External validation of the European association of urology recommendations for pelvic lymph node dissection in patients treated with robot-assisted radical prostatectomy.

INTRODUCTION

The European Association of Urology (EAU) guidelines recommend to perform extended pelvic lymph node dissection (ePLND) in all patients with a risk of lymph node invasion (LNI) higher than 5% estimated by the updated Briganti nomogram. However, this model has been developed in patients exclusively treated with open radical prostatectomy. No study has specifically assessed the accuracy of this model among men treated with robot-assisted radical prostatectomy (RARP). We hypothesized that EAU indications for PLND are accurate also among patients treated with RARP.

MATERIALS AND METHODS

We evaluated 615 patients treated with RARP and PLND between 2006 and 2012 at a single tertiary referral center. The predictive accuracy of the nomogram was quantified using the receiver operating characteristic-derived area under the curve, the calibration plot method, and decision curve analyses.

RESULTS

Median of lymph nodes (LNs) removed was 9 (interquartile range: 6-13). The rate of LNI was 5%. External validation of the Briganti nomogram showed good accuracy (81.8%). A nomogram-derived cutoff of 5% would allow the avoidance of 75% of PLND at the cost of missing of 19.4% of patients with LNI. When the same analyses were repeated in men with at least 10 and 15 LNs removed, the 5% cutoff was associated with a reduction in PLND and with an LNI missing rates of 67.6% and 59.3% and 17.4% and 6.2%, respectively. Similarly, the prediction accuracy increased to 81.2% and 85.3%, respectively. The decision curve analysis showed an increase in the net-benefit in the prediction range between 2.5% and 54%.

CONCLUSIONS

We report the first validation of the EAU guideline recommendation for PLND among patients exclusively treated with RARP. We demonstrated that the accuracy of Briganti nomogram is high, but the proposed 5% cutoff is valid only in the presence of adequate ePLND.