The quality of evidence is downgraded by study limitations (high loss to follow-up and differential loss to follow-up in the compared groups).
A Cochrane review included 8 studies with a total of 3 988 adult subjects. Study durations varied from 4 to 24 weeks. Participants had rheumatoid arthritis for an average of 9.2 years, their mean age was 54 years, and most of them (73%) were women.
Celecoxib versus placebo: Participants who received celecoxib (200 mg or 400 mg daily) showed significant clinical improvement and reported less pain compared with those receiving placebo, but results were inconclusive for improvement in physical function (table ). There was no statistically significant difference in the incidence of gastroduodenal ulcers or short-term serious adverse events between the groups. There were fewer withdrawals among people who received celecoxib. Cardiovascular events (myocardial infarction, stroke) were not reported. However, regulatory agencies warn of increased cardiovascular event risk associated with celecoxib.
| Outcome | Relative effect (95% CI) | Risk eith placebo | Risk with celecoxib (95% CI) | NNTB/NNTH (95% CI) | Participants (studies) |
|---|---|---|---|---|---|
| * higher scores means worse functional ability ; ** statistically significant heterogeneity I² = 86% | |||||
| Clinical improvement (ACR 20) | RR 1.53 (1.25 to 1.86) | 288 per 1000 | 441 per 1000 (360 to 536) | NNTB 7 (5 to 13) | 873 (2 studies) |
| Pain (VAS from 0 to 100) | - | Mean pain 60 | Mean pain was 11 points lower (14.04 lower to 7.96 lower) | NNTB 4 (3 to 6) | 706 (1 study) |
| Physical function (HAQ scale 0 to 3*) | - | Mean change in physical function -0.1 | Mean change in physical function was 0.1 point better (0.29 better to 0.1 worse)** | - | 706 (1 study) |
| Incidence of gastroduodenal ulcers ≥ 3 mm | Peto OR 1.26 (0.44 to 3.63) | 40 per 1000 | 51 per 1000 (17 to 142) | - | 392 (1 study) |
| Short-term serious adverse events | Peto OR 0.87 (0.28 to 2.69) | 22 per 1000 | 19 per 1000 (6 to 55) | - | 706 (1 study) |
| Total withdrawals | RR 0.61 (0.52 to 0.72) | 563 per 1000 | 343 per 1000 (293 to 405) | NNTH 5 (4 to 7) | 706 (1 study) |
Celecoxib versus traditional NSAIDs (tNSAIDs: amtolmetin guacyl, diclofenac, ibuprofen, meloxicam, nabumetone, naproxen, pelubiprofen): Pooled analysis showed no statistically significant difference between celecoxib and tNSAIDs in ACR20 improvement criteria (RR 1.10, 95% CI 0.99 to 1.23; 4 studies, n=1 981), in pain (MD -1.59, 95% CI -3.83 to 0.65; 3 studies, n=1 504), or in self-reported physical function assessed using the HAQ (MD 0.00, 95% CI -0.13 to 0.13; 2 studies, n=849). People who received celecoxib had a lower incidence of gastroduodenal ulcers ≥ 3 mm compared with those who received tNSAIDs (RR 0.22, 95% CI 0.15 to 0.32; 5 studies, n=1 568). There were 7% fewer withdrawals among people who received celecoxib (RR 0.73, 95% CI 0.62 to 0.86; 6 studies, n=2 639) compared to tNSAIDs. Results were inconclusive for short-term serious adverse events (Peto OR 0.71, 95% CI 0.39 to 1.28; 5 studies, 2 154 participants) and cardiovascular events (Peto OR 1.13, 95% CI 0.07 to 18.33; 1 study, n=149) between celecoxib and tNSAIDs.
Contraindications to the systemic use of celecoxib include coronary heart disease, diagnosed disorders of the cerebral circulation, heart failure, and peripheral vascular disease.
A Cochrane review