Risk Based Surveillance after Surgical Treatment of Renal Cell Carcinoma

PURPOSE: We assessed the accuracy of the UISS (UCLA Integrated Staging System) to predict the postoperative recurrence of renal cell carcinoma. We also evaluated whether including patient age and tumor histology would improve clinical decision making. MATERIALS AND METHODS: We analyzed the records of 1,630 patients treated with nephrectomy at a single academic center. The accuracy of the UISS model to predict early (12 months or less) and late (more than 60 months) recurrence after surgery was compared with a new model including patient age and disease histology. RESULTS: The new model and the UISS model showed high accuracy to predict early recurrence after surgery (AUC 0.84, 95% CI 0.81-0.88 and 0.83, 95% CI 0.80-0.87, respectively). In patients diagnosed with low risk tumor types (eg papillary type 1 and chromophobe lesions) the average risk of early recurrence significantly decreased in each UISS risk category when tumor histology was added to the predictive model (low risk 1.6% vs 0.6%, intermediate risk 5.5% vs 1.9% and high risk 45% vs 22%). Kaplan-Meier analysis showed no difference in the risk of late recurrence among the UISS risk categories. CONCLUSIONS: The UISS model should be applied to tailor the early followup protocol after nephrectomy. Patients with low risk histology deserve less stringent followup regardless of the UISS risk category. Our results do not support a risk stratification model to design a surveillance protocol after 5 years postoperatively.