Preoperative Predictors of Extraprostatic Extension of Prostate Cancer (pT3a) in a Contemporary Indian Cohort

We present our experience with robotic radical prostatectomy and analyse the predictors for extracapsular extension (EPE) in a cohort of Indian patients. Four hundred fifty-three patients underwent robotic radical prostatectomy from May 2010 to October 2015. Post hoc analysis of prospectively maintained data was done. Multivariable regression analysis was used on variables which had significant association with EPE on univariable analysis. Receiver operating characteristic (ROC) curves were plotted for the independent predictors of EPE. Mean age was 64.62 ± 6.44 years. Mean BMI 26.03 ± 4.01 kg/m². Median prostate-specific antigen (PSA) was 13.38 ng/ml (IQR 8.1-23.5). Seventy-seven (16.99%) patients had Gleason >7. Mean core positivity (percentage of cores positive) was 48,93 ± 27.29% and mean maximum percentage of tumour in a core was 33,04 ± 31.16%. Positive surgical margins were present in 141/453 (31.15%). Lymph nodes were positive in 21.2% of those who had nodes removed (77/364). Maximum cancer in a core (p < 0.001), core positivity (p = 0.002) and Gleason score ≥4 + 3 (p = 0.028) were significant predictors of EPE, independent of BMI, PSA, PNI and clinical stage. Receiver operating characteristic (ROC) analysis for core positivity showed an area under the curve (AUC) of 0.775, and a 76% core positivity predicted EPE with a sensitivity of 65% and a specificity of 87.1%, respectively. ROC curve for maximum cancer in a core showed an area under the curve of 0.898. Seventy per cent cancer in a single core predicted EPE with a sensitivity of 80.6% and a specificity of 85.9% Indian patients present with more advanced disease, higher PSA and have higher incidence of EPE. Maximum cancer in a core, ratio of positive cores and Gleason score ≥4 + 3 are predictors of EPE independent of PSA and clinical stage.