UE; NIHGuide for the Care and Use of Laboratory Animals, U.S. NationalResearch Council, 2011). All efforts were made to minimize thenumber of animals used and their suffering.
Histopathology and IHCFive-micrometer-thick sections of formalin-fixed, paraffin-
embedded mouse tissues were stained with H&E. Based onlymphoid follicle morphology, a pathologic score was attributedto the spleen and lymph nodes of each 10–12-month-old mouseanalyzed (normal ¼ 0; reactive ¼ 1; reactive > atypical ¼ 2;atypical > reactive ¼ 3; atypical ¼ 4; atypical > lymphomatous ¼5; lymphomatous > atypical¼6; lymphomatous¼7). Slideswereanalyzed by a certified hematopathologist (M. Ponzoni ) and two
investigators who were blinded to the experimental group. Thefollowing antibodies were used: anti-B220 (RA3-6B2, Serotec),anti-Ki67 (SP6, Neo Markers), anti-CD3 (1F4, Bio-Rad), anti-BCL6 (Rabbit polyclonal, Santa Cruz Biotechnology), anti-BCL2(C21, Santa Cruz Biotechnology), and anti-Multiple Myeloma 1/Interferon Regulatory Factor 4 protein (MUM1/IRF4; 3E4,BioLegend; ref. 31).
Tumor transplantationA total of 107 cells (5 � 106 splenocytes plus 5 � 106 lymph
node cells) from 10- to 12-month-old Il1r8�/�/lpr (n ¼ 8) orIl1r8þ/þ/lpr (n¼ 7)mice were injected i.p., s.c., or i.v. into C57BL/6J, nude or SCID mice (n ¼ 17 recipients of Il1r8�/�/lpr cells;n ¼ 16 recipients of Il1r8þ/þ/lpr cells). Recipient animals weresacrificed when clinical signs (enlargement of mandibular lymphnodes or abdomen) were evident or 12 to 20 months aftertransplantation, and organs were collected for histologic andmolecular analysis. The genotype of several tissues of all recipientmice was analyzed for lpr and Il1r8 mutations by PCR (14).
Western blot analysisWestern blot analysis of purified B-cell lysates (30 mg total
proteins) was performed with the following antibodies: rabbitanti-p100/p52 (CS4882, 1:1,000, overnight at 4�), mouse anti-Phospho-p65 (CS3036, 1:1,000, overnight at 4�C), rabbit anti-p65 (CS8242, 1:1,000, 2 hours RT; Cell Signaling Technology);anti-beta-actin-HRP (SIGMA A3852, 1:10,000, 2 hours roomtemperature), followed by anti-rabbit-HRP (Sigma A0545,1:5,000) or anti-mouse-HRP (Sigma A3682, 1:5,000), using10% or 4%–12% gradient precast gels (GenScript).
Real-time PCR and real-time PCR arrayTotal RNA frommouse spleen-purifiedB cells,DLBCL cell lines,
and B cells from healthy donor buffy coats was isolated with acolumn-based kit followed by DNAse treatment (Promega; forPCR array) or TRI reagent (Sigma-Aldrich; for PCR). RNA wasretrotranscribed and cDNA used for gene-expression analysis byreal-time PCR and real-time PCR array (Bio-Rad Prime PCRARRAY code:10034381).
Real-time PCR was performed in QuantStudium 7 Flex(AppliedBiosystems, ThermoFisher) or 7900SequenceDetectionSystem (Applied Biosystems), in duplicate using PowerSybr Green PCR Master Mix (Applied Biosystems) and primers(300 nmol/L) in MicroAmp optical 96-well plates (25 mL). Thefollowing primer pairs were purchased from Invitrogen: Nfkbizfor 50-GCGCTCTCGTATGTCC-30; Nfkbiz Rev 50-AGACTGCC-GATTCCTC-30; GAPDH for 50-GCAAAGTGGAGATTGTTGCCAT-30; GAPDH Rev 50-CCTTGA CTGTGCCGTTGAATTT-30 (28);human IL1R8 For: 50-CCGACCTTTTGG TGAACCTGA-30; humanIL1R8 Rev: 50-TGGCCCTCAAAGGTGATGAAG-30; Universal actinFor: 50-CCCAAGGCCAACCGCGAGAAGAT-30; Universal actinRev: 50-GTCCCGGCCAGCCAGGTCCAG-30. Experiments wererepeated at least twice. The expression of the target gene wasnormalizedusingGAPDHorb-actin cDNAexpressionof the samesample and run, and reported as 2(�DCT).
For real-time PCR array, the analysis of 84 NF-kB signalingtarget genes was performed as described (32). Data were reportedas 2(�DCT), relative to the average of six housekeeping genes. Ofnote, the specific assay for Fas mRNA expression is designedwithin exons 1 and2, and it recognizes bothwild-type andmutantFas (33).
IL1R8 in Autoimmunity-Associated B-cell Lymphoma
www.aacrjournals.org Cancer Immunol Res; 7(6) June 2019 875
IgH gene rearrangement analysisIgH gene rearrangement was investigated by Southern blot anal-
ysis of genomic DNA from different organs of one 20-month-oldwild-typemouse transplanted s.c.with total lymphnode, andspleencells collected from one 11-month-old Il1r8�/�/lpr mouse. DNA(5 mg) was extracted from spleen, lymph nodes, and solid lesions(100 mg each), digested with EcoRI or StuI and subsequentlyhybridized with a 32P-labeled DNA probe PJ3 representing the JH4region of the IgH locus, as described (34).
Cell cultureB cells were isolated from 100 � 106 splenocytes using a
B-cell isolation kit (Miltenyi Biotec), plated in 48-well plate at1 � 106/mL and cultured overnight.
Human DLBCL cell lines SU-DHL-2 (ATCC CRL-2956), SU-DHL-4 (ATCC CRL-2957), SU-DHL-6 (ATCC CRL-2959), SU-DHL-8 (ATCC CRL-2961), RC-K8 (DSMZ ACC-561), and RIVA(RI-1; DSMZ ACC-585) were received in 2013, expanded andfrozen, and then thawed and grown for 10 days in RPMI or IMDM(RIVA, RC-K8) medium supplemented with 10%–20% FCS, 2mmol/L L-glutamine, and streptomycin (100 U/mL) before theexperiments. DLBCL cell lines were not authenticated, were rou-tinely tested forMycoplasma contamination, and onlyMycoplasma-free cells were used for flow cytometry and molecular analysis.
Flow-cytometric analysisMouse spleen B cells overnight cultured in medium or in the
presence of LPS (100ng–1mg/mL, Sigma-Aldrich)were incubatedwith anti-mouse CD86 (GL1, eBioscience) and anti-mouse CD19(1D3, BD Biosciences) for 30 minutes at 4�C for surface stainingand analyzed by FACSCanto II (Becton Dickinson).
IL1R8 cell-surface staining on human cells was performed withbiotinylated goat anti-human IL1R8/SIGIRR (R&D Systems),followed by Alexa 647–conjugated streptavidin (MolecularProbes, Invitrogen), and analyzed with FACSCanto I flow cyt-ometer (BD Biosciences). Results are reported as mean fluores-cence intensity (MFI) normalized on fluorescence minus one.Diva software (BD Pharmingen) and FlowJo (Tree Star) were usedfor data acquisition and analysis, respectively.
Analysis of IL1R8 and IL37 expression in human DLBCLPublic gene-expression data of DLBCL were retrieved from
GEO. In the first study (GSE43677) samples of na€�ve B cells(n ¼ 8), germinal center (GC) B cells (n ¼ 13), post-GC B cells(n¼ 9), tonsils (n¼ 10), and DLBCL (n¼ 12) were analyzed. Inthe second study (GSE32018), gene-expression profiling wasconducted in a series of B-cell non-Hodgkin lymphomapatients [17 CLL, 22 DLBCL, 23 follicular lymphoma (FL),24 mantle cell lymphoma (MCL), 15 marginal zone lympho-ma-MALT type (MALT), 13 nodal marginal zone lymphoma(NMZL)], and 7 freshly frozen lymph nodes. Differentialexpression analysis was performed using limma (version3.26.8; ref. 35). For prognosis evaluation, expression andclinical data of 98 DLBCL cases selected from 220 lymphomasamples (GSE4475) were used. DLBCL patients treated withradiotherapy were excluded.
TheGene Set Enrichment Analysis (GSEA) softwarewas used toperform the overrepresentation analysis with gene sets comingfrom the Molecular Signature Database (36). The entire datama-trix containing normalized gene-expression values (log scale) wasused, and the expression profile of the IL1R8 gene was tested as
continuous phenotype label. The analysis was performed choos-ing the Pearson correlation as the metric to investigate gene setsenriched by genes correlated with the expression profile of IL1R8.The Reactome database, belonging to the C2 collection (c2.cp.reactome.v6.1), was used. Resulting gene sets were consideredsignificantly enriched according to the false discovery rate (FDR)threshold of 5%.
Statistical analysisStatistical differences in mouse mortality and lymphoma inci-
dence rates were analyzedwith theMantel–Cox test and the Fishertest, respectively. The Mann–Whitney test or Student t test withWelch correction was performed as specified. Survival analysis ofhuman DLBCL was performed using the Kaplan–Meier andMantel–Cox tests. The median gene-expression value was usedto classify patients into IL1R8low and IL1R8high or IL37low andIL37high gene-expression groups. A P < 0.05 was consideredstatistically significant. Statistical analysis was performed usingGraphPad Prism software (GraphPad Software).
ResultsIL1R8 deficiency is associated with severe lymphadenopathyand lymphoma in lpr mice
We previously observed that 6-month-old Il1r8�/�/lpr micewere affected by enhanced lymphoproliferation and lymph fol-licle hyperplasia comparedwith Il1r8þ/þ/lprmice (16). In order toaddress whether this benign lymphoproliferation eventuallyevolved to malignancy, we analyzed survival and followed theevolution of lymphoid organs in older animals. As shown bysurvival curves reported in Fig. 1A, Il1r8�/�/lpr mice reached theendpoints earlier than Il1r8þ/þ/lprmice, and mortality was 100%(23/23) at 15 months of age in Il1r8�/�/lpr mice compared with22% (6/27) in Il1r8þ/þ/lpr mice (P ¼ 0.0001, Mantel–Cox test).Splenomegaly and lymphadenomegalyweremore pronounced in10- to 12-month-old Il1r8�/�/lprmice compared with Il1r8þ/þ/lprmice of the same age (Fig. 1B). The spleenweight was significantlyincreased in both groups compared with wild-type or Il1r8�/�
mice, and in Il1r8�/�/lprmice it was significantly greater (4-fold)than in Il1r8þ/þ/lpr mice (Fig. 1C).
Histopathologic analysis of the spleens of 12- to 14-month-oldIl1r8�/�/lprmice showed an enlargement of the white pulp and acomplete loss of the normal architecture of the organ in mostanimals (Fig. 2A). In the spleens of 12- to 14-month-old Il1r8þ/
þ/lpr mice, we observed a moderate enlargement of the whitepulp, but the architecture of the organ remained recognizabledespite the presence of enlarged germinal centers (Fig. 2A). Sim-ilarly, most (62.5%; 20/32) lymph nodes from 10- to 12-month-old Il1r8�/�/lprmice presented abnormal histologic architecture,without any evident follicle (Fig. 2B). In contrast, lymph nodesfrom 10- to 12-month-old Il1r8þ/þ/lpr mice were enlarged butgenerally retained a preserved normalmorphology of the follicles(Fig. 2B). As shown in Fig. 2C and D, the pathologic score oflymphoid follicles (based on the presence of normal, reactive,atypical, or lymphomatous follicles) was significantly higher in10- to 12-month-old Il1r8�/�/lpr mice compared with 12- to 14-month-old Il1r8þ/þ/lprmice (P¼0.0001 in spleen andP¼0.0022in lymph nodes), and the diagnosis of lymphoma was mostlylimited to Il1r8�/�/lpr mice. Splenic and lymph nodal plasmacy-tosis occurred in the spleen and lymph nodes of both Il1r8þ/þ/lprand Il1r8�/�/lpr mice, in agreement with the role of TLR ligands
Riva et al.
Cancer Immunol Res; 7(6) June 2019 Cancer Immunology Research876
and autoantigens in inducing cellular differentiation into matureplasma cells and plasma cell expansion (37, 38).
Development of DLBCL in Il1r8�/�/lpr miceHistopathologic analysis showed that a diffuse organ replace-
ment by large cells in spleen and lymph nodes consistent with adiagnosis of lymphoma occurred in 13 of 26 Il1r8�/�/lprmice and3 of 23 Il1r8þ/þ/lprmice (P ¼ 0.0073, Fisher test; Fig. 2E). In 8 of13 mice with lymphoma (61.5%), large cells were observed inliver, lung, kidney, and gut, indicating the development ofDLBCL. Immunostaining for B220 highlighted a diffuse infiltra-tion by large neoplastic B cells (Fig. 3A). CD3þ T cells appeareddistributed throughout the tissue, rather than clustering withinand around follicles of the splenic white pulp in lymphoma-bearing Il1r8�/�/lpr mice (Fig. 3B). Neoplastic B lymphocyteswere relatively monomorphic, with large nuclei and abundantcytoplasm; within this population, high mitotic rate as well asdiffusely elevatedKi67 immunoreactivity confirmed the increasedproliferation rate of these lymphomas (Fig. 3C). In addition,neoplastic B cells were immunoreactive for Bcl-2, suggesting anactivation of an antiapoptotic machinery, and negative for Bcl-6and MUM1 (Fig. 3D–F, respectively). In Il1r8�/�/lpr lymphnodes, we observed lesions with similar features and in a fewcases (3/13, 23%), bone marrow of Il1r8�/�/lpr mice showedsmall and focal areas of DLBCL.
DLBCL was rarely diagnosed in Il1r8þ/þ/lprmice older than 12months (3/23, 13%)aswell. The vastmajority of Il1r8þ/þ/lprmiceactually showed an irregular enlargement of germinal centerswithpredominance of intermediate cells, without fulfilling therequired criteria for FL. In addition, the presence of few, scatteredlarge B220þ cells was consistent with atypical, lymphoprolifera-tive, potentially preneoplastic disorder (Fig. 3A–C).
Il1r8�/�/lpr DLBCL are transplantable and oligoclonalIn order to further demonstrate the malignant capability of
lesions developing in aged Il1r8�/�/lpr and Il1r8þ/þ/lpr mice,splenocytes and lymph node cells collected from eight andseven donors, respectively, were injected through differentroutes in immunodeficient (nude and SCID) or immunocompe-tent mice (17 recipients of Il1r8�/�/lpr and 16 recipients ofIl1r8þ/þ/lpr cells). Irrespective of the immunocompetence ofthe recipient mouse and route of injection, after 4 to 20 monthsof observation, cells from 4 of 8 Il1r8�/�/lpr mice generatedhistologically confirmed parental DLBCL in recipient mice
(Fig. 4A). In contrast, mice injected with cells collected from theseven Il1r8þ/þ/lpr donor mice did not develop lymphoma inrecipient mice up to 20 months of observation. Genotyping forlpr and Il1r8 gene modifications was performed by PCR analysison genomic DNA of several tissues of all recipient mice. Genotyp-ing of the recipient's organs affected by lymphoma (spleen andlymph nodes) and control tissues (skeletal muscle) confirmedthat malignant cells originated from the Il1r8�/�/lpr donors(Fig. 4B).
Because lymphoma lesions infiltratedmore thanone lymphoidorgan in Il1r8�/�/lprmice, we next assessed whether these lesionsoriginated from a common B-cell clone. Southern blot analysiswas performed to detect immunoglobulin (Ig) heavy-chain gene(IgH) rearrangements in DLBCL developed in different organs ofone recipient wild-type mouse transplanted with splenocytes andlymph node cells of one Il1r8�/�/lpr donor. The analysis revealedbands of IgH rearrangement of different sizes in the spleen, lymphnodes, and other organs, indicating that multiple B-cell cloneswere transformed in the donor mouse (Fig. 4C and D).
These results indicate that lymphomas of Il1r8�/�/lpr micecan be transplanted in wild-type recipient mice, giving rise tolymphoma.
Constitutive activation of the NF-kB pathway in Il1r8�/�/lprB cells
Hyperactivation of the NF-kB pathway and overexpression ofNFKBIZ are hallmarks of a subtype ofDLBCL in humans (39–41).IL1R8 dampensNF-kB activation induced by TLR and IL1R familymembers (42), and Fasmutations affect B-cell activation (16). Inaddition, we previously showed that IL1R8 deficiency significant-ly increased B-cell proliferation upon exposure to RNA and DNAimmune complexes and other TLR agonists (16). To furtherinvestigate the NF-kB pathway in B cells of Il1r8�/�/lpr mice, weanalyzed NF-kB activation and the expression of specific NF-kBregulated genes in CD19þ cells purified from the spleen of12-month-old Il1r8þ/þ, Il1r8�/�, Il1r8þ/þ/lpr, and Il1r8�/�/lprmice, in resting conditions or after stimulation with LPS.Noncanonical and canonical NF-kB activation can be monitoredby the cleavage of p100 (Nfkb2) into p52 fragment, and phos-phorylation of p65 (RelA), respectively. In contrast to wild-typemice, we observed activation of the noncanonical NF-kB pathwayin both Il1r8þ/þ/lpr and Il1r8�/�/lpr mice, whereas the canonicalpathway appeared mostly activated in the Il1r8þ/þ/lpr mice(Fig. 5A–C). We did not observe any significant difference in
Figure 1.
IL1R8 deficiency increases the severity of the lymphoproliferative disorder of lprmice.A,Mortality rate at 15 months of age in Il1r8�/�/lprmice (100%; n¼ 23)and in Il1r8þ/þ/lprmice (22%; n¼ 27) (Mantel–Cox test P < 0.0001). B, Spleen (bottom) andmandibular lymph nodes (top) from 10- to 12-month-old Il1r8þ/þ,Il1r8�/�, Il1r8þ/þ/lpr, Il1r8�/�/lprmice. C, Spleen weights of 10- to 12-month-old Il1r8þ/þ (n¼ 6), Il1r8�/� (n¼ 6), Il1r8þ/þ/lpr (n¼ 30), and Il1r8�/�/lpr (n¼ 33)mice (unpaired Student t test withWelch correction; mean and SD are indicated).
IL1R8 in Autoimmunity-Associated B-cell Lymphoma
www.aacrjournals.org Cancer Immunol Res; 7(6) June 2019 877
Il1r8�/� mice at this time point compared with Il1r8þ/þmice inthe absence of ex vivo stimulation.
Next, we used a real-time PCR array to analyze 84 genes knownto be targets of the NF-kB signaling pathway (SupplementaryTable S1).We compared the results obtained fromnonstimulatedB cells collected from 4 wild-type mice, 4 Il1r8�/� mice, 5 Il1r8þ/
þ/lpr mice, and 5 Il1r8�/�/lpr mice (Fig. 5D; SupplementaryTable S2). Of the 14 genes dysregulated in at least one group,13 genes were upregulated, and only one was downregulated ascompared with wild-type animals, again suggesting constitutivehyperactivation of this pathway in Il1r8þ/þ/lpr and Il1r8�/�/lpr Bcells from aged mice. Most of the NF-kB targets were upregulatedin both Il1r8þ/þ/lpr and Il1r8�/�/lpr mice, including proinflam-matory genes (e.g., Il1b, Ifng, Csf1, Stat1, Il12b) and genes asso-ciated with proliferation or antiapoptosis (e.g., Ccnd1; Supple-mentary Table S2). The Bcl2a1a gene coding for an antiapoptoticprotein necessary for cell transformation and growth in anaplasticlymphoma (43) was downregulated in lprmice (fold difference¼0.47 and P ¼ 0.04 as shown in Supplementary Table S2), andalthough IL1R8 deficiency did not influence its expression (fold
difference¼ 0.93; P¼ 0.8), expression was restored in Il1r8�/�/lprmice (fold difference ¼ 1.08 and P ¼ 0.8 as shown in Supple-mentary Table S2).
We then analyzed a secondary response gene prototypicallyinduced by TLRs and regulated byNF-kB, namely,Nfkbiz. In basalconditions, we observed low expression ofNfkbizmRNA in B cellsisolated from wild-type mice; however, Nfkbiz was significantlyinduced in Il1r8�/� mice, and this induction was sustained inIl1r8�/�/lpr mice (Fig. 5E), suggesting that a TLR-dependentNF-kB secondary response is constitutively activated in Il1r8�/�
and Il1r8�/�/lpr mice.In agreement with dysregulated activation of the NF-kB
pathway, flow cytometry analysis showed that overnight-cultured, spleen-purified Il1r8�/�/lpr B cells had increasedexpression of CD86, an activation marker downstream ofTLR activation (44, 45), compared with wild-type, Il1r8�/�,and Il1r8þ/þ/lpr B cells, both in basal conditions and after LPSstimulation (Fig. 5F).
Taken together, these results show that both Fas and IL1R8deficiencies contribute to constitutive dysregulated NF-kB
Figure 2.
IL1R8 deficiency is associated with increased susceptibility to lymphoma development in lprmice. A and B, Histopathologic analysis of the spleen (A) andlymph nodes (B) of 10- to 12-month-old Il1r8þ/þ, Il1r8�/�, Il1r8þ/þ/lpr, and Il1r8�/�/lprmice stained with H&E (400�; Axioskop 40microscope equipped withAxioCamMRc camera and AxioVision Rel. 4.8 acquisition software; Zeiss). C and D, Pathologic score of the spleen (C) and lymph nodes (D) of 10- to 12-month-old Il1r8þ/þ (n¼ 2), Il1r8�/� (n¼ 2), Il1r8þ/þ/lpr (n¼ 20), Il1r8�/�/lpr (n¼ 26) mice (unpaired Student t test; mean and SD are indicated). E, Incidence ofDLBCL in Il1r8þ/þ/lpr (3/23) and Il1r8�/�/lpr (13/26) mice (Fisher test).
Riva et al.
Cancer Immunol Res; 7(6) June 2019 Cancer Immunology Research878
signaling and increased B-cell activation with few differencesassociated with IL1R8 deficiency (e.g., induction of Nfkbiz). Thedouble mutation rendered the cells highly responsive to TLRactivation, as measured by CD86 expression.
IL1R8 expression is downmodulated in human DLBCL cellsand correlates with prognosis
To assess the relevance of these results to human disease, wefirst studied the expression of IL1R8 in human lymphoma celllines compared with normal circulating mature B cells. As shownin Fig. 6A and B, all DLBCL cell lines analyzed expressed lowerIL1R8mRNA and protein, respectively, compared with peripheralblood B cells.
Next, we studied IL1R8 expression in public gene-expressiondata of DLBCL retrieved from GEO, comparing different nor-mal resting and activated B-cell populations and lymphomas.In the first study analyzed (GSE43677), the expression of IL1R8was significantly downregulated in DLBCL samples versus na€�veB cells (logFC ¼ �0.43, adj. P ¼ 1.08E�04), GC B cells(logFC ¼ �0.21, adj. P ¼ 1.70E�02), post-GC B cells (logFC
¼ �0.9, adj. P ¼ 1.12E�09), and tonsils (logFC ¼ �0.62, adj.P ¼ 3.64E�08; Fig. 6C). The second study (GSE32018) showeda significant downregulation of IL1R8 expression in DLBCLversus lymph node control samples (logFC ¼ �1.34, adj.P ¼ 1.65E�02), but also versus FL, an indolent form thatmay transform into DLBCL (logFC ¼ �0.66, adj. P ¼3.45E�02; Fig. 6D).
In a third study (GSE4475), the expression of IL1R8 wasanalyzed together with clinical data to evaluate a correlation withprognosis. DLBCL patients were divided into IL1R8low and IL1R8-high based on the median gene expression. The resulting Kaplan–Meier curve showed that patients with IL1R8 expression above themedian value had significantly prolonged overall survival [hazardratio (HR)¼ 2.2; 95% CI, 1.2–3.8; P¼ 0.006; Fig. 6E), comparedwith patients below the median. In addition, the GSEA analysisretrieved a total of 60 pathways significantly enriched by genespositively correlated with IL1R8 gene-expression profile (Supple-mentary Table S3). Among these, the apoptotic process and theDNA damage response were two of the most enriched pathways(NES ¼ 2.02, FDR q ¼ 0.005 for the apoptosis process; NES ¼
Figure 3.
Il1r8�/�/lprmice develop DLBCL lesions. IHC analysis of B220 (A), CD3 (B), and Ki67 (C) in the spleen of 10- to 12-month-old Il1r8þ/þ, Il1r8�/�, Il1r8þ/þ/lpr, andIl1r8�/�/lprmice (40�). IHC analysis of bcl-2 (D), bcl-6 (E), and MUM-1 (F) in the spleen of 10- to 12-month-old Il1r8�/�/lprmice (representative images from 5mice analyzed; 400�; Axioskop 40microscope equipped with AxioCamMRc camera and AxioVision Rel. 4.8 acquisition software; Zeiss).
IL1R8 in Autoimmunity-Associated B-cell Lymphoma
www.aacrjournals.org Cancer Immunol Res; 7(6) June 2019 879
1.85, FDR q ¼ 0.01 for the P53-dependent G1 DNA damageresponse) with a total of 70 and 29 genes, respectively, belongingto the core enrichment (Supplementary Tables S4 and S5). Theseresults show a positive coregulation of the apoptotic process andDNA damage response genes with our gene of interest, suggestinga putative activation of the apoptotic process and DNA damageresponse in profiles with high IL1R8 expression compared withthose with low expression.
Because IL1R8 is required for the anti-inflammatory activityof IL37 in inflammatory conditions triggered by TLRligands (18, 46), we finally investigated whether IL1R8 and IL37were coregulated in DLBCL. In the GSE43677 and GSE32018studies, the expression of IL37 was significantly downregulated inDLBCL samples compared with normal B cells (logFC ¼ �0.18,adj. P ¼ 3.73E�02 for na€�ve B cells, logFC ¼ �0.25, adj. P ¼1.65E�03 for GC B cells) or FL cells (logFC ¼ �0.37, adj. P ¼4.30E�03), respectively, similarly to IL1R8 (Fig. 6F, G). However,in contrast to what was observed for IL1R8, the overall survivalwas not affected by IL37 expression in the GSE4475 study (HR,0.6; 95% CI, 0.4–1.1; P¼ 0.1), indicating that the regulatory roleof IL1R8 affects additional pathways.
These results indicate that IL1R8 is poorly expressed in DLBCLcompared with healthy GC B cells and other B-cell lymphomas,
and that lower IL1R8 expression is associated with shorter overallsurvival.
DiscussionIL1R8 is known to act as a negative regulator of NF-kB and JNK
activation following stimulation of IL1R familymembers or TLRs.We herein showed that the increased susceptibility to lympho-proliferation observed in lpr mice deficient of IL1R8 is alsoassociated with frequent development of DLBCL. The aggressivelymphomas that developed in Il1r8�/�/lpr mice were transplant-able and oligoclonal, possibly originating from multiple B-cellclones. In addition, we showed that IL1R8 expression is down-regulated in human DLBCL cells in comparison with peripheralblood, GC B cells, and other lymphomas. Expression also corre-lated with overall survival, suggesting that IL1R8 silencing inDLBCL might contribute to dysregulated NF-kB activation, afrequent occurrence observed in DLBCL, lymphoproliferation,and transformation.
FAS-deficient lpr mice are a model of ALPS and SLE. FAS is aproapoptotic TNF receptor superfamilymember, highly expressedon GC B cells. Mutations in the genes encoding FAS or its ligandcause massive accumulation of autoreactive B and T cells,
Figure 4.
DLBCL lesions are transplantable and oligoclonal. A, Histopathologic analysis of the spleen of a 12-month-old Il1r8�/�/lpr donor mice compared with thespleen of Il1r8þ/þ recipient mice stained with H&E (Axioskop 40, Zeiss, 200 and 400� for left and right). B,Genomic analysis by PCR of lpr and Il1r8 targetedgenes in organs of one recipient mouse 6.5 months after transplantation with Il1r8�/�/lpr spleen and lymph node cells. A and B, Representative images of 1/8Il1r8�/�/lpr donors and 1 of 8 recipients in which the transplanted cells generated a lymphoma. C and D, Southern blot analysis of Ig genes shows rearrangementand oligoclonal expansion of B cells in one recipient mouse of Il1r8�/�/lpr spleen and lymph node cells. Genomic DNA from different organs and tissues of therecipient animal was digested with EcoRI (C) or StuI (D). Yellow arrows indicate clonal bands.
Riva et al.
Cancer Immunol Res; 7(6) June 2019 Cancer Immunology Research880
resulting in ALPS in humans (12). In addition, FAS mutation hasbeen foundassociatedwithperforindeficiency inone caseofALPSand lymphoma (47), whereas in mice, increased lymphomadevelopment was observed in lpr mice deficient of SPARC, thegene coding for osteonectin (48). In a previous study, we dem-onstrated that IL1R8 deficiency was associated with amore severelymphadenopathy at 6 months of age in FAS-deficient lprmice (16). This phenotype was mainly due to overactivation ofDCs, B cells, and CD4þ T cells upon stimulation with lupusautoantigens, possibly through TLR7 engagement (16). Indeed,chromatin antigens in immune complexes can potently engageboth the BCR and TLRs in B cells, leading to overstimulation anddefective apoptosis of B cells, as well as to secondary inflamma-tion (5). FAS mutations have also been observed in humanlymphomas, indicating that longer lymphocyte survival mayallow accumulation of additional oncogenic events (4).
In addition to these autoimmunity-dependent mechanisms,genetic alterations affecting components of the NF-kB signalingpathways have been shown to occur frequently in DLBCL. Con-stitutiveNF-kBpathway activity is observed in almost all activated
B-cell–like (ABC) types of DLBCL and in a large fraction ofgerminal center B-cell (GCB)–DLBCLs and is associated with theproliferation, differentiation, and survival ofmalignant lymphoidcells (49, 50). Among mutations of the NF-kB signaling pathwayin B-cell lymphomas, MYD88mutations have emerged as one ofthe most frequently recurring (40). MyD88 is an adaptor proteinthat mediates TLR and IL1R signaling. Gain-of-function muta-tions of MYD88 confer a cell survival advantage during theevolution of DLBCL by promoting NF-kB and JAK/STAT3 signal-ing (40). IL1R8 tunes TLR and IL1R-dependent signaling byinterfering with the recruitment of TIR-containing adaptor mole-cules (51) and IL1R8 deficiency in mice is associated with uncon-trolled inflammatory responses both in infectious and sterileconditions (42). Furthermore, genetic inactivation of IL1R8 wasobserved to cause earlier, more disseminated, and aggressiveleukemia in the Em-TCL1 mouse model of CLL (26). In thismodel, the neoplastic transformation of B cells has an incidenceof 100% and is mediated by the overexpression of the TCL1oncogene; the absence of IL1R8 exacerbates CLL progression,but its impact on the B-cell transformation has not been
Figure 5.
Dysregulated NF-kB activation in Il1r8þ/þ/lpr and Il1r8�/�/lpr mice. A, Western blot of spleen B cells with the indicated antibodies. b-Actin expressionwas analyzed as internal control. B and C, Densitometric signal ratios of p52/p100 and phospho-p65/p65 shown in A (Mann–Whitney test). n ¼ 4Il1r8þ/þ, n ¼ 5 Il1r8�/�, n ¼ 4 Il1r8þ/þ/lpr, n ¼ 6 Il1r8�/�/lpr mice. D, Real-time PCR array of NF-kB signaling target genes. Expression data are shownonly for the genes for which a fold difference (FD) > 2 was observed in at least one comparison between two groups of mice (see SupplementaryTable S2 for individual data). In the graph, a two-color scale formatting scheme was used to format cells: red is the maximum expression; blue,minimum. Each column represents one sample (from one mouse). E, Nfkbiz mRNA expression in purified B cells (unpaired Student t test with Welchcorrection). A–E, One experiment performed. F, FACS analysis of CD86 expression in overnight-cultured purified B cells in basal condition and afterLPS stimulation. Top, representative histograms. Bottom, results are reported as MFI normalized on fluorescence minus one. One representativeexperiment with B cells collected from 3 to 6 mice (1 or 2 replicates per mouse) out of 2 performed is shown (unpaired Student t test with Welchcorrection). B, C, E, F, mean and SD are shown.
IL1R8 in Autoimmunity-Associated B-cell Lymphoma
www.aacrjournals.org Cancer Immunol Res; 7(6) June 2019 881
investigated (26). Finally, IL1R8 in associationwith IL18Ra servesa receptor chain for IL37, a cytokine that provides an anti-inflammatory environment in the aging bonemarrow, preventingoncogenic transformation of B-cell progenitors (52). Our resultsshow that Il1r8þ/þ/lprmice spontaneously developed DLBCL at avery low frequency and at late age (12–18 months). IL1R8deficiency increased the frequency and the severity of the disease,
as well as accelerated the onset of disease to 8 to 12months of age,indicating that IL1R8 has also a role in the neoplastic transfor-mation of B cells, and not only in the progression of establishedB-cell leukemia or lymphoma.
The pathogenesis of lymphoma seen in patients with autoim-mune diseases is complex and involves different factors contrib-uting to lymphomagenesis, including both disease activity and
Figure 6.
IL1R8 is downmodulated in humanlymphoma cell lines. A, Real-timePCR analysis of IL1R8 mRNAexpression (unpaired Student ttest of each line versus normal Bcells withWelch correction; meanand SD are indicated; the highestP value is reported) and B, flow-cytometric analysis of IL1R8protein expression (top,representative histograms;bottom, MFI quantification) inhuman lymphoma cell linescompared with circulating B(CD19þ) cells from 3 to 4 healthydonors (unpaired Student t test ofeach line versus normal B cells;mean and SD are indicated; thehighest P value is reported); 3 to4 experiments were performedwith each cell line. Results areexpressed as arbitrary units (A)and as MFI normalized onfluorescence minus one (B,bottom). C, Normalized logintensity of IL1R8 probe (218921_at;GSE43677) in DLBCL versusnormal B cells (na€�ve B cells,germinal center (GC) B cells, postGC B cells, and tonsil samples).D,Normalized log intensity of IL1R8probe (A_23_P84344; GSE32018)in DLBCL versus FL, MCL, MALT,NMZL, CLL, and lymph nodesamples. E, Kaplan–Meier survivalcurve of DLBCL patients (n¼ 98)with low and high IL1R8 geneexpression (218921_at probe)within DLBCL specimens (HR, 2.2;95% CI, 1.2–3.8; P¼ 0.006). F,Normalized log intensity of IL-37probe (218921_at; GSE43677) inDLBCL versus normal B cells (naiveB cells, germinal center (GC) Bcells, post GC B cells and tonsilsamples).G, Normalized logintensity of IL-37 probe(A_23_P84344; GSE32018) inDLBCL vs. FL, MCL, MALT, NMZL,CLL and lymph node samples.
Riva et al.
Cancer Immunol Res; 7(6) June 2019 Cancer Immunology Research882
immunosuppression, as well as disease-specific mechanisms andmechanisms unique to lymphoma subtype (8). In the currentstudy, we show that the Il1r8�/�/lpr mouse model recapitulatedautoimmunity-associated lymphomagenesis, suggesting that theabsence of a negative regulator of the ILR- or TLR-MyD88 axis inan autoimmune-prone background is sufficient for the neoplastictransformation of B cells. These results are in line with data inhumans, showing that aggressive B-cell lymphomas (particularlyDLBCL) are more frequently associated with autoimmune con-ditions thanmore indolent lymphomas (particularly FL; ref. 4). Itshould be noted that IL1R8 deficiency in this model was notrestricted to B cells and might also impact on antitumor immu-nity, as shown in other models (27, 28). Therefore, our resultsmay underestimate the effect of selective IL1R8 deficiency intumor B cells.
Western blot analysis of p52 and phospho-p65 demonstrat-ed that both canonical and noncanonical NF-kB pathways areconstitutively activated in Il1r8þ/þ/lpr mice. Moreover, theTLR-induced NF-kB–regulated Nfkbiz gene was constitutivelyactivated in Il1r8�/� mice. A combination of these pathwaysmay contribute to the activation of distinct NF-kB target genesand B-cell activation observed in lymphoma prone Il1r8�/�/lprmice. We previously observed that Il1r8�/�/lpr B cells had anincreased proliferation rate after stimulations with autoanti-gens acting through TLR7, TLR9, and other TLR ligands, com-pared with lpr B cells (16). In the present article, we described ahigh mitotic rate and diffusely elevated Ki67 immunoreactivityin Il1r8�/�/lpr spleen, indicating an increased proliferationrate, associated with immunoreactivity for Bcl-2, suggestingactivation of the antiapoptotic machinery. Thus, the resultspresented here are in line with the view that the lack of a tunerof TLR and IL1R signaling–dependent NF-kB activationcould impinge upon B-cell transformation in the contextof lymphoproliferative syndrome. Indeed, other oncogenicevents circumventing negative feedback mechanisms thatattenuate NF-kB signaling, such as inactivation of the deubi-quitinase A20, are associated with autoimmunity and lym-phoma development (40, 53).
DLBCLs developed in Il1r8�/�/lpr mice were characterized bythe presence of a monomorphic population of large B cells inlymphoid tissues. Neoplastic B cells displayed high proliferationrates and showed widespread involvement of distant organsincluding gut, liver, lung, and kidney. Histopathologic analysisand immunostaining for CD3, B220, Bcl-2, and Ki67 of spleenand lymph node specimens of Il1r8�/�/lpr mice documentedsharply separated masses constituted by DLBCL arising withina background characterized by atypical lymphoproliferative dis-order. Excessive lymphoproliferation associated with activationof antiapoptotic mechanisms were potentially responsible formultiple independent transformation events resulting in thepolyclonal (or oligoclonal) development of different primaryfoci of DLBCL, as suggested by the detection of different bandsof IgH rearrangement in the same mouse or in the same organ.Our results suggest that FAS deficiency was responsible forpolyclonal B-cell expansion and very rarely lymphoma transfor-mation, and the addition of the deficiency of IL1R8, causinghyperactivation of the MyD88–NF-kB axis, in response to auto-antigens, resulted in tumors that resemble humanABC-DLBCL. InIl1r8�/�/lpr mice, these tumors emerged sporadically and thuswere likely to have acquired the additional oncogenic hits nec-essary to give rise to DLBCL.
In the present study, we show that the expression of IL1R8 inhuman DLBCL was downmodulated compared with peripheralblood or GC B cells from healthy donors and correlated withoverall survival. The molecular mechanisms underlying IL1R8downmodulation in humanDLBCL are still undefined, and couldinclude promoter methylation, as observed in human gastriccarcinomas (54), alternative splicing leading to aberrant proteinexpression, as described in colorectal cancer (55), or promoterhypoacetylation as suggested by the analysis of hematologiccancer cell lines frompublicly available data sets (Ensembl, UCSCGenome Browser). In addition, it was reported that genomicmethylation affects IL1R8 expression, as azacytidine treatmentof CLL cell lines restored IL1R8 mRNA expression (56). Veryrarely, nonsense and somatic nonsynonymous mutations havebeen observed in the IL1R8 coding sequence as shown by Whole-Exome Sequencing data from The Cancer Genome Atlas and insequenced samples within DLBCL patients (Dalla-Favera R., per-sonal communication), but the functional consequences of thesemutations or polymorphisms need to be investigated. The apo-ptotic process and DNA damage response were among the path-ways significantly enriched by genes positively correlated withIL1R8 gene expression. This suggests that higher IL1R8 expressionmight be associated with increased apoptotic activity and bettercontrol of DNA damage, and as a consequence, with a lessaggressive phenotype of lymphoma cells, thus leading to betterprognosis.
Patients affected by DLBCL show different clinical courses,making prediction of prognosis and successful therapy difficult,leading to only 50%of patients being effectively treated (57). Ourresults demonstrate that IL1R8 activity limits B-cell activation andmalignant transformation induced by autoimmune stimulationand contribute to the identification of genes and molecular path-ways that could represent targets for novel therapeutic approachesin DLBCL treatment.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors' ContributionsConception and design: C. Garlanda, F. Riva, M.T.S. Bertilaccio, A. Mantovani,M. MuzioDevelopment of methodology: F. Riva, D. Supino, N. Polentarutti,A. Innocenzi, A. Anselmo, M. MuzioAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.):F. Riva, D. Supino, M.T.S. Bertilaccio, M. Massara,F. Pasqualini, A. Innocenzi, A. Anselmo, F. Caligaris-Cappio, M. MuzioAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Garlanda, F. Riva, M. Ponzoni, M.T.S. Bertilaccio,M. Massara, R. Carriero, M. MuzioWriting, review, and/or revision of the manuscript: C. Garlanda, F. Riva,M. Ponzoni, H.-J. Anders, M. MuzioAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): N. Polentarutti, F. Pasqualini, H.-J. AndersStudy supervision: C. Garlanda, F. Riva, M. MuzioOther (helped with experimental work): G. Simonetti, T. Veliz-Rodriguez
AcknowledgmentsThe study was supported by the European Commission (ERC project PHII-
669415; FP7 project TIMER HEALTH-F4-2011-281608), Ministero dell'Istru-zione, dell'Universit�a e della Ricerca (MIUR; project PRIN 2015YYKPNN;project FIRB RBAP11H2R9), Associazione Italiana Ricerca sul Cancro (AIRCIG 19014 to A. Mantovani and AIRC 5 � 1000 9962 to A. Mantovani andC. Garlanda; AIRC IG 16777 and 13042 to M. Muzio; AIRC 5 � 1000 9965to M. Muzio and F. Caligaris-Cappio), CARIPLO (project 2010-0795 to
IL1R8 in Autoimmunity-Associated B-cell Lymphoma
www.aacrjournals.org Cancer Immunol Res; 7(6) June 2019 883
F. Caligaris-Cappio, M. Muzio, and C. Garlanda), the Italian Ministry of Health(Ricerca Finalizzata, RF-2013-02355470 to C. Garlanda), and the DeutscheForschungsgemeinschaft (AN 372/24-1 and 27-1 to H.-J. Anders).
The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked
advertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.
Received October 6, 2018; revised January 28, 2019; accepted April 17, 2019;published first April 24, 2019.
References1. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation.
Nature 2008;454:436–44.2. Balkwill F, Charles KA, Mantovani A. Smoldering and polarized inflam-
mation in the initiation and promotion of malignant disease. Cancer Cell2005;7:211–7.
3. Coussens LM, Zitvogel L, Palucka AK. Neutralizing tumor-promotingchronic inflammation: a magic bullet? Science 2013;339:286–91.
4. GronbaekK, StratenPT, Ralfkiaer E, Ahrenkiel V, AndersenMK,HansenNE,et al. Somatic Fas mutations in non-Hodgkin's lymphoma: associationwith extranodal disease and autoimmunity. Blood 1998;92:3018–24.
5. Dias C, Isenberg DA. Susceptibility of patients with rheumatic diseases toB-cell non-Hodgkin lymphoma. Nat Rev Rheumatol 2011;7:360–8.
6. Poppema S, Maggio E, van den Berg A. Development of lymphoma inautoimmune lymphoproliferative syndrome (ALPS) and its relationship toFas gene mutations. Leuk Lymphoma 2004;45:423–31.
7. Carbone A, Tripodo C, Carlo-Stella C, Santoro A, Gloghini A. The role ofinflammation in lymphoma. Adv Exp Med Biol 2014;816:315–33.
8. Baecklund E, Smedby KE, Sutton LA, Askling J, Rosenquist R. Lymphomadevelopment in patients with autoimmune and inflammatory disorders—what are the driving forces? Semin Cancer Biol 2014;24:61–70.
9. Shaffer AL, Rosenwald A, Staudt LM. Lymphoidmalignancies: the dark sideof B-cell differentiation. Nat Rev Immunol 2002;2:920–32.
10. Pasqualucci L, Dalla-Favera R. The genetic landscape of diffuse large B-celllymphoma. Semin Hematol 2015;52:67–76.
11. Singer GG, Carrera AC, Marshak-Rothstein A, Martinez C, Abbas AK.Apoptosis, Fas and systemic autoimmunity: the MRL-lpr/lpr model.Curr Opin Immunol 1994;6:913–20.
12. Fisher GH, Rosenberg FJ, Straus SE, Dale JK, Middleton LA, Lin AY, et al.Dominant interfering Fas gene mutations impair apoptosis in a humanautoimmune lymphoproliferative syndrome. Cell 1995;81:935–46.
13. Garlanda C, Dinarello CA, Mantovani A. The interleukin-1 family: back tothe future. Immunity 2013;39:1003–18.
14. GarlandaC, Riva F, Polentarutti N, Buracchi C, SironiM,DeBortoliM, et al.Intestinal inflammation inmice deficient in Tir8, an inhibitory member ofthe IL-1 receptor family. Proc Natl Acad Sci U S A 2004;101:3522–6.
15. Wald D, Qin J, Zhao Z, Qian Y, Naramura M, Tian L, et al. SIGIRR, anegative regulator of Toll-like receptor-interleukin 1 receptor signaling.Nat Immunol 2003;4:920–7.
16. Lech M, Kulkarni OP, Pfeiffer S, Savarese E, Krug A, Garlanda C, et al.Tir8/Sigirr prevents murine lupus by suppressing the immunostimulatoryeffects of lupus autoantigens. J Exp Med 2008;205:1879–88.
17. BulekK, Swaidani S,Qin J, Lu Y,GulenMF,Herjan T, et al. The essential roleof single Ig IL-1 receptor-related molecule/Toll IL-1R8 in regulation of Th2immune response. J Immunol 2009;182:2601–9.
18. Nold-Petry CA, Lo CY, Rudloff I, Elgass KD, Li S, Gantier MP, et al. IL-37requires the receptors IL-18Ralpha and IL-1R8 (SIGIRR) to carry out itsmultifaceted anti-inflammatory program upon innate signal transduction.Nat Immunol 2015;16:354–65.
19. XiaoH,GulenMF,Qin J, Yao J, Bulek K, KishD, et al. The Toll-interleukin-1receptor member SIGIRR regulates colonic epithelial homeostasis, inflam-mation, and tumorigenesis. Immunity 2007;26:461–75.
20. Drexler SK, Kong P, Inglis J, Williams RO, Garlanda C, Mantovani A, et al.SIGIRR/TIR-8 is an inhibitor of Toll-like receptor signaling in primaryhuman cells and regulates inflammation inmodels of rheumatoid arthritis.Arthritis Rheum 2010;62:2249–61.
21. Gulen MF, Kang Z, Bulek K, Youzhong W, Kim TW, Chen Y, et al. Thereceptor SIGIRR suppresses Th17 cell proliferation via inhibition of theinterleukin-1 receptor pathway and mTOR kinase activation. Immunity2010;32:54–66.
22. Russell SE, Stefanska AM, KubicaM, Horan RM,Mantovani A, Garlanda C,et al. Toll IL-1R8/single Ig IL-1-related receptor regulates psoriasiform
inflammation through direct inhibition of innate IL-17A expression bygammadelta T cells. J Immunol 2013;191:3337–46.
23. Batliwalla FM, Li W, Ritchlin CT, Xiao X, Brenner M, Laragione T, et al.Microarray analyses of peripheral blood cells identifies unique geneexpression signature in psoriatic arthritis. Mol Med 2005;11:21–9.
24. Garlanda C, Riva F, Veliz T, Polentarutti N, Pasqualini F, Radaelli E, et al.Increased susceptibility to colitis-associated cancer ofmice lacking TIR8, aninhibitory member of the interleukin-1 receptor family. Cancer Res 2007;67:6017–21.
25. XiaoH, YinW, KhanMA, GulenMF, ZhouH, ShamHP, et al. Loss of singleimmunoglobulin interlukin-1 receptor-related molecule leads toenhanced colonic polyposis in Apc(min) mice. Gastroenterology 2010;139:574–85.
26. BertilaccioMT, Simonetti G, Dagklis A, RocchiM, Rodriguez TV, ApollonioB, et al. Lack of TIR8/SIGIRR triggers progression of chronic lymphocyticleukemia in mouse models. Blood 2011;118:660–9.
27. Campesato LF, Silva APM, Cordeiro L, Correa BR, Navarro FCP, Zanin RF,et al. High IL-1R8 expression in breast tumors promotes tumor growth andcontributes to impaired antitumor immunity. Oncotarget 2017;8:49470–83.
28. MolgoraM, Bonavita E, Ponzetta A, Riva F, BarbagalloM, Jaillon S, et al. IL-1R8 is a checkpoint in NK cells regulating anti-tumour and anti-viralactivity. Nature 2017;551:110–4.
29. Leadbetter EA, Rifkin IR, Hohlbaum AM, Beaudette BC, ShlomchikMJ, Marshak-Rothstein A. Chromatin-IgG complexes activate B cellsby dual engagement of IgM and Toll-like receptors. Nature 2002;416:603–7.
30. Means TK, Latz E,Hayashi F,MuraliMR,GolenbockDT, Luster AD.Humanlupus autoantibody-DNA complexes activate DCs through cooperation ofCD32 and TLR9. J Clin Invest 2005;115:407–17.
31. ScielzoC, BertilaccioMT, SimonettiG,Dagklis A, tenHackenE, Fazi C, et al.HS1 has a central role in the trafficking and homing of leukemic B cells.Blood 2010;116:3537–46.
32. Fonte E, Agathangelidis A, ReverberiD,Ntoufa S, Scarfo L, Ranghetti P, et al.Toll-like receptor stimulation in splenic marginal zone lymphoma canmodulate cell signaling, activation andproliferation.Haematologica 2015;100:1460–8.
33. Adachi M, Watanabe-Fukunaga R, Nagata S. Aberrant transcriptioncaused by the insertion of an early transposable element in an intronof the Fas antigen gene of lpr mice. Proc Natl Acad Sci U S A 1993;90:1756–60.
34. Bichi R, Shinton SA, Martin ES, Koval A, Calin GA, Cesari R, et al. Humanchronic lymphocytic leukemia modeled in mouse by targeted TCL1expression. Proc Natl Acad Sci U S A 2002;99:6955–60.
35. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powersdifferential expression analyses for RNA-sequencing and microarray stud-ies. Nucleic Acids Res 2015;43:e47.
36. SubramanianA, TamayoP,MoothaVK,Mukherjee S, Ebert BL,GilletteMA,et al. Gene set enrichment analysis: a knowledge-based approach forinterpreting genome-wide expression profiles. Proc Natl Acad Sci U S A2005;102:15545–50.
37. Bernasconi NL, Traggiai E, Lanzavecchia A. Maintenance of serologicalmemory by polyclonal activation of humanmemory B cells. Science 2002;298:2199–202.
38. De Groof A, Hemon P, Mignen O, Pers JO, Wakeland EK, RenaudineauY, et al. Dysregulated lymphoid cell populations in mouse models ofsystemic lupus erythematosus. Clin Rev Allergy Immunol 2017;53:181–97.
39. Compagno M, Lim WK, Grunn A, Nandula SV, Brahmachary M, Shen Q,et al. Mutations of multiple genes cause deregulation of NF-kappaB indiffuse large B-cell lymphoma. Nature 2009;459:717–21.
Riva et al.
Cancer Immunol Res; 7(6) June 2019 Cancer Immunology Research884
40. Ngo VN, Young RM, Schmitz R, Jhavar S, Xiao W, Lim KH, et al. Onco-genically active MYD88 mutations in human lymphoma. Nature 2011;470:115–9.
41. Nogai H,Wenzel SS, Hailfinger S, GrauM, Kaergel E, Seitz V, et al. IkappaB-zeta controls the constitutive NF-kappaB target gene network and survivalof ABC DLBCL. Blood 2013;122:2242–50.
42. Molgora M, Barajon I, Mantovani A, Garlanda C. Regulatory role of IL-1R8in immunity and disease. Front Immunol 2016;7:149.
43. Piva R, Pellegrino E, Mattioli M, Agnelli L, Lombardi L, Boccalatte F, et al.Functional validation of the anaplastic lymphoma kinase signature iden-tifies CEBPB and BCL2A1 as critical target genes. J Clin Invest 2006;116:3171–82.
44. Minguet S, Dopfer EP, Pollmer C, Freudenberg MA, Galanos C, Reth M,et al. Enhanced B-cell activationmediated by TLR4 and BCR crosstalk. Eur JImmunol 2008;38:2475–87.
45. Jiang W, Lederman MM, Harding CV, Rodriguez B, Mohner RJ, Sieg SF.TLR9 stimulation drives naive B cells to proliferate and to attainenhanced antigen presenting function. Eur J Immunol 2007;37:2205–13.
46. Li S, Neff CP, Barber K, Hong J, Luo Y, Azam T, et al. Extracellular forms ofIL-37 inhibit innate inflammation in vitro and in vivo but require the IL-1family decoy receptor IL-1R8. Proc Natl Acad Sci U S A 2015;112:2497–502.
47. Clementi R, Dagna L, Dianzani U, Dupre L, Dianzani I, Ponzoni M, et al.Inherited perforin and fas mutations in a patient with autoimmunelymphoproliferative syndrome and lymphoma. N Engl J Med 2004;351:1419–24.
48. Sangaletti S, Tripodo C, Vitali C, Portararo P, Guarnotta C, Casalini P, et al.Defective stromal remodeling and neutrophil extracellular traps in lym-
phoid tissues favor the transition from autoimmunity to lymphoma.Cancer Discov 2014;4:110–29.
49. Lim KH, Yang Y, Staudt LM. Pathogenetic importance and therapeuticimplications of NF-kappaB in lymphoid malignancies. Immunol Rev2012;246:359–78.
50. Pasqualucci L, Dalla-Favera R. Genetics of diffuse large B-cell lymphoma.Blood 2018;131:2307–19.
51. Qin J, Qian Y, Yao J, Grace C, Li X. SIGIRR inhibits interleukin-1 receptor-and toll-like receptor 4-mediated signaling through different mechanisms.J Biol Chem 2005;280:25233–41.
52. Henry CJ, Casas-Selves M, Kim J, Zaberezhnyy V, Aghili L, Daniel AE, et al.Aging-associated inflammation promotes selection for adaptive oncogenicevents in B cell progenitors. J Clin Invest 2015;125:4666–80.
53. Kato M, Sanada M, Kato I, Sato Y, Takita J, Takeuchi K, et al. Frequentinactivation of A20 in B-cell lymphomas. Nature 2009;459:712–6.
54. Liu Z, Zhang J, Gao Y, Pei L, Zhou J, Gu L, et al. Large-scale characterizationof DNA methylation changes in human gastric carcinomas with andwithout metastasis. Clin Cancer Res 2014;20:4598–612.
55. Zhao J, Bulek K, Gulen MF, Zepp JA, Karagkounis G, Martin BN, et al.Human colon tumors express a dominant-negative form of SIGIRR thatpromotes inflammation and colitis-associated colon cancer in mice.Gastroenterology 2015;149:1860–71.
56. Vilia MG, Fonte E, Veliz Rodriguez T, Tocchetti M, Ranghetti P, Scarfo L,et al. The inhibitory receptor toll interleukin-1R 8 (TIR8/IL-1R8/SIGIRR) isdownregulated in chronic lymphocytic leukemia. Leuk Lymphoma 2017;58:2419–25.
57. Campo E, Swerdlow SH, Harris NL, Pileri S, Stein H, Jaffe ES. The 2008WHO classification of lymphoid neoplasms and beyond: evolving con-cepts and practical applications. Blood 2011;117:5019–32.
www.aacrjournals.org Cancer Immunol Res; 7(6) June 2019 885
IL1R8 in Autoimmunity-Associated B-cell Lymphoma
CANCER RESEARCH | TUMOR BIOLOGYAND IMMUNOLOGY
Lymphoma Angiogenesis Is Orchestrated byNoncanonical Signaling PathwaysMarleen Gloger1, Lutz Menzel1, Michael Grau2, Anne-Clemence Vion3, Ioannis Anagnostopoulos4,Myroslav Zapukhlyak2, Kerstin Gerlach1, Thomas Kammert€ons5, Thomas Hehlgans6, Maria Zschummel7,Georg Lenz2, Holger Gerhardt3, Uta E. H€opken7, and Armin Rehm1
ABSTRACT◥
Tumor-induced remodeling of the microenvironment relies onthe formation of blood vessels, which go beyond the regulation ofmetabolism, shaping amaladapted survival niche for tumor cells. Inhigh-grade B-cell lymphoma, angiogenesis correlates with poorprognosis, but attempts to target established proangiogenic path-ways within the vascular niche have been inefficient. Here, weanalyzed Myc-driven B-cell lymphoma–induced angiogenesis inmice. A few lymphoma cells were sufficient to activate the angio-genic switch in lymph nodes. A unique morphology of densemicrovessels emergedwithout obvious tip cell guidance and relianceon blood endothelial cell (BEC) proliferation. The transcriptionalresponse of BECs was inflammation independent. ConventionalHIF1a or Notch signaling routes prevalent in solid tumors were notactivated. Instead, a nonconventional hypersprouting morphologywas orchestrated by lymphoma-provided VEGFC and lymphotoxin(LT). Interference with VEGF receptor-3 and LTb receptor signal-ing pathways abrogated lymphoma angiogenesis, thus revealingtargets to block lymphomagenesis.
Significance: In lymphoma, transcriptomes and morphogenicpatterns of the vasculature are distinct from processes in inflam-mation and solid tumors. Instead, LTbR and VEGFR3 signalinggain leading roles and are targets for lymphomagenesis blockade.
Graphical Abstract: http://cancerres.aacrjournals.org/content/canres/80/6/1316/F1.large.jpg.
A unique angiogenic milieu exists in lymph nodes, whereby VEGFR3 and LTβ receptor take the lead in lymphoma angiogenesis
when VEGFA availability is disturbed.
ECMFRC
VEGFA
HEV
VEGFR2
VEGFR3
VEGFC
VEGFA
2. Proliferation
1. Induction
Notch
LTβR
LTα1/β2
Hif1α↔
Collagenfibers
Eµ-Myc
Eµ-Myc
Eµ-Myc
IntroductionThe concept of reciprocal cross-talk between malignant B cells and
their microenvironment in secondary lymphoid organs (SLO) issupported by genetic signatures found in high-grade B-cell non-Hodgkin lymphoma (B-NHL). Gene expression profiling (GEP) ofprimary diffuse large B-cell lymphoma (DLBCL) samples showed thatdifferences in the microenvironment affect patient survival afterchemotherapy treatment. The stromal-1 signature, related to extra-cellular matrix (ECM) deposition and histiocytic infiltration, is asso-ciated with favorable outcome, whereas the angiogenesis-related sig-nature (stromal-2 signature) is prognostically unfavorable (1). Themicroanatomic correlate of this angiogenesis-related signature is anincreased microvessel density (MVD; refs. 2, 3).
An essential structural and functional component of the lymphomastroma consists of the blood and lymphatic vasculature, which besidesservicing the local cells, shapes the tumor survival niche. Antiangio-genic strategies are based on the activities of VEGF familymolecules incancer-associated angiogenesis (4, 5). Lymphoma cells themselvesexhibit proangiogenic activity as they release VEGFs, but attempts tocombinemultimodal chemo-/immunotherapies withVEGF inhibitorshave not been beneficial in B-cell lymphoma (6, 7). Therefore, it isplausible that other non-VEGF angiogenic pathways prevail and causeenhanced vascular assembly. Lymph nodes (LN) are equipped with
1Translational Tumorimmunology, Max Delbr€uck Center for Molecular Medicine,Berlin, Germany. 2Department of Medicine A, and Cluster of Excellence EXC1003, University Hospital M€unster, M€unster, Germany. 3Integrative VascularBiology Lab, Max Delbr€uck Center for Molecular Medicine, Berlin, Germany.4Institute of Pathology, Charit�e-University Medicine Berlin, Berlin, Germany.5Institute of Immunology, Charit�e -University Medicine Berlin, Berlin, Germany.6Regensburg Center for Interventional Immunology, University HospitalRegensburg, Regensburg, Germany. 7Microenvironmental Regulation in Auto-immunity and Cancer, Max Delbr€uck Center for Molecular Medicine, Berlin,Germany.
Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).
M. Gloger and L. Menzel contributed equally to this article.
Corresponding Authors: Armin Rehm, Max Delbr€uck Center for MolecularMedicine, Robert-R€ossle-Str., Berlin 13125, Germany. Phone: 0049-30-9406-3817; Fax: 0049-30-9406-3124; E-mail: [email protected]; and Uta E.H€opken, [email protected]
Cancer Res 2020;80:1316–29
doi: 10.1158/0008-5472.CAN-19-1493
�2020 American Association for Cancer Research.
AACRJournals.org | 1316
specialized endothelial cell (EC) subsets that line the blood andlymphatic vasculature. A comparison of transcriptomes of fibroblasticreticular cells (FRC), blood ECs (BEC), and lymphatic ECs (LEC)taken from na€�ve LNs and LNs exposed to an inflammatory milieushowed marked immune cell recruitment in the latter (8). However, itremains unclear whether exposure to lymphoma has the same impacton stromal cells as inflammation.
Earlier studies analyzing lymphoma-engendered alterations in ECsrelied on syngeneic or xenogeneic lymphoma cells transplanted sub-cutaneously. In these models, marked adaptations of the lymphaticvasculature occurred, which correlated with the increased availabilityof VEGFC from tumor-associated macrophages (9). However, intro-ducing cells subcutaneously or into xenogeneic hosts does not neces-sarily reflect the dissemination behavior of aggressive B-cell lympho-ma (10). Intravenously administered Myc-driven B-cell lymphomacells enter distinctivemicroanatomic compartments of LNs and spleenvia CCR7- and CD62L-regulated mechanisms, indicating that tumorcell immigration ismediated via high endothelial venules (HEVs) (11).HEVs are segments of postcapillary venules and are characterized byspecialized BECs that express chemokines and adhesion moleculesrequired for extravasation of lymphocytes into LN parenchyma (12).Within the niche, lymphoma cells stimulate reciprocal cross-talk withCD45�/gp38þ FRCs in a lymphotoxin b receptor (LTbR)–dependentmanner (13). During an immune response, activated lymphotoxin a(LT)�expressing B cells directly induce differentiation of ECs into aHEV phenotype (14, 15). The precise contribution of morphogeniccells and molecular factors to angiogenesis is likely to vary dependingon the stimuli, and it remains unclear whether lymphoma B-cellexposure engenders a condition that phenocopies high MVD. Here,we explored lymphoma-induced LN vascular reprogramming in anEm-Myc lymphoma-transplanted mouse model. GEP of FRCs, LECs,and BECs and comparison of angiogenic factors expressed by stromaand lymphoma cells revealed that mechanisms different from con-ventional solid tumor angiogenesis are active in B-cell lymphoma–exposed LNs.
Materials and MethodsMice
The full list of mice used is given in Supplementary Materials andMethods. All experiments were conducted in compliance with theinstitutional guidelines of the Max Delbr€uck Center for MolecularMedicine (Berlin, Germany) and approved by the Landesamt f€urGesundheit und Soziales Berlin, Germany (G0104/16; G0052/12;G0373/13; G 0058/19).
Human tissue specimens and IHCFormalin-fixed paraffin-embedded biopsy specimens from DLBCL
of not otherwise specified (NOS) subtype with either a BCL2 or BCL6rearrangement (single hit), high-grade B-cell lymphoma with MYCand BCL2 or BCL6 rearrangements (double hit), from Burkitt lym-phoma (BL) as well as from nonneoplastic palatine tonsils, andnonneoplastic lymph nodes (LN), were retrieved from the archivesof the Institute of Pathology, Charit�e-University Medicine (Berlin,Germany).
Multiple tissue arrays (MTA) were obtained from US Biomax andcontained various DLBCL specimens that were not further diagnosedaccording to cytogenetic rearrangements. IHC was performed asdescribed in SupplementaryMaterials andMethods. The study involv-ing primary human tissues was conducted according to the declarationof Helsinki and in accordance with local ethical guidelines.
Cell lines and primary B-NHL cellsThe human B-NHL cell lines Su-DHL-4 and OCI-Ly7 (DLBCL),
JeKo-1 [mantle cell lymphoma (MCL)] were obtained from DSMZ;Raji cells (Burkitt lymphoma) were from ATCC. Upon receipt, all celllines were expanded over 1 to 2 weeks, and aliquots were immediatelyfrozen in liquid nitrogen. Gene expression analysis was performed 2 to3 days after thawing.Mycoplasma testing was not performed. Humanumbilical vein cells (HUVEC) were purchased from Promocell andcultured in endothelial cell medium over 2 to 3 passages before use.
Patient-derived DLBCL and MCL xenograft samples (PDX) wereobtained from Dana-Farber Cancer Institute (PRoXe depository,Boston, MA) and used after a single NOD.Cg-Prkdcscid Il2rg tm1Wjl/SzJ (NSG) mouse passage. All primary cell lines were directlyobtained from public depositories or from a commercial supplier andnot additionally authenticated before use. Primary B cells (CD19þ
CD45þ CD69� CD4� CD8� CD14� 7AAD�) from healthy donorswere purified from PBMCs via a Ficoll gradient and further sorted byflow cytometry.
HUVEC activation assayHUVECs were starved overnight in serum and growth factor–free
endothelial basal medium (Promocell) at a cell density of 1� 104/cm2
in a 24-well plate. Cell cultures were supplemented with VEGFA165
and VEGFC, in addition LTa1b2 was added (all 10 ng/mL) for24 hours. For VEGFC, a 13,5 kDa non-disulfide linked homodimericprotein consisting of two 116 amino acid polypeptide chains waschosen.
In a spheroid-based angiogenesis assay, HUVECs (at passage 3 forall experiments) were cultured for two days, trypsinized, washed inPBS, and resuspended in basal medium mixed with methocel stocksolution (0.012% in basal medium) in a 4:1 ratio. Twenty-five micro-liters of cell suspension was pipetted dropwise on nonadherent culturedishes, turned upside-down to form hanging drop cultures. Collagenmix was prepared on ice using Collagen type I diluted in 10� PBS in a4:1 ratio, pH 7. Spheroids were obtained after 24 hours, mixed inmethocel–collagen medium, plated in cell culture dishes, and poly-merized for 30 minutes at 37�C. Spheroids were stimulated withVEGFA, VEGFC, LTa1/b2 (all 25 ng/mL), and anti–VEGFR2 anti-body (50 ng/mL, clone: 89106 R&D Systems) for 24 hours. Spheroidswere fixed in 4% PFA, images were recorded with transmissionmicroscopy, and sprouts were analyzed with ImageJ.
Tumor cell transferSingle-cell suspensions were prepared and transferred intravenous-
ly exactly as described previously (13); at least 2 to 6 independentlymphoma clones derived from different Em-Myc mice were tested.Tumor load in recipient mice was determined by spleen weight, or byflow cytometric analysis. Transplantation of MCA313 primary fibro-sarcoma cells was performed by subcutaneous injection of 1� 106 cellsin PBS.
Antibody and in vivo LTbR inhibitor treatmentLTbR-blocking immunoglobulin (LTbR-Ig; 100 mg; Biogen Idec) or
IgG1 isotype control antibody MOPC21 was injected twice intraper-itoneally (100 mg), exactly as described previously (16).
Mice were injected intraperitoneally on days 1 and 6 after Em-Myclymphoma cell administration either with 0.8 mg rat anti–VEGFR2antibody (cloneDC101), orwith isotype control IgG1 (both fromBioXCell). After fibrosarcoma transplantation, once small subcutaneousnodules became palpable, antibodies were injected on two occasions at5-day intervals.
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1317
Pharmacologic inhibitionThe VEGFR3 kinase inhibitor SAR131675 (SelleckChem) was
dissolved in DMSO and further diluted in 0.6% methylcellulose/0.25% Tween 80. Animals were treated for 6 consecutive days with100 mg/kg b.w. of the drug per os, starting at day 2 after Em-Myclymphoma transplantation, or when fibrosarcoma became palpable.
Gene expression profilingLN stromal cells were sorted from Wt controls and from animals
transplanted with Em-Myc lymphoma cells. The average lymphomaload in LNs was 4% to 10% of all lymphocytes. Stromal cell purity aftersorting was >95%.
For each sample, 80 ng RNA was pooled from 4 to 6 independentsorting experiments and reverse-transcribed using the Illumina TotalPrep RNA Amplification Kit. The biotin-labeled transcripts werehybridized to Illumina Mouse WG-6 v2.0 Expression BeadChips andprocessed for detection. Differential expression was evaluated aslog2-fold change (FC). A detailed description of microarray datageneration, qRT-PCR, and bioinformatic processing is given in Sup-plementary Methods and were performed essentially as describedpreviously (16). Data are deposited under GEO repository accessionnumber GSE126033.
Adoptive splenocyte transfer and transmigration assaySplenocytes were labeled with SNARF-1 fluorescent dye (Thermo
Fisher Scientific), and then, 3 � 107 cells per recipient mouse wereinjected intravenously. Four hours after cell transfer, mice weresacrificed, inguinal LNs were dissected, and cells were analyzed byflow cytometry.
In vivo proliferation assayDetection of bromodeoxyuridine (BrdU) incorporation into pro-
liferating ECs inmicewas done essentially as described previously (16).
Vessel permeability and perfusion assayFITC-coupled dextran polymers (10, 40, and 150 kDa; all Sigma-
Aldrich) diluted in PBS were injected intravenously (0.5 mg) intocontrol, Em-Myc cell-transferred, or subcutaneously fibrosarcoma-bearing mice. Vessel perfusion was examined by intravenous injectionof fluorescence-coupled Isolectin GS-IB4 (Thermo Fisher Scientific).
Inguinal LNs or fibrosarcoma tumors were fixed in 4% parafor-maldehyde/PBS for immunohistology.
Statistical analysisStatistical data were evaluated using GraphPad Prism (Version 6)
software. The confidence level was 95%, with a significance level of 5%(a¼ 0.05). Results are expressed as the arithmeticmeans� SEM.Datacomparison with P values of �0.05 was considered statistically sig-nificant. P values were calculated by Wilcoxon signed-rank test,Mann–Whitney U test for nonnormally distributed data, a two-tailed unpaired Student t test for normally distributed data, or a pairedStudent t test, as indicated.
ResultsThemurine Em-Myc lymphomamodelmimics human high-gradeB-NHL–induced stromal angiogenesis
In humans, areas of high MVD were seen in aggressive B-NHLsamples (Fig. 1A andB). Comparedwith nonneoplastic tonsils or LNs,the number of large vessels in theDLBCL samples was not affected, butthe number of small vessels was increased 1.5- to 3-fold (Fig. 1C).
These ratios were higher in high-grade B-NHL (MYC and BCL2 orBCL6 double translocated) and Burkitt lymphoma compared withsingle-hit DLBCL or DLBCL with unknown rearrangements. Smallvessels were rarely PNAd positive (2/30 DLBCL cases), indicating thatthey represented a capillary phenotype, but not HEVs (SupplementaryFig. S1). Functionally, this lower vessel differentiation might result inimpaired immigration of immunoprotective effector T cells (17).
To address themechanisms of lymphoma-induced angiogenesis, weused transgenic Em-Myc mice that spontaneously develop lymphomathat mimic important aspects of human high-grade B-cell lympho-mas (13, 18). Em-Myc lymphoma cells were transferred by intravenousinjection into Wt mice. Between 9 and 12 days after lymphoma celltransfer (mean 5%–15% of all CD45þ cells), LNs rapidly developed a3-fold increase in small vessels compared with an untreated cohort(Fig. 1D and E). This process was already visible by days 5 to 8,indicating that angiogenesis might be an early prerequisite for lym-phoma progression. Anti-CD31 staining of LNs from spontaneouslydiseased Em-Myc transgenic mice with a high tumor burden revealedthe occurrence of a majority of small vessels, which had substantialmorphologic similarities both to LNs from Em-Myc lymphoma-transplanted mice and to human DLBCL (Fig. 1F).
The murine Em-Myc model therefore mimics microanatomicaspects of lymphoma-induced vascular alterations that occur inhuman high-grade B-cell lymphoma.
Transcriptional modifications of LN vascular stromal cells occurearly after Em-Myc lymphoma challenge
Stromal populations from inguinal LNs derived from mice wereidentified by differential expression of gp38, the adhesion moleculeCD31, and the LEC marker Lyve1. The blood vasculature was clearlydistinguishable from lymphatics by the absence of Lyve1 staining and,additionally, by a higher intensity staining of CD31 (Fig. 2A).
FRCs, LECs, and BECs were sorted from stroma-enriched fractions(CD45�) of pooled LNs from untreated and Em-Myc lymphoma–transplanted animals (Fig. 2B). Mice were sacrificed when tumor cellstypically made up less than 10% of leukocytes in the LNs (days 8–12).FRCs exhibited the highest gene expression of the chemokines Ccl19,Ccl21, and of Il7. BECs expressed a high level of ESAM, and LECsexpressed the marker gene Lyve1 (Fig. 2C and D). Using wholetranscriptomeGEP relative changes in gene expression fromuntreatedstromal cells were compared with those from lymphoma-challengedmice. Subset-specific expression of the expected gene markers wasconfirmed (Supplementary Fig. S2A). Gene expression of lymphomaB-cell–specific genes, such as Ptprc (CD45R/B220), Ms4a1 (CD20),and Cxcr5within datasets from stromal subsets was rigorously exclud-ed (Supplementary Fig. S2B).
In all three subsets, there were significant overall differences inexpression between the treated and untreated mice (Fig. 2E). A totalof 710 genes were selectively upregulated in lymphoma-exposedcells. An overlap of 114 genes that were upregulated in all threesubpopulations occurred (Fig. 2F and G). Tumor induction andprogression has been linked to a chronic type of inflammatorymilieu. Therefore, we explored whether lymphoma-dependent stro-mal cells showed expression of genes characteristic of a LN-specificinflammation signature (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc¼GSE15907; ref. 8). Unexpectedly, no significant upre-gulation of inflammation signature genes was observed (Supple-mentary Fig. S2C and S2D).
On the other hand, similar to an inflammatory condition, geneontology (GO) analysis (Supplementary Table S1A) as well asGSEA-confirmed enriched signatures related to proliferation and
Gloger et al.
Cancer Res; 80(6) March 15, 2020 CANCER RESEARCH1318
Figure 1.
Murine Em-Myc B-cell lymphoma phenocopy high microvessel density in human DLBCL and Burkitt lymphoma. A, Anti-CD31 staining of nonneoplastic tonsils andDLBCL. B, Anti-CD34 staining on single tissue sections or on MTAs from high-grade B-cell lymphoma subtypes (n.SLO, nonneoplastic lymph nodes, n ¼ 11; tonsils,n¼ 5; DLBCL, NOSwithout cytogenetic data (mix), n¼ 36; DLBCL (single hit), n¼ 13; high-grade B-cell lymphoma (double hit), n¼ 3; Burkitt lymphoma, n¼ 8); threelow-power fields per case and tissue slide were counted. C, Vessels were grouped according their sizes, small <30 mm and large >30 mm. D,Wt mice (n > 10) wereadoptively transferred with Em-Myc lymphoma cells intravenously. At least three clones were used. Recipients were sacrificed early, day 5–8 (low tumorload, n ¼ 6) and late, day 9–12 (medium, n ¼ 5; control, n ¼ 4). LNs stained with an anti-CD31 antibody (day 9). One to four sections per LN were quantitativelyevaluated. E, Vessels were grouped as in C. F, LN from control mice (n ¼ 3) and from diseased transgenic Em-Myc transgenic mice (n ¼ 4) were stained as inD, inset shows amplification of MVDs in lymphoma bearing animals. In C and E, the ratios between small and large vessels are represented; error bars, mean� SEM;Student t test. Scale bar, 100 mm. Asterisks, statistical significance. � , P � 0.01; �� , P � 0.01; ��� , P � 0.001; n.s., nonsignificant.
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1319
cell cycle in all three stromal cell subsets (Supplementary Fig. S3A–S3C; Supplementary Table S1B). Phenotypically, the proportion ofBECs, LECs and FRCs in S-phase (EdUþ) was substantially higherthan in those from untreated LNs (Supplementary Fig. S3D).Collectively, the angiogenic switch in lymphoma-bearing LNs ischaracterized by proliferation of BECs. Gene expression and pro-liferation data confirmed the role of lymphangiogenesis duringlymphoma progression. Lymphangiogenesis has been established
as a structural pathogenetic factor supporting lymph flow andlymphoma dissemination (9, 19–21).
As an indication of structural alterations, in FRCs, the KEGG genesets correlated with “Collagen_Formation” and “Extracellular_Matrix_Organization” were significantly enriched with downregu-lated genes (Supplementary Table S1B). Both signatures containgenes (e.g., Collagen genes; Timp1, 2) that determine the formationand stability of the ECM. Among other functions that sustain tissue
Figure 2.
Lymph node vascular stroma cellresponse to Em-Myc lymphoma chal-lenge. A, Immunofluorescent stainingof inguinal LN sections. Images wererecorded by confocal microscopy.Scale bar, 500 mm. A higher magnifi-cation of amarked area is shownon theright; scale bar, 20 mm. B, Flow cyto-metry sorting strategy for the isolationof the vascular stromal cell subsetsFRCs (gp38þ CD31�), LECs (gp38þ
CD31þ), and BECs (gp38�CD31þ) fromCD45� Ter119� stromal cell–enrichedfractions of pooled LNs. Data are rep-resentative of at least 10 independentexperiments, 4–6 tumor cell clones and>10mice per replicate. Numberswithinthe gates indicate percent of stromalcell subsets. C, Anti-ESAM staining(red line) on viable (7-AAD�) CD45�
Ter119� gp38�CD31þBECs.D, Stromalcells from untreated mice were sub-jected to qRT-PCR analysis (pooledRNA from 5 independent flow cytome-try sorting experiments; n¼ 4–5mice/sort). Gene expression depicted rela-tive to Gapdh. E, Volcano plots showdifferential expression of genes inlymphoma versus control (FC log2>1.5; P � 0.01). F, Venn plot depictsgenes that were selectively upregu-lated in tumor-exposed vascular stro-ma cells versus controls (P� 0.001; FClog2 � 1.5). G, A heatmap showingdifferential gene expression of 114jointly upregulated genes; changes areindicated by the color scale. In E–G, aStudent t test was applied.
Gloger et al.
Cancer Res; 80(6) March 15, 2020 CANCER RESEARCH1320
homeostasis, one important function of the ECM includes VEGFAgradient shaping and integrin signaling, which are both involved inregulating the balance between stalk cell proliferation and tip cellmigration in the formation of the vascular branches (22). Thus, thissignature may indicate a perturbance of ECM function regardingguidance of angiogenesis.
BECs assemble into microvessels associated with a negative clinicalprognosis; thus, we focused on blood vessels in greater detail.
Regulation of a proangiogenic milieu in Em-Myc lymphoma–bearing LNs involves noncanonical pathways
Inguinal LNs were removed when tumor load was low (5%–10%of all CD45þ cells) or medium (10%–15%), RNA was extracted andanalyzed using angiogenesis gene-specific qRT-PCR arrays(Fig. 3A). With a medium tumor load, among the strongestupregulation of mRNA compared with nontumor bearing LNs wasnoted for Vegfc (>4-fold); VEGFC is a ligand for VEGFR2 andVEGFR3 and functions as a growth factor for lymphangiogenesisand for BECs (23).
GEP on nontransformed follicular B cells derived from controlmiceand Em-Myc lymphoma cells from transplanted mice was performed.The latter exhibited gene sets characteristically upregulated in humanDLBCL as well (FC > 1, P < 0.05; Supplementary Fig. S4A and S4B).Applying an angiogenesis-related GSEA signature, several strongimmunomodulators and proangiogenic factors were found to besubstantially upregulated in lymphoma cells. This list contained genesencoding secreted proteins such as Lgals1 (Galectin-1) and Lgals9(Galectin-9) involved in suppressing T lymphocytes and niche for-mation (Fig. 3B; ref. 24).
Vascular remodeling at the disease onset was analyzed by confocalmicroscopy on thick tissue sections (100 mm). Because transplantedEm-Myc lymphoma cells home preferentially to the paracortical T-cellzone of LNs in a CCR7-dependent manner, we costained the T-cellarea (CD3þ) and quantified the CD31þ 3D surface area within it. LECs(Lyve1þ) were largely absent from the paracortical T-cell zone;however, they were also found to be expanded (SupplementaryFig. S4C; Fig. 3D). Lymphoma induced an expansion of HEVs in theLNs of both the Em-Myc transgenic and transplanted animals, and theyhad many more small vessels and capillaries extending into the T-cellzone than control LNs (Fig. 3C and D).
During sprouting angiogenesis, growing capillaries are spearheadedby specialized ECs termed “tip cells” (25, 26). We analyzed the BEC-specific gene expression from lymphoma-challenged LNs against arefined signature of tip cell genes; however, no strong enrichment withrespect to the BEC-specific ranking of genes was found (Fig. 3E;Supplementary Fig. S4D). Likewise, the KEGG “VEGF SignalingPathway,” which relates to VEGFA/VEGFR2–induced signaling, wasnot enriched (Fig. 3F; Supplementary Fig. S4E). In contrast, a genesignature of distinct transcriptional responses to VEGFC/VEGFR3stimulation was modestly enriched among the top 10 genes up- ordownregulated in BECs (Fig. 3G; ref. 27).
GSEA indicated a significant enrichment of genes downregulatedby the hypoxia-related HIF1a gene among the genes upregulated inBECs (Fig. 3H), suggesting a lack of activation of this signaling route.Tissue staining using the hypoxia marker pimonidazole was negativein LNs from lymphoma-transplanted mice and scant in LNs fromEm-Myc transgenic mice, which was in striking contrast to the markedhypoxia demonstrated in subcutaneously grown fibrosarcoma tissue(Supplementary Fig. S4F). A Notch target gene signature was down-regulated in BECs (Fig. 3I), indicating the lack of a Notch-dependentfeedback loop from tip to stalk cells.
A stem cell transcriptional pattern was enriched among the BEC-specific genes. A “Stemness signature” (28) of 100 genes was selected,35 of which were upregulated (FC > 1; P � 0.001) among the BEC-specific genes (Fig. 3J). Genes related to spindle assembly and kinet-ochore formation, for example, Bub1, Mad2l1, and Nusap1 weresignificantly different. The forkhead TF Foxm1, which binds to themajority of kinetochore and cell division gene promoters (29), wasupregulated (FC ¼ 2.6), indicating that it has a coordinating functionin lymphoma-induced BEC reprogramming.
Em-Myc lymphoma induces aberrant tip cell and filopodiaformation in LNs
We visualized ECs in LNs from tamoxifen-inducible Cdh5-GFPreporter mice. The vascular surface area increased 2-fold upon lym-phoma induction. Correspondingly, vessel length increased about2-fold (mean control: 16.687 mm; mean Em-Myc: 32.767 mm; P ¼0.05). A complex vascular network intermingled with dense loops withsmaller diameters, and concomitantly, a 3-fold increased frequency ofbranching points was seen (Fig. 4A–C). However, upon normalizationof vessel length, distances between branches and loops were notsignificantly changed, indicating that the leading morphologic effectwas an increase in vascular density. The isolectin GS-IB4 was infusedsystemically. Perfused vessels were quantified using the ratio of alldouble-stained IB4þ/GFPþ vessels relative to all GFPþ vessels. In total,about 80% of the vasculature was found to be open for intraluminalflow (Fig. 4D).
Blood vessels of lymphoma-challenged and untreated LNs weremainly impervious to fluorescent 10, 40, and 150 kDa dextrans,whereas vessels in fibrosarcoma tissue leaked 10 and 40 kDa forms(Supplementary Fig. S5A–S5C). In addition, mice were inoculatedwith Em-Myc B cells, and then transplanted with SNARF-1–labeledsplenocytes. Total numbers of CD4þ and CD8þ T cells, monocytes(CD11bþ), and B lymphocytes (B220þ) were about 2-fold higher inlymphoma-bearing LNs (Supplementary Fig. S5D).
In Cdh5-GFP mice, lymphoma-exposed LNs exhibited numerousslender filopodial bursts that emanated from the lateral sides of vessels,but only rarely extended from the leading edge where tip cells areusually located (Fig. 4E). Several cellular protrusions reminiscent ofblebs normally associated with apoptosis, cell migration, and divisionin ontogeny (30, 31) were observed. Next, to explore tip cell occurrencefurther anti-Esm1 immunostainings were performed (Fig. 4F). Esm1expression was found to be EC (Cdh5-GFPþ) restricted, but ratherdistributed throughout the blood vasculature. This pattern is consis-tent with the observation that Esm1 expression is restricted to tip cellsin the retina, but widely distributed throughout carcinoma vasculaturewith a disturbance of the VEGFA gradient (32). The Notch ligand Dll4was variably expressed in ECswithfilopodia (Supplementary Fig. S5E).Collectively, disseminated localization of Esm1þ sprouting ECstogether with variable Dll4 expression are distinguishable from highlyordered morphologies and functions of tip cells in developing organs.
LTbR signaling is required for Em-Myc lymphoma-inducedangiogenesis in LNs
To determine the role of LTbR–LTab signaling, we transplanted Bcells from Em-Myc and Em-Myc � Lta�/� mice. Using a cell-cyclegene-specific qRT-PCR array, we found 34 of 78 genes upregulated inEm-Myc lymphoma-exposed BECs (Fig. 5A). In contrast, this prolif-eration signature was not obtained when Lta-deficient Em-Myc lym-phoma cells were transplanted (4/78 genes upregulated; Fig. 5B).
When Em-Myc lymphoma-bearing mice received the inhibitorydecoy receptor protein LTbR-Ig, a significant delay in the growth of
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1321
lymphomawas observed. In addition, the vascular surface area (CD31þ)and the proliferation rate (BrdUþ) of BECs were almost normalized tothose from control mice (Fig. 5C–F). In lymphoma-bearing Cdh5-GFPmice, the dense andhyperplastic GFPþ vasculaturewas not visiblewhen
animals were LTbR-Ig-treated (Fig. 5G); in particular, the number ofbranching points was about half that in controls (Fig. 5H and I); loss ofcuboid ECs with a concomitant decrease in luminal PNAd staining wasconfirmed (Supplementary Fig. S6A).
Figure 3.
The proangiogenic milieu in lympho-ma-bearing LNs leads to abnormal ves-sel density and morphology. A, qRT-PCR array analysis with RNA from totalLNs of either control or Em-Myc lym-phoma recipient mice with low andmedium tumor burden (n ¼ 3 indepen-dent experiments; n¼ 10 pooled mice/group). Altered gene expression levelsare depicted as x-fold expression (log2)normalized to control LNs (x-axis, base-line). Bars, means � SEM; t tests wereused to test if means are different from0. � , P � 0.05; �� , P � 0.01; ��� , P �0.001. B, GEP of sorted follicular B cellsfrom controls and Em-Myc lymphomacells from transplanted mice (n ¼ 3independent experiments; n¼ 3 differ-ent Em-Myc clones). A representative“Angiogenesis”-related GSEA signa-ture (top) and a signature of genesencoding secreted proteins is shown(bottom). Expression changes aredepicted according to the color scale.C, Confocal microscopy analysis ofinguinal LN vessels from control andEm-Myc lymphoma-recipient mice atdays 8 to 12. A dashed line indicatesthe T-cell zone. Representative imagesare maximum intensity projections of100-mm z-stacks of anti-CD31–stainedsections. D, Quantification of CD31þ
surface area within the paracorticalT-cell zone (CD3þ); 1 to 2 paracorticalT-cell zones per LN were analyzedand 3D quantitative analysis of CD31þ
surface integrates all z-stacks; n ¼ 4–5 mice per condition. In D, bars aremean values; Mann–Whitney U test.E, F, and H–J, GSEA of microarray datafrom BECs as displayed in Fig. 2 wasperformed. G, A heatmap of genesfrom a VEGFC156S/VEGFR3 gene signa-ture in Em-Myc lymphoma challengedBECs compared with untreated BECs(FC cutoff �0.2 or >0.2). Expressionchanges of 10 most up- or downregu-lated genes are indicated by the colorscale.
Gloger et al.
Cancer Res; 80(6) March 15, 2020 CANCER RESEARCH1322
Figure 4.
Angiocrine effects of Em-Myc lymphoma are characterized by a hypersprouting phenotype.A–C, LN sections (100 mm in z) from control (n¼ 4) or Em-Myc lymphomatransplanted (n¼ 9) Cdh5-GFP reporter mice (tamoxifen induced) were analyzed by confocal microscopy. 3D reconstructions were obtained from z-stacks (4� 4tiles; 63� objective; 540� 540� 100 mm). Shown are maximum intensity projections and quantitative analysis of GFPþ surface area relates to single fields of view.B and C, Number of loops (B) and number of branching points (C) per 4� 4 tiles normalized per vessel length; data points represent distances (mm) between loopsand branching points. Bars aremean values; Mann–WhitneyU test, �� , P�0.01. Scale bar, 100 mm.D,Vascular perfusion ofCdh5-GFP reportermice transplantedwithEm-Myc lymphoma cells (days 9–11 after lymphoma cell injection) was determined by intravenously injecting 100 mg isolectinB4 (IB4). Representative fluorescentimages of perfused and nonperfused blood vessels with isolectin (IB4, red) and Cdh5-GFP (green) are presented on the top; quantification of patent loops andcapillaries (GFPþ/IB4þ vessels relative to total GFPþ vessels) is given below. Bar graphs represent mean values for size-grouped vessels, and numbers in the barsindicate the number of vessels analyzed. n > 2 independent experiments; n¼ 5 animals analyzed. E, Visualization of atypical tip cells, aberrant filopodia, and blebs inLNs from untreated (middle) and lymphoma-recipient Cdh5-GFP reporter mice (right). Representative images from at least n¼ 4 mice per group are shown. Scalebar, 10 mm. In A–E, all vascular analysis refers to the T-cell zone. F, Anti-ESM1 staining of LN vasculature from lymphoma-recipient Cdh5-GFP reporter mice. At leastfour mice were investigated; shown are representative images. Scale bar, 5 and 25 mm.
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1323
Figure 5.
BEC-specific LTbR signaling is crucial for aggressive lymphoma-induced LN angiogenesis. A and B, Em-Myc and Em-Myc � Lta�/� tumor cells were intravenouslytransferred intoWt recipients (n¼ 3 independent experiments; n > 10 mice/group; control, n¼ 10 mice). At days 8 to 12, a qRT-PCR cell cycle–associated gene arraywas performedwith RNA from BECs. Gene expressions were calculated relative to Gapdh; the scatter plot shows all genes with a cutoff P < 0.05 and a FC > 2 (log10)highlighted in red. A Student t test was applied. C, Treatment of Em-Myc lymphoma-transferred mice with 100 mg LTbR-Ig (n¼ 11) intraperitoneally, or with controlmouse IgG1 (MOPC21; n¼ 11). Tumor load of LNs was assessed at days 8 to 12 by flow cytometry. Bars indicate the ratios of tumor cells in controls, set arbitrarily at 1,and LTbR-Ig–treated animals. n¼ 5 independent experiments;Wilcoxon signed-rank test.D, Immunostaining of inguinal LN vessels (CD31þ) fromMOPC21 and LTbR-Ig–treated lymphoma-bearingmice. Scale bar, 500 mm. Representative images are maximum intensity projections of 100 mm z-stacks. Analysis of the vasculature inD–J in the paracortical T-cell zone. E,CD31þ surface expression; n¼ 4 independent experiments, with 5 to 6 animals per group, and 1 to 2 T-cell zones per section andper LN analyzed. Bars, mean values; Student t test. F, Lymphoma recipient mice were injected with BrdU for 3 days. CD31þ gp38� BECs were analyzed by flowcytometry for BrdUþ uptake relative to MOPC controls, set arbitrarily to 1 (n ¼ 6 independent experiments, with 4–5 mice pooled per group and per condition);Wilcoxon signed-rank test.G,Em-Myc lymphomaB cellswere transferred intoCdh5-GFPmice. Animalswere treated as inC (LTbR-Ig, n¼ 7;MOPC21, n¼6). At days 8to 12, LN sectionswere analyzed.H and I,Quantitation of vessels based on 3D reconstruction analysis of confocal imaging z-stacks (4� 4 tiles, 63� objective; 540�540� 100mm). Shownaremaximum intensity projections. Scale bar, 100mm.Graphs showing endothelial surface areas (GFPþ) andnumber of branching points.n>3independent experiments. J,Cdh5-Cre/ERT2 and tamoxifen-induced Cdh5-Cre/ERT2 x LtbRfl/flmicewere challenged with lymphoma B cells. Tumor load of LNswasquantified at days 8 to 12; n¼ 8–9 animals; n¼ 2 independent experiments.H–J,Bars, mean values; Mann–WhitneyU test.K,Vascular surfacewas quantitated as in E.Bars, mean values; Student t test.
Gloger et al.
Cancer Res; 80(6) March 15, 2020 CANCER RESEARCH1324
To exclude that the growth reduction was caused by indirect effectson other LTbR-expressing cell populations (33), Cdh5-cre/ERT2 �LtbRfl/fl mice were generated. Reduction in LTbR expression in BECsfrom tamoxifen-treated mice was confirmed by flow cytometry (Sup-plementary Fig. S6B and S6C). After administration of lymphomacells, a substantial decrease inCD31þ vascular surface area, tumor loadand ICAM1 expression was observed compared with Cdh5-cre/ERT2controls (Fig. 5J and K; Supplementary Fig. S6D).
VEGFR3 signaling is induced by lymphoma growth andstimulates LN angiogenesis
We examined the growth kinetics of Em-Myc lymphoma in micetreated with a blocking anti–VEGFR2 antibody. Compared withcontrols, similar lymphoma growth rates, identical vascular surfaceareas (CD31þ), as well as branching points were obtained inlymphoma-exposed LNs (Fig. 6A–C). In contrast, anti–VEGFR2antibody treatment caused a substantial tumor and vascular growthretardation in subcutaneously grown fibrosarcoma (SupplementaryFig. S7A–S7C).
Next, we separated LN stromal cell subsets, Em-Myc lymphomacells, and nontransformed leukocytes. Upregulation of Vegfc was seenin FRCs, but the strongest gene expressionwas seen in the lymphomaBcells (Fig. 6D). Vegfd gene expression was reduced (SupplementaryFig. S7D). Correspondingly, VEGFC serum levels weremuch higher inEm-Mycmice (Fig. 6E). In accordance, several human lymphoma celllines and the patient-derived lymphoma samples (PDX) expressedmuch higher VEGFA and VEGFC mRNAs compared with normal Bcells (Supplementary Fig. S7E).
Stimulation of spheroid-grown HUVECs by VEGFC–inducedsprouts at lower rate than treatment with VEGFA (Fig. 6F). AlthoughLTa1b2 alone hadno stimulatory effect on sprouting, the combinationwith VEGFC revealed a synergistic growth effect (mean 25%). Thiseffect was not seen in the combination of LTa1b2 with VEGFA(Fig. 6F). Stimulation of HUVECs with VEGFC caused a strongVEGFR3 downregulation, which could be substantially amelioratedby a concomitant LTa1b2 stimulation (Fig. 6G). We conclude thatLTbR signaling might interfere with VEFR-3 endocytosis or recycling,a general mechanism of receptor tyrosine kinase regulation (34).
In whole-tissue homogenates from lymphoma-bearing LNs, a4-fold upregulation of VEGFC protein expression was detected. Theactive variant of the VEGFC protein is generated after enzymaticcleavage of a propeptide. Upon lymphoma challenge, there wasincreased conversion to the mature protein in spleen and LNs,indicating that lymphoma had induced enzymatic activities necessaryfor generating biologically active VEGFC (Fig. 6H and I). Geneexpression of collagen- and calcium-binding EGF domains 1 (Ccbe1)was strongly enhanced in Em-Myc lymphoma cells, but not in LNstroma (Fig. 6J). CCBE1 protein localizes together with pro-VEGFCand ADAMTS3 on ECM and EC surfaces, where a cleavage complex isassembled allowing the mature VEGFC to activate VEGFR3 (35).Importantly, fully processed VEGFC is not only lymphangiogenic, butalso angiogenic (36, 37).
VEGFR3 is activated by VEGFC and -D and is highly expressed inleading-edge ECs that undergo sprouting (38). We detected Vegfr3gene expression in BECs and LECs already in steady-state conditions,and observed a 2-fold induction in Vegfr3 expression in LECs uponlymphoma challenge (Fig. 6K). This upregulation is consistent withVEGFR3 signaling in LECs as a driving force for lymphangiogenesis.Vegfr2 mRNA was upregulated in LECs and BECs, which corre-sponded to a modest increase in VEGFR2 protein expression onBECs, but not in LECs (Supplementary Fig. S7F and S7G). To confirm
the VEGFR3 specific pro-proliferative activity in vitro, HUVECsgrown in spheroids were treated with VEGFA, VEGFC, and LTa1b2.In the presence of a VEGFR2–blocking antibody, VEGFC plusLTa1b2 still stimulated the formation of sprouts, whichwas abrogatedin the combination of VEGFA and LTa1b2 (Supplementary Fig. S7H).Next, inducible Vegfr3-GFP reporter mice (Tg(Flt4-tm2.1-cre ERT2)Rosa26-mTmG) were transplanted with lymphoma cells. The highestfrequency of GFP signal was found in LECs, followed by BECs, asconfirmed by anti–VEGFR3 antibody detection (Fig. 6L; Supplemen-tary Fig. S7I). We conclude that in vivo even a small population ofVEGFR3 carrying BECs can gain a leading role in lymphoma-inducedangiogenesis, independent from VEGFR2 activity.
Treatment with the VEGFR3 kinase inhibitor SAR131675, whichhas a much higher selectivity for VEGFR3 than for VEGFR1/2,inhibited both lymphoma progression and vascular expansion stim-ulated by lymphoma (Fig. 6M). A direct antilymphoma effect of thedrug was excluded since spleen located Em-Myc cells were spared(Supplementary Fig. S7J). Concomitantly, expansion of the lymphaticvasculature (Lyve1þ) was retarded (Supplementary Fig. S7K and S7L).Treatment with another VEGFR3 kinase inhibitor, MAZ51 decreasedthe density of the vascular surface (CD31þ) in the T-cell zone 0.5-fold(Supplementary Fig. S7M). In fibrosarcoma, SAR131675 had no effecton tumor development (Fig. 6N).
Taken together, VEGFR3þBECs in angiogenic vessels sensitized thevasculature for proliferation and differentiation inhibition by specifickinase inhibitors.
DiscussionWe elucidated the mechanisms regulating the angiogenic switch
and the morphogenesis of new blood vessels in aggressive B-celllymphoma. Using the transgenic Em-Myc mouse model, we detecteda substantial morphological similarity with human aggressive lym-phoma characterized by a high MVD, a feature that has an adverseprognostic impact on patient survival (2, 3). LymphomaB cells providegrowth stimuli that induce a marked proliferation of BECs. Thislymphoma cell cross-talk with BECs in turn translates to vesselhypersprouting and capillary growth.
The clinical importance of angiogenesis for growth of solid tumorsis well recognized (4, 39, 40). Therapeutic concepts from solid tumorstargeting the VEGFA/VEGFR1/2 axis have been adopted for DLBCLand mantle cell lymphoma combination therapies, resulting in ratherdisappointing clinical outcomes (6, 7). Here, we show that BECproliferation and angiogenesis proceeded in a VEGFR2–independentmanner, a finding that would not have been anticipated from pre-cedents set from studies of solid malignancies. Instead, we identifiedVEGFR3 as the leadingmediator of blood vessel expansion, a signalingprocess fueled by lymphoma cell- and FRC-derived VEGFC. Althougha much smaller fraction of BECs than LECs expressed VEGFR3,treatment with VEGFR3 inhibitors was sufficient to normalize angio-genesis and attenuate lymphoma expansion.
It is worthy to note that a limitation of pharmacologic inhibitoradministration to block receptor tyrosine kinase activity is the spec-ificity of their activity. Selectivity of VEGFR3 drug targetingmight varywith the active concentration of SAR131675, MAZ51, or any othertyrosine kinase inhibitor, respectively (41). In addition, ligand-independent activation of VEGFR3 in angiogenesis may lead toinefficiency of an anti–VEGFR3 antibody treatment (42). In contrast,a BEC-specific genetic deletion of the VEGFR3 would allow moredefinite conclusions regarding the mechanism of lymphoma-inducedangiogenesis. Here, we chose a drug application in our animal model
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1325
Figure 6.
Lymphoma-induced angiogenesis is governed by the VEGFC/VEGFR3 signaling axis. A, Em-Myc B cells were intravenously transferred into Wt recipients. Animalswere administeredwith anti–VEGFR2 antibody orwith an isotype control antibody at a dosage of 100mg intraperitoneally. At days 8 to 11, lymphomacells in LNswerequantified by flow cytometry. (Continued on the following page.)
Gloger et al.
Cancer Res; 80(6) March 15, 2020 CANCER RESEARCH1326
that has the potential to reveal a therapeutic solution directly trans-latable into clinical studies.
The overexpression of the VEGFR3 ligands VEGFC by lymphomacells and macrophages was interpreted to be the inducer of thelymphatic vasculature, but not of blood vessels (9, 21, 43). TheVEGFC/VEGFR3 axis is important for lymphatic sinus growth andincreased lymph flow that correlated with metastasis of solid tumorsand lymphoma cell dissemination, respectively (19, 41). In contrast, weshowed previously that the preferred dissemination route of trans-planted Em-Myc B cells involves the LN T-cell zone, which can beaccessed via HEVs (13). We also demonstrate structural remodelingand transcriptional reprogramming of lymphoma-exposed BECs,which are the major structural constituents of microvessels. To bringlymphangiogenesis and blood vessel growth in agreement, we envisionthat our intravenous transfer model reports on early steps of lym-phoma B-cell homing and niche formation. These early steps aredecisive for further LN remodeling and subsequently, lymphatic sinusgrowth and lymphoma cell dissemination. In LN, lymphatic and bloodvasculature growth proceed in a VEGFR3–dependent manner.
Our results draw parallels with infections and solid tumors, butthey also reveal clear distinctions from both. In clear contrast tomany solid tumors and their LN metastasis, lymphoma-challengedBECs acquired gene signatures that were not enriched for Notch,Vegfa, or Hif1a pathway genes. LN remodeling in inflammationproceeds in sequential overlapping phases (12, 15, 44). In theinitiation phase, a rapid blood vasculature growth is dependent onIL1b-secreting DCs that stimulate VEGFA release from FRCs (45).In contrast, we here examined LN from lymphoma challenged miceat days 8 to 12, when VEGFA has likely declined and instead,VEGFC has gained a leading role.
The TF HIF1a is induced by hypoxia and functions as the maintranscriptional regulator of tumor-associated vascular remodel-ing (46). In conjunction with the weak pimonidazole staining seenin lymphoma-bearing LNs, our data point to a micromilieu in whichhypoxia is not a driving force for angiogenesis. We conclude that LNshave a distinct vascular microanatomy that involves highly specializedHEVs and capillaries that can accommodate an autochthonous tumorwithout changing the tissue-specific oxygen metabolism.
Consistent with the inflammatory response upon infection, Em-Myclymphoma-challenged LNs rapidly begin HEV remodeling (15, 33).One study shows that blocking VEGFA or its cognate receptorVEGFR2 does not inhibit LCMV-induced vascular expansion inLNs (15). This observation parallels a mechanism that we detectedin lymphoma-induced angiogenesis; we also could not find evidencefor a functional role of VEGFA in this process. During viral infection, Bcell–derived LTa1b2 contributes to remodeling of theHEVnetwork inthe acute phase (15). The infection model examined the role ofactivated B cells in an environment perturbed by virus-inducedstructural disruption of the reticular network. In the current study,LT-expressing lymphoma B cells were being added to a largely intactLN microstructure. Thus, a functional B cell–dependent LTa1b2–LTbR signaling loopwas present. Immunopharmacologic interferencewith LTbR signaling on BECs retarded angiogenesis substantially anddelayed progression of B-cell lymphoma. Although LTa1b2 did notinduce HUVEC proliferation and sprouting alone, we envision that itacts as amorphogen stimulating HEV or capillary branching and thus,may synergize with the growth factor VEGFC.
Interestingly, VEGFD–overexpressing tumor cells induced down-regulation of BMP4 in HEV cells (47). Loss of BMP4 expressionresulted in severe remodeling of HEVs, in contrast to our lymphomamodel a key mediator of this process, VEGFD, was not increasedin tumor-exposed LNs. Likewise, BMP4 gene expression in BECs wasnot significantly altered, indicating that these linked pathogeneticfactors are unlikely to cooperate in lymphoma-induced vascularremodeling.
We note that branching points after LTbR-Ig treatment werestrongly diminished. Regarding the hierarchy of this interaction,it has been shown that LTbR stimulation of murine FRC-typecells upregulates VEGF expression, suggesting that the LTbR-LTa1b2 axis on FRCs regulates VEGF levels and thus, EC prolif-eration (44).
Despite large similarities to infection-induced LN remodeling, wenote that inflammation is not the cause for our observations. Acomparison of our transcriptome dataset with microarray data frompublished results characterizing stromal response to inflammationshowed no enrichment of an inflammation signature (8).
(Continued.) Data points represent the ratios of tumor cells among all CD45þ cells (in percent) in isotype treated mice versus anti-VEGFR2–treated animals.; n¼ 2independent experiments.B andC,CD31þ vascular surface areas and branching points in T-cell zones from inguinal LNs taken from animals treatedwith an isotype oran anti–VEGFR2 antibody. Quantification was done on 3D-reconstructed images as in Fig. 3C. Bars, mean values. A Mann–Whitney U test was used. n.s.,nonsignificant.D, LNs from control (n¼ 3–5) and lymphoma exposedmice (n¼ 3–5)were dissected and stroma cell subsetswere isolated.Macrophages (CD11bþF4/80þ), T cells (CD3þ), normal B cells (B220þ), and Em-Myc lymphoma cells (B220þFSC-Ahigh tumor) were also sorted and subjected to qRT-PCR analysis. Error bars,mean� SEM; n¼ 3–5 independent experiments were performed; Student t test. E, Sera from animals treatedwith (n¼ 8) or without Em-Myc lymphoma cells (n¼ 8)were analyzed by a VEGFC–specific ELISA. Data points depict the calculated amount of VEGFC relative to total protein amounts in 1mL serum. Bars, mean values; n¼3 independent experiments; Student t test. F, Sprouting of HUVECs performed in spheroid cultures. Cellswere grown in basal growthmedium alone (control), or withsupplements; in addition, a combination of these factors was used for a 24-hour culture. n¼ 40–55 spheroids counted per treatment condition; data points indicatemean values per experiment; n¼ 5–8 experiments were performed; Student t test. G, HUVEC growth factors and LTa1b2 (LT) were added to basal growth medium(control) for 6 hours. VEGFR3 surface density was analyzed by flow cytometry and expressed as geo mean fluorescence intensity. n¼ 8 independent experiments;Student t test. H, LNs and spleen protein lysates were generated from control (n¼ 6–7) and Em-Myc lymphoma-transferred mice (n¼ 6–7). VEGFCwas detected inimmunoblot, and anti-GAPDH was used as a loading control. Images were cropped for presentation and displayed bands are representative of all samples. I, Theratios of the different trimming forms of VEGFC relative toGAPDH are presented as bar graphs; n > 3 independent experiments, with n¼ 3 different tumor cell clones.Bars,meanvalues. J,Geneexpression ofCcbe1 as determinedbyqRT-PCR. FRCs, BECs, LECs, andnormal B cells fromcontrolmice (n¼3–5)were sorted, or cellswererecovered from Em-Myc lymphoma-transplanted mice (n ¼ 3–6); n > 3 independent experiments. K, Vegfr3 gene expression; control (n ¼ 3–5), Em-Myc lymphomatransplantedmice (n¼ 3–5); at least three independent experiments. Error bars, mean� SEM. I–K,AMann–WhitneyU test was applied. L, Vegfr3-GFP reportermicechallenged with Em-Myc lymphoma cells (n ¼ 3 mice) for 8 to 12 days. FRCs, BECs, and LECs were differentiated by anti-CD31 and anti-gp38 staining. Shown arerepresentative histograms of VEGFR3 (GFPþ)-expressing cell subsets. On the right, VEGFR3 expression in vascular subsets from control and Em-Myc lymphoma-bearingmice assessed by immunostaining and flow cytometry analysis. Error bars formean fluorescence values indicatemean� SEM. n¼ 3mice/group.M,Wtmicereceived a total of 2–4� 105 Em-Myc tumor B cells intravenously. At days 1 to 7, animalswere administeredwith SAR131675 (n¼4mice) orwith carrier solution (n¼6)orally. At days 8 to 9, lymphoma cells in LNs were quantified by flow cytometry relative to all leukocytes. Bars (mean values), the ratios of control and SAR131675-treated animals. On the right, CD31þ vascular surface areas in inguinal LNs from theseanimalswere analyzedby confocalmicroscopy; n¼ 3 independent experiments.N, Wt mice were inoculated subcutaneously with MCA313 fibrosarcoma cells. Upon palpable tumor growth, animals were treated with SAR131675 or with carriersolution orally for up to 16 days. Tumor weight of resected tumors was measured after sacrificing the animals. M and N, A Mann–Whitney U test was applied.n.s., nonsignificant.
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1327
Morphologic analysis revealed a hypersprouting phenotype oflymphoma-associated vessels. GSEA of BECs did not indicate anenrichment of a tip cell or Notch signaling pathway gene signature,which would have been expected from ontogeny or other tumorangiogenesis models (48–50). Angiogenic sprouting is led by tip cellsthat sense a VEGFA guide cue and produce Dll4. This Notch ligandacts on stalk cells preceding the tip that proliferate but are preventedfrom adopting a tip cell differentiation (22). It has also been shown thatVEGFR3 localizes in endothelial sprouts of several solid tumors and inthe developing mouse retina in leading tip cells. In the absence ofVEGFR2 activity, VEGFR3–transmitted signals can sustain somedegree of angiogenesis (38). When Notch signaling is disrupted,increased sprouting in conjunction with induced Vegfr3 expressionoccurs (51). We note that Esm1 and Dll4 expression normally asso-ciatedwith a tip cell phenotypewas irregular in lymphoma vasculature,contrasting to studies in highly ordered developing tissues like reti-na (42, 49, 52). Thus, in mice unopposed VEGFR3 activity that isreleased from a Notch signaling brake can result in a hyperactivesprouting phenotype, as seen in our Em-Myc lymphoma model.
In summary, we showed that lymphoma angiogenesis proceeds inunique transcriptional and morphogenic programs, which are clearlydistinct from solid tumor and inflammation-induced structural remo-deling. Identification of the VEGFR3 and the LTbR signaling pathwaysas regulators of lymphoma angiogenesis strongly suggests that combi-natorial therapies targeting vessel densities should be reconsidered.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors’ ContributionsConception and design:M. Gloger, M. Zschummel, G. Lenz, U.E. H€opken, A. RehmDevelopment of methodology: M. Gloger, L. Menzel, K. Gerlach, T. Kammert€ons,M. Zschummel, G. Lenz, H. Gerhardt, A. RehmAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): M. Gloger, L. Menzel, A.-C. Vion, I. Anagnostopoulos, K. Gerlach,T. Hehlgans, U.E. H€opken, A. RehmAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): M. Gloger, L. Menzel, M. Grau, A.-C. Vion,M. Zapukhlyak, M. Zschummel, G. Lenz, H. Gerhardt, U.E. H€opken, A. RehmWriting, review, and/or revision of the manuscript:M.Gloger, L. Menzel, M. Grau,A.-C. Vion, I. Anagnostopoulos, M. Zapukhlyak, T. Kammert€ons, G. Lenz,H. Gerhardt, U.E. H€opken, A. RehmAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): I. Anagnostopoulos, A. RehmStudy supervision: U.E. H€opken, A. Rehm
AcknowledgmentsThe study was funded by a grant from Deutsche Krebshilfe to A. Rehm and U.E.
H€opken (Grant No. 111918). We thank Daniel Besser (MDC) for help in bioinfor-matic analysis; Christian Friese (Charit�e, Berlin) for expert technical assistance; AnjeSporbert and Matthias Richter for microscopic image analysis (Advanced LightMicroscopy Platform; MDC); Ralf Adams (MPI, M€unster, Germany); and TaijaM€akinen (Uppsala University, Sweden) for Cre-deleter mouse strains.
The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.
Received May 13, 2019; revised September 4, 2019; accepted January 8, 2020;published first January 13, 2020.
References1. Lenz G, Wright G, Dave SS, Xiao W, Powell J, Zhao H, et al. Stromal gene
signatures in large-B-cell lymphomas. N Engl J Med 2008;359:2313–23.2. Cardesa-Salzmann TM, Colomo L, Gutierrez G, Chan WC, Weisenburger D,
Climent F, et al. Highmicrovessel density determines a poor outcome in patientswith diffuse large B-cell lymphoma treated with rituximab plus chemotherapy.Haematologica 2011;96:996–1001.
3. Ganjoo KN,Moore AM,Orazi A, Sen JA, Johnson CS, AnCS. The importance ofangiogenesis markers in the outcome of patients with diffuse large B celllymphoma: a retrospective study of 97 patients. J Cancer Res Clin Oncol2008;134:381–7.
4. Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases. Nature 2000;407:249–57.
5. Kerbel R, Folkman J. Clinical translation of angiogenesis inhibitors. Nat RevCancer 2002;2:727–39.
6. Ganjoo KN, An CS, Robertson MJ, Gordon LI, Sen JA, Weisenbach J, et al.Rituximab, bevacizumab and CHOP (RA-CHOP) in untreated diffuse large B-cell lymphoma: safety, biomarker and pharmacokinetic analysis. LeukLymphoma 2006;47:998–1005.
7. Seymour JF, Pfreundschuh M, Trneny M, Sehn LH, Catalano J, Csinady E, et al.R-CHOP with or without bevacizumab in patients with previously untreateddiffuse large B-cell lymphoma: final MAIN study outcomes. Haematologica2014;99:1343–9.
8. Malhotra D, Fletcher AL, Astarita J, Lukacs-Kornek V, Tayalia P, Gonzalez SF,et al. Transcriptional profiling of stroma from inflamed and resting lymph nodesdefines immunological hallmarks. Nat Immunol 2012;13:499–510.
9. Song K, Herzog BH, Sheng M, Fu J, McDaniel JM, Chen H, et al. Lenalidomideinhibits lymphangiogenesis in preclinical models of mantle cell lymphoma.Cancer Res 2013;73:7254–64.
10. Ruan J, Luo M, Wang C, Fan L, Yang SN, Cardenas M, et al. Imatinib disruptslymphoma angiogenesis by targeting vascular pericytes. Blood 2013;121:5192–202.
11. Hopken UE, Rehm A. Homeostatic chemokines guide lymphoma cells to tumorgrowth-promoting niches within secondary lymphoid organs. J Mol Med 2012;90:1237–45.
12. Webster B, Ekland EH, Agle LM, Chyou S, Ruggieri R, Lu TT. Regulation oflymph node vascular growth by dendritic cells. J Exp Med 2006;203:1903–13.
13. Rehm A, Mensen A, Schradi K, Gerlach K, Wittstock S, Winter S, et al.Cooperative function of CCR7 and lymphotoxin in the formation of a lympho-ma-permissive niche within murine secondary lymphoid organs. Blood 2011;118:1020–33.
14. Onder L, Danuser R, Scandella E, Firner S, Chai Q, Hehlgans T, et al. Endothelialcell-specific lymphotoxin-beta receptor signaling is critical for lymph node andhigh endothelial venule formation. J Exp Med 2013;210:465–73.
15. Kumar V, Scandella E, Danuser R, Onder L, Nitschke M, Fukui Y, et al. Globallymphoid tissue remodeling during a viral infection is orchestrated by a B cell-lymphotoxin-dependent pathway. Blood 2010;115:4725–33.
16. Heinig K, Gatjen M, Grau M, Stache V, Anagnostopoulos I, Gerlach K, et al.Access to follicular dendritic cells is a pivotal step inmurine chronic lymphocyticleukemia B-cell activation and proliferation. Cancer Discov 2014;4:1448–65.
17. Hindley JP, Jones E, Smart K, Bridgeman H, Lauder SN, Ondondo B, et al. T-celltrafficking facilitated by high endothelial venules is required for tumor controlafter regulatory T-cell depletion. Cancer Res 2012;72:5473–82.
18. Reimann M, Lee S, Loddenkemper C, Dorr JR, Tabor V, Aichele P, et al. Tumorstroma-derived TGF-beta limits myc-driven lymphomagenesis via Suv39h1-dependent senescence. Cancer Cell 2010;17:262–72.
19. Habenicht LM, Kirschbaum SB, Furuya M, Harrell MI, Ruddell A. Tumorregulation of lymph node lymphatic sinus growth and lymph flow inmice and inhumans. Yale J Biol Med 2017;90:403–15.
20. Ruddell A,Mezquita P, Brandvold KA, Farr A, Iritani BM. B lymphocyte-specificc-Myc expression stimulates early and functional expansion of the vasculatureand lymphatics during lymphomagenesis. Am J Pathol 2003;163:2233–45.
21. Ruddell A, Harrell MI, Furuya M, Kirschbaum SB, Iritani BM. B lymphocytespromote lymphogenous metastasis of lymphoma and melanoma. Neoplasia2011;13:748–57.
22. Mettouchi A. The role of extracellular matrix in vascular branching morpho-genesis. Cell Adh Migr 2012;6:528–34.
23. Alitalo K, Tammela T, Petrova TV. Lymphangiogenesis in development andhuman disease. Nature 2005;438:946–53.
Gloger et al.
Cancer Res; 80(6) March 15, 2020 CANCER RESEARCH1328
24. Lykken JM, Horikawa M, Minard-Colin V, Kamata M, Miyagaki T, Poe JC,et al. Galectin-1 drives lymphoma CD20 immunotherapy resistance: valida-tion of a preclinical system to identify resistance mechanisms. Blood 2016;127:1886–95.
25. del Toro R, Prahst C, Mathivet T, Siegfried G, Kaminker JS, Larrivee B, et al.Identification and functional analysis of endothelial tip cell-enriched genes.Blood 2010;116:4025–33.
26. Gerhardt H, Golding M, Fruttiger M, Ruhrberg C, Lundkvist A, Abramsson A,et al. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia.J Cell Biol 2003;161:1163–77.
27. Dieterich LC,Ducoli L, Shin JW,DetmarM.Distinct transcriptional responses oflymphatic endothelial cells to VEGFR-3 and VEGFR-2 stimulation. Sci Data2017;4:170106.
28. Shats I, Gatza ML, Chang JT, Mori S, Wang J, Rich J, et al. Using a stem cell-based signature to guide therapeutic selection in cancer. Cancer Res 2011;71:1772–80.
29. Thiru P, Kern DM, McKinley KL, Monda JK, Rago F, Su KC, et al. Kinetochoregenes are coordinately up-regulated in human tumors as part of a FoxM1-relatedcell division program. Mol Biol Cell 2014;25:1983–94.
30. Gebala V, Collins R, Geudens I, Phng LK, Gerhardt H. Blood flow drives lumenformation by inverse membrane blebbing during angiogenesis in vivo. Nat CellBiol 2016;18:443–50.
31. Reichman-FriedM, Raz E. Blood, blebs and lumen expansion. Nat Cell Biol 2016;18:366–7.
32. Rocha SF, Schiller M, Jing D, Li H, Butz S, Vestweber D, et al. Esm1 modulatesendothelial tip cell behavior and vascular permeability by enhancing VEGFbioavailability. Circ Res 2014;115:581–90.
33. LuTT, Browning JL. Role of the lymphotoxin/LIGHT system in the developmentand maintenance of reticular networks and vasculature in lymphoid tissues.Front Immunol 2014;5:47.
34. Lemmon MA, Schlessinger J. Cell signaling by receptor tyrosine kinases. Cell2010;141:1117–34.
35. Jha SK, Rauniyar K, Karpanen T, Leppanen VM, Brouillard P, Vikkula M, et al.Efficient activation of the lymphangiogenic growth factor VEGF-C requires theC-terminal domain of VEGF-C and the N-terminal domain of CCBE1. Sci Rep2017;7:4916.
36. Joukov V, Sorsa T, Kumar V, Jeltsch M, Claesson-Welsh L, Cao Y, et al.Proteolytic processing regulates receptor specificity and activity of VEGF-C.EMBO J 1997;16:3898–911.
37. Rauniyar K, Jha SK, JeltschM. Biology of vascular endothelial growth factor C inthe morphogenesis of lymphatic vessels. Front Bioeng Biotechnol 2018;6:7.
38. Tammela T, Zarkada G, Wallgard E, Murtomaki A, Suchting S, Wirzenius M,et al. Blocking VEGFR-3 suppresses angiogenic sprouting and vascular networkformation. Nature 2008;454:656–60.
39. Hanahan D, Folkman J. Patterns and emerging mechanisms of the angiogenicswitch during tumorigenesis. Cell 1996;86:353–64.
40. Tabchi S, Blais N. Antiangiogenesis for advanced non-small-cell lung cancer inthe era of immunotherapy and personalized medicine. Front Oncol 2017;7:52.
41. Hsu MC, Pan MR, Hung WC. Two birds, one stone: double hits on tumorgrowth and lymphangiogenesis by targeting vascular endothelial growth factorreceptor 3. Cells 2019;8. DOI: 10.3390/cells8030270.
42. Benedito R, Rocha SF, Woeste M, Zamykal M, Radtke F, Casanovas O, et al.Notch-dependent VEGFR3 upregulation allows angiogenesis without VEGF-VEGFR2 signalling. Nature 2012;484:110–4.
43. Lohela M, Bry M, Tammela T, Alitalo K. VEGFs and receptors involved inangiogenesis versus lymphangiogenesis. Curr Opin Cell Biol 2009;21:154–65.
44. Chyou S, Ekland EH, Carpenter AC, Tzeng TC, Tian S, Michaud M, et al.Fibroblast-type reticular stromal cells regulate the lymph node vasculature.J Immunol 2008;181:3887–96.
45. Benahmed F, Chyou S,DasoveanuD,Chen J, KumarV, Iwakura Y, et al.MultipleCD11cþ cells collaboratively express IL-1beta to modulate stromal vascularendothelial growth factor and lymph node vascular-stromal growth. J Immunol2014;192:4153–63.
46. BristowRG,Hill RP.Hypoxia andmetabolism.Hypoxia, DNA repair and geneticinstability. Nat Rev Cancer 2008;8:180–92.
47. Farnsworth RH, Karnezis T, Shayan R, Matsumoto M, Nowell CJ, Achen MG,et al. A role for bone morphogenetic protein-4 in lymph node vascularremodeling and primary tumor growth. Cancer Res 2011;71:6547–57.
48. Cao Z, Ding BS, Guo P, Lee SB, Butler JM, Casey SC, et al. Angiocrine factorsdeployed by tumor vascular niche induce B cell lymphoma invasiveness andchemoresistance. Cancer Cell 2014;25:350–65.
49. Hellstrom M, Phng LK, Hofmann JJ, Wallgard E, Coultas L, Lindblom P, et al.Dll4 signalling through Notch1 regulates formation of tip cells during angio-genesis. Nature 2007;445:776–80.
50. Ridgway J, Zhang G,Wu Y, Stawicki S, LiangWC, Chanthery Y, et al. Inhibitionof Dll4 signalling inhibits tumour growth by deregulating angiogenesis. Nature2006;444:1083–7.
51. Siekmann AF, Lawson ND. Notch signalling limits angiogenic cell behaviour indeveloping zebrafish arteries. Nature 2007;445:781–4.
52. Suchting S, Freitas C, le Noble F, Benedito R, Breant C, Duarte A, et al. TheNotchligand Delta-like 4 negatively regulates endothelial tip cell formation and vesselbranching. Proc Natl Acad Sci U S A 2007;104:3225–30.
AACRJournals.org Cancer Res; 80(6) March 15, 2020 1329
Lymphoma Angiogenesis Uses Noncanonical Angiocrine Pathways
440 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Selective Inhibition of HDAC3 Targets Synthetic Vulnerabilities and Activates Immune Surveillance in Lymphoma Patrizia Mondello 1 , Saber Tadros 2 , Matt Teater 1 , Lorena Fontan 1 , Aaron Y. Chang 3 , Neeraj Jain 2 , Haopeng Yang 2 , Shailbala Singh 4 , Hsia-Yuan Ying 1 , Chi-Shuen Chu 5 , Man Chun John Ma 2 , Eneda Toska 6 , Stefan Alig 7 , Matthew Durant 1 , Elisa de Stanchina 8 , Sreejoyee Ghosh 2 , Anja Mottok 9 , Loretta Nastoupil 2 , Sattva S. Neelapu 2 , Oliver Weigert 7 , Giorgio Inghirami 10 , José Baselga 5 , Anas Younes 11 , Cassian Yee 4 , Ahmet Dogan 12 , David A. Scheinberg 3 , Robert G. Roeder 5 , Ari M. Melnick 1 , and Michael R. Green 2 , 13 , 14
RESEARCH ARTICLE
ABSTRACT CREBBP mutations are highly recurrent in B-cell lymphomas and either inactivate its histone acetyltransferase (HAT) domain or truncate the protein. Herein, we
show that these two classes of mutations yield different degrees of disruption of the epigenome, with HAT mutations being more severe and associated with inferior clinical outcome. Genes perturbed by CREBBP mutation are direct targets of the BCL6–HDAC3 onco-repressor complex. Accordingly, we show that HDAC3-selective inhibitors reverse CREBBP -mutant aberrant epigenetic programming, resulting in: (i) growth inhibition of lymphoma cells through induction of BCL6 target genes such as CDKN1A and (ii) restoration of immune surveillance due to induction of BCL6-repressed IFN pathway and antigen-presenting genes. By reactivating these genes, exposure to HDAC3 inhibitors restored the ability of tumor-infi ltrating lymphocytes to kill DLBCL cells in an MHC class I and II–dependent manner, and synergized with PD-L1 blockade in a syngeneic model in vivo . Hence, HDAC3 inhibition represents a novel mechanism-based immune epigenetic therapy for CREBBP -mutant lymphomas.
SIGNIFICANCE: We have leveraged the molecular characterization of different types of CREBBP muta-tions to defi ne a rational approach for targeting these mutations through selective inhibition of HDAC3. This represents an attractive therapeutic avenue for targeting synthetic vulnerabilities in CREBBP -mutant cells in tandem with promoting antitumor immunity.
1 Division of Hematology/Oncology, Department of Medicine, Weill Cornell Medical College, New York, New York. 2 Department of Lymphoma & Myeloma, The University of Texas MD Anderson Cancer Center, Houston, Texas. 3 Molecular Pharmacology Program, Memorial Sloan Kettering Cancer Center, New York, New York. 4 Department of Melanoma Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas. 5 Laboratory of Biochemistry and Molecular Biology, The Rockefeller University, New York, New York. 6 Department of Human Oncology and Pathogenesis, Memorial Sloan Kettering Cancer Center, New York, New York. 7 Department of Internal Medicine III, University Hospital of the Ludwig-Maximilians-University Munich, Munich, Germany. 8 Antitumor Assessment Core Facility, Memorial Sloan Kettering Cancer Center, New York, New York. 9 Institute of Pathology, University of Würzburg and Com-prehensive Cancer Center Mainfranken, Würzburg, Germany. 10 Department of Pathology and Laboratory Medicine, Weill Cornell Medical College, New York, New York. 11 Department of Medicine, Memorial Sloan Kettering Can-cer Center, New York, New York. 12 Department of Pathology and Laboratory Medicine, Memorial Sloan Kettering Cancer Center, New York, New York. 13 Department of Genomic Medicine, The University of Texas MD Anderson
Cancer Center, Houston, Texas. 14 Center for Cancer Epigenetics, The University of Texas MD Anderson Cancer Center, Houston, Texas. Note: Supplementary data for this article are available at Cancer Discovery Online (http://cancerdiscovery.aacrjournals.org/). P. Mondello and S. Tadros contributed equally as co–fi rst authors of this article. A.M. Melnick and M.R. Green contributed equally as co–senior authors of this article. Corresponding Authors: Michael R. Green, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Unit 903, Houston, TX 77030. Phone: 713-745-4244; E-mail: [email protected] ; and Ari M. Melnick, Department of Medicine, Weill Cornell Medicine, 413 East 69th Street, New York, NY 10021. Phone: 646-962-6725; E-mail: [email protected] Cancer Discov 2020;10:440–59 doi: 10.1158/2159-8290.CD-19-0116 ©2020 American Association for Cancer Research.
MARCH 2020�CANCER DISCOVERY | 441
INTRODUCTIONDiffuse large B-cell lymphoma (DLBCL) and follicular
lymphoma are the two most frequent subtypes of non-Hodgkin lymphoma. These diseases originate from germi-nal center B (GCB) cells, which are at a stage of development that naturally allows for the proliferation and affinity matu-ration of antigen-experienced B cells to produce terminally differentiated memory B cells or plasma cells. The germinal center (GC) reaction is regulated by B cell–intrinsic activa-tion and suppression of genes by master regulators such as the BCL6 transcription factor (1), and extrinsically via the interaction of GCB cells with follicular helper T (TFH) cells and other immune cells within the GC (2). The BCL6 transcription factor is critical for GCB-cell development and coordinately suppresses the expression of large sets of genes by recruiting SMRT/NCOR corepressor complexes containing HDAC3 (3), the LSD1 histone demethylase (4), and tethering a noncanonical polycomb repressor 1–like complex in cooperation with EZH2 (5). These genes are normally reactivated to drive GC exit and terminal differen-tiation in the GC light zone, but the epigenetic control of these dynamically regulated GC transcriptional programs is
perturbed in DLBCL and follicular lymphoma via the down-stream effects of somatic mutation of chromatin-modifying genes (CMG; ref. 6).
The second most frequently mutated CMG in both DLBCL and follicular lymphoma is the CREBBP gene, which encodes a histone acetyltransferase that activates transcription via acetylation of histone H3 at lysine 27 (H3K27Ac) and other residues. We have previously found that these mutations arise as early events during the genomic evolution of follicular lym-phoma and reside in a population of tumor-propagating cells, often referred to as common progenitor cells (CPC; ref. 7). We have also noted an association between CREBBP inactivation and reduced expression of MHC class II in human and murine lymphomas (7, 8). The expression of MHC class II is critical for the terminal differentiation of B cells through the GC reaction (9). The interaction with T helper cells via MHC class II results in B-cell costimulation through CD40 that drives NFκB activation and subsequent IRF4-driven suppression of BCL6. However, in B-cell lymphoma, tumor antigens may also be presented in MHC class II and recognized by CD4+
T cells that drive an antitumor immune response (10, 11). The active suppression of MHC class II expression in B-cell
Mondello et al.RESEARCH ARTICLE
442 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
lymphoma may therefore be driven by evolutionary pressure against MHC class II–binding tumor antigens, as recognized in other cancers (12). In support of this notion, the reduced expression of MHC class II has been found to be associated with poor outcome in DLBCL (13, 14).
Recently, MHC class II expression has been defined as an important component of IFNγ-related signatures that are predictive of the activity of PD-1–neutralizing antibodies (14–17). This is consistent with a prominent role for CD4+ T cells in directing antitumor immunity and responses to immunotherapy (18). Despite this, current immunothera-peutic strategies largely rely on the preexistence of an inflam-matory microenvironment for therapeutic efficacy. Here, we have characterized the molecular consequences of CREBBP mutations and identified BCL6-regulated cell cycle, differ-entiation, and IFN signaling pathways as core features that are aberrantly silenced at the epigenetic and transcriptional level. We show that HDAC3 inhibition specifically restores these pathways, thus suppressing growth and most criti-cally enabling T cells to recognize and kill lymphoma cells. Together, these findings highlight multiple mechanisms by which selective inhibition of HDAC3 can drive tumor-intrinsic killing as well as activate IFNγ signaling and antitu-mor immunity that extends to both CREBBP wild-type and CREBBP-mutant tumors.
RESULTSCREBBPR1446C Mutations Function in a Dominant Manner to Suppress BCL6 Coregulated Epigenetic and Transcriptional Programs
In B-cell lymphomas, the CREBBP gene is predominantly targeted by point mutations that result in single amino acid substitutions within the lysine acetyltransferase (KAT) domain (7, 19), with a hotspot at arginine 1446 (R1446) that leads to a catalytically inactive protein (20, 21). However, all of the prior studies characterizing the effects of CREBBP muta-tion have been performed using knockout (KO) or knock-down of Crebbp, resulting in loss of protein (LOP; refs. 8, 19, 22–25). Furthermore, mutations of R1446 have not been documented in any lymphoma cell line. We therefore opted to investigate whether there may be unique functional con-sequences of KAT domain hotspot mutations in CREBBP. To achieve this, we utilized CRISPR/Cas9-mediated gene editing with two unique guide RNAs (gRNA) to introduce the most common CREBBP mutation, R1446C, into a CREBBP wild-type cell line bearing the t(14;18)(q21;q32) translocation RL (Fig. 1A). This allowed us to generate clones from each gRNA that had received the constructs but remained wild-type (CREBBPWT), those that edited their genomes to intro-duce the point mutations into both alleles (CREBBPR1446C), and those that acquired homozygous frameshift mutations resulting in LOP (CREBBPKO; Supplementary Fig. S1). These isogenic sets of clones differ only in their CREBBP mutation status, and therefore allow for detailed functional characteri-zation in a highly controlled setting.
Western blot analysis confirmed that the CREBBPR1446C pro-tein was still expressed and that CREBBPKO mutations resulted in a complete loss of protein expression (Fig. 1B). Western
blotting with densitometry of the H3K27Ac mark revealed significant reduction in H3K27Ac in CREBBPR1446C cells versus isogenic CREBBPWT controls (P < 0.001; Fig. 1C). Although CREBBPKO cells also showed lower H3K27Ac abundance than isogenic CREBBPWT controls, this reduction was not statisti-cally significant (P = 0.106). We performed chromatin immu-noprecipitation sequencing (ChIP-seq) for H3K27Ac to define the physical location of these changes and identify potentially deregulated genes. This revealed 2,022 regions with signifi-cantly reduced acetylation, and 2,304 regions with signifi-cantly increased acetylation in CREBBPR1446C cells compared with isogenic CREBBPWT controls (fold change >1.5, q-value < 0.01; Fig. 1D; Supplementary Fig. S2; Supplementary Table S1). This pattern was mirrored by another CREBBP-catalyzed mark, H3K18Ac, and the loss or gain of histone acetylation was also accompanied by gain or loss of H3K27me3, respec-tively (Supplementary Fig. S3A–S3D). Regions with loss of H3K27Ac were observed to normally bear this mark in human GCB cells (ref. 24; Fig. 1D), suggesting that CREBBPR1446C mutations lead to loss of a normal GCB epigenetic program. Notably, CREBBPKO resulted in a reduction of H3K27Ac in these regions also, but at a lower magnitude than that observed with CREBBPR1446C mutations (Fig. 1D and E). This was not observed for regions with increased H3K27Ac, which showed little change in CREBBPKO cells.
Using RNA sequencing (RNA-seq), we observed broad changes in transcription, with 766 genes showing signifi-cantly increased expression and 733 genes showing signifi-cantly decreased expression in CREBBPR1446C cells compared with CREBBPWT isogenic controls (fold change >1.5, q-value < 0.01; Fig. 1F; Supplementary Table S2). The genes that were proximal to regions of H3K27Ac loss showed a coordinate reduction in transcript abundance and vice versa (FDR < 0.001; Fig. 1G), suggesting that these broad changes in transcrip-tion are directly linked to altered promoter/enhancer activity as a result of differential H3K27Ac. Importantly, we observed significant enrichment of the transcriptional signature induced by shRNA knockdown of CREBBP in murine B cells or human DLBCL cell lines (ref. 8; Fig. 1H). However, an even more significant enrichment was observed for the signature of genes we defined as being lost in association with CREBBP mutation in primary human CREBBP mutant follicular lym-phoma B cells (ref. 7; FDR < 0.001; Fig. 1H). Consistent with our CRISPR cell line results, primary follicular lymphoma B cells with CREBBP LOP mutations were observed to have significantly less repression of this signature by single-sample gene set enrichment analysis (ssGSEA) compared with those with CREBBP KAT domain mutation (P = 0.039; Supplemen-tary Fig. S4).
We used epigenetic and transcriptional profiles and hyper-geometric analysis to gain insight into biological functions disrupted by CREBBPR1446C mutations (Fig. 1I; Supplementary Tables S3 and S4). This confirmed a significant enrichment of genes downregulated in patients with CREBBP-mutant lymphoma, and further revealed an enrichment of BCL6–SMRT and BCL6 targets, consistent with the proposed role of CREBBP in opposing BCL6-mediated transcriptional repression. The biological functions of these genes included B-cell receptor (BCR) and NFκB signaling in addition to IFN signaling and antigen presentation (Fig. 1I). In line with the
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 443
enrichment of antigen-presenting pathways and the genome-wide differences in H3K27Ac patterning, we observed more severe reduction of H3K27Ac over the MHC class II locus in CREBBPR1446C versus CREBBPKO cells (Fig. 1J). This was asso-ciated with an approximately 2-fold reduction of cell surface MHC class II in CREBBPKO cells compared with isogenic CREBBPWT controls, but an approximately 37-fold reduction in CREBBPR1446C cells (Fig. 1K). Notably, EP300 is expressed at a high level in all of the CRISPR-modified cell lines, and ChIP-qPCR showed that the MHC class II locus is bound by both CREBBP and EP300 in these cells (Supplementary Fig. S5A–S5C). Moreover, inhibition of EP300 activity in CREBBPKO cells reduced the expression of MHC class II to levels similar to that in CREBBPR1446C cells (Supplementary Fig. S5D), suggesting that the redundant activity of EP300 may partially sustain MHC class II expression in CREBBPKO
cells. These data further support a stronger epigenetic and transcriptional suppression of BCL6 target genes, including those involved in antigen presentation, with CREBBPR1446C
mutations compared with CREBBPKO.To confirm that the CREBBP KAT mutation is biologi-
cally active in primary GC B cells in a similar manner to that observed in CRISPR-engineered cell lines, we generated a novel genetically engineered mouse model with Cre-inducible expression of CrebbpR1447C (equivalent to human R1446C). The endogenous Crebbp locus was engineered to switch from WT to mutant gene expression using a floxed inversion cas-sette (Supplementary Fig. S6A). These animals were crossed to mice engineered with the Cγ1-Cre allele to specifically induce recombination in GC B cells and bone marrow transplanted into irradiated CD45.1 recipients (Supplementary Fig. S6B and S6C). Transplanted mice were immunized with sheep
Figure 1. Detailed molecular characterization of CREBBPR1446C and CREBBPKO mutations using isogenic CRISPR/Cas9-modified lymphoma cells. A, A diagram shows the CRISPR/Cas9 gene-editing strategy. Two guides were designed that were proximal to the R1446 codon, with PAM sites highlighted in yellow. A single-stranded homologous recombination (HR) template was utilized that encoded silent single-nucleotide changes that interfered with the PAM sites but did not change the protein coding sequence, and an additional single-nucleotide change that encoded the R1446C mutation. B, A representative Western blot shows that the CREBBPR1446C protein is expressed at similar levels to those of CREBBPWT, whereas CREBBPKO results in a complete loss of protein expression as expected. The level of H3K27Ac shows a more visible reduction in CREBBPR1446C cells compared with isogenic CREBBPWT cells than that observed in CREBBPKO cells. C, Quantification of triplicate Western blot experiments shows that there is a significant reduc-tion of H3K27Ac in CREBBPR1446C cells compared with CREBBPWT cells (t test, P < 0.001). A reduction is also observed in CREBBPKO cells, but this was not statistically significant (t test, P = 0.106). D, Heat maps show the regions of significant H3K27Ac loss (n = 2,022, above) and gain (n = 2,304, below) in CREBBPR1446C cells compared with isogenic WT controls. The regions with reduced H3K27Ac in CREBBPR1446C cells can be seen to normally bear this mark in GCB cells. E, Density plots show that the degree of H3K27Ac loss (above) is most notable in CREBBPR1446C cells compared with isogenic WT cells, whereas CREBBPKO cells show an intermediate level of loss. Regions with H3K27Ac gain (below) in CREBBPR1446C cells showed fewer changes in CREBBPKO cells. (continued on next page)
H3K
27Ac
loss
R14
46C
vs.
WT
(n =
2,0
22)
H3K
27Ac
gai
nR
1446
C v
s. W
T (n
= 2,
304)
WT
D
5 kb
Normal GCB
5 kb
R1446C
5 kb 5 kb
KO
5 kb 5 kb 5 kb 5 kb
0 2 4 6 8Normalized reads (RPKM)
gRNA1 gRNA2 CB4 CB5gRNA1 gRNA2 gRNA1 gRNA2
E
0
2
4
6
8
10
−5 kb LociMidpoint
5 kb
0
1
2
3
4
5
6
7
H3K
27Ac
loss
Mea
n no
rmal
ized
read
s (R
PKM
)H
3K27
Ac g
ain
Mea
n no
rmal
ized
read
s (R
PKM
)
−5 kb Locimidpoint
5 kb
WTKOR1446C
WTKOR1446C
1.00
1.25
0.75
0.50
0.25
0
gRNA
1gR
NA2
gRNA
1gR
NA2
gRNA
1gR
NA2
WT R1446C KO
Rel
ativ
e H
3K27
Ac/H
3 ra
tio
***
N.S.C
H3K27Ac
H3 (Total)20
2015
15
B
gRNA
1gR
NA2
gRNA
1gR
NA2
gRNA
1gR
NA2
WT R1446C KO
CREBBP
Actin280
4050
A
I RA T L C R P R F F HYE H V
I CA T L C R P R F F HYE H V
A A A
WT DNA
WT protein
Guide 1
Guide 2
HR template
Translation
Mondello et al.RESEARCH ARTICLE
444 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
M 100
80
60
40
20
0
Perc
ent s
urvi
val
0 5 10 15Overall survival
WTLOPKAT P.M.
P = 0.118
0 5 10 15Failure-free survival
L 100
80
60
40
20
0
Perc
ent s
urvi
val
P = 0.026
WTLOPKAT P.M.
HLA-DR
gRNA1
gRNA2
R1446C
KO
gRNA1
gRNA2gRNA1gRNA2
WT
0 10 4 105103
K
J chr6: 32,500,000 32,750,000
HLA-DRA HLA-DRB5HLA-DRB6
HLA-DRB1HLA-DQA1
HLA-DQB1 HLA-DQA2HLA-DQB2
HLA-DOB
gRNA1
MHC class II locus
gRNA2
gRNA1
gRNA2
gRNA1
gRNA2
WT
KOR
1446
C
[0-40]
[0-40]
[0-40]
[0-40]
[0-40]
[0-40]
REACTOME_CYTOKINE_SIGNALING_IN_IMMUNE_SYSTEMKEGG_CYTOKINE_CYTOKINE_RECEPTOR_INTERACTION
REACTOME_INTERFERON_GAMMA_SIGNALING
PID_IL4_2PATHWAYPID_BCR_5PATHWAY
BCL6_TARGETSBCL6-SMRT_TARGETS
CREBBP_mut_down_278_Green_FL
NFkB_all_OCILy3_Ly10
REACTOME_ANTIGEN_PROCESSING_AND_PRESENTATIONREACTOME_CLASS_I_MHC
H3K
27Ac
gai
n
H3K
27Ac
loss
Expr
essi
on g
ain
Expr
essi
on lo
ss
10.
050.
010.
001
1e-0
5
PNearest
gene
I (i)
(ii)
(iii)
(iv)
H
−0.5
0.0
Prim
ary
FL B
-cel
lC
REB
BP s
igna
ture
ES
NES = −1.64FDR < 0.001
−10
0
10
Log 2
(R14
46C
/ W
T)
1 5K 10K 15K 20K 23,250
−0.4
0.0
0.2 NES = −1.3FDR = 0.0171
−0.4
0.0
0.3NES = −1.22FDR = 0.0334
Ranked genes
Lym
phom
a ce
ll lin
esh
RN
A si
gnat
ure
ES
Mur
ine
B-ce
llsh
RN
A si
gnat
ure
ES
−10
0
10
Ranked genes
Log 2
(R14
46C
/WT)
1 5K 10K 15K 23,250
Zero cross at 8,635
G
−0.6
0.1
NES = −1.86FDR < 0.001
−0.1
0.5 NES = 1.7FDR < 0.001
Up inR1446C
Down inR1446C
HK2
7ac_
loss
gene
s,n
= 1,
299
ES
HK2
7ac_
gain
gene
s,n
= 1,
333
ES
F
gRNA1
gRNA2
−4 −2 0 2 4
WTgR
NA1
gRNA2
R1446C
Upr
egul
ated
n=
766
Dow
nreg
ulat
edn
= 73
3Log2 ratio (vs. mean WT)
IL4R
FCER2
CD74SP100
HLA-E
CREB1HLA-A
HLA-DPB1HLA-DOAHLA-DMBHLA-DOBHLA-DRAHLA-BHLA-DPA1HLA-FHLA-DQA2HLA-DRB5HLA-DMATAP1HLA-DQA1HLA-DRB1HLA-DQB1TAP2CIITASOCS1
gRNA1
gRNA2
KO
Figure 1. (Continued) F, A heat map of RNA-seq data shows that there are a similar number of genes with increased (n = 766) and decreased (n = 733) expression in CREBBPR1446C cells compared with isogenic WT controls. The CREBBPKO cells again show an intermediate level of change, with expression between that of CREBBPWT and CREBBPR1446C cells. G, Gene set enrichment analysis of the genes most closely associated with regions of H3K27Ac gain (above) or loss (below) shows that these epigenetic changes are significantly associated with coordinately increased or decreased expression in CREBBPR144C cells compared with isogenic WT controls, respectively. H, Gene set enrichment analysis shows those genes that were previously found to be downregulated following shRNA-mediated knockdown of CREBBP in murine B cells (top) or human lymphoma cell lines (middle) are also reduced in CREBBPR1446C-mutant cells compared with CREBBPWT cells. However, the most significant enrichment was observed for the signature of genes that we found to be significantly reduced in primary human follicular lymphoma (FL) with CREBBP mutation compared with CREBBP WT tumors. I, Hypergeomet-ric enrichment analysis identified sets of genes that were significantly overrepresented in those with altered H3K27Ac or expression in CREBBPR1446C cells. This included (i) gene sets associated with CREBBP mutation in primary tumors, (ii) BCL6 target genes, (iii) BCR and IL4 signaling pathways, and (iv) gene sets involving immune responses; antigen presentation and IFN signaling were significantly enriched. J, ChIP-seq tracks of the MHC class II locus on chromosome 6 are shown for isogenic CREBBPWT (blue), CREBBPR1446C (red), and CREBBPKO (orange) cells with regions of significant H3K27Ac loss shaded in gray. A significant reduction can be observed between CREBBPWT and CREBBPR1446C cells, with CREBBPKO cells harboring an intermediate level H3K27Ac over these loci. K, Flow cytometry for HLA-DR shows that reduced H3K27Ac over the MHC class II region is associated with changes of cell-surface protein expression. An approximately 2-fold reduction is observed in CREBBPKO cell compared with CREBBPWT, but a dramatic, approximately 39-fold reduction is observed in CREBBPR1446C cells. L, Kaplan–Meier plots show the failure-free survival in 231 previously untreated patients with folli-cular lymphoma according to their CREBBP mutation status. Nonsense/frameshift mutations that create a LOP are associated with a significantly better failure-free survival compared with KAT domain point mutations (KAT P.M.; log-rank P = 0.026). M, Kaplan–Meier plots show the overall survival in 231 previously untreated patients with follicular lymphoma according to their CREBBP mutation status. Patients with LOP mutations have a trend toward better overall survival, but this is not statistically significant (log-rank P = 0.118).
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 445
red blood cells and sacrificed 10 days later when the GC reac-tion was at its peak (Supplementary Fig. S6C), whereupon flow cytometry was performed to assess B-cell populations. As expected, there was no perturbation of earlier stages of B-cell development (Supplementary Fig. S6D and S6E). Cre recombination was validated in sorted GC B cells (Supple-mentary Fig. S6F). Cγ1Cre;CrebbpR1447C/+ GC B cells showed significantly reduced MHC class II expression compared with Cγ1Cre;Crebbp+/+ controls (P = 0.02, Supplementary Fig. S6G), whereas (as expected) there was no difference in MHC II among naïve B cells from either genotype (Supplementary Fig. S6H). Finally, Cγ1Cre;CrebbpR1447C/+ mice manifested a significantly skewed light-zone:dark-zone ratio in favor of increased dark-zone cells (P = 0.01; Supplementary Fig. S6I–S6L), which are the cells that most reflect the actions of BCL6, although the overall abundance of GC B cells was not increased (Supplementary Fig. S6M and S6N). These results confirm in primary cells our MHC class II findings from CRISPR/Cas9-edited clones and hence their value as a plat-form for mechanistic studies.
Finally, mutations in CREBBP have been previously associ-ated with adverse outcome in follicular lymphoma and are incorporated into the M7-FLIPI prognostic index (26). How-ever, in these analyses, all CREBBP mutations were considered collectively without discriminating between KAT domain point mutations and nonsense/frameshift LOP mutations. We therefore reevaluated these data in light of our observed functional differences between these mutations. This showed that there was a significant difference in failure-free survival (Fig. 1L) between patients with these two classes of CREBBP mutations. This was not significant for overall survival (Fig. 1M). Specifically, patients bearing KAT domain point mutations in CREBBP (22% of which were R1446 mutations) had a signifi-cantly reduced failure-free survival compared with patients with LOP mutations in CREBBP (log-rank P = 0.026), with no other clinical factors being significantly different between these groups (Supplementary Table S5). Collectively, these results provide the first direct experimental description of the role of CREBBPR1446C mutations in lymphoma B cells. Our data suggest that CREBBP KAT domain mutations may have a potential dominant repressive function on loci that can be targeted by redundant acetyltransferases, thereby driving a more profound molecular phenotype that is associated with a worse patient outcome.
Synthetic Dependence on HDAC3 in CREBBP-Mutant DLBCL Is Independent of Mutation Subtype
Our genomic analysis showed that BCL6 target genes were significantly enriched among those with reduced H3K27Ac and gene expression in CREBBPR1446C cells. We therefore evaluated CREBBP and BCL6 binding over these regions using data from normal GC B cells (24) and found that the regions with CREBBPR1446C mutation–associated H3K27Ac loss are bound by both proteins at significantly higher levels than H3K27Ac peaks that remain unchanged with CREBBP mutation (CREBBP Wilcoxon P = 1.2−41; BCL6 Wilcoxon P = 2.83−52) or peaks with increased H3K27Ac in CREBBPR1446C mutants (CREBBP Wilcoxon P = 3.01−137; BCL6 Wilcoxon P = 3.17−128; Fig. 2A). This suggests that
these genes may be antagonistically regulated by CREBBP and BCL6, the latter of which mediates gene repression via recruitment of HDAC3-containing NCOR/SMRT complexes (3, 27). The epigenetic suppression of gene expression in CREBBP-mutant cells may therefore be dependent upon HDAC3-mediated suppression of BCL6 target genes. Using a selective HDAC3 inhibitor, BRD3308 (28), we found that CREBBPR1446C and CREBBPKO clones showed greater sen-sitivity to HDAC3 inhibition compared with isogenic WT controls in cell-proliferation assays (Fig. 2B). We confirmed this as being an on-target effect of BRD3308 by performing shRNA knockdown of HDAC3 and observing a greater effect on cell proliferation in CREBBPR1446C cells compared with isogenic controls (Fig. 2C; Supplementary Fig. S7). More-over, HDAC3 inhibition was able to efficiently promote the accumulation of H3K27Ac in a dose-dependent manner in both CREBBPWT and CREBBPR1446C cells, as compared with the inactive chemical control compound BRD4097 (Fig. 2D). This suggests that the increased sensitivity to HDAC3 inhibi-tion in CREBBPR1446C cells may be linked with an acquired addiction to an epigenetic change driven by CREBBP muta-tion. We posited that one of these effects may be the suppres-sion of CDKN1A (encoding p21) expression, which is a key BCL6 target gene (29) that has reduced levels of H3K27Ac in both CREBBPR1446C and CREBBPKO cells (Fig. 2E). In support of this, we observed a marked induction of p21 expression by BRD3308 (Fig. 2F) and observed that shRNA-mediated silencing of CDKN1A partially rescued the effect of BRD3308 on cell proliferation (Fig. 2G). Therefore, CREBBP mutations, regardless of type, sensitize cells to the effects of HDAC3 inhibition in part via the induction of p21.
We aimed to confirm this trend using a larger panel of DLBCL cell lines with either WT (n = 6) or mutant CREBBP (n = 6). This revealed a significantly lower effective dose 50 (ED50) of BRD3308 in CREBBP mutant compared with WT cell lines (P = 0.002, Fig. 2H–I). We did not observe this trend using the nonspecific HDAC inhibitors romidep-sin and SAHA (Supplementary Fig. S8). Notably, none of these cell lines harbor KAT domain missense mutations in CREBBP, providing further evidence that sensitivity to HDAC3 selective inhibition is independent of mutation type (i.e., KAT domain missense vs. nonsense/frameshift). Furthermore, we also observed the induction of p21 expres-sion and apoptosis in CREBBP WT cell lines, although to a lesser degree (Supplementary Fig. S9). To gain greater insight into this, we performed RNA-seq of CREBBP WT (OCI-Ly1) and CREBBP-mutant (OCI-Ly19, OZ) cell lines treated with BRD3308. Notably, the ability of HDAC3 inhi-bition to induce the expression of genes involved in the terminal differentiation of B cells was conserved in both WT and mutant cell lines (Fig. 2J). These results are consistent with the role of BCL6 in controlling checkpoints, terminal differentiation, and other functions (1) and point toward induction of these transcriptional programs as contribut-ing to the antilymphoma response induced by HDAC3 inhibition in both the CREBBP WT and CREBBP-mutant settings. Although targetable by HDAC3 inhibition in WT cells, the BCL6–HDAC3 target gene set is more significantly perturbed in the context of CREBBP mutation leading to an enhanced cell-intrinsic effect of HDAC3 inhibition.
Mondello et al.RESEARCH ARTICLE
446 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
BRD3308 concentration (µmol/L)
H
I
P = 0.002
Cel
l pro
lifer
atio
n %
0
30
60
90
120
0 0.1 1 10 20
Cel
l pro
lifer
atio
n (%
)BR
D33
08/v
ehic
le
p21 shRNA1p21 shRNA2
gRNA1 gRNA2
20
15
10
5
0
25
72 h
ED 5
0 (µm
ol/L
) J
− + − − + − − + − − + −− +− − +− − +− − +−
**
** **
* *** **
* *
0
30
60
90
120CREBBP mut
CREBBP WT
CREBBP mut CREBBP WT
GC terminal differentiation genes (n = 1,495)
SUDHL6RIVA OZ
OCI-Ly1
0
OCI-Ly1
9
FARAGE
SUDHL5
DoHH2U29
32
OCI-Ly1
OCI-Ly3
OCI-Ly1
8
G
gRNA1 gRNA2WT R1446C
FARAGESUDHL6RIVA
SUDHL5OCI-Ly19OZ
OCI-Ly1OCI-Ly10
OCI-Ly18
OCI-Ly3U2932DoHH2
−0.1
0.1
0.3
0.5
Enric
hmen
t sco
re (E
S) NES = 1.26FDR < 0.001
−0.1
0.1
0.3
0.5
Enric
hmen
t sco
re (E
S) NES = 1.27FDR < 0.001
CREBBP WT CREBBP mut
Up inBRD3308-treated
Down inBRD3308-treated
Up inBRD3308-treated
Down inBRD3308-treated
EgRNA1
gRNA2
gRNA1
gRNA2
gRNA1
gRNA2
WT
R14
46C
KO
[0-18]
[0-18]
[0-18]
[0-18]
[0-18]
[0-18]
CDKN1A
FBRD3308
gRNA1 gRNA2gRNA1 gRNA2WT R1446C
CDKN1AACTIN
− + − + − + − +
20
40
D WT R1446CBRD4097 µmol/LBRD3308 µmol/L
00
100 1
0 05
010
00
100 1
0 05
010
H3K27acTotal H3
H3K27acTotal H3
gRNA1
gRNA220
2015
15
20
2015
15
A CREBBP BCL6CB4 CB5 CB4 CB5
5 kb 5 kb 5 kb 5 kb
H3K
27Ac
loss
R14
46C
vs.
WT
(n =
2,0
22)
H3K
27Ac
gai
nR
1446
C v
s. W
T (n
= 2
,304
)
0
5K27 gainK27 loss
0
5
CREBBP BCL6
Normalized reads (RPKM)0 10.5 1.5
Normalized reads (RPKM)0 10.5 1.5
C**
***
WT gRNA1 gRNA2
R1446C
Cel
l pro
lifer
atio
n %
(kno
ckdo
wn/
cont
rol)
0
40
60
80
100
20
shHDAC3 #1shHDAC3 #2
******
***
B
BRD3308 concentration (µmol/L)
Cel
l pro
lifer
atio
n %
0
30
60
90
120
0 0.1 1 10 20
gRNA1gRNA2
gRNA1gRNA2
gRNA1gRNA2
WT
R14
46C
KO
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 447
HDAC3 Inhibition Is Active against Primary Human DLBCL
Our observation that HDAC3 inhibition can affect both CREBBP mutant and WT B cells led us to test its effi-cacy in primary patient-derived xenograft (PDX) models of DLBCL. To achieve this, we first expanded CREBBP WT tumors in vivo and transitioned them to our novel in vitro organoid system for exposure to BRD3308 (30). All tumors that were tested showed a dose-dependent reduction in cell viability when cultured with BRD3308, compared with the vehicle control (Fig. 3A). We therefore treated 3 of these CREBBP WT tumors in vivo with either 25 mg/kg or 50 mg/kg of BRD3308, which led to a significant reduction in growth of PDX tumors treated with either dose compared with vehicle control (Fig. 3B). We were able to obtain only a single primary human CREBBP-mutant (R1446C) DLBCL PDX model, which could be grown only by renal capsule implantation. Treatment of these tumors with 25 mg/kg BRD3308 and monitoring by weekly MRI (Fig. 3C) showed a significant reduction in growth compared with vehicle control–treated tumors (Fig. 3D). In line with our cell line models, quantitative PCR analysis of treated PDX tumors showed an induction of BCL6 target genes with a role in B-cell terminal differentiation, including IRF4, PRDM1, CD138, and CD40 compared with vehicle-treated tumors (Supplementary Fig. S10). Therefore, selective inhibition of HDAC3 may be a rational approach for targeting the aber-rant epigenetic silencing of BCL6 target genes in primary human B-cell lymphoma.
Selective Inhibition of HDAC3 Reverts the Molecular Phenotype of CREBBP Mutations
We aimed to take a deeper look at the molecular conse-quences of HDAC3 inhibition by performing H3K27Ac ChIP-seq and RNA-seq of CREBBPR1446C-mutant cells after exposure to BRD3308, as compared with the inactive negative control compound BRD4097. This showed that selective HDAC3 inhi-bition promoted the gain of H3K27Ac over a broad number of regions (n = 6,756; Fig. 4A). Strikingly, HDAC3 inhibition significantly either restored or further increased the abun-dance of H3K27Ac at 28% (558/1,975) of sites that became deacetylated in CREBBPR1446C (hypergeometric enrichment P = 3.02−178), consistent with the role of HDAC3 in opposing CREBBP functions (Fig. 4B). Indeed, a more quantitative
analysis indicated HDAC3 inhibition coordinately increased the level of H3K27Ac over the same loci that showed reduced levels in CREBBPR1446C compared with CREBBPWT cells, although this restoration of H3K27Ac was not sufficient to completely revert the epigenomes of CREBBPR1446C cells to the level that was observed in CREBBPWT cells (Fig. 4C). In line with the role of HDAC3 and BCL6 in transcriptional repression, BRD3308 induced an expression profile that was markedly skewed toward gene upregulation (n = 1,467 vs. 208 genes downregulated; Fig. 4D; Supplementary Table S6 and S7), including upregulation of IFN-responsive genes such as antigen-presenting machinery and PD-L1 (CD274). Notably, the genes with increased expression following HDAC3 inhibi-tion were significantly enriched for those that lose H3K27Ac in CREBBPR1446C-mutant compared with WT (FDR < 0.001; Fig. 4E), further supporting the conclusion that HDAC3 inhi-bition directly counteracts changes associated with CREBBP mutation. A quantitative assessment of ChIP-seq signal indi-cated that enhancers manifested greater gain of H3K27Ac as compared with promoters in cells exposed to HDAC3i (Wilcoxon P < 0.001; Fig. 4F), although MHC class II genes also showed coordinate increases in promoter H3K27Ac (Fig. 4G). Analysis of critical gene loci that are deregulated by CREBBP mutation, such as the MHC class II gene cluster and CIITA, exemplify this induction of H3K27ac (Fig. 4H–I). We validated the increased H3K27Ac and expression of these genes in independent experiments wherein CREBBPR1446C or isogenic control cells were treated with BRD3308 or vehicle, H3K27Ac was measured by ChIP-qPCR (Supplementary Fig. S11), and transcript abundance was measured by qPCR (P <0.001; Fig. 4J). We further validated that this was an on-target effect of BRD3308 by performing shRNA-mediated knock-down of HDAC3, which resulted in the increased expression of these genes relative to control shRNA (Fig. 4K). Together, these data indicate that the aberrant mutant CREBBP epige-netic and transcriptional program can be restored by selective pharmacologic inhibition of HDAC3.
HDAC3 Inhibition Counteracts BCL6 Target Gene Repression in Lymphoma Cells Including IFN Response, Irrespective of CREBBP Mutation Status
IFN signaling and antigen-presenting genes have not been well investigated as downstream targets of BCL6–HDAC3
Figure 2. Synthetic dependence upon BCL6 and HDAC3 in CREBBP-mutant cells. A, A heat map shows that regions with reduced H3K27Ac in CREBBPR1446C cells compared with CREBBPWT cells (above) are bound by both CREBBP and BCL6 in normal germinal center B cells. This binding is not observed over regions with increased H3K27Ac in mutant cells. B, Isogenic CREBBPR1446C and CREBBPKO cells have a greater sensitivity to BRD3308, a selective HDAC3 inhibitor, compared with CREBBPWT cells. C, Knockdown of HDAC3 with two unique shRNAs shows a similar preference toward limiting cell proliferation in CREBBPR1446C cells compared with WT. Data are shown relative to control shRNA in the same cell lines (*, P < 0.05; ***, P < 0.001). D, Representative Western blots show a dose-dependent increase in H3K27Ac in both CREBBPWT and CREBBPR1446C cells treated with BRD3308, compared with the control compound BRD4097. E, ChIP-seq tracks of H3K27Ac show that CREBBPKO and CREBBPR1446C both have reduced levels over the CDKN1A locus compared with isogenic CREBBPWT cells. Regions that are statistically significant are shaded in gray. F, A representative Western blot shows that CDKN1A is induced at the protein level by treatment with 10 µmol/L BRD3308 in both CREBBPWT and CREBBPR1446C cells. G, Knockdown of CDKN1A (p21) using two unique shRNAs partially rescued the proliferative arrest of cells treated with BRD3308. This rescue was more significant in CREBBP-mutant cells compared with WT. Data are displayed relative to vehicle-treated cells (*, P < 0.05; **, P < 0.01; ***, P < 0.001). H, The difference in sensitivity to BRD3308 between CREBBP WT (blue) and CREBBP mutant (yellow to red) was validated in a large panel of DLBCL cell lines. I, The ED50 concentrations for each cell line from H are shown, colored by CREBBP mutation status. The ED50 for CREBBP-mutant (red) cell lines was significantly lower than that observed for CREBBP WT cell lines (blue; t test, P = 0.002). J, Gene-set enrichment analysis of “Germinal Center Terminal Differentiation” signature genes shows that these genes are coordinately induced in both CREBBP WT (above) and mutant (below) DLBCL cell lines by BRD3308 treatment compared with control.
Mondello et al.RESEARCH ARTICLE
448 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Figure 3. BRD3308 is effective in primary DLBCL. A, The sensitivity of primary DLBCL tumors to BRD3308 was evaluated by expanding them in vivo, followed by culture in our in vitro organoid model with different concentrations of BRD3308. A dose-dependent decrease in cell viability was observed in all 6 tumors with increasing concentrations of BRD3308. B, Treatment of three unique DLBCL xenograft models in vivo with 25 mg/kg (green) or 50 mg/kg (orange) of BRD3308 significantly reduced tumor growth compared with vehicle (black; **, P < 0.01; ***, P < 0.001). C, Representative MRI images of renal capsule implanted PDX tumors from a CREBBPR1446C-mutant DLBCL at the beginning (day 0) and day 14 of treatment. Tumor is outlined in yellow. D, Quantification of tumor volume by MRI images, normalized to the pretreatment volume for the same tumor, shows that BRD3308 treatment signifi-cantly reduces tumor growth (*, P < 0.05; **, P < 0.01).
PrimaryDLBCL
0
20
40
60
80
100
% C
ell v
iabi
lity
COBIL-11
DB-SCDB-RA
DR-NY2DANA
VehicleBRD3308 1.25 µmol/LBRD3308 2.5 µmol/LBRD3308 5 µmol/LBRD3308 10 µmol/L
In vivoexpansion
In vitroorganoid growth
A
***
1,500
1,000
500
0
Tum
or v
olum
e (m
m3 )
11 14 17 19 24 27 31Time after tumor injection (days)
DR-NY2
25 mg/kg BRD3308Vehicle
50 mg/kg BRD3308
**
800
600
400
200
0Tu
mor
vol
ume
(mm
3 )
8 11 14 18 21 24 29Time after tumor injection (days)
1,000 DANA
** ***
4,000
3,000
2,000
1,000
013 16 20 23 27 30 34 37
Tum
or v
olum
e (m
m3 )
Time after tumor injection (days)
TONYB
0
1
2
3
Rel
ativ
e tu
mor
vol
ume
Day 7 Day 14 Day 21
* ** *
VehicleBRD3308
++
++
++
DC
Day 14 Day 14Day 0 Day 0Vehicle BRD3308 (25 mg/kg)
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 449
Figure 4. HDAC3 inhibition counteracts the molecular phenotype of CREBBP mutation. A, A heat map shows the regions with significantly increased (above, n = 6,756) or decreased (below, n = 1,916) H3K27Ac in CREBBPR1446C cells treated with BRD3308 compared with those treated with the control compound, BRD4097. Experimental duplicates are shown for each clone. B, A river plot shows that a large fraction of the regions with significantly reduced H3K27Ac in CREBBPR1446C cells compared with CREBBPWT cells had significantly increased H3K27Ac following BRD3308 treatment. C, A density plot shows the regions with reduced H3K27Ac in CREBBPR1446C compared with CREBBPWT cells. The level of H3K27Ac over these regions is increased in CREBBPR1446C cells treated with BRD3308 compared with control (BRD4097), but does not reach the level observed in CREBBPWT cells. D, A heat map shows the genes with increased (above, n = 1,467) or decreased (below, n = 209) expression following BRD3308 treatment. Duplicate experiments are shown for each of the two CREBBPR1446C clones. IFN-responsive genes, including those with a role in antigen processing and presentation, can be observed to increase in expression following BRD3308 treatment. E, Gene set enrichment analysis shows that the set of genes with reduced H3K27Ac in associa-tion with CREBBP mutation has coordinately increased expression following BRD3308 treatment. F, A density plot illustrates the relative change in promoter (red) and enhancer (blue) H3K27Ac following treatment with BRD3308, with the enhancer distribution being significantly more right-shifted (increased) compared with promoter regions. (continued on next page)
R1446C gRNA1 R1446C gRNA2 R1446C gRNA1 R1446C gRNA2BRD4097 BRD3308
5 kb 5 kb5 kb 5 kb5 kb5 kb5 kb 5 kb
BRD
3308
H3K
27Ac
gai
n (n
= 6,
756)
BRD3308H3K27Ac loss
(n = 1,916)
NormalizedH3K27Ac(RPKM)
0
2
4
6
8
A
Baseline(R1446C vs. WT)
BRD3308vs. BRD4097
Gain(n = 2,299)
Loss(n = 1,998)
Gain in BRD3308 (n = 607)
Non-differentialin BRD3308 (n = 3,528)
Loss in BRD3308 (n = 162)
B
0
1
2
3
4
5
6
7
Distance from diffbind midpt
Nor
mal
ized
H3K
27Ac
(FPK
M)
−5 kb LociMidpoint
5 kb
WTBRD4097BRD3308
R1446CgRNA1 R1446C
gRNA2
C
CIITAHLA-DRAHLA-DPB1
HLA-DOAFCER2
IRF7
STAT1
IFNGR1
CD274CTSB
HSPA2
IRF5IL4R HLA-F
HLA-DQA2SOCS1
HLA-DQA1HLA-EHLA-DPA1HLA-DRB1HLA-DRB5TAP1TAP2HLA-DMAIL13RA1
HLA-DQB1HLA-BHLA-DMB
HLA-DOBIRF6
BRD4097 BRD3308
Upr
egul
ated
n=
1,46
7
Downregulatedn = 209
Log2 ratio(vs. meancontrol)
−4−2024
R1446C
D
−1.5 −1.0 −0.5 0.0 0.5 1.0 1.5
0.0
0.2
0.4
0.6
1.0
Log2 (BRD3308 H3K27ac/BRD4097 H3K27ac)D
ensi
ty
PromoterH3K27Ac
EnhancerH3K27Ac
WilcoxonP < 1e-300
0.8
FE
NES = 1.22FDR < 0.001
−6
−2
2
6
Zero cross at 14265
0
0.25
0.5
ES
Up inBRD3308
Down inBRD3308
Ranked genes1 5K 10K 15K 23250
HK27ac_loss genesn = 1,299
RL
R14
46C
log
FC(B
RD
3308
/Con
trol)
complexes, but were enriched in genes suppressed by CREBBP mutation and restored by HDAC3 inhibition. Given their critical role in antitumor immunity, we evaluated whether HDAC3 inhibition may be sufficient to restore or promote the expression of these BCL6-repressed immune signatures. Using MHC class II protein expression on CREBBPR1446C-mutant cells as a proxy for the CREBBP/BCL6 counter-regulated IFN signaling pathway, we evaluated the activity of HDAC inhibitors for promoting the expression of immune response genes. Although HDAC inhibitors with broader spe-cificities were able to induce MHC class II expression, selec-tive inhibition of HDAC3 was sufficient for the robust and maximal (>10-fold) restoration of MHC class II expression in CREBBPR1446C cells (Fig. 5A; Supplementary Fig. S12) in line
with our observation that these genes are silenced by BCL6–HDAC3 complexes. These selective inhibitors each show some inhibition of other HDACs (28), and their activity in opposing BCL6 function may in part be linked to inhibition of HDAC1/2, which are recruited by BCL6 via CoRest and NuRD (4, 31, 32). However, HDAC1/2 have important func-tions in normal hematopoiesis (33), and hence compounds that target these enzymes induce toxic effects against these cells that are not elicited by BRD330829 (29), suggesting that selective inhibition of HDAC3 may limit hematologic toxicities that are observed with pan-HDAC inhibitors. Fur-thermore, although some of the less specific HDAC inhibitors were toxic to normal human CD4+ and CD8+ T cells, the selec-tive inhibition of HDAC3 was not (Fig. 5B; Supplementary
Mondello et al.RESEARCH ARTICLE
450 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Fig. S13). This suggests that selective HDAC3 inhibition may be capable of eliciting immune responses by on-target on-tumor induction of antigen presentation without on-target off-tumor killing of T cells.
Based upon our observations with MHC class II, the mag-nitude of this induction appeared to be dependent upon the baseline of expression. We therefore hypothesized that CREBBP mutation status may determine the baseline expres-sion for IFN and antigen-presenting pathways, but may not be a prerequisite for this response due to the conserved activ-ity of BCL6–HDAC3 in regulating this axis in WT cells. We comprehensively evaluated this theory using our RNA-seq data from CREBBPWT and CREBBPR1446C cells treated with either the HDAC3 inhibitor BRD3308 or the inactive control compound BRD4097. Consistent with our observations of MHC class II protein expression, BRD3308 treatment coor-dinately induced the transcript expression of MHC class II and IFN pathway genes in both CREBBPWT and CREBBPR1446C CRISPR-edited cells (Fig. 5C). This trend included a signifi-cant enrichment for the genes that were epigenetically sup-pressed in association with CREBBP mutations, resulting in their significant increase in expression in CREBBPWT cells (Fig. 5D), similar to our observations from CREBBPR1446C cells (Fig. 4E). Moreover, we observed similar enrichment of transcriptionally induced pathways in both CREBBPWT and
CREBBPR1446C cells that included the same pathways that were suppressed by CREBBP mutation (Fig. 5E; Supplementary Tables S7–S10). Consistent with the almost exclusive func-tion of HDAC3 as a BCL6 corepressor in GC B cells (3), and the importance of BCL6 activity in both CREBBP WT and CREBBP-mutant cells, we observed highly significant enrich-ment for genes regulated by BCL6–SMRT complexes among genes with induced H3K27Ac and expression after BRD3308 treatment (Fig. 5E). Also significantly enriched were canoni-cal BCL6 target gene sets such as p53-regulated genes, and signaling through BCR, CD40, and cytokines such as IL4 and IL10. Finally, we observed significant enrichment for BCL6 target gene sets linked to IFN signaling, antigen presentation via MHC class II, and PD-1 signaling.
Although there were conserved patterns of gene activation in both CREBBPWT and CREBBPR1446C cells, we observed that the magnitude of this induction was greatest in CREBBPR1446C
cells, which started from a lower baseline of expression linked to mutation-associated epigenetic suppression (Fig. 5F). We identified IRF1 as a BCL6-regulated transcription factor that is critical for IFN responses, and is induced by HDAC3 inhibition in CREBBPWT and CREBBPR1446C cells (Fig. 5C). We therefore hypothesized that IRF1 may contribute to the different magnitude of induction in MHC class II genes following HDAC3 inhibition in these two genetic contexts.
11,000,000chr16: 10,960,000
R14
46C
gR
NA1
R14
46C
gR
NA2
BRD4097
BRD3308
BRD4097
BRD3308
CIITA
I
HLA-D
RACIIT
APD-L1 IRF1
0
3
6
9
12R
elat
ive
trans
crip
t abu
ndan
ce(B
RD
3308
/Veh
icle
)
WT gRNA1WT gRNA2
R1446C gRNA1R1446C gRNA2
*** **
**
*
**
******
J
8
6
4
2
0
Rel
ativ
e tra
nscr
ipt a
bund
ance
(shH
DAC
3/co
ntro
l)
HLA-D
RCIIT
APD-L1 IRF1
HDAC3sh1HDAC3sh2
HDAC3sh1HDAC3sh2
R1446C gRNA1 R1446C gRNA2
**
****
***
** ** * *
*** ***
**
***
****
**
K
HLA-DRB1 HLA-DQA1 HLA-DQB1
R14
46C
gR
NA1
R14
46C
gR
NA2
BRD4097
BRD3308
BRD4097
BRD3308
chr6: 32,550,000 32,600,000 32,650,000HG
HLA-DPA1HLA-DPB1 HLA-DMAHLA-DMB HLA-DRB4HLA-DRA HLA-DRB1HLA-DPA1 HLA-DPB1HLA-DMA HLA-DMBHLA-DRA HLA-DPA1HLA-DOA HLA-DMAHLA-DMB HLA-DRB1HLA-DRA HLA-DPA1HLA-DPB1 HLA-DOAHLA-DMB HLA-DRB4HLA-DRA HLA-DMAHLA-DPA1 HLA-DRAHLA-DPB1 HLA-DOBHLA-DPB1 HLA-DPA1HLA-DMA HLA-DOAHLA-DOB HLA-DRB3HLA-DRB1 HLA-DPB1HLA-DMA HLA-DRB1HLA-DPA1 HLA-DMB
CIITA
HLA-DPA1 HLA-DPA1HLA-DMB HLA-DOAHLA-DOB HLA-DQB1HLA-DRB1
HLA-DPA1HLA-DOBHLA-DOAHLA-DPA1HLA-DOAHLA-DPB1HLA-DPA1HLA-DQB1HLA-DPA1HLA-DMAHLA-DOBHLA-DOAHLA-DMBHLA-DPA1HLA-DPA1HLA-DMBHLA-DQB1HLA-DRAHLA-DOBHLA-DPA1HLA-DPB1HLA-DMAHLA-DQA1HLA-DRB5HLA-DRA
−1 0 1
Log2 (BRD3308 H3K27ac/BRD4097 H3K27ac):
R1446CgRNA1
R1446CgRNA1
Figure 4. (Continued) G, A heat map shows the change in H3K27Ac at the promoter regions of MHC class II genes following BRD3308 treatment of CREBBPR1446C cells, showing a coordinate increase. H and I, Regions with significantly increased H3K27Ac (shaded in gray) included those within the MHC class II and CIITA gene loci. J, The increased expression of candidate genes within the IFN signaling and antigen-presenting pathways was confirmed by qPCR. Increased expression was observed in both CREBBPWT and CREBBPR1446C cells following BRD3308 treatment, but the level of induction was much higher in CREBBPR1446C cells. Data are shown relative to vehicle-treated cells (t test, *, P < 0.05; **, P < 0.01; ***, P < 0.001). K, The on-target role of HDAC3 in the induction of candidate genes was confirmed by shRNA-mediated knockdown of HDAC3 and qPCR analysis of gene expression. Knockdown of HDAC3 was able to induce the expression of all genes, which is shown relative the control shRNA (t test, *, P < 0.05; **, P < 0.01; ***, P < 0.001).
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 451
Figure 5. HDAC3 inhibition induces IFN signaling and antigen presentation in both CREBBP WT and CREBBP-mutant cells. A, Flow cytometry was per-formed for HLA-DR following exposure to a selection of HDAC inhibitors at 10 µmol/L for 72 hours. This shows that HDAC inhibitors with range specifici-ties are able to induce MHC class II, but HDAC3-selective inhibition using BRD3308 is sufficient for this effect. B, Dose titrations of histone deacetylase inhibitors from A with peripheral blood CD4+ and CD8+ T cells from healthy donors. C, A heat map of IFN responsive and antigen-presenting genes from RNA-seq data shows an increased expression in both CREBBPWT and CREBBPR1446C cells. Data represent duplicate experiments for each clone and are normalized to control-treated cells from the same experiment. D, Gene set enrichment analysis of the genes that have reduced H3K27Ac in CREBBPR1446C cells shows that the expression of these same genes is coordinately increased by BRD3308 treatment in CREBBPWT cells. E, A heat map of hypergeomet-ric enrichment analysis results of RNA-seq data shows that BRD3308 induces the induction of similar gene sets in both CREBBPWT and CREBBPR1446C cells. F, A density strip plot, normalized to the mean expression in control (BRD4097)-treated CREBBPWT cells, shows the relative expression of the set of genes with reduced H3K27Ac in CREBBPR1446C cells. This shows that these genes are induced by BRD3308 in CREBBPWT cells, resulting in expression lev-els greater than baseline. Furthermore, CREBBPR1446C cells can be observed to start below baseline, with the induction by BRD3308 resulting in expres-sion levels similar to that observed in control-treated CREBBPWT cells. The four samples per condition represent duplicate experiments in each of the two clones for each genotype. G, The firefly luciferase luminescence of two unique IRF1 reporters (R1 and R2) is shown, normalized to Renilla luciferase from a control vector and shown as fold change compared with untreated cells. CREBBPWT cells show increased IRF1 activity following IFNγ treatment (posi-tive control; gray), but not following treatment with BRD3308 (green). In contrast, CREBBPR1446C cells show increased IRF1 activity following BRD3308 treatment, to a level that is similar to that observed with IFNγ treatment (t test vs. control-treated cells, **, P < 0.01; ***, P < 0.001). H, The role of IFNγ in inducing MHC class II expression following BRD3308 in CREBBPR1446C cells was assessed with a blocking experiment. Blocking IFNγ with a neutralizing antibody (αIFNγ) significantly reduced the induction of MHC class II, as measured by flow cytometry for HLA-DR, but the induction by BRD3308 with αIFNγ remained significantly higher than vehicle with αIFNγ (t test, ***, P < 0.001).
Concentration (µmol/L)
0
20
40
60
80
100
120
0
20
40
60
80
100
120
00.0
80.1
60.3
10.6
31.2
52.5
05.0
010
.00 00.0
80.1
60.3
10.6
31.2
52.5
05.0
010
.0020
.00
CD4+ T cells CD8+ T cells
Cel
l pro
lifer
atio
n %
20.00
Control (BRD4097)BRD3308
CI-994BRD2492
VorinostatEntinostat
B
IRF1CD274CIITAHLA-DMBHLA-DMAHLA-DOBHLA-DOAHLA-DPB1HLA-DPA1HLA-DQB1HLA-DQA1HLA-DRB5HLA-DRB4HLA-DRB3HLA-DRB1HLA-DRA
−4−2024
Log2 FPKM(BRD3308/BRD4097)
WT R1446CC
2.5
2.0
1.5
1.0
0.5
0Fold
cha
nge
IRF1
repo
rter
(Luc
ifera
se/re
nilla
)
R1 R2 R1 R2 R1 R2 R1 R2WT gRNA1 WT gRNA2 R1446C
gRNA1R1446CgRNA2
BRD3308 10 µmol/LIFNγ 10 ng/mL
*** *** **** ** **
*** *** *** ****** ***
G2,000
1,500
1,000
500
0
HLA
-DR
gM
FI
*** ***
R1446CgRNA1
R1446CgRNA2
Vehicle + αIFNγBRD3308 + αIFNγ
Vehicle + IsotypeBRD3308 + Isotype******
H
0
0.25
0.5
ES
NES = 1.16FDR = 0.001
−4
0
4
8
Ranked genes1 5K 10K 15K 23250
Zero cross at 9595
HK27ac_loss genes (n = 1,299)
Up inBRD3308
Down inBRD3308
RL
WT
log
FC(W
T BR
D33
08/B
RD
4097
)
D
−5
0
5
Log 2
(vs.
mea
n W
T BR
D40
97)
BRD4097 BRD4097BRD3308 BRD3308WT R1446C
F
WT
R1446
C 1 0.050.010.0011e-05
P
REACTOME_PD1_SIGNALING
LU_IL4_SIGNALING
REACTOME_MHC_CLASS_II_ANTIGEN_PRESENTATION
KEGG_ANTIGEN_PROCESSING_AND_PRESENTATION
PID_IL4_2PATHWAY
PID_P53_DOWNSTREAM_PATHWAY
BCL6-SMRT_TARGETS
E
A
BRD4097
CI-994BRD2492
BRD3308
EntinostatVorinostat
HLA-DRH
DAC
1H
DAC
2H
DAC
3H
DAC
8H
DAC
6
IC50 <1 µmol/LYes No WT gRNA1
0 104 105
R1446C gRNA1
0 104 105
WT gRNA2
0 104 105
R1446C gRNA2
0 104 105
Mondello et al.RESEARCH ARTICLE
452 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
We investigated this by measuring IRF1 activity in a lucif-erase reporter assay in CREBBPWT and CREBBPR1446C-mutant cells. Treatment of CREBBPR1446C cells with BRD3308 led to a significant increase of IRF1 activity that was similar in magnitude to that induced by IFNγ treatment (Fig. 5G). In contrast, this effect was not observed in isogenic CREBBPWT control cells following HDAC3 inhibition, despite these cells showing similarly increased IRF1 activity in response to IFNγ treatment. Exposing CREBBPR1446C cells to IFNγ-neutralizing antibodies only partially ameliorated MHC class II induction after BRD3308 in CREBBP-mutant cells (Fig. 5H). This sug-gests that preferential induction of antigen presentation by HDAC3 inhibition in CREBBP-mutant cells likely depends on a combination of mechanisms, including directly opposing BCL6 repression of these genes as well as secondary induc-tion through IFNγ signaling (which is also directly regulated by BCL6).
HDAC3 Inhibition Restores MHC Class II Expression in Human DLBCL Cell Lines and Patient Specimens
The frequency of MHC class II loss in DLBCL exceeds the frequency of CREBBP mutations in this disease (13, 21), through unknown mechanisms. The ability of HDAC3 inhibition to induce MHC class II expression in CREBBP WT DLBCL cells may therefore have potentially broad implica-tions for immunotherapy. Using RNA-seq data from WT (OCI-Ly1) and mutant (OCI-Ly19 and OZ) DLBCL cell lines (Fig. 6A; Supplementary Tables S11 and S12) and qPCR analysis of PDX tumors treated in vivo (Supplementary Fig. S14A–S14D), we confirmed that gene expression induced by HDAC3 inhibition was largely conserved in both contexts. The expression of CD274 (encoding PD-L1) was significantly higher in 4 of 4 PDX tumors treated with 25 mg/kg BRD3308 in vivo compared with vehicle control–treated tumors, and MHC class II expression was significantly higher in 3 of 4 PDX tumors (Supplementary Fig. S14A–S14D). We extended upon this using IHC staining for MHC class II in tumors from a CREBBP R1446C–mutant PDX model (DFBL; Fig. 6B) and a CREBBP WT PDX model that was MHC class II negative at baseline (DR-NY2, Fig. 6C). In tumors from both models, in vivo treatment with BRD3308 led to a marked induction of MHC class II expression. In a broader panel of CREBBP WT and CREBBP-mutant cell lines, we observed that a core set of genes including HLA-DR, CIITA, and CD274 had consistently higher expression in BRD3308-treated cells compared with the matched control, but with a higher mag-nitude of induction in CREBBP-mutant cell lines (Fig. 6D). This trend was also observed by flow cytometry for MHC class II, which extends upon our observations in CRISPR/Cas9-modified cells by showing a reproducible increase in expression in a larger panel of DLBCL cell lines, and a higher magnitude of induction in CREBBP-mutant cell lines (Fig. 6E). Together, these results show that selective inhibition of HDAC3 using BRD3308 can promote the expression of IFN and antigen-presenting pathway genes in both the CREBBP WT and CREBBP-mutant settings. However, the magnitude of induction is greatest in CREBBP-mutant cells, owing in part to the preferential induction of IRF1 activity in these cells.
HDAC3 Inhibition Drives T-cell Activation and Immune Responses
IFN signaling and antigen presentation are central to anti-tumor immune responses. We therefore investigated whether HDAC3 inhibition could promote antigen-dependent antitu-mor immunity. To test this, we implanted OCI-Ly18 DLBCL cells into NSG mice, and once tumors formed we injected human peripheral blood mononuclear cells (PBMC), includ-ing T cells, to expose them to antigens presented by the tumor cells. These are likely to be histocompatibility antigens rather than tumor neoantigens, but are nonetheless pre-sented and recognized through MHC–T-cell receptor (TCR) interactions. After in vivo priming, tumor-infiltrating lym-phocytes (TIL) were cocultured in vitro with OCI-Ly18 cells that were pretreated for 72 hours with increasing concentra-tions of BRD3308 to assess the effects on T-cell activation and tumor-cell killing (Fig. 7A). The DLBCL cells that were epigenetically primed for antigen presentation by BRD3308 significantly increased the activation of CD4+ T cells, as deter-mined by CD69 expression (P < 0.05; Fig. 7B). As in prior experiments, we observed cell-intrinsic effects of BRD3308 on OCI-Ly18 cells in the absence of TILs, resulting in declin-ing cell viability with higher concentrations of the inhibitor. However, the effects of BRD3308 were markedly and sig-nificantly increased in the presence of TILs, consistent with T cell–directed killing of the tumor cells (P < 0.001; Fig. 7C). To confirm that this killing was dependent on MHC–TCR interactions, we also performed this experiment in the pres-ence of blocking antibodies for MHC class I, MHC class II, or both. Blocking one or the other of MHC class I or II rescued some of the cytotoxicity observed in this assay, but blocking both MHC class I and class II together completely abro-gated the TIL-associated effect (Fig. 7C). These data show that HDAC3 inhibition can potentiate antitumor immune responses that are likely to be antigen-dependent because they are driven by MHC–TCR interactions. Our identification of the IFN signaling pathway, and IFNγ itself, as a central component of the response to HDAC3 inhibition led us to test whether IFNγ levels may rise as a result of treatment. An ELISpot analysis of the IFNγ levels from the TIL and OCI-Ly18 coculture experiment revealed a striking and sig-nificant increase in IFNγ levels, even at the lowest concentra-tions of HDAC3 inhibitor (Fig. 7D; Supplementary Fig. S15).
Considering the manner in which to best harness the antilymphoma immunity effect of HDAC3 inhibitors, we noted that induction of MHC class II is mechanistically linked to IFNγ-associated PD-L1 upregulation, which could potentially limit maximal antitumor response. We therefore posited that PD-1/PD-L1 blockade may be an attractive com-bination regimen. To test this hypothesis, we used a murine lymphoma transplantation model in which splenocytes from IµBcl6;Ezh2Y641F mice (5) were transplanted into irradi-ated syngeneic WT recipients (Fig. 7E). Aggressive tumors formed within these mice, which we treated with either vehi-cle control, BRD3308 alone, αPD-L1 alone, or a BRD3308 + αPD-L1 combination. Treatment with BRD3308 led to a significant increase in serum IFNγ within these mice, which was also observed with αPD-L1 treatment and was even further enhanced in an additive manner by the combination
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 453
Figure 6. Induction of IFN-responsive and antigen-presenting genes in DLBCL cell lines and patient-derived xenografts. A, A heat map shows signifi-cantly upregulated (above) and downregulated (below) genes in BRD3308-treated DLBCL cell lines that are CREBBP WT (OCI-Ly1) or CREBBP-mutant (OCI-Ly19 and OZ), expressed as a log2 ratio to vehicle control–treated cells. The observed changes were consistent between WT and mutant cell lines, and included upregulation of IFN-responsive and antigen-presenting genes. B, MHC class II was assessed on vehicle control (left) and BRD3308-treated (25 mg/kg; right) tumors from a CREBBP R1446C–mutant PDX model, showing a visible increase in expression in the BRD3308-treated tumors. These images are representative of 4 tumors per group. C, An MHC class II–negative DLBCL patient-derived xenograft model was treated in vivo with either 25 mg/kg or 50 mg/kg of BRD3308. IHC staining was performed for MHC class II, revealing a robust induction of MHC class II expression that was relative to the dose of treatment. These images are representative of 6 tumors per group. D, qPCR was used to validate the gene expression changes of select IFN-responsive genes following BRD3308 treatment across an extended panel of CREBBP WT and CREBBP-mutant DLBCL cell lines. These genes were uniformly increased in both genetic contexts, but with a higher magnitude of increase in CREBBP-mutant cell lines. One-tailed Student t test, *, P < 0.05; **, P < 0.01; **, P < 0.001. E, The induction of MHC class II expression by BRD3308 was measured in an extended panel of DLBCL cell lines by flow cytometry. Data are plotted as a fold change of the mean fluorescence intensity (MFI) of HLA-DR in BRD3308-treated versus control-treated cells. We observed uniformly increased MHC class II expression in all cell lines, but with higher magnitude in CREBBP mutants.
−4−2024
B2MCD74IFNGR1HLA-DRB1HLA-DPB1CTSBHLA-DMBHLA-DRAHLA-DQB1HLA-EHLA-DMAMT2ASOCS1HSPA2IRF5
HLA-DOA
ITGB3IL13RA1IRF6
FCER2
Log2 (BRD3308/vehicle)
OZ
OCI-Ly1
OCI-Ly1
9
A
MH
C C
lass
II
Treatin vivo
HumanDLBCL
VehicleBRD330825 mg/kg
DFBL (CREBBPR1446C )
50µm
100
µm
B
VehicleBRD330825 mg/kg
BRD330850 mg/kg
100
µmol
/L1,
000
µmol
/L
DR-NY2 (CREBBPWT )
MH
C c
lass
II
C
1815129630
HLA-D
RCIIT
APD-L1 IRF1
Rel
ativ
e tra
nscr
ipt a
bund
ance
(BR
D33
08/v
ehic
le)
FARAGEOCI-Ly19SUDHL6OZ
CREBBP WT CREBBP mut
OCI-Ly7OCI-Ly1SUDHL4
OCI-Ly18
D
Fold
cha
nge
HLA
-DR
MFI CREBBP WT CREBBP mut
OCI-Ly7
OCI-Ly1
SUDHL4
FARAGE
OCI-Ly1
9
SUDHL6OZ
OCI-Ly1
8
5
4
3
2
1
0
E
Mondello et al.RESEARCH ARTICLE
454 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
150
DMSO
140
130
120
CD
69 M
FI(C
D4+ c
ells
)
110
1001 5 10
BRD3308 (µmol/L)15
* * *
*
BOCI-LY18
OCI-LY18 + TIL + Isotype
OCI-LY18 + TIL + αMHC-I
OCI-LY18 + TIL + αMHC-II
OCI-LY18 + TIL + αMHC-I + αMHC-II
C
00
30
60
90
120C
ell v
iabi
lity
(%)
BRD3308 (µmol/L)0.1 1 10 20
*
****
120
DMSO
90
60
30
INFγ
SFC
01 5 10
BRD3308 (µmol/L)15
*** ***** ***
D
Human DLBCL cells(OCI-Ly18)
HumanT cells
HumanTILs
HDAC3ior vehicle
T-cell activation
DLBCL viability
PrimedDLBCL
Fresh DLBCL cells(OCI-Ly18)
A
Splenocytes
DonorEzh2Y641 × lmBcl6
Tail-vein injectionof splenocytes intoirradiated C57BL/6
Lymphomaonset
E F80
*********
****
UntreatedBRD3308αPD-L1BRD3308+αPD-L1
60
40
20
0
INFγ
(pg/
mL)
Vehicle
CD
8 100 µmol/L
BRD3308 αPD-L1BRD3308+
αPD-L1G
50 µmol/L
CD
4
I
100 µmol/L
50 µmol/L
PD-L
1
K
100 µmol/L
50 µmol/L
B220
M
100 µmol/L
50 µmol/L
H3,000
*****
**
2,000
1,000
0
Untreated
WT splenocytes
BRD3308
αPD-L1
BRD3308+ αPD-L1
CD
8 gM
FI
Ezh2
Y641
× lµ
Bcl6
L5,0004,0003,0002,0001,000
0
PD-L
1 gM
FI
*****
J3,000
2,000
1,000
0
CD
4 gM
FI
****
***
N5,0004,0003,0002,0001,000
0
B220
gM
FI
***
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 455
treatment (Fig. 7F). Immunofluorescence staining for CD4 and CD8 showed significantly increased infiltration within the spleens from BRD3308-treated mice, which was further increased by the combination therapy (Fig. 7G–J). The ability of BRD3308 and αPD-L1 combination to increase TILs was likely linked with the IFNγ-mediated PD-L1 induction that was observed in BRD3308-treated tumors (Fig. 7K and L). In line with this interaction, BRD3308 and αPD-L1 each showed a small degree of single-agent efficacy, but the combination led to the most significant reduction in B220 cells within the spleens of tumor-bearing mice (Fig. 7M and N). Together, these data show that HDAC3 inhibitor–mediated reversal of the BCL6-repressed IFN pathway leads to joint induction of MHC class II and PD-L1, which, although eliciting signifi-cant antitumor immune response, can be further enhanced by combination with PD-1/PD-L1 blockade to yield a more potent immunotherapy strategy that is superior to immune checkpoint inhibition alone.
DISCUSSIONPrecision medicine and immunotherapy have led to sig-
nificant breakthroughs in a variety of cancers, but have lacked success in B-cell lymphoma. For precision medicine, this is largely due to a paucity of “ actionable” genetic alterations or, rather, the lack of current therapeutic avenues to target the mutations that have been defined as being important for disease biology. For immunotherapy, the mechanisms driving lack of response or early progression are not well understood, but are likely to be underpinned by the complex immune microenvironment and genetic alterations that drive immune escape. The exception to both of these statements is the use of PD-1–neutralizing antibodies in classic Hodgkin lymphoma, which opposes genetically driven immune suppression by the malignant Reed–Sternberg cells through DNA copy-number gain of the PD-L1 locus (34) and elicits responses in the majority of patients (35). This stands as an example of the potential success that could be achieved by the characteriza-tion and rational therapeutic targeting of genetic alterations and/or the neutralization of immune escape mechanisms. However, we are not yet able to successfully target some of the most frequently mutated genes or overcome the barriers of inadequate response to immunotherapy in the most com-mon subtypes of B-cell lymphoma. These are important areas
of need if we hope to further improve the outcomes of these patients.
The CREBBP gene is mutated in approximately 15% of DLBCL (21) and approximately 66% of follicular lymphoma (7), and is therefore a potentially high-yield target for preci-sion therapeutic approaches. Our use of CRISPR/Cas9 gene editing to generate isogenic lymphoma cell lines that differ only in their CREBBP mutation status allowed us to perform detailed characterization of the epigenetic and transcriptional consequences of these mutations. Using this approach, we identified for the first time functional differences between the most frequent KAT domain point mutation, R1446C muta-tions, and frameshift mutations that result in KO. Although similar regions of the genome showed reduced H3K27Ac in R1446C and KO mutants compared with isogenic WT con-trols, the magnitude of these changes was markedly reduced in CREBBP KO. This suggests that acetyltransferases such as EP300 may compensate for the loss of CREBBP protein in the setting of CREBBP nonsense/frameshift mutations, con-sistent with observations of functional redundancy between Crebbp and Ep300 in B cell–specific conditional KO mice (25, 36). But the presence of a catalytically inactive CREBBPR1446C
protein may dominantly suppress histone acetylation by pre-venting the participation of redundant acetyltransferases such as EP300 in transcriptional activating complexes. This yields a more severe epigenetic and transcriptional pheno-type in CREBBPR1446C mutants compared with CREBBPKO
mutants, and an inferior clinical outcome.Despite differences in the magnitude of molecular changes
between R1446C and KO cells, these mutations yielded a sim-ilar degree of synthetic vulnerability to HDAC3 inhibition. This was likely driven by the increased suppression of BCL6 target genes that we observed in both contexts, including CDKN1A (encoding p21). This gene has also been highlighted as a critical nexus in the oncogenic potential of EZH2 (37), which cooperates with BCL6 to silence gene expression (5). One of the important mechanisms for BCL6-mediated gene silencing is through the recruitment of HDAC3 as part of the NCOR/SMRT complex (3), thereby highlighting HDAC3 inhibition as a rational therapeutic avenue for counteract-ing BCL6 activity. Virtually all of the HDAC3 corepressor complexes present in DLBCL cells are bound with BCL6, suggesting that HDAC3 inhibitors’ effect is largely explained by their depression of the subset of BCL6 target genes
Figure 7. HDAC3 inhibition induces antigen-dependent immune responses. A, A schematic of the generation of antigen-specific T cells and epigenetic priming of DLBCL cells. A human DLBCL cell line (OCI-Ly18) was engrafted into immunodeficient mice and allowed to establish. Human T cells were then engrafted, exposing them to tumor antigens prior to harvesting of the tumor-infiltrating T-cell (TIL) fraction. These TILs were cultured with fresh DLBCL cells that had been epigenetically primed with different concentrations of BRD3308, and the cell viability of the DLBCL cells was measured after 72 hours. B, TIL and DLBCL coculture resulted in activation of the CD4+ T cells in a dose-dependent manner, as measured by flow cytometry for the CD69 activation marker. Data represent the fold change in CD69 expression compared with vehicle-treated DLBCL cells (t test vs. DMSO control, *, P < 0.05). C, The cell viability of DLBCL cells in TIL coculture experiments was measured by CellTiter-Blue assay. Treatment with BRD3308 resulted in some cell killing through cell-intrinsic mechanisms in the absence of TILs (black). The addition of TILs at a 1:1 ratio led to a significant increase in cell death of the DLBCL cells. This was partially reduced by blocking of either MHC class I or MHC class II using neutralizing antibodies. Blocking of MHC class I and class II together completely eliminated the TIL-associated increase in cell death, suggesting that killing was mediated through MHC–TCR interactions (t test, *, P < 0.05; ***, P < 0.001). D, The production of IFNγ was measured by ELISpot and found to increase in cultures with epigenetically primed DLBCL cells (t test vs. DMSO control, **, P < 0.01; ***, P < 0.001). SFC, spot-forming cells. E, A syngeneic BCL6-dependent lymphoma model for in vivo testing of BRD3308 and PD-L1–blocking antibodies. Splenocytes were taken from Ezh2Y641;IµBcl6 mice and injected into irradiated WT recipients that were treated upon the onset of lymphoma. F, Serum IFNγ levels measured in mice following treatment. G–N, Representative immunofluorescence images of mouse spleens fol-lowing treatment and quantification of mean fluorescence intensities (MFI) from multiple mice for CD8 (G and H), CD4 (I and J), PD-L1 (K and L), and B220 (M and N), showing increased T-cell infiltration following treatment with BRD3308 and cooperation with αPD-L1 in eliminating B220+ tumor cells within the spleen (t test, *, P < 0.05; **, P < 0.01; ***, P < 0.001).
Mondello et al.RESEARCH ARTICLE
456 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
regulated through this mechanism (3). Importantly, as a variety of GCB-derived malignancies rely on BCL6 function independently of CREBBP or EZH2 mutation (38, 39), oppos-ing this function through HDAC3 inhibition may also be effective in tumors without these genetic alterations. We have shown preliminary evidence in the primary setting using DLBCL PDX models.
One of the important transcriptional programs that is regulated by BCL6 is the IFN signaling pathway (3), which we observed to be significantly repressed in CREBBP-mutant cells. It has been also long been recognized that IFNγ cooper-ates with lymphocytes to prevent cancer development (40). IFN signaling supports T cell–driven antitumor immunity via a variety of mechanisms, including the induction of anti-gen presentation on MHC class II (41), and the expression of MHC class II–restricted tumor antigens is required for the spontaneous or immunotherapy-induced rejection of tumors (14). We have shown that the selective inhibition of HDAC3 is sufficient for broadly restoring the reduced H3K27Ac and gene expression that is associated with CREBBP muta-tion, including the IFN signaling and antigen-presenting programs. This was in part driven by the increased produc-tion of IFNγ following HDAC3 inhibition, but also via the induced expression and activity of the IRF1 transcription factor. Together, these factors led to a robust restoration of MHC class II expression on CREBBP-mutant cells and drove dose-dependent potentiation of antitumor T-cell responses. However, we also noted that the same molecular signature is promoted by HDAC3 inhibition in CREBBP WT cells, also with an associated increase of MHC class II expression. As with the cell-intrinsic effects of HDAC3 inhibition, the effects on immune interactions may therefore be active in both CREBBP WT and CREBBP-mutant cells as a result of the conserved molecular circuitry controlling these pathways in each genetic context.
A variety of HDAC inhibitors were capable of restoring antigen presentation in our models, and clinical responses are observed with these agents in patients with relapsed/refrac-tory follicular lymphoma and DLBCL. However, grade 3 to 4 hematologic toxicities such as thrombocytopenia, anemia, and neutropenia were frequent, and the responses tended not to be durable (42, 43). We posit that specific inhibition of HDAC3 may be accompanied by reduced toxicity, as HDAC1 and HDAC2 have important roles in hematopoiesis (33), and avoiding the inhibition of these HDACs may therefore avoid the undesired hematologic effects associated with pan-HDAC inhibition. In line with this, we observed that BRD3308 was less toxic to CD4+ and CD8+ T cells than pan-HDAC inhibi-tors, and was able to elicit MHC-dependent T-cell responses against a CREBBP WT DLBCL cell line in vitro. Although these responses are likely driven by histocompatibility antigens rather than tumor neoantigens, they support the premise that selective HDAC3 inhibition is capable of promoting antigen presentation and immune responses. We also speculate that the clinically observed lack of durability for HDAC inhibitors may be the result of adaptive immune suppression through mechanisms such as PD-L1 induction, which dampens T-cell responses through the PD-1 receptor (44), as well as direct toxicity of pan-HDAC inhibitors to T cells. We found evidence for adaptive immune suppression within our model systems,
showing that HDAC3 inhibition leads to increased IFNγ pro-duction and the upregulation of PD-L1 expression. This is in line with recent observations that CD274 (encoding PD-L1) is a BCL6-supressed gene (45) and with a well-characterized role for PD-L1 as an IFNγ-responsive gene (44). In other cancers in which a florid antigen-driven immune response and concomi-tant adaptive immune suppression via PD-L1 exist within the tumor microenvironment, blockade of the PD-1 receptor is an effective therapeutic strategy (15, 17). Recent studies have also shown that the efficacy of PD-1 blockade is inextricably linked with the existence of an IFN-driven immune response and expression of MHC class II (15, 17). We have shown some evidence for this in a syngeneic, BCL6-driven murine model of B-cell lymphoma, wherein the combination of HDAC3 inhibi-tion and αPD-L1 led to enhanced levels of CD4+ and CD8+ T-cell infiltration and clearance of tumor cells. Together, these observations suggest that the greatest potentiation of antitumor immunity in GCB-derived malignancies may be achieved through stimulation of IFN signaling and MHC class II expression by HDAC3 inhibition, in combination with the blockade of adaptive immune suppression using PD-L1/PD-1 neutralizing antibodies. However, this concept requires further exploration in future studies.
In conclusion, this work defines a molecular circuit that controls the survival and differentiation of lymphoma B cells and their interaction with T cells. This circuit is antagonisti-cally regulated by CREBBP and BCL6, and can be pushed toward promoting tumor cell death and antitumor immunity via selective inhibition of HDAC3. This highlights HDAC3 inhibition as an attractive therapeutic avenue, which may be broadly active in follicular lymphoma and DLBCL due to the near-ubiquitous role of BCL6, but which may have enhanced potency in CREBBP-mutant tumors.
METHODSFor detailed methods, please refer to the Supplementary Methods.
CRISPR/Cas9 Modification of Lymphoma CellsThe RL cell line (ATCC CRL-2261) was modified by electroporation
with one of two unique gRNA sequences in the pSpCas9(BB)-2A-GFP vector (Addgene plasmid # 48138, gift from Feng Zhang, Massachu-setts Institute of Technology; ref. 46) with a single-stranded oligo-nucleotide donor template. Single GFP-positive cells were sorted 3 to 4 days after transfection, and colonies expanded and evaluated for changes in the targeted region using Sanger sequencing. The process was repeated until point mutants were retrieved from cells treated with each of the two gRNAs, totaling 742 single clones. Potential off-target sites of each gRNA were determined by BLAST, and all sites with ≥16/20 nucleotide match to either of the gRNA sequences were interrogated by Sanger sequencing (Supplementary Table S13). All cells were maintained as subconfluent culture in RPMI medium with 10% FBS and penicillin/streptomycin and revalidated by Sanger sequencing prior to each set of experiments. Basic phenotyping of the cell lines is presented in Supplementary Fig. S16A–S16E.
ChIP-seqCells were washed and fixed in formaldehyde and chromatin
sheared by sonication. An antibody specific to H3K27Ac (Active Motif) was coupled to magnetic protein G beads and incubated with chromatin overnight, and immunoprecipitation was performed.
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 457
Input controls were reserved for comparison. Nucleosomal DNA was isolated and used either as a template for qPCR (ChIP-qPCR) or to generate NGS libraries using KAPA Hyper Prep Kits (Roche) and TruSeq adaptors (Bioo Scientific) using 6 cycles of PCR enrichment. Libraries were 6-plexed and sequenced with 2 × 100bp reads on a HiSeq-4000 (Illumina). The data were mapped using BWA, peaks called using MACS2, and differential analyses performed using Diff-Bind. For GSEA, the gene with the closest transcription start site to the peak was used. The statistical thresholds for significance were q < 0.05 and fold change > 1.5.
RNA-seqTotal RNA was isolated using AllPrep DNA/RNA Kits (Qiagen)
and evaluated for quality on a Tapestation 4200 instrument (Agi-lent). Total RNA (1 µg) was used for library preparation with KAPA HyperPrep RNA Kits with RiboErase (Roche) and TruSeq adapters (Bioo Scientific). Libraries were validated on a Tapestation 4200 instrument (Agilent), quantified by Qubit dsDNA Kit (Life Tech-nologies), 6-plexed, and sequenced on a HiSeq4000 instrument at the MD Anderson Sequencing and Microarray Facility using 2 × 100 bp reads. Reads were aligned with STAR, and differential gene expres-sion analysis performed with DEseq2. The statistical thresholds for significance were q < 0.05 and fold change > 2.
Cell Proliferation AssaysCells were seeded in 96-well plates at 50,000 cells/100 µL/well with
either vehicle (DMSO 0.1%) or increasing concentrations of drugs. Cell viability was assessed with the fluorescent redox dye CellTiter-Blue (Promega). The reagent was added to the culture medium at 1:5 dilution, according to the manufacturer’s instructions. Procedures to determine the effects of certain conditions on cell proliferation and apoptosis were performed in three independent experiments. The two-tailed Student t test and Wilcoxon rank-sum test were used to estimate the statistical significance of differences among results from the three experiments. Significance was set at P < 0.05. The PRISM software was used for the statistical analyses.
PDX and In Vitro Organoid StudiesA CREBBP R1446C–mutant tumor (DFBL13727) was obtained
from the public repository of xenografts (47), expanded for 12 weeks in 1 mouse by surgical implantation into the renal capsule, and then implanted into the renal capsule of 12 additional mice for efficacy studies. Tumors were allowed to grow for 6 weeks, the size measured by MRI, and mice randomized to treatment and control groups to similar average tumor sizes. Mice were treated with BRD3308 (25 mg/kg) or vehicle control twice per day, 5 days on and 2 days off for a total of 3 weeks, and tumor size was assessed by MRI every 7 days. Two mice per group were euthanized on day 10 and the tumors har-vested for biomarker analysis. Mice were cared for in accordance with guidelines approved by the MD Anderson Institutional Animal Care and Use Committee.
For CREBBP WT tumors (NY-DR2, DANA, and TONY), 6-week-old female NSG mice were implanted subcutaneously and treat-ments started when tumors reached 100 mm3. Mice (12/group) were randomized and dosed via oral gavage with BRD3308 (25 mg/kg) or control vehicle (0.5% methylcellulose, 0.2% Tween 80) twice daily for 21 consecutive days. Mice were cared for in accordance with guidelines approved by the Memorial Sloan Kettering Cancer Center Institutional Animal Care and Use Committee and Research Animal Resource Center.
For in vitro organoid culture, the tumors were dissociated in single cells, stained with carboxyfluorescein diacetate succinimidyl ester (CFSE), washed, and mixed with irradiated 40LB cells at a 10:1 ratio of Primary:40LB. The cell mixture was then used to fabricate organoids in a 96-well plate as described previously (48), with 20 µL
organoids containing 3% silicate nanoparticles and a 5% gelatin in Iscove’s Modified Dulbecco’s Medium (IMDM) solution. The orga-noids were cultured in IMDM containing 20% FBS supplemented with antibiotics and normocin (Invivogen) for 6 days, doubling the volume of medium after 3 days. The cell mixture was exposed to four 1:3 serial dilutions of BRD3308 starting at 5 µmol/L or vehicle con-trol (DMSO) in triplicate for 6 days, treating a second time at 3 days. After 6 days of exposure, cell viability and proliferation were assessed by flow cytometry using DAPI staining gating on CFSE-positive cells.
Ex Vivo Killing AssayThe OCI-LY18 cells were subcutaneously implanted in NSG mice
and allowed to become established before the mice were injected with PBMC. After 2 weeks, the tumors were dissociated to single cells and CD3+ TILs were positively selected. T cells were expanded with a single administration of immunomagnetic microbeads coated with mouse anti-human CD3 and CD28 mAbs and rhIL2 and rhIL15 for 5 days. OCI-LY18 cells were treated with either DMSO or 10 µmol/L BRD3308 for 3 days, washed, and resuspended in fresh media with or without T cells at a ratio of 1:10. For MHC blocking, OCI-LY18 cells were treated with 10 µg/mL of either isotype Ig or blocking antibody against HLA-ABC W6/32, HLA-DR/DP/DQ, or the combination of the two. After 24-hour coculture, cell viability was analyzed using CellTiter-Blue.
Immunocompetent IlBcl6;Ezh2Y641F ModelSix-week-old female C56BL/6J mice were irradiated at day −1 and 0
with 250 rad, and after 2 hours transplanted with 1 × 106 splenocytes from IµBcl6;Ezh2Y641F mice (5) and 0.2 × 106 bone marrow cells from a healthy age-matched donor. Three weeks after transplantation, mice became sick and treatment was initiated with either vehicle or BRD3308 (25 mg/kg twice daily, every day for 21 days) via oral gav-age, with or without αPD-L1 delivered by intraperitoneal injection (250 µg twice weekly for 4 administration). Mice were monitored daily and euthanized when they became moribund. Spleen and liver were measured, weighed, and analyzed by flow cytometry for evaluating the disease. Immunofluorescence staining and imaging were performed at the Molecular Cytology Core Facility of Memo-rial Sloan Kettering Cancer Center (New York, NY). Sections were stained with either anti-CD4 (R&D Systems, catalog no. AF554, 2 µg/mL) or anti–PD-L1 (R&D Systems catalog no. AF1019, 2 µg/mL) or B220 (BD Bioscience, catalog no. 550286, 0.3 µg/mL) or anti-CD8 (eBiosciences catalog no. 14–0808, 2.5 µg/mL) for 5 hours, followed by 60-minute incubation with biotinylated horse anti-goat IgG (for CD4 and PD-L1; Vector Laboratories, catalog no. BA-9500) or bioti-nylated goat anti-rat IgG (for CD8 and B220; Vector Laboratories, catalog no. BA9400) at 1:200 dilution. The detection was performed with Streptavidin-HRP D (part of DAB Map kit, Ventana Medical Systems), followed by incubation with Tyramide Alexa 488 (Invit-rogen, catalog no. B40953), and counterstaining with DAPI (Sigma Aldrich, catalog no. D9542, 5 µg/mL). The slides were scanned with a Pannoramic Flash P250 Scanner (3DHistech) using a 20×/0.8NA objective lens. The fluorescence channels were imaged using DAPI, FITC, and TRITC filters sequentially with manually adjusted expo-sure times. Images were then exported into .tifs using Caseviewer (3DHistech) to be analyzed.
Data DepositionThe RNA-seq and ChIP-seq data are available at the Gene Expres-
sion Omnibus database under accession number GSE142357.
Disclosure of Potential Conflicts of InterestA.Y. Chang is a postdoctoral fellow at Pfizer Inc. (but was not dur-
ing the time of this manuscript study). S.S. Neelapu is a consultant at Kite/Gilead, Novartis, Incyte, Legend Biotech, Merck, Celgene,
Mondello et al.RESEARCH ARTICLE
458 | CANCER DISCOVERY�MARCH 2020 AACRJournals.org
Unum Therapeutics, Pfizer, Precision Biosciences, Cell Medica, Allo-gene, and Calibr and reports receiving commercial research grants from Kite/Gilead, Novartis, Merck, BMS, Cellectis, Poseida, Karus, Acerta, Allogene, and Unum Therapeutics. J. Baselga is an EVP, Oncol-ogy R& D, at AstraZeneca, is a member of the Board of Directors at Foghorn, Varian Medical Systems, and Bristol-Myers Squibb, is a consultant for Grail, Aura, Infinity, PMV Pharma, Juno, and Seragon, reports receiving a commercial research grant from Roche, and has ownership interest (including patents) in Apogen, Tango, and Venthera. A. Younes is a consultant at Xynomics, BMS, Merck, and Janssen. A. Dogan is a consultant at Roche, Seattle Genetics, EUSA Pharma, Takeda, and Corvus Pharmaceuticals, reports receiv-ing a commercial research grant from Roche, and is a consultant/advisory board member for Amgen. D.A. Scheinberg is on the SAB at Eureka, Oncopep, KLUS, Sellas, Actinium, and Pfizer, reports receiv-ing a commercial research grant from BMS, and has ownership inter-est (including patents) in IOVA. A.M. Melnick is a consultant for Epizyme and Constellation, reports receiving commercial research grants from Janssen and Daiichi Sankyo, and is a consultant/ advisory board member for KDAC. M.R. Green is a consultant at Verastem Oncology, has ownership interest (including patents) in KDAC Therapeutics, and is a consultant/advisory board member for KDAC Therapeutics. No potential conflicts of interest were dis-closed by the other authors.
Authors’ ContributionsConception and design: P. Mondello, S. Tadros, E. de Stanchina, J. Baselga, A.M. Melnick, M.R. GreenDevelopment of methodology: P. Mondello, S. Tadros, H.-Y. Ying, E. Toska, A.M. Melnick, M.R. GreenAcquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): P. Mondello, S. Tadros, L. Fontan, A.Y. Chang, N. Jain, H. Yang, S. Singh, H.-Y. Ying, C.-S. Chu, S. Alig, M. Durant, E. de Stanchina, S. Ghosh, A. Mottok, S.S. Neelapu, O. Weigert, G. Inghirami, A. Younes, A. Dogan, R.G. Roeder, A.M. MelnickAnalysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P. Mondello, S. Tadros, M. Teater, L. Fontan, A.Y. Chang, H. Yang, H.-Y. Ying, C.-S. Chu, M.C.J. Ma, S. Alig, S. Ghosh, A. Mottok, L. Nastoupil, S.S. Neelapu, O. Weigert, J. Baselga, A. Dogan, D.A. Scheinberg, R.G. Roeder, A.M. Melnick, M.R. GreenWriting, review, and/or revision of the manuscript: P. Mondello, S. Tadros, M. Teater, H.-Y. Ying, A. Mottok, L. Nastoupil, S.S. Nee-lapu, G. Inghirami, J. Baselga, A. Younes, C. Yee, D.A. Scheinberg, A.M. Melnick, M.R. GreenAdministrative, technical, or material support (i.e., reporting or organizing data, constructing databases): P. Mondello, M.C.J. Ma, L. Nastoupil, S.S. Neelapu, A.M. Melnick, M.R. GreenStudy supervision: A. Younes, D.A. Scheinberg, A.M. Melnick, M.R. Green
AcknowledgmentsThis work was supported by R01 CA201380 (to M.R. Green), R01
CA055349 (to D.A. Scheinberg), R01 CA178765 (to R.G. Roeder), U54 OD020355 01 (to E. de Stanchina and G. Inghirami), NCI SPORE P50 CA192937 (to A. Dogan and A. Younes), the MD Ander-son Cancer Center (P30 CA016672) and Memorial Sloan Kettering Cancer Center (P30 CA008748) NCI CORE Grants, the Chemo-therapy Foundation (to A.M. Melnick), the Star Cancer Consortium (to A.M. Melnick), the Jaime Erin Follicular Lymphoma Research Consortium (to A.M. Melnick, M.R. Green, and S.S. Neelapu), the Schweitzer Family Fund (to M.R. Green), and the Futcher Founda-tion (to L. Nastoupil and M.R. Green). H. Yang is a Fellow of the Leukemia and Lymphoma Society.
Received January 29, 2019; revised November 11, 2019; accepted January 3, 2020; published first January 8, 2020.
REFERENCES 1. Hatzi K, Melnick A. Breaking bad in the germinal center: how deregu-
lation of BCL6 contributes to lymphomagenesis. Trends Mol Med 2014;20:343–52.
2. Mesin L, Ersching J, Victora GD. Germinal center B cell dynamics. Immunity 2016;45:471–82.
3. Hatzi K, Jiang Y, Huang C, Garrett-Bakelman F, Gearhart MD, Giannopoulou EG, et al. A hybrid mechanism of action for BCL6 in B cells defined by formation of functionally distinct complexes at enhancers and promoters. Cell Rep 2013;4:578–88.
4. Hatzi K, Geng H, Doane AS, Meydan C, LaRiviere R, Cardenas M, et al. Histone demethylase LSD1 is required for germinal center for-mation and BCL6-driven lymphomagenesis. Nat Immunol 2019;20: 86–96.
5. Beguelin W, Teater M, Gearhart MD, Calvo Ferná ndez MT, Goldstein RL, Cá rdenas MG, et al. EZH2 and BCL6 cooperate to assemble CBX8-BCOR complex to repress bivalent promoters, mediate germinal center formation and lymphomagenesis. Cancer Cell 2016;30:197–213.
6. Green MR. Chromatin modifying gene mutations in follicular lym-phoma. Blood 2018;131:595–604.
7. Green MR, Kihira S, Liu CL, Nair RV, Salari R, Gentles AJ, et al. Muta-tions in early follicular lymphoma progenitors are associated with suppressed antigen presentation. Proc Natl Acad Sci U S A 2015;112: E1116–25.
8. Jiang Y. Ortega-Molina A, Geng H, Ying HY, Hatzi K, Parsa S, et al. CREBBP inactivation promotes the development of HDAC3-depend-ent lymphomas. Cancer Discov 2017;7:38–53.
9. Allen CD, Okada T, Cyster JG. Germinal-center organization and cel-lular dynamics. Immunity 2007;27:190–202.
10. Khodadoust MS, Olsson N, Wagar LE, Haabeth OA, Chen B, Swami-nathan K, et al. Antigen presentation profiling reveals recognition of lymphoma immunoglobulin neoantigens. Nature 2017;543:723–7.
11. Khodadoust MS, Olsson N, Chen B, Sworder B, Shree T, Liu CL, et al. B cell lymphomas present immunoglobulin neoantigens. Blood 2019;133:878–81.
12. Marty R, Thompson WK, Salem RM, Zanetti M, Carter H. Evolution-ary pressure against MHC class II binding cancer mutations. Cell 2018;175:416–28.
13. Rimsza LM, Roberts RA, Miller TP, Unger JM, LeBlanc M, Braziel RM, et al. Loss of MHC class II gene and protein expression in diffuse large B-cell lymphoma is related to decreased tumor immunosurveillance and poor patient survival regardless of other prognostic factors: a follow-up study from the Leukemia and Lymphoma Molecular Profil-ing Project. Blood 2004;103:4251–8.
14. Alspach E, Lussier DM, Miceli AP, Kizhvatov I, DuPage M, Luoma AM, et al. MHC-II neoantigens shape tumour immunity and response to immunotherapy. Nature 2019;574:696–701.
15. Rodig SJ, Gusenleitner D, Jackson DG, Gjini E, Giobbie-Hurder A, Jin C, et al. MHC proteins confer differential sensitivity to CTLA-4 and PD-1 blockade in untreated metastatic melanoma. Sci Transl Med 2018;10:eaar3342.
16. Ayers M, Lunceford J, Nebozhyn M, Murphy E, Loboda A, Kauf-man DR, et al. IFN-gamma-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest 2017;127:2930–40.
17. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 2014;515:568–71.
18. Borst J, Ahrends T, Babala N, Melief CJM, Kastenmuller W. CD4(+) T cell help in cancer immunology and immunotherapy. Nat Rev Immunol 2018;18:635–47.
19. Garcia-Ramirez I, Tadros S, Gonzá lez-Herrero I, Martín-Lorenzo A, Rodríguez-Herná ndez G, Moore D, et al. Crebbp loss cooperates with Bcl2 over-expression to promote lymphoma in mice. Blood 2017; 129:2645–56.
Targeting HDAC3 in Lymphoma RESEARCH ARTICLE
MARCH 2020�CANCER DISCOVERY | 459
20. Mullighan CG, Zhang J, Kasper LH, Lerach S, Payne-Turner D, Phillips LA, et al. CREBBP mutations in relapsed acute lymphoblas-tic leukaemia. Nature 2011;471:235–9.
21. Pasqualucci L, Dominguez-Sola D, Chiarenza A, Fabbri G, Grunn A, Trifonov V, et al. Inactivating mutations of acetyltransferase genes in B-cell lymphoma. Nature 2011;471:189–95.
22. Hashwah H, Schmid CA, Kasser S, Bertram K, Stelling A, Manz MG, et al. Inactivation of CREBBP expands the germinal center B cell com-partment, down-regulates MHCII expression and promotes DLBCL growth. Proc Natl Acad Sci U S A 2017;114:9701–6.
23. Horton SJ, Giotopoulos G, Yun H, Vohra S, Sheppard O, Bashford-Rogers R, et al. Early loss of Crebbp confers malignant stem cell prop-erties on lymphoid progenitors. Nat Cell Biol 2017;19:11093–104.
24. Zhang J, Vlasevska S, Wells VA, Nataraj S, Holmes AB, Duval R, et al. The CREBBP acetyltransferase is a haploinsufficient tumor suppres-sor in B-cell lymphoma. Cancer Discov 2017;7:322–37.
25. Meyer SN, Scuoppo C, Vlasevska S, Bal E, Holmes AB, Holloman M, et al. Unique and shared epigenetic programs of the CREBBP and EP300 acetyltransferases in germinal center B cells reveal targetable dependencies in lymphoma. Immunity 2019;51:535–47.
26. Pastore A, Jurinovic V, Kridel R, Hoster E, Staiger AM, Szczepanow-ski M, et al. Integration of gene mutations in risk prognostication for patients receiving first-line immunochemotherapy for follicular lymphoma: a retrospective analysis of a prospective clinical trial and validation in a population-based registry. Lancet Oncol 2015;16: 1111–22.
27. Li J, Wang J, Wang J, Nawaz Z, Liu JM, Qin J, et al. Both corepressor proteins SMRT and N-CoR exist in large protein complexes contain-ing HDAC3. EMBO J 2000;19:4342–50.
28. Wagner FF, Lundh M, Kaya T, McCarren P, Zhang YL, Chattopadhyay S, et al. An isochemogenic set of inhibitors to define the therapeutic potential of histone deacetylases in beta-cell protection. ACS Chem Biol 2016;11:363–74.
29. Phan RT, Saito M, Basso K, Niu H, Dalla-Favera R. BCL6 interacts with the transcription factor Miz-1 to suppress the cyclin-dependent kinase inhibitor p21 and cell cycle arrest in germinal center B cells. Nat Immunol 2005;6:1054–60.
30. Tian YF, Ahn H, Schneider RS, Yang SN, Roman-Gonzalez L, Melnick AM, et al. Integrin-specific hydrogels as adaptable tumor organoids for malignant B and T cells. Biomaterials 2015;73:110–9.
31. Fujita N, Jaye DL, Geigerman C, Akyildiz A, Mooney MR, Boss JM, et al. MTA3 and the Mi-2/NuRD complex regulate cell fate during B lymphocyte differentiation. Cell 2004;119:75–86.
32. Huang C, Gonzalez DG, Cote CM, Jiang Y, Hatzi K, Teater M, et al. The BCL6 RD2 domain governs commitment of activated B cells to form germinal centers. Cell Rep 2014;8:1497–508.
33. Wilting RH, Yanover E, Heideman MR, Jacobs H, Horner J, van der Torre J, et al. Overlapping functions of Hdac1 and Hdac2 in cell cycle regulation and haematopoiesis. EMBO J 2010;29:2586–97.
34. Green MR, Monti S, Rodig SJ, Juszczynski P, Currie T, O’Donnell E, et al. Integrative analysis reveals selective 9p24.1 amplification, increased PD-1 ligand expression, and further induction via JAK2 in nodular sclerosing Hodgkin lymphoma and primary mediastinal large B-cell lymphoma. Blood 2010;116:3268–77.
35. Ansell SM, Lesokhin AM, Borrello I, Halwani A, Scott EC, Gutierrez M, et al. PD-1 blockade with nivolumab in relapsed or refractory Hodgkin’s lymphoma. N Engl J Med 2015;372:311–9.
36. Xu W, Fukuyama T, Ney PA, Wang D, Rehg J, Boyd K, et al. Global transcriptional coactivators CREB-binding protein and p300 are highly essential collectively but not individually in peripheral B cells. Blood 2006;107:4407–16.
37. Beguelin W, Rivas MA, Calvo Ferná ndez MT, Teater M, Purwada A, Redmond D, et al. EZH2 enables germinal centre formation through epigenetic silencing of CDKN1A and an Rb-E2F1 feedback loop. Nat Commun 2017;8:877.
38. Valls E, Lobry C, Geng H, Wang L, Cardenas M, Rivas M, et al. BCL6 antagonizes NOTCH2 to maintain survival of human follicular lym-phoma cells. Cancer Discov 2017;7:506–21.
39. Cardenas MG, Yu W, Beguelin W, Teater MR, Geng H, Goldstein RL, et al. Rationally designed BCL6 inhibitors target activated B cell dif-fuse large B cell lymphoma. J Clin Invest 2016;126:3351–62.
40. Shankaran V, Ikeda H, Bruce AT, White JM, Swanson PE, Old LJ, et al. IFNgamma and lymphocytes prevent primary tumour development and shape tumour immunogenicity. Nature 2001;410:1107–11.
41. Minn AJ, Wherry EJ. Combination cancer therapies with immune check-point blockade: convergence on interferon signaling. Cell 2016;165:272–5.
42. Ogura M, Ando K, Suzuki T, Ishizawa K, Oh SY, Itoh K, et al. A multicentre phase II study of vorinostat in patients with relapsed or refractory indolent B-cell non-Hodgkin lymphoma and mantle cell lymphoma. Br J Haematol 2014;165:768–76.
43. Crump M, Coiffier B, Jacobsen ED, Sun L, Ricker JL, Xie H, et al. Phase II trial of oral vorinostat (suberoylanilide hydroxamic acid) in relapsed diffuse large-B-cell lymphoma. Ann Oncol 2008;19:964–9.
44. Zaidi MR, Merlino G. The two faces of interferon-gamma in cancer. Clin Cancer Res 2011;17:6118–24.
45. Peng C, Hu Q, Yang F, Zhang H, Li F, Huang C, et al. BCL6-mediated silencing of PD-1 ligands in germinal center B cells maintains follicu-lar T cell population. J Immunol 2019;202:704–13.
46. Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F, et al. Genome engineering using the CRISPR-Cas9 system. Nat Protoc 2013;8:2281–308.
47. Townsend EC, Murakami MA, Christodoulou A, Christie AL, Kö ster J, DeSouza TA, et al. The public repository of xenografts enables discov-ery and randomized phase II-like trials in mice. Cancer Cell 2016;29:574–86.
48. Purwada A, Jaiswal MK, Ahn H, Nojima T, Kitamura D, Gaharwar AK, et al. Ex vivo engineered immune organoids for controlled germinal center reactions. Biomaterials 2015;63:24–34.
CLINICAL CANCER RESEARCH | CLINICAL TRIALS: TARGETED THERAPY
Phase I Study of TAK-659, an Investigational, DualSYK/FLT3 Inhibitor, in Patientswith B-Cell Lymphoma A C
Leo I. Gordon1, Jason B. Kaplan1, Rakesh Popat2, Howard A. Burris III3, Silvia Ferrari4, Sumit Madan5,Manish R. Patel6, Giuseppe Gritti4, Dima El-Sharkawi2, Ian Chau7, John A. Radford8,Jaime P�erez de Oteyza9, Pier Luigi Zinzani10, Swaminathan Iyer11, William Townsend2, Reem Karmali1,Harry Miao12, Igor Proscurshim12, Shining Wang12, Yujun Wu12, Kate Stumpo12, Yaping Shou12,Cecilia Carpio13, and Francesc Bosch13
ABSTRACT◥
Purpose: TAK-659 is an investigational, dual SYK/FLT3 inhib-itor with preclinical activity in B-cell malignancy models. This first-in-human, dose-escalation/expansion study aimed to determine thesafety, tolerability, MTD/recommended phase II dose (RP2D), andpreliminary efficacy of TAK-659 in relapsed/refractory solid tumorsand B-cell lymphomas.
Patients and Methods: Patients received continuous, once-dailyoral TAK-659, 60–120 mg in 28-day cycles, until disease progres-sion or unacceptable toxicity. The study applied an accelerateddose-escalation design to determine the MTD and RP2D. In theexpansion phase, patients with lymphoma were enrolled in fivedisease cohorts at the MTD.
Results: Overall, 105 patients were enrolled [dose escalation,n ¼ 36 (solid tumors, n ¼ 19; lymphoma, n ¼ 17); expansion,n¼ 69]. TheMTDwas 100mg once daily. TAK-659 absorptionwas
fast (Tmax �2 hours) with a long terminal half-life (�37 hours).Exposure generally increased with dose (60–120 mg), with mod-erate variability. The most common treatment-related adverseevents were generally asymptomatic and reversible elevations inclinical laboratory values. Among 43 response-evaluable patientswith diffuse large B-cell lymphoma, 8 (19%) achieved a completeresponse (CR) with an overall response rate (ORR) of 28% [23%intent-to-treat (ITT)]. Responses were seen in both de novo andtransformed disease and appeared independent of cell-of-originclassification. Among 9 response-evaluable patients with follicularlymphoma, 2 (22%) achieved CR with an ORR of 89% (57% ITT).
Conclusions: TAK-659 has single-agent activity in patients withB-cell lymphoma. Further studies of the drug in combination,including an evaluation of the biologically optimal and safestlong-term dose and schedule, are warranted.
IntroductionSpleen tyrosine kinase (SYK) is a nonreceptor cytoplasmic kinase
expressed primarily in hematopoietic cells. It is an essential componentof the signaling machinery involved in B-cell receptor (BCR)�medi-ated signaling (1–3).Once activated, SYKpropagates theBCR signal byassociating with adaptor proteins and phosphorylating signalingintermediates, such as B-cell linker protein, Bruton tyrosine kinase(BTK), and phospholipase Cg2 (PLC-g2), leading to cell proliferation,differentiation, and survival (4, 5). SYK also appears to play a role innonimmune functions, including cellular adhesion, bone metabolism,and platelet function (6).
Aberrant SYK-mediated signaling from the BCR has been impli-cated in the pathogenesis of several B-cell malignancies (7, 8), includ-ing upregulated SYK mRNA and protein levels (4), and constitutiveSYK activation (9). Consequently, SYK is an attractive target in thetreatment of B-cell malignancies mediated by BCR signaling (8–10).Recent data also suggest that SYK is a component of the cellularsignaling cascade associated with Epstein–Barr virus (EBV) latencyand transformation, and is involved with cell–cell and cell–matrixinteractions (11), which may be relevant in EBV-positive posttrans-plant lymphoproliferative disorder (PTLD). In addition, evidencesuggests that the SYK pathway may also be implicated in select solidtumors (12–15).
TAK-659 is an investigational, oral, reversible, and potent dualinhibitor of SYK and FMS-like tyrosine kinase 3 (FLT3; ref. 16). TAK-659 has demonstrated inhibitory activity in preclinical models ofdiffuse large B-cell lymphoma (DLBCL), in a RL follicular lymphoma(FL) cell line (17), and in chronic lymphocytic leukemia (CLL)cells (18). Preclinical studies inmultiple syngeneic or xenograftmodelshave also shown that TAK-659 administration can result in reductions
1Northwestern University Feinberg School of Medicine and the Robert H. LurieComprehensive Cancer Center, Chicago, Illinois. 2NIHR/UCLH Clinical ResearchFacility, University College London Hospitals NHS Foundation Trust, London,United Kingdom. 3Sarah Cannon Research Institute/Tennessee Oncology, Nash-ville, Tennessee. 4Ospedale Papa Giovanni XXIII, Bergamo, Italy. 5Cancer Ther-apy and Research Center at University of Texas Health Science Center, SanAntonio, Texas. 6Florida Cancer Specialists/Sarah Cannon Research Institute,Sarasota, Florida. 7Royal Marsden Hospital, Sutton, Surrey, United Kingdom.8The University of Manchester and the Christie NHS Foundation Trust, Manche-ster Academic Health Science Centre, Manchester, United Kingdom. 9HospitalUniversitario Madrid Sanchinarro, Universidad CEU San Pablo, Madrid, Spain.10Institute of Hematology “Seragnoli,” University of Bologna, Bologna, Italy.11Houston Methodist Cancer Center, Houston, Texas. 12Millennium Pharmaceu-ticals, Inc., Cambridge, Massachusetts, a wholly owned subsidiary of TakedaPharmaceutical Company Limited. 13University Hospital Vall d’Hebron, Barce-lona, Spain.
Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).
ClinicalTrials.gov identifier: NCT02000934.
Current address for J.B. Kaplan: NorthShore University Hospitals, Evanston,Illinois; and current address for D. El-Sharkawi, Royal Marsden Hospital, Sutton,Surrey, United Kingdom.
CorrespondingAuthor: Leo I. Gordon, NorthwesternUniversity Feinberg Schoolof Medicine, and the Robert H. Lurie Comprehensive Cancer Center, 676 N. St.Clair Street, Suite 850, Chicago, IL 60611. Phone: 312-695-4546; Fax: 312-695-6189; E-mail: [email protected]
Clin Cancer Res 2020;26:3546–56
doi: 10.1158/1078-0432.CCR-19-3239
�2020 American Association for Cancer Research.
AACRJournals.org | 3546
in immunosuppressive cell populations, such as myeloid-derivedsuppressor cells (MDSC) and regulatory T cells, which are mediatedby both SYK and FLT3 signaling (19), suggesting an immune-modulating effect (20, 21). It is not clear whether antagonizing FLT3signaling in parallel will contribute further to the clinical activity ofTAK-659 in B-cell malignancies. To date, FLT3 inhibitors have noreported activity in B-cell lymphomas.
On the basis of these preclinical data, we conducted a first-in-human study of TAK-659 to determine the safety, tolerability, andMTD/recommended phase II dose (RP2D) of TAK-659 in patientswith solid tumors and B-cell lymphomas. Four dose levels of TAK-659were explored in patients with relapsed/refractory solid tumors orlymphomas (60, 80, 100, and 120 mg). Additional patients withlymphoma were treated with TAK-659 at the MTD in expansioncohorts. This manuscript primarily focuses on the results for patientswith lymphoma.
Patients and MethodsStudy design
This was a phase I, multicenter, open-label, dose-escalation, andexpansion study of single-agent TAK-659. The primary objective wasto determine the MTD/RP2D of TAK-659 administered orally oncedaily. Secondary objectives were to characterize the pharmacokineticsof TAK-659 and to assess the preliminary antitumor activity of TAK-659 in patients with relapsed/refractory B-cell lymphoma treated at theMTD/RP2D.
Patients received oral TAK-659 (tablets) continuously once daily in28-day cycles. Patients were treated until disease progression orunacceptable toxicity.
The safe starting dose of TAK-659 based on nonclinical GoodLaboratory Practice-compliant toxicology data for this first-in-humanstudy was estimated to be 80 mg once daily (see SupplementaryInformation). However, 60 mg once daily was selected as the startingdose based on available tablet strengths at study start.
The phase I portion of the study applied an accelerated dose-escalation design to determine the MTD/RP2D. One patient wasplanned to be enrolled at the 60 mg starting dose and the twosubsequent dose levels of 120 and 200 mg. In the initial single-patient cohorts, if either a dose-limiting toxicity (DLT; defined inSupplementary Table S1) or a related grade ≥2 adverse event (AE)
occurred in cycle 1, the current and subsequent dose levels were to betested in ≥3 patient cohorts. Dose escalation was then to follow astandard 3þ3 design.
In expansion, patients with lymphoma were enrolled to five diseasecohorts to receive TAK-659 at theMTD/RP2D, including CLL (n¼ 12planned), DLBCL (n ¼ 12–25 planned based on Simon’s two-stagedesign; ≥3 responses in 12 patients required to proceed to the secondstage), indolent non-Hodgkin lymphoma (iNHL; n ¼ 10–23 plannedbased on Simon’s two-stage design; ≥5 responses in 10 patientsrequired to proceed to the second stage), mantle cell lymphoma (MCL;n ¼ 16 planned), and PTLD (n ¼ 16 planned). Prior to cycle 1,approximately 12 patients with iNHL were planned to complete asingle-dose, 7-day, pharmacokinetic run-in to obtain at least 8 half-life�evaluable patients.
PatientsThe dose-escalation phase enrolled adults with confirmed meta-
static and/or advanced solid tumors or lymphoma. Patients hadmeasurable or evaluable disease. The expansion phase enrolledpatients with CLL, DLBCL, iNHL, MCL, or PTLD who had received≥1 prior line(s) of therapy. Measurable disease was determined by theRECIST version 1.1 for solid tumors, or by the modified InternationalWorking Group criteria for malignant lymphoma (IWG 2007), orInternational Workshop on CLL criteria (22–24; SupplementaryInformation). Additional eligibility criteria included an Eastern Coop-erative Oncology Group (ECOG) performance status of 0–1, adequateorgan function, and recovery from prior therapy. Patients with lym-phoma in the expansion cohorts could have been previously treatedwith BCR pathway inhibitors not directly targeting SYK. Details oninclusion/exclusion criteria are reported in the SupplementaryInformation.
The study was conducted in compliance with the Declaration ofHelsinki, International Conference on Harmonization Good ClinicalPractice standards, and applicable regulatory requirements. Relevantinstitutional review boards or ethics committees approved all aspectsof the study, and all authors had access to primary clinical trial data. Allpatients provided written informed consent. The trial is registered atClinicalTrials.gov (NCT02000934).
AssessmentsAEs were assessed throughout the study and graded according to
the NCI Common Terminology Criteria for Adverse Events version4.03. Plasma and urine samples were collected (SupplementaryInformation) for TAK-659 concentration measurements by high-performance liquid chromatography with tandem mass spectrom-etry detection. During dose escalation, responses were assessed atcycles 2, 4, 6, and every three cycles thereafter in patients withlymphoma. In expansion, responses were assessed at every even-numbered cycle through cycle 12, every four cycles through cycle24, and every 6 months thereafter.
DLBCL cell of origin was determined by IHC when available (locallaboratory) and was classified as germinal center B-cell (GCB) or non-GCB (Supplementary Information; ref. 25).
Statistical analysisStatistical analyses were primarily descriptive without formal
hypothesis testing. Median time to response, duration of response,and progression-free survival (PFS) were estimated using Kaplan–Meier methodology. Plasma and urine TAK-659 pharmacokineticparameters in dose-escalation patients were derived using noncom-partmental analysis with Phoenix WinNonlin 7.0 (Certara).
Translational Relevance
In this phase I study, the MTD of TAK-659 was 100 mg oncedaily. The safety and activity of TAK-659 at theMTDwas evaluatedin an additional 69 patients with lymphoma (expansion phase).Common treatment-emergent adverse events were generallyasymptomatic and reversible elevations in clinical laboratoryvalues. TAK-659 showed single-agent activity across differenthistologic subtypes. Responses occurred in both de novo andtransformed disease and appeared independent of cell-of-originclassification.
These data suggest that targeting spleen tyrosine kinase (the roleof FLT3 is unclear) with TAK-659 has clinical activity in B-cellmalignancies. Further studies are needed to elucidate the exactmechanism of this effect and to further evaluate TAK-659 mono-therapy or combination therapy in patients with relapsed/refrac-tory B-cell malignancies.
Phase I Study of TAK-659 in Lymphoma
AACRJournals.org Clin Cancer Res; 26(14) July 15, 2020 3547
Dose escalation was conducted according to a modified doseescalation rule, with 1 patient only for the first three cohorts, then3–6 patients evaluated at each subsequent dose level, up to six plannedascending dose cohorts. Approximately 21–30 evaluable patients wereplanned to be enrolled according to the modified 3þ3 dose-escalationscheme. The MTD/RP2D cohort was planned to have at least 6patients, including a minimum of 3 patients with lymphoma or CLL.
The estimated sample size of 12 for the CLL expansion cohort wasbased on the following considerations: 12 patients would provide 61%power to detect a statistically significant difference between theuninteresting (20%) and interesting (50%) overall response rates(ORR) based on the exact one-sample binomial test.
A Simon two-stage design was built into the largest expansioncohorts, DLBCL and iNHL, for futility stopping because these twoindications were considered the priority for development based on thepreclinical experience. The estimated sample sizes for the DLBCL andiNHL expansion cohorts were based on a Simon two-stage designusing the following parameters: a one-sided test at the significance levelof a ¼ 0.1, a power of 80%, a null hypothesis of an ORR of ≤20% forDLBCL and ≤35% for iNHL, and an alternative hypothesis of an ORRof≥40% forDLBCL and≥60% for iNHL. A total of 12–25 patients withDLBCL (14–29 patients projected on the basis of a 15% dropout rate)and 10–23 patients with iNHL (14–27 patients projected on the basis ofa 15% dropout rate) would be needed for the expansion cohorts.
The estimated sample size of 16 for the MCL and EBV-positivePTLD expansion cohorts was based on the following considerations:16 patients in each cohort would provide 77% power to detect astatistically significant difference between the uninteresting (20%) andinteresting (50%) ORRs based on the exact one-sample binomial test.
Data Sharing StatementTakeda makes patient-level, deidentified datasets and associated
documents available after applicable marketing approvals andcommercial availability have been received, an opportunity for theprimary publication of the research has been allowed, and othercriteria have been met as set forth in Takeda's Data Sharing Policy
(see https://www.takedaclinicaltrials.com/ for details). To obtainaccess, researchers must submit a legitimate academic researchproposal for adjudication by an independent review panel, whowill review the scientific merit of the research and the requestor'squalifications and conflict of interest that can result in potentialbias. Once approved, qualified researchers who sign a data sharingagreement are provided access to these data in a secure researchenvironment.
ResultsPatients and treatment
A total of 105 patients (36 dose escalation, 69 expansion) wereenrolled by the data cutoff of April 9, 2018. In dose escalation,36 patients (solid tumors, n ¼ 19; lymphoma, n ¼ 17) receivedTAK-659 at the following doses: 60 mg (n ¼ 10), 80 mg (n ¼ 4),100 mg (n ¼ 15), and 120 mg (n ¼ 7). In both dose escalation (n ¼17) and expansion (n ¼ 69), a total of 86 relapsed/refractorypatients with lymphoma [DLBCL, n ¼ 53; iNHL, n ¼ 21 (FL, n¼ 14); CLL, n ¼ 6; MCL, n ¼ 5; PTLD, n ¼ 1] received TAK-65960–120 mg once daily. A total of 84 patients (80%; 46 with DLBCL)received TAK-659 at the MTD/RP2D of 100 mg once daily. Inthis manuscript, when DLBCL and iNHL/FL are discussed, we arereferring to patients who were enrolled into both dose-escalationand dose-expansion cohorts.
Baseline characteristics and demographics of enrolled patients areshown in Table 1 and disease subtypes are shown in SupplementaryTable S2. Patients with DLBCL had a median time from initialdiagnosis of 15 months, 57% of patients with DLBCL were Ann Arborstage III–IV at study entry, 60% had evidence of extranodal involve-ment, and 19% received prior autologous transplant (prior allogeneictransplantation was excluded). Patients with FL had a median timefrom initial diagnosis of 53 months, 78% of patients with FL were AnnArbor stage III–IV at initial diagnosis, 86% had three or more priorlines of therapy, and 50% received a prior autologous transplant (priorallogeneic transplantation was excluded).
Table 1. Baseline demographics and disease characteristics by tumor type.
LymphomaDLBCL FL All lymphomasaSolid tumorsa
n ¼ 19 n ¼ 53 n ¼ 14 n ¼ 86All patientsN ¼ 105
Age, median (range) years 59 (38–80) 60 (23–84) 57 (33–82) 65 (23–85) 63 (23–85)Gender, male, n (%) 9 (47) 35 (66) 8 (57) 54 (63) 63 (60)ECOG performance status, 0, n (%) 8 (42) 17 (32) 7 (50) 29 (34) 37 (35)Disease characteristics
TNM or Ann Arbor stage III–IV, n (%) 15 (79) 30 (57) 8 (57) 49 (57) 64 (61)Months since diagnosis, median (range) 27 (2–144) 15 (0–257) 53 (15–99) 21 (0–257) 23 (0–257)Nodal sites ≥ 5, n (%) - 18 (34) 7 (50) 30 (35) 30 (29)Bulky disease at entry, n (%) - 4 (8) 1 (7) 7 (8) 7 (7)Bone marrow involvement at entry, n (%) - 7 (13) 5 (36) 24 (28) 24 (23)Genetic classification (DLBCL only)
Double/triple hit - 7 (13) - - -GCB/non-GCBb - 29 (55)/8 (15) - - -
Prior treatmentLines of prior therapy, median (range) 3 (1–11) 3 (1–9) 3 (2–9) 3 (1–9) 3 (1–11)Prior autologous transplant, n (%) - 10 (19) 7 (50) 18 (21) 18 (17)
Abbreviations: DLBCL, diffuse large B-cell lymphoma; ECOG, Eastern CooperativeOncologyGroup; FL, follicular lymphoma; GCB, germinal center B-cell; TNM, tumornode metastasis.aSee Supplementary Table S2 for details of specific diseases.bUnknown in 16 patients.
Gordon et al.
Clin Cancer Res; 26(14) July 15, 2020 CLINICAL CANCER RESEARCH3548
At data cutoff, 79 of 86 patients with lymphoma (92%) haddiscontinued treatment; mostly due to disease progression (n ¼ 35;Supplementary Table S3). Twenty-eight patients (33%) discontinuedtreatment due to a treatment-emergent AE (TEAE). Of patients withlymphoma receiving TAK-659 100 mg, 23 of 78 patients (29%)proceeded through cycle 1 without dose modification (all causesincluded, regardless of whether they were disease related or treatmentrelated) and 19 of 54 (35%) proceeded through cycle 2 without dosemodification. There were 25 on-study deaths (defined as death thatoccurs between the first dose and 28 days after the last dose of studydrug) in patients with lymphoma (29%), of which 4were considered byinvestigators to be related or possibly related to study drug [respiratoryfailure, multiorgan failure, disseminated varicella, and pneumocystisjirovecii pneumonia (PJP) infection, all in patients receiving TAK-659100 mg once daily]; all 4 patients (DLBCL, n ¼ 2; B-cell lympho-plasmacytic lymphoma/immunocytoma, n ¼ 1; mucosa-associatedlymphoid tissue, n ¼ 1) had confounders and alternative etiologies,and 1 had concurrent progressive disease. The four related deathsoccurred within 3.6 months of the initial dose of TAK-659, with twooccurring within the first 2 months. Two of the 3 patients withoutconcurrent disease progression initially presented with infection andwere later positive for viral reactivation; 1 had PJP and cytomegalo-virus (CMV) reactivation, and 1 had nonspecific pneumonitis followedby varicella. These infectious processes, among other factors, led todeath. Prophylaxismeasures (notmandatory for every patient) againstPJP and viral infection were later introduced to the protocol. Thethird patient without concurrent disease progression who came ontothe study with an existing dry cough, did not benefit from therecommended prophylactic measures because they were not startedon antibiotics until after documentation of the PJP.
Dose-escalation phaseIn dose escalation, the safety of four TAK-659 dose levels was
assessed sequentially in 36 patients (19 solid tumor, 17 lymphoma;Supplementary Table S4). The observed safety in the single-patientcohorts at the initial three dose levels allowed for escalation based onstandard 3þ3 escalation criterion. The tested dose levels included theinitial starting dose of 60 mg (n¼ 10), 80mg (n¼ 4), 100mg (n¼ 15),and 120 mg (n ¼ 7). One DLT [increased aspartate aminotransferase(AST), grade 3] occurred in 1 of 10 patients at the 60-mgdose level, andnoDLTs occurred at either the 80-mg dose level or in the initial cohortenrolled at the 100-mg dose level. At the 120-mg dose level, DLTsoccurred in 4 of 7 patients, including stomatitis (grade 3), generalizededema (grade 3), and increased lipase in 2 patients (grades 3 and 4).The 120-mg dose was thus confirmed to have exceeded the MTD,and the MTD was determined to be 100 mg. Expanded testing ofthe 100-mg dose was undertaken with 12 additional patients wherea single DLT of grade 3 hypophosphatemia was identified; thus the100-mg dose was determined to be the MTD per the standard 3þ3dose-escalation criteria (see Supplementary Table S4). Among the15 patients who were treated at the 100-mg dose, 10 (67%) received ≥2cycles, 4 (27%) received ≥4 cycles, 4 (27%) received ≥6 cycles, 2 (13%)received ≥12 cycles, and 1 (7%) received ≥24 cycles.
Herein, we report the results for the 86 patients with lymphomawhoreceived TAK-659 during dose escalation or expansion.
Treatment exposure and safety in patients with lymphomaAll 86 patientswith lymphoma received≥1dose ofTAK-659 (60mg,
n¼ 4; 80mg, n¼ 3; 100mg, n¼ 78; 120mg, n¼ 1) and were evaluablefor safety analysis. The median number of cycles of TAK-659 receivedwas 2.0 (range, 1�46); 37% of patients received ≥4 cycles of treatment.
All patients experienced at least one TEAE, and 80 (93%) experiencedat least one grade ≥3 TEAE (Fig. 1). The most common TEAEs(occurring in ≥35% of patients; Fig. 1) were elevations in transaminase(increased AST 64%), pyrexia (60%), increased amylase (44%), diar-rhea (44%), hypophosphatemia (40%), anemia (36%), increased bloodcreatine phosphokinase (CPK; 36%), and increased lipase (35%). Thechanges in laboratory parameters, includingAST, amylase, bloodCPK,lipase, and hypophosphatemia, resulted in no symptoms and werereversible upon withholding study drug or with phosphate repletionfor hypophosphatemia. The total incidence of increased blood CPK islikely underreported, because collection of this laboratory parameterwas only implemented in a later protocol amendment. Themajority ofthe pyrexia, anemia, and diarrhea events were grade 1 or 2. The mostcommon grade ≥3 TEAEs (occurring in >15% of patients) wereincreased amylase and hypophosphatemia (both 26%), neutropenia(24%), anemia, increased blood CPK, and increased lipase (16% each),and thrombocytopenia (15%). Serious AEs (SAEs) were reported in 67patients (78%); themost common were pyrexia (23%) and pneumonia(12%).
Lactate dehydrogenase (LDH) levels increased from baseline innearly all patients, but no clinical correlates (e.g., histology, diseaseburden, response to therapy) were found to be associated with the levelof increase. Laboratory-parameter changes in the three most frequenttreatment-related AEs (AST, amylase, and lipase elevations) and LDHover the first three cycles of treatment showed that the increasesgenerally occurred early (approximately 1 week into treatment) andremained relatively stable over time (Fig. 2). There were no knownepisodes of tumor lysis syndrome and therefore no actions were takenclinically for increases in LDH. TAK-659 was only interrupted perprotocol for grade 4 increases in amylase or lipase, or grade 3 increasesin both amylase and lipase. Transaminase increases were mostly mildand often involved a single enzyme [either AST or alanine amino-transferase (ALT)]. TAK-659 was only interrupted per protocol forgrade 4 elevations in AST or ALT with significant bilirubin elevation(grade ≥3), or grade 3 elevations in both enzymes with significantbilirubin elevation (grade ≥3); dose interruption for these reasons wasinfrequent.
Overall, 81 patients (94%) had at least one treatment-related AE and65 (76%) had at least one grade ≥3 treatment-related AE (Supple-mentary Table S5). Diarrhea (23%), pyrexia (22%), and periorbitaledema (21%) were the most common (occurring in ≥20% of patients)nonlaboratory-parameter and nonhematologic treatment-relatedAEs, most of which were grade 1 or 2. The most common hematologictreatment-related grade ≥3 AEs were neutropenia (17%) and throm-bocytopenia (6%); anemia was only reported in 1 patient (1%).Treatment-related SAEs were experienced by 29 patients (34%); eachtreatment-related SAE occurred in 1 patient only, except for pyrexia(n ¼ 8), pneumonia (n ¼ 6), pneumonitis (n ¼ 4), and febrileneutropenia, PJP, and sepsis (all n ¼ 2).
PharmacokineticsTAK-659 plasma concentration–time profiles on day 1 of cycle 1
and at steady state (day 15 of cycle 1) after oral once-daily adminis-tration in patients with lymphoma are presented in Fig. 3. TAK-659exposure generally increased with dose over the 60–120-mg range.Median time to maximum plasma concentration (Tmax) was approx-imately 2 hours. Variability of dose-normalized, steady-state exposure[area under the concentration–time curve for a dosing interval(AUCtau)/dose] was 20%.
Following oral once-daily administration of TAK-659 for 15 days,the overall mean accumulation was 1.90-fold (range, 1.42–2.41) and
Phase I Study of TAK-659 in Lymphoma
AACRJournals.org Clin Cancer Res; 26(14) July 15, 2020 3549
the overallmean peak/trough ratio on day 15 of cycle 1was 4.85 (range,2.33–10.2). Approximate steady-state pharmacokinetic conditionswere reached by day 8, based on visual inspection of trough concen-trations obtained during cycle 1. Overall geometric mean renal clear-ance was 12.1 L/hour (range, 2.95–25.6), and mean ratio of TAK-659renal clearance to apparent clearance was 0.306 (range, 0.070–0.530),indicating that renal clearance accounted for at least 30% of systemicclearance. Geometric mean terminal half-life was approximately36.6 hours based on data from 14 iNHL run-in patients with serialpharmacokinetic sampling up to 168 hours after a single oral dose ofTAK-659 100 mg.
Antitumor activityResponse to TAK-659 was evaluated in the response-evaluable
population (defined as those who received at least one dose ofTAK-659 and had at least one follow-up assessment), per the statisticalanalysis plan and is shown inTable 2 alongwith response in the intent-to-treat (ITT) lymphoma population. Eight of 43 response-evaluablepatients with DLBCL (19%) achieved a complete response (CR) and 4(9%) achieved a partial response (PR), resulting in an ORR (CRþPR)of 28% in response-evaluable patients (Fig. 4A) and 23% in the ITTpopulation. Median time to response was 53 days (range, 27–168).Among the 12 responders, 9 had a response lasting for≥16weeks and 4were ongoing on study at the data cutoff; median duration of response
was not estimable (NE; range, 1–1,182) and median duration oftreatment in patients with DLBCL with an objective response was435.5 days (range, 53–1,354; Fig. 4B). Two patients with DLBCL withan objective response stopped treatment to undergo allogeneichematopoietic stem cell transplantation (one after cycle 4, one aftercycle 7). Median PFS was 50 days [95% confidence interval (CI), 44–62] in all patients with DLBCL; median was not reached in DLBCLresponders (95% CI, 279–NE; one progressive disease event and theremaining patients were censored) and was 50 days (95%CI, 34–54) inDLBCL nonresponders (Supplementary Fig. S1). Responses were seenin patients with de novo DLBCL and in patients with transformedDLBCL; responses appeared to be independent of DLBCL cell-of-origin classification (Supplementary Table S6).
Among 15 evaluable patients with iNHL, 4 achieved CR [mucosa-associated lymphoid tissue lymphoma (n ¼ 1), FL (n ¼ 2), smalllymphocytic lymphoma (n¼ 1)] and 7 achieved PR [FL (n¼ 6), nodalmarginal zone B-cell lymphoma (n¼ 1)] resulting in anORRof 73% inresponse-evaluable patients and 52% in the ITT population. Of the 11responders, 5 had a response lasting≥16weeks. Among the 9 response-evaluable patients with FL, the ORR was 89% (CR, n ¼ 2; PR, n ¼ 6;57% in the ITT population). Median time to response was 58.5 days(range, 52–148), median duration of response was 137 days (range,1–856), and median duration of treatment in responding patients was174 days (range, 84–896). At the time of data cutoff, 1 patient was still
Figure 1.
Most common TEAEs in patients with lymphoma. Bar lengths may appear different than the corresponding values of frequency due to rounding. ALT, alanineaminotransferase; AST, aspartate aminotransferase; CPK, creatine phosphokinase.
Gordon et al.
Clin Cancer Res; 26(14) July 15, 2020 CLINICAL CANCER RESEARCH3550
on study and 1 patient had transitioned to hematopoietic stem celltransplantation after 8months on treatment.Median PFSwas 201 days(95%CI, 61–907) in all patients with FLwith 3 patients censored due tothe date of the last adequate assessment (Supplementary Fig. S2). Of 5evaluable patients with CLL, 3 (60%) achieved PR and 2 had stabledisease (ORR of 50% in the ITT population).
Of the 6 enrolled patients with CLL, 5 were tested for mutations.None of the 5 patients had BTK mutations and 2 had a PLC-g2mutation (one S707F mutation and one T706I mutation). All 3 CLLresponders had prior ibrutinib and/or idelalisib exposure, includingthe patient with the PLC-g2 S707F mutation. The first responderachieved a best response of PR to ibrutinib in the fourth line and wasrefractory to idelalisib in the fifth line. The second responder wasrefractory to ibrutinib in the second line of treatment. The lastresponder had received rituximab and venetoclax as a second-linetreatment with a best response of CR and was later refractory toibrutinib in the fourth line.
DiscussionWe conducted a phase I, first-in-human study of TAK-659 in
patients with advanced solid tumors or lymphoma. The rationale forinclusion of patients with solid tumors in dose-escalation only wasmultifold: (i) it was exploratory based on emerging data; (ii) itexpedited enrollment and completion of the escalation phase; and(iii) the results of the solid tumor response assessment, with only 1 of19 patients responding, did not justify opening a dedicated solid tumor
cohort in expansion. The role of SYK in the pathogenesis of B-cellmalignancies is well established (4–10), and targeting SYK wasexpected to drive clinical benefit in the target study population ofthis trial (patients with lymphoma). However, it is uncertain whetherantagonizing FLT3 signaling in parallel would contribute further to theclinical activity of TAK-659 in B-cell malignancies.
TheMTDof TAK-659was determined to be 100mg once daily, andthis dose was selected for expansion. Clinical responses were observedat all four doses evaluated, suggesting the therapeutic window of TAK-659 ranges from 60 to 100mg once daily. Toxicity wasmanageable andthere was single-agent antitumor activity in patients with relapsed/refractory B-cell lymphoma across different histologic subtypes. Inparticular, TAK-659 demonstrated a relatively high rate of CR (two-thirds of responders achieved CR), a long duration of treatment(median of 435.5 days among responders at data cutoff), and amedianresponse duration that was NE (range, 1–1182) in later-line DLBCL.TAK-659 also showed encouraging single-agent antitumor activity inresponse-evaluable patients with FL with an ORR of 89% (57% ITT),albeit in a small group of patients and with short duration of response.
The contribution of dual SYK and FLT-3 inhibition to the toxicityprofile of TAK-659 versus SYK inhibition alone is difficult to ascertainbased on the available data from this study. The most commonnonlaboratory AEs that occurred in at least 30% of patients includedpyrexia, diarrhea, fatigue, cough, nausea, and asthenia; these eventswere largely grade 1 and 2. The laboratory elevations in transaminase,pancreatic enzymes, and CPK were not associated with symptoms,and were manageable with dose interruptions and modifications.
Grade 1 Grade 2 Grade 3 Grade 4
AST Grade ranges assigned to LDH; Grade 1, >ULN - 3x ULN; Grade 2, >3x - 5x ULN; Grade 3, >5x - 20x ULN; and Grade 4, > 20x ULN.
85 84 79 74 70 63 61 53 49 34 21 16
0102030405060708090
100
1 8 15 22 29 36 43 50 57 64 71 78
Pat
ient
s (%
)
Days
Amylase
Pts (n)
86 84 79 74 70 63 61 53 49 34 21 16
0102030405060708090
100
1 8 15 22 29 36 43 50 57 64 71 78
Pat
ient
s (%
)
Days
AST
Pts (n)
86 84 79 74 70 63 61 53 49 34 21 16
0102030405060708090
100
1 8 15 22 29 36 43 50 57 64 71 78
Patie
nts
(%)
Days
Lipase
Pts (n)
85 84 78 73 69 60 58 52 49 34 20 15
0102030405060708090
100
1 8 15 22 29 36 43 50 57 64 71 78P
atie
nts
(%)
Days
LDH
Pts (n)
Figure 2.
Laboratory-parameter grade changes over three cycles in patients with lymphoma receiving TAK-659. AST, aspartate aminotransferase; LDH, lactate dehydro-genase; Pts, patients; ULN, upper limit of normal.
Phase I Study of TAK-659 in Lymphoma
AACRJournals.org Clin Cancer Res; 26(14) July 15, 2020 3551
LDH was universally elevated in patients treated with TAK-659and the mechanism is not known. Although asymptomatic elevationof these enzymes was also reported with the SYK inhibitorentospletinib (26–28), it is unclear whether this is a target-based classeffect or an individual drug effect. The most common hematologic AEwas neutropenia, though a significant number of patients had marrowinvolvement at baseline.Most patients (94%) experienced a treatment-related AE; however, the overall incidence of hematologic treatment-related AEs reported in this study was relatively low. Treatment-related grade ≥3 neutropenia and thrombocytopenia was reported in17% and 6% of patients, respectively, and treatment-related grade ≥3anemia was reported in 1 patient only.
Increased susceptibility to infection was identified as a potential riskwith TAK-659 based on the nonclinical toxicology findings of lym-phoid depletion andmyelosuppression (Millennium Pharmaceuticals,Inc., Cambridge, MA, a wholly owned subsidiary of Takeda Pharma-ceutical Company Limited, data on file). In this study, opportunisticinfection and viral reactivation events were seen, including 4 patients
(5%) with treatment-emergent PJP (all grade 3 and related) and 12(14%)with treatment-emergent CMV infection, ofwhich 9 (11%)wereconsidered related to TAK-659. In the 12 patients with CMV reac-tivation, 1 patient had concurrent grade 2 pneumonia, 1 patient hadconcurrent grade 1 pneumonia/grade 2 pneumonitis, and anotherpatient had concurrent grade 3 PJP; however, there were no associatedreports of retinitis, esophagitis, colitis, hepatitis, or encephalitis.Interestingly, the incidence of CMV reactivation was higher in Europe(all 12 cases were reported in either Spain or Italy) compared with theUnited States where no cases were reported for the duration of thestudy. To manage the risk of infection, we initiated prophylacticmeasures during the study against PJP in high-risk patients, as wellas CMVmonitoring and preemptive treatment, as clinically indicatedfollowing local standard practice.
TAK-659 showed encouraging single-agent clinical activity in patientswith lymphomas of different histologies, including DLBCL, iNHL (FL),CLL, and MCL. Durable and deep responses were observed in approx-imately 20% of patients with DLBCL, which is not insignificant in this
400
A
B
300
200
100
Mea
n (S
D) p
lasm
a TA
K-65
9 co
ncen
tratio
n (n
g/m
L)M
ean
(SD
) pla
sma
TAK-
659
conc
entra
tion
(ng/
mL)
Nominal time (hours)
0
400
300
200
100
0
0 4 8 12 16 20 24
Nominal time (hours)0 4 8 12 16 20 24
TAK-659 60 mg QD (n = 4)TAK-659 80 mg QD (n = 3)TAK-659 100 mg QD (n = 9)TAK-659 120 mg QD (n = 1)
TAK-659 60 mg QD (n = 4)TAK-659 80 mg QD (n = 1)TAK-659 100 mg QD (n = 4)TAK-659 120 mg QD (n = 1)
Figure 3.
Plasma concentration–time profiles of TAK-659 inpatients with lymphoma following oral once-dailyadministration on cycle 1 day 1 (A) and cycle 1 day 15(B). QD, once daily.
Gordon et al.
Clin Cancer Res; 26(14) July 15, 2020 CLINICAL CANCER RESEARCH3552
heavily pretreated patient population (Fig. 4B). The longer than 12-month duration of treatment in responding patients by the data cutoff,and the CR rate of 19% (two-thirds of responders had CR; 15% ITT)reportedhere forDLBCLcould be attributable to the role of SYK, not onlyin the direct targeting of tumor cells but also in local immunity, ahypothesis supported by preclinical data, which demonstrate thatTAK-659 exhibits immunomodulatory effects that control tumorgrowth (19). These effects include inhibition of tumor-associated T-regulatory cells and MDSCs, as well as upregulation of antitumorigenicmacrophages (19). Furthermore, responses to TAK-659 in patients withDLBCL appeared to be independent of cell-of-origin classification, asresponses were observed in both patients with GCB and non-GCB;however, more data are needed before making confirmatory statementsabout where this agent would fit in lymphoma treatment with respect tocell of origin. Antitumor activity was also observed in patients with CLLwho received prior ibrutinib and/or prior idelalisib therapy, suggestingthat TAK-659 could be beneficial in treating B-cell malignancies lesssensitive to BTK inhibitors, although larger studies are needed to confirmthis observation.
For TAK-659–treated response-evaluable patients with lymphoma,the ORR of 40% (31% ITT), including the 12 patients (18%; 14% ITT)who achieved CR, is notable; response-evaluable patients with FLresponded particularly well, with an ORR of 89% (57% ITT). Whilecross-study comparisons have major limitations, in an initial phase IIstudy of the SYK inhibitor fostamatinib (29), activity was limited, withan ORR of 3% reported in relapsed/refractory patients with DLBCL;13% of patients reported at least stable disease (30). In a phase II studyof entospletinib (800mg twice a day), no patients achievedCR and 61%achieved PR in relapsed/refractory CLL (26), and the ORR in iNHLwas reported to be 13% (26). In a similar phase II study of entospletinib(800 mg twice a day), the ORR in patients with FL was 17% (all werePRs) with 51% achieving stable disease (27). In a phase I, dose-escalation study of cerdulatinib in patients with relapsed/refractoryB-cellmalignancies, of 13 patients with FL, 1 patient achievedCR and 1achieved PR; none of the 16 patients with DLBCL responded (31). In aphase II study, cerdulatinib has demonstrated response rates of 43%–61% in relapsed/refractory B- and T-cell NHL [61% in CLL/SLL, 50%in FL, and 43% in PTCL (4 CRs and 2 PRs in 14 patients; ref. 31)].Moreover, preliminary phase II data of cerdulatinib in relapsed/refractory FL have demonstrated an ORR of 46% as a single-agentand 67% in combination with rituximab (32).
The pharmacokinetic properties of TAK-659 in this study sup-ported continuous, oral once-daily dosing. However, later studies ofTAK-659 are investigating alternative dosing schedules (e.g., inter-mittent). TAK-659 has favorable pharmaceutical properties, namely,low protein binding (free fraction of 0.55 in human plasma), goodsolubility, fast absorption (Tmax �2 hours), and a relatively longterminal half-life in humans (�37 hours).
Unfortunately, no informative pharmacodynamic data were gen-erated in this study to support the dose determination. SYK and FLT3inhibition by TAK-659 and correlation with response could not beassessed in this study. Details of our efforts to evaluate downstreamsignaling of SYK and FLT-3 as potential pharmacodynamicmarkers ofTAK-659 activity are provided in the Supplementary Materials. Tostudy the pharmacodynamic effects of TAK-659 inhibition of BCR andFLT3 signaling, modulation of multiple proximal signal transducerswas evaluated in cell lines byflowcytometry, including phosphorylatedSYK, FLT3, BLNK, and BTK.Onlymodest regulation by TAK-659wasobserved (Supplementary Fig. S3A–S3D), which limited their utility asa pharmacodynamic measure of activity in a clinical application.Both SYK and FLT3 are known to regulate AKT/mammalian targetof rapamycin signaling (33, 34); thus, the phosphorylation of S6kinase on serine 235/236 was also evaluated as a potential TAK-659pharmacodynamic readout. TAK-659 treatment was shown todecrease the phosphorylation of S6 kinase in multiple cell linesconstitutively expressing phosphorylated SYK, with maximal inhi-bition seen in cell lines also harboring a FLT3-internal tandemdeletion mutation (Supplementary Fig. S3E). Thus, there was alack of significant change compared with dimethyl sulfoxide, butTAK-659 treatment decreased the phosphorylation of S6 kinase. Inprimary human samples, this assay was shown to be variable due tothe quality and handling of the primary samples and, therefore, thedata were not conclusive.
Although TAK-659 100mg once daily was determined as theMTD,it remains an open question whether this is a biologically optimal doseand safest long-term dose. Due to the aggressive nature of DLBCL andthe need for immediate disease control, the level of dose interruption inthe first two cycles may greatly influence efficacy. Only 29% of patientswith lymphoma dosed at TAK-659 100 mg proceeded through cycle 1without dose modification. Common reasons for dose modificationincluded neutropenia, hypophosphatemia, bloodCPK increased, amy-lase increased, periorbital edema, and pyrexia. A dose of TAK-659 that
Table 2. Best antitumor response and duration of treatment.
ITT Response-evaluable patients
Tumor type nORR,n (%) n
ORR,n (%)
CR,n (%)
PR,n (%)
Median timeto response,days (range)
Median durationof response,days (range)
Median duration oftreatment in respondingpatients, days (range)
All lymphoma 86 27 (31) 67 27 (40) 12 (18) 15 (22) 57 (27–168) 856 (1–1182) 189 (53–1354)DLBCL 53 12 (23) 43 12 (28) 8 (19) 4 (9) 53 (27–168) NE (1–1182) 435.5 (53–1354)iNHLa 21 11 (52) 15 11 (73) 4 (27) 7 (47) 60 (52–148) 176 (1–856) 185 (64–896)
FL 14 8 (57) 9 8 (89) 2 (22) 6 (67) 58.5 (52–148) 137 (1–856) 174 (84–896)CLL 6 3 (50) 5 3 (60) 0 3 (60) 57 (56–107) NE (1–59) 138 (104–155)MCL 5 1 (20) 3 1 (33) 0 1 (33) 54 (54–54) 109 (109–109) 149 (149–149)PTLD 1 0 1 0 0 0 - - -
Abbreviations: CLL, chronic lymphocytic leukemia; CR, complete response; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; iNHL, indolent non-Hodgkin lymphoma; ITT, intent-to-treat; MCL, mantle cell lymphoma; NE, not estimable; ORR, overall response rate; PR, partial response; PTLD, posttransplantlymphoproliferative disease.aIncludes nodal marginal zone B-cell lymphoma (n ¼ 2), mucosa-associated lymphoid tissue, B-cell lymphoplasmacytic lymphoma/immunocytoma, B-cell smalllymphocytic lymphoma, and splenic marginal zone lymphoma (n ¼ 1 each).
Phase I Study of TAK-659 in Lymphoma
AACRJournals.org Clin Cancer Res; 26(14) July 15, 2020 3553
results in substantial dose modification during initial treatment—andthus potentially compromises efficacy—may be suboptimal. Toaddress this, lower doses and/or intermittent schedules are beinginvestigated (NCT03123393, NCT03357627, NCT02954406).
On the basis of these findings, further investigation of TAK-659 iswarranted in patients with relapsed/refractory B-cell lymphoma(NCT02954406, NCT03357627), and in early-line combination withstandard chemotherapy (NCT03742258). Given the established role ofSYK in BCR-driven pathogenesis of B-cell malignancies, additionaltranslational research to identify potential patient selection markersmay also prove productive.
Disclosure of Potential Conflicts of InterestR. Popat is an employee/paid consultant for, reports receiving speakers’ bureau
honoraria from, and reports receiving other remuneration from Takeda. S. Madan
reports receiving speakers’ bureau honoraria from and is an advisory board member/unpaid consultant for Takeda. G. Gritti reports receiving speakers’ bureau honorariafrom IQVIA. D. El-Sharkawi reports receiving speakers’ bureau honoraria fromAbbvie and Janssen, and is an advisory board member/unpaid consultant for Abbvie.I. Chau is an employee/paid consultant for Bristol-Myers Squibb, Eli-Lilly, MSD,AstraZeneca, Roche, Merck-Serono, Oncologie International, Pierre Fabre, andBayer; and reports receiving commercial research grants from Eli-Lilly and JanssenCilag. J.A. Radford reports receiving commercial research grants from Takeda andreports receiving speakers bureau honoraria from speaker engagements. J.P. deOteyza is an employee/paid consultant for and reports receiving commercial researchgrants from Takeda. S. Iyer reports receiving commercial research grants fromTakeda, Rhizen, Seattle Genetics, and Merck. R. Karmali reports receiving othercommercial research support from Takeda, BMS, Kite/Gilead, BMS/Celgene/Juno;reports receiving speakers bureau honoraria fromKite/Gilead, AstraZeneca, BeiGene;and is an advisory board member/unpaid consultant for BMS/Celgene/Juno,Kite/Gilead, and Karyopharm. H. Miao, I. Proscurshim, and S. Wang are employ-ees/paid consultants for Millennium Pharmaceuticals Inc., Cambridge, MA, USA, a
DLBCL(n = 39)
iNHL(n = 14)
CLL(n = 5)
MCL(n = 3)
* *
* * **
**
*
*
*
FL FLFLFL FLFLFLFLFL−100
−50
0
50
100
Targ
et le
sion
cha
nge
from
bas
elin
e (%
)
PDSDPRCR
* PD due to new lesion
A
FL
FL
FL
FLFLFLFLFL
FL
*
*
**
*
*
**
**
*
*
*
*
**
*
*
**
*
*
*
*
*
*
*
0 90 180 270
MCL(n = 3)
CLL(n = 5)
iNHL(n = 15)
DLBCL(n = 43)
Days
//
Start of response, → ongoing*
PD
SD
PR (n = 15)
CR (n = 12)
→→ → →
360 720 1,080 1,440
B
Figure 4.
A, Antitumor response to TAK-659 inresponse-evaluable patients: best per-centage change from baseline in thesum of the products of the diametersamong all patients with lymphomawho received at least one dose ofstudy drug and at least one follow-up assessment. B, Response and timeon study by patient. CLL, chronic lym-phocytic lymphoma; CR, completeresponse; DLBCL, diffuse large B-celllymphoma; FL, follicular lymphoma;iNHL, indolent non-Hodgkin lympho-ma; MCL, mantle cell lymphoma; ORR,overall response rate; PD, progressivedisease; PR, partial response; SD, sta-ble disease.
Gordon et al.
Clin Cancer Res; 26(14) July 15, 2020 CLINICAL CANCER RESEARCH3554
wholly owned subsidiary of Takeda Pharmaceutical Company Limited. Y. Wu is anemployee/paid consultant for Millennium Pharmaceuticals Inc., a wholly ownedsubsidiary of Takeda Pharmaceutical Company. K. Stumpo is an employee/paidconsultant for Millennium Pharmaceuticals Inc., a wholly owned subsidiary ofTakeda Pharmaceutical Company, and holds ownership interest (including patents)in AstraZeneca, Teva Pharmaceuticals, GSK, BMS, and Pfizer. C. Carpio reportsreceiving speakers bureau honoraria from Takeda. F. Bosch reports receivingcommercial research grants from Hoffman La Roche, Janssen, Celgene, Gilead, andNovartis; reports receiving speakers’ bureau honoraria from Abbvie, Janssen,AstraZeneca, and Roche; and is an advisory board member/unpaid consultant forAbbvie, Janssen, AstraZeneca, and Gilead. No potential conflicts of interest weredisclosed by the other authors.
Authors’ ContributionsConception and design: L.I. Gordon, H.A. Burris, S. Iyer, Y. Shou, C. CarpioDevelopment of methodology: L.I. Gordon, H. Miao, Y. ShouAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): L.I. Gordon, J.B. Kaplan, R. Popat, H.A. Burris, M.R. Patel, G. Gritti,D. El-Sharkawi, I. Chau, J.A. Radford, J.P. de Oteyza, P. Luigi Zinzani, S. Iyer,W. Townsend, R. Karmali, H. Miao, I. Proscurshim, S. Wang, Y. Shou, C. Carpio,F. BoschAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): L.I. Gordon, J.B. Kaplan, H.A. Burris, M.R. Patel,G. Gritti, I. Chau, S. Iyer, W. Townsend, R. Karmali, H. Miao, I. Proscurshim,S. Wang, Y. Wu, Y. Shou, F. BoschWriting, review, and/or revision of the manuscript: L.I. Gordon, J.B. Kaplan,R. Popat, H.A. Burris, S. Ferrari, S. Madan, M.R. Patel, G. Gritti, D. El-Sharkawi,I. Chau, J.A. Radford, J.P. de Oteyza, P. Luigi Zinzani, S. Iyer, W. Townsend,
R. Karmali, H. Miao, I. Proscurshim, S. Wang, Y. Wu, K. Stumpo, Y. Shou,C. Carpio, F. BoschAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): H. Miao, K. Stumpo, C. CarpioStudy supervision: L.I. Gordon, H.A. Burris, M.R. Patel, S. Iyer, R. Karmali, H. Miao,I. Proscurshim, K. Stumpo, Y. ShouOther (clinical care of patients and delivery of treatment): W. Townsend
AcknowledgmentsThe authors thank the study participants and their families, and the staff at all
study centers. Research was sponsored by Millennium Pharmaceuticals, Inc., Cam-bridge, MA, USA, a wholly owned subsidiary of Takeda Pharmaceutical CompanyLimited. The authors also acknowledge Emily Sheldon-Waniga (Millennium Phar-maceuticals, Inc.) for statistical analyses, HelenWilkinson, PhD (FireKite, anAshfieldCompany, part of UDG Healthcare plc), for medical writing support, which wasfunded by Millennium Pharmaceuticals, Inc., and Janice Y. Ahn, PhD, and MarcelKuttab PharmD (Millennium Pharmaceuticals, Inc.) for editorial support in com-pliance with Good Publication Practice 3 ethical guidelines (Battisti et al, Ann InternMed 2015;163:461–4). Dr. Popat is supported by the National Institute for HealthResearch University College London Hospitals Biomedical Research Centre.
The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.
Received October 22, 2019; revised March 11, 2020; accepted April 17, 2020;published first April 23, 2020.
References1. Cheng AM, Rowley B, Pao W, Hayday A, Bolen JB, Pawson T. SYK tyrosine
kinase required for mouse viability and B-cell development. Nature 1995;378:303–6.
2. Cornall RJ, Cheng AM, Pawson T, Goodnow CC. Role of Syk in B-celldevelopment and antigen-receptor signaling. Proc Natl Acad Sci U S A 2000;97:1713–8.
3. Turner M, Mee PJ, Costello PS, Williams O, Price AA, Duddy LP, et al. Perinatallethality and blocked B-cell development inmice lacking the tyrosine kinase Syk.Nature 1995;378:298–302.
4. Buchner M, Fuchs S, Prinz G, Pfeifer D, Bartholome K, Burger M, et al.Spleen tyrosine kinase is overexpressed and represents a potential thera-peutic target in chronic lymphocytic leukemia. Cancer Res 2009;69:5424–32.
5. Cheng S, Coffey G, Zhang XH, Shaknovich R, Song Z, Lu P, et al. SYK inhibitionand response prediction in diffuse large B-cell lymphoma. Blood 2011;118:6342–52.
6. Mocsai A, Ruland J, Tybulewicz VL. The SYK tyrosine kinase: a crucial player indiverse biological functions. Nat Rev Immunol 2010;10:387–402.
7. GururajanM, Jennings CD, Bondada S. Cutting edge: constitutive B cell receptorsignaling is critical for basal growth of B lymphoma. J Immunol 2006;176:5715–9.
8. Leseux L, Hamdi SM, Al Saati T, Capilla F, Recher C, Laurent G, et al.Syk-dependent mTOR activation in follicular lymphoma cells. Blood 2006;108:4156–62.
9. Gobessi S, Laurenti L, Longo PG, Carsetti L, Berno V, Sica S, et al. Inhibition ofconstitutive and BCR-induced Syk activation downregulates Mcl-1 and inducesapoptosis in chronic lymphocytic leukemia B cells. Leukemia 2009;23:686–97.
10. Chen L, Monti S, Juszczynski P, Daley J, Chen W, Witzig TE, et al. SYK-dependent tonic B-cell receptor signaling is a rational treatment target in diffuselarge B-cell lymphoma. Blood 2008;111:2230–7.
11. CenO, KannanK, Huck Sappal J, Yu J, ZhangM, ArikanM, et al. Spleen tyrosinekinase inhibitor TAK-659 prevents splenomegaly and tumor development in amurine model of Epstein-Barr virus-associated lymphoma. mSphere 2018;3:e00378–18.
12. Du ZM, Kou CW, Wang HY, Huang MY, Liao DZ, Hu CF, et al. Clinicalsignificance of elevated spleen tyrosine kinase expression in nasopharyngealcarcinoma. Head Neck 2012;34:1456–64.
13. Hao YT, Peng CL, Zhao YP, Sun QF, Zhao XG, Cong B. Effect of spleen tyrosinekinase on nonsmall cell lung cancer. J Cancer Res Ther 2018;14:S100–4.
14. Katz E, Dubois-Marshall S, Sims AH, Faratian D, Li J, Smith ES, et al. A gene onthe HER2 amplicon, C35, is an oncogene in breast cancer whose actions areprevented by inhibition of Syk. Br J Cancer 2010;103:401–10.
15. Yu Y, Suryo Rahmanto Y, Shen YA, Ardighieri L, Davidson B, Gaillard S, et al.Spleen tyrosine kinase activity regulates epidermal growth factor receptorsignaling pathway in ovarian cancer. EBioMedicine 2019;47:184–94.
16. Lam B, Arikawa Y, Cramlett J, Dong Q, de Jong R, Feher V, et al. Discovery ofTAK-659 an orally available investigational inhibitor of spleen tyrosine kinase(SYK). Bioorg Med Chem Lett 2016;26:5947–50.
17. Huck J, Kabir S, Kuida K, Kannan K, Shou Y, Petrich A. Preclinical and clinicaldevelopment of TAK-659: an investigational SYK and FLT3 kinase inhibitor.Paper presented at 12th International Ultmann Chicago Lymphoma Sympo-sium; 24–25 April 2015; W Chicago City Center, Chicago, IL.
18. PurroyN, Carabia J, Abrisqueta P, Egia L, AguiloM, Carpio C, et al. Inhibition ofBCR signaling using the Syk inhibitor TAK-659 prevents stroma-mediatedsignaling in chronic lymphocytic leukemia cells. Oncotarget 2017;8:742–56.
19. Kannan K, Riley J, Zhang M, Farrell P, Bailey B, Creson J, et al. TAK-659, a dualSYK/FLT3 inhibitor, leads to complete and sustained tumor regression andimmune memory against tumor cells upon combination with anti-PD-1 agent.Eur J Cancer 2016;69:S92.
20. Gabrilovich DI, Nagaraj S. Myeloid-derived suppressor cells as regulators of theimmune system. Nat Rev Immunol 2009;9:162–74.
21. Pan PY, Wang GX, Yin B, Ozao J, Ku T, Divino CM, et al. Reversion of immunetolerance in advanced malignancy: modulation of myeloid-derived suppressorcell development by blockade of stem-cell factor function. Blood 2008;111:219–28.
22. Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero G, Dohner H,et al. Guidelines for the diagnosis and treatment of chronic lymphocyticleukemia: a report from the International Workshop on Chronic LymphocyticLeukemia updating the National Cancer Institute-Working Group 1996 guide-lines. Blood 2008;111:5446–56.
23. Cheson BD, Pfistner B, Juweid ME, Gascoyne RD, Specht L, Horning SJ, et al.Revised response criteria formalignant lymphoma. JClinOncol 2007;25:579–86.
24. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al.New response evaluation criteria in solid tumours: revised RECIST guideline(version 1.1). Eur J Cancer 2009;45:228–47.
25. Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie J, Ott G, et al.Confirmation of the molecular classification of diffuse large B-cell lymphoma byimmunohistochemistry using a tissue microarray. Blood 2004;103:275–82.
AACRJournals.org Clin Cancer Res; 26(14) July 15, 2020 3555
Phase I Study of TAK-659 in Lymphoma
26. Sharman J,HawkinsM,KolibabaK, BoxerM,Klein L,WuM, et al. An open-labelphase 2 trial of entospletinib (GS-9973), a selective spleen tyrosine kinaseinhibitor, in chronic lymphocytic leukemia. Blood 2015;125:2336–43.
27. Andorsky DJ, Kolibaba KS, Assouline S, Forero-Torres A, Jones V, Klein LM,et al. An open-label phase 2 trial of entospletinib in indolent non-Hodgkinlymphoma and mantle cell lymphoma. Br J Haematol 2019;184:215–22.
28. Burke JM, Shustov A, Essell J, Patel-Donnelly D, Yang J, Chen R, et al. An open-label, phase II trial of entospletinib (GS-9973), a selective spleen tyrosine kinaseinhibitor, in diffuse large B-cell lymphoma. Clin Lymphoma Myeloma Leuk2018;18:e327–31.
29. Friedberg JW, Sharman J, Sweetenham J, Johnston PB, Vose JM, Lacasce A, et al.Inhibition of Syk with fostamatinib disodium has significant clinical activity innon-Hodgkin lymphoma and chronic lymphocytic leukemia. Blood 2010;115:2578–85.
30. Flinn IW, Bartlett NL, Blum KA, Ardeshna KM, LaCasce AS, Flowers CR,et al. A phase II trial to evaluate the efficacy of fostamatinib in patients with
relapsed or refractory diffuse large B-cell lymphoma (DLBCL). Eur J Cancer2016;54:11–7.
31. Hamlin PA, Flinn IW, Wagner-Johnston N, Burger JA, Coffey GP, Conley PB,et al. Efficacy and safety of the dual SYK/JAK inhibitor cerdulatinib in patientswith relapsed or refractory B-cell malignancies: results of a phase I study. Am JHematol 2019;94:E90–93.
32. Smith SD, Munoz J, Stevens DA, Smith SM, Feldman TA, Ye JC, et al. Rapid anddurable responses with the SYK/JAK Inhibitor cerdulatinib in a phase 2 study inrelapsed/refractory follicular lymphoma—alone or in combination with ritux-imab. Hematol Oncol 2019;37:61–62.
33. Kazi JU, Ronnstrand L. FMS-like tyrosine kinase 3/FLT3: from basic science toclinical implications. Physiol Rev 2019;99:1433–66.
34. Chen L, Monti S, Juszczynski P, Ouyang J, Chapuy B, Neuberg D, et al. SYKinhibition modulates distinct PI3K/AKT- dependent survival pathways andcholesterol biosynthesis in diffuse large B cell lymphomas. Cancer Cell 2013;23:826–38.
Clin Cancer Res; 26(14) July 15, 2020 CLINICAL CANCER RESEARCH3556
Gordon et al.
CLINICAL CANCER RESEARCH | PRECISION MEDICINE AND IMAGING
Tumor Microenvironment Composition and SevereCytokine Release Syndrome (CRS) Influence Toxicity inPatients with Large B-Cell Lymphoma Treated withAxicabtagene Ciloleucel A C
Rawan Faramand1,2,3, Michael Jain1,2,3, Verena Staedtke4, Hiroshi Kotani5, Renyuan Bai6, Kayla Reid5,Sae Bom Lee2, Kristen Spitler5, Xuefeng Wang7, Biwei Cao7, Javier Pinilla2,8, Aleksander Lazaryan1,2,Farhad Khimani1,2, Bijal Shah2,8, Julio C. Chavez2,8, Taiga Nishihori1,2, Asmita Mishra1,2, John Mullinax2,9,Ricardo Gonzalez2,9, Mohammad Hussaini2,10, Marian Dam1, Brigett D. Brandjes1, Christina A. Bachmeier1,Claudio Anasetti1,2,3, Frederick L. Locke1,2,3, and Marco L. Davila1,2,3
ABSTRACT◥
Purpose: One of the challenges of adoptive T-cell therapy is thedevelopment of immune-mediated toxicities including cytokinerelease syndrome (CRS) and neurotoxicity (NT). We aimed toidentify factors that place patients at high risk of severe toxicityor treatment-related death in a cohort of 75 patientswith large B-celllymphoma treated with a standard of care CD19 targeted CART-cell product (axicabtagene ciloleucel).
Experimental Design: Serum cytokine and catecholamine levelswere measured prior to lymphodepleting chemotherapy, on the dayof CAR T infusion and daily thereafter while patients remainedhospitalized. Tumor biopsies were taken within 1 month prior toCAR T infusion for evaluation of gene expression.
Results: We identified an association between pretreatmentlevels of IL6 and life-threatening CRS and NT. Because the riskof toxicity was related to pretreatment factors, we hypothesizedthat the tumor microenvironment (TME) may influence CART-cell toxicity. In pretreatment patient tumor biopsies, geneexpression of myeloid markers was associated with highertoxicity.
Conclusions: These results suggest that a proinflammatorystate and an unfavorable TME preemptively put patients at riskfor toxicity after CAR T-cell therapy. Tailoring toxicity man-agement strategies to patient risk may reduce morbidity andmortality.
IntroductionTwo anti-CD19 CAR T-cell products, axicabtagene ciloleucel
(axi-cel) and tisangenlecleucel, are approved by the FDA on the basisof clinical trials that report durable remissions in approximately 40%of
patients with relapsed or refractory large B-cell lymphoma (R/R LBCL;refs. 1, 2). Despite clinical activity in a poor prognosis LBCL popu-lation, thewide spread utilization of these products is limited by uniquetoxicities caused by the en-masse activation of tumor-reactive T cellsinducing a large release of cytokines causing cardiovascular, pulmo-nary, and neurologic toxicities (1, 3–7). These toxicities, termedcytokine release syndrome (CRS) and/or neurotoxicity (NT), are amajor source of morbidity and, in some patients, mortality (1–3).
To date, animal models have demonstrated that the severity ofimmune toxicities is associated with IL6, IL1, and nitric oxide pro-duced by macrophages (8–11). However, there remains a need tovalidate these observations and use accessible laboratory tests todiagnose, monitor, and/or prognosticate toxicities. We and othershave identified the utility of laboratory measurements of C reactiveprotein (CRP) and ferritin in the diagnosing and/or monitoring ofCRS (12–16). Furthermore, several cytokines have been associatedwith severe CRS and/or NT, including IL1b, IL6, IL15, angiopoietin1&2, and GM-CSF supporting the development of targeted therapiesto mitigate these toxicities (1, 12, 13, 17, 18).
Here, we report findings of serum and tumor analysis of 75 patientswith aggressive R/R LBCL treated at our institution with axi-cel. Wecorrelated cytokines and other serummarkers with the development oftoxicities. We found that clinical outcomes, including objectiveresponses and toxicity, are similar to the ZUMA-1 registrational trialeven though the majority of the patients we treated would not havebeen eligible (1, 5). Further, we identified several cytokines andlaboratory markers which correlate with severe immune mediatedtoxicities as previously reported in patients treated under the pivotalclinical trials (1, 3, 15, 16, 19). We observed that a subset of patientswith elevated IL6 levels prior to lymphodepleting chemotherapy are at
1Department of Blood andMarrowTransplantation andCellular Immunotherapy,H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida. 2MorsaniCollege of Medicine, University of South Florida, Tampa, Florida. 3Department ofImmunology, H. Lee Moffitt Cancer Center and Research Institute, Tampa,Florida. 4Department of Neurology, Johns Hopkins University School of Med-icine, Baltimore, Maryland. 5Clinical Science Division, H. Lee Moffitt CancerCenter and Research Institute, Tampa, Florida. 6Department of Neurosurgery,Johns Hopkins University School ofMedicine, Baltimore, Maryland. 7Departmentof Biostatistics and Bioinformatics, H. Lee Moffitt Cancer Center and ResearchInstitute, Tampa, Florida. 8Department of Malignant Hematology, H. Lee MoffittCancer Center and Research Institute, Tampa, Florida. 9Department of Sarcoma,H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida. 10Depart-ment of Pathology, H. Lee Moffitt Cancer Center and Research Institute, Tampa,Florida.
Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).
R. Faramand, M. Jain, and V. Staedtke contributed equally as the co-seniorauthors of this article.
Corresponding Author: Marco L. Davila, Moffitt Cancer Center, 12902 MagnoliaDrive, FOB-3, Tampa, FL 33612. Phone: 813-745-7202; E-mail:[email protected]
Clin Cancer Res 2020;XX:XX–XX
doi: 10.1158/1078-0432.CCR-20-1434
�2020 American Association for Cancer Research.
AACRJournals.org | 1
high-risk for death from CRS. These patients experienced significanttoxicities despite management with standard early cytokine-blockadeand/or steroids as per guidelines (4, 6). By performing gene expressionstudies on baseline tumor tissue samples, we describe immune path-ways that correlate with toxicities. Our work suggests that recognitionof an unfavorable tumor microenvironment (TME) and baselineproinflammatory state can identify patients who are at highest riskof developing severe toxicity after CAR T-cell therapy.
Materials and MethodsSpecimen collection protocol
This is a sample study to evaluate factors associated with thedevelopment of immune-mediated toxicities in patients treated withaxi-cel at H. Lee Moffitt Cancer Center. The study is open for allpatients treated with commercial axi-cel at our institution. Theinstitutional review board reviewed and approved the protocol. Allclinical investigation was conducted according to the Declaration ofHelsinki principles. All patients provided written informed consent. Atotal of 75 patients were enrolled and underwent leukapheresis and areincluded in this study. Patients were lymphodepleted with fludarabine30 mg/m2/day � 3 days and cyclophosphamide 500 mg/m2/day �3 days (1, 5). Patients received bridging treatment between the time ofleukapheresis and lymphodepleting chemotherapy at the discretion ofthe treating physician (1, 5).
All patients had histologically confirmed large B-cell lymphoma(LBCL) or transformed indolent lymphoma. Seventy patients had R/Rdisease following two lines of systemic therapy and met the FDA labelfor the administration of commercial axi-cel. Five patients also meteligibility criteria having failed two or more lines of systemic therapyand were enrolled in ZUMA-9, which is an expanded access studyfor patients who did not meet commercial release specifications(NCT03153462).
Toxicity grading and managementCRS was defined and graded using the ASTCT consensus grading
guidelines (4). Although some patients included in this cohort wereenrolled prior to the publication of the ASTCT consensus gradingguidelines (4),medical recordswere reviewed and patients were gradedretrospectively. Severe CRS was defined as grade � 3 toxicity. Theinstitutional management guidelines were adapted from the CAR-T-cell-therapy-associated TOXicity (CARTOX) Working Group guide-lines and are included in Supplementary Fig. S1 (6).
During the sample collection study, guidelines for the terminologyand grading of neurologic toxicity changed from Common Termi-nology Criteria for Adverse Events (CTCAE) v4.03 to CAR T-cellencephalopathy syndrome (CRES) and most recently to immuneeffector cell-associated neurotoxicity syndrome (ICANS). To simplythe terminology, we used the general term neurologic toxicity toencompass all neurologic toxicity related to the administration ofaxi-cel. For the majority of patients (n ¼ 49), neurologic toxicity wasgraded using the CARTOX guidelines (6). The remainder of patientshad a daily ICANS assessment whereas hospitalized (n¼ 25) and weregraded using the ASTCT guidelines (4). One patient was gradedaccording to CTCAE v4.03. Severe toxicity was defined as grade� 3 NT. Management NT was per institutional guidelines, whichwere adapted from the CARTOX working group and are included inSupplementary Fig. S2 (6).
Clinical responsePatients underwent response assessment using standard of care PET
and/or CT scans at baseline prior to lymphodepleting chemotherapy.To evaluate for response, repeat scans were completed at approxi-mately 4 weeks, 3 months, 6 months, and 12 months following axi-celinfusion and as clinically indicated thereafter. Tumor response wasdetermined by the treating physician per Lugano 2014 classifica-tion (20). The objective response rate at day 90 was defined as eithercomplete remission (CR) or partial remission (PR) by Lugano clas-sification and does not reflect ongoing best response. Progression and/or relapse were determined by the treating physician according toradiographic or clinical criteria and confirmed by biopsywhen feasible.
Serum studiesSerum samples were collected at baseline (within 30 days of lym-
phodepleting chemotherapy), and then daily starting the day of axi-celinfusion until discharge or up to day 60, whichever occurred first. Fifty-twopatients had samples available for baseline analysis. For themajorityof patients (n ¼ 38), baseline samples were collected within 1 week oflymphodepleting chemotherapy. The remainder of baseline samples(n ¼ 14) were collected at the time of apheresis. Peak levels weredetermined as the highest levels starting at day one post infusion untillast day of hospitalization. Cytokines analyzed include GM-CSF, IL1b,IL2, IL6, IL15, IFNg , TNFa, and angiopoietin 1&2. Serumwas analyzedusing the Ella automated simple plex immunoassay system (Protein-Simple). GM-CSF samples were not available for baseline analysis. Atwo-fold dilution of serum sample was loaded at 50 mL per well of aSimple Plex cartridge testing four analytes at a time. Each cytokine wastested in triplicate with average values reported in pg/mL. Using patientserum samples that were previously analyzed with the Luminex andreportedbyPark and colleagues, we confirmed the reliability of cytokineanalysis by the Ella (19). We demonstrated that cytokine values (IFNg ,TNFa, IL6) measured on the Ella correlated strongly with cytokinevalues measured on the Luminex (Supplementary Figs. S3A–S3D). TheDepartment of Pathology and Laboratory Medicine at H. Lee MoffittCancer Center used standard clinical tests to measure serum CRP andferritin levels at baseline, on the day of CAR T infusion and daily whilepatients were hospitalized. Serum samples for catecholamine analysiswere collected at baseline and then daily starting the day of axi-celinfusion in thirty patients. Catecholamine measurements were com-pleted as described by Staedtke and colleagues (21).
GenomicsTumor biopsies were taken within 1month prior to axi-cel infusion
and reviewed by a hematopathologist for tumor content. RNA was
Translational Relevance
Patient and disease factors present prior to chimeric antigenreceptor T (CAR T) cell infusion influence the development ofimmune mediated toxicities. We identify an association between aproinflammatory tumor microenvironment and severe cytokinerelease syndrome (CRS) and neurotoxicity in patients with largeB-cell lymphoma treated with a standard of care CAR T product.We report a novel finding of elevated peak noradrenaline levels inpatients with severe CRS. Further, we describe the role of myeloidcells and regulatory T cells in the lymphoma tumor microenvi-ronment of patients with severe toxicities. Our understanding ofthe impact of myeloid cells and the tumor microenvironment onmediating toxicities is evolving and warrants further investigationin a clinical trial setting.
Faramand et al.
Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH2
extracted from retrospective formalin-fixed paraffin-embedded mate-rial (n ¼ 4) or prospectively frozen material (n ¼ 32). To increasesample size, seven patient tumor samples were collected outside of the75-patient cohort. All patients signed informed written consent. RNAexpression was measured by the Nanostring IO360 panel consisting of770 genes found in the TME in cancer or RNA-seq as indicated at aminimum depth of 80M reads per sample. Nanostring analysis usednSolver to identify cell types, pathway scores and differential geneexpression between groups. All Nanostring and RNA-seq data havebeen deposited to the Gene Expression Omnibus data repository(GSE153439).
Nucleic acid-sequencing libraries were prepared using the NuGenFFPE RNA-Seq Multiplex System (Tecan US, Inc.). Briefly, 50 ng ofDNase treated RNA was used to generate cDNA and a strand-specificlibrary following the manufacturer's protocol. Library molecules con-taining ribosomal RNA sequences were depleted using the NuGenAnyDeplete probe-based enzymatic process. The final libraries wereassessed for quality on the Agilent TapeStation (Agilent Technologies,Inc.), and qRT-PCR for library quantification was performed using theKapa Library Quantification Kit (Roche Sequencing). The librarieswere sequenced on the Illumina NextSeq 500 sequencer with a 75-basepaired-end run to generate 80 to 100 million read pairs per sample.
Read adapters were detected using BBMerge (v37.88; ref. 22) andsubsequently removedwith cutadapt (v1.8.1) (23). Processed raw readswere then aligned to human genome HG19 using STAR (v2.5.3a;ref. 24). Gene expression was evaluated as read count at gene level withHTSeq (v0.6.1) and Gencode gene model (25). Gene express data werethen normalized and differential expression between experimentalgroups were evaluated using DEseq2 (26). Gene set enrichmentanalysis (GESA) for RNA-seq expression profiles was performed usingthe Broad GSEA software version 3.0 (http://software.broadinstitute.org/gsea) with default settings and the phenotype label as the permu-tation type. The Molecular Signatures Database (MSigDB) version6.2 gene set collections were evaluated in the GSEA.
Statistical analysisComparisons of cytokine levels (log2 transformed) between different
NT and CRS patient groups were performed using univariable logisticregression as well as two-tailed Wilcoxon rank sum test. Potentialconfounders were adjusted using multivariable logistic regression anal-ysis. Cumulative incidences curves stratified by cytokine levels wereestimated using the “cmprsk” package in R. Spearman's rank ordercorrelation was used to measure the strength of the association betweentwo variables. P values ≤0.05 were considered significant, unless oth-erwise indicated. Statistical analyses were carried out in R and Prism.
ResultsPatient characteristics
Seventy-five patients with LBCL who received lymphodepletingchemotherapy followed by infusion of axi-cel are included in thisstudy. Baseline patient characteristics are summarized inTable 1 (1, 5).The median follow up time was 11 months at the time of data cut off,September 30, 2019. Median age was 63 years (range 23–79 years).They received amedian of three prior lines of therapy (range 2–8), anda subset had relapsed or progressed after autologous stem cell trans-plant (n ¼ 11). Forty-eight patients (64%) required bridging therapybetween the time of leukapheresis and lymphodepleting chemother-apy. The majority of patients (n ¼ 26) received bridging chemother-apy. Ten patients received radiation as part of bridging therapy. Elevenpatients received high dose steroids alone. Thirty-six patients (48%)
would not have been eligible for enrollment in the ZUMA-1 trial basedupon disease or comorbidity factors at the time of apheresis. Factorsthat excluded patients from ZUMA-1 include renal insufficiency,decreased ejection fraction � 50%, platelet count � 75,000 cells/mL,symptomatic pleural effusions, chronic hepatitis B or C, prior historyof CNS disease, prior allogeneic transplant and venous thromboem-bolism within 6 months.
Clinical characteristics, response, and toxicity assessmentUnivariate (Supplementary Table S1) logistic regression was con-
ducted to test the association between baseline characteristics andsevere CRS and NT. None of the baseline factors analyzed weresignificantly associated with either severe CRS or NT. In a multivar-iable logistic regression model (Supplementary Table S2), no baselinepatient or tumor factors including age, ECOG at apheresis, stage, norbridging chemotherapy were associated with severe CRS or NT.
Sixty-eight patients were evaluated for response at day 90. Theoverall response rate at day 90 after CART-cell infusion was 53%, with29 (42.6%) patients in CR at day 90 (Table 2). Most patients developedCRS (n ¼ 72) with 16% developing severe (grade ≥3) CRS. Of thosewho had CRS, 31 patients (43%) had a fever within 24 hours of axi-celinfusion. Fifty patients (67%) developed NT with an incidence ofsevere NT of 31%. The median time to maximum toxicities was 4 and6 days respectively for CRS and NT. Forty-three patients (57%) weretreated with tocilizumab and 41 (55%) required steroids for themanagement of toxicities. Four patients (5%) died directly as a resultof, or in part attributable to, severe CAR T-mediated toxicities. Nopatients died as a result of severe NT. Nine patients (13%) died within90 days as a result of disease progression.
Laboratory measurements associated with toxicity in patientstreated with axi-cel
Univariate logistic regressionwas used to test the association of eachcytokine to toxicity for baseline, day zero, and peak. Baseline levels of
Table 1. Patient characteristics.
Patient characteristics N ¼ 75
Age - Median (Range) years 63 (23–79)Male sex, no. (%) 50 (67)Histology, no. (%)De novo DLBCL 50 (67)Transformed indolent lymphoma 25 (33)Bulky disease ≥10 cm, no. (%) 12 (16)Ann Arbor stage III/IV, no. (%) 62 (83)IPI ≥ 3 at apheresis, no. (%) 51 (69)Lines of therapy ≥3, no. (%) 46 (61)Bridging therapy, no. (%) 48 (64)
Chemotherapy 26 (54)Radiation � chemotherapy 10 (21)Steroids only 11 (23)Other 6 (12.5)
Prior autologous HSCT, no. (%) 11 (19)Not eligible for ZUMA-1, no. (%) 36 (48)
Renal insufficiency 6 (17)VTE within 6 months 5 (14)Thrombocytopenia 5 (14)Symptomatic effusion 3 (8)Other 17 (47)
Abbreviations: HSCT, hematopoietic stem cell transplantation; IPI, internationalprognostic index; VTE, venous thromboembolism.
Mechanisms of Toxicity in Patients Treated with Axi-cel
AACRJournals.org Clin Cancer Res; 2020 3
IL6 (P¼ 0.013), ANG2/ANG1 (P¼ 0.0056), ANG2 (P¼ 0.0190), andferritin (P¼ 0.0491) were associated with severe NT (Fig. 1A–D). Day0 levels of IL15 (P¼ 0.0407) and ANG2/ANG1 (P¼ 0.0248) were alsoassociated with severe NT. Peak levels of IL6 (0.0128), ANG2/ANG1(P ¼ 0.0016), IFNg (P ¼ 0.0064), and IL15 (0.0006) were associatedwith severe NT (Fig. 1E–H). Lower peak levels of ANG1 wereassociated with severe NT (Fig. 1I; P ¼ 0.028). Although it did notmeet statistical significance, there was a trend towards significant ofpeak levels of GM-CSF and severe NT as shown in Fig. 1J (P¼ 0.051).In a multivariate analysis (Supplementary Table S2), ANG2/ANG1was the only baseline factor or cytokine significant for severe NT (P¼0.0154).
Levels of IL6 (P ¼ 0.046) were the only baseline value associatedwith severe CRS (Fig. 2A). Serum peak levels of IL15 (P ¼ 0.0285),ANG2/ANG1 (P ¼ 0.0003), IFNg (P ¼ 0.0011), IL6 (P ¼ 0.0003),TNFa (P ¼ 0.0082), and GM-CSF (0.0172) were higher in patientswho had, or went on to develop, severe CRS (Fig. 2B–G). Usingspearman's correlation coefficient, baseline IL6 levels correlatedstrongly with day 0 IL6 levels (Fig. 4SA; r ¼ 0.71) and less so withpeak IL6 levels (Fig. 4SB; r ¼ 0.46). We recently identified a role forcatecholamine secretion by CAR T cells and myeloid cells in animalmodels of CRS and wanted to determine if serum catecholamine levelswere higher in patients with severe toxicity (21). Using univariatelogistic regression, higher peak levels of noradrenaline (NAD) wereassociated with grade ≥3 CRS (P ¼ 0.0234), but not with severe NT(Fig. 2H). Using patient case examples of grades 0, 1, 2, and 4 CRS, weobserve NAD levels increasing in patients with more severe toxicity(Supplementary Figs. S5A–S5D).
Baseline elevated levels of IL6 correlatewith severe CRS andNTAlthough levels of IL6 following CAR T-cell infusion have been
consistently associated with CRS and/or NT, there is limited data onthe association between baseline IL6 and toxicity (1, 3, 10, 27, 28). Theassociation between IL6 and severe immune mediated toxicity may be
mediated by myeloid cells (29, 30). In a multivariable model (Sup-plementary Table S2), IL6 was the only baseline value that wasassociated with severe CRS (P ¼ 0.03898). Patient characteristicscomparing the subset of patients with baseline elevated IL6 demon-strates that these patients all had stage III/IV disease and requiredbridging chemotherapy, which was not allowed in ZUMA-1 (Supple-mentary Table S3; refs. 1, 5). These patients appeared to have partic-ularly aggressive lymphoma. Median onset to CRS and NT was earlierthan for the overall cohort at 1 and 4 days, respectively. We observed89% of patients who had IL6 levels� 40 pg/mL died within 90 days ofCAR-T infusion with 56% developing grade 3 or higher CRS (Sup-plementary Table S4). Figure 3 demonstrates a cumulative incidenceof death in patients with baseline IL6 levels � 40 pg/mL versus thosewith baseline levels IL6 levels <40 pg/mL.
Gene expression of lymphoma tissue correlated with toxicitiesin patients treated with axi-cel
For the purpose of correlating toxicity with tumor intrinsic features,we evaluated gene expression profiles of baseline tumor tissue collectedwithin 1month of axi-cel infusion. Thirty-six patients had a limited setgene expression profiling (Nanostring) panel of 770 genes. We cor-related severity of NT and CRS to lymphoma gene expression todetermine what tumor cellular pathways may impact outcomes. Todetermine the association between the TME and toxicities, cell typescores were calculated for T cell and macrophages. Patients whoexperienced severe NT had lower T-cell type score (P < 0.001) andhigher macrophage score (P < 0.01; Fig. 4A). Furthermore, those withsevere NT had less expression of the regulatory T (Treg) marker Foxp3(P¼ 0.00001;Fig. 4B).We found that patients with severeNT (n¼ 10)had a different gene expression profile than patients with less severeNT (n ¼ 26; Fig. 4C). An overlapping set of baseline biopsies hadRNA-SEQ done and were also analyzed for correlation to toxicity. Avolcano plot (Fig. 4D) demonstrates the differentially expressed genesin the cohort who developed grade ≥3 NT (n ¼ 11) versus those whohad grade 0 to 2 toxicities (n¼ 14). GSEA plots (Fig. 4E) are positivelycorrelated with genes differentially expressed between conventionaland regulatory T cells in patients with severe NT. Severe NT alsopositively correlated with classically activated macrophages (M1) andalternatively activated macrophages (M2; Fig. 4F). Although thenumber of severe CRS cases was smaller than NT, there were stilldifferences in gene expression in the subset of patients with severe ≥ 3CRS (n ¼ 5; Supplementary Fig. S6).
Elevated baseline IL6 in a patientWe describe the management of the only patient in our cohort
with a baseline IL6 level � 40 pg/mL who remained alive at time ofdata cut off. This patient had a history of double hit diffuse large Bcell lymphoma (DLBCL) with primary refractory disease followingthree lines of therapy. Prior to axi-cel infusion, he requiredbridging chemotherapy due to symptomatic bulky disease (Sup-plementary Fig. S7A). On the day following infusion of axi-cel, hedeveloped grade 2 CRS with fevers and hypotension requiringtocilizumab. In addition to tocilizumab, he was treated with onedose of dexamethasone due to concern for severe toxicities basedon bulky disease. Despite improvement on day two with resolutionof fevers, he had recurrence of symptoms by the third day.Considering the patient had high risk features of toxicity includingbulky disease, elevated baseline level of IL6 as well as resistance totocilizumab and dexamethasone, he was treated with high dosemethylprednisolone despite being classified as Grade 2 CRS perASTCT guidelines.
Table 2. Clinical endpoints.
N ¼ 75
CRSMedian time to CRS 2 daysMedian time to max CRS 4 daysCRS all grades, no. (%) 72 (96)Grade ≥3 CRS, no. (%) 12 (16)Grade 5 CRS, no. (%) 3 (4)Use of tocilizumab, no. (%) 43 (57)Use of steroids, no. (%) 41 (55)
NeurotoxicityMedian time to NT 5 daysMedian time to max NT 6 daysNT all grades, no. (%) 50 (67)Grade ≥3 NT, no. (%) 23 (31)
D90 response (N ¼ 68)CR þ PR, no. (%) 36 (53)Complete response, no. (%) 29 (43)NRM, no. (%) 4 (6)Disease related mortality, no. (%) 9 (13)
Note: CRS and NT were graded prospectively. CRS was defined and gradedusing the ASTCT grading guidelines (4). Neurologic toxicity was graded usingthe CAR T-cell-therapy-associated (CARTOX) working group guidelines (6).Tumor response was determined by the treating physician per Lugano 2014classification (20).Abbreviation: NRM, nonrelapse mortality.
Faramand et al.
Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH4
Per institutional andCARTOXworking group guidelines, high dosemethylprednisolone is indicated for patients with grade 4 CRS (6).However, in this case, early and aggressive management of CRS likelycontributed to resolution of CRS to grade 0 without impacting efficacyas patient had partial response on day 30 scans (SupplementaryFig. S7B). Real time cytokine data were not available for the remainderof patients included in this study and therefore did not impact thediagnosis or management of immune-mediated toxicities. Althoughthis is a single case, it demonstrates that real time cytokine monitoringcan be a helpful tool alongwith clinical and standard laboratory data inguiding management.
DiscussionWe report the role of cytokine analysis in identifying patients with
aggressive B-cell lymphoma who are at risk of developing severeimmune-mediated toxicity after treatment with standard of care
axi-cel. Using gene expression of the lymphoma TME, we identifiedimmune cells that correlate with severe toxicity.We observed that bothmyeloid cells and regulatory T cells correlated with severity of CRS andNT. Further, we report on the role of catecholamines in mediatingCAR T-cell therapy-related toxicities (21). It is important to note thatmany patients included in this study would not have been eligible fortreatment in the pivotal ZUMA-1 study and reflect a real-worldpopulation. Despite patients having more aggressive disease, the ratesof severe grade ≥3 CRS and NT were comparable to those reported inZUMA-1, as has been reported elsewhere (1, 5).
Several biomarkers have been associated with CRS and/or NT andmost recently reported to impact efficacy (1, 12, 13, 17). Althoughcytokines are elevated after infusion of CAR T cells, there is novalidated cut off of a single biomarker that is predictive of severetoxicities. The utility of cytokines and lab values such as CRP varydepending on several factors including which CARTproduct was usedand underlying disease (10, 13). Comparing toxicities across the
Figure 1.
Serummarkers associatedwith severe NT. Patients receiving standard of care CD19 CART-cell therapy (axicabtagene ciloleucel) for LBCL had prospective collectionof serum formeasurement of cytokine levels using apoint-of-care device (Ella Simple Plex - Protein Simple).A,Baseline (n¼ 52) levels of IL6 indicates samples drawnprior the start of lymphodepleting chemotherapy.B andC,Baseline (n¼ 51) levels of ANG1 andANG2/ANG1 drawnprior the start of lymphodepleting chemotherapy.Cytokinesweremeasuring usingElla Simple Plex - Protein Simple as described inMaterials andMethods.D, Baseline levels of ferritin (n¼69) collected prior to start oflymphodepleting chemotherapy. Ferritin (ng/dL) measuring using standard lab equipment (Roche Cobas). E–I, Max (n ¼ 75) indicates the highest level of thecytokine measured during daily sampling in the first 30 days after CAR T therapy fusion. P value calculated using univariate logistic regression. Logarithmtransformation is used (A–J) to allow visualization of data points on the graphs.
Mechanisms of Toxicity in Patients Treated with Axi-cel
AACRJournals.org Clin Cancer Res; 2020 5
pivotal CAR T trials is challenging as there is variability in the studiesamong the CRS grading systems. In this report, the majority oftoxicities were graded according to the most recent ASTCT consensusgrading to provide uniform grading that allows for easier comparisonsof toxicities across various CAR T constructs and clinical trials (4).
Consistent with prior published reports, we demonstrate thatbiomarkers of endothelial activation, ANG1 and ANG2, play a rolein the development of severe CRS and NT (13, 15). ANG1 and ANG2are ligands that have opposing functions on endothelial activationwithANG2 promoting capillary leak and ANG1 promoting endothelialstability (31). Levels of ANG2 and ANG 2/ANG1 ratio were highereven prior to the start of lymphodepleting chemotherapy in patientswho subsequently developed grade ≥3 NT. Furthermore, we observethat peak levels of cytokines which activate endothelial cells includingIL6, IFNg, TNFa in addition to ANG2/ANG1 are elevated in patientswho develop severe grade � 3 NT. These findings support a modelproposed by Gust and colleagues, whereby CAR T cells initiate acascade of proinflammatory and endothelial activating cytokines inpatients with baseline tumor endothelial dysfunction leading to aforward loop of further endothelial activation and subsequent severeNT (12, 13). Although CRS and NT are overlapping syndromes, theyhave distinct features and therefore do not respond to the same clinicalmanagement schemas. Because prophylactic IL6 receptor blockade hasnot been shown to be effective in ameliorating neurotoxicity, clinicaltrials aimed at modulating endothelial activation to prevent severe NTare warranted (1, 5, 32, 33).
We report that recipient myeloid cells and Tregs may play animportant role in the pathogenesis of severe NT and CRS, and proposeamodel in which severity of CART toxicity is driven by the interactionbetween infused CART cells and these recipient cells (8, 9). The role ofrecipient Tregs and immune-mediated toxicities and clinical responsein patients with LBCL treated with CAR T is not yet well defined.Reducing systemic inflammation and polarizing the TME towards a T-cell infiltration phenotype are potential strategies to lower toxicity.However, the impact of Tregs in the TME on efficacy is not yet fullyelucidated and requires further evaluation.
Figure 2.
Serummarkers associatedwith severe CRS.A, Baseline levels of IL6 (n¼ 52) associate with severe CRS.B–G,Max (n¼ 75) indicates the highest level of the cytokinemeasured during daily sampling in the first 30 days after CAR T therapy fusion. P value calculated using univariate logistic regression. H, Samples (n ¼ 30) wereanalyzed for catecholamine levels (noradrenaline, NAD) as described in ref. 21. Logarithm transformation is used (A–H) to allow visualization of data points on thegraphs.
Figure 3.
Cumulative incidence of events stratified by IL6 levels. Events were defined asdeath from progressive disease or death from toxicity and/or infection afterinfusion of axi-cel. Patient samples were collected prior to lymphodepletingchemotherapy and stratified by baseline IL6 level � 40 or <40 pg/mL.
Faramand et al.
Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH6
We describe a novel observation between peak levels of NAD andsevere CRS. In preclinical models, we previously reported that T-cell–activated macrophages secrete high levels of catecholamines thatenhance the inflammatory response through a positive feedbackloop (21). Metyrosine, a tyrosine hydroxylase inhibitor, is a catechol-amine inhibitor that is used to treat patients with pheochromocytoma.Pretreatment of mice with Metyrosine reduced the production ofcatecholamines by macrophages (21). Cosentino and colleagues haveshown that catecholamine synthesis is effectively induced through aprotein kinase C-dependent mechanism that triggers the de novoexpression of tyrosine hydroxylase and subsequent activation in Tand B lymphocytes (34). Furthermore, catecholamines surges can beseen in patients who develop immune-mediated toxicities, resulting ina hyperactivated immune system. Our study is limited by the smallsample size of patients who had serum available for catecholamineanalysis (n ¼ 30) and therefore warrants validation of these observa-tions in larger studies.
Although we and others have described the role of cytokines inassociation with CRS and NT, there is no defined cytokine level that
can predict which patients are at highest risk of developing severetoxicity. During the course of the pivotal axi-cel trials, there was achange in management from delaying intervention until toxicitiesbecame severe (grade� 3) to earlier intervention aimed at preventingtoxicities from becoming severe by providing early cytokine blockadetherapy (1, 5). Despite this change to earlier intervention, somepatients reported in this cohort still died of or with severe CRS,suggesting that there may be disease factors in these patients thatpromote severe toxicity. We identified elevated IL6 levels (≥40 pg/mL)to be associated with severe toxicity and death in patients treated withaxi-cel therapy. Themajority of patients (n¼ 5) died of disease-relatedfactors suggesting that elevated IL6 may be an indicator of refractorydisease. However, three patients died of treatment-related mortality,suggesting a different mechanism for the source of inflammation.Future studies evaluating the role of IL6 and baseline myeloid infil-tration in the TME in a larger patient cohort are planned. Cytokineswere evaluated in a research lab with the Ella automated simple plexassay (ProteinSimple) as described in the Materials and Methods.Future studies using cytokine cut off values to guide clinical
Figure 4.
Gene expression of lymphoma tissue correlatedwith toxicities in patient treatedwith axi-cel.A–C. Baseline tumor biopsieswere takenwithin 1 month prior to axi-celinfusion (n¼ 36). 10 patients experienced severe grade 3-4NTand26patients experienced non-severeNT.A,Cell type score for T cells andmacrophages.B,Examplegenes include CD3g (T cell marker) and FOXP3 (Tregmarker). C,Gene expressionwasmeasuredwith the Nanostring IO360 panel, which include 770 genes, andwasanalyzed by nSolver. Heatmap of T cell genes. Each column represents a patient sample and genes as shown as Z transformed expressionD. RNAseqwas performedonanoverlapping set of baseline biopsies thatwas prospectively snap frozen (n¼ 11 gr.3-4,n¼ 14 gr.0-2). Volcanoplot for differentially expressedgenes based onNTseverity. E and F, Enrichment plots of genes analyzed for immunologic GSEA signatures including resting Tregs, macrophages, M1 andM2macrophages respectively.
Mechanisms of Toxicity in Patients Treated with Axi-cel
AACRJournals.org Clin Cancer Res; 2020 7
management would require Clinical Laboratory ImprovementAmendments (CLIA) laboratory validation. Although these findingswere observed in a small number of patients and will need to bevalidated in a larger cohort, they provide the rationale for designingrisk-adapted clinical trials to mitigate toxicity with prophylacticagents. An analysis is planned to create and prospectively validate apredictive model to risk stratify patients and guide management ofsevere toxicities. Furthermore, due to the financial toxicity of CAR T,future predictive models incorporating baseline cytokines will informwhich patients can be safely treated in the outpatient setting.
Because of the retrospective nature of this study, we are limited todescribing observations of significant associations of baseline tumorand patient factors with toxicity. Our observations suggest that theincidence of severe life-threatening CRS following CD19 CAR T-celltherapy is influenced by baseline tumor characteristics that are presentprior to the infusion of CAR T cells. On the basis of these findings wepropose a model in which inflammation characterized by elevatedcytokine levels, specifically IL6, primes myeloid cells that are furtheractivated upon CAR T cell infusion to release toxic amounts ofcytokines and catecholamines leading to severe toxicity. We couldnot find a significant correlation between baseline IL6 and myeloidgene expression, however only 10 patients had baseline serum cyto-kines in addition to tissue biopsy samples limiting this analysis.Cytokines and other potential predictive markers of toxicity will needto be evaluated in a larger prospective cohort.
Our understanding of the impact of myeloid cells, a proinflam-matory TME, and endothelial dysfunction on mediating toxicitiesand outcomes is evolving and warrants further investigation in aclinical trial setting. The identification of mechanisms of immune-mediated toxicity and recognition of early predictive markers ofsafety are important for our understanding of these toxicities andrecognition of CRS and NT as separate entities. We have confirmedthat patients’ cytokine profile at baseline as related to the TME playsa key role in toxicity and provides insight into a novel mechanism oftoxicity involving myeloid cells and catecholamines. Emergingbiologic data such as reported here is vital for designing clinicaltrials aimed at further reducing the rates of severe toxicities in high-risk patients.
Disclosure of Potential Conflicts of InterestM. Jain reports personal fees from Kite/Gilead and Novartis outside the submitted
work. V. Staedtke reports grants from NCI 5K08CA230179, Sontag DistinguishedScientist Award, and Emerson Research during the conduct of the study; personal feesfrom Gilbert Family Foundation - Gene Therapy Initiative (scientific board member)outside the submitted work; and has a patent for JHU Reference: C14930 – ‘Drugs toreduce cytokine storms (sepsis) following application of biotherapeutic agents’ issued(The terms of these arrangements are managed by the Johns Hopkins University inaccordance with its conflict of interest policy.). R. Bai reports grants from DODW81XWH1810236 New Investigator Award during the conduct of the study; has apatent for JHU Reference: C14930 – ‘Drugs to reduce cytokine storms (sepsis)following application of biotherapeutic agents’; issued (The terms of these arrange-ments aremanaged by the Johns Hopkins University in accordance with its conflict ofinterest policy.). J. Pinilla reports personal fees from Janssen, Abbvie, Pharmacyclics,Astra Zeneca, Takeda, and Sanofi outside the submitted work. A. Lazaryan reportsother from EUSA Pharma LLC (consultancy) and Gilead (equities) outside thesubmitted work. B. Shah reports grants and personal fees from Kite/Gilead (advisory,grant), Novartis (advisory), BMS/Celgene/Juno (advisory), and Precision Biosciences(advisory) during the conduct of the study; personal fees from Adaptive (advisory),Pharmacyclics/Jansen (advisory), Astra Zeneca (advisory), Beigene (advisory),Amgen (advisory), and Pfizer (advisory), and grants from Jazz (Invest Init Trial),Servier (Invest Init Trial), and Incyte (grant) outside the submitted work. J.C. Chavezreports personal fees from Kite/Giled, Genentech, Morphosys, Novartis, Bayer,personal fees from Astrazeneca, Beigene, and Celgene outside the submitted work.T. Nishihori reports nonfinancial support from Karyopharm (research support to the
institution (drug supply for clinical trial)) and other from Novartis (research supportto the institution (clinical trial)) outside the submitted work. J. Mullinax reports otherfrom Iovance Biotherapeutics (research support) outside the submitted work.M. Hussaini reports personal fees from Stemline Therapeutics (Adboard for Stemlineon BPCDN), Boston Biomedical (consultancy for tissue processing), AdaptiveBiotechnologies (Adboard, Speaker's Bureau; consultant. Lymphoid MRD testing),Amgen (AdBoard; Consultant for MRD), Guidepoint (phone interviews regardingCancer Diagnostics/Pathology), and Decibio (phone interviews regarding CancerDiagnostics/Pathology) outside the submitted work. M. Dam reports personal feesfrom Archbow Consulting Group (logistics and process of cell therapy) outside thesubmitted work. C.A. Bachmeier reports personal fees from Kite Pharma, Novartis,and Legend Biotech outside the submitted work. F.L. Locke reports personal fees fromNovartis (compensated scientific advisor), Allogene (compensated scientific advisor),Calibr (compensated scientific advisor),Wugen (compensated scientific advisor), andGammaDelta Therapeutics (compensated scientific advisor) during the conduct ofthe study; grants, nonfinancial support, and other fromKite/Gilead (noncompensatedscientific advisor; research support not related to this manuscript; writing assistancenot related to this manuscript; travel and lodging), personal fees and other fromBMS/Celgene (compensated scientific advisor including travel and lodging), CellularBiomedicine Group Inc. (consultant; travel and lodging; grant options (expired andnot exercised)), Novartis (compensated scientific advisor), Allogene (compensatedscientific advisor), Calibr (compensated scientific advisor), Wugen (compensatedscientific advisor), and Gamma Delta Therapeutics (compensated scientific advisor)outside the submitted work; and has a patent for CAR T Cells with EnhancedMetabolic Fitness (62/939,727) pending, a patent for Methods of Enhancing CAR TCell Therapies (62/892,292) pending, and a patent for Evolutionary Dynamics ofNon-Hodgkin Lymphoma CAR-T cell therapy (62/879,534) pending. M.L. Davilareports grants and other from Atara (Licensing fees), personal fees from Novartis,Kite/Gilead, Celyad, and PACT Pharma, and other from Precision Biosciences(stock), Adaptive Biotechnologies (stock), Bellicum (stock options), and AnixaBiosciences (stock options) outside the submitted work; and a patent for cytokinemonitoring pending.
Authors’ ContributionsR. Faramand: Conceptualization, data curation, formal analysis, investigation,
methodology, writing-original draft, writing-review and editing. M. Jain:Conceptualization, data curation, software, formal analysis, investigation,methodology, writing-original draft, writing-review and editing. V. Staedtke:Conceptualization, data curation, methodology and writing-original draft.H. Kotani: Data curation, investigation, methodology, writing-original draft,writing-review and editing. R. Bai: Methodology, writing-review and editing.K. Reid: Data curation, investigation, methodology, writing-review and editing.S.B. Lee: Data curation, investigation, methodology, writing-original draft,writing-review and editing. K. Spitler: Investigation, methodology, writing-reviewand editing.X.Wang: Software, formal analysis, methodology, writing-original draft,writing-review and editing. B. Cao: Formal analysis, writing-original draft, writing-review and editing. J. Pinilla: Writing-review and editing. A. Lazaryan: Writing-review and editing. F. Khimani:Writing-review and editing.B. Shah:Writing-reviewand editing. J.C. Chavez: Writing-review and editing. T. Nishihori: Writing-reviewand editing. A. Mishra: Writing-review and editing. J. Mullinax: Data curation,writing-review and editing. R. Gonzalez: Data curation, writing-review and editing.M. Hussaini: Data curation, writing-review and editing. M. Dam: Data curation,writing-review and editing.B.D. Brandjes:Data curation, writing-review and editing.C.A. Bachmeier: Data curation, investigation, methodology, writing-review andediting. C. Anasetti: Conceptualization, resources, formal analysis, supervision,writing-review and editing. F.L. Locke: Conceptualization, data curation, formalanalysis, supervision, investigation, methodology, writing-original draft, writing-review and editing. M.L. Davila: Conceptualization, resources, data curation,formal analysis, supervision, investigation, visualization, methodology, writing-original draft, project administration, writing-review and editing. No potentialconflicts of interest were disclosed by the other authors.
AcknowledgmentsThis work has been supported in part by the Molecular Genomics and
Biostatistics Core Facilities at the H. Lee Moffitt Cancer Center & ResearchInstitute, an NCI designated Comprehensive Cancer Center (P30-CA076292).A patent application on CRS prevention listing V. Staedtke and R. Bai as co-inventors has been provisionally filed by Johns Hopkins University. The terms ofthese arrangements are managed by Johns Hopkins University in accordance withits conflict of interest policies. We are grateful to the contributions from the
Faramand et al.
Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH8
clinical staff and clinical research staff, as well as the patients and their families forparticipating in this clinical study.
The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked
advertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.
Received April 21, 2020; revised June 14, 2020; accepted July 13, 2020;published first July 15, 2020.
References1. Neelapu SS, Locke FL, Bartlett NL, Lekakis LJ, Miklos DB, Jacobson CA, et al.
Axicabtagene ciloleucel CART-cell therapy in refractory large B-cell lymphoma.N Engl J Med 2017;377:2531–44.
2. Schuster SJ, Bishop MR, Tam CS, Waller EK, Borchmann P, McGuirk JP, et al.Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma.N Engl J Med 2019;380:45–56.
3. Davila ML, Riviere I, Wang X, Bartido S, Park J, Curran K, et al. Efficacy andtoxicity management of 19–28z CAR T cell therapy in B cell acute lymphoblasticleukemia. Sci Transl Med 2014;6:224ra25.
4. Lee DW, Santomasso BD, Locke FL, Ghobadi A, Turtle CJ, Brudno JN, et al.ASBMT consensus grading for cytokine release syndrome and neurologictoxicity associated with immune effector cells. Biol Blood Marrow Transplant2019;25:625–38.
5. Locke FL, Ghobadi A, Jacobson CA, Miklos DB, Lekakis LJ, Oluwole OO, et al.Long-term safety and activity of axicabtagene ciloleucel in refractory large B-celllymphoma (ZUMA-1): a single-arm, multicentre, phase 1–2 trial. Lancet Oncol2019;20:31–42.
6. Neelapu SS, Tummala S, Kebriaei P, Wierda W, Gutierrez C, Locke FL, et al.Chimeric antigen receptor T-cell therapy - assessment and management oftoxicities. Nat Rev Clin Oncol 2018;15:47–62.
7. Alvi RM, Frigault MJ, Fradley MG, Jain MD, Mahmood SS, Awadalla M, et al.Cardiovascular events among adults treated with chimeric antigen receptorT-cells (CAR-T). J Am Coll Cardiol 2019;74:3099–108.
8. Giavridis T, van der Stegen SJC, Eyquem J, Hamieh M, Piersigilli A, Sadelain M.CAR T cell-induced cytokine release syndrome is mediated bymacrophages andabated by IL-1 blockade. Nat Med 2018;24:731–8.
9. Norelli M, Camisa B, Barbiera G, Falcone L, Purevdorj A, Genua M, et al.Monocyte-derived IL-1 and IL-6 are differentially required for cytokine-releasesyndrome and neurotoxicity due to CAR T cells. Nat Med 2018;24:739–48.
10. Teachey DT, Lacey SF, Shaw PA, Melenhorst JJ, Maude SL, Frey N, et al.Identification of predictive biomarkers for cytokine release syndrome afterchimeric antigen receptor T-cell therapy for acute lymphoblastic leukemia.Cancer Discov 2016;6:664–79.
11. Pennell CA, Barnum JL, McDonald-Hyman CS, Panoskaltsis-Mortari A, RiddleMJ, Xiong Z, et al. Human CD19-targetedmouse T cells induce B cell aplasia andtoxicity in human CD19 transgenic mice. Mol Ther 2018;26:1423–34.
12. Gust J,HayKA,HanafiLA, LiD,MyersonD,Gonzalez-Cuyar LF, et al. Endothelialactivation and blood-brain barrier disruption in neurotoxicity after adoptiveimmunotherapy with CD19 CAR-T cells. Cancer Discov 2017;7:1404–19.
13. Hay KA, Hanafi LA, Li D, Gust J, Liles WC, Wurfel MM, et al. Kinetics andbiomarkers of severe cytokine release syndrome after CD19 chimeric antigenreceptor-modified T-cell therapy. Blood 2017;130:2295–306.
14. Hirayama AV, Gauthier J, Hay KA, Voutsinas JM, Wu Q, Gooley T, et al. Theresponse to lymphodepletion impacts PFS in patients with aggressive non-Hodgkin lymphoma treated with CD19 CAR T cells. Blood 2019;133:1876–87.
15. Santomasso BD, Park JH, SalloumD,Rivi�ere I, Flynn J,Mead E, et al. Clinical andbiologic correlates of neurotoxicity associated with CAR T cell therapy inpatients with B-cell acute lymphoblastic leukemia (B-ALL). Cancer Discov2018;8:958–71.
16. Locke FL, Neelapu SS, Bartlett NL, Siddiqi T, Chavez JC, Hosing CM, et al. Phase1 results of ZUMA-1: a multicenter study of KTE-C19 anti-CD19 CAR T celltherapy in refractory aggressive lymphoma. Mol Ther 2017;25:285–95.
17. Kochenderfer JN, Somerville RPT, Lu T, Shi V, Bot A, Rossi J, et al. Lymphomaremissions caused by anti-CD19 chimeric antigen receptor T cells are associatedwith high serum interleukin-15 levels. J Clin Oncol 2017;35:1803–13.
18. Sterner RM, Sakemura R, CoxMJ, Yang N, Khadka RH, Forsman CL, et al. GM-CSF inhibition reduces cytokine release syndrome and neuroinflammation butenhances CAR-T cell function in xenografts. Blood 2019;133:697–709.
19. Park JH, Rivi�ere I, Gonen M, Wang X, S�en�echal B, Curran KJ, et al. Long-termfollow-up of CD19 CAR therapy in acute lymphoblastic leukemia. N Engl J Med2018;378:449–59.
20. Cheson BD, Fisher RI, Barrington SF, Cavalli F, Schwartz LH, Zucca E, et al.Recommendations for initial evaluation, staging, and response assessment ofHodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol2014;32:3059–68.
21. Staedtke V, Bai RY, KimK, DarvasM, DavilaML, Riggins GJ, et al. Disruption ofa self-amplifying catecholamine loop reduces cytokine release syndrome. Nature2018;564:273–7.
22. Bushnell B, Rood J, Singer E. BBMerge - accurate paired shotgun read mergingvia overlap. PLoS One 2017;12:e0185056.
23. MartinM. Cutadapt removes adapter sequences fromhigh-throughput sequenc-ing reads. 2011 2011;17:3.
24. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR:ultrafast universal RNA-seq aligner. Bioinformatics 2013;29:15–21.
25. Anders S, Pyl PT, Huber W. HTSeq–a Python framework to work with high-throughput sequencing data. Bioinformatics 2015;31:166–9.
26. Love MI, Huber W, Anders S. Moderated estimation of fold change anddispersion for RNA-seq data with DESeq2. Genome Biol 2014;15:550.
27. Brentjens RJ, Davila ML, Riviere I, Park J, Wang X, Cowell LG, et al. CD19-targeted T cells rapidly induce molecular remissions in adults with chemo-therapy-refractory acute lymphoblastic leukemia. Sci Transl Med 2013;5:177ra38.
28. Porter DL, Levine BL, Kalos M, Bagg A, June CH. Chimeric antigen receptor-modified T cells in chronic lymphoid leukemia. N Engl J Med 2011;365:725–33.
29. Azzaoui I, Uhel F, RossilleD, PangaultC,Dulong J, Le Priol J, et al. T-cell defect indiffuse large B-cell lymphomas involves expansion of myeloid-derived suppres-sor cells. Blood 2016;128:1081–92.
30. De Vlaeminck Y, Gonz�alez-Rasc�on A, Goyvaerts C, Breckpot K. Cancer-associated myeloid regulatory cells. Front Immunol 2016;7:113.
31. Fiedler U, Scharpfenecker M, Koidl S, Hegen A, Grunow V, Schmidt JM, et al.The Tie-2 ligand angiopoietin-2 is stored in and rapidly released upon stimu-lation from endothelial cell weibel-palade bodies. Blood 2004;103:4150–6.
32. Locke FL, Neelapu SS, Bartlett NL, Lekakis LJ, Jacobson CA, Braunschweig I,et al. Preliminary results of prophylactic tocilizumab after axicabtageneciloleucel(axi-cel; KTE-C19) treatment for patients with refractory, aggressive non-hodgkin lymphoma (NHL). Blood 2017;130:1547.
33. Topp M, Van Meerten T, Houot R, Minnema MC, Milpied N, Lugtenburg PJ,et al. Earlier steroid use with axicabtagene ciloleucel (Axi-Cel) in patients withrelapsed/refractory large B cell lymphoma. American Society of Hematology,Washington, DC; 2019.
34. Cosentino M, Fietta AM, Ferrari M, Rasini E, Bombelli R, Carcano E, et al.HumanCD4þCD25þ regulatory T cells selectively express tyrosine hydroxylaseand contain endogenous catecholamines subserving an autocrine/paracrineinhibitory functional loop. Blood 2007;109:632–42.
AACRJournals.org Clin Cancer Res; 2020 9
Mechanisms of Toxicity in Patients Treated with Axi-cel
![[RTF]goaprintingpress.gov.ingoaprintingpress.gov.in/downloads/7374/7374-13-SIII-EOG...Keni Brothers -Ganga INiwas -General Souza Bran, dao Road, Panaji. Karmali Virjee -Afonso de Albuquerque](https://static.fdocuments.in/doc/165x107/5aa857b27f8b9a86188b6f51/rtf-brothers-ganga-iniwas-general-souza-bran-dao-road-panaji-karmali-virjee.jpg)
