Computerized nuclear morphometry of renal cell carcinomas.

Abstract

The aim of this study was the evaluation of the usefulness of nuclear morphometry in the pathology of renal cell carcinoma by computer-aided image analysis and the statistical comparison of nuclear morphometric parameters with the histologic grade of the tumors. 44 cases of renal cell carcinoma, classified by two independent pathologists into 4 histologic grades (I-IV), were examined. The following 5 nuclear morphometric parameters were measured in a large number of randomly selected nuclei of each case: Major axis length, area, elongation, roundness and compactness. The statistical evaluation was performed using one-way analysis of variance between the four groups of the histologic grades (I-IV). Between them, there was a statistically very significant difference of the mean value of all the evaluated parameters. The values of the estimated parameters, with the exception of roundness and compactness, showed a strong tendency to increase in proportion to histologic grade. Our results suggest that image analysis is a reproducible and objective method for the grading of renal cell carcinoma, and it can be helpful in the unbiased evaluation of such tumors.