Discordance Between Intraoperative Frozen and Final Pathology of Borderline Ovarian Tumors [13P]

INTRODUCTION: When faced with a concerning adnexal mass intraoperatively, frozen pathology can guide intraoperative management. Borderline tumors are challenging to diagnose with frozen section compared to other ovarian pathology. Little is known about the effects of intraoperative diagnosis of ovarian pathology on surgical decision-making. We aimed to determine the rate of discordance between intraoperative diagnosis of a borderline tumor and final pathology and the effect on surgical management. METHODS: All gynecologic surgical cases wherein frozen ovarian pathology was collected between 2004 and 2014 were reviewed. Cases with frozen pathology consistent with a borderline tumor were further evaluated. For cases of borderline diagnosis by frozen pathology, the false positive and negative rates, sensitivity, specificity, and positive and negative predictive values of the intraoperative diagnosis for ovarian frozen sections were determined. Impact on surgical management was investigated. RESULTS: 1139 cases were sent for frozen ovarian pathology; 50 were concerning for a borderline tumor by frozen pathology. Two cases diagnosed as borderline tumors by frozen pathology returned benign on final pathology with unneeded unilateral salpingo-oophorectomies. The false positive rate between frozen and final pathology for borderline tumors was 4% (2/50). The false negative rate was 12.3% (8/65). The sensitivity, specificity, positive and negative predictive value of frozen ovarian pathology for borderline tumors was 77.1%, 92.2%, 93.0% and 98.9%, respectively. CONCLUSION: The false positive rate of intraoperative frozen and final pathology for borderline tumors is 4% in this study. Further studies are needed to evaluate the impact of intraoperative diagnosis on surgical management and reproductive health.