Comparison of morphometry and DNA flow cytometry with standard prognostic factors in bladder cancer.

In 83 bladder cancer patients with adequate follow-up (mean 13 years, range 9-22) the prognostic value of morphometric, DNA flow cytometric and clinical parameters was assessed. Paraffin embedded material was used in flow cytometry. Univariate life-table analysis showed the statistically significant relation of clinical stage, histological grade, mean nuclear area, the Standard Deviation (SD) of nuclear area, mean maximal nuclear diameter, mean nuclear perimeter and the volume corrected mitotic index (M/V index) to survival when bladder cancer deaths alone were used in the analysis. The recurrence of bladder cancer could be predicted with the M/V index. Survival analysis with Cox's regression model pointed to primary tumour clinical stage as the most important prognostic factor of crude survival. Histologically, grade and the SD of nuclear area were the best prognostic factors. Primary tumour stage and histological grade were the best predictors of death from bladder cancer. In Cox's model, histoquantitative methods are almost as good as clinical staging in predicting prognosis. DNA flow cytometry of paraffin embedded material offered no advantage over clinical stage, histological grade or morphology in assessing prognosis.