Diagnostic Discordance in Intraoperative Frozen Section Diagnosis of Ovarian Tumors: A Literature Review and Analysis of 871 Cases Treated at a Japanese Cancer Center

Background This study examined the accuracy and pitfalls associated with frozen section diagnosis of primary ovarian tumors and ovarian metastases based on the 2014 World Health Organization classification (WHO) criteria and proposed improvements from a pathologist's perspective. Methods We microscopically reviewed 871 cases of primary ovarian tumor (N = 802) and ovarian metastasis (N = 69) and compared the results of frozen sections with the final diagnosis. Malignant potential concordance (benign, borderline, or malignant) and specific discordant diagnosis rates were analyzed. Finally, we conducted a unique literature review of specific diagnostic errors in the frozen section diagnosis of primary ovarian tumors. Results Of 802 primary ovarian tumors, 50 (6.2%) cases showed discordant diagnoses in which mucinous carcinoma (40.5%), low-grade serous carcinoma (LGSC; 31.3%), and mucinous borderline tumor (18.4%) were frequently misinterpreted. Of 69 ovarian metastases, all 4 cases of low-grade appendiceal mucinous neoplasm (LAMN) were misdiagnosed as primary ovarian mucinous tumor. A literature review revealed that mucinous/serous borderline tumor or carcinoma accounted for approximately 70% of 217 reported discordant diagnoses. Conclusion In the present study, the concordance rate of malignant potential of the tumor was comparable to that previously reported. Even in the 2014 WHO classification, primary ovarian mucinous borderline tumor/carcinoma and LGSC still comprised the majority of discordant cases. Grossing methods that reduce sampling error are required. LAMN was frequently misinterpreted as a benign or borderline ovarian mucinous tumor. To prevent this error, a differential algorithm integrating clinical information and gross findings should be developed.