Endometriosis does not confer improved prognosis in ovarian carcinoma of uniform cell type.

The role of endometriosis in ovarian cancer, disease progression, and survival is a subject of active investigation. A series of 144 ovarian cancers with clear cell or endometrioid histology or associated endometriosis, all classified on the basis of strict histologic criteria, was evaluated to further explore the relationship between endometriosis-associated ovarian cancer and age at presentation, FIGO stage, histology, presence of synchronous primary disease elsewhere in the mullerian tract, and survival. Patients with endometrioid carcinomas were significantly younger (mean, 52 y) in comparison with patients with either clear cell carcinoma (mean, 55 y) or mixed tumors (mean, 59 y; P=0.002). Clear cell carcinoma presented as low-stage disease (FIGO I) in 33% of cases compared with endometrioid carcinomas in 97% of cases and mixed carcinomas in 27% of cases. Endometriosis was associated with 53% of clear cell carcinomas, 33% of endometrioid carcinomas, and 45% of mixed tumors (P<0.001). Synchronous primary tumors, observed in 31% of endometrioid tumors, 5% of mixed tumors, and in 2% of clear cell tumors (P<0.001), were unlikely to be associated with endometriosis (P=0.04). Univariate analysis of the aggregate cohort demonstrated that the single best overall predictor of disease-free survival was FIGO stage at presentation (P<0.001), followed by histologic subtype (P=0.003). Endometriosis did not have a significant relationship with disease-free survival (P=0.7). We conclude that the link between endometriosis and ovarian cancer is much stronger for clear cell carcinoma than for other histologic subtypes (P<0.001). Furthermore, when uniform histologic criteria are applied, true mixed endometrioid and clear cell carcinomas are uncommon; most endometriosis-associated mixed tumors are heterogenous mixtures of endometrioid, mucinous, and serous histology with areas of clear cell cytoplasm. Endometriosis per se does not appear to predict prognosis in clear cell and endometrioid tumors, with the possible exception of tumors with mixed histology. Until more data are collected, pathologists should classify ovarian tumors with mixed histology as a separate and potentially unique biological and prognostic group.