The authors attempted to determine the potential prognostic value of several ultrastructural morphometric parameters, including nuclear, nucleolar, and cytoplasmic features, that could be used in the objective and reproducible histological grading of clear cell renal cell carcinoma. Several nuclear and cytoplasmic parameters were assessed by ultrastructural morphometry in 26 consecutive cases of clear cell renal cell carcinoma. The nuclear and nucleolar sizes, the number of nucleoli per nuclear section and the number of marginated nucleoli, Fuhrman's nuclear grade, and Robson's stage were recorded. In addition, the proportion of cytoplasmic components was semiquantitatively estimated and compared to light microscopic appearance. Follow-up ranged from 5 to 1 5 years (mean = 10 years). Statistical evaluations were performed by means of the Pearson or Spearman correlation coefficient tests, and differences in survival were estimated, using the Mantel - Cox proportional risk method. Differences in survival among patients with a mean nuclear area over and under 160 μm2, and among those with a mean nucleolar area over and under 10 μm2, were statistically significant. (Cutoff points were selected at the median value for both parameters; Mantel - Cox test: X2 = 7.102, p < .01; and X2 = 11.096, p < .001, respectively). Fuhrman's nuclear grade (p < .01) and tumor stage at diagnosis (p < .001) were also related to survival. These data suggest that, out of all the ultrastructural morphometric features, nucleolar area is the most useful in the reproducible and accurate grading of clear cell renal cell carcinoma.
- Electron microscopy
- Renal cell carcinoma