PURPOSE: While cancer risk is increased among children with chromosomal birth defects, less is known about associations between specific non-chromosomal structural birth defects and specific childhood cancers. To investigate these, we established a population-based retrospective cohort of >10 million children by pooling statewide registry data from four U.S. states (Texas, Michigan, North Carolina, and Arkansas) for the period 1992-2013.
METHODS: Individual level data from birth certificates, birth defects registries, and cancer registries were linked in each state; demographic and diagnostic variables were harmonized; and the data were pooled for the overall analysis. We used Cox proportional hazards models to evaluate associations between 60 birth defects and 31 childhood cancers when there were five or more comorbid cases. A hazard ratio (HR) and 95% confidence interval (CI) was calculated for each birth defect-childhood cancer (BD-CC) pairwise combination, adjusted for maternal age, infant sex, and state. The false discovery rate (FDR) was computed via the Benjamini-Hochberg procedure to account for multiple comparisons.
RESULTS: We identified 517,548 children with non-chromosomal structural birth defects and 14,774 children with cancer. The risk of any cancer was increased among children with any non-chromosomal structural defect compared to children without any birth defect (HR=2.6, 95% CI 2.4-2.7). Of 2,511 potential BD-CC combinations, we tested 606 where there were ≥5 comorbid cases and identified 496 BD-CC associations with significantly elevated HRs at a 5% FDR. No significant inverse associations were identified for any BD-CC combination. Notably, hepatoblastoma, astrocytoma, ependymoma, and extracranial germ cell tumors were each strongly associated with several birth defects. For example, the risk of hepatoblastoma was increased among children with atrial septal defects (HR=12.5, 95% CI 7.9-19.7) and craniosynostosis (HR=15.4, 95% CI 7.6-31.3). Astrocytoma and ependymoma were associated with central nervous system (CNS) defects (HR=6.7, 95% CI 4.6-9.8 and HR=7.4, 95%CI 3.5-15.7, respectively). Elevated risk of extracranial germ cell tumors was observed among children with CNS defects (HR=22.5, 95% CI 10.9-46.4) and obstructive genitourinary defects (HR=32.4, 95% CI 16.2-64.6).
CONCLUSIONS: By pooling registry data across four U.S. states, we were able to evaluate specific BD-CC patterns and report several novel associations. Our findings suggest that children with non-chromosomal birth defects have a significantly elevated risk of several childhood cancers. These findings may inform research into the etiologies of childhood cancer, as well as new cancer surveillance protocols for children with non-chromosomal birth defects. This work was supported by the Cancer Prevention Research Institute of Texas.
