SAT0071 Rofecoxib shows consistent efficacy in oa clinical trials, regardless of specific patient demographic and disease factors

Background Objectives Analyses were performed to demonstrate the generalizability of the efficacy of rofecoxib (MK-0966, VIOXXTM) in patients with different demographic or baseline disease characteristics, or with OA in nonsignal joints or multiple locations. Methods Data were combined for patients with OA of the knee or hip receiving placebo, 12.5 or 25 mg rofecoxib once daily in three 6-week placebo controlled trials, the only trials to date using all 3 treatments. Predefined analyses were performed after categorising patients according to baseline demographics (age, gender, height, weight BMI) and baseline disease characteristics (ARA functional class, joint tenderness, joint stiffness, WOMAC functional subscale, unilateral/bilateral joint involvement, number of joint groups involved). Three primary endpoints: Pain Walking on Flat Surface (WOMAC, 0- 100 mm VAS), Patient Global Assessment of Response to Therapy (0 to 4 point Likert Scale), and Investigator Global Assessment of Disease Status (0 to 4 point Likert Scale) were analysed. Efficacy treating OA in patients with 1, 2, 3, or 4, joint groups (among interphalangeal/first carpal-metacarpal joint, spine, hip or knee) affected was analysed for four end points, Patient and Investigator Global Assessments of Response to Therapy and of Disease Status (which provided data on overall aspects of OA, regardless of affected joint). Treatment by factor interactions were tested by ANOVA. Results Data from 1501 patients were included. For patients taking 12.5 or 25 mg rofecoxib, treatment effects were generally consistent across subgroups categorised by demographic factors: age (p = 0.08), gender (p = 0.06), race (p = 0.46), prior therapy (p = 0.19), weight (p = 0.10), BMI (p = 0.087) or disease characteristics: baseline joint tenderness (p = 0.17), WOMAC functional subscale (p = 0.52) ARA functional class I, II, or III (p = 0.49), duration of OA (p = 0.37), or primary study joint (p = 0.27). Greater efficacy was seen in patients with study joint swelling; however, the difference from those without study joint swelling was not clinically meaningful. No significant interactions were observed among patients with unilateral/bilateral involvement or 1, 2, 3, or 4, joint groups affected. Conclusion No specific factor predicted treatment effect in this study. This study was funded by a grant from Merck& Co., Inc.