Clinicopathologic factors that influence the use and accuracy of frozen section diagnosis for mucinous ovarian tumors

Objectives: To determine clinicopathologic factors that influence the use and accuracy of frozen section diagnosis (FS) for mucinous ovarian tumors, including all types of tumors (benign, borderline, and malignant). Methods: A retrospective cohort analysis was conducted for 1,032 patients (662 benign, 272 borderline, and 98 malignant) from January 1997 to December 2010. Univariate and multivariate regression analyses were used to assess the influence of clinicopathologic factors on the possibility of misdiagnosis, especially underdiagnosis. Results: We found that 1.0% (6/662) of benign tumors were overdiagnosed; 21.3% (58/272) and 4.0% (11/272) of borderline tumors were underdiagnosed and overdiagnosed, respectively, whereas 27.6% (27/98) of malignant tumors were underdiagnosed. In addition, 8.2% (85/1,032) of the total cases were underdiagnosed, and 1.6% (17/1032) of the total cases were overdiagnosed with FS. In univariate and multivariate analyses, the associated tumor and the predominant component (predominantly liquid tumors) were significant predictors for the underdiagnosis of mucinous borderline ovarian tumor (MBOT) with FS. In univariate and multivariate analyses, the mode of surgery (laparoscopic operation) and the predominant component (predominantly liquid tumors) were significant predictors for the underdiagnosis of mucinous ovarian carcinoma with FS. Conclusions: FS was not as accurate as expected in the diagnosis of MBOT and mucinous ovarian carcinoma. Misdiagnosis, especially underdiagnosis, affects the surgical strategy for women with mucinous ovarian tumors because an underdiagnosis leads to inadequate conservative surgery and is a primary cause for problems with restaging; therefore, particular attention is required to minimize the risk of underdiagnosis, especially in mucinous ovarian tumors.