Histological Grading of Ovarian Clear Cell Adenocarcinoma

Abstract

In this study, we aimed to develop a histological grading system for ovarian clear cell adenocarcinoma (CCA), based on the tumor growth architectures. Cases were defined as Group A if ≥90% of a tumor examined was composed of well-differentiated tubulocystic and/or papillary architectures; Group C if at least 10% of the tumor was composed of very poorly differentiated histology (i.e. solid masses or individual infiltrating tumor cells with no or little glandular/papillary differentiation); and tumors not corresponding to the first 2 descriptions were defined as Group B. The interobserver reproducibility and prognostic value of the assigned groups were analyzed for 159 CCAs from 5 institutions. The level of agreement in assigning the groups between 2 pathologists was 88.7% (=0.82). After consensus was reached, 46 (29%), 79 (50%), and 34 (21%) tumors were classified in Groups A, B, and C, respectively. In early-stage cases [International Federation of Gynecology and Obstetrics (FIGO) stage I-II], Group A tumors had significantly better outcomes (100% 5-yr survival) than Group B tumors (82% 5-yr survival, P=0.024 by log-rank test) or Group C tumors (56% 5-yr survival, P=0.00054 by log-rank test). Moreover, early-stage Group B tumors had significantly better outcomes than Group C tumors (P<0.001 by a generalized Wilcoxon test). In advanced cases (FIGO stage III-IV), Group A tumors had significantly better outcomes than Group C tumors (52% vs. 16% 5-yr survival, respectively, P=0.043). Group A and C tumors defined with our system were identified to have favorable and unfavorable prognostic factors, respectively, independent of the clinical stage of the disease and presence of residual tumors after the initial surgery. The proposed grouping system could divide patients with CCA into 3 subgroups with distinct prognostic indications, providing a 3-tier histological grading system for ovarian CCA.