THE ROLE OF NUCLEAR MORPHOMETRY FOR PREDICTING DISEASE OUTCOME IN PATIENTS WITH LOCALIZED RENAL-CELL CARCINOMA

Background. More than one-third of patients with localized renal cell carcinoma (RCC) will have disease progression after nephrectomy. Prese nt histopathologic variables cannot accurately predict the outcome of individual patients. Methods. Nuclear morphometry was performed by an image analyzer on histologic sections from 39 specimens of pathologic T1 and T2 classification RCC. All patients underwent radical nephrecto my and were followed for a mean of 7.6 years. A univariate analysis an d then a multivariate stepwise regression method were used to correlat e results with patients' outcome. Results. The best predictors of dise ase free interval were mean nuclear elongation factor (MNEF) (P = 0.02 3), mean nuclear regularity factor (MNRF) (P = 0.034), and mean nuclea r area (MNA) (N = 0.038). Univariate analysis identified a significant correlation between patient survival and MNEF (P = 0.009), MNRF (P = 0.020) and MNA (P = 0.023). Combination of MNEF and MNA was even more strongly associated with survival(P = 0.0013). Multivariate analysis r evealed that MNA (P = 0.044) and MNEF (P = 0.045) correlated independe ntly with survival. Conclusion. These results suggest that nuclear mor phometry provides objective independent prognostic information for pat ients with localized RCC.