Abstract
PURPOSE: To evaluate concordance (C) between pre-operative hysteroscopic-directed sampling and final pathology in uterine cancers. METHODS: A retrospective cross-sectional evaluation of prospectively collected data of women who underwent hysterectomy for uterine malignancies and a previous hysteroscopic-guided biopsy was performed. Diagnostic concordance between pre-operative (hysteroscopic biopsy) and postoperative (uterine specimen) histology was evaluated. In endometrioid-endometrial cancers cases Kappa (k) statistics was applied to evaluate agreement for grading (G) between the preoperative and final pathology. RESULTS: A total 101 hysterectomies for uterine malignancies were evaluated. There were 23 non-endometrioid cancers: 7 serous (C:5/7, 71.4%); 10 carcinosarcomas (C:7/10, 70%, remaining 3 cases only epithelial component diagnosed); 3 clear cell (C:3/3, 100%); 3 sarcomas (C:3/3, 100%). In 78 cases an endometrioid endometrial cancer was found. In 63 cases there was a histological C (63/78, 80.8%) between hysteroscopic-guided biopsy and final pathology, while in 15 cases (19.2%) only hyperplasia (with/without atypia) was found preoperatively. Overall accuracy to detect endometrial cancer was 80.2%. In 50 out of 63 endometrial cancers (79.4%) grading was concordant. The overall level of agreement between preoperative and postoperative grading was "substantial" according to Kappa (k) statistics (k 0.64; 95% CI: 0.449-0.83; p < 0.001), as well as for G1 (0.679; 95% CI: 0.432-0.926; p < 0.001) and G3 (0.774; 94% CI: 0.534-1; p < 0.001), while for G2 (0.531; 95% CI: 0.286-0.777; p < 0.001) it was moderate. CONCLUSIONS: In our series we found an 80% C between pre-operative hysteroscopic-guided biopsy and final pathology, in uterine malignancies. Moreover, hysteroscopic biopsy accurately predicted endometrial cancer in 80% of cases and "substantially" predicted histological grading. Hysteroscopic-guided uterine sampling could be a useful tool to tailor treatment in patients with uterine malignancies.