The multitude of grading systems proposed for prostate adenocarcinoma illustrate the putative value of tumor grading as a predictor of prognosis. However, it also indicates the uncertainty and subjectivity of visual grading, as is confirmed by the results of several studies. Recent computer techniques enable fast image processing of microscoping images. Hence, a quantitative assessment of cellular and nuclear features can be obtained, eliminating subjectivity of visual grading. DNA-analysis and morphometric karyometry can be valuable for predicting prognosis in prostate tumors. For routine use, however, standardized preparatory techniques are mandatory. Up till now image analysis of microscopic images is a useful tool in addition to visual grading. Further standardization, inter-as well as intra-laboratory, is necessary to elaborate its value as a reproducible tool in tumor grading.Ever since Broders (1925) [15] introduced the grading of tumors on base of cytological and histological features, practically all tumors have been submitted to different grading systems to be used as predictors of tumor behavior. Many of the proposed grading systems, however, did not offer the desired accuracy in predicting malignant tumor behavior. Moreover, reproducibility of the grading systems in general is rather poor, since the criteria are often loosely defined and can be interpreted in many ways. On the other hand, benign or less malignant lesions can contain small malignant parts that can be overlooked. This heterogeneity is present in many tumors. Renal cell carcinoma is especially known for its multiple cell populations, and for this reason grading systems for renal cell carcinoma are also of doubtful value.Transitional cell carcinomas of the urinary tract are described by only a limited number of grading systems that are very similar in their description of malignancy. Prostate adenocarcinoma is the tumor of the urogenital tract for which a far greater number of grading systems has been proposed than for other urological tumors. The variety and large number of grading systems for prostate cancer caused us to select this type of tumor for further discussion. First, the five most widely used visual grading systems are reviewed. Second, the grading systems using image analysis parameters are discussed.Since no single tumor marker has so far allowed optimal prediction of tumor behavior, the fact that both visual grading and quantitative pathology have mutual interests means that they need to be applied in concert. Image analytical findings should merely be used in addition to pathological grading rather than as a completely different entity. When staining and computer techniques progress, the use of quantitative techniques will become more and more integrated into routine clinical pathology.
Grading of prostate adenocarcinomaThe first grading system for prostate adenocarcinoma was introduced by Broders in 1925 [15] and was based on the estimated percentage of glandular differentiation. Many grading systems did not show signific...