Purpose: We determined which clinical and molecular variables can predict cancer recurrence in patients following surgical management for localized renal cell carcinoma (RCC).
Materials and methods: From a custom kidney cancer tissue microarray containing tumors specimens from 350 patients 193 undergoing nephrectomy for localized RCC at our institution between 1989 and 2000 were identified. The array was then analyzed by immunohistochemistry for certain molecular markers, namely CA9, CA12, Ki67, gelsolin, p53, EpCAM, pTEN and vimentin. The medical records of these patients were then reviewed for age, sex, TNM stage, tumor size, nuclear grade, Eastern Cooperative Oncology Group (ECOG) performance status, recurrence status and, when applicable, time to recurrence. Cox regression analyses were done to determine clinical and molecular predictors of time to tumor recurrence.
Results: Of the patients 15% demonstrated evidence of tumor recurrence following nephrectomy (29 of 193). Univariate Cox regression demonstrated that tumor size, T stage, grade, ECOG performance status, Ki67, EpCAM and p53 were significantly associated with recurrence (p <0.05). A multivariate Cox regression model showed that T stage (p = 0.018), ECOG (p = 0.004) and p53 (p = 0.003) were the 3 most significant predictors. p53 expression correlated significantly with nuclear grade (Pearson correlation 0.22, p = 0.023) but not with any other clinical factors. Patients with localized tumors demonstrating mean p53 staining values above and below 20% of cells had recurrence rates of 37.7% and 14.4%, respectively (RR = 3.28, p = 0.018).
Conclusions: p53 is a significant molecular predictor of tumor recurrence, as identified in patients undergoing treatment for localized RCC.