Introduction: Extended lymph node dissection (ELND) compared to limited lymph node dissection (LLND) at time of prostatectomy improves staging and lymph node (LN) yield. The effect on biochemical recurrence (BCR) and survival is less well understood. We sought to evaluate the benefit of robotic ELND and LLND with respect to BCR.
Materials and methods: Between 2008-2012, 584 consecutive men with intermediate or high risk clinically localized adenocarcinoma of the prostate underwent robotic assisted radical prostatectomy (RARP) with concomitant LLND (n = 326) or ELND (n = 258). Survival estimates were made using the Kaplan-Meier method. Log-rank statistic was used for comparison of curves. BCR predictors were determined with multivariable Cox regression analysis. Chi-square and Wilcoxon rank-sum tests were used to compare discrete and continuous variables, respectively, across the two groups.
Results: Median follow up for ELND and LLND patients was 46 and 54 months, respectively. ELND yielded more LNs (20 versus 6, p < 0.0001) and had higher node positivity (15.1% versus 3.4%, p < 0.0001). BCR free survival (BCRFS) at 3 and 5 years for ELND and LLND was 85% and 75% (p = 0.01), and 76% and 67% (p = 0.10), respectively. In subgroup analysis, ELND was associated with higher 5 year BCRFS in node-negative patients (84% versus 68%, p = 0.0005) and in intermediate risk patients (93% versus 80%, p = 0.0002). In multivariable analysis, ELND was a significant predictor of BCRFS in node-negative (HR = 0.50, p = 0.003) and intermediate risk patients (HR = 0.54, p = 0.03).
Conclusions: ELND improves LN yield and detection of positive nodes. BCR analysis suggests a reduced risk of PSA failure for robotic ELND in intermediate risk and node-negative patients.