Does any correlation exist between the Gleason classification system and the computer-assisted microscope analysis of Feulgen-stained nuclei features in human prostate adenocarcinoma?

Abstract

Grading prostatic adenocarcinomas remains an important problem. Various systems exist (including those proposed by Gleason) but none of these systems seems able to reliably predict either the lethal potential of a tumor in an individual patient or the responsiveness of an individual tumor to various forms of therapy. The most frequently used grading system, as proposed by Gleason, is essentially based on the description of tumor growth pattern. The aim of the present work is therefore to investigate whether the quantitative description of morphonuclear features (including cell anaplasia) and the DNA ploidy level can contribute significant information to the Gleason grading, thus partly at least reducing its subjective nature. This quantitative description was carried out by means of the Feulgen-stained nuclei image cytometry computation of 24 variables in 101 prostatic adenocarcinomas. The results show that both DNA ploidy- and morphonuclear-related variables were weak discriminators for the various grades of the Gleason classification system, and particularly between the high (Gleason 4 and 5) and the other Gleason-grades, i.e. the low (Gleason 1 and 2) and intermediate (Gleason 3) ones. The morphonuclear evidence of anaplasia is thus not redundant data on tumor growth pattern and may be expected to provide additional diagnostic information.