California Childhood Leukemia Study Research Articles

Abstract A079: The impact of Indigenous American ancestry on the risk of acute lymphoblastic leukemia and the implications for future study designs

Abstract Acute lymphoblastic leukemia (ALL) is the most common childhood cancer, with Hispanic/Latino (H/L) children having up to 1.4 times the rate of ALL compared to their non-Hispanic White counterparts. This disparity has not been fully explained by environmental factors, suggesting the role of understudied genetic variants, particularly ones found in Indigenous American (IA) ancestry inherited by Latinos. In this study, we characterized the impact of IA ancestry on the frequency and effect size of known ALL risk alleles. We found that elevated risk of ALL in Latinos may be conferred by increased frequency of risk alleles due to IA ancestry. For instance, in an analysis of Latinos from the California Cancer Linkage Project (CCRLP; 1,930 cases, 2,103 controls) and the California Childhood Leukemia Study (CCLS; N = 605 cases, 515 controls), we found that, across independent known ALL loci (N=21), more loci than expected by chance showed a significant association between IA ancestry and increasing risk (p = 0.015). Compared to randomly-sampled, frequency-matched alleles from the genome, the risk alleles at all known ALL loci also showed a significant association with IA ancestry (46.9 % for known risk alleles vs. 43.5 % for matched alleles; ). The IA haplotype had 1.39 times the odds (95% CI: 1.35 – 1.43; ) of harboring the risk allele compared to the non-IA haplotype. Conversely, in a combined Latino sample of 2,535 cases and 9,035 controls, we found no evidence of GxAncestry interaction on ALL risk across all known loci (min P = 0.0026, prior to adjusting for multiple testing), suggesting that ALL risk alleles do not have increased effects on the IA ancestry background. This lack of synergism between local ancestry and genotype in ALL has methodological implications for gene discovery in admixed populations. To explore this, we implemented Tractor, a method that models genotype effects by ancestry, and thus is best powered in the presence of effect size heterogeneity by ancestry. Across known loci, we found that Tractor had 36.1% decreased power compared to standard GWAS. Taken together, our results suggest that the disproportionate burden of ALL in Latino populations may arise from the enrichment of risk alleles within the IA ancestry component, possibly driven by immunity- related selective pressures. Finally, future genetic studies of ALL in Latino patients may benefit from methods that incorporate contributions from ancestry to increase the statistical power in identifying associations, rather than leverage effect size heterogeneity across ancestries. Citation Format: Jalen Langie, Libby Morimoto, Xiaomei Ma, Catherine Metayer, Joseph L. Wiemels, Adam J. de Smith, Charleston W.K. Chiang. The impact of Indigenous American ancestry on the risk of acute lymphoblastic leukemia and the implications for future study designs [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl):Abstract nr A079.

Early-life tobacco exposure is causally implicated in aberrant RAG-mediated recombination in childhood acute lymphoblastic leukemia.

Acute lymphoblastic leukemia (ALL) is the most common cancer in children, yet few environmental risk factors have been identified. We previously found an association between early-life tobacco smoke exposure and frequency of somatic deletions of 8 leukemia driver genes among childhood ALL patients in the California Childhood Leukemia Study. To expand analysis genome-wide and examine potential mechanisms, we conducted tumor whole-genome sequencing in 35 ALL patients, including 18 with high prenatal tobacco exposure and 17 with low exposure as determined by established epigenetic biomarkers. High tobacco exposure patients had significantly more structural variants (P < .001) and deletions (P = .001) genome-wide than low exposure patients. Investigation of off-target RAG recombination revealed that 41% of deletions in the high tobacco exposure patients were putatively RAG-mediated (full RAG motif identified at one or both breakpoints) compared with only 21% in the low exposure group (P = .001). In a multilevel model, deletions in high tobacco exposure patients were 2.44-fold (95% CI:1.13-5.38) more likely to be putatively RAG-mediated than deletions in low exposure patients. No point mutational signatures were associated with prenatal tobacco exposure. Our findings suggest that early-life tobacco smoke exposure may promote leukemogenesis by driving development of somatic deletions in pre-leukemic lymphocytes via off-target RAG recombination.

Glyphosate in house dust and risk of childhood acute lymphoblastic leukemia in California

BackgroundResidential use of pesticides has been associated with increased risk of childhood acute lymphoblastic leukemia (ALL). We evaluated determinants of glyphosate concentrations in house dust and estimated ALL risk in the California Childhood Leukemia Study (CCLS). MethodsThe CCLS is a population-based case-control study of childhood leukemia in California. Among those < 8-years (no move since diagnosis/reference date), we collected dust (2001–2007) from the room where the child spent the most time while awake and measured > 40 pesticides. Three-to-eight years later, we collected a second sample from non-movers. We used Ultra-Performance Liquid Chromatography Tandem Mass Spectrometry to measure glyphosate (µg/g dust) for 181 ALL cases and 225 controls and for 45 households with a second dust sample. We used multivariable Tobit regression to evaluate determinants of glyphosate concentrations. Odds ratios (ORs) and 95 % confidence intervals (CI) were calculated for ALL and quartiles of the concentration (first samples) using unconditional logistic regression. We computed the within- and between-home variance and intraclass correlation coefficient (ICC). ResultsGlyphosate was frequently detected (cases: 98 %; controls: 99 %). Higher concentrations were associated with occupational pesticide exposure, nearby agricultural use, treatment for lawn weeds and bees/wasps, and sampling season. Increasing concentrations were not associated with ALL risk (adjusted ORQ4vsQ1 = 0.8, CI: 0.4–1.4). We observed similar null associations for boys and girls, Hispanics and non-Hispanic whites, and among those who resided in their home since birth (76 cases/117 controls) or age two (130 cases/176 controls). The ICC was 0.32 indicating high within-home temporal variability during the years of our study. ConclusionsWe observed higher concentrations in homes associated with expected predictors of exposure but no association with childhood ALL risk. Due to continuing use, potential exposure to young children is high. It will be important to evaluate risk in future studies with multiple dust measurements or biomarkers of exposure.

Periconceptional folate intake influences DNA methylation at birth based on dietary source in an analysis of pediatric acute lymphoblastic leukemia cases and controls

Periconceptional folate intake is associated with the establishment of DNA methylation in offspring; however, variations in this relation by food sources compared with folic acid supplements are not described. Also, maternal folate intake is associated with decreased risk of pediatric acute lymphoblastic leukemia (ALL), but the mechanism is not known. We evaluated the relation between periconceptional folate intake by source and DNA methylation at birth in a cohort of pediatric ALL cases and controls in an epigenome-wide association study. Genome-wide DNA methylation status obtained from archived neonatal blood spots from pediatric ALL cases (n=189) and controls (n=205) in the California Childhood Leukemia Study (CCLS) from 1995-2008 was compared with periconceptional folate from total, food, and supplemental sources using multivariable linear regression. Further stratification was performed by income, education, ethnicity, and total folate intake. We evaluated variable DNA methylation response to periconceptional folate by ALL case status through an interaction term. Two significant differentially methylated probes (DMPs) were associated with food and supplemental periconceptional folate intake in all subjects (n=394). The top differentially methylated region at the promoter region of DUSP22(dual specificity phosphatase 22) demonstrated DNA hypermethylation in ALL cases but not in controls in response to total and food folate intake. We further identified 8 interaction term DMPs with variable DNA methylation response to folate intake by ALL case status. Further stratification of the cohort by education and ethnicity revealed a substantially higher number of DMPs associated with supplemental folic acid intake in Hispanic subjects with lower income and educational level. We identified modest associations between periconceptional folate intake and DNA methylation differing by source, including variation by ALL case status. Hispanic subjects of lower income and education appear uniquely responsive to periconceptional folate supplementation.

Associations between early-life and in utero infections and cytomegalovirus-positive acute lymphoblastic leukemia in children.

Childhood infections and cytomegalovirus (CMV) are associated with pediatric acute lymphoblastic leukemia (ALL). CMV dysregulates the host immune system and alters the immune response to subsequent antigenic exposures. We suspect that this immune dysregulation contributes to increased numbers of symptomatic infections in childhood allowing for expansion of pre-leukemic clones. We explored the association between childhood infections, maternal infections during pregnancy and CMV-positive ALL. Using a droplet digital PCR assay, we screened diagnostic ALL bone marrow samples from the California Childhood Leukemia Study (1995-2015) for the presence of CMV DNA identifying CMV-positive and CMV-negative cases. We performed a case-only analysis (n=524) comparing the number and types of childhood infections and maternal infections during pregnancy between CMV-positive and CMV-negative ALL cases using logistic regression. With increasing numbers of infections in the first 12 months of life, children were more likely to classify to the highest tertile of CMV DNA in the bone marrow at diagnosis (OR: 1.04, 95% CI: 1.01-1.08). Specifically, those reporting cough or flu in the first 12 months were more likely to be CMV-positive at ALL diagnosis (OR: 2.15, 95% CI: 1.06-4.37 and OR: 2.06, 95% CI: 1.17-3.63 respectively). Furthermore, those with a history of maternal infection during pregnancy were more likely to be CMV-positive (OR: 2.12, 95% CI: 1.24-3.62). We hypothesize that children with underlying immune dysregulation develop more symptomatic infections in childhood and ultimately CMV-positive ALL; this underlying immune dysregulation may be due to early immune system alterations via CMV exposure (in utero or early infancy) proposing a potential link between CMV and ALL etiology.

Imputation of Below Detection Limit Missing Data in Chemical Mixture Analysis with Bayesian Group Index Regression.

There is growing scientific interest in identifying the multitude of chemical exposures related to human diseases through mixture analysis. In this paper, we address the issue of below detection limit (BDL) missing data in mixture analysis using Bayesian group index regression by treating both regression effects and missing BDL observations as parameters in a model estimated through a Markov chain Monte Carlo algorithm that we refer to as pseudo-Gibbs imputation. We compare this with other Bayesian imputation methods found in the literature (Multiple Imputation by Chained Equations and Sequential Full Bayes imputation) as well as with a non-Bayesian single-imputation method. To evaluate our proposed method, we conduct simulation studies with varying percentages of BDL missingness and strengths of association. We apply our method to the California Childhood Leukemia Study (CCLS) to estimate concentrations of chemicals in house dust in a mixture analysis of potential environmental risk factors for childhood leukemia. Our results indicate that pseudo-Gibbs imputation has superior power for exposure effects and sensitivity for identifying individual chemicals at high percentages of BDL missing data. In the CCLS, we found a significant positive association between concentrations of polycyclic aromatic hydrocarbons (PAHs) in homes and childhood leukemia as well as significant positive associations for polychlorinated biphenyls (PCBs) and herbicides among children from the highest quartile of household income. In conclusion, pseudo-Gibbs imputation addresses a commonly encountered problem in environmental epidemiology, providing practitioners the ability to jointly estimate the effects of multiple chemical exposures with high levels of BDL missingness.

The Effect of Cytomegalovirus on Pediatric Acute Lymphoblastic Leukemia

Background: Recent evidence supports the role of cytomegalovirus (CMV) in the development of childhood ALL. The underlying mechanism and CMV's role in the leukemic cell phenotype is unknown, but CMV typically interacts with the host immune system allowing the virus to survive in a latent state; it may be that this immune dysregulation affects the risk of ALL. This study aims to explore the association of CMV and ALL at the time of leukemia diagnosis, using AML cases as a control and further, to determine whether CMV affects certain subgroups of ALL patients such as specific ethnicities, age groups, or cytogenetic subtypes.Methods: Pediatric diagnostic leukemia bone marrow samples obtained from the California Childhood Leukemia Study were screened for the presence of CMV DNA using a custom-designed droplet digital PCR assay. A total of 869 cases were analyzed including 125 AML cases and 744 ALL cases. Demographic and clinical features were compared between patients found to be CMV positive (cases with any detectable CMV positive droplets) and those who were CMV negative (cases with no detectable CMV positive droplets). The effect of the level of CMV viral DNA load was also assessed. For a subset of cases (n=61), Affymetrix Array gene expression data were available and differential gene expression performed to compare CMV positive cases with high viral load to CMV negative cases. Odds ratios and confidence intervals were estimated using logistic regression.Results: ALL cases were more likely to be CMV positive compared to AML (OR: 2.50; CI 1.00, 5.47, p = 0.039 for CMV highest quintile vs. CMV negative). Within ALL cases, B-cell ALL (B-ALL) was significantly associated with CMV positivity compared to T-cell ALL (T-ALL) (OR: 2.93; CI: 1.01, 8.52, p = 0.048 for CMV highest tertile vs. CMV negative). Further subtype analysis of B-ALL cases revealed CMV positivity to be significantly associated with high hyperdiploidy, one of the most common ALL subtypes, when compared to ETV6-RUNX1 (OR: 2.52; CI:1.34, 4.73, p = 0.004 for highest CMV tertile vs. CMV negative). CMV positive B-ALL cases were also more likely to harbor deletions of EBF1, a B-cell development gene, compared to CMV negative cases (OR: 6.10; CI: 1.09, 34.06, p = 0.04 for CMV 2 nd tertile vs. CMV negative; OR: 5.54; CI: 1.13, 27.18, p = 0.03 for CMV highest tertile vs. CMV negative). Differential gene expression analysis revealed 830 genes to be significantly differentially expressed between the highest quintile of CMV positive cases and CMV negative cases, and gene ontology analysis revealed upregulation of processes involved in viral infection and replication. Specifically, cytokine signaling pathways including IL-1, IL-8, and IL-7 were upregulated in CMV positive cases while Th1 and the pathway facilitating crosstalk between dendritic cells and natural killer cells were downregulated. Interestingly, B-cell receptor signaling was also upregulated in CMV positive cases.Conclusion: Our results support the hypothesis that CMV plays an enhanced role in leukemia development in specific subtypes of ALL, and not in AML development. The ability of CMV to interact with the host immune system, highlights immune dysregulation as a potential mechanism by which CMV contributes to risk of ALL. Gene expression analysis on a subset of cases revealed differentially expressed genes to be enriched in pathways involved in immune response, suggesting a potential role for active CMV infection in the leukemic phenotype. The patterns of up- and downregulation in these pathways were consistent with the host response to CMV. Additionally, acute CMV infection has been shown to promote B-cell activation and proliferation. Further, CMV seropositive individuals have been reported to have altered immune responses even with the virus in a latent state highlighting the virus's effect on B-cells. In our study we also found B-cell receptor signaling to be upregulated in CMV positive cases. This is consistent with the known effects of CMV in a typical host, but in patients with leukemia it provides an interesting potential link between CMV infection and development of pediatric ALL. These intriguing results require validation and warrant continued investigation of the role of CMV in pediatric leukemia development. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Large Multi-Ethnic Genome Wide Association Study Reveals Novel Genetic Associations with Childhood Acute Lymphoblastic Leukemia at Chromosomes 8q24 and 17q12

Genome-wide association studies (GWAS) have revealed heritable genetic risk factors for childhood acute lymphoblastic leukemia (ALL), highlighting the role of lymphoid development and cell cycle control genes in disease etiology. Most studies have focused on non-Latino white populations; however, the highest global rates of ALL are found in Latinos in the US and Latin America. To identify novel genetic susceptibility alleles, we carried out a large, multi-ethnic GWAS of ALL in the Californian population.Childhood ALL cases were identified using the California Cancer Registry, and age, gender, and ethnicity-matched controls were identified from the Department of Vital Statistics. Newborn bloodspots from California-born children were obtained from the California Biobank Program. Genome-wide SNP genotyping was carried out for 1,949 Latino, 1,184 non-Latino white, and 130 African-American cases as well as 3,506 controls, using the Affymetrix Axiom World Latin array. Additional controls (n=12,471) from the Kaiser Resource for Genetic Epidemiology Research on Aging Cohort (GERA) were genotyped on the same array. For increased statistical power, we carried out a meta-analysis of the three ethnicity-stratified GWAS. Replication of putative novel loci was performed in two independent datasets: the European-ancestry based Children's Oncology Group (COG) (959 cases, plus 2624 controls from the Wellcome Trust Case-Control Consortium), and in Latino subjects from the California Childhood Leukemia Study (CCLS) (530 cases, 511 controls). Local imputation across top association loci was carried out, using 1000 Genomes Project SNP data as a reference, to identify causal variants.Our childhood ALL GWAS meta-analysis of 3,263 cases and 15,977 controls identified two novel genome-wide significant SNP loci (P<5.0 x 10-8) at chromosomes 8q24 (OR=1.27, 95% CI:1.17-1.38) and 17q12 (OR=1.18, 95% CI:1.11-1.25) that replicated in COG and CCLS datasets. The 8q24 locus lies within a 2.5Mb region containing known GWAS hits for multiple phenotypes, including several cancer types and developmental traits. Imputation narrowed the locus to a ~100kb intergenic region, including top hit SNP rs4617118, that demonstrates chromatin looping with the MYC oncogene in lymphoblastoid cell lines. The 17q12 association peak encompasses the lymphoid transcription factor IKZF3, as well as ZPBP2, GSDMB, and ORMDL3. Top SNP rs2290400 disrupts a BLIMP-1 binding motif, and is a highly significant expression quantitative trait locus (eQTL) for both GSDMB and ORMDL3 . Rs2290400 is also a GWAS hit for type 1 diabetes and asthma, with the ALL risk allele conferring protection against these autoimmune conditions. In addition to novel loci, we replicated established GWAS hits in ARID5B, IKZF1, CEBPE, PIP4K2A, and CDKN2A at genome-wide significance, and in GATA3 at P=5.2x10-6. Recently identified loci at LHPP and ELK3 also replicated, at P=5.7x10-6 and P=2.1x10-3 respectively.Capitalizing on the large proportion of Latinos in our study, we utilized the reduced linkage disequilibrium of this genetically admixed population in combination with fine-mapping to identify new putative causal variants at known ALL-associated loci.Our large, multi-ethnic GWAS of childhood ALL has identified two novel genetic associations, at loci that function in hematopoiesis (17q12) and B-cell proliferation (MYC), in addition to pinpointing causal variants at known ALL loci. Further work is required to elucidate the biological mechanisms underlying these associations. Our results suggest that analysis of larger ALL datasets may yield additional genetic risk loci of moderate effect size. DisclosuresMa:Incyte Corp.: Consultancy.

Cytokine Levels at Birth in Children Who Developed Acute Lymphoblastic Leukemia.

Prenatal immune development may play an important role in the etiology of childhood acute lymphoblastic leukemia (ALL). Seven cytokines, IL1β, IL4, IL6, IL8, GM-CSF, TNFα, and VEGF, were analyzed in blood spots collected at birth from 1,020 ALL cases and 1,003 controls participating in the California Childhood Leukemia Study. ORs and 95% confidence intervals (95% CI) associated with an interquartile range increment in cytokine levels were calculated using logistic regression, adjusting for sociodemographic and birth characteristics. We found that patients with ALL were born with higher levels of a group of correlated cytokines than controls [IL1β: OR of 1.18 (95% confidence interval [CI], 1.03-1.35); IL8: 1.19 (1.03-1.38); TNFα: 1.15 (1.01-1.30); VEGF: 1.16 (1.01-1.33)], especially among children of Latina mothers (ORs from 1.31 to 1.40) and for ALL with high hyperdiploidy (ORs as high as 1.27). We found that neonatal cytokine levels were correlated with neonatal levels of endogenous metabolites which had been previously associated with ALL risk; however, there was no evidence that the cytokines were mediating the relationship between these metabolites and ALL risk. We posit that children born with altered cytokine levels are set on a trajectory towards an increased risk for subsequent aberrant immune reactions that can initiate ALL. This is the first study to evaluate the interplay between levels of immunomodulatory cytokines at birth, prenatal exposures, and the risk of childhood ALL.

Epigenetic Biomarkers of Prenatal Tobacco Smoke Exposure Are Associated with Gene Deletions in Childhood Acute Lymphoblastic Leukemia.

Parental smoking is implicated in the etiology of acute lymphoblastic leukemia (ALL), the most common childhood cancer. We recently reported an association between an epigenetic biomarker of early-life tobacco smoke exposure at the AHRR gene and increased frequency of somatic gene deletions among ALL cases. Here, we further assess this association using two epigenetic biomarkers for maternal smoking during pregnancy-DNA methylation at AHRR CpG cg05575921 and a recently established polyepigenetic smoking score-in an expanded set of 482 B-cell ALL (B-ALL) cases in the California Childhood Leukemia Study with available Illumina 450K or MethylationEPIC array data. Multivariable Poisson regression models were used to test the associations between the epigenetic biomarkers and gene deletion numbers. We found an association between DNA methylation at AHRR CpG cg05575921 and deletion number among 284 childhood B-ALL cases with MethylationEPIC array data, with a ratio of means (RM) of 1.31 [95% confidence interval (CI), 1.02-1.69] for each 0.1 β value reduction in DNA methylation, an effect size similar to our previous report in an independent set of 198 B-ALL cases with 450K array data [meta-analysis summary RM (sRM) = 1.32; 95% CI, 1.10-1.57]. The polyepigenetic smoking score was positively associated with gene deletion frequency among all 482 B-ALL cases (sRM = 1.31 for each 4-unit increase in score; 95% CI, 1.09-1.57). We provide further evidence that prenatal tobacco-smoke exposure may influence the generation of somatic copy-number deletions in childhood B-ALL. Analyses of deletion breakpoint sequences are required to further understand the mutagenic effects of tobacco smoke in childhood ALL.

Untargeted metabolomics of newborn dried blood spots reveals sex-specific associations with pediatric acute myeloid leukemia

The etiology of pediatric acute myeloid leukemia (AML) is largely unknown, but evidence for mutations in utero and long latency periods suggests that environmental factors play a role. Therefore, we used untargeted metabolomics of archived newborn dried blood spots (DBS) to investigate neonatal exposures as potential causal risk factors for AML. Untargeted metabolomics profiling was performed on DBS punches from 48 pediatric patients with AML and 46 healthy controls as part of the California Childhood Leukemia Study (CCLS). Because sex disparities are suggested by differences in AML incidence rates, metabolite features associated with AML were identified in analyses stratified by sex. There was no overlap between the 16 predictors of AML in females and 15 predictors in males, suggesting that neonatal metabolomic profiles of pediatric AML risk are sex-specific. In females, four predictors of AML were putatively annotated as ceramides, a class of metabolites that has been linked with cancer cell proliferation. In females, two metabolite predictors of AML were strongly correlated with breastfeeding duration, indicating a possible biological link between this putative protective risk factor and childhood leukemia. In males, a heterogeneous metabolite profile of AML predictors was observed. Replication with larger participant numbers is required to validate findings.

Bayesian Group Index Regression for Modeling Chemical Mixtures and Cancer Risk.

There has been a growing interest in the literature on multiple environmental risk factors for diseases and an increasing emphasis on assessing multiple environmental exposures simultaneously in epidemiologic studies of cancer. One method used to analyze exposure to multiple chemical exposures is weighted quantile sum (WQS) regression. While WQS regression has been demonstrated to have good sensitivity and specificity when identifying important exposures, it has limitations including a two-step model fitting process that decreases power and model stability and a requirement that all exposures in the weighted index have associations in the same direction with the outcome, which is not realistic when chemicals in different classes have different directions and magnitude of association with a health outcome. Grouped WQS (GWQS) was proposed to allow for multiple groups of chemicals in the model where different magnitude and direction of associations are possible for each group. However, GWQS shares the limitation of WQS of a two-step estimation process and splitting of data into training and validation sets. In this paper, we propose a Bayesian group index model to avoid the estimation limitation of GWQS while having multiple exposure indices in the model. To evaluate the performance of the Bayesian group index model, we conducted a simulation study with several different exposure scenarios. We also applied the Bayesian group index method to analyze childhood leukemia risk in the California Childhood Leukemia Study (CCLS). The results showed that the Bayesian group index model had slightly better power for exposure effects and specificity and sensitivity in identifying important chemical exposure components compared with the existing frequentist method, particularly for small sample sizes. In the application to the CCLS, we found a significant negative association for insecticides, with the most important chemical being carbaryl. In addition, for children who were born and raised in the home where dust samples were taken, there was a significant positive association for herbicides with dacthal being the most important exposure. In conclusion, our approach of the Bayesian group index model appears able to make a substantial contribution to the field of environmental epidemiology.

Assessment of Grouped Weighted Quantile Sum Regression for Modeling Chemical Mixtures and Cancer Risk.

Individuals are exposed to a large number of diverse environmental chemicals simultaneously and the evaluation of multiple chemical exposures is important for identifying cancer risk factors. The measurement of a large number of chemicals (the exposome) in epidemiologic studies is allowing for a more comprehensive assessment of cancer risk factors than was done in earlier studies that focused on only a few chemicals. Empirical evidence from epidemiologic studies shows that chemicals from different chemical classes have different magnitudes and directions of association with cancers. Given increasing data availability, there is a need for the development and assessment of statistical methods to model environmental cancer risk that considers a large number of diverse chemicals with different effects for different chemical classes. The method of grouped weighted quantile sum (GWQS) regression allows for multiple groups of chemicals to be considered in the model such that different magnitudes and directions of associations are possible for each group of chemicals. In this paper, we assessed the ability of GWQS regression to estimate exposure effects for multiple chemical groups and correctly identify important chemicals in each group using a simulation study. We compared the performance of GWQS regression with WQS regression, the least absolute shrinkage and selection operator (lasso), and the group lasso in estimating exposure effects and identifying important chemicals. The simulation study results demonstrate that GWQS is an effective method for modeling exposure to multiple groups of chemicals and compares favorably with other methods used in mixture analysis. As an application, we used GWQS regression in the California Childhood Leukemia Study (CCLS), a population-based case-control study of childhood leukemia in California to estimate exposure effects for many chemical classes while also adjusting for demographic factors. The CCLS analysis found evidence of a positive association between exposure to the herbicide dacthal and an increased risk of childhood leukemia.

Untargeted metabolomics profiles of newborn dried blood spots and pediatric acute myeloid leukemia: a pilot study

Background: Leukemia is the most common childhood cancer. Compared to acute lymphoblastic leukemia, which makes up 80% of pediatric leukemia cases, acute myeloid leukemia (AML) is rarer and has lower survival rates. The etiology of pediatric AML is largely unknown, but evidence for the requirement of multiple mutations in utero and long latency periods (of up to 15 years) suggests that environmental factors play a role in causing the disease. Untargeted metabolomics of archived newborn dried blood spots (DBS) provides a means for retrospective investigation of in utero exposures as potential causal risk factors for rare pediatric diseases. Using DBS, we discovered untargeted metabolomics profiles at birth, which are associated with subsequent development of AML in childhood.Methods: Following AML diagnosis, the California Department of Public Health provided archived newborn DBS from 48 pediatric patients and 46 healthy controls matched by sex, ethnicity, and age as part of the California Childhood Leukemia Study (CCLS). Using a single 4.7-mm punch of DBS, untargeted metabolomics profiling was performed with liquid chromatography high-resolution mass spectrometry (LC-HRMS). Metabolomics features associated with AML were identified in analyses stratified by sex.Results: An ensemble of feature selection methods found 8 predictors of AML in females with fold-changes ranging from 0.84-1.88 and 16 different predictors of AML in males with fold-changes ranging 0.75 to 1.24. Two of the metabolites positively associated with AML in females were putatively annotated as ceramides, a class of metabolites that has been linked with cancer cells.Conclusions: Untargeted metabolomics of DBS revealed sex-specific predictors of pediatric AML, suggesting different early life biology. Replication with larger numbers of subjects is required to validate the findings.

Predisposing germline mutations in high hyperdiploid acute lymphoblastic leukemia in children.

High hyperdiploidy (HD) is the most common cytogenetic subtype of childhood acute lymphoblastic leukemia (ALL), and a higher incidence of HD has been reported in ALL patients with congenital cancer syndromes. We assessed the frequency of predisposing germline mutations in 57 HD-ALL patients from the California Childhood Leukemia Study via targeted sequencing of cancer-relevant genes. Three out of 57 patients (5.3%) harbored confirmed germline mutations that were likely causal, in NBN, ETV6, and FLT3, with an additional six patients (10.5%) harboring putative predisposing mutations that were rare in unselected individuals (<0.01% allele frequency in the Exome Aggregation Consortium, ExAC) and predicted functional (scaled CADD score ≥ 20) in known or potential ALL predisposition genes (SH2B3, CREBBP, PMS2, MLL, ABL1, and MYH9). Three additional patients carried rare and predicted damaging germline mutations in GAB2, a known activator of the ERK/MAPK and PI3K/AKT pathways and binding partner of PTPN11-encoded SHP2. The frequency of rare and predicted functional germline GAB2 mutations was significantly higher in our patients (2.6%) than in ExAC (0.28%, P = 4.4 × 10-3 ), an observation that was replicated in ALL patients from the TARGET project (P = .034). We cloned patient GAB2 mutations and expressed mutant proteins in HEK293 cells and found that frameshift mutation P621fs led to reduced SHP2 binding and ERK1/2 phosphorylation but significantly increased AKT phosphorylation, suggesting possible RAS-independent leukemogenic effects. Our results support a significant contribution of rare, high penetrance germline mutations to HD-ALL etiology, and pinpoint GAB2 as a putative novel ALL predisposition gene.

Germline GAB2 Mutations in Childhood Acute Lymphoblastic Leukemia

Recent studies using next-generation sequencing of selected individuals, such as those with familial leukemia or congenital syndromes, have identified rare and highly penetrant germline mutations that predispose to childhood acute lymphoblastic leukemia (ALL). High hyperdiploidy (HD), the most common cytogenetic subtype of childhood ALL, is enriched in children with RASopathies who develop ALL and, similarly, a high proportion of ALL patients with germline ETV6 or IKZF1 mutations presented with the HD subtype. Here, we aimed to identify novel predisposition genes in a selected group of HD-ALL patients.Targeted sequencing of 538 cancer-relevant genes was carried out using the UCSF500 Cancer Gene Panel in diagnostic bone marrow (i.e. tumor) DNA from 57 HD-ALL patients from the California Childhood Leukemia Study (CCLS). HD-ALL patients were selected based on absence of somatic KRAS or NRAS hotspot mutations detectable by Sanger sequencing, and absence of somatic copy number deletions from multiplex ligation-dependent probe amplification (MLPA) assays. After filtering out likely somatic mutations (mutant allele fraction <0.44), and restricting to variants with low frequency in unselected individuals (allele frequency <0.01% in the Exome Aggregation Consortium, ExAC) and with predicted functional effects (Combined Annotation Dependent Depletion, CADD score ≥20), we identified 151 putative predisposing mutations. Of 41 mutations of interest selected for validation, 37 (90.2%) were confirmed as germline in origin via Sanger sequencing of remission or newborn bloodspot DNA.Rare and predicted functional germline mutations in known (NBN, SH2B3, ETV6, CREBBP, MSH6) or suspected (MLL, ABL1, FLT3, MYH9) ALL predisposition genes were identified in nine out of 57 patients (15.8%). Three additional patients harbored germline mutations in the GRB2-associated binding protein 2 (GAB2), a known binding partner of PTPN11-encoded SHP2 and activator of the ERK/MAPK and PI3K/AKT pathways. Two GAB2 mutations, a missense mutation S592F and frameshift mutation P621fs, were predicted to be highly functional (CADD scores = 34 and 36 respectively) and absent in ExAC. Frequency of rare and damaging GAB2 mutations was significantly higher in our patient set (2.6%) than in ExAC (0.28%, P = 2.70 x 10-6). We replicated this finding in sequencing data from 309 ALL patients in the TARGET (Therapeutically Applicable Research to Generate Effective Treatments) project (0.81% vs. 0.28%, P = 0.015).Patient GAB2 mutations were cloned into HEK293 cells and, following EGF stimulation, we found that the P621fs mutation reduced SHP2 binding and ERK1/2 phosphorylation but increased AKT phosphorylation. This suggested possible Ras-independent leukemogenic effects, supported by a lack of somatic Ras pathway mutations in the three GAB2 mutant patients. Additional functional analyses and sequencing of larger patient cohorts will be required to elucidate the role of germline GAB2 mutations in childhood ALL. DisclosuresNo relevant conflicts of interest to declare.

Allergies and Childhood Acute Lymphoblastic Leukemia: A Case-Control Study and Meta-analysis.

Background: Allergic disease is suspected to play a role in the development of childhood acute lymphoblastic leukemia (ALL). Studies conducted over the last several decades have yielded mixed results.Methods: We examined the association between allergy, a common immune-mediated disorder, and ALL in the California Childhood Leukemia Study (CCLS), a case-control study of 977 children diagnosed with ALL and 1,037 matched controls (1995-2015). History of allergies in the first year of life was obtained from interviews, mainly reported by mothers. Logistic regression analyses were conducted to estimate ORs and 95% confidence intervals (CIs), controlling for birth order, daycare attendance, and mode of delivery. In addition, we conducted meta-analyses with data from the CCLS and 12 published studies and employed a new method to estimate between-study heterogeneity (R_b).Results: Overall, no associations were observed between childhood ALL risk and specific allergy phenotypes or any allergy, as a group. However, having any allergy was associated with an increased risk of ALL among the youngest study participants. In the meta-analysis random-effects models, reduced odds of ALL were associated with hay fever (metaOR = 0.65; 95% CI, 0.47-0.90); however, restricting the analysis to studies that used medical records for assessment of allergy or recently published studies led to null or attenuated results.Conclusions: Overall, our findings do not support a clear association between allergy and childhood ALL.Impact: The degree to which epidemiologic studies can inform the relationship between allergies and risk of childhood ALL is limited by R_b. Cancer Epidemiol Biomarkers Prev; 27(10); 1142-50. ©2018 AACR.

Untargeted Metabolomics of Archived Dried Blood Spots Reveals Lipid Modulation at Birth Associated with Pediatric Acute Lymphoblastic Leukemia

Acute lymphoblastic leukemia (ALL) is the most common form of childhood leukemia. Metabolomics analysis of archived dried blood spots (DBS) collected at birth provides a snapshot of early life exposures that can provide insights into the causes and underlying prenatal biology of ALL.We received single 4.7-mm DBS punches (equivalent to ~8 µL whole blood) collected between 1985 and 2005 for over 300 ALL cases and their matched controls as part of the California Childhood Leukemia Study. Matching was performed on date of birth and gender and roughly matched on ethnicity. DBS were extracted with acetonitrile and analyzed with an Agilent 1290 UHPLC system connected to a 6550 QTOF HRMS (Santa Clara, US) in ESI (-) mode. Analysis was performed on subjects stratified by age at diagnosis as a means of distinguishing between potentially different ALL phenotypes (early: 1-5 years; late: 6-14 years). In order to minimize the number of false positives, associations between metabolite features and ALL were determined with an ensemble of statistical methods, namely, linear regression p-values, bootstrapped regularized logistic regression (LASSO), and random forest.There was no overlap in features selected for associations with early diagnosis (9 features) and late diagnosis (24 features) of ALL. Annotated metabolites were primarily fatty acid and glycerophospholipids. Several of the selected metabolites were associated with the time to ALL diagnosis, suggesting that they reflect effects of disease progression rather than causal exposures. Interestingly, several metabolites selected in the late diagnosis group were highly associated with breastfeeding practice, a known risk factor of ALL. Since DBS are collected within 48 hours of birth, they include nutrients from one or more infant feedings. These results suggest that lipids associated with neonatal nutrition may be involved with initiation of ALL in early life and should be a focus of future targeted research.

GWAS in childhood acute lymphoblastic leukemia reveals novel genetic associations at chromosomes 17q12 and 8q24.21

Childhood acute lymphoblastic leukemia (ALL) (age 0–14 years) is 20% more common in Latino Americans than non-Latino whites. We conduct a genome-wide association study in a large sample of 3263 Californian children with ALL (including 1949 of Latino heritage) and 3506 controls matched on month and year of birth, sex, and ethnicity, and an additional 12,471 controls from the Kaiser Resource for Genetic Epidemiology Research on Aging Cohort. Replication of the strongest genetic associations is performed in two independent datasets from the Children’s Oncology Group and the California Childhood Leukemia Study. Here we identify new risk loci on 17q12 near IKZF3/ZPBP2/GSDMB/ORMDL3, a locus encompassing a transcription factor important for lymphocyte development (IKZF3), and at an 8q24 region known for structural contacts with the MYC oncogene. These new risk loci may impact gene expression via local (four 17q12 genes) or long-range (8q24) interactions, affecting function of well-characterized hematopoietic and growth-regulation pathways.

Correlates of Prenatal and Early-Life Tobacco Smoke Exposure and Frequency of Common Gene Deletions in Childhood Acute Lymphoblastic Leukemia.

Tobacco smoke exposure has been associated with risk of childhood acute lymphoblastic leukemia (ALL). Understanding the relationship between tobacco exposures and specific mutations may yield etiologic insights. We carried out a case-only analysis to explore whether prenatal and early-life tobacco smoke exposure influences the formation of leukemogenic genomic deletions. Somatic copy number of 8 genes frequently deleted in ALL (CDKN2A, ETV6, IKZF1, PAX5, RB1, BTG1, PAR1 region, and EBF1) was assessed in 559 pretreatment tumor samples from the California Childhood Leukemia Study. Parent and child's passive tobacco exposure was assessed using interview-assisted questionnaires as well as DNA methylation in aryl-hydrocarbon receptor repressor (AHRR), a sentinel epigenetic biomarker of exposure to maternal smoking during pregnancy. Multivariable Poisson regressions were used to test the association between the smoking exposures and total number of deletions. Deletion burden varied by subtype, with a lower frequency in high-hyperdiploid and higher frequency in ETV6-RUNX1 fusion ALL. The total number of deletions per case was positively associated with tobacco smoke exposure, in particular for maternal ever-smoking (ratio of means, RM, 1.31; 95% CI, 1.08-1.59), maternal smoking during pregnancy (RM, 1.48; 95% CI, 1.12-1.94), and during breastfeeding (RM, 2.11; 95% CI, 1.48-3.02). The magnitude of association with maternal ever-smoking was stronger in male children compared with females (Pinteraction = 0.04). The total number of deletions was also associated with DNA methylation at the AHRR epigenetic biomarker (RM, 1.32; 95% CI, 1.02-1.69). Our results suggest that prenatal and early-life tobacco smoke exposure increase the frequency of somatic deletions in children who develop ALL. Cancer Res; 77(7); 1674-83. ©2017 AACR.