Evaluating the diagnostic performance of preoperative endometrial biopsies in patients diagnosed with high grade endometrial cancer: A study of the Society of Gynecologic Oncology (GOC) Community of Practice (CoP)

Abstract

BackgroundHigh grade cancers account for a disproportionate number of recurrences in patients with endometrial cancer. Accurately identifying these cases on endometrial biopsies allows for better surgical planning. This study evaluates the diagnostic accuracy of general pathologists (GP) compared to gynecological pathologists (GYNP) in interpreting preoperative biopsies. MethodsA retrospective cohort study was conducted of patients diagnosed with high grade endometrial cancer (HGEC) between 2012 and 2016 at eight Canadian cancer centres. Data was collected from medical records. Pre-operative biopsies were categorized into groups; biopsies read by GP, GYNP and GP reviewed by GYNP. Rates of HGEC on pre-operative biopsy were calculated. Fisher exact test was used to compare differences between the groups. Univariate logistic regression analysis was conducted for HGEC prediction. ResultsOf 1237 patients diagnosed with HGEC, 245 (19.8%) did not have a preoperative diagnosis of high-grade disease. Discordancy was identified in 91/287 (31.71%) of biopsies reported by GP, and in 114/910 (12.53%) of biopsies reported by a GYNP (p < 0.0001). Compared to GP, GYNP were 3.24 (CI 2.36–4.45) times more likely to identify high grade disease on preoperative biopsy. Patients whose biopsy was reported by a GYNP were more likely to have a comprehensive staging procedure (OR 1.77 CI 1.33–2.38) and less likely to receive adjuvant therapy (OR 0.71 CI 0.52–0.96). ConclusionGYNP are more likely to identify HGEC on pre-operative biopsies. Due to high rates of overall discordancy, it is possible that surgical staging procedures should not be based solely on preoperative biopsy. Further strategies to improve pre-operative biopsies' accuracy are needed.