Search

Background: Inhibition of constitutively expressed cyclooxygenase (Cox-1) is thought to play an important role in the gastrointestinal toxicity of nonsteroidal anti-inflammatory drugs (NSAID), while their therapeutic action may be due to inhibition of the enzyme Cox-2, which is specifically expressed at sites of inflammation. NSAIDs with high affinity and specifity for Cox-2 hold the promise of maintaining efficacy without the gastrointestinal side effects of conventional NSAIDs., Methods: We assessed the gastrointestinal tolerability of flosulide (20 mg twice a day), a highly selective Cox-2 inhibitor with that of naproxen (500 mg twice a day), which has equal affinity for Cox-1 and -2 in 19 patients with osteoarthrosis in a randomized, double blind, crossover endoscopy study. Subjects were treated for 2 weeks with a 2-week washout period. Gastroduodenal damage was primarily assessed as by Lanza (grades 0-4)., Results: No stomach damage was seen in 13 (68%) patients after flosulide and in 5 (37%) after naproxen (P < 0.001). Lanza scores were significantly lower after flosulide (0.58) than after naproxen (1.47) (P < 0.001; odds ratio, 84.4; 95% confidence interval, 1.45-4908). Flosulide was significantly better tolerated (P < 0.005) than naproxen., Conclusion: These results endorse the idea that highly selective Cox-2 inhibitors may be associated with lesser gastrointestinal side effects than conventional NSAIDs.

A number of studies show that the idea that inhibition of cyclooxygenase is the sole mechanism of NSAID-induced gastrointestinal damage is no longer tenable. We re-examined various aspects of the mechanism of small intestinal damage due to NSAIDs in rat. Subcellular organelle marker enzyme studies show selective alterations in mitochondrial and brush border marker enzymes. Electron microscopy shows changes compatible with uncoupling of mitochondrial oxidative phosphorylation. In vitro, all common acidic-NSAIDs (n = 15) were found to uncouple oxidative phosphorylation at concentrations (microM) easily achievable within intestinal epithelium. Experiments in bile duct ligated animals show that intact indomethacin within the gastrointestinal lumen is required for uncoupling. Relative importance and pathophysiological consequences of uncoupling and inhibition of cyclooxygenase were assessed following administration of R and S flurbiprofen: the former selectively uncouples whilst the latter is also an effective cyclooxygenase inhibitor. R flurbiprofen uncoupled in vitro and in vivo, increased intestinal permeability and caused mild intestinal inflammation, but had not significant effect on prostanoid levels and produced no ulcers. S flurbiprofen uncoupled and increased intestinal permeability equally but was associated with significant decreases in intestinal prostanoid levels, more inflammation and numerous ulcers. Collectively these studies suggest that uncoupling may underlie the "topical" phase of NSAID damage which leads to increased intestinal permeability and inflammation, but concomitant inhibition of cyclooxygenase is essential to drive the inflammation to ulcers.

Thirteen juvenile chronic arthritis patients with abdominal symptoms related to non-steroidal anti-inflammatory drug therapy were endoscoped before and after a 6-week course of either misoprostol or ranitidine therapy. Major presenting symptoms were generalized abdominal pain and nausea. Symptoms did not correlate well with endoscopic findings which revealed no evidence of ulceration and minimal erosive damage. Five patients had mild erythema or gastritis. Bleeding lesions were confined to small numbers of petechiae. Following treatment with either misoprostol or ranitidine, patients improved symptomatically without a corresponding improvement on endoscopic and histological examination of stomach and duodenum. Both treatments were well tolerated.

There is no doubt that nonsteroidal anti-inflammatory drugs (NSAIDs) cause gastrointestinal injury. The most serious consequences are gastric and duodenal ulcers which can cause bleeding and perforation, and which may lead to the premature death of 3000 to 4000 patients in the UK annually. The immediate actions of NSAIDs operate at a subcellular level; in particular altering of mitochondrial function which causes depletion of ATP and renders the cell vulnerable to oxidant stress. Secondary consequences follow, such as the inhibition of prostaglandin synthesis which delays cellular repair. While adaptation can be shown in volunteers despite continued NSAID ingestion, studies in patients suggest mucosal damage develops continuously and cumulatively even with low doses of aspirin. Histamine H2-receptor antagonists and proton pump inhibitors heal NSAID-related ulcers, though healing rates with H2-antagonists are slower in patients who continue NSAID treatment. They have little role in preventing damage. In addition to acid suppression, prostaglandin analogues cause bicarbonate secretion and enhance mucosal blood flow. They have a specific role in both prevention and treatment of NSAID-related damage. The use of misoprostol offers a rational approach to reduce the high prevalence of unwanted gastroduodenal damage from NSAIDs. On a purely financial basis more information is needed before routine coprescribing can be recommended. However, for any patient on NSAIDs with a previous ulcer or for patients aged over 60 years (where the risks and seriousness of complications are markedly increased), the use of misoprostol should be considered. Further developments in prostaglandin analogues may reduce their adverse effects and perhaps thereby improve their efficacy at symptom control.