P-063: Pathogenic germline variants in hereditary cancer genes in patients with Multiple Myeloma

Abstract

<h3>Background</h3> GWAS have identified common SNPs & rare high-penetrance variants that explain ~16% of the estimated heritability of multiple myeloma (MM)(PMID 30213928). Pathogenic/likely-pathogenic germline variants (PGV) in hereditary cancer genes (HCG) are common in adult cancer pts (~8%), but prevalence in MM is not known. The aim of our study is to investigate the occurrence of PGV in newly-diagnosed MM (NDMM) and describe clinical characteristics & outcomes of carriers. <h3>Methods</h3> We analyzed MMRF CoMMpass data & identified 895 NDMM pts with whole-exome sequencing of germline DNA. We used the clinical annotation pipeline from Sema4, a CLIA/CAP certified genetic testing laboratory, to identify pts with PGV according to ACMG variant classification guidelines. We compared clinical characteristics & disease phenotypes of PGV carriers vs non-carriers. We used Chi-Square and Fisher's Exact tests to assess statistical significance. <h3>Results</h3> We identified 83 PGV in 31 distinct HCG in 79 (8.8%) of 895 NDMM pts (83% European ancestry). Most PGV involved DNA damage repair (DDR) genes (78%), & homologous recombination (HR) genes were the most commonly mutated (34%). PGV in CHEK2 were most common (n=10,1.1% of all MM pts). 2 pts carried PGV in TP53, & 6 pts had mismatch repair (MMR) gene defects (1:149). MM pts with family history (FH) of hematologic malignancy (HM) in a 1st or 2nd-degree relative were more likely to carry PGV (22 vs 7.6%,OR=3.3,p<0.001), an association that remained significant in multivariate analysis (MVA) (OR=4.1,p<0.001). CHEK2 variants were leading drivers of this correlation (OR 18,adjusted p<0.01), & especially protein-truncating founder variant c.1100delC. Likelihood of MM dx before age 40 was significantly higher in PGV carriers (6.3 vs 1.8%,OR=3.7,p=0.025). 25% of those <40y/o carried PGV, but none of these were in DDR-HR genes, a notable difference with other age groups (0 vs 41%,p=0.02). 2/6 MMR PGV were detected in pts <40y/o. In univariate analysis, DDR-PGV carriers had a significant PFS1 advantage (median 52 vs 35 months,p=0.008) & a non-significant OS advantage (p=0.08). PFS1 difference remained significant in MVA adjusting for age, ISS stage, high-risk cytogenetics, treatment type & transplant status (OR 0.65,95% CI 0.44-0.97,p=0.03). <h3>Conclusions</h3> PGV in HCG were common (8.8%) in this large cohort of NDMM pts of predominantly European ancestry, especially in those with FH of HM (1:4, with high prevalence of CHEK2 variants), and in those <40y/o (1:4). Routine screening in high-prevalence subgroups may be warranted, as carriers can benefit from counseling & enrollment in early cancer detection programs. We observed a clinically & statistically significant PFS1 advantage in PGV carriers, possibly due to increased sensitivity to MM therapies, a well-described phenomenon in other cancers (PMID 33158305). Prospective validation of these findings is needed to better understand prognostic & therapeutic implications of PGV in MM.