Importance: Understanding germline and somatic status in patients with gynecologic cancers could improve risk assessment and guide therapeutic decision-making.
Objective: To evaluate the prevalence and landscape of germline pathogenic or likely pathogenic (P/LP) variants and explore whether these variants are associated with somatic phenotypes and cancer risk in unselected patients with gynecologic cancers.
Design, setting, and participants: This cross-sectional study retrospectively enrolled unselected patients in China with a gynecologic cancer, including ovarian, cervical, and endometrial, who underwent tumor-normal sequencing using a 520-gene panel from October 1, 2017, through May 31, 2021.
Exposure: Germline variants in gynecologic cancers.
Main outcomes and measures: The P/LP germline variant rates in 62 cancer predisposition genes were assessed using descriptive statistics. The associations of P/LP variant status with age, somatic profiles, and cancer risk were also investigated using the Fisher exact test or Student t test.
Results: A total of 1610 women (median [IQR] age, 54 [47-62] years; 1201 [74.6%] with stage III-IV disease) were included (945 with ovarian cancer, 307 with endometrial cancer, and 358 with cervical cancer). The prevalence of patients with P/LP variants was 20.5% (194 of 945) for ovarian cancer, 13.4% (41 of 307) for endometrial cancer, and 6.4% (23 of 358) for cervical cancer; 95.1% of the germline findings (n = 252) were potentially actionable, mainly in homologous recombination repair (HRR) and mismatch repair genes. Chinese patients with endometrial cancer had a higher rate of P/LP variants than a White population from The Cancer Genome Atlas (42 of 307 [13.7%] vs 24 of 367 [6.5%]; P = .003). In endometrial and cervical cancers, the prevalence of P/LP variants was 12.7% (30 of 237) and 4.8% (13 of 270), respectively, in patients diagnosed at age 45 years or older and increased to 25.0% (9 of 36; P = .09) and 12.0% (10 of 83; P = .04), respectively, for those with an onset age of less than 45 years. Mismatch repair P/LP variants were associated with a younger age at onset for ovarian cancer (46 vs 54 years; P = .02) and endometrial cancer (48 vs 57 years; P < .001), while HRR P/LP variants were associated with a younger age at onset for cervical cancer (46 vs 52 years; P = .04). Carriers of HRR P/LP variants had more prevalent somatic TP53 variants and less common somatic variants in oncogenic driver genes vs noncarriers. BRCA1/2 P/LP variants were also associated with moderate risks for endometrial and cervical cancer.
Conclusions and relevance: This study delineates the landscape of germline P/LP variants in Chinese women with gynecologic cancers. The findings highlight the hereditary factor in cervical cancer that has long been neglected and suggest the importance of next-generation sequencing-based genetic testing with a large gene panel for gynecologic cancers.