Translational Cancer Mechanisms and Therapy

Lobular Carcinomas In Situ Display IntralesionGenetic Heterogeneity and Clonal Evolution in theProgression to Invasive Lobular CarcinomaJu Youn Lee1, Michail Schizas2, Felipe C. Geyer1, Pier Selenica1, Salvatore Piscuoglio1,3,Rita A. Sakr2, Charlotte K.Y. Ng1,3,4, Jose V. Scarpa Carniello2, Russell Towers2, Dilip D. Giri1,Victor P. de Andrade1, Anastasios D. Papanastasiou1, Agnes Viale5, Reuben S. Harris6,David B. Solit5,7, Britta Weigelt1, Jorge S. Reis-Filho1,7, and Tari A. King2

Abstract

Purpose: Lobular carcinoma in situ (LCIS) is a preinvasivelesionof thebreast.We sought todefine its genomic landscape,whether intralesion genetic heterogeneity is present in LCIS,and the clonal relatedness between LCIS and invasive breastcancers.

Experimental Design: We reanalyzed whole-exomesequencing (WES) data and performed a targeted ampliconsequencing validation of mutations identified in 43 LCIS and27 synchronous more clinically advanced lesions from 24patients [9 ductal carcinomas in situ (DCIS), 13 invasivelobular carcinomas (ILC), and 5 invasive ductal carcinomas(IDC)]. Somatic genetic alterations, mutational signatures,clonal composition, and phylogenetic treeswere defined usingvalidated computational methods.

Results:WES of 43 LCIS lesions revealed a genomic profilesimilar to that previously reported for ILCs, with CDH1

mutations present in 81% of the lesions. Forty-two percent(18/43) of LCIS were found to be clonally related to synchro-nous DCIS and/or ILCs, with clonal evolutionary patternsindicative of clonal selection and/or parallel/branched pro-gression. Intralesion genetic heterogeneity was higher amongLCIS clonally related to DCIS/ILC than in those nonclonallyrelated to DCIS/ILC. A shift from aging to APOBEC-relatedmutational processes was observed in the progression fromLCIS to DCIS and/or ILC in a subset of cases.

Conclusions: Our findings support the contention thatLCIS has a repertoire of somatic genetic alterations similar tothat of ILCs, and likely constitutes a nonobligate precursor ofbreast cancer. Intralesion genetic heterogeneity is observed inLCIS and should be considered in studies aiming to developbiomarkers of progression from LCIS to more advancedlesions.

IntroductionLobular carcinoma in situ (LCIS) is a preinvasive lesion of the

breast, which is often multifocal and bilateral (1). Over the last

three decades, LCIS has been clinically perceived as a risk indicatorand managed accordingly (1). There is, however, burgeoningphenotypic and genetic evidence to suggest that LCIS is a non-obligate precursor of invasive breast cancer, akin to ductal carci-noma in situ (DCIS; ref. 2).

LCIS and invasive lobular carcinomas (ILC) are phenotypi-cally and genetically similar. Both lesions are preferentially ofthe luminal-A molecular subtype [i.e., estrogen receptor (ER)-positive, HER2-negative, low-grade, and low-proliferation] andharbor recurrent gains of 1q and losses of 16q, encompassingthe CDH1 gene locus, as well as recurrent CDH1 somaticmutations (1, 3–7). In fact, loss of E-cadherin, the proteinproduct of the CDH1 gene, is a hallmark feature of these lesions(3, 6) and has been shown to result in the development of ILCsin conditional mouse models (8). Analyses of the genomicfeatures of ILCs by The Cancer Genome Atlas (TCGA) consor-tium (6) and individual investigators (9) have revealed thegenes most commonly mutated in this subtype of breast cancerand identified molecular differences between invasive ductalcarcinomas (IDC) of no special type and ILCs, including ahigher rate of FOXA1 mutations and a lower rate of GATA3mutations in those with lobular histology. Additional whole-exome sequencing (WES; ref. 7) and targeted (10) sequencinganalyses focused on paired LCIS and ILCs demonstrated com-parable rates of mutations affecting CDH1, PIK3CA, and CBFB,among other genes.

1Department of Pathology, Memorial Sloan Kettering Cancer Center, New York,New York. 2Department of Surgery, Memorial Sloan Kettering Cancer Center,New York, New York. 3Institute of Pathology and Medical Genetics, UniversityHospital Basel, Basel, Switzerland. 4Department of Biomedicine, University ofBasel, Basel, Switzerland. 5Center for Molecular Oncology, Memorial SloanKettering Cancer Center, New York, New York. 6Howard Hughes MedicalInstitute, Masonic Cancer Center, Department of Biochemistry, Molecular Biol-ogy and Biophysics, University of Minnesota, Minneapolis, Minnesota. 7HumanOncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center,New York, New York.

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

J.Y. Lee, M. Schizas, F.C. Geyer, and P. Selenica contributed equally to this article.

B. Weigelt, J.S. Reis-Filho, and T.A. King jointly directed the work.

Corresponding Authors: Jorge S. Reis-Filho, Memorial Sloan Kettering CancerCenter, 1275 York Avenue, New York, NY 10065. Phone: 212-639-8054; E-mail:reisfilj@mskcc.org; and Tari A. King, Dana-Farber Cancer Institute, Brigham andWomen's Hospital, Harvard Medical School, Boston, MA 02215. Phone: 617-632-3891; E-mail: tking7@bwh.harvard.edu

doi: 10.1158/1078-0432.CCR-18-1103

�2018 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 25(2) January 15, 2019674

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Previous studies havedemonstrated that synchronous LCIS andinvasive breast cancers may be clonally related and share acommon ancestral lesion (4, 7, 10). In most studies, however,clonal relatedness was inferred using limited genomic informa-tion derived from copy-number (4) or targeted sequencing anal-yses (10). By combining copy-number and WES data, Begg andcolleagues provided evidence of clonal relatedness between LCISand associated lesions (7). These studies, however, did not inves-tigate the basis of the clonal relatedness between LCIS and ILC,and whether the progression from LCIS to ILC would involve theselection of specific subclones or happen through multiclonalinvasion (11, 12). Given that not only invasive breast cancers (13)but also preinvasive lesions (11) may be genetically heteroge-neous at diagnosis, and that tumor progression/stromal invasionmay stem from clonal selection (11, 13), it is plausible that LCISmay display intralesion genetic heterogeneity and that the pro-gression from LCIS to more clinically advanced lesions, such asDCIS or invasive breast cancer, may result from the selection ofpre-existing subclones.

Here, we performed a reanalysis of WES data generated froma unique series of frozen LCIS samples from prospectivelyaccrued, consecutive patients subjected to prophylactic or ther-apeutic mastectomy, previously published by Begg and collea-gues (7). We performed a high-depth targeted capture sequenc-ing validation of the mutations identified by WES in that study,using the same DNA samples, and employed state-of-the-artbioinformatics algorithms with a Bayesian clustering model(PyClone) to infer subclone structure and with construction ofclone-based phylogeny, seeking to define the clonal composi-tion and mutational processes in LCIS synchronously diag-nosed with ILC, DCIS, and/or IDC, and to ascertain whetherchanges in the clonal composition are observed in the progres-sion from LCIS to DCIS or ILC.

Materials and MethodsSubjects and samples

This study is basedon the 24 caseswith availableWESout of the30 cases previously subjected to microarray-based comparativegenomic hybridization and/or WES by Begg and colleagues (7).Eight out of the 24 cases included in this study were also included

in the targeted sequencing analysis previously reported by Sakrand colleagues (10). The cases subjected to WES include 43 LCISand 27 synchronous more clinically advanced lesions (Table 1;Supplementary Methods).

ImmunohistochemistryImmunohistochemistry for ER, progesterone receptor (PR),

and HER2 was performed essentially as previously described(ref. 11; Supplementary Methods) and analyzed according to theAmerican Society of Clinical Oncology/College of AmericanPathologists guidelines (14, 15).

WES data analysisPreviously generated (7) WES data from tumor-normal DNA

samples were retrieved and reanalyzed. Neoplastic samples weresequenced to a median depth of 192� (range, 95�–369�), andmatched normal samples were sequenced to a median depth of154� (range, 105�–238�; Supplementary Table S1).

WES data analysis was performed as described in Ng andcolleagues (16) and detailed in the Supplementary Methods. Inbrief, after aligning the reads to the reference human genomeGRCh37, somatic genetic alterations were detected using state-of-the-art bioinformatics algorithms, and filters were subsequentlyapplied. In addition to the identification of single-nucleotidevariants (SNV) and insertions and deletions (indels), for samplesfrom a given patient, mutations that were identified in at least onesample were subsequently interrogated in all related samples(SupplementaryMethods). Given thatCDH1 germlinemutationshave been shown to be causative of familial gastric and breastcancer syndrome (17), the germline DNA samples from eachpatient were evaluated for the presence of pathogenic CDH1germline mutations (Supplementary Methods). The potentialfunctional effect of each somatic mutation was defined using acombination ofmutation function predictors shown to have highnegative predictive value (18), as previously described (19), andgenes were annotated according to their presence in three cancergene datasets, Kandoth and colleagues (20), the Cancer GeneCensus (21), and Lawrence and colleagues (22). Allele-specificcopy-number alterations (CNAs) and loss of heterozygosity(LOH) for specific genes were defined using FACETS (23), aspreviously described (16), andpurity andploidy estimationswerecalculated using ABSOLUTE (ref. 24; Supplementary Methods).

Targeted amplicon resequencing validation of somaticmutations

A validation of the mutations found with WES was performedfor cases with sufficient DNA material (n ¼ 11), using a custom-designed AmpliSeq panel on an Ion Torrent Personal GenomeMachine. This validation was not included in Begg and colleagues(7). Out of 4,061 somatic mutations identified by WES, 1,796were investigated in 5 LCIS, 5 DCIS, 8 ILCs, and 2 IDCs from cases1, 2, 4, 7, 8, 10, 11, 12, 13, 14, and15.One thousand four hundredninety-two (83%) mutations were successfully validated. Muta-tions that had sufficient coverage in the validation experiment(minimumof 50 reads) but were not validated (allele frequency <1%) were excluded from the list of mutations used in the down-stream analyses.

Clonality analysisTo infer the clonal relatedness between synchronous lesions,

we defined the "clonality index" as the probability of two lesions

Translational Relevance

We investigated the somatic genetic alterations affecting allprotein coding genes in lobular carcinoma in situ (LCIS) andsynchronously diagnosed ductal carcinomas in situ and inva-sive lobular (ILC) or ductal carcinomas. Our analysesrevealed that LCIS is a genetically advanced lesion, oftendisplaying intralesion genetic heterogeneity, with minor sub-clones of LCIS becoming the dominant clone in ILCs. AnAPOBEC-related mutational signature coupled with overex-pression of APOBEC3B was found to be present in LCISsubclones progressing tomore advanced lesions. Our findingssupport the notion that LCIS is a nonobligate precursor of ILCand suggest that the development of robust molecular pre-dictors of the risk of LCIS progression/evolution into moreaggressive forms of breast cancer may benefit from the assess-ment of intralesion genetic heterogeneity in LCIS.

Clonal Composition of LCIS and Synchronous Lesions

www.aacrjournals.org Clin Cancer Res; 25(2) January 15, 2019 675

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

sharing mutations not expected to have co-occurred by chancebased on a previously validated method (ref. 25; SupplementaryMethods).

Clonal frequenciesTo estimate the clonal architecture and composition of the

lesions fromeachpatient,mutant allelic fractions fromall somaticmutationswere adjusted for tumor cell content, ploidy, local copynumber, and sequencing errors using PyClone, as previouslydescribed (ref. 26; Supplementary Methods).

Truncal and branch mutationsFor each patient displaying at least one LCIS sample clonally

related to other lesions (LCIS, DCIS, or ILC), we categorized themutations into truncal and branch using PyClone (ref. 26; Sup-plementary Methods). Truncal mutations were defined as thoseconcurrently present in the modal populations of all LCIS andtheir clonally related other lesions from a given patient. Branchmutations were defined as those comprising all nontruncalmutations.

Measure of diversityTo quantitate the intralesion genetic heterogeneity of each

sample analyzed, we used the Shannon diversity index (27) andGini–Simpson index (ref. 28; Supplementary Methods).

Phylogenetic tree constructionMaximum parsimony trees were built using binary presence/

absence matrices built from the somatic genetic alterations,

including synonymous and nonsynonymous SNVs, indels, andCNAs, within the clonally related lesions from each patient,essentially as described by Murugaesu and colleagues (refs. 16,29; Supplementary Methods). We have also employed Treeomicsas an alternative approach for the reconstruction of phylogenetictrees (30). Treeomics reconstructs phylogenies using a Bayesianinference model and determines the probability that a variant iseither present or absent in a given sample.

Reverse transcription quantitative PCRTotal RNA was extracted using TRIZOL and reverse tran-

scribed using SuperScript VILO Master Mix (Life Technologies;Thermo Fisher Scientific) according to the manufacturer'sinstructions from cases for which sufficient frozen tissue sam-ples were available. Reverse transcription quantitative PCR (RT-qPCR) was performed to analyze the expression levels ofAPOBEC3B, APOBEC3H, and REV1 genes using TaqManAssay-on-Demand (Supplementary Methods).

Mutational frequencies of TCGA ILCs and luminal-A cancersTCGA luminal-A–invasive breast cancers (31) and ILCs (6)

and their mutations were retrieved from the "Final Full BRCASample Summary" and "Mutations - Publicly accessible MAFarchives" at https://tcga-data.nci.nih.gov/docs/publications/brca_2012/ and https://tcga-data.nci.nih.gov/docs/publications/brca_2015/, including all nonsilent, non-RNA mutations for 209luminal-A primary invasive breast cancers and 127 ILCs. Previousstudies have demonstrated the equivalence between the TCGApipeline and the pipeline employed in this study for mutationdetection (19, 32).

Table 1. Clinicopathologic characteristics of the 24 patients included in the study

Breast Frozen tissue Type Tumor size Lymph node Sakr and Begg andCase ID laterality blocks analyzed (n) LCISa DCISb ILC IDC (invasive, mm) status ER PR HER2 colleagues (10) colleagues (7)

1 Left 16 ü ü N/A N/A þ þ þ N Y2 Right 81 ü ü 21 þ þ þ – Y Y3 Left 52 ü ü ü 18 – þ þ – N Y4 Right 28 ü ü 28 – þ – – Y Y5 Left 27 ü ü 23 – þ þ – N Y6 Left 28 ü ü 16 þ þ þ – N Y7 Left 27 ü ü ü ü 15 – þ þ – N Y8 Left 26 ü ü 60 þ þ þ – Y Y9 Right 54 ü ü 37 þ þ þ – Y Y10 Left 24 ü ü N/A N/A þ þ – N Y11 Right 97 ü ü ü 14 þ þ þ – N Y12 Right 52 ü ü 30 þ þ þ – N Y13 Left 50 ü N/A N/A þ þ – N Y

Right ü ü 17 – þ þ – N Y14 Left 50 ü N/A N/A þ þ – N Y

Right ü ü 7.5 – þ þ – N Y15 Left 50 ü ü ü 18 þ þ þ – N Y

Right ü ü 35 þ þ þ – N Y16 Left N/A ü ü N/A N/A þ þ – N Y17 Left N/A ü N/A N/A þ þ – N Y18 Left N/A ü ü 16 N/A þ þ – Y Y19 Left N/A ü ü N/A N/A þ þ – N Y20 Left N/A ü N/A N/A þ þ – N Y21 Left N/A ü N/A N/A þ þ – N Y22 Right N/A ü ü 21 N/A þ þ – Y Y23 Right N/A ü N/A N/A þ þ – Y Y24 Right N/A ü ü 13 N/A þ þ – Y Y

Abbreviations: ER, estrogen receptor; N, the case was not analyzed in the previous study; N/A, not available; PR, progesterone receptor; Y, the case was analyzed inthe previous study; þ, positive; –, negative.aAll LCIS are of classic type.bAll DCIS are of grade 2.

Lee et al.

Clin Cancer Res; 25(2) January 15, 2019 Clinical Cancer Research676

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Mutational signaturesTo define the mutational signatures involved in the develop-

ment of LCIS, DCIS, and ILCs, we employed deconstructSigs (33)basedon the set ofmutational signatures "signature.cosmic" (34).

Statistical analysisAnalyses were performed using R. For comparisons between

categorical variables, the Fisher's exact testwas employed,whereasfor continuous variables, the Student t test and Mann–WhitneyU test were employed as appropriate. A hypergeometric test wasperformed to estimate the statistical significance of the enrich-ment for cancer genes [genes present in at least one of the cancergene lists by Lawrence and colleagues (Cancer 5000-S; ref. 22),Kandoth and colleagues (20), and/or Cancer Gene Census (21);n ¼ 745] in the genes with truncal mutations (n ¼ 559) andbranch mutations (n ¼ 2,452). For the hypergeometric test, thetotal number of genes in the genome used was 18,986, as definedas the number of protein-coding genes by the HUGO GeneNomenclature Committee. The representation (enrichment)factors and the P values of the hypergeometric tests wereprovided for the analyses performed. All tests were two-sided,and P values < 0.05 were considered statistically significant,adjusted for multiple comparisons where specified.

ResultsLCIS displays a repertoire of somatic genetic alterationsconsistent with those of ILCs and luminal-A–like breast cancers

This study consists of a reanalysis of previously describedWES data (7), followed by a previously unpublished targetedamplicon sequencing validation of approximately 1,800 selectedmutations, from 43 LCIS and synchronous DCIS (n ¼ 9), ILCs(n ¼ 13), or IDCs (n ¼ 5) from 24 patients (Table 1). Threepatients underwent bilateral mastectomy, 1 was therapeutic forbilateral breast cancer and 2 patients underwent contralateralprophylactic mastectomy; these 3 patients were found to havebilateral LCIS (Table 1). All LCIS lesions were of classic type, andall DCIS were of intermediate nuclear grade. For those patientswith invasive lesions, tumor size and ER, PR, and HER2 status ininvasive tumor cells are described in Table 1. Notably, all invasivecarcinomas were ER-positive/HER2-negative.

Somatic mutation analysis of the 43 LCIS lesions revealed amedian of 20 nonsynonymous somatic mutations/lesion (range,5–333) and a mutation rate of 0.39 mutations/Mb (Figs. 1Aand 2A–D), comparable with the number of nonsynonymoussomatic mutations and the mutation rates of 209 luminal-A–invasive breast cancers (31) and 127 ILCs (6) from TCGA [i.e., 27somaticmutations/lesion (range, 7–203) and0.52mutations/Mbin luminal-A and 29 somatic mutations/lesion (range, 1–1,080)and 0.56 mutations/Mb in ILCs; Mann–Whitney U test, P > 0.1].Consistent with the notion that CDH1 inactivation is a driver oflesions with lobular histologic features (1), we observed patho-genic mutations affecting the CDH1 gene in 35 of 43 (81%) LCIS,of which all but three were somatic; patient 13, who had threedistinct foci of LCIS, was found to harbor a CDH1 germlinemutation. All but two CDH1 mutations were coupled with LOHof thewild-type allele [77% (33/43) of all LCIS analyzed; Fig. 1A].Moreover, all LCIS cases lacked E-cadherin expression by immu-nohistochemical analysis. LCIS lacking CDH1mutations did notharbor mutations or deletions affecting genes coding for addi-tional proteins that comprise the cadherin–catenin complex, such

as CTNNB1 (b-catenin), CTNNA1 (a-catenin), or CTNND1(p120-catenin), nor somatic or germline genetic alterations inRHOA (Supplementary Data File S1), a gene that has beenimplicated in the biology of gastric cancer (35), and whosealterations result in neoplastic cells displaying discohesivenessakin to that caused by CDH1 loss of function.

Additional genes identified by TCGA to be significantly mutat-ed in ILCs (6), such as PIK3CA, TBX3, FOXA1, andMAP3K1, werealso found to be recurrently somaticallymutated in LCIS (Figs. 1Aand 2E); however, TP53 somatic mutations, and PTEN somaticmutations and homozygous deletions, present in 8%, 7%, and6% of ILCs analyzed by TCGA (6), were not found in any of theLCIS analyzed here. Notably, TP53 mutations were significantlymore frequently found in luminal-A–invasive breast cancers fromTCGA than in the LCIS analyzed here [12% (25/209) vs. 0%(0/43), Fisher's exact test, P ¼ 0.019; Fig. 2E]. Moreover, genesidentified by TCGA to be significantly mutated in luminal-A–invasive breast cancers, including CBFB, GATA3, NCOR1, andMED23, were also found to be recurrently mutated in LCIS.Interestingly, however, CBFB was found to be mutated in 19%(8/43) of LCIS, a rate significantly higher than that in 2% (2/127)of ILCs and 2% (5/209) of luminal-A breast cancers from TCGA(Fisher's exact tests, P < 0.01, Fig. 2E). Gene CNA analysis revealedrecurrent losses of 16q and gains of 1q (Fig. 1B), a pattern alsoobserved in ILCs (4, 6) and luminal-A–invasive breast cancers(31). Taken together, our findings demonstrate that LCIS syn-chronously diagnosed with more advanced lesions in thisstudy is a genetically advanced, neoplastic lesion often driven byE-cadherin loss of function, with a spectrum of somatic geneticalterations affecting genes commonly altered in ILCs and luminal-A–invasive breast cancers.

LCIS is often clonally related to DCIS and ILCsWES of 9 DCIS (a noninvasive precursor lesion perceived

clinically to be more advanced than LCIS; refs. 36, 37), 13 ILCs,and 5 IDCs collected synchronouslywith the LCIS analyzed abovedemonstrated that overall these lesions displayed similar numberof mutations/case, mutation rates, repertoires of CNAs, andnonsynonymous somaticmutations to those of the LCIS analyzedin this study (Figs. 1 and 2A–D), with exception ofCDH1 somaticmutations that were exclusively found in LCIS and ILCs.

We reasoned that the somatic mutations and CNAs found inanatomically distinct foci of LCIS, ILC, DCIS, and IDC couldprovide a basis for defining their clonal relatedness. Consistentwith the analysis reported by Begg and colleagues (7), but basedon distinct bioinformatics and biostatistical approaches (Supple-mentary Methods), here we demonstrate that all multifocal LCISoriginating in the same breast quadrant (8/8 samples, 4 patients;cases 4, 7, 9, and 23) were clonally related, harboring severalidentical somatic mutations and CNAs (Fig. 3A; SupplementaryFigs. S1–S3). Sixty-seven percent (16/24) of multifocal LCISaffecting distinct quadrants of the breast were also clonally related(Fig. 3A). Further, 10 of 13 (77%) ILCs and 5 of 9 (56%) DCISsamples were found to be clonally related to at least one syn-chronous LCIS analyzed (Fig. 3A). Interestingly, none of the fiveIDCs studiedwere found to be clonally related to a LCIS (Fig. 3A);however, in all three cases where synchronous DCIS and IDCsamples were analyzed, the DCIS and IDC were found to beclonally related (Fig. 3A; Supplementary Figs. S1 and S2). Asexpected, no clonal relatedness was observed between lesionsarising in distinct breasts (bilateral cases; Fig. 3A; Supplementary

Clonal Composition of LCIS and Synchronous Lesions

www.aacrjournals.org Clin Cancer Res; 25(2) January 15, 2019 677

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Figure 1.

Landscape of somatic genetic alterations of LCIS and associated lesions. A, Heatmap illustrating the recurrent (n � 2) somatic mutations and selected geneamplifications in LCIS (n ¼ 43), ILCs (n ¼ 13), DCIS (n ¼ 9), and IDCs (n ¼ 5), subjected to WES. Cases are shown in columns, grouped according tohistologic category color-coded according to the legend, and genes in rows. Somatic mutations affecting cancer genes listed in Kandoth and colleagues (20),the Cancer Gene Census (21), and/or Lawrence and colleagues (22) are ordered from top to bottom in decreasing order of frequency, followed by selectedgene amplifications. Genes highlighted in bold and/or red represent significantly mutated genes in ILCs and/or luminal-A–invasive breast cancers from TCGAbreast cancer studies (6, 31). B, Heatmap illustrating the copy number alterations found in LCIS (n ¼ 43), ILC (n ¼ 13), DCIS (n ¼ 9), and IDC (n ¼ 5). For A and B,mutation type, copy-number states, and/or type of lesion are indicated according to the color keys on the right of the figure.

Lee et al.

Clin Cancer Res; 25(2) January 15, 2019 Clinical Cancer Research678

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Figure 2.

Comparison of mutation rate and frequency of mutations at gene level affecting LCIS and ILCs and luminal-A breast cancers from The Cancer Genome Atlas (TCGA)breast cancer study. Boxplots showing the (A andB)mutation burden and (C andD)mutation rate (mutation/Mb) in LCIS samples (n¼ 43) and ILC (n¼ 13), DCIS (n¼ 9),and IDC (n ¼ 5) from this study and ILC (n ¼ 127) and luminal-A breast cancers (n ¼ 209) from TCGA (31). NS, not significant; � , P > 0.05; �� , P > 0.1 (Mann–WhitneyU test). E, Heatmap depicting the most recurrently mutated genes affecting cancer genes identified in LCIS samples from this study and ILCs (6) and luminal-A–invasive breast cancers (31) from TCGA. Cases are shown in columns, and genes in rows. Fisher's exact test comparisons of mutational frequencies of the mutated geneswere performedbetweenLCIS from this study (n¼ 43) and 127 ILCs (6) and 209 luminal-Abreast cancers (31) fromTCGA. The significantly differentmutation frequenciesbetween LCIS and TCGA ILCs and/or luminal-A breast cancers (TCGA) are highlighted with an asterisk, where �� , P < 0.01 and �, P < 0.05 (Fisher's exact test).

Clonal Composition of LCIS and Synchronous Lesions

www.aacrjournals.org Clin Cancer Res; 25(2) January 15, 2019 679

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Figure 3.

Clonal relatedness and intralesion genetic heterogeneity in LCIS, ILC, and DCIS. A, Schematic representation of the anatomical locations (breast quadrants)of all sequenced samples in each patient and their clonal relatedness. Clonally-related lesions are connected by orange or green lines, whereas those in blackrepresent lesions without a clonal relationship with any other lesion from the respective patient. In cases of unilateral LCIS, only the left or right breast wasrepresented, and in cases of bilateral LCIS, both breasts were schematically depicted. Boxplot illustrating the distribution of (B) the Shannon diversity index and (C)the Gini–Simpson diversity index in LCIS not clonally related to DCIS/ILC (n ¼ 25) and in LCIS clonally related to DCIS and/or ILC (n ¼ 18). The coloreddots indicate the cases. P values of unpaired t test with Welch correction are indicated at the top of each figure. NS, not significant; P value > 0.05.

Lee et al.

Clin Cancer Res; 25(2) January 15, 2019 Clinical Cancer Research680

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Figs. S1 and S2). In addition, the clonal relatedness reported bySakr and colleagues for the 8 pairs of LCIS and ILC was confirmedin this study (Supplementary Table S2).

Taken together, our findings indicate that the majority ofmultifocal LCIS lesions are clonally related and that the presenceof these lesions in distinct quadrants of the breast does not predicttheir clonal relatedness. LCIS and synchronous DCIS and/or ILCare often clonally related, corroborating the notion (4, 7, 38, 39)that LCIS is a nonobligate precursor of more clinically advancedlesions, in particular ILCs. Furthermore, no evidence of clonalitybetween LCIS and IDC was observed here, suggesting that directprogression from CDH1-mutant LCIS to IDC is an uncommonbiological phenomenon.

LCIS foci displaying intralesion genetic heterogeneity are morelikely to progress to ILCs

Recent studies have demonstrated that intratumor geneticheterogeneity may be present in noninvasive lesions includingDCIS (11, 40) andpreinvasive lesions arising in other organs (e.g.,of the esophagus; ref. 41). In such cases, all neoplastic cells harborthe founder genetic events (i.e., truncalmutations), and subclonalpopulations of cancer cells display additional genetic alterations(i.e., branch mutations; ref. 42). We posited that LCIS wouldharbor intralesion genetic heterogeneity and that LCIS lesionswhen clonally related to DCIS or ILC would be associated with ahigher level of intralesion genetic heterogeneity than LCIS notclonally related to more advanced lesions.

To test this hypothesis, we resolved the clonal composition ofLCIS, DCIS, and/or ILC samples by applying a Bayesian clusteringmodel (PyClone; ref. 26) tomutant allele fractions, incorporatingtumor cellularity, ploidy, and local copy number obtained fromABSOLUTE (24) and/or FACETS (ref. 23; Supplementary Meth-ods). This analysis revealed that all but two (89%; 16/18) LCISclonally related to DCIS/ILC but only 40% (10/25) of LCIS notclonally related to DCIS/ILC displayed intralesion geneticheterogeneity at the sequencing depth analyzed (Fisher's exacttest, P ¼ 0.0016; Supplementary Figs. S4A and S4B). Thesefindings were further corroborated by an analysis of the Shannonand Gini–Simpson diversity indices (27, 28, 43), which demon-strated that as a group LCIS clonally related to DCIS and/or ILC(n ¼ 18) displayed significantly higher intralesion genetic het-erogeneity than LCIS not clonally related to more advancedlesions (n ¼ 25; Mann–Whitney U test, P ¼ 0.005; Fig. 3B andC; Supplementary Figs. S4C and S4D). Interestingly, in case 4,composed of two LCIS and one ILC, all sharing a commonancestor, the LCIS lesion displaying heterogeneity was found tobe the likeliest direct precursor of the ILC (Fig. 4A).

Given the intralesion genetic heterogeneity observed in LCIS, inparticular in those related tomore advanced lesions, we sought todefine whether the branch mutations found in these lesionswould affect "passenger" genes or genes significantly mutatedin cancer (20–22). Contrary to the notion that heterogeneitywould primarily affect passenger genetic events, both truncal andbranch nonsynonymous somatic mutations detected in LCISclonally related to the other lesions were found to target genessignificantly enriched for known cancer drivers (refs. 20–22;hypergeometric test, representation factor ¼ 2.09, P < 0.01, andhypergeometric test, representation factor¼ 1.5, P < 0.01, respec-tively; Supplementary Figs. S4E and S4F). Importantly, however,in agreement with previous multiregion analyses that suggestedthat most of the driver genetic alterations are early truncal events

(13, 16, 44), the enrichment for cancer genes was higher in theconstellation of truncal than in branch mutations. Truncal muta-tions included genes found to be significantly mutated in ILCsand/or luminal-A–invasive breast cancers, including CDH1,PIK3CA, MAP3K1, CBFB, SF3B1, RUNX1, and FOXA1 (Supple-mentary Data File S1), whereas branch mutations includedGATA3, PIK3CA, ERBB2, and KMT2C.

Given the clonal relatedness of LCIS with DCIS and ILC, weposited that progression fromLCIS toDCIS/ILC could result in theselection of specific subclones harboring private genetic altera-tions (11, 12, 40). In 29% (4/14) of cases where LCIS was clonallyrelated to DCIS or ILC, we observed that a selected populationfrom the LCIS became dominant in the respective DCIS or ILC(Fig. 4; Supplementary Fig. S1),whereas in the remaining10 cases,our findings suggested parallel progression between LCIS, DCIS,and/or ILC. In two cases (cases 4 and 10), aminor subclone fromaLCISwas the likeliest substrate for the development of theDCIS orthe ILC (Fig. 4). In cases 1, 11, and16, thebiological chronologyofthe LCIS and DCIS could not be resolved on the basis of thesequencing data available (Supplementary Fig. S1). Analysis ofthe genes affected by branch somatic mutations restricted to, orenriched in, the DCIS/ILC samples clonally related to LCISrevealed that in the progression from LCIS to DCIS or ILC, knowncancer driver genes were affected by somatic mutations [e.g.,MAP3K1 (2 cases),RUNX1,NCOR1, ARID1A, and TBX3 (2 cases)]or LOH of the wild-type allele (Fig. 1; Supplementary Fig. S4F;Supplementary Data File S1).

Taken together, our results demonstrate that LCIS clonallyrelated to DCIS/ILC more frequently displays intralesion geneticheterogeneity than LCIS not clonally related to more advancedlesions, that both truncal and branch mutations are enriched forknown cancer drivers, and that known cancer genes are likelytargeted by somatic genetic events in the progression fromLCIS tomore clinically advanced lesions.

Shifts in mutational processes are linked to progression fromLCIS to DCIS and ILCs

There is evidence to suggest that the mutational processes thatshape the mutational spectra of tumors may change duringevolution (16, 45). Hence, we sought to define whether changesinmutational spectrawere observed in the transition fromLCIS toDCIS/ILC. Given that truncal mutations are likely reflective ofbiological phenomena that took place prior to or during thedevelopment of LCIS, and that branch mutations in DCIS/ILClikely stem from mutational processes involved in tumor main-tenance and progression, we compared the mutational spectrumof truncal and branchmutations in cases where LCIS was clonallyrelated to DCIS/ILC. Both truncal and branch mutations werefound to be enriched for C>T transitions in the NpCpG context,consistent with a signature ascribed to aging (46), and C>Gtransversions and C>T transitions in the TpCpW context, sugges-tive of themutational processes caused by APOBECDNA cytosinedeaminase activity (47), the latter being predominately found inthe branchmutations of case 4 (Fig. 5A) and emerging in theDCISof case 1 and ILC of case 18 (Figs. 5B and 5C). Akin to thevariations in mutational processes observed in the progression ofother cancer types (29, 45), in-depth analysis of cases 1, 4, and 18revealed that a mutational process consistent with the APOBECsignature was active in the progression from LCIS to DCIS or ILC(Fig. 5). Moreover, the mRNA levels of APOBEC3B, a DNAcytosine deaminase that has been causally implicated in the

Clonal Composition of LCIS and Synchronous Lesions

www.aacrjournals.org Clin Cancer Res; 25(2) January 15, 2019 681

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

development of APOBEC signature mutations in cancer (47, 48),were significantly higher in samples displaying an APOBECmutational process than in those displaying an aging signature(Fig. 5D). These observations combined to indicate that, at least ina subset of cases, the APOBEC mutational process is likely to becontributing to the development of more advanced lesions.

DiscussionHere, we provide direct evidence of the neoplastic and non-

obligate precursor nature of at least a subset of LCIS. By perform-ing a clonal decomposition and clonal relatedness analysis ofLCIS and synchronously diagnosed DCIS, ILCs, and/or DCIS, we

Figure 4.

Clonal composition of clonally related LCIS and DCIS or ILC and potential clonal selection during progression. A and B, Decomposition of genetically distinctclones and clonal evolution in lesions from (A) case 4 and (B) case 10 using the results from PyClone (26). On the top left, a schematic representation ofthe quadrants from which each sequenced lesion was sampled is shown, and on the top right, the clonal frequency heatmap of mutations within the lesionsof each case, grouped by their inferred clonal/subclonal structure (clusters). Nonsynonymous somatic mutations are shown. The clusters inferred by PyCloneare shown below the clonal frequency heatmap, and the Shannon index measuring intralesion genetic heterogeneity for each lesion is specified withinparentheses after the sample names in the heatmap. On the bottom left, the parallel coordinates plot generated by PyClone and, in the middle right, a cluster-basedphylogenetic tree based on the clusters identified by PyClone are shown. The color of the trunk and branches matches the color of their respective clustersshown in the parallel coordinates plot. On the bottom right, a histologic lesion-based phylogenetic tree constructed using Treeomics (30) is depicted. Themutationsaffecting cancer genes (colored in orange) and the hotspot mutations (colored in blue) that define a given clone are illustrated alongside the branches.The length of the branches is proportional to the number of mutations that distinguish a given clone from its ancestor. The numbers alongside the branchesrepresent the total number of somatic mutations.

Lee et al.

Clin Cancer Res; 25(2) January 15, 2019 Clinical Cancer Research682

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

Figure 5.

Mutational signatures of trunk and branch mutations in LCIS clonally related DCIS or ILC. Evolution of the mutational processes with a schematicrepresentation of the subclone structure in (A) case 4, (B) case 1, and (C) case 18. Each black line represents the acquisition of somatic genetic alterationsthat define a given clone, and each arrow depicts the divergence of a cell population from one lesion to another along with the acquisition of a set ofsomatic genetic alterations. The mutational signature representative of newly acquired mutations by a given subclone is depicted adjacent to each circle. The piechart depicts the proportion of mutational signatures detected, and signature 1 (aging, blue), signature 2 (APOBEC, violet), and signature 13 (APOBEC, green)are shown, with the remaining mutational signatures merged as "Others" (dark gray). The number alongside the branches is the total number of somatic mutations.D, Reverse transcription quantitative PCR (RT-qPCR) of APOBEC3B (left), APOBEC3H (middle), and REV1 (right) genes in samples displaying the APOBEC-related and aging-related signatures, where tissue samples were available for RNA extraction. The error bars represent the SD of mean of RT-qPCR data (n ¼ 3).

www.aacrjournals.org Clin Cancer Res; 25(2) January 15, 2019 683

Clonal Composition of LCIS and Synchronous Lesions

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

have observed that LCIS can display intralesion genetic hetero-geneity and be clonally related to DCIS and ILCs, whereas pro-gression from LCIS to IDC is likely a rare event. Notably, LCISclonally related to ILCs and/or DCIS was found to display higherlevels of intralesion genetic heterogeneity than LCIS that were notclonally related to amore advanced lesion, and evidence of clonalselection in the progression from LCIS to ILCs and/or DCIS wasdocumented in a subset of patients. In these patients, the APOBECmutational process, which has been implicated in genetic insta-bility and intratumor genetic heterogeneity, appears to be presentlater in the evolution of LCIS and may be involved in its pro-gression to more advanced lesions. Interestingly, the samplesenriched for APOBEC mutation process displayed higher expres-sion levels of APOBEC3B, whose activity has been shown tobe mutagenic (47). Therefore, one may hypothesize that in asubset of LCIS, upregulation of APOBEC3B results in increasedmutagenesis and intratumor genetic heterogeneity, ultimatelypromoting subclonal expansions and progression to ILC.

LCIS has been historically considered a less advanced lesion ascompared with DCIS, and is usually managed conservatively, notmandating surgical excision (1). Accordingly, in the latest versionof the tumor–node–metastasis (TNM) staging system, LCIS is nolonger staged as an in situ carcinoma (pTis) as DCIS is (49). Itshould be noted that although we detected clonal relatednessbetween LCIS and DCIS, and the LCIS as the potential substratefor the development of the DCIS (i.e., case 10), the directionalityof the evolution was not clear in three cases (i.e., cases 1, 11, and16). Hence, we cannot rule out the possibility that in a subset ofcases, LCISmay have arisen from a preexistent DCIS or a commonprecursor (e.g., flat epithelial atypia). In fact, due to themolecularsimilarities between low-grade LCIS andDCIS (2), inactivation ofCDH1 in a DCIS subclone would be the likeliest explanation forsuch aphenotypic shift. Bidirectional progressionbetween lesionsof lobular (atypical lobular hyperplasia and LCIS) and ductalphenotype (atypical ductal hyperplasia and DCIS) is entirelyconsistent with the proposed concept of a low nuclear gradebreast neoplasia family (2), which encompasses a group oflow-grade, ER-positive neoplasms of the breast that not uncom-monly affect the same segment of the breast, if not the sameterminal ductal-lobular unit, and share a remarkably similargenomic landscape, having concurrent 1q gains and 16q losses,and PIK3CA mutations, as their genetic signature (2). Alterna-tively, both LCIS andDCISmight arise from a common precursor(i.e., case 1), such asflat epithelial atypia (2). Taken together, thesefindings support the notion that the progression of LCIS andDCISmight be bidirectional or that these lesions may evolve in parallelfrom a common ancestor.

Our findings demonstrate that LCIS displays a genomic land-scape comparable to that of invasive breast cancers of luminal-Asubtype (31) and/or of lobular histology (6), lesions unequivo-cally more advanced and that mandate therapeutic intervention.Akin to ILCs (6), LCIS harbors recurrent biallelic inactivation ofCDH1 (77%) and recurrent mutations affecting genes commonlymutated in breast cancer, including PIK3CA, FOXA1, and TBX3,among other genes. It should be noted, however, that geneticalterations affecting TP53 and PTEN, previously found as recur-rent events in ILCs (6), were not identified in the LCIS samplesanalyzed in this study. These differences might be related to thefact that our cohort included only classic LCIS, but given thatprogression may occur via clonal selection, and that not onlytruncal, but also branchmutations are enriched for known cancer

genes, it is plausible that acquisition of genetic alterations,including those resulting in inactivation of these two bona fidetumor-suppressor genes,may play a role in the progression to ILC.Indeed, we (7, 11) and others (13) have demonstrated previouslythat loss of PTENmay be associated in the progression fromDCISto IDC.

The finding that LCIS is unlikely clonally related to IDCs is incontrast with previous publications, including that fromBegg andcolleagues (7), who reported two LCIS lesions clonally related toIDCs based on a limited number of shared mutations (one andthree mutations in Patients 9 and 14, respectively), which issubstantially lower than the number of shared mutationsobserved in clonally related LCIS-ILC or LCIS-DCIS lesions inthis study (median, 12; range, 2–171). The mutations describedby Begg and colleagues (7) found to be shared between LCIS andIDC samples may have constituted sequencing artifacts, germlinemutations, or common SNPs (Supplementary Table S2; Supple-mentary Data File S1), as they were filtered out in our moreconservative somaticmutationanalysis. Althoughwedidnot detectdirect clonal relatedness between LCIS and IDC in this study, wecannot ruleout thepossibility that a subsetof synchronous IDCandLCISmay share a commonearlyprecursoror thatductal lesions andLCIS may arise from a common earlier precursor lesion andundergo parallel evolution. In addition, it is also plausible that asubset of LCIS may stem from DCIS harboring 16q losses, but theCDH1 inactivation takes place later in the evolution of the lesion.

Althoughourfindings defineLCIS as a nonobligate precursor ofILC, they do not imply that changes in the clinicalmanagement ofpatients presenting with LCIS are necessary, as the rate of subse-quent breast cancer development in a large cohort of patients witha diagnosis of LCIS as reported by the Surveillance, Epidemiology,and End Results (SEER) database demonstrates a risk of approx-imately 1% per year (50). Nonetheless, our studymight provide aframework for the identification of markers to define LCIS casesthat have a greater likelihood to progress. Although some of thepathologic characteristics of LCIS, such as volume of disease, areassociated with a greater likelihood of progression to DCIS/ILC,there has yet to be a validated biomarker to predict the behavior ofclassic LCIS. Based on our results, one could posit that assessingthe levels of intralesion genetic heterogeneity and/or APOBEC3Bactivity in LCIS may help select patients that should be counseledmore proactively toward surgical excision and/or hormonal che-moprevention, akin to the current management of low- to inter-mediate-grade DCIS. Increasingly, treatment of early ER-positivebreast cancer relies on pathologic features, tumor burden, andgenomic profiles; our findings suggest that with continued inves-tigation, a combination of clinical features, histologic classifica-tion, assessment of volume of disease, and intralesion geneticheterogeneity may allow a more personalized risk assessment forpatients with LCIS.

Our study has important limitations. The prospective accrual offrozen samples of LCIS adequate for detailed molecular studies isremarkably challenging; hence, the sample size of the presentstudy is small. In addition, our studymay not be representative ofincidental cases of LCIS, given that the patients included in thisstudywere accrued in a prospective protocol for themultiregionalsampling of prophylactic and/or therapeutic mastectomies frompatients with a previous diagnosis of LCIS. Moreover, we onlyperformed WES analysis, hence we cannot rule out that non-coding alterations and/or epigenetic changes may play a role inthe development and progression of LCIS. More comprehensive

Lee et al.

Clin Cancer Res; 25(2) January 15, 2019 Clinical Cancer Research684

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

analyses may also be required to define the alternative drivers ofCDH1 wild-type LCIS. Finally, we used tumor bulk sequencingand state-of-the-art computational approaches to infer the clonespresent in each sample/case and their phylogeny. Single-cellsequencing analyses of LCIS and synchronous lesions are war-ranted to confirm our findings and provide direct evidence of theclonal composition of LCIS and of clonal selection in the evolu-tion to more advanced lesions.

Despite these limitations, this proof-of-principle study demon-strates that LCIS is a neoplastic nonobligate precursor ofDCIS andILC, with a repertoire of somatic genetic alterations similar to thatof ILCs and luminal-A–invasive breast cancers, but lacking TP53and PTENmutations. LCIS at diagnosis often displays intralesiongenetic heterogeneity, and, in a subset of cases, the progressionfrom LCIS to DCIS and ILC may involve the selection of clones,which may harbor distinct active mutational processes such asAPOBEC. Our findings suggest that early documentation ofintralesion genetic heterogeneity may be central to developingrobust molecular predictors of the risk of LCIS progression/evolution into more aggressive forms of breast cancer.

Data availabilityWES data have been deposited in the database of Genotypes

and Phenotypes (dbGaP) under the accession phs001006.v1.p1.

Disclosure of Potential Conflicts of InterestR.S. Harris holds ownership interest (including patents) in and is a

consultant/advisory board member for ApoGen Biotechnologies. No poten-tial conflicts of interest were disclosed by the other authors.

DisclaimerThe content is solely the responsibility of the authors and does not neces-

sarily represent the official views of the NIH.

Authors' ContributionsConception and design: V.P. de Andrade, D.B. Solit, B. Weigelt, J.S. Reis-Filho,T.A. KingDevelopment of methodology: J.Y. Lee, S. Piscuoglio, C.K.Y. Ng, D.D. Giri,V.P. de Andrade, D.B. Solit, J.S. Reis-Filho, T.A. KingAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): R.A. Sakr, J.V. Scarpa Carniello, R. Towers, D.D. Giri,V.P. de Andrade, A.D. Papanastasiou, A. Viale, D.B. Solit, J.S. Reis-Filho,T.A. KingAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.Y. Lee, M. Schizas, F.C. Geyer, P. Selenica,S. Piscuoglio, R.A. Sakr, C.K.Y. Ng, R.S. Harris, D.B. Solit, B. Weigelt,J.S. Reis-FilhoWriting, review, and/or revision of the manuscript: J.Y. Lee, M. Schizas,F.C.Geyer, R.A. Sakr, C.K.Y.Ng, J.V. ScarpaCarniello,D.D.Giri, V.P. deAndrade,A. Viale, R.S. Harris, D.B. Solit, B. Weigelt, J.S. Reis-Filho, T.A. KingAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): M. Schizas, D.D. Giri, D.B. SolitStudy supervision: D.B. Solit, B. Weigelt, J.S. Reis-Filho, T.A. King

AcknowledgmentsJ.S. Reis-Filho is funded in part by the Breast Cancer Research Foundation.

S. Piscuoglio is funded by the Swiss National Science Foundation (Ambizionegrant number PZ00P3_168165). R.S. Harris is the Margaret Harvey ScheringLand Grant Chair for Cancer Research, a Distinguished McKnight UniversityProfessor, and an Investigator of theHowardHughesMedical Institute. Researchreported in this publication was supported in part by 2009 Komen CareerCatalyst Award (T.A. King), 2012 Komen Investigator Initiated Research Award(T.A. King), Susan G. Komen for the Cure, and by a Cancer Center SupportGrant of the NIH/NCI (grant no. P30CA008748).

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received April 9, 2018; revised July 26, 2018; accepted August 31, 2018;published first September 5, 2018.

References1. King TA, Reis-Filho JS. Lobular neoplasia. Surg Oncol Clin N Am

2014;23:487–503.2. Lopez-Garcia MA, Geyer FC, Lacroix-Triki M, Marchio C, Reis-Filho JS.

Breast cancer precursors revisited: molecular features and progressionpathways. Histopathology 2010;57:171–92.

3. Weigelt B, Geyer FC, Natrajan R, Lopez-Garcia MA, Ahmad AS, Savage K,et al. Themolecular underpinning of lobular histological growth pattern: agenome-wide transcriptomic analysis of invasive lobular carcinomas andgrade- and molecular subtype-matched invasive ductal carcinomas of nospecial type. J Pathol 2010;220:45–57.

4. Andrade VP, Ostrovnaya I, Seshan VE, Morrogh M, Giri D, Olvera N, et al.Clonal relatedness between lobular carcinoma in situ and synchronousmalignant lesions. Breast Cancer Res 2012;14:R103.

5. Lu YJ, Osin P, Lakhani SR, Di Palma S, Gusterson BA, Shipley JM.Comparative genomic hybridization analysis of lobular carcinoma insitu and atypical lobular hyperplasia and potential roles for gains andlosses of genetic material in breast neoplasia. Cancer Res 1998;58:4721–7.

6. Ciriello G, Gatza ML, Beck AH, Wilkerson MD, Rhie SK, Pastore A, et al.Comprehensive molecular portraits of invasive lobular breast cancer. Cell2015;163:506–19.

7. Begg CB, Ostrovnaya I, Carniello JV, Sakr RA, Giri D, Towers R, et al. Clonalrelationships between lobular carcinoma in situ and other breast malig-nancies. Breast Cancer Res 2016;18:66.

8. Derksen PW, Liu X, Saridin F, van der Gulden H, Zevenhoven J, Evers B,et al. Somatic inactivation of E-cadherin andp53 inmice leads tometastaticlobular mammary carcinoma through induction of anoikis resistance andangiogenesis. Cancer Cell 2006;10:437–49.

9. DesmedtC, ZoppoliG,GundemG,PruneriG, LarsimontD, ForniliM, et al.Genomic characterization of primary invasive lobular breast cancer. J ClinOncol 2016;34:1872–81.

10. Sakr RA, SchizasM, Carniello JV, Ng CK, Piscuoglio S, Giri D, et al. Targetedcapturemassively parallel sequencing analysis of LCIS and invasive lobularcancer: Repertoire of somatic genetic alterations and clonal relationships.Mol Oncol 2016;10:360–70.

11. Martelotto LG, Baslan T, Kendall J, Geyer FC, Burke KA, Spraggon L, et al.Whole-genome single-cell copy number profiling from formalin-fixedparaffin-embedded samples. Nat Med 2017;23:376–85.

12. Casasent AK, Edgerton M, Navin NE. Genome evolution in ductal carci-noma in situ: invasion of the clones. J Pathol 2017;241:208–18.

13. Yates LR, Gerstung M, Knappskog S, Desmedt C, Gundem G, Van Loo P,et al. Subclonal diversification of primary breast cancer revealed by multi-region sequencing. Nat Med 2015;21:751–9.

14. Hammond ME, Hayes DF, Dowsett M, Allred DC, Hagerty KL, Badve S,et al. American Society of Clinical Oncology/College of American Pathol-ogists guideline recommendations for immunohistochemical testing ofestrogen and progesterone receptors in breast cancer. J Clin Oncol2010;28:2784–95.

15. Wolff AC,HammondME,HicksDG,DowsettM,McShane LM, Allison KH,et al. Recommendations for human epidermal growth factor receptor 2testing in breast cancer: American Society of Clinical Oncology/College ofAmerican Pathologists clinical practice guideline update. J Clin Oncol2013;31:3997–4013.

16. Ng CKY, Bidard FC, Piscuoglio S, Geyer FC, Lim RS, de Bruijn I, et al.Genetic heterogeneity in therapy-naive synchronous primary breast can-cers and their metastases. Clin Cancer Res 2017;23:4402–15.

Clonal Composition of LCIS and Synchronous Lesions

www.aacrjournals.org Clin Cancer Res; 25(2) January 15, 2019 685

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

17. Richards FM, McKee SA, Rajpar MH, Cole TR, Evans DG, Jankowski JA,et al. Germline E-cadherin gene (CDH1) mutations predispose tofamilial gastric cancer and colorectal cancer. Hum Mol Genet1999;8:607–10.

18. Martelotto LG, Ng C, De Filippo MR, Zhang Y, Piscuoglio S, Lim R, et al.Benchmarking mutation effect prediction algorithms using functionallyvalidated cancer-related missense mutations. Genome Biol 2014;15:484.

19. Ng CKY, Piscuoglio S, Geyer FC, Burke KA, Pareja F, Eberle CA, et al. Thelandscape of somatic genetic alterations in metaplastic breast carcinomas.Clin Cancer Res 2017;23:3859–70.

20. Kandoth C, McLellan MD, Vandin F, Ye K, Niu B, Lu C, et al. Mutationallandscape and significance across 12 major cancer types. Nature2013;502:333–9.

21. Futreal PA, Coin L, Marshall M, Down T, Hubbard T, Wooster R, et al. Acensus of human cancer genes. Nat Rev Cancer 2004;4:177–83.

22. Lawrence MS, Stojanov P, Mermel CH, Robinson JT, Garraway LA, GolubTR, et al. Discovery and saturation analysis of cancer genes across 21tumour types. Nature 2014;505:495–501.

23. Shen R, Seshan VE. FACETS: allele-specific copy number and clonalheterogeneity analysis tool for high-throughput DNA sequencing. NucleicAcids Res 2016;44:e131.

24. Carter SL, Cibulskis K, Helman E, McKenna A, Shen H, Zack T, et al.Absolute quantification of somatic DNA alterations in human cancer.Nat Biotechnol 2012;30:413–21.

25. Schultheis AM, Ng CK, De Filippo MR, Piscuoglio S, Macedo GS, Gatius S,et al. Massively parallel sequencing-based clonality analysis of synchro-nous endometrioid endometrial and ovarian carcinomas. J Natl CancerInst 2016;108:djv427.

26. Roth A, Khattra J, Yap D, Wan A, Laks E, Biele J, et al. PyClone: statisticalinference of clonal population structure in cancer. Nat Methods2014;11:396–8.

27. Shannon CE. The mathematical theory of communication. 1963. MDComput 1997;14:306–17.

28. Simpson EH. Measurement of diversity. Nature 1949;163:688.29. MurugaesuN,WilsonGA, BirkbakNJ,Watkins T,McGranahanN,Kumar S,

et al. Tracking the genomic evolution of esophageal adenocarcinomathrough neoadjuvant chemotherapy. Cancer Discov 2015;5:821–31.

30. Reiter JG,Makohon-Moore AP,Gerold JM, Bozic I, Chatterjee K, Iacobuzio-Donahue CA, et al. Reconstructing metastatic seeding patterns of humancancers. Nat Commun 2017;8:14114.

31. Cancer Genome Atlas Network. Comprehensive molecular portraits ofhuman breast tumours. Nature 2012;490:61–70.

32. Weigelt B, Bi R, Kumar R, Blecua P, Mandelker DL, Geyer FC, et al. Thelandscape of somatic genetic alterations in breast cancers from ATMgermline mutation carriers. J Natl Cancer Inst 2018;110:1030–4.

33. Rosenthal R, McGranahan N, Herrero J, Taylor BS, Swanton C.DeconstructSigs: delineating mutational processes in single tumorsdistinguishes DNA repair deficiencies and patterns of carcinomaevolution. Genome Biol 2016;17:31.

34. Nik-Zainal S, Davies H, Staaf J, Ramakrishna M, Glodzik D, Zou X, et al.Landscape of somatic mutations in 560 breast cancer whole-genomesequences. Nature 2016;534:47–54.

35. Kakiuchi M, Nishizawa T, Ueda H, Gotoh K, Tanaka A, Hayashi A, et al.Recurrent gain-of-function mutations of RHOA in diffuse-type gastriccarcinoma. Nat Genet 2014;46:583–7.

36. Schnitt SJ, Allred C, Britton P, Ellis IO, Lakhani SR, MorrowM, et al. Ductalcarcinoma in situ. In: Lakhani SR, Ellis IO, Schnitt SJ, Tan PH, van de VijverMJ, editors. WHO classification of tumours of the breast. Lyon: IARC Press;2012. p. 90–4.

37. Vincent-Salomon A, Lucchesi C, Gruel N, Raynal V, Pierron G, GoudefroyeR, et al. Integrated genomic and transcriptomic analysis of ductal carcino-ma in situ of the breast. Clin Cancer Res 2008;14:1956–65.

38. Wagner PL, Kitabayashi N, Chen YT, Shin SJ. Clonal relationship betweenclosely approximated low-grade ductal and lobular lesions in the breast: amolecular study of 10 cases. Am J Clin Pathol 2009;132:871–6.

39. Aulmann S, Penzel R, Longerich T, Funke B, Schirmacher P, Sinn HP.Clonality of lobular carcinoma in situ (LCIS) and metachronous invasivebreast cancer. Breast Cancer Res Treat 2008;107:331–5.

40. CasasentAK, SchalckA,GaoR, Sei E, LongA, PangburnW, et al.Multiclonalinvasion in breast tumors identified by topographic single cell sequencing.Cell 2018;172:205–17 e12.

41. Maley CC, Galipeau PC, Finley JC,Wongsurawat VJ, Li X, Sanchez CA, et al.Genetic clonal diversity predicts progression to esophageal adenocarcino-ma. Nat Genet 2006;38:468–73.

42. McGranahan N, Swanton C. Biological and therapeutic impact of intra-tumor heterogeneity in cancer evolution. Cancer Cell 2015;27:15–26.

43. Almendro V, Kim HJ, Cheng YK, Gonen M, Itzkovitz S, Argani P, et al.Genetic and phenotypic diversity in breast tumor metastases. Cancer Res2014;74:1338–48.

44. Yates LR, Knappskog S,WedgeD, Farmery JHR,Gonzalez S,Martincorena I,et al. Genomic evolution of breast cancer metastasis and relapse. CancerCell 2017;32:169–84 e7.

45. de Bruin EC, McGranahan N, Mitter R, Salm M, Wedge DC, Yates L, et al.Spatial and temporal diversity in genomic instability processes defines lungcancer evolution. Science 2014;346:251–6.

46. AlexandrovLB,Nik-Zainal S,WedgeDC,Aparicio SA, Behjati S, BiankinAV,et al. Signatures of mutational processes in human cancer. Nature 2013;500:415–21.

47. Swanton C, McGranahan N, Starrett GJ, Harris RS. APOBEC enzymes:mutagenic fuel for cancer evolution and heterogeneity. Cancer Discov2015;5:704–12.

48. Burns MB, Lackey L, Carpenter MA, Rathore A, Land AM, Leonard B, et al.APOBEC3B is an enzymatic source of mutation in breast cancer. Nature2013;494:366–70.

49. Hortobagyi GN, Connolly JL, D'Orsi CJ, Edge SB, Mittendorf EA, Rugo HS,et al. Breast. AJCCcancer stagingmanual. 8th ed.NewYork: Springer; 2017.

50. Wong SM, King T, Boileau JF, Barry WT, Golshan M. Population-basedanalysis of breast cancer incidence and survival outcomes in women diag-nosed with lobular carcinoma in situ. Ann Surg Oncol 2017;24:2509–17.

Clin Cancer Res; 25(2) January 15, 2019 Clinical Cancer Research686

Lee et al.

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103

2019;25:674-686. Published OnlineFirst September 5, 2018.Clin Cancer Res   Ju Youn Lee, Michail Schizas, Felipe C. Geyer, et al.   Lobular CarcinomaHeterogeneity and Clonal Evolution in the Progression to Invasive

Display Intralesion GeneticIn SituLobular Carcinomas

  Updated version

  10.1158/1078-0432.CCR-18-1103doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://clincancerres.aacrjournals.org/content/suppl/2018/09/05/1078-0432.CCR-18-1103.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://clincancerres.aacrjournals.org/content/25/2/674.full#ref-list-1

This article cites 47 articles, 10 of which you can access for free at:

  Citing articles

  http://clincancerres.aacrjournals.org/content/25/2/674.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.org

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/25/2/674To request permission to re-use all or part of this article, use this link

on July 5, 2020. © 2019 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2018; DOI: 10.1158/1078-0432.CCR-18-1103