Prognostic significance of DNA cytometry in comparison with histologic measurements in malignant melanomas.

Abstract

In the present study we evaluated the prognostic impact of both DNA cytometry and quantitative histology in patients with malignant melanoma (MM). In contrast to previous investigations on sections, rapid image analysis (IA) of imprint specimens was performed to measure DNA cytometric features; 34 cases of stage I MM with low (< 1.5 mm, n = 20) and higher tumor thickness (TTH) (> 1.5 mm, n = 14) were analyzed. We found significant correlations between cytometric features and TTH, which is accepted as the most important prognostic criterion in MM. Higher TTH was closely correlated with the rate of markedly aneuploid nuclei, which is indicated by the 5c exceeding rate (5cER; r, 0.89; p < 0.001). The gain of chromatin in thicker tumors was accompanied by higher mean nuclear area (MAREA; r, 0.60; p < 0.001) and mean DNA content (MDNA; r, 0.58; p < 0.001). Additional evidence for prognostic significance of cytometry was obtained by preliminary survival analysis of 16 cases (four patients died within 2 1/2 years and 12 patients survived for at least 4 1/2 years). Applying multivariate stepwise discriminant analysis, a combination of TTH, level of invasion (LEV), 2c deviation index (2cDI), and modified standard deviation of the DNA values (SDNAM) proved to be most effective. One hundred percent of the cases were correctly classified as survivors or nonsurvivors. TTH and the 5cER were found to be the best univariate criteria for prognosis. In the U test according to Mann and Whitney, a significant discrimination of survivors and nonsurvivors was possible using either TTH or the 5cER, respectively (p < 0.02). Thus, we would like to conclude that IA of imprint specimens can be recommended as a rapid and simple additional method for grading.