ABSTRACT Background ENZA-proposed INN (MDV3100) increased median survival by 4.8 mo (P Methods Pts were randomized 2:1 to ENZA 160 mg/d or PBO, stratified by baseline ECOG & mean Brief Pain Inventory score (BPI). Efficacy endpoints included Overall Survival (OS), radiographic progression-free survival (rPFS), & time to PSA progression (TTPP). In a post-hoc analysis, pts were evaluated by oral daily CS baseline use. Using a stepwise multivariate Cox proportional-hazards analysis, associations between the OS and the prespecified baseline factors ( treatment [Trt], age, region, ECOG, BPI, visceral disease, number of chemotherapy regimens, progression type at entry, LDH, PSA, hemoglobin, alkaline phosphatase) and CS use were evaluated. Results 800 pts were randomized to ENZA & 399 to PBO. ∼30% of pts in each arm were on CS at baseline. Multivariate analysis showed baseline oral CS use was associated with the OS outcome, with median OS of 11 mo (95% CI: 10, 13) for pts on CS vs median OS not met (18, NM) for pts not on baseline CS (HR = 0.54; 95% CI: 0.45, 0.64)). Subgroup analysis showed for pts on CS median OS was 12.3 mo on ENZA vs 9.3 mo on PBO (P = 0.0116, HR 0.70) and for pts not on CS median was not met on ENZA vs 15.8 mo on PBO (P Results of Stepwise Multivariate Analysis of Overall Survival Variable Parameter Estimates HR for Death (95% CI) Coefficient P-value Treatment (ENZA vs PBO) -0.54 ± 0.090 0.58 (0.49-0.70) Mean pain score: -0.25 ± 0.098 0.0091 0.78 (0.64-0.94) Progression at study entry: PSA only vs radiographic -0.35 ± 0.094 0.0002 0.70 (0.59-0.85) Visceral disease at screening (No vs Yes) -0.42 ± 0.097 0.66 (0.54-0.80) Baseline hemoglobin -0.03 ± 0.003 0.97 (0.97-0.98) Baseline lactate dehydrogenase (LDH) 0.002 ± 0.000 1.002 (1.001-1.002) Baseline corticosteroid use (No vs Yes) -0.62 ± 0.091 0.54 (0.45-0.64) 1. Survival for pts who had not died by the time of analysis was censored at the date the pt was last known to be alive. 2. confidence interval Conclusions Use of CS at baseline was associated with reduced OS in the multivariate model. This finding should be further explored in future studies. While pts on baseline CS had worse outcomes, ENZA was consistently superior to PBO on OS, rPFS, and TTPP regardless of baseline CS use. Disclosure H.I. Scher: Consultant/Advisor Amgen, Aragon, BMS, Dendreon, Exelixis, Foundation, Janssen, J & J, MDV, Millennium, Novartis, Ortho Biotech, Sanofi Aventis. Research support from Aragon, BMS, Exelixis, Medivation, Orthobiotech & Veridex, F. Saad: Grants: Astellas; Consulting fee or honorarium: Astellas, Medivation, Johnson & Johnson; Board membership: Medivation, Astellas, Johnson & Johnson. M. Taplin: Consulting fee or honorarium: Medivation, JNJ, Tokai; Support for travel to meetings for the study or other purposes: Medivation; Travel/accomodations/meeting expenses unrelated to activities listed: JNJ, Tokai. C.N. Sternberg: Consultancy fee or honorarium: Johnson & Johnson, Astellas, Amgen, Sanofi-Aventis. A.J. Armstrong: Consultancy: Amgen, BMS, Sanofi-Aventis; Grants/grants pending: Medivation, Janssen Biotech, Sanofi-Aventis, Imclone, BMS, Pfizer, Novartis, Active Biotech, J & J/Veridex, Dendreon; Payment for lectures: Sanofi-Aventis, Dendreon, Janssen Biotech, Amgen. M. Hirmand: Employement: Medivation; Stock/stock options: Medivation, B. Selby: Employement: Medivation; Stock/stock options: Medivation; Payment for writing or reviewing the manuscript: Medivation; Fees for participation in review activities such as data monitoring boards, statistical analysis, ent point committies, Medivation. J.S. de Bono: Consultancy: Medivation, Astellas. All other authors have declared no conflicts of interest.
Read full abstract