3248 - Biochemical Recurrence is Not a Surrogate for Overall Survival in Localized Prostate Cancer: Data from 2 Phase III Trials

A. Nabid¹, N. Carrier¹, E. Vigneault², A. G. Martin², T. V. Nguyen³, J. P. Bahary⁴, P. Vavassis⁵, B. Bahoric⁶, M. A. Brassard⁷, S. T. Vass Jr⁸, R. Archambault⁹, F. Vincent¹⁰, R. Bettahar¹¹, M. Duclos¹², D. R. Wilke¹³, and L. Souhami¹²; ¹Centre hospitalier universitaire de Sherbrooke, Sherbrooke, QC, Canada, ²CHU de Québec – Université Laval, Radiation Oncology, Québec, Canada, Quebec, QC, Canada, ³Centre hospitalier universitaire de Montréal, Montréal, QC, Canada, ⁴Department of Radiation Oncology, Centre Hospitalier de lUniversité de Montréal, Montreal, Quebec, Montreal, QC, Canada, ⁵Maisonneuve-Rosemont Hospital, Montreal, QC, Canada, ⁶Jewish General Hospital, Montreal, QC, Canada, ⁷CIUSSS du Saguenay-Lac-Saint-Jean, Chicoutimi, QC, Canada, ⁸CSSS Chicoutimi, Chicoutimi, QC, Canada, ⁹Gatineau, Gatineau, QC, Canada, ¹⁰Centre hospitalier regional de Trois-Rivieres, Trois-Rivieres, QC, Canada, ¹¹CSSS Rimouski-Neigette, Rimouski, QC, Canada, ¹²McGill University Health Centre, Montreal, QC, Canada, ¹³Capital Health, Halifax, NS, Canada

Purpose/Objective(s): Aiming to confirm that biochemical recurrence (BR) is not a surrogate for overall survival (OS) in patients (pts) treated for intermediate risk prostate cancer (IRPC) and high-risk prostate cancer (HRPC), we analysed data from our two prospective phase III randomized trials. Materials/

Methods: From October 2000 to September 2010, 1230, pts were randomized in two phase III trials. 600 pts with IRPC were treated using prostate radiotherapy (RT) alone (200 pts) or with 6 months (m) ADT (400 pts) and 630 pts with HRPC were treated with pelvic + prostate RT and 18 or 36 m ADT. The Chi-squared test compared BR rates between cohorts. The log rank test compared the OS rates between IRPC vs. HRPC pts and in pts with or without BR by considering all follow-up values. OS rates were estimated with Kaplan Meier curve.
Results: With a median follow up of 16.3 years(y) (interquartile range 13.0 à 19.5), 722/ 1230 (58.7%) pts had died. Median age at randomization was 71 y and 82 y at death. The 10-y OS rates in IRPC vs HRPC were respectively 73% (69-77) vs. 63% (60-67). HRPC pts died significantly faster than IRPC pts [Hazard Ratio (95% confidence interval) = 1.30 (1.12-1.51), p<0.001]. 153 IRPC pts (25.5%) developed BR at a median time of 6.5 y post-randomisation and 89/153 had died. 47 pts (7.8%) presented BR in the first 5 y post-randomisation, 84 (14%) between 6 and 10 y and 22 (3.7%) after 10 y. The rate of BR was significantly higher in pts not receiving ADT: 79/202 (39.1%) vs 74/398 (18.6%), p<0.001. 447/600 (74.5%) pts were free from BR at a median FU of 13.4 years. 49 IRPC pts (8.2%) died in the first 5 y post-randomisation, 112 (19%) between 6 and 10 y and 144 (24%) after 10 y. There was no significant difference in OS between pts with vs without BR [10-y OS: 79% (73-86) vs 71% (67-75), HR (95%C-I) = 0.93 (0.72-1.19), p=0.5]. 197 HRPC pts (31.3%) developed BR at a median time of 5.9 y post-randomisation (IQR 3.5-7.9) and 139/197 had died. 91 pts (15.2%) presented BR in the first 5 years post-randomisation, 80 (13.3%) between 5 and 10 y, and 26 (4.3%) after 10 y. The rate of BR was significantly lower in pts receiving 36 vs 18 m: 85/310 (27.4%) vs. 112/320 (35.0%), p=0.048. 433/630 (68.7%) pts were free from BR at a median FU of 15.9 y (IQR 12.9-18.7). 72 HRPC pts (12%) died in the first 5 y post-randomisation, 155 (26%) between 6 and 10 y and 190 (32%) after 10 y. There was no significant difference in OS between pts with vs without BR [10-y OS: 63% (57-70) vs. 63% (59-68), HR (95% CI) = 0.92 (0.75-1.13), p=0.4].
Conclusion: Based on the long-term follow up of 1230 pts with localized PC, we confirm that BR is not a surrogate for OS in IRPC or HRPC.