Risk-adapted scoring model to identify candidates benefiting from adjuvant chemotherapy after radical nephroureterectomy in nonmetastatic upper tract urothelial carcinoma with ≤pT2.

Introduction: After radical nephroureterectomy (RNU), adjuvant chemotherapy (AC) is recommended in either muscle invasive or lymph node positive upper urinary tract urothelial carcinoma (UTUC). However, optimal patient selection remains to be studied. We propose a risk-adapted scoring model for selecting patients for AC in localised UTUC with ≤pT2. Material and methods: The model was based on 7 risk factors modified from the risk stratification system in the European Association of Urology guideline for localised UTUC. Each risk factor indicated one point; total scores were used to categorise patients as at low or high risk for disease recurrence. We applied our model to 135 patients with localised UTUC with ≤pT2, who underwent RNU without AC. Recurrence-free survival and cancer-specific survival were analysed based on risk group. Results: A risk score of ≥4 points indicated high risk (33/135 patients [24.4%]). The accuracy of predicting recurrence was 82.9% (95% confidence interval [CI]: 75.5-88.9%) with a negative predictive value of 93.1% (95% CI: 87.9-96.2%). Disease recurred in 51.5% of high-risk patients and 6.9% of low-risk patients. Multivariate analysis indicated that high-risk was independently associated with recurrence and cancer-specific death (hazard ratio [HR] = 10.20, 95% CI: 3.94-26.44%, HR = 8.72, 95% CI: 2.47-30.73%, all p <0.001, respectively). Conclusions: The risk-adapted scoring model might be an effective way for selecting patients who may benefit from AC after RNU in nonmetastatic UTUC with ≤pT2. These results should be validated in a larger, prospective study.