Connectivity of the Cerebello-Thalamo-Cortical Pathway in … · PharmacologicalAnalysis...

Original Investigation | Oncology

Connectivity of the Cerebello-Thalamo-Cortical Pathway in Survivorsof Childhood Leukemia Treated With Chemotherapy OnlyNicholas S. Phillips, MD, PhD; Shelli R. Kesler, PhD; Matthew A. Scoggins, PhD; John O. Glass, MS; Yin Ting Cheung, PhD; Wei Liu, PhD; Pia Banerjee, PhD;Robert J. Ogg, PhD; Deokumar Srivastava, PhD; Ching-Hon Pui, MD; Leslie L. Robison, PhD; Wilburn E. Reddick, PhD; Melissa M. Hudson, MD; Kevin R. Krull, PhD

Abstract

IMPORTANCE Treatment with contemporary chemotherapy-only protocols is associated with riskfor neurocognitive impairment among survivors of childhood acute lymphoblastic leukemia (ALL).

OBJECTIVE To determine whether concurrent use of methotrexate and glucocorticoids isassociated with interference with the antioxidant system of the brain and damage and disruption ofglucocorticoid-sensitive regions of the cerebello-thalamo-cortical network.

DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study was conducted from December2016 to July 2019 in a single pediatric cancer tertiary care center. Participants included survivors ofchildhood ALL who were more than 5 years from cancer diagnosis, age 8 years or older, and treatedon an institutional chemotherapy-only protocol. Age-matched community members were recruitedas a control group. Data were analyzed from August 2017 to August 2020.

EXPOSURE ALL treatment using chemotherapy-only protocols.

MAIN OUTCOMES AND MEASURES This study compared brain volumes between survivors andindividuals in a community control group and examined associations among survivors ofmethotrexate and dexamethasone exposure with neurocognitive outcomes. Functional andeffective connectivity measures were compared between survivors with and without cognitiveimpairment. The Rey-Osterrieth complex figure test, a neurocognitive evaluation in which individualsare asked to copy a figure and then draw the figure from memory, was scored according to publishedguidelines and transformed into age-adjusted z scores based on nationally representative referencedata and used to measure organization and planning deficits. β values for neurocognitive testsrepresented the amount of change in cerebellar volume or chemotherapy exposure associated with1 SD change in neurocognitive outcome by z score (mm3/1 SD in z score for cerebellum, mm3/[g×hr/L]for dexamethasone and methotrexate AUC, and mm3/intrathecal count for total intrathecal count).

RESULTS Among 302 eligible individuals, 218 (72%) participated in the study and 176 (58%) hadusable magnetic resonance imaging (MRI) results. Among these, 89 (51%) were female participantsand the mean (range) age was 6.8 (1-18) years at diagnosis and 14.5 (8-27) years at evaluation. Of 100community individuals recruited as the control group, 82 had usable MRI results; among these, 35(43%) were female individuals and the mean (range) age was 13.8 (8-26) years at evaluation. Therewas no significant difference in total brain volume between survivors and individuals in the controlgroup. Survivors of both sexes showed decreased mean (SD) cerebellar volumes compared with thecontrol population (female: 70 568 [6465] mm3 vs 75 134 [6780] mm3; P < .001; male: 77 335 [6210]mm3 vs 79 020 [7420] mm3; P < .001). In female survivors, decreased cerebellar volume wasassociated with worse performance in Rey-Osterrieth complex figure test (left cerebellum: β = 55.54;SE = 25.55; P = .03; right cerebellum: β = 52.57; SE = 25.50; P = .04) and poorer dominant-hand

(continued)

Key PointsQuestion Are changes in glucocorticoid

receptor–rich brain structures of the

cerebello-thalamo-cortical network

associated with altered network

communication and neurocognitive

performance in survivors of childhood

acute lymphoblastic leukemia (ALL)

treated using chemotherapy-only

protocols?

Findings In this cross-sectional study of

176 survivors of childhood ALL treated

with chemotherapy only, a significant

interaction association was found

among female survivors between

glucocorticoid exposure and altered

functional and effective connectivity of

the cerebellum and dorsolateral

frontal cortex.

Meaning These findings suggest that

female survivors of ALL may be at

increased risk of altered connectivity

from glucocorticoid exposure during

therapy and may benefit from

posttreatment strategies that repair

altered network communication in the

cerebello-thalamo-cortical network.

+ Supplemental content

Author affiliations and article information arelisted at the end of this article.

Open Access. This is an open access article distributed under the terms of the CC-BY License.

JAMA Network Open. 2020;3(11):e2025839. doi:10.1001/jamanetworkopen.2020.25839 (Reprinted) November 20, 2020 1/16

Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 09/03/2021

https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.25839&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.25839

Abstract (continued)

motor processing speed (ie, grooved pegboard performance) (left cerebellum: β = 82.71; SE = 31.04;P = .009; right cerebellum: β = 91.06; SE = 30.72; P = .004). In female survivors, increased numberof intrathecal treatments (ie, number of separate injections) was also associated with WorseRey-Osterrieth test performance (β = −0.154; SE = 0.063; P = .02), as was increased dexamethasoneexposure (β = −0.0014; SE = 0.0005; P = .01). Executive dysfunction was correlated with increasedglobal efficiency between smaller brain regions (Pearson r = −0.24; P = .01) compared withindividuals without dysfunction. Anatomical connectivity showed differences between impaired andnonimpaired survivors. Analysis of variance of effective-connectivity weights identified a significantinteraction association (F = 3.99; P = .02) among the direction and strength of connectivity betweenthe cerebellum and DLPFC, female sex, and executive dysfunction. Finally, no effective connectivitywas found between the precuneus and DLPFC in female survivors with executive dysfunction.

CONCLUSIONS AND RELEVANCE These findings suggest that dexamethasone exposure wasassociated with smaller cerebello-thalamo-cortical regions in survivors of ALL and that disruption ofeffective connectivity was associated with impairment of executive function in female survivors.

JAMA Network Open. 2020;3(11):e2025839. doi:10.1001/jamanetworkopen.2020.25839

Introduction

Among survivors of childhood acute lymphoblastic leukemia (ALL), treatment with contemporarychemotherapy-only protocols, which include high doses of methotrexate and glucocorticoids (ie,prednisone, hydrocortisone, and dexamethasone), is associated with increased risk forneurocognitive impairment.1 Previously, we examined supratentorial connectivity in survivors of ALLtreated with chemotherapy only and found that poor network segregation and specialization of brainregions were associated with increased numbers of triple-therapy (ie, methotrexate, glucocorticoid,and cytarabine) intrathecal administrations and executive dysfunction.2 Additionally, amongsurvivors of childhood ALL, female sex has been associated with higher frequency of executivedysfunction.3 The pathophysiology underlying this association remains unclear.

Concomitant use of glucocorticoids and methotrexate is associated with blocking or depletingkey enzymes in the glutathione antioxidant pathway via direct or indirect action. Glucocorticoids areassociated with decreased activity of glutathione peroxidase in the antioxidant pathway, which isassociated with decreased cellular ability to clear oxidants.4,5 Glucocorticoids are indirectlyassociated with increased sensitivity of brain cells to glutamate toxicity via presynaptic glutamaterelease.6,7 Excess glutamate is associated with inhibition among cystine/glutamate transporters,which is associated with decreased levels of glutathione and accumulation of lipid peroxides andreactive oxygen species.8 Methotrexate is associated with decreased glucose-6-phosphatedehydrogenase activity in the pentose phosphate cycle and glutathione reductase activity, which isimportant for preventing intracellular accumulation of peroxide.9 This double hit, in glucocorticoidreceptor–rich brain regions, may be associated with antioxidant/oxidant imbalance and accumulationof peroxide. Consistent with our model, we previously reported the association of increasedcerebrospinal fluid levels of oxidized phosphatidylcholine (a marker associated with oxidative stress)with induction treatment and consolidation among survivors of ALL treated with chemotherapyonly.10 These increased levels were associated with neurocognitive deficits in working memory,organization, and attention.11 We suggested that neural networks rich in glucocorticoid receptorsmay be particularly at risk of damage from methotrexate and glucocorticoid exposure in survivorsof ALL.

The distribution of glucocorticoid receptors in the brain is not homogenous.12 The hippocampusand cerebellum have the highest concentrations of glucocorticoid receptors. We have previouslydemonstrated13 that survivors of childhood ALL treated on chemotherapy-only protocols had smaller

JAMA Network Open | Oncology Cerebello-Thalamo-Cortical Pathway Connectivity in Chemotherapy-Treated Childhood Leukemia




hippocampal, cerebellar, anterior cingulate, precuneus, and dorsolateral prefrontal cortex (DLPFC)regions compared with age-matched healthy individuals in control groups. This finding is consistentwith studies that showed that glucocorticoid exposure is associated with smaller brain structuresand attenuated functional activity in populations without cancer.14-17 Additionally, studies haveshown that these structures are associated with each other structurally and functionally.18-21

Increasing evidence suggests that the cerebellum may regulate activity and facilitate effectiveneural communication by synchronizing functionally associated supratentorial neural clusters in thecerebral cortex.22-24 These networks organize to facilitate economical signaling, and they servespecific cognitive functions.25,26 Alterations to these networks are associated with inefficientinformation processing and decreased neurocognitive performance.27 Evidence suggests that thefrontoparietal network has increased associations with cerebellar integrity compared with otherneural networks.22,23 Treatment-induced injury to these cerebello-thalamo-cortical networks couldtherefore be associated with disruption of synchronized signaling and neurocognitive impairment.28

Evidence for the role of the cerebello-thalamo-cortical network in neurocognitive function has beendescribed by Schmahmann and Sherman,29 who found that cerebellar damage was associated withdeficits in a constellation of cognitive functions, including working memory, cognitive flexibility,visuospatial integration, language, and global intelligence.

In this study, we performed a targeted, hypothesis-driven network analysis to investigateassociations among functional and effective connectivity, treatment exposures, and neurocognitivefunction in survivors of childhood ALL. We previously found that being a survivor of ALL wasassociated with decreased size in regions of the cerebello-thalamo-cortical network compared withage-matched individuals who did not have ALL.13 We hypothesized that decreased size in theseregions would be associated with neurocognitive impairment and would vary by treatment intensity.Furthermore, we hypothesized that the cerebellum facilitates neural communication (ie, effectiveconnectivity) between the DLPFC and the precuneus and that this would be associated withexecutive dysfunction in survivors of ALL.

Methods

Study Design and ParticipantsThis cross-sectional study protocol was approved by the institutional review board of St JudeChildren’s Research Hospital, and all participants provided written informed consent or assent withinformed consent from the parent or guardian. The study follows the Strengthening the Reporting ofObservational Studies in Epidemiology (STROBE) reporting guideline.

In this study, conducted from December 2016 to July 2019, participants were recruited amongsurvivors of ALL treated in the St Jude Children’s Research Hospital Total Therapy Study XV (TOTXV)protocol.30 Survivors were considered eligible if they had received a cancer diagnosis 5 or more yearspreviously and were aged 8 years or older. The TOTXV protocol was a risk-stratified, institution-based, chemotherapy-only protocol that omitted cranial radiation therapy in all patients.31 Patientsin this protocol were treated with oral dexamethasone and intrathecal glucocorticoids andmethotrexate. This study evaluated intravenous methotrexate, oral dexamethasone, and intrathecalglucocorticoids and methotrexate. Exclusion criteria included death prior to long-term follow-up,secondary cancers or relapse requiring cranial radiation or additional chemotherapy, unrelatedcentral nervous system injury or disease, lack of proficiency in English, and loss of eligibility forpediatric follow-up. Age- and sex-matched individuals who were aged 6 to 25 years were recruitedfrom the local community for a control population and were excluded if not proficient in English or ifdiagnosed with central nervous system injury or disease. Individuals in the control population werematched to the study patient population by socioeconomic status (SES) using the Barratt simplifiedmeasure of social status.




http://www.equator-network.org/reporting-guidelines/strobe/

Pharmacological AnalysisBlood samples were drawn at 0, 6, 23, and 42 hours surrounding each administration of intravenoushigh-dose methotrexate infusion. Blood samples for dexamethasone were collected prior to oraladministration and at 1, 2, 4, and 8 hours after the morning dose on days 1 and 8 of reinduction I inweeks 7 and 8 of continuous therapy.32 Dexamethasone and methotrexate levels were measured inserum and quantified as area under curve (AUC), as previously described.33,34

ImagingSurvivors were assessed once during long-term follow-up, from 5 to 10 years after diagnosis.Survivors and individuals in the control population were imaged using structural magnetic resonanceimaging (MRI), including a T1-weighted set of imaging results acquired using sagittal 3-dimensionalmagnetization-prepared rapid acquisition with gradient echo (MPRAGE) sequence (repetition time[TR]/echo time [TE]/inversion time = 1980/2.32/1100 ms), with an imaging resolution of 1.0 mmisotropic. In survivors, resting-state functional MRI (fMRI) was obtained during 6 minutes of eyes-open rest on a 3T scanner (Siemens Trio or Skyra) using a single-shot T2*-weighted echo-planar pulsesequence (TR, 2.06 seconds; TE, 30 milliseconds; field of view, 192 mm; matrix, 64 × 64; slicethickness, 4 mm). Resting-state images were processed using statistical parametric mapping withSPM statistical software version 8 (Wellcome Centre for Human Neuroimaging), as previouslydescribed (eAppendix in the Supplement).27,35-37

Neurocognitive TestingStandardized neuropsychological testing was conducted within 1 day of brain imaging in survivorsand the control population. A certified psychological examiner conducted testing under supervisionof a board-certified neuropsychologist (K.R.K). As part of a larger battery of tests, the examineradministered the Delis-Kaplan Executive Function System38 and processing-speed and working-memory subtests from Wechsler Intelligence Scale for Children Fourth Edition and Wechsler AdultIntelligence Scale Fourth Edition.39 General results from these tests were reported in a 2016 study.40

Raw scores were transformed into age-corrected standard scores based on population referencevalues. Dominant hand motor processing speed was evaluated using grooved pegboard testperformance, a neurocognitive evaluation of eye-hand coordination and motor speed in whichindividuals are asked to insert grooved pegs into matching grooved holes as quickly as possible,scored according to published guidelines and transformed into age-adjusted z scores based onnationally representative reference data. Executive dysfunction was defined as an age-adjusted zscore below the 10th percentile of the nationally representative z score for that age group in letter/number switch, color/word switch, verbal fluency, digit span backward, Rey-Osterrieth complexfigure, or 20 questions measures. In the Rey-Osterrieth complex figure test, individuals are asked tocopy a complex figure and then draw it from memory immediately and again after a small delay. Itwas scored according to published guidelines and transformed into age-adjusted z scores based onnationally representative reference data; it was used to measure organization and planning deficits.Neurocognitive outcomes are presented by test measurement and grouped, for ease of comparison,by the theoretical construct used in a 2016 study.32

Functional Connectivity AnalysisOur aim was to evaluate specific a priori functional subnetworks (ie, modules) within a brain network(ie, connectome). Accordingly, we defined subnetworks of the cerebello-thalamo-cortical networkas bilateral cerebellar Crus I and II, bilateral thalamus, bilateral DLPFC, and bilateral precuneus usingthe automated anatomical atlas.41 To reduce confounders and improve power, we comparednetwork connectivity in survivors who had executive dysfunction with that of individuals withoutexecutive dysfunction. We also examined a language network (consisting of bilateral Brodmann areas40, 44, and 45) as a control comparison, given that this network has relatively low glucocorticoidreceptor distribution and survivors demonstrate few if any expressive or receptive language





problems. Functional connectivity matrices for the regions of interest were obtained using Conntoolbox version 18.4 (Alfonso Nieto-Castanon), as previously described.22,36,37 Global efficiency ofthe bilateral cerebello-thalamo-cortical network was calculated using a permutation distribution,with the P value calculated by determining the proportion of times the permutation mean differencewas greater than the actual mean difference (out of 2000 permutations); permutation meandifferences outside the confidence interval were considered significant.

Effective Connectivity AnalysisTo test the hypothesis that cerebellar activity is associated with connectivity between the DLPFC andthe precuneus, we performed an analysis of effective connectivity. We used bayesian networkanalysis to identify directional and weighted network structure among the regions of interest (ie,cerebellum, precuneus, and DLPFC). This technique has been used to identify biologically relevantstructure-function dependencies.42,43 Time courses were extracted from a 50-elementindependent-component analysis. Effective connectivity was estimated using a bayesian networkscore–based structure-learning algorithm for survivors with impairment (ie, score on any executivefunction test less than −1.3) and survivors without impairment, stratified by sex, to identifydifferences in network structure across subgroups. Additionally, bayesian network analysis wasperformed on the entire cohort to find the best-fit common model and to estimate connectivityweights for subsequent 2-way analysis of variance (ANOVA) comparison (by sex and by impaired vsnot impaired) (eAppendix in the Supplement).

Statistical AnalysisGeneral linear models were used to compare demographic characteristics and brain volumesbetween survivors and members of the control group. Among survivors, sex-stratified multivariablelinear models were used to test associations among morphometric measurements and serumconcentration of dexamethasone and methotrexate AUC, adjusting for age at diagnosis, age atassessment, and intracranial volume. Regions of interest were identified a priori based onglucocorticoid receptor distribution and inclusion in the cerebello-thalamo-cortical pathway. Amultivariable generalized linear model was used to examine associations between morphometricmeasurements and neurocognitive scores. β values for neurocognitive tests represent the amount ofchange in cerebellar volume or chemotherapy exposure associated with 1 SD change inneurocognitive outcome by z score (mm3/1 SD in z score for cerebellum, mm3/[g×hr/L] fordexamethasone and methotrexate AUC, and mm3/intrathecal count for total intrathecal count).Statistical models included age at diagnosis, age at assessment, intracranial volume, dexamethasoneserum levels, methotrexate serum levels, and number of intrathecal injections received duringtherapy as covariates. Correlation coefficients were calculated between global efficiency metrics fortreatment and neurocognitive outcomes. All results were corrected for false discovery rate byBenjamini, Hochberg, and Yekutieli method and closed testing procedures as applicable. Statisticalanalysis was conducted from August 2017 to August 2020 using RStudio Team 2017, includingRStudio statistical software version 1.1 (RStudio PBC). Two-sided t tests and 1-sided generalized linearmodeling were conducted with a statistical significance level of P < .05.

Results

Demographic and Anatomic DifferencesOf 408 survivors treated at St. Jude on the TOTXV protocol, 302 individuals (74.0%) were eligible forthis study, and of 218 individuals who participated (72.2%), 38 individuals (17.4%) had MRIs withartifacts or incomplete images and 4 individuals had incomplete segmentation. The remaining 176survivors of ALL included in this study had a mean (range) age of 6.8 (1-18) years at diagnosis and 14.5(8-27) years at evaluation; 89 [51%] were female individuals. These survivors of ALL were comparedwith 82 age- and SES-matched community members as a control population. The mean (range) age





of the control population was 13.8 (8-26) years at evaluation, and 35 [43%] were femaleindividuals.(Table 1). There were no significant differences in demographic characteristics betweengroups. Mean SES for the control population was within 10% of patient SES. Survivors were a mean(SD) of 7.7 (1.7) years from diagnosis. No difference was found in intracranial volume betweensurvivors and individuals in the control group, suggesting comparable global brain volumes. Mean(SD) cerebellar volume in survivors, compared with the age-matched control population, wassignificantly smaller in both sexes (male: 77 334 [6210] mm3 vs 79 019 [7420] mm3; P < .001; female:70 568 [6465] mm3 vs. 75 134 [6780] mm3; P < .001).

Table 1. Characteristics and Comparison of Anatomic Locations Associated With the Cerebello-Thalamo-CorticalNetwork Between Survivors and Control Population

Characteristic

Mean (SD)

P valueSurvivors (n = 176) Control population (n = 82)Demographic characteristic

Sex, No. (%)

Male 87 (49.4) 47 (57.3) .47

Female 89 (50.6) 35 (42.7) .86

Race/Ethnicity, No. (%)

White 162 (92.0) 73 (89.0) .43

Black or other 14 (8.0) 9 (11.0)

Age, y

At diagnosis 6.8 (4.5) NA NA

At evaluation 14.5 (4.8) 13.8 (4.8) .85

Time since diagnosis, y 7.7 (1.7) NA NA

Anatomic location

Female

Cerebellum volume, mm3

Left 70 611 (6540) 75 083 (6030) <.001

Right 70 525 (6390) 75 186 (7530) <.001

Thalamus volume, mm3

Left 7457 (801) 7658 (870) .11

Right 7546 (971) 7812 (805) .07

Precuneus thickness, mm

Left 2.48 (0.35) 2.65 (0.22) <.001

Right 2.59 (0.17) 2.59 (0.17) <.001

DLPFC thickness, mm

Left 2.88 (0.19) 2.96 (0.18) .014

Right 2.88 (0.20) 2.98 (0.17) .003

Male

Cerebellum volume, mm3

Left 77 437 (6030) 80 602 (7380) .003

Right 77 232 (6390) 77 437 (7460) .002

Thalamus volume, mm3

Left 8001 (829) 8171 (1002) .14

Right 8142 (907) 8344 (917) .10

Precuneus thickness, mm

Left 2.60 (0.18) 2.73 (0.20) <.001

Right 2.57 (0.17) 2.71 (0.17) <.001

DLPFC thickness, mm

Left 2.86 (0.17) 3.00 (0.19) <.001

Right 2.85 (0.16) 2.98 (0.21) <.001Abbreviations: DLPFC, dorsolateral prefrontal cortex;NA, not applicable.




Neurocognitive OutcomesExecutive function scores and processing speed were significantly decreased for survivors comparedwith population reference values (Table 2). For example, scores for survivors in the executivefunction test for verbal fluency were significantly decreased compared with expected populationvalues (male survivors: mean [SD] age-adjusted z score, −0.49 [1.03]; P < .001; female survivors:mean [SD] age-adjusted z score, −0.25 [0.93]; P = .04). Scores for survivors in the dominant-handprocessing-speed test were significantly decreased compared with expected population values(male survivors: mean [SD] age-adjusted z score, −1.48 [1.50]; P < .001; female survivors: mean [SD]age-adjusted z score, −1.16 [1.57]; P < .001). Additionally, scores for survivors in the Rey-Osterriethcomplex figure test were significantly decreased compared with expected population values (malesurvivors: mean [SD] age-adjusted z score, −2.44 [2.37]; P < .001; female survivors: mean [SD]age-adjusted z score, −2.33 [2.45]; P < .001). Among female survivors, decreased bilateral cerebellarvolume was associated with poorer Rey-Osterrieth complex figure performance (left cerebellum:β = 55.54; SE = 25.55; P = .03; right cerebellum: β = 52.57; SE = 25.50; P = .04), as was increaseddexamethasone exposure (β = −0.0014; SE = 0.0005; P = .01) and increased number of tripleintrathecal chemotherapy treatments (β = −0.154; SE = 0.063; P = .02). For female survivors,decreased bilateral cerebellar volume was associated with poorer dominant-hand motor-processingspeed (ie, grooved pegboard performance) (left cerebellum: β = 82.71; SE = 31.04; P = .009; rightcerebellum: β = 91.06; SE = 30.72; P = .004) (Table 3). Among male survivors, decreased bilateralcerebellar volumes were associated with poorer symbol search processing speed (left cerebellum:β = 49.57; SE = 18.06; P = .007; right cerebellum: β = 47.15; SE = 18.31; P = .01), decreased rightcerebellar volume was associated with poorer digit backward performance (β = 35.04; SE = 17.06;

Table 2. Neurocognitive Outcomes of Survivors of Childhood Acute Lymphoblastic Leukemia After Completion of Therapy Compared With National Reference Values

Outcome

Male (n = 108) Female (n = 104)

Mean (SD)a P valueb,c Impairment (95% CI)b,d Mean (SD)a P valueb,c Impairment (95% CI)b,dMale vs femaleP valueb,c

Executive function

Flexibility

Number-letter switch −0.74 (1.14) <.001 30.48 (21.67-39.28) −0.38 (1.23) .02 24.04 (15.83 to 32.25) .13

Color-word switch −0.34 (0.95) .002 14.85 (7.92-21.79) −0.07 (1.03) .68 13.59 (6.97-20.21) .19

Fluency

Verbal −0.49 (1.03) <.001 31.73 (22.79-40.68) −0.25 (0.93) .04 16.35 (9.24-23.45) .21

Categorical 0.03 (1.15) .78 16.35 (9.24-23.45) −0.01 (0.99) .92 12.5 (6.14-18.86) .84

Working memory

Digit backward −0.33 (1.03) .004 20 (12.35-27.65) −0.25 (0.96) .04 14.42 (7.67-21.18) .80

Spatial backward −0.05 (0.98) .64 10.48 (4.62-16.33) −0.02 (1.00) .86 14.42 7.67-21.18) .85

Organization and planning

Rey-Osterrieth complexfigure copy

−2.44 (2.37) <.001 62.86 (53.62-72.10) −2.33 (2.45) <.001 54.37 (44.75-63.99) .84

20 Questions −0.12 (1.11) .31 19.23 (11.66-26.81) −0.24 (0.94) .04 16.5 (9.34-23.67) .68

Tower −0.18 (0.87) .06 13.21 (6.76-19.65) −0.05 (0.83) .68 7.69 (2.57-12.81) .49

Processing speed

Motor

Dominant hand −1.48 (1.50) <.001 48.57 (39.01-58.13) −1.16 (1.57) <.001 35.92 (26.66-45.19) .28

Visual

Symbol search −0.25 (1.06) .029 17.92 (10.62-25.23) 0.14 (0.99) .27 8.65 (3.25-14.06) .04

Visual-motor

Digit symbol −0.70 (0.90) <.001 25.71 (17.35-34.07) −0.10 (0.93) .42 8.65 (3.25-14.06) <.001

Number sequencing −0.25 (0.99) .02 19.05 (11.54-26.56) −0.18 (1.15) .22 17.31 (10.04-24.58) .80

Letter sequencing −0.47 (1.18) <.001 24.76 (16.51-33.02) −0.23 (1.12) .09 16.35 (9.24-23.45) .28

a Age-adjusted z scores referenced to nationally representative reference values.b Corrected for false-discovery rate.

c Comparison of group with expected population value (μ = 0; σ = 1.0) for test.d A score below the 10th percentile of the age-adjusted z score.




P = .04), and decreased left cerebellar volume was associated with poorer dominant-hand motor-processing speed (β = 56.98; SE = 26.88; P = .04). In male survivors, increased triple intrathecaltreatments was associated with poor dominant-hand motor speed (β = −0.059;SE = 0.044; P = .02).

Functional Connectivity AnalysisOf 176 survivors, 23 individuals (13%) (including 14 female and 9 male survivors) showed completelydisconnected nodes in the left and right hemisphere cerebello-thalamo-cortical networks, even atthe lowest-density threshold required to remove false-positive connections. Among 23 survivorswith completely disconnected nodes, mean (range) age at diagnosis of ALL was 5.4 (2.1-9.2) years forfemale survivors and 9.6 (3.9-18.2) years for male survivors. Female survivors with disconnectednodes had decreased right DLPFC thickness compared with individuals in the control population(mean difference, −0.336; P = .045). Male survivors with disconnected nodes had decreased rightcerebellar volume compared with individuals in the control population, as measured by volume ratiocontrolling for intracranial volume (0.044 vs 0.0294; P = .03). This fragmentation was notsignificantly different in survivors with executive dysfunction compared with those withoutexecutive dysfunction.

Among all study survivors, mean (SD) global efficiency of the bilateral cerebello-thalamo-cortical network was increased compared with a control network that had lower glucocorticoidreceptor distribution (ie, Brodmann areas 40, 44, and 45) in both sexes (females: 0.20 [<0.01] vs 0.12[<0.01]; P < .001; males: 0.20 [<0.01] vs 0.12 [<0.01]; P < .001). Global efficiency of the entirecerebello-thalamo-cortical network was higher in survivors with executive dysfunction (meandifference, 0.0042; 95% CI, −0.0042 to 0.0040; P = .01). Within-module-degree z score of leftDLPFC showed increased differences among survivors with impairment compared with survivorswithout impairment (Figure 1). In female survivors, global efficiency was inversely correlated withcognitive flexibility (Pearson r = −0.24; P = .02) and visual-motor processing (Pearson r = −0.27;P = .04). Individuals with executive dysfunction had increased global efficiency between smaller

Table 3. Clinically Significant Neurocognitive Deficits in Survivors Associated With Decreased Cerebellar Volume and Chemotherapy, Controlling for IntracranialVolume and Age at Diagnosis

Outcome

Cerebellum

Dexamethasone AUC Total intrathecal count Methotrexate AUCLeft Right

β (SE)a P value β (SE)a P value β (SE)b P value β (SE)c P value β (SE)b P valueDigit backward

Male 34.40(17.45)

.05 35.04(17.06)

.04 0.0003(0.0005)

.53 0.020(0.030)

.50 0.0012(0.0093)

.90

Female 7.80(18.51)

.68 7.22(18.50)

.70 −0.0005(0.0004)

.18 −0.006(0.045)

.90 0.0087(0.0113)

.44

Rey-Osterrieth complex figure

Male 2.57(38.35)

.95 7.05(37.58)

.85 0.0010(0.0011)

.36 0.005(0.040)

.90 −0.0029(0.0202)

.89

Female 55.54(25.55)

.03 52.57(25.50)

.04 −0.0014(0.0005)

.01 −0.154(0.063)

.02 −0.0009(0.0304)

.98

Dominant hand motor speed

Male 56.98(26.88)

.04 50.92(25.53)

.06 0.0011(0.0008)

.15 −0.059(0.044)

.02 −0.0053(0.0135)

.69

Female 82.71(31.04)

.009 91.06(30.72)

.004 −0.0001(0.0006)

.83 −0.090(0.070)

.20 0.0242(0.0173)

.17

Symbol search

Male 49.57(18.06)

.007 47.15(18.31)

.01 0.0004(0.0005)

.42 −0.054(0.026)

.06 0.0019(0.0094)

.84

Female 4.90(19.66)

.80 1.13(19.66)

.95 −0.0007(0.0004)

.08 −0.042(0.045)

.35 −0.0027(0.0117)

.82

Abbreviation: AUC, area under the curve.a mm3/1-SD in z score.

b mm3/(g×hr/L).c mm3/intrathecal count.




brain regions (Pearson r = −0.24; P = .01) compared with individuals without dysfunction.Associations between control language network and neurocognitive function were not significant.

Effective Connectivity AnalysisUsing estimated effective connectivity network structure derived from bayesian network analysis,we found that DLPFC activity was conditionally dependent on activity in the cerebellum andprecuneus in all cohorts without impairment (ie, there were no effects of sex). Estimated networkstructure was consistent in male survivors with impairment. However, in female survivors withimpairment, there were differences in network connectivity and directionality (Figure 2).Specifically, cerebellar activity was conditioned by DLPFC activity, and there was no connectivitybetween DLPFC activity and precuneus activity. In ANOVA of connectivity weights derived frombayesian network analysis of the entire cohort, we identified a significant interaction associationamong the direction and strength of connectivity between the cerebellum and DLPFC, female sex,and executive dysfunction (F = 3.99; P = .02).

Discussion

This cross-sectional study found decreased cerebellar volume in survivors of childhood ALL, whowere treated with intravenous methotrexate, oral dexamethasone, and intrathecal glucocorticoidsand methotrexate, compared with a control population. Decreased cerebellar volume was associatedwith poorer performance in tests of working memory, organization and planning, and motor andvisual processing speed. Cerebello-thalamo-cortical network functional connectivity measures wereassociated with poorer neurocognitive outcomes. Moreover, effective connectivity in femalesurvivors with impaired executive function was significantly different compared with femalesurvivors without impairment and male survivors with or without impairment. Our results suggestthat the cerebellum and precuneus are associated with DLPFC activity for all groups except femalesurvivors with impairment, who have differences in both directionality and connectivity. Consistentwith the graph theory analysis, this highlights a different mechanism of network dysfunction infemale survivors with impairment. Additionally, we found that global efficiency (a measure of totalnetwork integration and information exchange) was increased in the cerebello-thalamo-corticalnetwork compared with a control network with fewer glucocorticoid receptors. High efficiencyindicates high network integration and capacity for parallel information processing. However,beyond an optimal range, increased efficiency suggests excessive long-range connections, which aremetabolically costly.44,45 This study’s findings suggest that glucocorticoids and methotrexate may

Figure 1. Surface Image of Cerebello-Thalamo-Cortical Functional Connectivity Differences in Survivors

Between-group difference, z score

0.02 3.41

LR

Purple shades indicate greater between-groupdifference of individuals with executive functionimpairment compared with those without executivefunction impairment. Areas in the left dorsolateralprefrontal cortex show increased differences insurvivors with impairment compared with survivorswithout impairment. Image shown in neurologicconvention. L indicates, left; R, right.




be associated with disruption of substrates of the cerebello-thalamo-cortical neural network andwith neurocognitive impairments seen in survivors of childhood ALL.

In children, healthy development includes a period of overconnectivity followed by a period ofsynaptic pruning.46 As such, higher global efficiency can be associated with altered or poorly prunedstructural networks.47 Higher global efficiency is also associated with greater metabolic cost andshows an inverted, U-shaped association with cognitive function (ie, too little and too muchconnectivity are associated with poor neurocognitive performance).44,48 In female survivors in ourstudy, global efficiency was inversely correlated with cognitive flexibility and visual-motor processingspeed. In comparison, a 2016 study49 of 99 healthy preadolescent children found that globalefficiency was positively associated with performance on measures of visual-motor processing.Moreover, higher global functional connectivity has been found among patients with mild cognitiveimpairment.50 Our results suggest that treatment may be associated with disruption of optimal brainnetwork organization and function and with impairment of executive function seen in survivorsof ALL.

Bilateral cerebello-thalamo-cortical networks appeared disconnected at the lowest-densitythreshold for a subset of survivors and could not be validly compared between groups byhemisphere. The fragmentation was not significantly different in survivors with executivedysfunction compared with those without executive dysfunction, but fragmentation may affectother neurocognitive functions we did not assess. It is biologically implausible that any brain regionwould be completely disconnected from all other regions. Thus, our finding that several survivorshad such low connectivity that a common minimum density could not be found supports ourhypothesis that disruption of the cerebello-thalamo-cortical network may be associated withneurocognitive problems seen in survivors of ALL. This finding could also reflect graph-definingproperties, such as size (ie, number of nodes) of the cerebello-thalamo-cortical network, which is

Figure 2. Bayesian Network Analysis of Cerebello-Thalamo-Cortical Network of Glucocorticoid–Sensitive Regions in Survivors

Malenonimpaired

Cerebellum

Precuneus

DLPFC

Impaired

Cerebellum

Precuneus

DLPFC

Femalenonimpaired

Cerebellum

Precuneus

DLPFC

Impaired

Cerebellum

Precuneus

DLPFC

DLPFC

Thalamus

Precuneus

Cerebellum

Anteriorcingulate gyrus

Hippocampus/entorhinal cortex

Our results suggest that the cerebellum and precuneus are associated with DLPFC(dorsolateral prefrontal cortex) activity for all groups except female survivors withimpairment (shaded box), who have differences in both directionality and connectivity.Gray brain regions indicate cingulum and ventricle and are used for neuroanatomicalreference for the anterior cingulate gyrus, which is a substructure of the cingulum; bold

arrows, neural connections between regions; nodes (ie, circles), brain regions; directededges (ie, thin arrows), highest-probability connection and direction of interactionbetween regions; absence of edge (ie, absence of thin arrow), no probability ofconnectedness.




known to affect connectome measurement.30 However, evidence for altered connectivity wasfurther suggested by bayesian network analysis, which found that impairment in female survivorswas associated with a lack of effective connectivity between precuneus and DLPFC.

We found similarities and differences in neurocognitive-associated network changes by sex.Significant associations were found between dominant hand motor processing speed and leftcerebellar volume in male and female survivors. In male survivors, visual processing speed wasassociated with bilateral cerebellar volume. These findings agree with a 2014 study51 thatdemonstrated an association between the cerebellum and motor/visual processing performance. Inour study, poor Rey-Osterrieth complex figure performance was associated with decreasedcerebellar volume and increased dexamethasone exposure in female survivors but not in malesurvivors. This is noteworthy given the results of the bayesian analysis, which suggested that infemale survivors, impairment was associated with altered effective connectivity among the DLPFC,precuneus, and cerebellum. Lobule IX and Crus I and II of the cerebellum have been shown to projectto the default mode network and may be associated with this disruption.23,52 Unfortunately, thecaudal vermis, which contains lobule IX, was not reliably captured in many of our study’s fMRIs;therefore, we were not able demonstrate this lobule’s effective connectivity to the precuneus to testthis hypothesis.

Sex differences were seen in this study, and these differences in outcomes may be associatedwith postnatal testosterone surge seen in newborn boys.53 A 1996 animal study54 demonstrated thatandrogen receptor binding mediates the downregulation of glucocorticoid receptor expression inCA1 pyramidal cells of the hippocampus in rats. This may indicate that androgen exposure in youngmen is associated with decreased sensitivity to glucocorticoid receptor–mediated effects.Additionally, a 2017 in vitro study55 found that 17-β estradiol levels were associated with modulatingthe effects of oxidative stress and may have protective associations for pubertal girls. This wouldsuggest that prepubertal girls and individuals with hypogonadism may have increased sensitivity tointrathecal triple therapy and dexamethasone. An association between decreased hormone receptorexpression and decreased functional connectivity between DLPFC and precuneus has been seen insurvivors of breast cancer.21 Unfortunately, we did not obtain sex hormone levels and were unable tofully test this hypothesis. Nonetheless, sex differences in our findings may indicate that boys andgirls require different strategies to improve neurocognitive outcomes.

Clinical implications of this study may include the need for strategies to investigate the use ofn-acetyl cysteine (n-AC), which may be able to rescue the glutathione antioxidant system inindividuals receiving chemotherapy. Studies56,57 have shown that n-AC has a good safety profile andis tolerable at doses necessary for central nervous system penetrance. Additionally, N-methyl-D-aspartate (NMDA) receptor antagonist has been associated with reversing of cortisol-inducedglutamatergic synaptic pruning and may be associated with the decreased oxidative injury associatedwith glucocorticoids.58 However, studies are needed to investigate if these neuroprotective agentscan decrease oxidative injury without compromising the efficacy of cancer therapy. Transcranialelectrical stimulation devices have been associated with altered activity of resting-state networksand regulation of NMDA receptor activity and may offer a potential therapy for survivors aftercompletion of therapy.59-62 Additionally, Children’s Oncology Group long-term follow-up guidelinesdo not include dexamethasone or intrathecal hydrocortisone as potential risk factor for cognitive lateeffects.63 The results of our study suggest that this may need to be reconsidered.

LimitationsThis study has some limitations, including the lack of availability of neurocognitive testing in thecontrol population. We did not have measures for biomarkers associated with oxidative injury beforeor during therapy or high-resolution MRI results that could help exclude pretreatment oxidativeinjury associated with disease or genetic predisposition. Additionally, we cannot exclude potentialeffects of cytarabine on brain volumes.64 Furthermore, we performed a subgroup analysis ofsurvivors with or without impairment for the resting-state fMRI analysis to test our hypothesis that




cerebello-thalamo-cortical network disruption was associated with executive dysfunction. This typeof subgroup analysis can increase risk of false-positive errors if done incorrectly. However, our studywas designed to accommodate this type of analysis, based on empirical evidence and defined a priorihypothesis. These factors may reduce the probability of a type I error. Additionally, this study is asingle-institution, cross-sectional analysis and, as such, is not representative of every chemotherapy-only ALL regimen. Doses of methotrexate and dexamethasone in this study were consistent with thecurrent standard of care. Future studies should investigate other glucocorticoids to determinewhether other formulations produce similar effects in those populations. These studies shouldinclude measures of short- and long-term memory in addition to the measures of working memoryincluded in this study.13 Neuroimaging studies should include resting-state fMRI with completecoverage of the cerebellum.

Conclusions

These findings suggest that intrathecal glucocorticoids and methotrexate and oral dexamethasonemay be associated with disrupted functional connectivity and poorer neurocognitive outcomes infemale survivors of childhood ALL. Furthermore, the findings suggest that cerebellum activity isassociated with neural activity of the DLPFC and disruption of the association between the two isassociated with executive dysfunction.

ARTICLE INFORMATIONAccepted for Publication: September 4, 2020.

Published: November 20, 2020. doi:10.1001/jamanetworkopen.2020.25839

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Phillips NSet al. JAMA Network Open.

Corresponding Author: Nicholas S. Phillips, MD, PhD ([email protected]) and Kevin R. Krull, PhD([email protected]), Department of Epidemiology and Cancer Control, St Jude Children’s Research Hospital,262 Danny Thomas Pl, MS 735, Memphis, TN 38105-3678.

Author Affiliations: Department of Epidemiology and Cancer Control, St Jude Children’s Research Hospital,Memphis, Tennessee (Phillips, Banerjee, Robison, Hudson, Krull); Department of Oncology, St Jude Children’sResearch Hospital, Memphis, Tennessee (Phillips, Pui, Hudson); Now with School of Nursing, University of Texasat Austin (Kesler); Department of Neuro-oncology, University of Texas MD Anderson Cancer Center, Houston(Kesler); Department of Diagnostic Imaging, St Jude Children’s Research Hospital, Memphis, Tennessee (Scoggins,Glass, Ogg, Reddick); School of Pharmacy, Faculty of Medicine, Chinese University of Hong Kong, Hong Kong,China (Cheung); Department of Biostatistics, St Jude Children’s Research Hospital, Memphis, Tennessee (Liu,Srivastava); Department of Psychology, St Jude Children’s Research Hospital, Memphis, Tennessee (Krull).

Author Contributions: Drs Phillips and Krull had full access to all of the data in the study and take responsibility forthe integrity of the data and the accuracy of the data analysis. This cross-sectional study was developed by thelead investigator (Dr Phillips) in collaboration with the principal investigator (Dr Krull).

Concept and design: Phillips, Cheung, Ogg, Srivastava, Reddick, Hudson, Krull.

Acquisition, analysis, or interpretation of data: Kesler, Scoggins, Glass, Liu, Banerjee, Ogg, Srivastava, Pui, Robison,Reddick, Hudson, Krull.

Drafting of the manuscript: Phillips, Kesler, Glass, Srivastava, Reddick, Krull.

Critical revision of the manuscript for important intellectual content: Kesler, Scoggins, Cheung, Liu, Banerjee, Ogg,Pui, Robison, Hudson, Krull.

Statistical analysis: Phillips, Kesler, Scoggins, Liu, Srivastava, Krull.

Obtained funding: Ogg, Hudson, Krull.

Administrative, technical, or material support: Glass, Cheung, Banerjee, Pui, Robison, Reddick, Hudson, Krull.

Supervision: Banerjee, Ogg, Srivastava, Krull.





https://jamanetwork.com/journals/jamanetworkopen/pages/instructions-for-authors#SecOpenAccess/?utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.25839

mailto:[email protected]

mailto:[email protected]

Conflict of Interest Disclosures: Dr Phillips reported receiving grants from the National Institutes of Health (NIH)National Cancer Institute (NCI) during the conduct of the study. Dr Glass reported receiving grants from the NCIand National Institute of Mental Health (NIMH) during the conduct of the study. Dr Ogg reported receiving a grantfrom National Institute of Child Health and Human Development. Dr Pui reported receiving grants from the NIHduring the conduct of the study and personal fees from Amgen, Servier, Erytech, and Adaptive Biotechnologiesoutside the submitted work. Dr Reddick reported receiving grants from the NIMH and NCI during the conduct ofthe study. Dr Hudson reported receiving grants from the NCI during the conduct of the study. Dr Krull reportedreceiving grants from the NIH during the conduct of the study. No other disclosures were reported.

Funding/Support: Dr Krull was supported by grant No. MH085849 from the National Institute of Mental Healthand grant No. 1 T32 CA225590 from the NCI. Drs Hudson and Robison were supported by grant No. 5 U01CA195547-04 from the NCI. Dr Ogg was supported by grant No. S10RR029005 from the NIH. This study wassupported by the American Lebanese Syrian Associated Charities and by grant No. CA21765 from the NCI.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection,management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; anddecision to submit the manuscript for publication.

Additional Contributions: E. Brannon Morris III, MD (Department of Neurology, Medical College of Georgia),helped develop the cerebello-thalamo-cortical network hypothesis and was not compensated for this work.

REFERENCES1. Krull KR, Brinkman TM, Li C, et al. Neurocognitive outcomes decades after treatment for childhood acutelymphoblastic leukemia: a report from the St Jude lifetime cohort study. J Clin Oncol. 2013;31(35):4407-4415. doi:10.1200/JCO.2012.48.2315

2. Kesler SR, Ogg R, Reddick WE, et al. Brain network connectivity and executive function in long-term survivorsof childhood acute lymphoblastic leukemia. Brain Connect. 2018;8(6):333-342. doi:10.1089/brain.2017.0574

3. Jacola LM, Krull KR, Pui CH, et al. Longitudinal assessment of neurocognitive outcomes in survivors ofchildhood acute lymphoblastic leukemia treated on a contemporary chemotherapy protocol. J Clin Oncol. 2016;34(11):1239-1247. doi:10.1200/JCO.2015.64.3205

4. McIntosh LJ, Hong KE, Sapolsky RM. Glucocorticoids may alter antioxidant enzyme capacity in the brain:baseline studies. Brain Res. 1998;791(1-2):209-214. doi:10.1016/S0006-8993(98)00115-2

5. Weydert CJ, Cullen JJ. Measurement of superoxide dismutase, catalase and glutathione peroxidase in culturedcells and tissue. Nat Protoc. 2010;5(1):51-66. doi:10.1038/nprot.2009.197

6. Behl C, Lezoualc’h F, Trapp T, Widmann M, Skutella T, Holsboer F. Glucocorticoids enhance oxidative stress–induced cell death in hippocampal neurons in vitro. Endocrinology. 1997;138(1):101-106. doi:10.1210/endo.138.1.4835

7. Osborne DM, Pearson-Leary J, McNay EC. The neuroenergetics of stress hormones in the hippocampus andimplications for memory. Front Neurosci. 2015;9:164. doi:10.3389/fnins.2015.00164

8. Tobaben S, Grohm J, Seiler A, Conrad M, Plesnila N, Culmsee C. Bid-mediated mitochondrial damage is a keymechanism in glutamate-induced oxidative stress and AIF-dependent cell death in immortalized HT-22hippocampal neurons. Cell Death Differ. 2011;18(2):282-292. doi:10.1038/cdd.2010.92

9. Babiak RM, Campello AP, Carnieri EG, Oliveira MB. Methotrexate: pentose cycle and oxidative stress. CellBiochem Funct. 1998;16(4):283-293. doi:10.1002/(SICI)1099-0844(1998120)16:4<283::AID-CBF801>3.0.CO;2-E

10. Miketova P, Kaemingk K, Hockenberry M, et al. Oxidative changes in cerebral spinal fluid phosphatidylcholineduring treatment for acute lymphoblastic leukemia. Biol Res Nurs. 2005;6(3):187-195. doi:10.1177/1099800404271916

11. Caron JE, Krull KR, Hockenberry M, Jain N, Kaemingk K, Moore IM. Oxidative stress and executive function inchildren receiving chemotherapy for acute lymphoblastic leukemia. Pediatr Blood Cancer. 2009;53(4):551-556.doi:10.1002/pbc.22128

12. Webster MJ, Knable MB, O’Grady J, Orthmann J, Weickert CS. Regional specificity of brain glucocorticoidreceptor mRNA alterations in subjects with schizophrenia and mood disorders. Mol Psychiatry.2002;7(9):985-994, 924. doi:10.1038/sj.mp.4001139

13. Phillips NS, Cheung YT, Glass JO, et al. Neuroanatomical abnormalities related to dexamethasone exposure insurvivors of childhood acute lymphoblastic leukemia. Pediatr Blood Cancer. 2020;67(3):e27968. doi:10.1002/pbc.27968

14. Oei NY, Elzinga BM, Wolf OT, et al. Glucocorticoids decrease hippocampal and prefrontal activation duringdeclarative memory retrieval in young men. Brain Imaging Behav. 2007;1(1-2):31-41. doi:10.1007/s11682-007-9003-2




https://dx.doi.org/10.1200/JCO.2012.48.2315

https://dx.doi.org/10.1089/brain.2017.0574


https://dx.doi.org/10.1016/S0006-8993(98)00115-2

https://dx.doi.org/10.1038/nprot.2009.197

https://dx.doi.org/10.1210/endo.138.1.4835

https://dx.doi.org/10.1210/endo.138.1.4835

https://dx.doi.org/10.3389/fnins.2015.00164

https://dx.doi.org/10.1038/cdd.2010.92

https://dx.doi.org/10.1002/(SICI)1099-0844(1998120)16:4%3C283::AID-CBF801%3E3.0.CO;2-E

https://dx.doi.org/10.1177/1099800404271916

https://dx.doi.org/10.1177/1099800404271916

https://dx.doi.org/10.1002/pbc.22128

https://dx.doi.org/10.1038/sj.mp.4001139



https://dx.doi.org/10.1007/s11682-007-9003-2

https://dx.doi.org/10.1007/s11682-007-9003-2

15. Liu S, Wang Y, Xu K, et al. Voxel-based comparison of brain glucose metabolism between patients withCushing’s disease and healthy subjects. Neuroimage Clin. 2017;17:354-358. doi:10.1016/j.nicl.2017.10.038

16. Ilg L, Klados M, Alexander N, Kirschbaum C, Li SC. Long-term impacts of prenatal synthetic glucocorticoidsexposure on functional brain correlates of cognitive monitoring in adolescence. Sci Rep. 2018;8(1):7715. doi:10.1038/s41598-018-26067-3

17. Burkhardt T, Lüdecke D, Spies L, Wittmann L, Westphal M, Flitsch J. Hippocampal and cerebellar atrophy inpatients with Cushing’s disease. Neurosurg Focus. 2015;39(5):E5. doi:10.3171/2015.8.FOCUS15324

18. Margulies DS, Vincent JL, Kelly C, et al. Precuneus shares intrinsic functional architecture in humans andmonkeys. Proc Natl Acad Sci U S A. 2009;106(47):20069-20074. doi:10.1073/pnas.0905314106

19. Zhang S, Li CS. Functional connectivity mapping of the human precuneus by resting state fMRI. Neuroimage.2012;59(4):3548-3562. doi:10.1016/j.neuroimage.2011.11.023

20. Kumar J, Iwabuchi SJ, Völlm BA, Palaniyappan L. Oxytocin modulates the effective connectivity between theprecuneus and the dorsolateral prefrontal cortex. Eur Arch Psychiatry Clin Neurosci. 2020;270(5):567-576. doi:10.1007/s00406-019-00989-z

21. Chen H, Ding K, Zhao J, Chao HH, Li CR, Cheng H. The dorsolateral prefrontal cortex is selectively involved inchemotherapy-related cognitive impairment in breast cancer patients with different hormone receptorexpression. Am J Cancer Res. 2019;9(8):1776-1785.

22. Marek S, Siegel JS, Gordon EM, et al. Spatial and temporal organization of the individual human cerebellum.Neuron. 2018;100(4):977-993.e7. doi:10.1016/j.neuron.2018.10.010

23. Buckner RL, Krienen FM, Castellanos A, Diaz JC, Yeo BT. The organization of the human cerebellum estimatedby intrinsic functional connectivity. J Neurophysiol. 2011;106(5):2322-2345. doi:10.1152/jn.00339.2011

24. Krienen FM, Buckner RL. Segregated fronto-cerebellar circuits revealed by intrinsic functional connectivity.Cereb Cortex. 2009;19(10):2485-2497. doi:10.1093/cercor/bhp135

25. Mesulam MM. Large-scale neurocognitive networks and distributed processing for attention, language, andmemory. Ann Neurol. 1990;28(5):597-613. doi:10.1002/ana.410280502

26. Bressler SL. Large-scale cortical networks and cognition. Brain Res Brain Res Rev. 1995;20(3):288-304. doi:10.1016/0165-0173(94)00016-I

27. Kesler SR, Gugel M, Pritchard-Berman M, et al. Altered resting state functional connectivity in young survivorsof acute lymphoblastic leukemia. Pediatr Blood Cancer. 2014;61(7):1295-1299. doi:10.1002/pbc.25022

28. Person AL, Raman IM. Synchrony and neural coding in cerebellar circuits. Front Neural Circuits. 2012;6:97. doi:10.3389/fncir.2012.00097

29. Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain. 1998;121(Pt 4):561-579.doi:10.1093/brain/121.4.561

30. Pui CH, Relling MV, Sandlund JT, Downing JR, Campana D, Evans WE. Rationale and design of Total TherapyStudy XV for newly diagnosed childhood acute lymphoblastic leukemia. Ann Hematol. 2004;83(suppl 1):S124-S126.

31. Pui CH, Campana D, Pei D, et al. Treating childhood acute lymphoblastic leukemia without cranial irradiation.N Engl J Med. 2009;360(26):2730-2741. doi:10.1056/NEJMoa0900386

32. Krull KR, Cheung YT, Liu W, et al. Chemotherapy pharmacodynamics and neuroimaging and neurocognitiveoutcomes in long-term survivors of childhood acute lymphoblastic leukemia. J Clin Oncol. 2016;34(22):2644-2653. doi:10.1200/JCO.2015.65.4574

33. Bhojwani D, Sabin ND, Pei D, et al. Methotrexate-induced neurotoxicity and leukoencephalopathy in childhoodacute lymphoblastic leukemia. J Clin Oncol. 2014;32(9):949-959. doi:10.1200/JCO.2013.53.0808

34. Yang L, Panetta JC, Cai X, et al. Asparaginase may influence dexamethasone pharmacokinetics in acutelymphoblastic leukemia. J Clin Oncol. 2008;26(12):1932-1939. doi:10.1200/JCO.2007.13.8404

35. Kesler SR, Blayney DW. Neurotoxic effects of anthracycline- vs nonanthracycline-based chemotherapy oncognition in breast cancer survivors. JAMA Oncol. 2016;2(2):185-192. doi:10.1001/jamaoncol.2015.4333

36. Kesler SR, Wefel JS, Hosseini SM, Cheung M, Watson CL, Hoeft F. Default mode network connectivitydistinguishes chemotherapy-treated breast cancer survivors from controls. Proc Natl Acad Sci U S A. 2013;110(28):11600-11605. doi:10.1073/pnas.1214551110

37. Bruno J, Hosseini SM, Kesler S. Altered resting state functional brain network topology in chemotherapy-treated breast cancer survivors. Neurobiol Dis. 2012;48(3):329-338. doi:10.1016/j.nbd.2012.07.009

38. Delis DC, Kramer JH, Kaplan E, Holdnack J. Reliability and validity of the Delis-Kaplan Executive FunctionSystem: an update. J Int Neuropsychol Soc. 2004;10(2):301-303. doi:10.1017/S1355617704102191




https://dx.doi.org/10.1016/j.nicl.2017.10.038

https://dx.doi.org/10.1038/s41598-018-26067-3

https://dx.doi.org/10.1038/s41598-018-26067-3

https://dx.doi.org/10.3171/2015.8.FOCUS15324

https://dx.doi.org/10.1073/pnas.0905314106

https://dx.doi.org/10.1016/j.neuroimage.2011.11.023

https://dx.doi.org/10.1007/s00406-019-00989-z

https://dx.doi.org/10.1007/s00406-019-00989-z

https://www.ncbi.nlm.nih.gov/pubmed/31497358

https://dx.doi.org/10.1016/j.neuron.2018.10.010

https://dx.doi.org/10.1152/jn.00339.2011

https://dx.doi.org/10.1093/cercor/bhp135

https://dx.doi.org/10.1002/ana.410280502

https://dx.doi.org/10.1016/0165-0173(94)00016-I


https://dx.doi.org/10.3389/fncir.2012.00097

https://dx.doi.org/10.1093/brain/121.4.561



https://dx.doi.org/10.1056/NEJMoa0900386




https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamaoncol.2015.4333&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.25839

https://dx.doi.org/10.1073/pnas.1214551110

https://dx.doi.org/10.1016/j.nbd.2012.07.009

https://dx.doi.org/10.1017/S1355617704102191

39. Strauss ES, Sherman EMS, Spreen O. A Compendium of Neuropsychological Tests: Administration, Norms, andCommentary. 3rd ed. Oxford University Press; 2006.

40. Cheung YT, Sabin ND, Reddick WE, et al. Leukoencephalopathy and long-term neurobehavioural,neurocognitive, and brain imaging outcomes in survivors of childhood acute lymphoblastic leukaemia treated withchemotherapy: a longitudinal analysis. Lancet Haematol. 2016;3(10):e456-e466. doi:10.1016/S2352-3026(16)30110-7

41. Tzourio-Mazoyer N, Landeau B, Papathanassiou D, et al. Automated anatomical labeling of activations in SPMusing a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage. 2002;15(1):273-289. doi:10.1006/nimg.2001.0978

42. Arnoux A, Toba MN, Duering M, et al. Is VLSM a valid tool for determining the functional anatomy of the brain:usefulness of additional bayesian network analysis. Neuropsychologia. 2018;121:69-78. doi:10.1016/j.neuropsychologia.2018.10.003

43. Wang L, Durante D, Jung RE, Dunson DB. Bayesian network-response regression. Bioinformatics. 2017;33(12):1859-1866. doi:10.1093/bioinformatics/btx050

44. Achard S, Bullmore E. Efficiency and cost of economical brain functional networks. PLoS Comput Biol. 2007;3(2):e17. doi:10.1371/journal.pcbi.0030017

45. Kesler SR, Noll K, Cahill DP, Rao G, Wefel JS. The effect of IDH1 mutation on the structural connectome inmalignant astrocytoma. J Neurooncol. 2017;131(3):565-574. doi:10.1007/s11060-016-2328-1

46. Supekar K, Musen M, Menon V. Development of large-scale functional brain networks in children. PLoS Biol.2009;7(7):e1000157. doi:10.1371/journal.pbio.1000157

47. Qin B, Wang L, Zhang Y, Cai J, Chen J, Li T. Enhanced topological network efficiency in preschool autismspectrum disorder: a diffusion tensor imaging study. Front Psychiatry. 2018;9:278. doi:10.3389/fpsyt.2018.00278

48. Kesler SR, Gugel M, Huston-Warren E, Watson C. Atypical structural connectome organization and cognitiveimpairment in young survivors of acute lymphoblastic leukemia. Brain Connect. 2016;6(4):273-282. doi:10.1089/brain.2015.0409

49. Kim DJ, Davis EP, Sandman CA, et al. Children’s intellectual ability is associated with structural networkintegrity. Neuroimage. 2016;124(Pt A):550-556. doi:10.1016/j.neuroimage.2015.09.012

50. Badhwar A, Tam A, Dansereau C, Orban P, Hoffstaedter F, Bellec P. Resting-state network dysfunction inAlzheimer’s disease: a systematic review and meta-analysis. Alzheimers Dement (Amst). 2017;8:73-85. doi:10.1016/j.dadm.2017.03.007

51. Deluca C, Golzar A, Santandrea E, et al. The cerebellum and visual perceptual learning: evidence from a motionextrapolation task. Cortex. 2014;58:52-71. doi:10.1016/j.cortex.2014.04.017

52. Habas C, Kamdar N, Nguyen D, et al. Distinct cerebellar contributions to intrinsic connectivity networks.J Neurosci. 2009;29(26):8586-8594. doi:10.1523/JNEUROSCI.1868-09.2009

53. Clarkson J, Herbison AE. Hypothalamic control of the male neonatal testosterone surge. Philos Trans R SocLond B Biol Sci. 2016;371(1688):20150115. doi:10.1098/rstb.2015.0115

54. Kerr JE, Beck SG, Handa RJ. Androgens modulate glucocorticoid receptor mRNA, but not mineralocorticoidreceptor mRNA levels, in the rat hippocampus. J Neuroendocrinol. 1996;8(6):439-447. doi:10.1046/j.1365-2826.1996.04735.x

55. Surico D, Ercoli A, Farruggio S, et al. Modulation of oxidative stress by 17 β-estradiol and genistein in humanhepatic cell lines in vitro. Cell Physiol Biochem. 2017;42(3):1051-1062. doi:10.1159/000478752

56. Duailibi MS, Cordeiro Q, Brietzke E, et al. N-acetylcysteine in the treatment of craving in substance usedisorders: systematic review and meta-analysis. Am J Addict. 2017;26(7):660-666. doi:10.1111/ajad.12620

57. McClure EA, Gipson CD, Malcolm RJ, Kalivas PW, Gray KM. Potential role of N-acetylcysteine in themanagement of substance use disorders. CNS Drugs. 2014;28(2):95-106. doi:10.1007/s40263-014-0142-x

58. Moda-Sava RN, Murdock MH, Parekh PK, et al. Sustained rescue of prefrontal circuit dysfunction byantidepressant-induced spine formation. Science. 2019;364(6436):eaat8078.

59. Keeser D, Meindl T, Bor J, et al. Prefrontal transcranial direct current stimulation changes connectivity ofresting-state networks during fMRI. J Neurosci. 2011;31(43):15284-15293. doi:10.1523/JNEUROSCI.0542-11.2011

60. Peña-Gómez C, Sala-Lonch R, Junqué C, et al. Modulation of large-scale brain networks by transcranial directcurrent stimulation evidenced by resting-state functional MRI. Brain Stimul. 2012;5(3):252-263. doi:10.1016/j.brs.2011.08.006




https://dx.doi.org/10.1016/S2352-3026(16)30110-7

https://dx.doi.org/10.1016/S2352-3026(16)30110-7

https://dx.doi.org/10.1006/nimg.2001.0978

https://dx.doi.org/10.1016/j.neuropsychologia.2018.10.003

https://dx.doi.org/10.1016/j.neuropsychologia.2018.10.003

https://dx.doi.org/10.1093/bioinformatics/btx050

https://dx.doi.org/10.1371/journal.pcbi.0030017

https://dx.doi.org/10.1007/s11060-016-2328-1

https://dx.doi.org/10.1371/journal.pbio.1000157

https://dx.doi.org/10.3389/fpsyt.2018.00278




https://dx.doi.org/10.1016/j.dadm.2017.03.007

https://dx.doi.org/10.1016/j.dadm.2017.03.007

https://dx.doi.org/10.1016/j.cortex.2014.04.017

https://dx.doi.org/10.1523/JNEUROSCI.1868-09.2009

https://dx.doi.org/10.1098/rstb.2015.0115

https://dx.doi.org/10.1046/j.1365-2826.1996.04735.x

https://dx.doi.org/10.1046/j.1365-2826.1996.04735.x

https://dx.doi.org/10.1159/000478752

https://dx.doi.org/10.1111/ajad.12620

https://dx.doi.org/10.1007/s40263-014-0142-x



https://dx.doi.org/10.1523/JNEUROSCI.0542-11.2011

https://dx.doi.org/10.1016/j.brs.2011.08.006

https://dx.doi.org/10.1016/j.brs.2011.08.006

61. Fiori V, Kunz L, Kuhnke P, Marangolo P, Hartwigsen G. Transcranial direct current stimulation (tDCS) facilitatesverb learning by altering effective connectivity in the healthy brain. Neuroimage. 2018;181:550-559. doi:10.1016/j.neuroimage.2018.07.040

62. Huang YJ, Lane HY, Lin CH. New treatment strategies of depression: based on mechanisms related toneuroplasticity. Neural Plast. 2017;2017:4605971. doi:10.1155/2017/4605971

63. Children’s Oncology Group. Long-term follow-up guidelines for survivors of childhood, adolescent and youngadult cancers. Updated October 2018. Accessed January 17, 2020. http://www.survivorshipguidelines.org/pdf/2018/COG_LTFU_Guidelines_v5.pdf

64. Patel RS, Rachamalla M, Chary NR, Shera FY, Tikoo K, Jena G. Cytarabine induced cerebellar neuronal damagein juvenile rat: correlating neurobehavioral performance with cellular and genetic alterations. Toxicology. 2012;293(1-3):41-52. doi:10.1016/j.tox.2011.12.005

65. Rubinov M, Sporns O. Complex network measures of brain connectivity: uses and interpretations.Neuroimage. 2010;52(3):1059-1069. doi:10.1016/j.neuroimage.2009.10.003

SUPPLEMENT.eAppendix. Supplemental Methods






https://dx.doi.org/10.1155/2017/4605971

http://www.survivorshipguidelines.org/pdf/2018/COG_LTFU_Guidelines_v5.pdf

http://www.survivorshipguidelines.org/pdf/2018/COG_LTFU_Guidelines_v5.pdf

https://dx.doi.org/10.1016/j.tox.2011.12.005


Connectivity of the Cerebello-Thalamo-Cortical Pathway in … · PharmacologicalAnalysis...

Documents

Transcript of Connectivity of the Cerebello-Thalamo-Cortical Pathway in … · PharmacologicalAnalysis...