Nerve-sparing (NS) procedures in radical prostatectomy have gained importance as they show better functional outcomes in terms of erectile function (1). It is recommended that the neurovascular bundle (NVB) should be removed if residual tumor is suspected. Alternatively, intraoperative fresh frozen section (FFS) may help in this decision. However, standard FFS techniques do not provide a direct visual impression of the positive surgical margin (PSM) for guidance regarding secondary resection (SR) at the GNB (2-4). We therefore intend to optimize precision for SR using a personalized 3D printed prostate model in robotic-assisted radical prostatectomy (RARP).
We prospectively included 100 patients in which NS-RARP was performed between September 2018 and August 2021 at the University Hospital Basel and Cantonal Hospital Basel-Land. Image data of the preoperative pelvic MRI scan were used to render and print a personalized 3D prostate model. PIRADS lesions (> grade 3) and potential capsular involvement were color-characterized in the model. The model was used during surgery to allow the pathologist to mark a positive surgical margin (PSM) on the model and to help guide the surgeon for selective SR (SSR). Functional outcome was assessed using IIEF-5 prior and one year after surgery.
17.0% (17/100) of all patients had PSM in final histology with a rate of 15.4% in pT2 and 25% in pT3 tumors, respectively. PSM was identified during FFS in 23.0% (23/100) of patients, and SSR achieved a tumor-free surgical margin on final histology in 19 of these 23 patients (82.6%), with SSR containing tumor tissue in 9 of 23 patients (39.1%). Intraoperative PSM was associated with a PIRADS lesion in 11 patients (48%). PSA was undetectable 6 weeks postoperatively in 18/19 (94.7%) patients with successful SSR but only in 76.5 % of patients with PSM in final pathology. False negative rate of FFS was 16.9% (13/77). Despite 6/8 (75%) patients with biochemical recurrence (BCR) having a PSM successfully converted, only 1 had image proven local recurrence.
3D printed prostate models are a promising approach to precision SSR and improve communication between surgeons and pathologists. Successful conversion rates (82.6%) were similar to procedures using NEUROSafe (86%). The sensitivity of MRI, the false-negative rate of conventional FFS and cost effectiveness remain critical limitations.