NSAIDs and faecal blood loss in elderly patients with osteoarthritis: is plasma half-life relevant?

Nonsteroidal anti-inflammatory drugs (NSAIDs) vary in their degree of gastrointestinal (GI) toxicity. NSAIDs with longer half-lives are of particular concern as they may be more toxic in the elderly.

To compare the GI toxicity, by measurement of faecal blood loss, of short, intermediate and long half-life NSAID treatments compared with control in elderly patients with osteoarthritis.

Twenty-three patients, mean age 69 years, with osteoarthritis requiring NSAID treatment, received treatment with diclofenac 100 mg/day, naproxen 750 mg/day and piroxicam 20 mg/day, representing a short, medium and long half-life NSAID respectively, in a double-blind, randomised, three way, cross-over block design. In each case, a three week washout control phase was followed by active treatment phases of two weeks each with three week washout between treatment phases.

Faecal blood loss, collected over 72 hours at the end of each treatment phase, was measured by 51Cr-labelled erythrocyte method. Comparison was made of mean 24 hour faecal blood loss with each treatment compared with control using repeated measures analysis of variance. Eighteen patients completed all phases of the study. Three patients were withdrawn due to GI bleeding; two during diclofenac treatment and one during treatment with piroxicam. Mean 24 hour faecal blood loss with diclofenac (0.53 mL +/- 0.21) was not significantly different from control (0.28 mL +/- 0.06), whereas it was significantly increased with naproxen (2.76 mL +/- 2.22) and piroxicam (1.16 mL +/- 0.62), p = 0.0013.

A short half-life NSAID was associated with lower GI toxicity than a medium and long half-life NSAID, as measured by faecal blood loss.