PD61-01 THE USE OF 3D VIRTUAL MODELS DURING ROBOT ASSISTED RADICAL PROSTATECTOMY ENSURES A REDUCTION OF POSITIVE SURGICAL MARGINS RATE

Abstract

INTRODUCTION AND OBJECTIVE: Positive surgical margins (PSM) after radical prostatectomy represent a worrying circumstance for both surgeons and patients, leading to a higher rate of biochemical recurrence (BCR). The aim of this study was to investigate the role of 3D model during robot assisted radical prostatectomy (RARP) in reducing the PSM rate. METHOD(S): We prospectively enrolled patients with localised prostate cancer (cT1e3, cN0, cM0) undergoing preoperative Multiarametric Magnetic Resonance Imaging (mpMRI) and subsequent RARP at our Centre from January 2016 to January 2020. Thanks to the support of bioengineers, starting from mpMRI bidimensional images, hyper accuracy 3D (HA3DTM) prostate models were realized. The first surgeon had the opportunity to use the models during the intervention in a cognitive or augmented reality setting. Demographic and perioperative variables were collected. The nerve sparing (NS) approach according to Pasadena classification and the PSM rate were evaluated both in 3D and non-3D groups. Modelling relied on multivariable logistic regression (MLR) models predicting the presence of PSM in the overall population and in 4 specific subgroups of patients: extra capsular extension patients, D'Amico high risk category, positive Digital Rectal Exam (DRE), cT3 patients. RESULT(S): 757 patients were enrolled. 3D technology was used in 160 patients during the procedure. In particular, both 3D cognitive (n=56, 35%) and 3D augmented reality (n=104, 65%) procedure were performed. No differences were observed in terms of preoperative and demographic characteristics and complications rates between 3D and non-3D groups. 3D patients were treated more often with full NS approach (20.6 vs. 12.7%, p <0.05) and had lower PSM rates (25 vs. 35.3%, p=0.01), compared to their non-3D counterpart. A subgroup analysis according to the type of NS showed no difference in terms of PSM rates between the two groups (33% vs. 36.8%, 22.9% vs. 32.1% and 22.7% vs.37.4% for respectively full, intermediate and standard NS approach, p >0.05). Finally, at MLR, the availability of 3D technology (OR: 0.5, CI:0.3-0.8, p=0.005) and the absence of extracapsular extension (ECE, OR: 0.6, CI:0.4-0.9, p=0.004) at mpMRI were independent predictors of lower rates of PSM. CONCLUSION(S): The use of 3D models during RARP allows to decrease the PSM rates. Moreover, it grants a more conservative NS approach, without compromising the PSM rate. However, the low sample size does not allow to identify a specific subgroup of patients in whom the advantage of 3D technology is maximized.