Despite being the most common over-the-counter treatment for pain related to knee osteoarthritis (OA), acetaminophen was found inferior to other commonly used pain medications in a large network meta-analysis. Intra-articular (IA) treatments were superior to oral nonsteroidal anti-inflammatory drugs (NSAIDs) for pain relief in this study.
Lead author Raveendhara R. Bannuru, MD, and colleagues at Tufts Medical Center in Boston, Massachusetts, wrote that the superiority of IA treatments is possibly due to a placebo effect. “The information from this study, along with the safety profiles and relative costs of included treatments, should be helpful to clinicians when making care decisions tailored to individual patient needs,” they wrote.
In the absence of definitive head-to-head trials of medications used to treat knee OA, the authors conducted a large meta-analysis to get a handle on the comparative effectiveness of commonly used drugs in this setting. They reviewed a total of 33,243 participants in 137 randomized controlled trials published between 1980 and 2014; each trial compared 2 or more treatments for knee OA and reported at least 1 measure of pain, function, or stiffness. Studies were drawn from searches of MEDLINE, EMASE, Web of Science, Google, Scholar, and Cochrane Central Register of Controlled Trials, as well as unpublished data. Treatments studied included acetaminophen, diclofenac, ibuprofen, naproxen, celecoxib, IA corticosteroids, IA hyaluronic acid, oral placebo, and IA placebo. An absolute change of 20 points on a scale from 0 to 100 on the OA Research Society International-Outcome Measures in Rheumatology responder criteria was deemed clinically significant.
Reported outcomes differed in these trials; 129 trials with 32,129 participants contributed to the analyses of pain-related outcomes, 76 trials (24,059 participants) contributed to the analyses of physical function outcomes, and 55 trials (18,267 participants) contributed to the analyses of stiffness outcomes. Age of participants ranged from 45 to 76 years, and the proportion of women ranged from 3% to 100%.
All interventions were statistically superior to placebo for pain-related outcomes. The most efficacious pain relief occurred with IA hyaluronic acid and the least efficacious was acetaminophen; effect sizes (standard mean differences) were 0.63 and 0.18, respectively. All treatments except acetaminophen met the prespecified criteria for clinically significant improvement. All active treatments except celecoxib were statistically superior to acetaminophen; IA treatments were more effective than oral treatments.
All interventions were significantly superior to oral placebo for function, except IA corticosteroids. Naproxen, ibuprofen, diclofenac, and celecoxib were statistically significantly better than acetaminophen for functional outcomes. IA hyaluronic acid was significantly better than IA placebo and IA corticosteroids.
All oral treatments were significantly superior to acetaminophen for stiffness. IA hyaluronic acid was significantly better than IA placebo for stiffness.
Traditional nonselective NSAIDs were associated with more gastrointestinal (GI) adverse events and withdrawals compared with oral placebo and acetaminophen; GI adverse events were similar between acetaminophen and celecoxib. The short exposure time of 2 to 3 months probably affected reporting on adverse cardiovascular events, the authors wrote. Transient local reactions were the most commonly reported adverse events related to IA treatments, and the frequency was similar among IA therapies.
The network meta-analysis did not investigate combination therapies, even though these are commonly used to treat knee OA. “The large number of possible therapy combinations and scarcity of trials comparing these treatments prompted us to restrict our analysis to monotherapies,” the authors wrote.
Bannuru RR, et al. Ann Intern Med. 2015;162(1):46-54.
Last modified: May 21, 2015