The prognostic value of morphometrical features and cellular DNA content in cis-platin treated late ovarian cancer patients

Abstract

In 73 CAP-1 treated stage III and IV ovarian cancers, the prognostic significance of morphometric features and cellular DNA content has been evaluated in comparison with histologic type, grade of differentiation and a number of clinical characteristics. Borderline tumours were excluded from the study. Median follow-up was 44 months, median survival time 36 months. Single features associated with prognosis were (in order of decreasing significance according to single variate analysis): FIGO stage (P = 0.0002), bulky disease (P = 0.004), standard deviation and mean of nuclear area (P = 0.0006 and P = 0.01), cellular DNA content (P = 0.01), mitotic activity index (P = 0.08) and volume percentage epithelium (P = 0.13, not quite significant). Tumours with a mean nuclear area greater than 70 micron2 (which occurred in 35% of the cases) were nearly all aneuploid. Multivariate analysis showed that the statistically most significant prognostic combination of features consisted of mean nuclear area, presence or absence of bulky disease and FIGO stage (in order of decreasing importance) (Mantel-Cox = 23.07, P less than 0.00001). A low value for the multivariate function of this combination of features was associated with a poor prognosis within 24 months, a high value with a favourable outcome. Another favourable combination of features appeared to be diploid cellular DNA content and a low mitotic activity index (11 patients, one died). However, even with the prognostically most favourable combination of these features, several patients died. Of all combinations of features investigated, only two were associated with an excellent prognosis (low mitotic activity index and low volume percentage epithelium). Cancers of 7 patients (10%) displayed such features, and none of them died during the follow-up period (minimally 20 and maximally 54 months). It is concluded that morphometric and flow cytometric analysis can provide significant and objective information to predict the prognosis of cis-platin-treated advanced ovarian cancer patients.