Frozen Section Biopsies of Uterine Smooth Muscle Tumors: A Multicenter Clinicopathologic Study of 86 Cases Emphasizing Diagnostic Pitfalls
Philip Ip, Ka-Yu Tse, Mandy Ng, Charlotte Leung, Alice Chan, Polly Tsui, Hextan Ngan, Annie Cheung. University of Hong Kong, Queen Mary Hospital, SAR, Hong Kong; United Christian Hospital, SAR, Hong Kong; Tuen Mun Hospital, SAR, Hong Kong; Kwong Wah Hospital, SAR, Hong Kong; Pamela Youde Nethersole Eastern Hospital, SAR, Hong Kong
Background: Frozen section (FS) biopsies for uterine smooth muscle tumors (U-SMT) are infrequently performed and detailed studying under this scenario is rarely done.
Design: Clinicopathologic features of 86 U-SMTs sent for FS between 1997 and 2012 from 5 hospitals were reviewed.
Results: Patients' age was 31 to 90 years (median, 44.5). Commonest preoperative finding was a mass (n=61, 71%) with 10 suspected ovarian tumors. Commonest reasons for request was extrauterine involvement with or without adhesions to other organs (n=43, 50%) or an atypical intraoperative appearance (n=28, 33%). Lesions were 0.5 to 40 cm (median, 15), mostly subserosal (n=46, 53%) or pedunculated (n=20, 23%). There were 82 leiomyomas including variants; 4 were leiomyosarcomas. In 87% (n=75) the FS concurred with paraffin diagnosis. The false positive and false negative rates were 1.2% (n=1) and 0%, respectively. Deferral rate was 10.5% (n=9). The false-positive case was due to misinterpretation of stromal hyalinization as tumor cell necrosis (TCN). That tumor also had atypia but no mitosis. This was diagnosed as malignant and the patient had hysterectomy and full staging. Another case with stromal hyalinization resulted in compartmentalization of tumor cells was misinterpreted as an epithelial tumor. However, extent of operation was unaltered. Accurate FS diagnosis of malignancy was made in all 4 leiomyosarcomas. Reasons for deferral were hypercellularity, epithelioid differentiation, necrosis of an uncertain type, atypia, myxoid change, unusual growth pattern or combinations thereof. Difficult tissue-cutting in 2 was due to adipose tissue and calcification. The sensitivity and specificity were 100% and 98.6%, respectively. The positive and negative predictive values were 80% and 100%, respectively.
Conclusions: FS of U-SMT is reasonably accurate. Intraoperative appearances are not specific. Alterations of stroma accounted for most cases of errors or deferrals; specifically, stromal hyalinization may potentially mimic TCN and this latter feature is unreliable in FS. Diagnosis of leiomyosarcoma is possible even if the tumor has only diffuse atypia and high mitotic rate. Malignancy should always be diagnosed with caution as it may lead to unnecessary overtreatment.
Category: Gynecologic & Obstetrics
Tuesday, March 5, 2013 1:00 PM
Poster Session IV # 243, Tuesday Afternoon