Search

Supplementary Excel Tables

Supplementary Tables and Figures

Genomic characterization of pediatric patients with acute myeloid leukemia (AML) has led to the discovery of somatic mutations with prognostic implications. Although gene-expression profiling can differentiate subsets of pediatric AML, its clinical utility in risk stratification remains limited. Here, we evaluate gene expression, pathogenic somatic mutations, and outcome in a cohort of 435 pediatric patients with a spectrum of pediatric myeloid-related acute leukemias for biological subtype discovery. This analysis revealed 63 patients with varying immunophenotypes that span a T-lineage and myeloid continuum designated as acute myeloid/T-lymphoblastic leukemia (AMTL). Within AMTL, two patient subgroups distinguished by FLT3-ITD and PRC2 mutations have different outcomes, demonstrating the impact of mutational composition on survival. Across the cohort, variability in outcomes of patients within isomutational subsets is influenced by transcriptional identity and the presence of a stem cell–like gene-expression signature. Integration of gene expression and somatic mutations leads to improved risk stratification.Significance:Immunophenotype and somatic mutations play a significant role in treatment approach and risk stratification of acute leukemia. We conducted an integrated genomic analysis of pediatric myeloid malignancies and found that a combination of genetic and transcriptional readouts was superior to immunophenotype and genomic mutations in identifying biological subtypes and predicting outcomes.This article is highlighted in the In This Issue feature, p. 549

Figure S1: Age distribution of the cohort. Figure S2: The distribution of bleomycin 3rd tertile. Figure S3: Distribution of cancer types treated without alkylating agents, platinum, anthracyclines, bleomycin, dactinomycin, epipodophyllotoxins, methotrexate, vinca alkaloids. Figure S4: Conceptual illustration for age- and therapy-related clonal hematopoiesis. Figure S5: Distribution of cell phenotypes harboring STAT3 Y640F mutation. Figure S6: Regression analysis on public dataset to reproduce the clone expansion pattern in main Figure 4A. Figure S7: Longitudinal view of the serial samples. Figure S8: Simulation analysis to reproduce the pattern in main Figure 4B. Figure S9: Comparison of clone growth rates and clone sizes between age- and therapy-related CH. Figure S10: Additional sequencing for PPM1D. Table S1: Gene panel. Table S3: Comparison between CH and sex, race and age. Table S4: Comparison between CH status and pathogenic germline mutations. Separately analyzed for cancer predisposition genes and DNA damage repair genes. Table S5: Enrichment of STAT3 mutations in Hodgkin lymphoma survivors. Table S6: Association between procarbazine exposure and mutation pattern in Hodgkin lymphoma survivors.

Table S2: List of CH mutations identified. Table S7: Mutations analyzed in serial analysis. Table S8: Clinical features of the survivors. Table S9: List of primers used in ddPCR validation.

We present the first comprehensive investigation of clonal hematopoiesis (CH) in 2,860 long-term survivors of pediatric cancer with a median follow-up time of 23.5 years. Deep sequencing over 39 CH-related genes reveals mutations in 15% of the survivors, significantly higher than the 8.5% in 324 community controls. CH in survivors is associated with exposures to alkylating agents, radiation, and bleomycin. Therapy-related CH shows significant enrichment in STAT3, characterized as a CH gene specific to survivors of Hodgkin lymphoma, and TP53. Single-cell profiling of peripheral blood samples revealed STAT3 mutations predominantly present in T cells and contributed by SBS25, a mutational signature associated with procarbazine exposure. Serial sample tracking reveals that larger clone size is a predictor for future expansion of age-related CH clones, whereas therapy-related CH remains stable decades after treatment. These data depict the distinct dynamics of these CH subtypes and support the need for longitudinal monitoring to determine the potential contribution to late effects. Significance: This first comprehensive CH analysis in long-term survivors of pediatric cancer presents the elevated prevalence and therapy exposures/diagnostic spectrum associated with CH. Due to the contrasting dynamics of clonal expansion for age-related versus therapy-related CH, longitudinal monitoring is recommended to ascertain the long-term effects of therapy-induced CH in pediatric cancer survivors. See related commentary by Collord and Behjati, p. 811. This article is highlighted in the In This Issue feature, p. 799

We present the first comprehensive investigation of clonal hematopoiesis (CH) in 2,860 long-term survivors of pediatric cancer with a median follow-up time of 23.5 years. Deep sequencing over 39 CH-related genes reveals mutations in 15% of the survivors, significantly higher than the 8.5% in 324 community controls. CH in survivors is associated with exposures to alkylating agents, radiation, and bleomycin. Therapy-related CH shows significant enrichment in STAT3, characterized as a CH gene specific to survivors of Hodgkin lymphoma, and TP53. Single-cell profiling of peripheral blood samples revealed STAT3 mutations predominantly present in T cells and contributed by SBS25, a mutational signature associated with procarbazine exposure. Serial sample tracking reveals that larger clone size is a predictor for future expansion of age-related CH clones, whereas therapy-related CH remains stable decades after treatment. These data depict the distinct dynamics of these CH subtypes and support the need for longitudinal monitoring to determine the potential contribution to late effects.Significance:This first comprehensive CH analysis in long-term survivors of pediatric cancer presents the elevated prevalence and therapy exposures/diagnostic spectrum associated with CH. Due to the contrasting dynamics of clonal expansion for age-related versus therapy-related CH, longitudinal monitoring is recommended to ascertain the long-term effects of therapy-induced CH in pediatric cancer survivors.See related commentary by Collord and Behjati, p. 811.This article is highlighted in the In This Issue feature, p. 799

To investigate the genomic evolution of metastatic pediatric osteosarcoma, we performed whole-genome and targeted deep sequencing on 14 osteosarcoma metastases and two primary tumors from four patients (two to eight samples per patient). All four patients harbored ancestral (truncal) somatic variants resulting in TP53 inactivation and cell-cycle aberrations, followed by divergence into relapse-specific lineages exhibiting a cisplatin-induced mutation signature. In three of the four patients, the cisplatin signature accounted for >40% of mutations detected in the metastatic samples. Mutations potentially acquired during cisplatin treatment included NF1 missense mutations of uncertain significance in two patients and a KIT G565R activating mutation in one patient. Three of four patients demonstrated widespread ploidy differences between samples from the sample patient. Single-cell seeding of metastasis was detected in most metastatic samples. Cross-seeding between metastatic sites was observed in one patient, whereas in another patient a minor clone from the primary tumor seeded both metastases analyzed. These results reveal extensive clonal heterogeneity in metastatic osteosarcoma, much of which is likely cisplatin-induced.Implications:The extent and consequences of chemotherapy-induced damage in pediatric cancers is unknown. We found that cisplatin treatment can potentially double the mutational burden in osteosarcoma, which has implications for optimizing therapy for recurrent, chemotherapy-resistant disease.

Osteosarcoma clinical history and capture validation sequencing data. Supplementary Table S1. Sample and treatment information for relapsed osteosarcoma patients. Supplementary Table S2. SJOS0011101 SNVs detected by capture validation and tumor purity. Supplementary Table S3. SJOS0011105 SNVs detected by capture validation and tumor purity. Supplementary Table S4. SJOS0011107 SNVs detected by capture validation and tumor purity. Supplementary Table S5. SJOS010 SNVs detected by capture validation and tumor purity.

Background Adult survivors of childhood cancer are at increased risk of cardiac late effects. Methods Using whole-genome sequencing data from 1870 survivors of European ancestry in the St. Jude Lifetime Cohort (SJLIFE) study, genetic variants were examined for association with ejection fraction (EF) and clinically assessed cancer therapy–induced cardiac dysfunction (CCD). Statistically significant findings were validated in 301 SJLIFE survivors of African ancestry and 4020 survivors of European ancestry from the Childhood Cancer Survivor Study. All statistical tests were 2-sided. Results A variant near KCNK17 showed genome-wide significant association with EF (rs2815063-A: EF reduction = 1.6%; P = 2.1 × 10-8) in SJLIFE survivors of European ancestry, which replicated in SJLIFE survivors of African ancestry (EF reduction = 1.5%; P = .004). The rs2815063-A also showed a 1.80-fold (P = .008) risk of severe or disabling or life-threatening CCD and replicated in 4020 Childhood Cancer Survivor Study survivors of European ancestry (odds ratio = 1.40; P = .04). Notably, rs2815063-A was specifically associated among survivors exposed to doxorubicin only, with a stronger effect on EF (3.3% EF reduction) and CCD (2.97-fold). Whole blood DNA methylation data in 1651 SJLIFE survivors of European ancestry showed statistically significant correlation of rs2815063-A with dysregulation of KCNK17 enhancers (false discovery rate Conclusions Leveraging the 2 largest cohorts of childhood cancer survivors in North America and survivor-specific polygenomic functional data, we identified a novel risk locus for CCD, which showed specificity with doxorubicin-induced cardiac dysfunction and highlighted dysregulation of KCNK17 as the likely molecular mechanism underlying this genetic association.

Supplementary Fig. S1-S3, Supplementary Table S1-S9 Supplementary Fig. S1. A flow diagram of the study population. Supplementary Fig. S2. Dynamics of telomere length differs between individuals with no history of cancer and survivors of childhood cancer due to exposures to chemotherapy and/or radiation treatment. Supplementary Fig. S3. Theoretical framework of relationships among telomere length, health behaviors and chronic health conditions.

Purpose:We aimed to analyze and compare leukocyte telomere length (LTL) and age-dependent LTL attrition between childhood cancer survivors and noncancer controls, and to evaluate the associations of LTL with treatment exposures, chronic health conditions (CHC), and health behaviors among survivors.Experimental Design:We included 2,427 survivors and 293 noncancer controls of European ancestry, drawn from the participants in St. Jude Lifetime Cohort Study (SJLIFE), a retrospective hospital-based study with prospective follow-up (2007–2016). Common nonneoplastic CHCs (59 types) and subsequent malignant neoplasms (5 types) were clinically assessed. LTL was measured with whole-genome sequencing data.Results:After adjusting for age at DNA sampling, gender, genetic risk score based on 9 SNPs known to be associated with telomere length, and eigenvectors, LTL among survivors was significantly shorter both overall [adjusted mean (AM) = 6.20 kb; SE = 0.03 kb] and across diagnoses than controls (AM = 6.69 kb; SE = 0.07 kb). Among survivors, specific treatment exposures associated with shorter LTL included chest or abdominal irradiation, glucocorticoid, and vincristine chemotherapies. Significant negative associations of LTL with 14 different CHCs, and a positive association with subsequent thyroid cancer occurring out of irradiation field were identified. Health behaviors were significantly associated with LTL among survivors aged 18 to 35 years (Ptrend = 0.03).Conclusions:LTL is significantly shorter among childhood cancer survivors than noncancer controls, and is associated with CHCs and health behaviors, suggesting LTL as an aging biomarker may be a potential mechanistic target for future intervention studies designed to prevent or delay onset of CHCs in childhood cancer survivors.See related commentary by Walsh, p. 2281

Supplemental Figure Legend

Studies focusing on genetic susceptibility of childhood Hodgkin lymphoma (HL) are limited.To identify genetic variants associated with childhood-onset HL vs adult-onset HL.This genetic association study was performed with 3 cohorts: the St Jude Lifetime Cohort Study (SJLIFE), initiated in 2007 with ongoing follow-up, and the original and expansion cohorts of the Childhood Cancer Survivor Study (CCSS), initiated in the 1990s with ongoing follow-up. Results of these genome-wide association studies (GWASs) were combined via meta-analysis. Data were analyzed from June 2021 to June 2022.Childhood HL was the focused outcome. Single-nucleotide variant (SNV, formerly single-nucleotide polymorphism) array genotyping and imputation were conducted for the CCSS original cohort, and whole-genome sequencing was performed for the SJLIFE and CCSS expansion cohort.A total of 1286 HL cases (mean diagnosis [SD] age, 14.6 [3.9] years), 6193 non-HL childhood cancer cases, and 369 noncancer controls, all of European ancestry, were included in the analysis. Using step-wise conditional logistic regression, the odds ratios (ORs) for each of the 3 independent SNVs identified in the human leukocyte antigen (HLA) locus were 1.80 (95% CI, 1.59-2.03; P = 2.14 × 10-21) for rs28383311, 1.53 (95% CI, 1.37-1.70; P = 2.05 × 10-14) for rs3129198, and 1.51 (95% CI, 1.35-1.69; P = 6.21 × 10-13) for rs3129890. Further HLA imputation revealed 9 alleles and 55 amino acid changes that potentially conferred HL susceptibility. In addition, 5 non-HLA loci were identified: (1) rs1432297 (OR, 1.29; 95% CI, 1.18-1.41; P = 2.5 × 10-8; r2 = 0.55; D' = 0.75 with previously reported rs1432295, REL); (2) rs2757647 (OR, 1.30; 95% CI, 1.18-1.42; P = 3.5 × 10-8; r2 = 0.59; D' = 0.83 with previously reported rs6928977, AHI1); (3) rs13279159 (OR, 1.33; 95% CI, 1.20-1.47; P = 1.7 × 10-8; r2 = 0.75; D' = 1.00 with previously reported rs2019960, PVT1); (4) rs3824662 (OR, 1.52; 95% CI, 1.33-1.73; P = 3.9 × 10-10; r2 = 0.91; D' = 1.00 with previously reported rs3781093, GATA3); and (5) rs117953624 (OR, 1.98; 95% CI, 1.56-2.51; P = 1.5 × 10-8; minor allele frequency, 0.02), a novel uncommon SNV mapped to PDGFD. Twelve of 18 previously reported genome-wide significant non-HLA SNVs (67%) were replicated with statistically significant results.In this genetic association study, a predominantly common and potentially unique genetic etiology was found between childhood-onset and adulthood-onset HL.

Our assessment of epigenetic age acceleration (EAA), calculated with DNA methylation (DNAm) data generated from bulk DNA derived from peripheral blood mononuclear cells (PBMC), supports accelerated aging in childhood cancer survivors (CCS). It is challenging to disentangle variation of DNAm at cell-type specific levels from the effects of age-dependent cell type composition, and bulk measurements may obfuscate the links between EAA and age-related outcomes (e.g., cardiomyopathy). Methylation profiling was generated using Infinium EPIC BeadChips on PBMC-derived DNA from CCS in the St. Jude Lifetime Cohort. Tensor composition analysis was employed to deconvolute bulk DNAm and infer DNAm at each leukocyte subtype level, i.e., a single n (individuals) by m (DNAm sites) matrix of observed DNAm data was deconvoluted into multiple n by m matrices of DNAm data. Epigenetic age (EA, using Levine’s clock) and EAA (residuals from the fit of a simple linear regression of EA on chronological age at blood draw) were calculated for bulk PBMCs and each cell subtype (CD4T, CD8T, B, natural killer, and monocyte). Cardiomyopathy (CMP) was assessed by echocardiography, and severity graded (2 = moderate, 3 = severe/disabling, 4 = life-threatening and 5 = fatal) using a modified version of the National Cancer Institute Common Terminology Criteria for Adverse Events. Cumulative doses of anthracyclines and mean heart radiation doses (heart-RT) calculated through radiation dosimetry were abstracted from medical records. Associations between EAA and CMP were evaluated by multivariable Cox regression. Cell-type specific EA was highly correlated with bulk EA with Pearson r2 between 0.63 (CD8T) and 0.79 (CD4T), but the linear regression lines of cell-type specific EA against bulk EA differed in both intercept and slope, suggesting heterogeneity across leukocyte subtypes. Cell-type specific EAA was moderately correlated with bulk EAA with Pearson r2 between 0.23 (monocyte) and 0.29 (CD8T). Among 2,044 CCS (median age = 33.7 years), 104 (5.1%) developed CMP (≥grade 3). Among bulk EAA and five cell-type specific EAAs, monocyte EAA was the only one significantly associated with CMP (hazard ratio per standard deviation increase in EAA = 1.25, 95% CI = 1.04-1.50, P = 0.018). Cell sorting of PBMC followed by DNAm is currently underway for further validation. We showed an in-depth view of the variability of EAA across leukocyte subtypes, and more importantly, demonstrated that monocyte EAA was associated with CMP risk. Our novel finding is plausible and consistent with the literature implicating monocyte-derived cardiac macrophages in cardiac remodeling, which can be induced by cardiotoxic cancer treatment exposures in CCS. Therapeutic strategies that prevent deleterious effects of monocytes contributing to adverse cardiac remodeling, while sparing their essential immune functions, may prevent or ameliorate CMP among CCS. Citation Format: Cheng Chen, Qian Dong, Qin Na, Nan Song, John Easton, Heather L. Mulder, Emily Walker, Geoffrey Neale, Emily R. Finch, Qian Li, Yutaka Yasui, Daniel A. Mulrooney, Melissa M. Hudson, Kirsten K. Ness, Jinghui Zhang, Xiang Chen, Hui Wang, Leslie L. Robison, Zhaoming Wang. Monocyte-specific epigenetic age acceleration and cardiomyopathy risk among survivors of childhood cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4363.

We previously showed that cancer treatment associated DNA methylation (DNAm) signatures were present decades following the cancer diagnosis in childhood cancer survivors (CCS) of European ancestry (EA) and that treatment associated DNAm sites mediated the causal pathway from specific treatment exposures to increased risk of chronic health conditions (CHCs). This new analysis further evaluated and compared the treatment and DNAm associations in CCS of African ancestry (AA) from the St. Jude Lifetime Cohort Study. Cancer treatments were abstracted from medical records. DNAm was measured using MethylationEPIC BeadChip with blood-derived DNA. Among 370 AA CCS (53.2% female, median age at blood draw=31.2 [range=18.4-65.1] years), treatments included alkylating agents (54.6%), anthracyclines (48.6%), epipodophyllotoxins (29.2%), corticosteroids (33.0%), and vincristine (61.6%), and radiation therapy (RT) to brain (26.5%), chest (27.6%), abdomen (22.7%), and pelvis (21.6%). Epigenome-wide association study (EWAS) for each treatment, using multivariable linear regression adjusting for sex, age at blood draw, leukocyte cell subtype proportions, genotype principal components and DNAm principal components, showed little inflation with genomic control (GC) factor between 1.1 (brain-RT) and 1.2 (chest-RT). A total of 93 5′-cytosine-phosphate-guanine-3′ (CpG) was associated with one or more cancer treatments (GC-adjusted P Citation Format: Qian Dong, Sarmistha Das, Cheng Chen, John Easton, Heather L. Mulder, Emily Walker, Geoffrey Neale, Deo Kumar Srivastava, I-Chan Huang, Jinghui Zhang, Melissa M. Hudson, Leslie L. Robison, Kirsten K. Ness, Nan Song, Zhaoming Wang. Variations of blood DNA methylation associated with cancer treatment exposures among childhood cancer survivors of African ancestry. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4514.

Background Adult childhood cancer survivors are at risk for frailty, including low muscle mass and weakness (sarcopenia). Using peripheral blood mitochondrial DNA copy number (mtDNAcn) as a proxy for functional mitochondria, this study describes cross-sectional associations between mtDNAcn and sarcopenia among survivors. Methods Among 1762 adult childhood cancer survivors (51.6% male; median age = 29.4 years, interquartile range [IQR] = 23.3-36.8), with a median of 20.6 years from diagnosis (IQR = 15.2-28.2), mtDNAcn estimates were derived from whole-genome sequencing. A subset was validated by quantitative polymerase chain reaction and evaluated cross-sectionally using multivariable logistic regression for their association with sarcopenia, defined by race-, age-, and sex-specific low lean muscle mass or weak grip strength. All statistical tests were 2-sided. Results The prevalence of sarcopenia was 27.0%, higher among female than male survivors (31.5% vs 22.9%; P Conclusions A growing body of evidence supports peripheral blood mtDNAcn as a biomarker for adverse health outcomes; however, this study is the first to report an association between mtDNAcn and sarcopenia among childhood cancer survivors.

Background Mounting evidence supports the occurrence of accelerating aging among long-term survivors of childhood cancer. We aimed to investigate epigenetic age acceleration (EAA) in survivors and evaluate associations between EAA, treatment exposures, health behaviors, and chronic health conditions (CHCs). Methods Genome-wide methylation data were generated with Infinium EPIC BeadChip on blood-derived DNA from 2139 survivors and 282 frequency matched controls from the St Jude Lifetime Cohort Study. EAAs were estimated as residuals from a linear regression of epigenetic age (Levine’s clock) against chronological age. Adjusted least square mean (ALSM) of EAA was calculated and compared between survivors and controls, across treatment exposures and health behaviors. Associations of EAA with 20 clinically assessed CHCs were evaluated with multivariable piecewise-exponential models. All statistical tests for P values below were 2-sided. Results EAA was statistically significantly higher in survivors than controls (ALSM = 0.63, 95% confidence interval [CI] = 0.26 to 1.01 vs −3.61, 95% CI = −4.43 to 2.80). In a multivariable model among survivors, statistically significantly higher EAA (P Conclusions EAA is statistically significantly higher in survivors of childhood cancer than in noncancer controls and is associated with specific treatment exposures, unfavorable health behaviors, and presence of specific CHCs.

PURPOSE To investigate cancer treatment plus pathogenic germline mutations (PGMs) in DNA repair genes (DRGs) for identification of childhood cancer survivors at increased risk of subsequent neoplasms (SNs). METHODS Whole-genome sequencing was performed on blood-derived DNA from survivors in the St Jude Lifetime Cohort. PGMs were evaluated in 127 genes from 6 major DNA repair pathways. Cumulative doses of chemotherapy and body region–specific radiotherapy (RT) were abstracted from medical records. Relative rates (RRs) and 95% CIs of SNs by mutation status were estimated using multivariable piecewise exponential models. RESULTS Of 4,402 survivors, 495 (11.2%) developed 1,269 SNs. We identified 538 PGMs in 98 DRGs ( POLG, MUTYH, ERCC2, and BRCA2, among others) in 508 (11.5%) survivors. Mutations in homologous recombination (HR) genes were significantly associated with an increased rate of subsequent female breast cancer (RR, 3.7; 95% CI, 1.8 to 7.7), especially among survivors with chest RT ≥ 20 Gy (RR, 4.4; 95% CI, 1.6 to 12.4), or with a cumulative dose of anthracyclines in the second or third tertile (RR, 4.4; 95% CI, 1.7 to 11.4). Mutations in HR genes were also associated with an increased rate of subsequent sarcoma among those who received alkylating agent doses in the third tertile (RR, 14.9; 95% CI, 4.0 to 38.0). Mutations in nucleotide excision repair genes were associated with subsequent thyroid cancer for those treated with neck RT ≥ 30 Gy (RR, 12.9; 95% CI, 1.6 to 46.6) with marginal statistical significance. CONCLUSION Our study provides novel insights regarding the contribution of genetics, in combination with known treatment-related risks, for the development of SNs. These findings have the potential to facilitate identification of high-risk survivors who may benefit from genetic counseling and/or testing of DRGs, which may further inform personalized cancer surveillance and prevention strategies.

To study the mechanisms of relapse in acute lymphoblastic leukemia (ALL), we performed whole-genome sequencing of 103 diagnosis-relapse-germline trios and ultra-deep sequencing of 208 serial samples in 16 patients. Relapse-specific somatic alterations were enriched in 12 genes (NR3C1, NR3C2, TP53, NT5C2, FPGS, CREBBP, MSH2, MSH6, PMS2, WHSC1, PRPS1, and PRPS2) involved in drug response. Their prevalence was 17% in very early relapse (36 months) groups. Convergent evolution, in which multiple subclones harbor mutations in the same drug resistance gene, was observed in 6 relapses and confirmed by single-cell sequencing in 1 case. Mathematical modeling and mutational signature analysis indicated that early relapse resistance acquisition was frequently a 2-step process in which a persistent clone survived initial therapy and later acquired bona fide resistance mutations during therapy. In contrast, very early relapses arose from preexisting resistant clone(s). Two novel relapse-specific mutational signatures, one of which was caused by thiopurine treatment based on in vitro drug exposure experiments, were identified in early and late relapses but were absent from 2540 pan-cancer diagnosis samples and 129 non-ALL relapses. The novel signatures were detected in 27% of relapsed ALLs and were responsible for 46% of acquired resistance mutations in NT5C2, PRPS1, NR3C1, and TP53. These results suggest that chemotherapy-induced drug resistance mutations facilitate a subset of pediatric ALL relapses.

Background: Subsequent thyroid cancer (STC) is one of the most common malignancies in childhood cancer survivors. We aimed to evaluate the polygenic contributions to STC risk and potential utility in improving risk prediction. Methods: A polygenic risk score (PRS) was calculated from 12 independent SNPs associated with thyroid cancer risk in the general population. Associations between PRS and STC risk were evaluated among survivors from St. Jude Lifetime Cohort (SJLIFE) and were replicated in survivors from Childhood Cancer Survivor Study (CCSS). A risk prediction model integrating the PRS and clinical factors, initially developed in SJLIFE, and its performance were validated in CCSS. Results: Among 2,370 SJLIFE survivors with a median follow-up of 28.8 [interquartile range (IQR) = 21.9–36.1] years, 65 (2.7%) developed STC. Among them, the standardized PRS was associated with an increased rate of STC [relative rate (RR) = 1.57; 95% confidence interval (CI) = 1.24–1.98; P < 0.001]. Similar associations were replicated in 6,416 CCSS survivors, among whom 121 (1.9%) developed STC during median follow-up of 28.9 (IQR = 22.6–34.6) years (RR = 1.52; 95% CI = 1.25–1.83; P < 0.001). A risk prediction model integrating the PRS with clinical factors showed better performance than the model considering only clinical factors in SJLIFE (P = 0.004, AUC = 83.2% vs. 82.1%, at age 40), which was further validated in CCSS (P = 0.010, AUC = 72.9% vs. 70.6%). Conclusions: Integration of the PRS with clinical factors provided a statistically significant improvement in risk prediction of STC, although the magnitude of improvement was modest. Impact: PRS improves risk stratification and prediction of STC, suggesting its potential utility for optimizing screening strategies in survivorship care.

Spitzoid melanoma is a specific morphologic variant of melanoma that most commonly affects children and adolescents, and ranges on the spectrum of malignancy from low grade to overtly malignant. These tumors are generally driven by fusions of ALK, RET, NTRK1/3, MET, ROS1 and BRAF1,2. However, in approximately 50% of cases no genetic driver has been established2. Clinical whole-genome and transcriptome sequencing (RNA-Seq) of a spitzoid tumor from an adolescent revealed a novel gene fusion of MAP3K8, encoding a serine-threonine kinase that activates MEK3,4. The patient, who had exhausted all other therapeutic options, was treated with a MEK inhibitor and underwent a transient clinical response. We subsequently analyzed spitzoid tumors from 49 patients by RNA-Seq and found in-frame fusions or C-terminal truncations of MAP3K8 in 33% of cases. The fusion transcripts and truncated genes all contained MAP3K8 exons 1–8 but lacked the autoinhibitory final exon. Data mining of RNA-Seq from the Cancer Genome Atlas (TCGA) uncovered analogous MAP3K8 rearrangements in 1.5% of adult melanomas. Thus, MAP3K8 rearrangements—uncovered by comprehensive clinical sequencing of a single case—are the most common genetic event in spitzoid melanoma, are present in adult melanomas and could be amenable to MEK inhibition. Rearrangements in MAP3K8 respond to MEK inhibition and represent the most common genetic driver in pediatric spitzoid melanoma, and in some adult melanomas lacking other MAPK alterations and for which clinical testing is warranted.

Survivors of childhood cancer are at risk for a spectrum of adverse outcomes including hematological malignancies and cardiovascular diseases, which in other populations have been associated with clonal hematopoiesis (CH), an age-related clonal expansion of hematopoietic stem cells. Here we present the first comprehensive CH analysis including 2,872 pediatric cancer survivors from the St. Jude Lifetime Cohort (median follow-up of 26 years from cancer diagnosis; median age of 31 years at blood draw) using deep sequencing (10,000 ×) over 39 CH genes on DNA derived from peripheral blood. As a comparison group, 324 age-, sex- and race/ethnicity-frequency-matched community controls (median age of 35 years at blood draw) were also analyzed. To detect rare CH variants, a machine-learning approach was applied, which characterized true variants as outliers in a genomic context-specific error profile. Digital droplet PCR validation was performed for 543 variants (323 positive [allele frequency: median 0.003, range 0.001 to 0.255] and 220 negative) out of 1,450 putative variants. The 907 untested variants were inferred using the 543 experimentally validated variants’ status as a training set. Analyses considered mutation status by the 555 positive variants and clinical data including age, sex, and detailed cancer treatment exposures. CH was detected in 15% (95% Confidence Interval (CI): 13.7 to 16.3) of survivors and 8.6% (95% CI: 5.6 to 11.7) of controls in an age-dependent manner: 10.6% (survivors) vs. 6.3% (controls) for ages in year 18 - 29; 13.5% vs. 5.8% for 30 - 39; 26.3% vs. 13.2% for 40 - 49; 28.1% vs. 17.2% for > 50. We also observed that the alterations in DNMT3A, TP53, STAT3, KRAS and TET2 became more prevalent as the survivors aged: combined proportion for the 5 genes, 61% for ages in year < 30; 72% for 30 - 40; 77% for 40 - 50; 87% for > 50. Analysis of longitudinal samples available for 73 survivors with 180 time points (median interval 4.6 years [range: 2.07 to 8.0 years]) confirmed that clones with these gene mutations had a higher growth rate than those without (p=0.034), suggesting fitness advantages in hematopoietic stem cell context. Multivariable analysis adjusted for treatment indicated that CH development is associated with cumulative dose of alkylating agents (Odds Ratio (OR) for 1st tertile 1.77 [95% CI: 1.22 to 2.57], 2nd tertile 2.72 [95% CI: 1.83 to 4.04], and 3rd tertile 2.92 [95% CI: 2.03 to 4.02], relative to the unexposed) and estimated radiotherapy-dose to active bone marrow (OR for 1st tertile 0.98 [95% CI: 0.68 to 1.40], 2nd tertile 1.06 [0.77 to 1.47], and 3rd tertile 1.57 [1.15 to 2.14], relative to the unexposed), supporting the hypothesis that CH development in survivors is primarily driven by prior therapy. This study identified accelerated CH development in pediatric cancer survivors, which may represent an important genomic biomarker predictive of future adverse health risks. Citation Format: Kohei Hagiwara, Zhaoming Wang, Haseeb Zubair, John Easton, Heather L. Mulder, Xiaotu Ma, Kristen K. Ness, Zhenghong Li, Daniel A. Mulrooney, Carmen L. Wilson, Yutaka Yasui, Melissa M. Hudson, Leslie L. Robison, Jinghui Zhang. Accelerated clonal hematopoiesis in survivors of childhood cancer: A report for the St. Jude Lifetime Cohort Study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 499.

Long Interspersed Element-1 (LINE-1 or L1) is the only autonomously active mobile genetic element in the human genome. It has been shown to mobilize (i.e., retrotranspose) in cancers, causing de novo somatic L1 insertions. Somatic L1 retrotransposition can cause aberrant splicing of tumor suppressor genes by utilizing cryptic splice donor (SD) sites such as the nucleotide 97 SD in the L1 5’UTR or disrupting repressive regulatory regions of proto-oncogenes, inducing aberrant activation, leading to cancer. In contrast to these reported mechanisms, here we report the first evidence of a driver oncogene activation due to somatic acquisition of an L1 promoter in a pediatric high-grade glioma (HGG). While investigating activation of FOXR2 (forkhead box R2), a known oncogene in CNS embryonal tumors, in a recurrent pediatric diffuse HGG that showed a methylation profile of CNS FOXR2-activated tumors, we identified overexpression of a non-canonical FOXR2 isoform. Transcription of FOXR2 initiated from ~11kb upstream of exon 2 at a site harboring a somatic structural variation (SV) reminiscent of an L1 insertion. Targeted PCR amplification and subsequent PacBio sequencing revealed a 5’ inverted/deleted L1 insertion (~3kb sequence) containing an intact L1 5’UTR, presence of a 3’ transduction sequence, two poly-A tails, and target site duplications. The 3’ transduction sequence matches the sequence downstream an intact polymorphic full-length L1 element at 6p24.1, the likely source L1 element that initiated the FOXR2 somatic insertion. Strikingly, tumor RNA-seq data reveals FOXR2 transcription initiated from the intact sense pol-II promoter of the inserted L1 5’UTR, generating a fusion transcript joining the L1 5’UTR 97 SD to a canonical splice acceptor within exon 2 of FOXR2. Targeted bisulfite sequencing of the FOXR2 L1 5’UTR revealed a hypomethylated state, further supporting its role in FOXR2 transcription initiation. FOXR2 activation is likely a tumor initiating event as the following was observed in the primary tumor that occurred two years prior: a matching recurrent tumor methylation profile, the presence of L1/FOXR2 fusion transcripts, and overexpression of FOXR2. Additional driver mutations, clonal TP53 R175H mutation and PDGFRA amplification, identified in the recurrent tumor but absent in the primary tumor, support the somatic L1 insertion as a tumor initiating event occurring prior to later acquired alterations. Our data represent the first example of oncogenic activation by L1 “promoter donation” via somatic retrotransposition as a novel tumor initiating mechanism. Future large-scale research and development of new computational methods for detecting rearranged L1 insertions are required to determine the prevalence of L1 “promoter donation” in cancer or other diseases. Citation Format: Xiaolong Chen, Diane A. Flasch, Bensheng Ju, Heather L. Mulder, John Easton, Lu Wang, Suzanne J. Baker, Jason Chiang, Jinghui Zhang. Oncogenic activation of FOXR2 driven by somatic acquisition of a LINE-1 promoter in pediatric high-grade glioma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1462.

Individuals with monogenic disorders can experience variable phenotypes that are influenced by genetic variation. To investigate this in sickle cell disease (SCD), we performed whole-genome sequencing (WGS) of 722 individuals with hemoglobin HbSS or HbSβ0-thalassemia from Baylor College of Medicine and from the St. Jude Children’s Research Hospital Sickle Cell Clinical Research and Intervention Program (SCCRIP) longitudinal cohort study. We developed pipelines to identify genetic variants that modulate sickle hemoglobin polymerization in red blood cells and combined these with pain-associated variants to build a polygenic score (PGS) for acute vaso-occlusive pain (VOP). Overall, we interrogated the α-thalassemia deletion −α3.7 and 133 candidate single-nucleotide polymorphisms (SNPs) across 66 genes for associations with VOP in 327 SCCRIP participants followed longitudinally over 6 years. Twenty-one SNPs in 9 loci were associated with VOP, including 3 (BCL11A, MYB, and the β-like globin gene cluster) that regulate erythrocyte fetal hemoglobin (HbF) levels and 6 (COMT, TBC1D1, KCNJ6, FAAH, NR3C1, and IL1A) that were associated previously with various pain syndromes. An unweighted PGS integrating all 21 SNPs was associated with the VOP event rate (estimate, 0.35; standard error, 0.04; P = 5.9 × 10−14) and VOP event occurrence (estimate, 0.42; standard error, 0.06; P = 4.1 × 10−13). These associations were stronger than those of any single locus. Our findings provide insights into the genetic modulation of VOP in children with SCD. More generally, we demonstrate the utility of WGS for investigating genetic contributions to the variable expression of SCD-associated morbidities.

Integrating somatic mutation analysis and gene expression profiling distinguishes pediatric AML subtypes with differential prognoses and clinical risks., Genomic characterization of pediatric patients with acute myeloid leukemia (AML) has led to the discovery of somatic mutations with prognostic implications. Although gene-expression profiling can differentiate subsets of pediatric AML, its clinical utility in risk stratification remains limited. Here, we evaluate gene expression, pathogenic somatic mutations, and outcome in a cohort of 435 pediatric patients with a spectrum of pediatric myeloid-related acute leukemias for biological subtype discovery. This analysis revealed 63 patients with varying immunophenotypes that span a T-lineage and myeloid continuum designated as acute myeloid/T-lymphoblastic leukemia (AMTL). Within AMTL, two patient subgroups distinguished by FLT3-ITD and PRC2 mutations have different outcomes, demonstrating the impact of mutational composition on survival. Across the cohort, variability in outcomes of patients within isomutational subsets is influenced by transcriptional identity and the presence of a stem cell–like gene-expression signature. Integration of gene expression and somatic mutations leads to improved risk stratification. Significance: Immunophenotype and somatic mutations play a significant role in treatment approach and risk stratification of acute leukemia. We conducted an integrated genomic analysis of pediatric myeloid malignancies and found that a combination of genetic and transcriptional readouts was superior to immunophenotype and genomic mutations in identifying biological subtypes and predicting outcomes. This article is highlighted in the In This Issue feature, p. 549

BackgroundEmerging evidence suggests that social determinants of health (SDOH) may influence health and wellness through an epigenetic mechanism in the general population. However, the social epigenomic approach has not yet been applied to survivors of childhood cancer, a vulnerable population with elevated risk for chronic health conditions (CHCs).MethodsStudy participants were drawn from the St Jude Lifetime Cohort, a hospital-based retrospective cohort with prospective follow up. DNA methylation (DNAm) profiling was generated based on blood derived DNA collected during follow-up visit. SDOH included educational attainment, personal income, and area deprivation index (ADI) based on baseline or follow-up questionnaires and geocoding. CHCs were clinically assessed with severity grade.ResultsWe included 258 childhood cancer survivors of African ancestry (AA) (median time from diagnosis=25.2 years, interquartile range [IQR]=19.9-32.1 years) and 1,618 survivors of European ancestry (EA) (median time from diagnosis=27.3, IQR=21.1-33.7 years). Through epigenome-wide association studies, we identified 130 SDOH-CpG associations including educational attainment (N=88), personal income (N=23), and ADI (N=19) at epigenome-wide significance level (P−8). There were 13 CpGs, commonly associated with all three SDOH factors, with attenuated remaining effect sizes (36.8-48.3%) after additionally adjusting body mass index and smoking, mapped to smoking-related genes including GPR55, CLDND1, CPOX, GPR15, AHRR, PRRC2B, and ELMSAN1. Among 130 SDOH-related CpGs, three independent CpGs (cg04180924, cg1120500, and cg27470486) had a significant combined mediation effect for educational attainment (%mediation=48.9%), and a single mediator cg08064403 was found with significant mediation effect for personal income (25.9%) and ADI (24.1%) on pulmonary diffusion deficit, which showed higher incidence in AA than in EA survivors implying racial disparity which is possibly due to more disadvantageous SDOH factors in AA than in EA.ConclusionsWe demonstrated striking DNAm signatures associated with multiple SDOH factors (educational attainment, personal income, and ADI) and many epigenome-wide significant CpG sites resembling the effect of smoking exposure. We also identified an exemplified racial health disparity in pulmonary diffusion deficit between AA and EA survivors and illuminated DNAm as potential mechanistic mediators for SDOH factors using a social epigenomic approach.

Neuroblastoma is a pediatric malignancy with heterogeneous clinical outcomes. To better understand neuroblastoma pathogenesis, here we analyze whole-genome, whole-exome and/or transcriptome data from 702 neuroblastoma samples. Forty percent of samples harbor at least one recurrent driver gene alteration and most aberrations, including MYCN, ATRX, and TERT alterations, differ in frequency by age. MYCN alterations occur at median 2.3 years of age, TERT at 3.8 years, and ATRX at 5.6 years. COSMIC mutational signature 18, previously associated with reactive oxygen species, is the most common cause of driver point mutations in neuroblastoma, including most ALK and Ras-activating variants. Signature 18 appears early and is continuous throughout disease evolution. Signature 18 is enriched in neuroblastomas with MYCN amplification, 17q gain, and increased expression of mitochondrial ribosome and electron transport-associated genes. Recurrent FGFR1 variants in six patients, and ALK N-terminal structural alterations in five samples, identify additional patients potentially amenable to precision therapy., Genomic analysis of neuroblastoma has revealed important disease etiology. In this study, the authors assembled whole genome, exome and transcriptome data from over 700 neuroblastomas and identified molecular signatures correlated with age, and rare, potentially targetable variants overlooked in smaller cohorts.

BackgroundIt is well-established that cancer treatment substantially increases risk of long-term adverse health outcomes among childhood cancer survivors. However, there is limited research on the underlying mechanisms. To elucidate the pathophysiology and a possible causal pathway from treatment exposures to cardiometabolic conditions, we conducted epigenome-wide association studies (EWAS) to identify DNA methylation (DNAm) sites associated with cancer treatment exposures and examined whether treatment-associated DNAm sites mediate associations between specific treatments and cardiometabolic conditions.MethodsWe included 2,052 survivors (median age 33.7 years) of European ancestry from the St. Jude Lifetime Cohort Study, a retrospective hospital-based study with prospective clinical follow-up. Cumulative doses of chemotherapy and region-specific radiation were abstracted from medical records. Seven cardiometabolic conditions were clinically assessed. DNAm profile was measured using MethylationEPIC BeadChip with blood-derived DNA.ResultsBy performing multiple treatment-specific EWAS, we identified 2,894 5’-cytosine-phosphate-guanine-3′ (CpG) sites mapped to 1,583 gene/regions associated with one or more cancer treatments at epigenome-wide significance level (P < 9×10−8). Among the treatment-associated CpGs, 298 were associated with obesity, 85 with hypercholesterolemia, 41 with hypertriglyceridemia, and four with abnormal glucose metabolism (False-Discovery-Rate-Adjusted PConclusionsIn childhood cancer survivors, prior cancer treatments are associated with DNAm variations present decades following the exposure. Treatment-associated DNAm sites may mediate the causal pathway from specific treatment exposures to certain cardiometabolic conditions, suggesting the utility of DNAm sites as risk predictors and potential mechanistic targets for future intervention studies.

Aggressive cancers often have activating mutations in growth-controlling oncogenes and inactivating mutations in tumor-suppressor genes. In neuroblastoma, amplification of the MYCN oncogene and inactivation of the ATRX tumor-suppressor gene correlate with high-risk disease and poor prognosis. Here we show that ATRX mutations and MYCN amplification are mutually exclusive across all ages and stages in neuroblastoma. Using human cell lines and mouse models, we found that elevated MYCN expression and ATRX mutations are incompatible. Elevated MYCN levels promote metabolic reprogramming, mitochondrial dysfunction, reactive-oxygen species generation, and DNA-replicative stress. The combination of replicative stress caused by defects in the ATRX–histone chaperone complex, and that induced by MYCN-mediated metabolic reprogramming, leads to synthetic lethality. Therefore, ATRX and MYCN represent an unusual example, where inactivation of a tumor-suppressor gene and activation of an oncogene are incompatible. This synthetic lethality may eventually be exploited to improve outcomes for patients with high-risk neuroblastoma., In most cancers, mutations that lead to oncogene activation and tumor suppressor inactivation synergize to promote tumorigenesis. However, in neuroblastomas, MYCN amplification and ATRX mutations are mutually exclusive and incompatible.

Background: Prior research suggests that social determinants of health (SDOH) may influence health through an epigenetic mechanism. However, the social epigenomic approach has not yet been applied to childhood cancer survivors, a population at high risk for chronic health conditions. We aim to investigate how SDOH factors contribute to racial disparity in pulmonary impairment with survivors from the St. Jude Lifetime Cohort Study. Methods: DNA methylation (DNAm) profile was generated with EPIC BeadChip using blood derived DNA. SDOH factors included educational attainment and personal income self-reported using a survey, and socioeconomic area deprivation index (ADI) geocoded using full home addresses. Clinically assessed pulmonary impairment included pulmonary diffusion deficits (PDD), restrictive pulmonary deficits (RPD) and obstructive pulmonary deficits (OPD). Epigenome-wide association study (EWAS) was performed to evaluate the association between DNAm at each CpG and each SDOH factor. Mediation analysis was conducted by treating each SDOH-associated CpG as a mediator, SDOH factor as an exposure, and specific pulmonary condition as the outcome. Genetically inferred races, i.e. survivors of European ancestry (EA) and African ancestry (AA), were considered. Results: The study included 258 AA (median time from cancer diagnosis [MTD]=25.2 years, interquartile range [IQR]=19.9-32.1 years) and 1,618 EA survivors (MTD=27.3, IQR=21.1-33.7 years). Compared to EA survivors, AA survivors had lower educational attainment (P Conclusions: DNAm signatures, many resembling the effect of tobacco use, are associated with educational attainment, personal income, and ADI. Our findings suggest that these DNAm are potential mechanistic mediators for the effects of SDOH factors on PDD risk. Through this mechanism, poor SDOH factors in AA survivors led to racial disparity in PDD. Citation Format: Nan Song, Jin-ah Sim, Qian Dong, Yinan Zheng, Lifang Hou, Zhenghong Li, Chia-Wei Hsu, Haitao Pan, Heather Mulder, John Easton, Emily Walker, Geoffrey Neale, Carmen L. Wilson, Kirsten K. Ness, Kevin R. Krull, Deo Kumar Srivastava, Yutaka Yasui, Jinghui Zhang, Melissa M. Hudson, Leslie L. Robison, I-Chan Huang, Zhaoming Wang. A social epigenomic investigation of racial disparity in pulmonary impairment among aging survivors of childhood cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 685.

Background: Epigenetic research of blood lipids has been conducted in the general population with some robust findings. However, there is lack of similar studies that explore epigenetic effects among childhood cancer survivors who have elevated risk of dyslipidemia. We aimed to utilize epigenome-wide approach to search for DNA methylation (DNAm) sites influencing lipid metabolism among survivors of childhood cancer. Methods: Epigenome-wide methylation data were generated with Infinium EPIC BeadChip on blood derived DNA of survivors from the St. Jude Lifetime Cohort Study. Chemotherapy agents and region-specific radiation exposures were abstracted from medical records. Two forms of dyslipidemia, i.e. hypertriglyceridemia (HTG) or hypercholesteremia (HCL) were clinically assessed. Multivariable logistic regression was used to evaluate associations between each DNAm site with incident cases of HTG or HCL adjusting for age, sex, and cancer treatment exposures. Analysis was stratified by genetically determined races, i.e. survivors of European ancestry (EA) and survivors of African ancestry (AA). Significant DNAm sites were correlated with mRNA expression of the mapped genes using RNA sequencing data available for blood samples of 87 survivors. Results: 2,052 EA (median age=34.7 years, 47.1% female, 26.6% HTG, 32.0% HCL) and 370 AA survivors (median age=32.1 years, 53.2% female, 15.9% HTG, 26.8% HCL) were included. Among EA survivors, six DNAm sites were associated with HTG risk at epigenome-wide significance level (P Conclusion: We identified one or more DNAm sites, residing in each of CPT1A, SLC43A1, and PHGDH genes, associated with dyslipidemia among EA survivors. Lack of associations among AA survivors may partially contribute to lower risk of dyslipidemia among AA survivors. The findings demonstrated epigenetic roles in dyslipidemia among survivors, suggesting epigenetic biomarkers for identification of survivors with higher dyslipidemia risk and new targets for future interventions. Citation Format: Qian Dong, Nan Song, Yadav Sapkota, Yinan Zheng, I-Chan Huang, Deo Kumar Srivastava, John Easton, Heather Mulder, Geoffrey Neale, Emily Walker, Jinghui Zhang, Melissa M. Hudson, Leslie L. Robison, Zhaoming Wang. Epigenome-wide association study of dyslipidemia in survivors of childhood cancer: A report from the St. Jude lifetime cohort [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 904.

Diffuse intrinsic pontine gliomas (DIPG) are lethal pediatric brain tumors that arise in the pons. To understand the spatial heterogeneity in DIPG, we performed ultra-deep sequencing and copy number profiling of 44 multi-region autopsy samples from 10 DIPG patients after phase I treatment with the PDGFR inhibitor crenolanib. Our study recapitulated the previously reported temporal order of driver mutation acquisition starting with histone H3 K27M followed by cell cycle (TP53 and PPM1D) or growth factor pathways (AVCR1, or PIK3R1), with mutations affecting PI3K and MAPK pathways being late, subclonal events. Further, we observed that aneuploidy was likely an early event as copy number heterogeneity was minimal in most patients. Compared to previous DIPG studies, we observed an enrichment in subclonal mutations and convergent evolution of PI3K alterations. This enrichment may have been due to selection for crenolanib resistance, or to the prior treatment history of the patients within the cohort. We also detailed intra-patient clonal composition among tumor samples and identified founder clones that were dominant in only a single location, which we considered to be non-migratory “occupier” clones, in four patients. By contrast, in six other cases, we observed clones that frequently invaded other sites. Interestingly, in one case with TP53 convergent evolution, aneuploidy clone carrying TP53 R273H mutation invades all other sequenced tumor sites whereas another clone with TP53 R248W does not. These results indicate the importance of understanding spatial heterogeneity within DIPG patients, which is crucial to the identification of effective treatments. Citation Format: Sasi Arunachalam, Samuel W. Brady, Xiaotu Ma, Karol Szlachta, Heather Mulder, John Easton, Christopher L. Tinkle, Cynthia Wetmore, Suzanne J. Baker, Jinghui Zhang. Spatial heterogeneity in diffuse intrinsic pontine gliomas treated with a PDGFR inhibitor [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 3047.

Single-cell RNA sequencing (scRNA-seq) is a powerful tool for characterizing the cell-to-cell variation and cellular dynamics in populations which appear homogeneous otherwise in basic and translational biological research. However, significant challenges arise in the analysis of scRNA-seq data, including the low signal-to-noise ratio with high data sparsity, potential batch effects, scalability problems when hundreds of thousands of cells are to be analyzed among others. The inherent complexities of scRNA-seq data and dynamic nature of cellular processes lead to suboptimal performance of many currently available algorithms, even for basic tasks such as identifying biologically meaningful heterogeneous subpopulations. In this study, we developed the Latent Cellular Analysis (LCA), a machine learning–based analytical pipeline that combines cosine-similarity measurement by latent cellular states with a graph-based clustering algorithm. LCA provides heuristic solutions for population number inference, dimension reduction, feature selection, and control of technical variations without explicit gene filtering. We show that LCA is robust, accurate, and powerful by comparison with multiple state-of-the-art computational methods when applied to large-scale real and simulated scRNA-seq data. Importantly, the ability of LCA to learn from representative subsets of the data provides scalability, thereby addressing a significant challenge posed by growing sample sizes in scRNA-seq data analysis.

Purpose: We aimed to analyze and compare leukocyte telomere length (LTL) and age-dependent LTL attrition between childhood cancer survivors and noncancer controls, and to evaluate the associations of LTL with treatment exposures, chronic health conditions (CHC), and health behaviors among survivors. Experimental Design: We included 2,427 survivors and 293 noncancer controls of European ancestry, drawn from the participants in St. Jude Lifetime Cohort Study (SJLIFE), a retrospective hospital-based study with prospective follow-up (2007–2016). Common nonneoplastic CHCs (59 types) and subsequent malignant neoplasms (5 types) were clinically assessed. LTL was measured with whole-genome sequencing data. Results: After adjusting for age at DNA sampling, gender, genetic risk score based on 9 SNPs known to be associated with telomere length, and eigenvectors, LTL among survivors was significantly shorter both overall [adjusted mean (AM) = 6.20 kb; SE = 0.03 kb] and across diagnoses than controls (AM = 6.69 kb; SE = 0.07 kb). Among survivors, specific treatment exposures associated with shorter LTL included chest or abdominal irradiation, glucocorticoid, and vincristine chemotherapies. Significant negative associations of LTL with 14 different CHCs, and a positive association with subsequent thyroid cancer occurring out of irradiation field were identified. Health behaviors were significantly associated with LTL among survivors aged 18 to 35 years (Ptrend = 0.03). Conclusions: LTL is significantly shorter among childhood cancer survivors than noncancer controls, and is associated with CHCs and health behaviors, suggesting LTL as an aging biomarker may be a potential mechanistic target for future intervention studies designed to prevent or delay onset of CHCs in childhood cancer survivors. See related commentary by Walsh, p. 2281

Background Sequencing errors are key confounding factors for detecting low-frequency genetic variants that are important for cancer molecular diagnosis, treatment, and surveillance using deep next-generation sequencing (NGS). However, there is a lack of comprehensive understanding of errors introduced at various steps of a conventional NGS workflow, such as sample handling, library preparation, PCR enrichment, and sequencing. In this study, we use current NGS technology to systematically investigate these questions. Results By evaluating read-specific error distributions, we discover that the substitution error rate can be computationally suppressed to 10−5 to 10−4, which is 10- to 100-fold lower than generally considered achievable (10−3) in the current literature. We then quantify substitution errors attributable to sample handling, library preparation, enrichment PCR, and sequencing by using multiple deep sequencing datasets. We find that error rates differ by nucleotide substitution types, ranging from 10−5 for A>C/T>G, C>A/G>T, and C>G/G>C changes to 10−4 for A>G/T>C changes. Furthermore, C>T/G>A errors exhibit strong sequence context dependency, sample-specific effects dominate elevated C>A/G>T errors, and target-enrichment PCR led to ~ 6-fold increase of overall error rate. We also find that more than 70% of hotspot variants can be detected at 0.1 ~ 0.01% frequency with the current NGS technology by applying in silico error suppression. Conclusions We present the first comprehensive analysis of sequencing error sources in conventional NGS workflows. The error profiles revealed by our study highlight new directions for further improving NGS analysis accuracy both experimentally and computationally, ultimately enhancing the precision of deep sequencing. Electronic supplementary material The online version of this article (10.1186/s13059-019-1659-6) contains supplementary material, which is available to authorized users.

Purpose: The risk of subsequent breast cancer among female childhood cancer survivors is markedly elevated. We aimed to determine genetic contributions to this risk, focusing on polygenic determinants implicated in breast cancer susceptibility in the general population. Experimental Design: Whole-genome sequencing (30×) was performed on survivors in the St Jude Lifetime Cohort, and germline mutations in breast cancer predisposition genes were classified for pathogenicity. A polygenic risk score (PRS) was constructed for each survivor using 170 established common risk variants. Relative rate (RR) and 95% confidence interval (95% CI) of subsequent breast cancer incidence were estimated using multivariable piecewise exponential regression. Results: The analysis included 1,133 female survivors of European ancestry (median age at last follow-up = 35.4 years; range, 8.4–67.4), of whom 47 were diagnosed with one or more subsequent breast cancers (median age at subsequent breast cancer = 40.3 years; range, 24.5–53.0). Adjusting for attained age, age at primary diagnosis, chest irradiation, doses of alkylating agents and anthracyclines, and genotype eigenvectors, RRs for survivors with PRS in the highest versus lowest quintiles were 2.7 (95% CI, 1.0–7.3), 3.0 (95% CI, 1.1–8.1), and 2.4 (95% CI, 0.1–81.1) for all survivors and survivors with and without chest irradiation, respectively. Similar associations were observed after excluding carriers of pathogenic/likely pathogenic mutations in breast cancer predisposition genes. Notably, the PRS was associated with the subsequent breast cancer rate under the age of 45 years (RR = 3.2; 95% CI, 1.2–8.3). Conclusions: Genetic profiles comprised of small-effect common variants and large-effect predisposing mutations can inform personalized breast cancer risk and surveillance/intervention in female childhood cancer survivors.

Background: The epigenetic clock becomes an important biomarker in aging research, where epigenetic age (EA) approximates chronological age with high accuracy and epigenetic age acceleration (EAA) demonstrates superior performance in predicting health outcomes in the general population. The utility of EAA among adult survivors of childhood cancer is largely unknown. We aimed to compare EAA between childhood cancer survivors and non-cancer controls, and evaluate associations between EAA, treatment exposures, health behaviors, and chronic health conditions (CHCs) among survivors. Methods: Genome-wide methylation data were generated with Infinium EPIC BeadChip on blood derived DNA from 2139 survivors and 282 community controls with no prior history of cancer. EAA was estimated as the residual value from the fit of a simple linear regression of EA (using Levine's clock) on chronological age (i.e., age at DNA sampling). Cumulative doses of chemotherapy and region-specific radiation exposures were abstracted from medical records. Health behaviors including physical activity, diet, tobacco smoking, and alcohol consumption were assessed from questionnaires and categorized as favorable, intermediate and unfavorable. Multivariable piecewise-exponential regression model was employed to evaluate associations of EAA with 59 clinically assessed CHCs. Results: EA was highly correlated with chronological age with the same Pearson coefficient (r = 0.88) in survivors and controls; however, EAA was significantly higher in survivors than controls, overall (adjusted least square mean [ALSM] = 0.63 vs. -3.61 years, P Conclusion: EAA is significantly higher in childhood cancer survivors than non-cancer controls and is associated with treatment exposures, health behaviors, and CHCs. Clinically, EAA may inform intervention strategies including health coaching to prevent or delay the onset of CHCs among survivors of childhood cancer. Citation Format: Na Qin, Zhenghong Li, Nan Song, John Easton, Heather Mulder, Emily Plyler, Geoffery Neale, Emily Walker, Carmen L. Wilson, Melissa M. Hudson, Yutaka Yasui, Leslie L. Robison, Jinghui Zhang, Kirsten K. Ness, Zhaoming Wang. Epigenetic age acceleration and chronic health conditions among adult survivors of childhood cancer: A report from the St. Jude Lifetime Cohort [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5414.

Liquid biopsy holds great promise in noninvasive diagnosis of cancers through detecting minute amounts of cell-free DNA released from cancer cells in non-solid biologic tissue such as peripheral blood. A critical bottleneck in developing liquid biopsy methods is the limited accuracy of current next-generation sequencing technology (NGS), evidenced by its high error rate (0.1%-1%, as of 2018). Through mathematical modeling of NGS errors, we have recently published a method to computationally suppress the current NGS error rate to between 10−5 and 10−4, two orders of magnitude lower than general reports. However, this error rate is a product of both PCR errors and instrument (i.e., sequencer) errors, and it is currently unknown how to separate these error sources. In this work, we developed a novel computational algorithm to precisely measure the errors caused by sequencers. By using 12 publicly available datasets from 10 sequencing centers (in America, Europe, and Asia), we discovered highly reproducible patterns of sequencer errors, including: 1) the overall sequencer error rate is 10−5; 2) at the flow-cell level, error rates are elevated in the bottom surface; 3) almost all flow cells have a small fraction of random tiles with a dramatically elevated error rate; 4) the elevated error rates appear to be enriched in some reaction cycles; 5) removal of these reaction cycles yields 5-fold lower error rates at some genomic loci, so that A>C, A>T, and C>G error types have error rates close to 10−6; and 6) sequencer errors have a pattern markedly distinct from PCR errors. We have implemented the above observations into a general-purpose algorithm, termed CleanDeepSeq2, to computationally suppress sequencer errors and to also effectively monitor sequencer anomalies. CleanDeepSeq2 was engineered for efficiency so that a dataset with ultra-deep sequencing (1,000,000X depth) can be processed in 1.5N minutes on a single CPU core, where N is the number of target regions. Similarly, WES (100X) and WGS (~30X) datasets can be processed in under 1 CPU hour in order to monitor instrument performance. Overall, we have developed a computational method that for the first time enabled precise measurement of sequencer errors. Our study revealed novel insights on sequencer errors that can lead to improved instrumentation, NGS chemistry, and ultimately higher DNA sequencing fidelity. In addition, our developed software can efficiently suppress sequencer errors in addition to previously discovered error sources. Citation Format: Eric Davis, Rain Sun, Ying Shao, Yanling Liu, Heather L. Mulder, Stephen V. Rice, John Easton, Jinghui Zhang, Xiaotu Ma. Uncovering instrument errors in next-generation sequencing by CleanDeepSeq2 [abstract]. In: Proceedings of the AACR Special Conference on Advances in Liquid Biopsies; Jan 13-16, 2020; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(11_Suppl):Abstract nr A57.

Double minute chromosomes are extrachromosomal circular DNA fragments frequently found in brain tumors. To understand their evolution, we characterized the double minutes in paired diagnosis and relapse tumors from a pediatric high-grade glioma and four adult glioblastoma patients. We determined the full structures of the major double minutes using a novel approach combining multiple types of supporting genomic evidence. Among the double minutes identified in the pediatric patient, only one carrying EGFR was maintained at high abundance in both samples, whereas two others were present in only trace amounts at diagnosis but abundant at relapse, and the rest were found either in the relapse sample only or in the diagnosis sample only. For the EGFR-carrying double minutes, we found a secondary somatic deletion in all copies at relapse, after erlotinib treatment. However, the somatic mutation was present at very low frequency at diagnosis, suggesting potential resistance to the EGFR inhibitor. This mutation caused an in-frame RNA transcript to skip exon 16, a novel transcript isoform absent in EST database, as well as about 700 RNA-seq of normal brains that we reviewed. We observed similar patterns involving longitudinal copy number shift of double minutes in another four pairs (diagnosis/relapse) of adult glioblastoma. Overall, in three of five paired tumor samples, we found that although the same oncogenes were amplified at diagnosis and relapse, they were amplified on different double minutes. Our results suggest that double minutes readily evolve, increasing tumor heterogeneity rapidly. Understanding patterns of double minute evolution can shed light on future therapeutic solutions to brain tumors carrying such variants. Electronic supplementary material The online version of this article (10.1007/s00401-018-1912-1) contains supplementary material, which is available to authorized users.

SUMMARYAggressive cancers often have activating mutations in growth-controlling oncogenes and inactivating mutations in tumor-suppressor genes. In neuroblastoma, amplification of the MYCN oncogene and inactivation of the ATRX tumor-suppressor gene correlate with high-risk disease and poor prognosis. Here we show that ATRX mutations and MYCN amplification are mutually exclusive across all ages and stages in neuroblastoma. Using human cell lines and mouse models, we found that elevated MYCN expression and ATRX mutations are incompatible. Elevated MYCN levels promote metabolic reprogramming, mitochondrial dysfunction, reactive-oxygen species generation, and DNA-replicative stress. The combination of replicative stress caused by defects in the ATRX–histone chaperone complex and that induced by MYCN-mediated metabolic reprogramming leads to synthetic lethality. Therefore, ATRX and MYCN represent an unusual example, where inactivation of a tumor-suppressor gene and activation of an oncogene are incompatible. This synthetic lethality may eventually be exploited to improve outcomes for patients with high-risk neuroblastoma.

Purpose Childhood cancer survivors are at increased risk of subsequent neoplasms (SNs), but the germline genetic contribution is largely unknown. We assessed the contribution of pathogenic/likely pathogenic (P/LP) mutations in cancer predisposition genes to their SN risk. Patients and Methods Whole-genome sequencing (30-fold) was performed on samples from childhood cancer survivors who were ≥ 5 years since initial cancer diagnosis and participants in the St Jude Lifetime Cohort Study, a retrospective hospital-based study with prospective clinical follow-up. Germline mutations in 60 genes known to be associated with autosomal dominant cancer predisposition syndromes with moderate to high penetrance were classified by their pathogenicity according to the American College of Medical Genetics and Genomics guidelines. Relative rates (RRs) and 95% CIs of SN occurrence by mutation status were estimated using multivariable piecewise exponential regression stratified by radiation exposure. Results Participants were 3,006 survivors (53% male; median age, 35.8 years [range, 7.1 to 69.8 years]; 56% received radiotherapy), 1,120 SNs were diagnosed among 439 survivors (14.6%), and 175 P/LP mutations were identified in 5.8% (95% CI, 5.0% to 6.7%) of survivors. Mutations were associated with significantly increased rates of breast cancer (RR, 13.9; 95% CI, 6.0 to 32.2) and sarcoma (RR, 10.6; 95% CI, 4.3 to 26.3) among irradiated survivors and with increased rates of developing any SN (RR, 4.7; 95% CI, 2.4 to 9.3), breast cancer (RR, 7.7; 95% CI, 2.4 to 24.4), nonmelanoma skin cancer (RR, 11.0; 95% CI, 2.9 to 41.4), and two or more histologically distinct SNs (RR, 18.6; 95% CI, 3.5 to 99.3) among nonirradiated survivors. Conclusion The findings support referral of all survivors for genetic counseling for potential clinical genetic testing, which should be prioritized for nonirradiated survivors with any SN and for those with breast cancer or sarcoma in the field of prior irradiation.