Objectives: Most cases of granulosa cell tumors (GCT) of the ovary are characterized by a relatively good outcome. However, some tumors behave aggressively and some tend to recur many years after the initial diagnosis. Tumor growth depends on cell proliferation and angiogenesis. Thus, proliferative indices and microvessel density were studied to determine possible valuable methods to assess the GCT patient's outcome.
Methods and study design: Paraffin-embedded tissue blocks were available for 60 patients with primary GCT and were investigated by immunostaining with monoclonal antibodies against PCNA, Ki-67 and factor VIII-related antigen. The follow-up was available for 51 patients and ranged from 25 to 206 months. A clinical follow-up distribution of patients was made: 8 patients with recurrence (group I); 6 patients who lived with no evidence of recurrence for 100 months or more (group II), and 37 patients alive with no evidence of recurrence in the follow-up period of less than 100 months (group III).
Results: There was a statistical correlation between PCNA and Ki-67 proliferative indices. A significant increase (P <0.05) of mean PCNA and Ki-67 proliferative indices and mean tumor size was seen in patients of Group I compared to those of Group II. The mean PCNA proliferative index positively correlated with the mean Ki-67 proliferative index for Groups I and II. Mean microvessel density showed a positive correlation with mean PCNA and Ki-67 proliferative indices and with mean tumor size for Group I, whereas it was negatively correlated with PCNA proliferative index and tumor size for Group II. A positive correlation was found between mean mitotic count and both proliferative indices only for Group II. The following features were indicative of a relatively poor prognosis: GCT measuring >9 cm in diameter, PCNA >4.0%, Ki-67 >1.2%, and diffuse, insular and sarcomatoid histologic patterns.
Conclusions: The findings support the importance of proliferative factors, tumor size and histologic patterns as possible prognostic indicators for estimating the biologic behavior of patients with GCT. Unfortunately, angiogenesis did not seem to be a useful determinant parameter of a possible aggressive behavior. However, a longer follow-up period with larger series may be required to assess the value of the parameters in prediction of patient survival.