Skip navigation
Acute Pain | Chronic Pain | General

Tenoxicam versus piroxicam, diclofenac and indomethacin for osteoarthritis

Clinical bottom line:

In the treatment of osteoarthritis, there is no difference between tenoxicam and piroxicam for efficacy, including pain relief. However, tenoxicam is associated with fewer adverse effects when compared with piroxicam, with a number-needed-to-harm of -12 (-67 to 7) for total number of patients experiencing adverse effects.

There appears to be no difference in efficacy or harm between tenoxicam, diclofenac and indomethacin. However, this assessment is based on a small dataset.


NSAIDs are widely used for symptom control in inflammatory rheumatic disease. There are five main groups of NSAIDs, the salicylates, arylalkanoic acids, anthranilic acids, pyrazolones and oxicams. Diclofenac and indomethacin are both acetic acid derivatives of arylalkanoic acid.

Systematic review

Riedemann PJ, Bersinic S, Cuddy LJ, Torrance GW, Tugwell PX. A study to determine the efficacy of safety and tenoxicam versus piroxicam, diclofenac and indomethacin in patients with osteoarthritis: a meta-analysis. J Rheumatol. 1993; 20: 2095-103.

Inclusion criteria were randomised, double-blind trials of tenoxicam vs. piroxicam, diclofenac or indomethacin; outcomes pain rating and/or physician global rating; extractable data for numerical analysis; treatment of at least 4 weeks. However, some non-randomised and non-blinded trials were included.

Mean and dichotomous data were extracted from trials, and were weighted according to sample size. Reviewers calculated odds ratios for physician global ratings of treatment. We present this data as a relative benefit and number-needed-to-treat with 95% confidence intervals. Pain ratings were based on means weighted by trial size. These were expressed as effect size, where effect size was calculated for each trial (difference between two treatments divided by the pooled standard deviation). Pooled effect size was a weighted mean of individual effect sizes (using variance to weight). All statistical outputs were with 95% confidence intervals.

Findings

Tenoxicam vs. piroxicam

Eleven of 13 trials compared tenoxicam 20 mg/day with piroxicam 20 mg/day. One trial compared 40 mg/day with 40 mg/day, and one trial compared tenoxicam 20 mg/day with piroxicam 40 mg/day.

Six trials looked at pain in 152 patients (pain on movement and pain after normal day's activity). Of these, four were included in the aggregate estimate. There was no significant difference between tenoxicam and piroxicam on this measure.

Ten trials looked at physician global assessment in 835 patients. Successful treatment was a physician rating of excellent or good. When tenoxicam was compared with piroxicam, the relative benefit was 1.1 (1 to 1.2) and the number-needed-to-treat was 12 (7 to 54) for successful treatment, suggesting no benefit of tenoxicam over piroxicam.

Assessment of tolerability (excellent or good tolerance was regarded as side effects absent or not pronounced). Seven trials looked at physician global assessment of tolerability. When tenoxicam was compared with piroxicam, the relative benefit was 1.1 (1 to 1.1) and the number-needed-to-treat was 20 (11 to 243) for successful treatment, suggesting no difference between tenoxicam and piroxicam. (Reviewers rate this as a marginally significant advantage with tenoxicam).

Tenoxicam vs. Diclofenac

Four of four trials (757 patients) compared tenoxicam 20 mg/day with diclofenac 100 mg/day. One trial had an additional tenoxicam 40 mg/day and placebo group.

Based on two trials which allowed data pooling, reviewers found no significant differences on pain measures and physician global assessment, but stated that group size was too small to detect a difference.

Reviewers judged tolerability to be the same.

Tenoxicam vs. indomethacin

Two trials looked at 221 patients. One compared tenoxicam 20 mg/day with indomethacin sustained release 75 mg/day, and one compared tenoxicam 20, 30, 40 mg/day with indomethacin 75 mg/ day (25 mg x 3).

Pain measures could not be pooled. Reviewers state that physician global assessment showed no difference, but sample sizes were small. Reviewers found tenoxicam to be significantly better tolerated than indomethacin.

Adverse effects

Two trials reported on self-reported adverse effects, five on direct questioning, and five were unspecified methods. Ten trials allowed data pooling of total adverse effects for tenoxicam vs. piroxicam. For total number of patients experiencing adverse effects, this gave a relative benefit of 0.86 (0.76 to 0.97), with a number-needed-to-harm of -12 (-67 to 7), suggesting that piroxicam causes more harm than tenoxicam. Reviewers noted a significantly lower rate of dropouts due to adverse effects with tenoxicam (17%).

Reviewers found no significant differences between tenoxicam and diclofenac for total number of patients with adverse effects.

Reviewers found that tenoxicam was associated with significantly more patients suffering adverse effects than indomethacin.

Related topics