Risk stratification of patients with nodal involvement in upper tract urothelial carcinoma: value of lymph-node density

To determine the risk factors associated with clinical outcome in patients with lymph node (LN)-positive urothelial carcinoma of the upper urinary tract (UTUC) treated with radical nephroureterectomy (RNU) and lymphadenectomy, focusing on the concept of LN density (LND). Patients undergoing RNU with regional lymphadenectomy were identified through multi-institutional databases. All pathology slides were re-evaluated by genitourinary pathologists unaware of the clinical data. The exposure variable used was LND (continuously coded and that of all possible thresholds) with recurrence-free and disease-specific survival (DSS) serving as the outcome measures. Of 432 patients undergoing RNU with lymphadenectomy, 135 (31%) had LN metastases. Within a median follow-up of 4.1 years, 90 of the 135 patients with LN metastases (68%) had disease recurrence and 76 (58%) died from UTUC. The mean (sem) 5-year recurrence-free and DSS probabilities were 27 (4)% and 33 (5)%, respectively. The median (range) LND was 50 (3-100)%. The most informative threshold for LND in relation to outcome was 30%. In multivariable analyses that adjusted for the effects of tumour stage and grade, patients with a LND of >= 30% were at greater risk of both cancer recurrence, with 5-year rates of 25 (5)% vs 38 (8)% (hazard ratio 1.8, P = 0.021) and mortality, with 5-year rates of 30 (6)% vs 48 (9)% (1.7, P = 0.032) compared to those with a LND of < 30%. Our results are primarily limited by a lack of standardization in the lymphadenectomy template. We evaluated the concept of LND for the first time in UTUC. LND provides additional prognostic information in patients with node-positive disease after RNU. The use of LND in clinical trials might provide an additional insight into the value of LN dissection in patients undergoing RNU.