THE ACCURACY OF A FROZEN-SECTION DIAGNOSIS OF BORDERLINE OVARIAN MALIGNANCY

While the accuracy of frozen section in the diagnosis of invasive ovar ian neoplasms has been previously addressed, the frozen section diagno sis of borderline ovarian tumors has been less well characterized, The distinction between benign and borderline lesions is critical to the proper operative management of these patients. The records of 48 patie nts that had a frozen section diagnosis of borderline ovarian malignan cy during surgical exploration between 1986 and 1993 were reviewed. Th irty-one patients were Stage I, 2 patients were Stage II, 10 patients were Stage III, and 5 patients were unstaged. Clarifying phrases were used frequently in the frozen section report, and these terms were cat egorized as ''rule out'' borderline tumor, borderline tumor, and ''at least'' borderline tumor. Of 33 cases with a frozen section report of borderline or at least borderline malignancy, no cases were subsequent ly found to be benign. Errors in the intraoperative management could h ave occurred in 3 of 48 cases (6.25%), when benign neoplasms were thou ght to be of borderline malignancy on frozen section, However, in each of these cases, the clarifying term rule out was used, indicating the equivocal nature of the frozen section findings, Thirteen of the 48 p atients (27.1%) were found to have a focus of invasive cancer within a borderline tumor on final pathologic review; the primary tumors in th ese 13 cases ranged in maximal diameter from 5 to 26 cm and were of va rying histologic types. The level of experience of the pathologist res ponsible for the frozen section did not influence the accuracy of froz en section determination noted in this study. A frozen section evaluat ion identifying a borderline ovarian malignancy is accurate in excludi ng the presence of benign pathology. It is crucial to understand the m eaning of specific clarifying terms used in frozen section diagnoses. (C) 1995 Academic Press, Inc.