Prevalence of germline cancer predisposition pathogenic variants in cancer cohorts.

Abstract

e22667 Background: Clinical practice guidelines recommend germline genetic testing for a broadening scope of incident adult malignancies. The objective of this study was to identify the frequency of germline pathogenic variants (PVs) in cancer cohorts. Methods: We conducted a systematic review and meta-analysis in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines of the proportion of individuals with incident cancer who harbor cancer germline predisposition PVs. We searched MEDLINE, Embase, and CENTRAL from inception through August 2025. The primary outcome evaluated was the percentage of those with incident malignancy who harbor a germline cancer predisposition PV. Meta-analysis was conducted using R software (Version 4.4.2 (6/14/2024)). Random effects models were reported. Risk of bias was evaluated using tools from the Joanna Briggs Institute. Publication bias was assessed by evaluating funnel plots. Results: In mixed cancer cohorts, 8 heterogeneous moderate-quality studies without significant evidence of publication bias revealed a pooled population of 40,436 individuals with 15.6% [95% CI 12 – 21] harboring a germline cancer predisposition PV and BRCA 2 (21%) BRCA 1(16%), CHEK2 (10%), and ATM (8%) PVs being most common. PV stage distribution was Stage 0 3.4% [95% CI 1.0-5.8], Stage I 5.5% [95% CI 1.0-25], Stage II 12.2% [95% CI 0.5-80], Stage III 37% [95% CI 1.0-90], and Stage IV 38.1% [95% CI 2.6-94]. Approximately 29% [95% CI 4.5%-79] of those with PVs did not meet NCCN criteria for germline cancer genetic testing. Variants of unknown significance were found in 43% [95% CI 26-62] and negative test results in 54% [95% CI 34-72%]. Where demographics were reported, mean age was 55 years, 14.9% of participants were non-Hispanic White, 2.7% Black, 4.3% Hispanic, 0.3% Asian, and the remaining classified as “other/unknown.” Approximately 64% of participants were women. Prevalence of PV in publications that reported on a singular cancer type was as follows: ovarian [14.1% 95% CI 6.4 -28] (3 studies, 4130 participants); colorectal cancer [13.1% 95% CI 9.7 -17] (3 studies 1899 participants); prostate [12% 95% CI 9.0-15] (12 studies 36,839 participants); endometrial cancer [11% 95% CI 5.1-20] (4 studies 1884 participants); pancreatic [10% 95% CI 5.3-18] (7 studies, 5087 participants); breast [9.1% 95% CI 7.0-12] (11 studies of 112,110 participants). Single cancer studies were heterogeneous, of low/moderate quality, and with no evidence of publication bias. Conclusions: Approximately 16% of those in a mixed-cancer cohort harbor cancer germline pathogenic variants, with trends towards increased prevalence by cancer stage. PV prevalence in select single cancer cohorts ranged from 9 – 14%.