Clinicopathologic analysis of 187 high-grade endometrial carcinomas of different histologic subtypes: similar outcomes belie distinctive biologic differences.

The clinical and histopathologic features of 187 high-grade endometrial cancers [FIGO grade 3 endometrioid (EC-3), serous (SC), and clear cell (CC)] were studied to determine whether clinicopathologic differences between these various histologic subtypes existed. The study group consisted of 89 EC-3s, 61 SCs, and 37 CCs. Treatment regimens were individualized. SCs and CCs were significantly more likely than EC-3s to occur in patients older than 65 years (P=0.03), and SCs tended to occur more frequently in patients of African descent than EC-3s and CCs (P=0.07), although this was not statistically significant. EC-3s had the highest rate of associated endometrial hyperplasia (P=0.05). SCs were most likely to have high-stage disease at presentation (>or=stage IIB; P=0.01), with peritoneal dissemination at diagnosis being much more common compared with EC-3s and CCs (P=0.004). Median follow-up was 39 months, and median overall survival was 47 months. Five-year survivals were 45% (EC-3), 36% (SC), and 50% (CC)-differences that were not statistically significant. In contrast, the impact of stage on survival was significant (P<0.001). Among all other factors evaluated, only age greater than 65 years was a negative predictor (risk ratio, 2.23; P<0.001), whereas a family history of cancer reduced the risk of death when controlling for stage (risk ratio, 0.54; P=0.005). When controlling for stage, race, reproductive history, personal history of cancer, histologic subtype, depth of myometrial invasion, lymphovascular invasion, presence of an endometrial polyp, presence of hyperplasia, or staging adequacy did not affect prognosis. High-grade endometrial cancers of different histologic subtypes treated in an individualized manner are associated with similar clinical outcomes, but differences in age at presentation, race distribution, association with hyperplasia, stage, and sites of tumor dissemination support the idea that these represent distinct disease entities as defined by traditional histopathologic classification of endometrial cancers.