Investigation of cancer predisposition in Fanconi anemia heterozygotes: A DiscovEHR cohort population study.

Abstract

10589 Background: Fanconi anemia (FA) is a cancer predisposition syndrome caused by biallelic pathogenic variants in one of 22 genes involved in DNA intrastrand crosslink repair. 20 of these genes have an autosomal recessive inheritance pattern. Of the FA genes, 5 ( BRCA1, BRCA2, BRIP1, PALB2, RAD51C) are known cancer predisposition genes (CPG) when inherited in monoallelic autosomal dominant manner, though the heterozygous predisposition status of the remaining 15 remains unclear. Large population-level exome sequencing projects with linked health records such as the DiscovEHR cohort serve as promising tools to establish sufficient power to investigate all FA genes. Quantifying cancer risk is important for the counseling and surveillance of FA heterozygotes, especially with the ever-increasing use of genetic testing identifying these individuals. Methods: 170,503 individuals enrolled in the DiscovEHR Cohort were analyzed for pathogenic/likely pathogenic variants in 22 FA genes and identified 5834 subjects, the case group. ICD10 phenotype classifications for these subjects were curated to Phecodes and ICD-O-3 Site Recodes for two control arms of analysis. The first control arm included all DiscovEHR subjects excluding the case group and those with a variant of uncertain significance in the genes of interest with a deleterious in-silico score. The second control arm included individuals from the Surveillance, Epidemiology, and End Results (SEER) database. Phecode analysis yielded odds ratios adjusted for age, sex, race, smoking status, and BMI. SEER analysis yielded Standardized Incidence Ratio scores adjusted for age, sex, and birth cohort. When both arms of the analysis were completed Phecodes and SEER analyses were matched and positive signals for cancer risk were defined by statistically significant ratios in both. Results: The 5 known CPGs demonstrated multiple signals for cancer predisposition as expected (p < 0.0001). Results for FANCA (p = 0.4231) and FANCC (p = 0.8142) heterozygotes validated recent results from our group showing no increased risk of cancer in relatives of FA patients with these genotypes. Ten other FA genes also showed no increased risk of cancer. 3 genes suggested a possible increased risk of specific cancers in heterozygotes and are currently being evaluated in a second large population exome cohort to validate these findings. Conclusions: This is the largest study to date on the cancer predisposition risk in Fanconi anemia heterozygotes. We have confirmed the known risk in 5 well-described CPGs and the lack of risk amongst FANCA and FANCC heterozygotes, the most common in population. Further analysis via an additional large population cohort is underway to validate any possible novel cancer predispositions in 3 genes. For the remaining FA genes, this population analysis supports previous literature on known CPGs and the lack of cancer predisposition in most FA genes.