Abstract
Ovarian cancer (OC) is the least survivable gynecological malignancy and presents late. Five-year survival for OC is around 45% increasing the need for innovative treatments. Checkpoint inhibitors have shown significant clinical efficacy in mismatch repair deficient (MMRd) cancers and could be a powerful treatment in OC. However, their application in OC is limited due to the lack of data on the prevalence of MMRd. The aim of our study was to conduct a systematic review of the literature and meta-analysis to provide an accurate estimate of the prevalence of MMRd in OC. We followed PRISMA guidelines throughout. Studies were identified by electronic searches of Medline, Embase, Cochrane CENTRAL and Web of Science followed by citation searching. Studies not written in English were excluded. All studies were reviewed by at least two independent reviewers. Proportions of test positivity were calculated by random and fixed-effects meta-analysis models. I(2) score was used to assess heterogeneity across studies. In total 54 studies were included with 17 532 analyzed for MMRd. The overall proportions of MMRd by immunohistochemistry and microsatellite instability analysis were 6.7% and 10.4%, respectively. MMRd was reported in all histotypes of epithelial OC but was most common in endometrioid OC. We estimate that on average 46.7% (95% CI: 28.8-65.4) of ovarian carcinomas showing MMRd by IHC had a germline path_MMR variant identified. OC in those with Lynch syndrome seems to present at an earlier age and stage. Studies however were generally of low quality and there was a high degree of heterogeneity. A significant minority (up to 16%) of OC displays MMRd and, therefore, could be amenable to checkpoint inhibition therapy. However, the current literature base is of limited quality and therefore high-quality prospective studies exploring MMRd in OC with the use of multimodal testing are required. In addition, trials researching efficacy of checkpoint inhibition in MMRd OC are needed.