Genetic and clinical factors associated with obesity among adult survivors of childhood cancer: A report from the St. Jude Lifetime Cohort

Abstract

Purpose We aimed to identify treatment and genetic factors associated with obesity among childhood cancer survivors. Methods Participants included 1996 survivors previously treated for cancer at St. Jude Children’s Research Hospital who survived ≥10 years from diagnosis (median age at diagnosis 7.2 years, median age at follow-up 32.4 years). Obesity was defined as a body mass index ≥30kg/m2. Factors associated with adult obesity were identified by subgroup-specific (cranial radiation (CRT) exposure status) multivariable logistic regression. Single nucleotide polymorphisms (SNPs) associated with obesity were identified by subgroup-specific exploratory genome-wide association analyses (two-stage resampling approach with type I error rate=5×10−6). Results Forty-seven percent and 29.4% of survivors treated with or without receive CRT were obese at evaluation, respectively. In multivariable analyses, abdominal/pelvic radiation exposure was associated with decreased prevalence of obesity among survivors regardless of cranial radiation (p<0.0001). The odds of obesity were increased among survivors treated with CRT who had also received glucocorticoids (p=0.014), or who were younger at diagnosis (p=0.013). Among survivors treated with CRT, 166 SNPs were associated with obesity. The strongest association was observed with rs35669975 (p=3.3×10−8) on 13q33.3, approximately 30kb downstream of FAM155A. SNPs within GLRA3, and near SOX11 and CDH18 were also identified. These genes have been implicated in neural growth, repair, and connectivity. Conclusion Obesity in childhood cancer survivors remains associated with previous CRT and glucocorticoid exposures. Genetic variants related to neural connectivity may modify the risk of obesity among survivors treated with cranial radiation. Validation of our findings in independent cohorts is required.

Impact of Age at Diagnosis and Hypothalamic Involvement on Body Mass Index Z-Score Change in Pediatric Brain Tumor Survivors

“…Our finding that there was no increase in BMI z-score in patients with chemotherapy and radiation therapy is inconsistent with that of prior studies which have found that injury to the CNS via radiation [2,3,5,7,8,20,24,27] and chemotherapy [3] is associated with increased likelihood for obesity. One possibility for the discrepancy is the difference in dose of radiation.…”

Obesity and Metabolic Syndrome Among Adult Survivors of Childhood Leukemia

“…Self-report data from the CCSS are supported by self-report measures in other cohort studies [10]. However, data from clinical cohorts, where BMI is ascertained using objective measures, indicate obesity rates ranging from 31% at a mean age of 22.3 years [11] to 42.8% at a median age of 32.4 years [12]. These rates are higher than expected given that the proportion of 20–39 year olds in the general population who are obese is 30.3% [13].…”

“…The most frequently documented risk factors for obesity following treatment for childhood ALL are CRT, younger age at diagnosis, and female sex [11, 8, 12]. In the CCSS, females who received CRT doses in excess of 20 Gy had the highest odds (OR 3.81, 95% CI 2.34–5.99) of obesity when compared to siblings [8].…”

“…While the exact function of the FTO gene is unknown, it has been associated with regulation of global metabolic rate, energy expenditure, energy homeostasis, body size, and body fat accumulation [35, 36]. Wilson et al [12] conducted a GWAS among 1,996 adult survivors of childhood cancer (42% were ALL survivors) and found potential genetic predictors of obesity on chromosomes 13 (FAM155A) , 2 ( SOX11 ), 4 (GLRA3) and 5 ( CDH18 and BASP 1) among those exposed to CRT. FAM155A is expressed in the hypothalamus and pituitary, GLRA3 codes for a receptor protein involved in signaling the glycine neurotransmitter, and CDH18 and SOX11 influence neuronal growth, repair and connectivity, so these findings may support the hypothesis that CRT induced neuronal damage influences obesity.…”

Evaluation of a two-step iterative resampling procedure for internal validation of genome-wide association studies

“…An example is the evaluation of genetic predictors of clinically ascertained outcomes in the SJLIFE cohort 9 , a study among childhood cancer survivors treated at St. Jude Children’s Research Hospital (SJCRH), who have survived 10 or more years from diagnosis and are at least 18 years of age. Because this study includes the largest cohort of childhood cancer survivors with prospective medical/clinical evaluation of health outcomes, it is extremely hard to find another cohort that has similarly ascertained health phenotypes 10 . In such situations, it is imperative that an innovative and robust internal validation approach is undertaken to validate the top SNPs identified through GWAS 10,11 .…”