Quantitative pathologic features as predictors of long-term survival in patients with advanced ovarian cancer treated with cisplatin.

Abstract

The prognostic value of morphometric and DNA flow cytometric features were studied and compared with FIGO stage, preoperative tumor load, residual disease status, Karnofsky index and classic pathologic features such as Broders' grade and histologic type in 58 FIGO stage III and IV adequately debulked ovarian patients with long-term follow-up. The mitotic activity index, volume percentage of epithelium, and mean and SD of nuclear area were assessed by interactive morphometry, and tumor material was routinely processed for DNA flow cytometric assessment of DNA ploidy and S-phase fraction. Survival analysis (Kaplan-Meier curves, Mantel-Cox test), revealed FIGO stage (P = 0.013) and the mean and SD of nuclear area to be significant prognosticators (P = 0.027 and P = 0.012, respectively). In multivariate survival analysis (Cox model), a multivariate combination of FIGO stage, preoperative tumor load and mean nuclear area was the best prognostic combination of features (P = 0.0034). These results confirm the findings of previous studies. We conclude that, in accord with previous studies, morphometric features are good predictors of survival after cisplatin treatment in advanced ovarian cancer, especially in combination with FIGO stage and preoperative tumor load.