TY - JOUR T1 - The significance of differences in prognostic value of quantitative pathologic features in FIGO stage III and IV serous adenocarcinoma of the ovary between a group of Danish patients and other groups JF - International Journal of Gynecologic Cancer JO - Int J Gynecol Cancer SP - 355 LP - 360 DO - 10.1046/j.1525-1438.1995.05050355.x VL - 5 IS - 5 AU - M. BRINKHUIS AU - O. MOGENSEN AU - P. BICHEL AU - J. P.A. BAAK Y1 - 1995/09/01 UR - http://ijgc.bmj.com/content/5/5/355.abstract N2 - Previous studies have shown DNA cytometric, morphometric and stereologic features to be strong prognosticators in advanced ovarian cancer. However, in one Danish study a new stereologic parameter, the volume-weighted mean nuclear volume (v̄v), was not of prognostic significance. In that study, sampling was performed in the whole tumor section, contrary to that in other studies where measurements were performed in the worst differentiated area (the measurement area). The aim of the present investigation was, therefore, to confirm or reject the hypothesis that random estimates of nuclear v̄v over the whole tumor, rather than in the measurement area, are the cause for the lack of prognostic value in the Danish patients. Additionally, the mitotic activity index (MAI), volume percentage of epithelium (VPE), mean nuclear area (MNA) and standard deviation of the nuclear area (SDNA), which have proved to be of strong prognostic significance in advanced ovarian cancer, were assessed in the measurement area. The MAI, VPE, MNA, SDNA and nuclear v̄v were well reproducible, but prognostically not significant. This study thus excludes differences in microscopic sampling as an explanation for the contrasting findings in the Danish study and other ones. Possible remaining explanations are: (i) differences in sampling at the macroscopic level; (ii) fixation differences; (iii) patient selection bias; and (iv) geographic differences in the degree of malignancy of advanced ovarian cancer in Denmark and the Netherlands. As to the latter, in comparison with two Dutch studies the Danish ovarian cancers in this series had residual disease <2 cm (33% vs 50%) less often, were less often well differentiated (4% vs 19%), and less often had favorable morphometric characteristics (6% vs 15%). Further studies are required to analyze whether other Danish patient groups with advanced ovarian cancers differ from the Dutch. ER -