Nonchromosomal birth defects and risk of childhood acute leukemia: An assessment in 15 000 leukemia cases and 46 000 controls from the Childhood Cancer and Leukemia International Consortium.

Although recent studies have demonstrated associations between nonchromosomal birth defects and several pediatric cancers, less is known about their role on childhood leukemia susceptibility. Using data from the Childhood Cancer and Leukemia International Consortium, we evaluated associations between nonchromosomal birth defects and childhood leukemia. Pooling consortium data from 18 questionnaire-based and three registry-based case-control studies across 13 countries, we used multivariable logistic regression models to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association between a spectrum of birth defects and leukemia. Our analyses included acute lymphoblastic leukemia (ALL, n = 13 115) and acute myeloid leukemia (AML, n = 2120) cases, along with 46 172 controls. We used the false discovery rate to account for multiple comparisons. In the questionnaire-based studies, the prevalence of birth defects was 5% among cases vs 4% in controls, whereas, in the registry-based studies, the prevalence was 11% among cases vs 7% in controls. In pooled adjusted analyses, there were several notable associations, including (1) digestive system defects and ALL (OR = 2.70, 95% CI: 1.46-4.98); (2) congenital anomalies of the heart and circulatory system and AML (OR = 2.86, 95% CI: 1.81-4.52) and (3) nervous system defects and AML (OR = 4.23, 95% CI: 1.50-11.89). Effect sizes were generally larger in registry-based studies. Overall, our results could point to novel genetic and environmental factors associated with birth defects that could also increase leukemia susceptibility. Additionally, differences between questionnaire- and registry-based studies point to the importance of complementary sources of birth defect phenotype data when exploring these associations.

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.

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.

Monitoring neurocognitive functioning in childhood cancer survivors: evaluation of CogState computerized assessment and the Behavior Rating Inventory of Executive Function (BRIEF).

BACKGROUND: Many childhood cancer survivors develop neurocognitive impairment, negatively affecting education and psychosocial functioning. Recommended comprehensive neuropsychological testing can be time- and cost- intensive for both institutions and patients and their families. It is important to find quick and easily administered surveillance measures to identify those in need of evaluation. METHODS: We evaluated, individually and in combination, the sensitivity and specificity of the 1) Behavior Rating Inventory of Executive Functioning-Metacognition Index (BRIEF-MCI), and 2) CogState Composite Index (computerized assessment of cognition) in identifying below grade-level performance on state-administered tests of reading and mathematics among childhood cancer survivors. RESULTS: The 45 participants (39% female) were a mean age of 7.1 ± 4.4 years at diagnosis, 14.0 ± 3.0 at evaluation, with a history of leukemia (58%), lymphoma (9%), central nervous system tumors (20%), and other tumors (13%). Impairment on the BRIEF-MCI was associated with low sensitivity (26% reading, 41% mathematics) but stronger specificity (88% reading, 96% mathematics). We found similar associations for the CogState Composite Index with sensitivity of 26% for reading and 29% for mathematics and specificity of 92% for both reading and mathematics. Combining the two measures did not improve sensitivity appreciably (47% reading, 59% mathematics) while reducing specificity (84% reading, 88% mathematics). CONCLUSIONS: While individuals identified from the BRIEF-MCI or CogState Composite would likely benefit from a full neuropsychological evaluation given the strong specificity, use of these measures as screening tools is limited. With poor sensitivity, they do not identify many patients with academic difficulties and in need of a full neuropsychological evaluation. Continued effort is required to find screening measures that have both strong sensitivity and specificity.

A task-based assessment of parental occupational exposure to pesticides and childhood acute lymphoblastic leukemia.

OBJECTIVES: Associations between parental occupational pesticide exposure and childhood acute lymphoblastic leukemia (ALL) vary across studies, likely due to different exposure assessment methodologies. METHODS: We assessed parental occupational pesticide exposure from the year before pregnancy to the child's third year of life for 669 children diagnosed with ALL and 1021 controls. We conducted expert rating using task-based job modules (JM) to estimate exposure to pesticides among farmer workers, gardeners, agricultural packers, and pesticide applicators. We compared this method to (1) partial JM using job titles and a brief description, but without completing the task-based questionnaire, and (2) job exposure matrix (JEM) linking job titles to the International Standard Classifications of Occupation Codes. We used unconditional logistic regression to calculate odds ratios (OR) and 95% confidence intervals (95% CI) for ALL cancer risk and pesticide exposure adjusting for child's sex, age, race/ethnicity and household income. RESULTS: Compared to complete JMs, partial JMs and JEM led to 3.1% and 9.4% of parents with pesticide exposure misclassified, respectively. Misclassification was similar in cases and controls. Using complete JMs, we observed an increased risk of ALL for paternal occupational exposure to any pesticides (OR=1.7; 95% CI=1.2, 2.5), with higher risks reported for pesticides to treat nut crops (OR=4.5; 95% CI=0.9, 23.0), and for children diagnosed before five years of age (OR=2.3; 95% CI: 1.3, 4.1). Exposure misclassification from JEM attenuated these associations by about 57%. Maternal occupational pesticide exposure before and after birth was not associated with ALL. CONCLUSIONS: The risk of ALL was elevated in young children with paternal occupational pesticide exposure during the perinatal period, using more detailed occupational information for exposure classification.

A task-based assessment of parental occupational exposure to organic solvents and other compounds and the risk of childhood leukemia in California.

PURPOSE: Data on parental occupational exposures and risk of childhood leukemia lack specificity. Using 19 task-based job modules, we examined the relationship between occupational exposure to organic solvents and other compounds and the risk of leukemia in children. METHODS: Latino (48%) and non-Latino (52%) children with acute lymphoblastic leukemia (ALL; n=670), acute myeloid leukemia (AML; n=104), and controls (n=1021) were enrolled in a study in California (2000-2008). Logistic regression models were used to estimate the odds ratios (ORs) and 95% confidence intervals (CIs), adjusted for socio-demographic factors. RESULTS: Among children with non-Latino fathers, none of the exposures evaluated were associated with risks of ALL and AML. In contrast, exposure to any organic solvents in Latino fathers was associated with an increased risk of childhood ALL (OR=1.48; 95% CI: 1.01-2.16); in multivariable analyses, the OR for chlorinated hydrocarbons was 2.28 (95% CI: 0.97-5.37) while the ORs were close to one for aromatic hydrocarbons, glycol ethers, and other hydrocarbon mixtures. We also observed an increased risk of ALL with exposure to combustion exhaust/polycyclic aromatic hydrocarbons (PAHs) (ORs=1.70; 95% CI: 1.16-2.57, and 1.46; 95% CI: 0.94-2.26 with and without adjustment for chlorinated hydrocarbons, respectively). Moderately elevated risks of ALL were seen with exposure to metals, paints, and wood dust, although not statistically significant. An increased risk was reported for asbestos based on small numbers of exposed Latino fathers. No associations were reported between maternal exposures to any exposures and childhood ALL and AML. CONCLUSIONS: Our data support associations between paternal occupational exposures to chlorinated hydrocarbons, combustion exhaust, metals, and possibly asbestos and the risk of ALL in the children of Latino fathers only.

Somatic Mutation Allelic Ratio Test Using ddPCR (SMART-ddPCR): An Accurate Method for Assessment of Preferential Allelic Imbalance in Tumor DNA.

The extent to which heritable genetic variants can affect tumor development has yet to be fully elucidated. Tumor selection of single nucleotide polymorphism (SNP) risk alleles, a phenomenon called preferential allelic imbalance (PAI), has been demonstrated in some cancer types. We developed a novel application of digital PCR termed Somatic Mutation Allelic Ratio Test using Droplet Digital PCR (SMART-ddPCR) for accurate assessment of tumor PAI, and have applied this method to test the hypothesis that heritable SNPs associated with childhood acute lymphoblastic leukemia (ALL) may demonstrate tumor PAI. These SNPs are located at CDKN2A (rs3731217) and IKZF1 (rs4132601), genes frequently lost in ALL, and at CEBPE (rs2239633), ARID5B (rs7089424), PIP4K2A (rs10764338), and GATA3 (rs3824662), genes located on chromosomes gained in high-hyperdiploid ALL. We established thresholds of AI using constitutional DNA from SNP heterozygotes, and subsequently measured allelic copy number in tumor DNA from 19-142 heterozygote samples per SNP locus. We did not find significant tumor PAI at these loci, though CDKN2A and IKZF1 SNPs showed a trend towards preferential selection of the risk allele (p = 0.17 and p = 0.23, respectively). Using a genomic copy number control ddPCR assay, we investigated somatic copy number alterations (SCNA) underlying AI at CDKN2A and IKZF1, revealing a complex range of alterations including homozygous and hemizygous deletions and copy-neutral loss of heterozygosity, with varying degrees of clonality. Copy number estimates from ddPCR showed high agreement with those from multiplex ligation-dependent probe amplification (MLPA) assays. We demonstrate that SMART-ddPCR is a highly accurate method for investigation of tumor PAI and for assessment of the somatic alterations underlying AI. Furthermore, analysis of publicly available data from The Cancer Genome Atlas identified 16 recurrent SCNA loci that contain heritable cancer risk SNPs associated with a matching tumor type, and which represent candidate PAI regions warranting further investigation.

Children's Cancer and Environmental Exposures: Professional Attitudes and Practices.

BACKGROUND: Epidemiologic studies worldwide have provided substantial evidence of the contributions of environmental exposures to the development of childhood cancer, yet this knowledge has not been integrated into the routine practice of clinicians who care for children with this disease. To identify the basis of this deficit, we sought to assess the environmental history-taking behavior and perceptions of environmental health among pediatric hematologists and oncologists. PROCEDURE: A web-based survey was sent from June to October 2012 to 427 pediatric oncologists, fellows, and nurse practitioners from 20 US institutions, with an overall response rate of 45%. RESULTS: Survey responses indicated that environmental exposures are of concern to clinicians. The vast majority of respondents (88%) reported receiving questions from families about the relationship between certain environmental exposures and the cancers they regularly treat. However, a lack of comfort with these topics seems to have limited their discussions with families about the role of environmental exposures in childhood cancer pathogenesis. Although 77% of respondents suspected that some of the cases they saw had an environmental origin, their methods of taking environmental histories varied widely. Over 90% of respondents believed that more knowledge of the associations between environmental exposures and childhood cancer would be helpful in addressing these issues with patients. CONCLUSIONS: Although limited in size and representativeness of participating institutions, the results of this survey indicate a need for increased training for hematology/oncology clinicians about environmental health exposures related to cancer and prompt translation of emerging research findings in biomedical journals that clinicians read.

Persistent organic pollutants in dust from older homes: learning from lead.

OBJECTIVES: We aimed to (1) evaluate the relation between home age and concentrations of multiple chemical contaminants in settled dust and (2) discuss the feasibility of using lead hazard controls to reduce children's exposure to persistent organic pollutants. METHODS: As part of the California Childhood Leukemia Study, from 2001 to 2007, we used a high-volume small surface sampler and household vacuum cleaners to collect dust samples from 583 homes and analyzed the samples for 94 chemicals with gas chromatography-mass spectrometry and inductively coupled plasma mass spectrometry. We evaluated relations between chemical concentrations in dust and home age with Spearman rank correlation coefficients. RESULTS: Dust concentrations of lead, polychlorinated biphenyls, organochlorine insecticides, and polycyclic aromatic hydrocarbons were correlated with home age (ρ > 0.2; P < .001), whereas concentrations of pyrethroid insecticides and polybrominated diphenyl ethers were not. CONCLUSIONS: Dust in older homes contains higher levels of multiple, persistent chemicals than does dust in newer homes. Further development of strategies to reduce chemical exposures for children living in older homes is warranted.

Potential role of selection bias in the association between childhood leukemia and residential magnetic fields exposure: a population-based assessment.

PURPOSE: Data from the Northern California Childhood Leukemia Study (NCCLS) were used to assess whether selection bias may explain the association between residential magnetic fields (assessed by wire codes) and childhood leukemia as previously observed in case-control studies. METHODS: Wiring codes were calculated for participating cases, n=310; and non-participating cases, n=66; as well as for three control groups: first-choice participating, n=174; first-choice non-participating, n=252; and replacement (non-first choice participating controls), n=220. RESULTS: Participating controls tended to be of higher socioeconomic status than non-participating controls, and lower socioeconomic status was related to higher wire-codes. The odds ratio (OR) for developing childhood leukemia associated with high wire-codes was 1.18 (95% CI: 0.85, 1.64) when all cases were compared to all first-choice controls (participating and non-participating). The OR for developing childhood leukemia in the high current category was 1.43 (95% CI: 0.91, 2.26) when participating cases were compared to first-choice participating controls, but no associations were observed when participating cases were compared to non-participating controls (OR=1.06, 95% CI: 0.71, 1.57) or to replacement controls (OR=1.06, 95% CI: 0.71, 1.60). CONCLUSIONS: The observed risk estimates vary by type of control group, and no statistically significant association between wire codes and childhood leukemia is observed in the California population participating in the NCCLS.

Determinants of polycyclic aromatic hydrocarbon levels in house dust.

Estimation of human exposures to polycyclic aromatic hydrocarbons (PAHs) is often desired for the epidemiological studies of cancer. One way to obtain information about indoor levels of PAHs is to measure these chemicals in house dust. In this study, we evaluated the predictive value of self-reported and geographic data for estimating measured levels of nine PAHs in house dust from 583 households in the Northern California Childhood Leukemia Study (NCCLS). Using multivariable linear regression models, we evaluated the effects on house-dust PAH concentrations from the following covariates: residential heating sources, smoking habits, house characteristics, and outdoor emission sources. House dust was collected from 2001 to 2007, using both high-volume surface samplers and household vacuum cleaners, and was analyzed for nine PAHs using gas chromatography-mass spectrometry. All nine PAHs were detected in more than 93% of dust samples, with median concentrations ranging from 14 to 94 ng/g dust. Statistically significant effects on PAH concentrations in house dust were found for gas heating, outdoor PAH concentrations, and residence age. Yet, the optimal regression model only explained 15% of the variation in PAH levels in house dust. As self-reported data and outdoor PAH sources were only marginally predictive of observed PAH levels, we recommend that PAH concentrations be measured directly in dust samples for use in epidemiological studies.

Is house-dust nicotine a good surrogate for household smoking?

The literature is inconsistent regarding associations between parental smoking and childhood leukemia, possibly because previous studies used self-reported smoking habits as surrogates for children's true exposures to cigarette smoke. Here, the authors investigated the use of nicotine concentrations in house dust as measures of children's exposure to cigarette smoke in 469 households from the Northern California Childhood Leukemia Study (1999-2007). House dust was collected by using high-volume surface samplers and household vacuum cleaners and was analyzed for nicotine via gas chromatography-mass spectrometry. Using multivariable linear regression, the authors evaluated the effects of self-reported parental smoking, parental demographics, house characteristics, and other covariates on house-dust nicotine concentrations. They observed that nicotine concentrations in house dust were associated with self-reported smoking for periods of months and years before dust collection. Furthermore, the authors found that the relation between nicotine dust levels and self-reported smoking varied by parental age and socioeconomic status. These findings suggest that house-dust nicotine concentrations reflect long-term exposures to cigarette smoke in the home and that they may be less biased surrogates for children's exposures to cigarette smoke than self-reported smoking habits.