Abstract
Patients with biallelic mutations in BRCA2 comprise Fanconi Anemia (FA) complementation group FA-D1. We analyzed the severity of the BRCA2 mutations in 26 literature cases and one from our National Cancer Institute Cohort. The functional consequence of each mutant allele was inferred from the Breast Cancer Information Core (BIC) database. Knowledge of gene structure was used to evaluate whether the reported mutations led to inactivation of the BRCA2 gene or diminution of the gene product. We examined genotype/phenotype/cancer correlations between specific BRCA2 mutations and specific malignancies in FA families. We also mapped the location of the FA-related BRCA2 mutations. Five of twenty-seven patients had birth defects consistent with VATER association, a 71,000-fold increase compared with 2.6/106 in the general population, and 5% of unclassified FA patients. Ten patients had acute myeloid leukemia (AML), four acute lymphocytic leukemia (ALL, one had AML as well), twelve brain tumors, seven Wilms tumor, and one neuroblastoma; six patients had ≥2 cancers, and only two patients had no cancer. IVS7+1G>A and IVS7+2T>G were associated with AML (odds ratio 15 compared to those with other mutations in BRCA2, p=0.02), and 886delGT and 6174delT with brain tumors (odds ratios infinite compared to those with other BRCA2 mutations, p=0.006 and 0.03 respectively). Compared with non-FA-D1 patients, the relative risk of any malignancy in FA-D1 was 66 (25/27 patients with 15 solid tumors and 13 leukemias; 3 patient had solid tumors and leukemia); the cumulative probability of any cancer was 97% by age 5. Compared with the age- and sex-adjusted general population, AML was increased 7000-fold, and any malignancy 3300-fold. While half of the families had relatives with BRCA2-type cancers, several of the alleles reported in probands were not associated with cancer in presumed carriers. Two of the reported FA-related mutant BRCA2 alleles were benign polymorphisms (one patient was FA-B with a concurrent mutation and polymorphism in BRCA2), five were missense variants of unknown clinical significance, three involved splice sites, and 20 were frameshift or truncating mutations. Missense mutations formed a distinct cluster between residues 2236 and 2729 of the BRCA2 protein (Pearson statistic p=0.01, range test p<0.001), while the truncation and frameshift mutations did not cluster. This small group of patients with biallelic mutations in BRCA2 is unique in the severity of the physical phenotype, and in the very early onset and inordinately high rates of leukemia and specific solid tumors. In a subset of patients, a single deleterious mutation coupled with a missense mutation seemed sufficient for cancer to occur. Our observations suggest that FA patients with mutations in BRCA2 are distinctly more severely affected than those in other complementation groups. Appropriate counseling strategies need to be developed for family members who are heterozygous for these mutations in BRCA2. BRCA2 regions which are not implicated in cancer in heterozygotes may nonetheless be critical interaction regions for the FA pathway downstream of FANCD2 ubiquitination.
Disclosure: No relevant conflicts of interest to declare.
Author notes
Corresponding author