Multiple published trials ^1,2,3 of a novel formulation of a topical diclofenac solution (TDS) have established efficacy in knee osteoarthritis (OA). Defining a clinically meaningful response remains controversial. In pain trials, a standard clinically significant response has been defined as ≥30% improvement in pain ('pain criterion'),⁴ whereas OA experts (OMERACT-OARSI) define criteria based on a composite of 3 outcome domains: pain, physical function (PF) and patient global assessment (PGA). ⁵ Based on this 'OA criteria', a responder should demonstrate at least (a) 50% improvement in score and absolute improvement of at least 20% of the domain scale, for pain or PF, OR (b) 20% improvement in score that is at least 10% of the scale, for at least 2 of the 3 co-domains.

We applied both the pain and OA criteria to the data of 1138 patients with knee OA from 3 double-blind TDS trials, 2 vehicle (V)-controlled trials^1,2 and one equivalence trial to oral diclofenac (OD).³ Using the OA criteria, TDS demonstrated consistent efficacy across all 3 trials¨D65.7%, 73.6% and 66.3% responders. Percentage of patients responding to TDS was significantly greater than the response to V (65.7 vs. 49.5%, P=0.02; and 73.6 vs. 59.1%, P=0.006) and not different from OD (66.3 vs. 69.8%, P=0.37). Using the pain criterion, TDS demonstrated consistent efficacy across all 3 trials--61.0%, 66.9% and 61.4% responders. This response was significantly greater than to V (61.0 vs. 43.9%, P=0.01; and 66.9 vs. 54.7%, P=0.03) and not significantly different from OD (61.4 vs. 66.8%, P=0.17). There was no statistical difference between the two clinically meaningful response criteria in predicting efficacy of TDS relative to V or OD (P=0.93).

Conclusion: A standard pain responder criterion may yield results similar to OA-specific criteria. TDS provides statistically meaningful pain relief equivalent to OD regardless of which criteria are utilized.