Real-world application of comprehensive genomic profiling for gynecological malignancies: a multicenter observational study

Abstract

Background: The actual status of comprehensive genomic profiling (CGP) applications in Japan has not been clarified. We conducted a multicenter study to investigate the real-world application of CGP in gynecological malignancies.

Methods: Nine designated cancer hospitals participated in this study. Patients who underwent CGP in 2020-2021 were assigned to the CGP group (n = 134). For the population that would have been eligible for CGP, patients who received initial treatment in 2015-2016 and were either alive with disease or died of disease at 5 years follow up were included in the control group (n = 316). We compared clinicopathological characteristics including tumor type (cervix, corpus, ovary, and others including sarcoma) and age. We also investigated the context of CGP-recommended treatment.

Results: The CGP group had significantly fewer cervical cases and more others cases (cervix/corpus/ovary/others: CGP, 22/44/56/12; control, 89/79/142/6; p = 0.0003). The CGP group was significantly younger than the control group (median: CGP, 54.0; control, 65.0; p < 0.0001). Subgroup analyses revealed that patients with cervical and ovarian cancers were significantly younger in the CGP group. Among the CGP group, 17 patients (12.7%) received CGP-recommended treatments, 15 of which were not covered by public insurance. The survival time after CGP in 17 patients was longer than in the other 117 cases (median 21 vs. 11 months).

Conclusion: There was significant selection bias in tumor type and age for the application of CGP for gynecological malignancies in clinical practice in Japan. While CGP often recommended drugs not covered by public insurance, prognosis can be improved by use of CGP.

Keywords: Age; Comprehensive genomic profiling; Gynecological malignancies; Selection bias; Tumor type.