Category: Laparoscopic/ Robotic: Prostate

MP18-19 - A Novel Side Specific mpMRI-based nomogram for the prediction of Extra Capsular Extension of Prostate Cancer and Update of the Incremental Nerve Sparing Algorithm

Sat, Sep 22
10:00 AM - 12:00 PM

Introduction & Objective :

Knowledge of prostate cancer extracapsular extension (ECE) preoperatively is of pivotal importance for planning nerve sparing radical prostatectomy (NSRP). We aimed to create a model that predicts side-specific ECE and to update our algorithm for incremental NS.


Methods :

data from 561 patients who underwent mpMRI before robot-assisted radical prostatectomy were identified. To develop a side-specific predictive model, we considered the two prostatic lobes singularly after excluding lobes with no biopsy documented tumor. The role of PSA, highest ipsilateral biopsy Gleason grade, highest ipsilateral percentage core involvement with the highest Gleason and ECE on MRI was investigated in a binary multivariable logistic regression. A nomogram was built based on the coefficients of the logit function. The leave-one-out cross validation was used for internal validation and calibration was graphically investigated. The decision curve analysis (DCA) was used to evaluate the net clinical benefit. The nomogram-derived probability (P(%)), after internal validation, was used as the independent variable on a regression tree to identify the most significant nodes for ECE prediction.


Results :

After excluding negative biopsy observations (n=293), the study population consisted of 829 side-specific observations. ECE was reported on MRI and final pathology in 115 (14%) and 142 (17.1%) cases, respectively. All variables in the model except highest percentage core involvement were predictors of ECE (all p≤0.006, tab. 1) and were considered to build a nomogram. After internal validation, the AUC was 82.11%. The model demonstrated excellent calibration and improved clinical risk prediction at the DCA, especially when compared to relying on MRI prediction of ECE alone. The regression tree identified three nodes: 10%, 21% and 73%. Four risk categories were defined. In the low (P(%)≤10), intermediate (10< P(%)≤21), high (21< P(%)≤73) and very high-risk (P(%)>73) groups, the ECE rate was 3.3%, 16%, 61.6% and 90%, respectively. Among those, ECE was focal in 41.7%, 31.7%, 7.9% and 0%, respectively. 


Conclusions :

We developed a nomogram that accurately predicts ECE and updated our previous algorithm for incremental NS during RARP. We suggest the grade 1 (intrafascial) NS to be performed in the low risk group. The interfascial approach, namely grade 2 and 3, should be performed in the intermediate and high-risk categories, respectively. Grade 4 (extrafascial) should be performed in the very high-risk group.

Alberto Martini

Fellow
Icahn School of Medicine at Mount Sinai
New York, New York

Akriti Gupta

New York, New York

Shivaram Cumarasamy

Icahn School of Medicine at Mount Sinai
New York, New York

Alp Tuna Beksac

Fellow
Icahn School of Medicine at Mount Sinai
New York, New York

Endourological Society Fellow in Minimally Invasive and Robotic Surgery at Icahn School of Medicine at Mount Sinai.

Sara C. Lewis

New York, New York

Kenneth G Haines

New York, New York

Ashutosh K. Tewari

Icahn School of Medicine at Mount Sinai
New York, New York