Comparison of nomograms predicting lymph node invasion in patients undergoing radical prostatectomy for prostate cancer.

Abstract

The aim of this study was to compare the performance of preoperative risk nomograms or detecting lymph node invasion in a cohort of men undergoing radical prostatectomy (RP). A retrospective analysis was performed on all men (n=145) who underwent RP between 2012 and 2015. Preoperative data was inputted to the Memorial Sloan-Kettering Cancer Centre (MSKCC), Partin 2011 and Briganti nomograms and the University of California San Francisco- Centre of the Prostate Risk Assessment tool (UCSF-CAPRA). The risk of lymph node involvement (LNI) was calculated and compared to final histology. One hundred three (71%) men underwent a lymph node dissection at RP. Ten (9.7%) demonstrated LNI. The median nodal yield was 15 nodes, with no difference between those with LNI and those without (19.5 vs 14.5, p=0.22).No patient classified as low risk on the UCSF-CAPRA score had evidence of LNI. In patients with LNI, no patient breached the 2% threshold for lymph node dissection (LND) on the MSKCC nomogram; four patients breached the 5% threshold on the Partin tables while three patients breached the 2.5% threshold for the Briganti nomogram. Nomograms produce useful information regarding risk of disease; however, they often have not been validated on different populations. Risk predictions need to be considered carefully and treatment decisions were made on a patient specific basis.