Abstract
Fuhrman grading of renal cell carcinoma focuses on features of nuclear size, nuclear shape, and nucleolar prominence. Despite the reported widespread usage of Fuhrman grading in clinical studies, there is debate as to the prognostic significance and reproducibility of its criteria. It has been noted that many pathologists rely on assessment of nucleolar prominence alone when grading renal cell carcinoma; however, the validity of this remains unconfirmed. This study was undertaken to determine the relationship of the 3 morphologic components of the Fuhrman grading system with one another and to determine which, if any of these, can be correlated with outcome for clear cell renal cell carcinoma. One hundred twenty-one organ-confined clear cell renal cell carcinomas were examined in this study. Parameters of nuclear size (area, major axis, perimeter) and nuclear shape (shape factor, nuclear compactness) were assessed by image analysis, whereas nucleolar prominence was assigned (grades 1 to 3) using the criteria of Fuhrman. On the basis of the predominant grade present, there were 17 nucleolar grade 1, 90 nucleolar grade 2, and 14 nucleolar grade 3 tumors. When the high-power field in each tumor with the worst nucleolar grade was assessed, there was 1 nucleolar grade 1, 68 nucleolar grade 2, and 52 nucleolar grade 3 tumors. Predominant and worst nucleolar grade correlated with all measures of nuclear size, but not nuclear shape. Worst nucleolar grade and all parameters of nuclear size were significantly associated with outcome. On multivariate analysis, worst nucleolar grade retained a significant association with survival when modeled with nuclear area. Neither worst nucleolar grade nor major nuclear axis/nuclear perimeter was significantly associated with survival when modeled together. In this study, we have shown that worst nucleolar grade and nuclear size are of prognostic significance for clear cell renal cell carcinoma. We have further shown the association of worst nucleolar grade with outcome to be independent of nuclear area, whereas it is a dependent variable when tested against other parameters of nuclear size. These findings indicate that worst nucleolar grading alone is a valid grading parameter for clear cell renal cell carcinoma.