Presentation Authors: Gaetan Devos*, Leuven, Belgium, Alberto Briganti, Milan, Italy, Jeffrey Karnes, Rochester, NY, Bertrand Tombal, Brussels, Belgium, Henk Van Der Poel, Amsterdam, Netherlands, Jochen Walz, Marseille, France, Alberto Bossi, Villejuif, France, Gert De Meerleer, Hein Van Poppel, Leuven, Belgium, Giansilvio Marchioro, Novara, Italy, Lisa Moris, Thomas Van den Broeck, Leuven, Belgium, Markus Graefen, Hamburg, Germany, Christian Gratzke, Munich, Germany, Rafael Sanchez-Salas, Paris, France, Martin Spahn, ZÃ¼rich, Switzerland, Wouter Everaerts, Leuven, Belgium, Paolo Gonterro, Turin, Italy, Steven Joniau, Leuven, Belgium
Introduction: According to the dÃ¢â‚¬â„¢Amico classification, high-risk prostate cancer (PCa)disease is defined as either preoperative PSA>20ng/ml, clinical tumor stage (cT) Ã¢â€°Â¥T2c or biopsy Gleason score(bGS) 8-10.The International Society of Urological Pathology (ISUP) has further stratified this high risk population in 2 subgroups depending whether they have bGS8 (Grade group 4) or 9-10 (Grade group 5). This study aims at investigating the impact of preoperative PSA and cT-stage in patients with very-high-risk bGS 9-10 (Grade group 5) PCa on cancer-specific survival (CSS) and overall survival (OS).
Methods: From the multicenter EMPaCT database, which contains detailed clinico-pathological and cancer outcome data of high-risk PCa patients treated with radical prostatectomy, we selected all patients with bGS9-10. In this subgroup, we investigated the impact of bGS (9 vs. 10), preoperative PSA (Ã¢â€°Â¤20ng/ml vs.>20ng/ml) and clinical T-stage (Ã¢â€°Â¤cT2c vs. T3a vs. T3b-4) on CSS and OS. Furthermore, the possible interaction of supplementary risk factors on CSS and OS in patients with bGS 9-10 were studies by subdividing patients into 6 subgroups: Ã¢â€°Â¤cT2c+Ã¢â€°Â¤20ng/ml, T3a+Ã¢â€°Â¤20ng/ml, T3b-4+Ã¢â€°Â¤20ng/ml, Ã¢â€°Â¤cT2c+>20ng/ml, T3a+>20ng/ml and T3b-4+>20ng/ml. We used Kaplan-Meier analysis with log rank tests to compare subgroups.
Results: A total of 1370 patients were included in the analysis of whom 487(35.5%) and 537(39.2%) had data on CSS and OS available respectively. Of all patients, 1233 (90%) had biopsy Gleason score 9 and 137 (10%) biopsy score 10. Most patients had a PSA-value Ã¢â€°Â¤20ng/ml (74.3%) and cT-stage Ã¢â€°Â¤cT2c(79.4%). We did not observe a significant impact of preoperative PSA independent of cT-stage (and vice-versa) on CSS and OS. Nor did we find significant differences in CSS or OS between the 6 subgroups (p=0.9 and p=0.99, respectively). Patients with bGS 9 had a trend towards better CSS vs.those with bGS10 (Mean CSS 145 vs. 96 mo., p=0.06).
Conclusions: In PCa patients with bGS 9-10, additional preoperative high-risk factors (PSA>20ng/ml and cT3b-4) had no measurable impact on CSS and OS. Therefore, bGS9-10 on itself appeared to be the single-most important presurgical predictor of OS and CSS.