Trans-capsaicin injections minimally improve pain in patients with knee degenerative joint disease

Reference

Stevens RM, Ervin J, Nezzer J, et al. Randomized, double-blind, placebo-controlled trial of intraarticular trans-capsaicin for pain associated with osteoarthritis of the knee. Arthritis Rheumatol 2019;71(9):1524-1533.

Synopsis

We have covered MANY studies of joint injections for patients with arthritic knees (hyaluronic acid, platelet-rich plasma, corticosteroids, ozone, prolotherapy, and so forth) and, at best, steroids provide short-term relief of pain. In this industry-led study, the researchers enrolled patients with radiographically confirmed moderately to severely painful degenerative joint disease of one knee. Pain had to be at least 5 on a 10-point scale in the afflicted knee and no more than 3 in the contralateral knee. The authors randomized the patients to receive intra-articular injections of placebo (n = 70), 0.5 mg trans-capsaicin (n = 34) or 1 mg trans-capsaicin (n = 71). The researchers used the Western Ontario and McMaster Universities Osteoarthritis Index and recorded pain-while-walking scores for each patient at baseline and for up to 24 weeks after injection. However, rather than compare the difference in the average scores for these measures, the researchers (also stockholders in the company that makes the trans-capsaicin) report the area under the curve for the changes from baseline until 12 weeks as the primary outcome. At clinicaltrials.gov (https://clinicaltrials.gov/ct2/show/NCT02558439), the authors' stated primary and secondary outcomes were the difference in average pain scores, with no mention of the nonstandard area under the curve. This is very odd, and may be an example of Procrustean data torturing (Mills JL. Data torturing. N Engl J Med 1993;329(16):1196-1199.). Oh, and rather than using 95% confidence levels, they used 90% confidence levels, increasing their chances of finding a statistically significant difference. After all these statistical gymnastics, the authors report that each dose of capsaicin was more effective than placebo, that the higher dose was most effective at 12 weeks, and only the higher dose maintained its effect at 24 weeks. Given the nonstandard assessment, it is unclear of these differences are clinically meaningful. Finally, the injection procedure must have been very painful. The treating physicians could pre-treat the patients with analgesics. They also cooled each knee for 15 minutes, then injected 15 mL of 2% lidocaine followed by another 30 minutes of cooling. After the "therapeutic" injection, they cooled the knee for an additional 30 to 60 minutes. Wow.