O. Nativ et al., THE ROLE OF NUCLEAR MORPHOMETRY FOR PREDICTING DISEASE OUTCOME IN PATIENTS WITH LOCALIZED RENAL-CELL CARCINOMA, Cancer, 76(8), 1995, pp. 1440-1444
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.