Papillary (chromophil) renal cell carcinoma: histomorphologic characteristics and evaluation of conventional pathologic prognostic parameters in 62 cases.

For more than two decades, papillary renal cell carcinoma has been recognized as a possible distinct clinicopathologic subtype of renal cell carcinoma (RCC). However, the histologic criteria for its diagnosis and the clinical outcome are still debated. In an attempt to clarify the diagnostic criteria and resolve issues pertaining to biologic potential, we have evaluated the histologic spectrum of 62 papillary RCCs and assessed significance of conventional pathologic prognostic parameters (Fuhrrman's nuclear grade [NG], pathologic stage [Robson and TNM], tumor size, multifocality, necrosis, and foam cells) and correlated these with outcome. The mean age of patients was 61.8 years (range 22-83), and males were more commonly affected (1.8:1). Grossly, most tumors were well circumscribed, averaged 6.7 cm in size (range 1.8-18), and were predominantly localized to the renal poles (polar vs. mid-renal, 3:1). Multifocality was a prominent feature (24 cases), and in three cases tumors were bilateral. Microscopically, papillary RCCs were predominantly papillary or tubulopapillary, often with a thick fibrous capsule, foam cells, necrosis, hemorrhage, and multifocality. Thirty-five percent of these tumors were low grade (NG I and II) and 65% high grade (NG III and IV). Sixteen of these tumors presented in a higher stage (stages III and IV), and the overall stage correlated with NG (chi 2, p = 0.009). Tumors were further distinguished by cytoplasmic features: eosinophilic (42%), basophilic (34%) and mixed (24%). Eosinophilic tumors were predominantly high grade, and basophilic tumors low grade (chi 2, p = 0.000). A mean follow-up of 57 months showed progression (metastasis, recurrence, or death due to disease) in 21%, whereas 63% were free of disease. Eleven percent died of unrelated causes, and 5% were lost to follow-up. Kaplan-Meier survival analysis showed that both high NG and stage were strongly associated with decreased survival (p = 0.0000 each), as were decreased foam cell (p = 0.0025) and vascular invasion (p = 0.0002). Comparison of 196 reported cases of papillary RCC, including the current series, with reported large series of conventional RCC indicates that papillary RCC usually presents at an early stage, and stage I (Robson) papillary RCC has better 5 year survival rates (87%-100%) than does RCC of the same stage (65-75%). The overall 5 years survival rate for papillary carcinoma (82-90%) was also higher than that of conventional RCC (44-54%). In a Cox proportional hazard regression model, TNM stage appeared to be the only significant variable (p = 0.0000, hazard ratio 10.1) in predicting outcome among papillary RCC. Based on this experience, we conclude that (a) papillary RCC is a malignant tumor, with a tendency to present at a lower stage, but with a distinct potential for progression and aggressive behavior; (b) stratification of these tumors according to cell type, amount of foam cells, presence or absence of vascular invasion, nuclear grade, and pathologic stage provides useful prognostic information; (c) the better 5-year survival rate of papillary RCC (overall and for stage I tumors) compared with that of conventional RCC suggests that it is a tumor with lower malignant potential. Thus, histologic subcategorization of a renal carcinoma as papillary RCC appears to have prognostic implications.