Accuracy of frozen section for the operative management of endometrial cancer.

OBJECTIVE: To assess the accuracy of intra-operative frozen section reports at identifying the features of high risk uterine disease compared with final histopathology.
DESIGN: Retrospective study.
METHODS: The records of 460 patients with uterine cancer registered with the Queensland Centre for Gynaecological Cancer between January 1, 1996 and December 31, 1998 were reviewed. Intra-operative frozen section was undertaken in 260 patients with endometrial adenocarcinoma. Frozen section pathology was compared with the final histopathology reports. Inter-observer reliability was assessed using percentage agreement and kappa statistics. Clinical notes were also reviewed to determine if errors resulted in sub-optimal patient care.
RESULTS: Respectively, tumour grade and depth of myometrial invasion were accurately reported in 88.6% of cases (expected 61.5%, Kappa 0.70) and 94.7% (expected 53.8%, Kappa 0.89). Errors were predominantly attributable to difficulties with respect to the interpretation of tumour grade. The error resulted in the patient receiving sub-optimal surgical management in only 11 cases (5.3%)
CONCLUSION: Frozen section is accurate at identifying the features of high risk uterine disease in the setting of endometrial cancer and can play an important role in directing primary operative management.