Prognostic value of karyometric and clinical characteristics in renal cell carcinoma. Quantitative assessment of tumor heterogeneity.

Abstract

The variation in tumor cell differentiation within one renal cell carcinoma, also termed tumor heterogeneity, renders visual tumor grading of these carcinomas difficult. Karyometric analysis enables description of nuclear characteristics of multiple tumor areas. Hence, karyometric analysis can be used to quantify tumor heterogeneity and thus may aid in a more objective grading of renal cell carcinoma. In 121 patients with renal cell carcinoma (tumors in International Union Against Cancer [UICC] stages I [5 cases], II [23 cases], III [33 cases], and IV [60 cases]), clinical and karyometric features were studied to obtain routinely applicable prognostic factors. Several parts of the tumor were analyzed to obtain a measure of tumor heterogeneity. Univariate and multivariate Cox regression analyses were used to determine the predictive value of karyometric features independent of tumor stage and other clinical characteristics. The Cox univariate regression analysis showed correlation of several clinical and karyometric characteristics with survival. Of the clinical characteristics, TNM stage, tumor size, weight reduction, and performance status were significantly associated with survival. The karyometric features, especially those measurements associated with tumor heterogeneity (e.g. differences in nuclear size or chromatin texture between tumor subpopulations) were of value in predicting prognosis. In the Cox multivariate regression analysis, the Robson and UICC stages proved to be the most powerful predictors of survival (P < 0.0001). Of the clinical features, weight reduction and performance score were the only characteristics offering additional information regarding tumor stage (P < 0.0001). From the karyometric analysis quantification of anisokaryosis in the tumor at time of diagnosis offered additional prognostic information. Moreover, the differences of karyometric features within the tumor presumably associated with tumor heterogeneity correlated with survival. Using the features from the multivariate analysis, prognostic groups could be defined. We conclude that karyometric analysis offers a useful means for quantifying tumor heterogeneity. Multivariate Cox analysis revealed additional value of a grading system based on karyometric analysis to tumor stage. Karyometric analysis can be a useful tool for stratification of patient populations.