Abstract
BACKGROUND: Survivors of childhood cancer are at risk of subsequent neoplasms (SNs) associated with exposure to radiotherapy and chemotherapy, as well as with genetic predisposition. We aimed to estimate the relative contributions of these risk factors to the total SN burden in survivor populations. METHODS: We analysed data from two retrospectively constructed cohorts with ongoing recruitment and prospective follow-up: the St Jude Lifetime Cohort (SJLIFE; 4401 participants; NCT00760656) and the Childhood Cancer Survivor Study (CCSS; 7943 participants; NCT01120353). We used multivariable piecewise-exponential models to calculate attributable fractions to assess the contributions of radiotherapy and chemotherapy exposures, genetic predisposition (comparing the top two tertiles with the lowest tertile of polygenic risk scores [PRSs] where the tertile is from external general population corresponding to SN outcome) and lifestyle factors (physical activity, smoking, alcohol consumption, obesity, and diet) to incident of the first occurrences of SNs as the primary outcome. FINDINGS: The study was conducted between Jan 1, 2024, and Sept 30, 2024. Of the 12 344 survivors eligible for analysis, median attained age was 33·0 years (IQR 24·1-42·1) in SJLIFE and 36·0 years (29·5-43·6) in CCSS; 6127 (49·6%) were men and 6217 (50·4%) were women. Most patients were White (10 907 [88·4%]). The median follow-up from primary cancer diagnosis was 24·2 years (IQR 11·7-35·4) in SJLIFE (from Sept 13, 2007 to April 20, 2020) and 28·0 years (8·9-37·2) in CCSS (from Jan 1, 1975 to Dec 31, 2023). Cancer treatments and genetic risk jointly contributed to a substantial proportion of incident SN cases with attributable fractions ranging from 30% (95% CI 6-49; sarcoma) to 92% (89-94; meningioma). Higher exposure levels of radiotherapy contributed most, particularly in older (≥35 years; SJLIFE proportion of SNs 44·7% [95% CI 41·9-47·5]) compared with younger (<35 years; 40·0% [37·1-43·3]) follow-up age periods. Elevated genetic risk based on the PRSs accounted for a notable proportion, ranging from 1% (95% CI 0-7; meningioma) to 52% (39-62; thyroid cancer), surpassing contributions of chemotherapies, ranging from 3% (1-6; SMNs) to 35% (19-49; sarcoma). Lifestyle factors contributed negligibly. INTERPRETATION: Cancer treatments and genetic predisposition are primary contributors to the risk of SNs in childhood cancer survivors, and lifestyle factors seem to have a minimal effect. These results highlight the crucial need to consider both treatment history and genetic factors in developing effective risk assessment and surveillance strategies for this vulnerable population. FUNDING: US National Institutes of Health and the American Lebanese Syrian Associated Charities.