- Email: [email protected]
Drug-induced gastrointestinal disorders
STOMACH
Drug-induced gastrointestinal disorders
principally by inhibiting prostaglandin synthesis. In other cases, impairment of defence mechanisms is less direct; for example, there is a greater incidence of Salmonella and Campylobacter enteritis with proton pump inhibitors because the main physiological role of stomach acid is to sterilize ingested food before passage to the intestine.1 Direct injury to the GI tract e examples include oesophageal damage by potassium preparations, gastroduodenal ulcers caused by cytotoxic drugs, and the ulceration and colitis associated with NSAIDs independent of prostaglandin inhibition. Changes in colonic bacterial flora e widespread use of antibiotics, particularly cephalosporins, is associated with an increased incidence of Clostridium difficile infection and pseudomembranous colitis. However, there may be overlap between the groups (e.g. aspirin exhibits topical toxicity and inhibits prostaglandin-dependent mucosal defences).
Anthony Norman Christopher J Hawkey
Abstract Adverse drug effects on the gastrointestinal (GI) tract can occur as a predictable result of a drug’s mode of action, by direct injury, through compromising GI defences, or as a consequence of changes in colonic bacterial flora. Non-steroidal anti-inflammatory drugs (NSAIDs) are the commonest cause of gastroduodenal injury, owing to inhibition of prostaglandin synthesis, and increase the risk of serious GI complications between 2.5- and 5-fold. There is considerable variation in risk across the recommended dose ranges for individual NSAIDs. Low doses of aspirin are associated with an increased risk of upper GI haemorrhage. COX-2 inhibitors largely spare the GI mucosa from injury, unless they are co-administered with aspirin when the GI safety benefits are abrogated. As the small intestine becomes more accessible to imaging modalities such as capsule endoscopy, small bowel enteropathy is suggested by hypoalbuminaemia and iron deficiency. Drug-induced colitis is an important problem, with antibiotics the commonest drug cause. Drugs can also exacerbate pre-existing inflammatory bowel disease. Strategies to minimize the adverse GI effects of drugs include giving the lowest dose of NSAID for the shortest time, selective COX-2 inhibitors for high-GI/ low-cardiovascular risk patients, and upper GI mucosal protection by co-prescription of NSAIDs with proton pump inhibitors.
Oesophageal disease Heartburn Drugs commonly associated with heartburn and/or oesophageal injury are listed in Table 1. Drugs that relax the lower oesophageal sphincter are particularly likely to cause problems because of the increased acid exposure of the oesophagus (often without mucosal injury). Patients should be managed by drug withdrawal when possible, though a few may require acid suppression. Endoscopy is not usually required if the onset is clearly related to the drug and the condition settles on drug cessation. Candida oesophagitis is a common cause of dyspepsia (heartburn) or odynophagia (painful swallowing) in patients taking corticosteroids, antibiotics or immunosuppressive drugs. The characteristic change on endoscopy is the formation of discrete white plaques.
Keywords adverse events; drugs; gastrointestinal injury; large bowel; NSAIDs; small intestine
Mucosal injury and strictures Tetracyclines, potassium chloride, certain bisphosphonates and quinidine are recognized as toxic to the oesophageal mucosa.
Almost any drug can affect the gastrointestinal (GI) tract. It has been estimated that 40% of adverse drug reactions affect the GI tract, and 4% of admissions to hospital are caused by druginduced disorders. This contribution highlights drug-induced GI disorders that are common, important or often missed. It is useful to consider drug effects on the GI tract in the four categories below. Consequences of pharmacological mode of action e adverse drug effects can occur as a predictable result of a drug’s mode of action. Thus, drugs with anticholinergic effects (e.g. antidepressants) reduce oesophageal sphincter pressure, resulting in reflux and heartburn; they also reduce colonic transit, resulting in constipation. Impairment of GI defences e the best-known examples are non-steroidal anti-inflammatory drugs (NSAIDs), which act
Common drug causes of oesophageal problems Relax lower oesophageal sphincter causing heartburn C Anticholinergic agents (e.g. procyclidine, trihexyphenidyl) C Tricyclic antidepressants C Calcium channel blockers C Nitrates C Phenothiazines Direct mucosal injury C Tetracyclines C Bisphosphonates Associated with strictures C Potassium chloridea C Quinidinea C NSAIDs
Anthony Norman MRCP is SpR in Gastroenterology at University Hospital, Nottingham, UK. Competing interests: none declared. Christopher J Hawkey DM FRCP is Professor of Gastroenterology at University Hospital, Nottingham, UK. Competing interests: none declared.
MEDICINE 39:3
a
Particularly modified-release preparations.
Table 1
162
Ó 2010 Published by Elsevier Ltd.
STOMACH
However, presentation may vary with different drugs. NSAIDs and tetracyclines are a relatively common cause of erosive or pill oesophagitis; potassium chloride and quinidine are more strongly associated with stricture, often in the mid-oesophagus. These problems are more common in elderly patients, particularly those with pre-existing oesophageal disease or massive left atrial enlargement, and are more likely to occur when patients swallow their pills in a reclining position. Sustained-release preparations of potassium and quinidine are particularly likely to cause problems. There is an association between bisphosphonates and reversible oesophagitis, but the risk is low if the medication is administered correctly (i.e. in an upright position).2
low doses for cardiovascular prophylaxis), this drug is now associated with as many ulcer bleeds as all other NSAIDs together. Overall, patients taking NSAIDs probably undergo hospitalization for gastrointestinal problems about once every 50e100 patientyears. In the UK, it is likely that at least 2000 patients per year die as a result of NSAID use (because of GI complications).9 Risk factors Patient-associated risk factors: concomitant risk factors are age and previous gastrointestinal bleeding.3,10e13 Drug-related risk factors: the main drug-related risk factors are dose, choice of NSAID, co-prescription of corticosteroids and coprescription of warfarin. The risk of ulcer complications increases with dose. A meta-analysis by Lewis et al. found significant differences in the odds ratio for upper gastrointestinal bleeding, with an increased risk from ibuprofen (odds ratio 1.7) to ketoprofen (34.9).14 In addition, there was an up to 8-fold variation in risk across the recommended dose ranges for each drug. Indometacin <1200 mg/day was no more harmful than paracetamol, but at doses >1800 mg/day, it had an odds ratio of 4.6. As a general rule, the GI toxicity of NSAIDs correlates with the degree of anti-inflammatory activity, so that those with high analgesic/low anti-inflammatory activity (e.g. ibuprofen) are less ulcerogenic than those with relatively high anti-inflammatory: analgesic properties (e.g. piroxicam).15 The GI risks with aspirin are becoming a major cause of ulcer complications as its use becomes more widespread. Furthermore, the combination of NSAIDs and aspirin has a multiplicative rather than a simple additive effect on risk. Even the low doses of aspirin widely used to prevent cardiovascular events are associated with an increased risk of upper GI haemorrhage, with a daily dose of 75 mg having an odds ratio of around 2, and a daily dose of 300 mg having an odds ratio of almost 4.16 At least some of this is attributable to the anti-haemostatic effects of aspirin. Clopidogrel is also associated with an increased incidence of upper GI haemorrhage. Co-prescription of clopidogrel and aspirin in patients with ischaemic heart disease further increases the risk of gastrointestinal bleeding by a factor of 15.17
Dyspepsia and gastroduodenal ulcers NSAIDs About one-third of patients take NSAIDs experience dyspepsia. In most cases, this occurs in the absence of ulceration. Some studies suggest that NSAID-induced dyspepsia is particularly common in patients infected with Helicobacter pylori. NSAIDs are the commonest cause of gastric and duodenal ulcers (Figure 1). Studies in which patients taking NSAIDs have undergone endoscopy regardless of symptoms have shown a high incidence of ulceration (about 20%). It seems likely, however, that some of these ulcers are relatively trivial or transient, or are superficial erosions mistakenly classified as ulcers. NSAIDs cause ulcers because of inhibition of prostaglandin synthesis. Diagnostic problems e many NSAID-associated ulcers are silent, and NSAIDs often cause dyspepsia without ulceration. Dyspepsia is thus a poor guide when deciding whether to perform endoscopy. Estimation of risk e the use of traditional NSAIDs increases the risk of serious GI complications between 2.5 and 5-fold.3e5 Recent studies have shown that the cumulative incidence of upper GI complications is between 0.92% and 1.4% after 12 months of NSAID use.6e8 With the increasing use of aspirin (particularly in
Cyclo-oxygenase (COX)-2 inhibitors Selective COX-2 inhibitors were heralded as ‘safer NSAIDs’ because they carried the anti-inflammatory benefits of COX-2 inhibition whilst retaining the gastroprotective benefits of COX-1. There have been four large outcome studies of rofecoxib (VIGOR study7), celecoxib (CLASS study6), lumiracoxib (TARGET study18) and etorecoxib (MEDAL programme19). TARGET showed a five-fold reduction in ulcer complications to very low levels that were similar to what is seen in the background population in patients not taking aspirin. All studies have similarly shown reductions in complicated and uncomplicated ulcers or in close surrogate measures to a greater or lesser extent in non-users of aspirin. In all studies aspirin use was associated with a loss of most, if not all, of the benefits of the COX-2 inhibitor. Proton pump inhibitors An important recent development has been the recommendation by NICE that all patients taking an NSAID or a COX-2 inhibitor should be given ulcer prophylaxis with a proton pump inhibitor
Figure 1 An NSAID-induced gastric ulcer with active bleeding. Courtesy of Clinical gastroenterology and hepatology, Mosby, 2005.2
MEDICINE 39:3
163
Ó 2010 Published by Elsevier Ltd.
STOMACH
(PPI).20 This advice reflects the marked fall in the cost of omeprazole (£2.04 for omeprazole 20 mg for 28 days). This is based largely on primary prophylaxis endoscopic studies and secondary prevention studies that have used ulcer bleeding as an endpoint. For those at particularly high risk of GI toxicity, there is evidence to promote the combination of COX-2 inhibitors with PPI therapy as this markedly reduces the incidence of recurrent GI ulcer bleeding.21 There is current concern that at least some PPIs may inhibit the activation of clopidogrel22 and lead to an increased incidence of thrombotic events.23 Giving the PPI in the morning and clopidogrel at night may avoid the problem.24 Prescribers should use the most up-to-date recommendations. H. pylori e the role of H. pylori in drug-induced GI ulceration is complex and there are few systematic trial data. Some studies have shown that eradication of H. pylori before NSAID use reduces the risks of endoscopic ulcers.25 Others have shown no difference26 or that those with continuing H. pylori infection have better rather than worse outcomes when acid-suppressing drugs are used.27
pivampicillin have been associated with peptic ulceration, though the evidence for the association is often informal. Recently, calcium channel antagonists have been implicated in peptic ulcer bleeding. Corticosteroids e it now seems likely that corticosteroids do not increase the risk of ulceration, except in patients taking NSAIDs.
Nausea and vomiting Nausea and vomiting are common adverse drug reactions that may occur as a direct effect of the drug or as a result of gut damage, including ulceration. The more common drug causes are listed in Table 2. Many drugs (e.g. levodopa, opioids, cisplatin, digoxin) induce symptoms by directly stimulating the chemoreceptor trigger zone in the area postrema (located on the dorsal surface of the medulla oblongata at the caudal end of the fourth ventricle). Chemotherapeutic agents probably cause nausea and vomiting by stimulating serotonin release from entero-chromaffin cells within the gut, which activates the vagus. Drugs such as potassium chloride, iron preparations and NSAIDs probably act via direct gastric irritation. In a few patients, these symptoms are accompanied by gastric or duodenal ulceration. Dopamine receptors in the area postrema are the target for anti-emetic drugs like chlorpromazine, haloperidol, domperidone and metoclopramide. Nausea is a challenging symptom to investigate and no drug should be judged blameless until the effect of stopping it has been assessed. Use of several drugs at one time, particularly in elderly patients, can cause nausea, and it is important to simplify drug regimens in such patients.
Recommendations for prophylaxis Patients with one or more risk factors for NSAID associated ulcers (80% of patients) should be advised to take preventive treatment, although this is often neglected. Many patients are given H2-receptor antagonists to protect them against NSAID damage, but there is good evidence to suggest that at conventional dosage these agents prevent only duodenal ulcer. Proton pump inhibitors (PPIs) have been shown to prevent aspirin- and NSAID-induced small upper gastrointestinal injury. There remains some concern over their use in patients treated with clopidogrel, because of abrogation of the anti-platelet effect. Misoprostol is as effective as PPIs but is poorly tolerated.28 COX-2 inhibitors reduce the risk of GI events compared to traditional NSAIDs. This benefit is lost in aspirin users. However, unlike NSAIDs, COX-2 inhibitors do not interfere with aspirin’s ability to inhibit platelet aggregation. Although there is concern about the effects of COX-2 inhibitors (and NSAIDs) on cardiovascular outcomes,29 a combination of a COX-2 inhibitor with aspirin may prove to be a more reliable anti-haemostatic strategy than a combination with a non-selective NSAID. For those at high risk of GI ulceration, addition of a PPI to a COX-2 inhibitor offers even greater GI protection.
Pancreatitis Alcohol is a common cause of pancreatitis. Other drug-related causes are uncommon, but the main culprits are: diuretics (thiazides, furosemide) sulphonamides azathioprine 6-mercaptopurine corticosteroids
Common drug causes of nausea and vomiting Locally irritant C Potassium chloride C Iron preparations C NSAIDs C Theophyllines C Azathioprine (?mechanism) C Metronidazole
Management NSAID-induced ulcers have been shown to heal with all standard anti-ulcer drug regimens. Healing is faster if the NSAID is stopped. Comparative studies show that omeprazole, 20 mg daily, achieves faster healing in gastric and duodenal ulcers than ranitidine, 150 mg twice daily, or misoprostol, 200 mg four times daily.30
Act via the CNS C Levodopa C Bromocriptine C Opioids C Digoxin C Chemotherapeutic agents (e.g. cisplatin, mustine, dacarbazine, cyclophosphamide)
Other causes It should be remembered that dyspepsia occurs with the nonaspirin salicylates (e.g. mesalazine) used in inflammatory bowel disease. For most drugs, the precise mechanism by which dyspepsia arises is unknown, though there are a few exceptions (e.g. erythromycin, which causes epigastric pain and general cramp as a consequence of its prokinetic activity as a motilin receptor agonist). Drugs such as cytotoxic agents, oral gold and
MEDICINE 39:3
Table 2
164
Ó 2010 Published by Elsevier Ltd.
STOMACH
NSAIDs sodium valproate exenatide and liraglutide L-asparaginase (about 7% of recipients).
Common drug causes of constipation and diarrhoea Constipation C Drugs with antimuscarinic effects (e.g. atropine, phenothiazines, tricyclic antidepressants) C Opioids C Mebeverine C Peppermint oil C Aluminium-containing antacids C Sucralfate C Gaviscon C Iron C Laxatives (chronic) Diarrhoea C b-blockers C Misoprostol C Antibiotics C Magnesium-containing antacids C Olsalazine C Mefenamic acid C Iron C Laxatives (acute) C Metformin C Angiotension-converting enzyme inhibitors C Statins
Constipation Opioids and drugs with antimuscarinic effects are the main causes of drug-induced constipation (Table 4). An interesting development is the use of peripheral opioid antagonist, naltrexone, to prevent constipation in patients in palliative care who are receiving opioids. Reduced bowel frequency also occurs with selective serotonin reuptake inhibitors (e.g. fluoxetine) and locally acting antispasmodics (e.g. mebeverine, peppermint oil). Codeine-containing analgesics are available over the counter, and some patients establish a selfdefeating cycle, using them for constipation-induced abdominal pain. Aluminium-containing antacids, sucralfate and Gaviscon also commonly constipate, as does iron (though this can also cause diarrhoea). Drug-induced megacolon from enteric neurotoxicity caused by vincristine or (possibly) laxatives is an uncommon but important cause of drug-induced constipation. Although fibre (dietary and prescribed) is used to treat constipation, it can cause pain, bloating and, occasionally, impaction if excessive amounts are ingested unaccompanied by adequate volumes of liquid. Insoluble dietary fibre is often used to treat constipation but it can cause pain, bloating and occasional impaction if excessive amounts are ingested unaccompanied by adequate volumes of liquid. Soluble fibre (porridge, bananas, isphagula husk) is better tolerated by many patients.
Table 3
Deliberate use of laxatives should always be considered in patients with diarrhoea, particularly when potassium is low. All salicylates can cause diarrhoea, but it is particularly likely with olsalazine, which has been shown to stimulate small intestinal secretion. Azathioprine can occasionally cause diarrhoea, sometimes in the context of toxic shock-like syndrome. This may cause diagnostic confusion when the drug is used for inflammatory bowel disease. It is increasingly recognized that drugs that suppress acid production (H2-antagonists and PPIs) predispose to enteric superinfection.
Melanosis coli Anthranoid laxatives (e.g. senna) are widely used laxatives of natural origin that pass unabsorbed to the large bowel, where they are metabolized to aglycones. These aglycones exert their laxative effect by damaging epithelial cells, which leads directly and indirectly to changes in absorption, secretion and motility. Damaged epithelial cells can be found as apoptotic bodies in the pigmented colonic mucosa, characteristic of pseudomelanosis coli. The belief that laxatives cause enteric neuronal damage leading to chronic megacolon is probably misplaced.
Malabsorption Diarrhoea
Many drugs (e.g. colchicine, methotrexate, laxatives) can cause malabsorption, though this is usually of little clinical significance. Sulfasalazine and phenytoin commonly cause folate malabsorption.
Drug-induced diarrhoea (Table 3) is common. The main causes are: antibiotics, which disturb the colonic flora and tend to select C. difficile (responsible for a growing epidemic of pseudomembranous colitis in hospitalized patients) b-blockers, which act by antagonizing antiperistaltic adrenergic stimulation bile acids, which have a direct irritant action in the colon bisoprostol, which stimulates intestinal secretion and motility. Other causes include the following. Magnesium-containing antacids (available over the counter); in a recent study, they accounted for 4% of patients referred for evaluation of diarrhoea. Generally, this was not detected in the initial history.
MEDICINE 39:3
Small intestinal injury The small intestine may be a more common site for NSAID toxicity than the well-recognized effects on the stomach and duodenum, although these effects may be more occult, and usually arise only after chronic (e.g. 6 months) use. An estimated 70% of NSAID users demonstrate intestinal inflammation when indirectly measured, for example by scintigraphy.31 Small bowel enteropathy, a combination of increased intestinal permeability and mild mucosal inflammation, is suggested by hypoalbuminaemia and iron deficiency. Studies of long-term NSAID users suggest 60e70% have an asymptomatic enteropathy.31,32
165
Ó 2010 Published by Elsevier Ltd.
STOMACH
Capsule endoscopy has allowed the demonstration that NSAIDs cause mucosal erosions in most subjects,33 whilst COX-2 inhibitors have a much reduced or absent effect.7,34 This may contribute to the lower incidence of anaemia seen in patients taking COX-2 inhibitors (Figure 2).
Drugs that cause colitis Antibiotics Amoxicillin Ampicillin Clindamycin Erythromycin Cephalosporins
Colonic ulcers NSAIDs are also recognized to cause sporadic colonic ulcers. Increasing use of nicorandil, which has long been associated with oral and anal ulceration, has led to recognition that it also causes colonic ulceration.35,36 The reversibility of the injury on cessation of therapy is well documented and the association is not in doubt. The mechanism, however, is yet to be established. In terms of clinical manifestations, the patient may develop anaemia, diarrhoea or weight loss, and even ileal perforation has been reported.37
Drugs causing ischaemic colitis Oral contraceptive pill Chemotherapeutic agents e 5-fluorouracil, cisplatin danazol Vasopressin Clindamycin Miscellaneous NSAIDs Methyldopa Penicillamine Gold (oral)
Colitis The most common drug-related causes of colitis are shown in Table 4. Mefenamic acid directly damages epithelial cells; clinical presentation ranges from simple reversible secretory diarrhoea to colitis. Other NSAIDs seldom cause diarrhoea and none has been conclusively shown to be a primary cause of colitis, though they may provoke relapse. Collagenous colitis is a rare condition that may be associated with NSAID use, but this may be a consequence rather than a cause because patients with collagenous colitis exhibit a greater incidence of arthritis, which may require NSAID administration. Antibiotics commonly cause loose stools or diarrhoea, and patients may present with severe pseudo-membranous colitis (see Figure 3) as a result of C. difficile selection.
Table 4
also been implicated in the exacerbation of inflammatory bowel disease.38 The mechanism is unknown. It is increasingly recognized that NSAIDs cause lower gastrointestinal tract problems, including erosions, ulcers and strictures. Presentations include blood and protein loss, perforation and obstruction. Minimizing the adverse gastrointestinal effects of drugs Drugs can get stuck in the oesophagus, particularly when in gelatine capsules; they should be taken with plenty of fluid. NSAID-associated gastroduodenal toxicity varies with drug and dosage; the best advice is to stop the NSAID whenever possible, minimize the dosage, or use ibuprofen, 1200 mg daily, or a COX-2 inhibitor.
Exacerbation of ulcerative colitis There is evidence that analgesic drugs (NSAIDs and paracetamol) provoke relapse of colitis in some patients. COX-2 drugs have
Figure 3 Appearance of pseudomembrane in sigmoid colon after broadspectrum antibiotics (cephalosporin), with proven Clostridium difficile toxin on stool testing. Courtesy of Clinical gastroenterology and hepatology, Mosby, 2005.2
Figure 2 Typical small bowel erosions in the terminal ileum of a healthy volunteer taking NSAIDs, with red halo resulting from hyperaemia and pale centre.
MEDICINE 39:3
166
Ó 2010 Published by Elsevier Ltd.
STOMACH
NICE recommends that COX-2 inhibitors be reserved for patients at ‘high risk’ of GI side-effects; an alternative might be to use COX-2 inhibitors in ‘average-risk’ patients and consider NSAIDs plus a PPI, or even a coxib plus a PPI in high-risk patients, based on individual circumstances.39 Avoiding modified-release preparations may reduce small and large bowel problems associated with NSAIDs and iron preparations. In patients with inflammatory bowel disease, the possibility of drug-induced diarrhoea caused by olsalazine or azathioprine, and disease relapse caused by paracetamol or NSAIDs, requires careful consideration. Diarrhoea caused by b-blockers is common and often not appreciated by prescribers. Prescribers of H2-antagonists and proton pump inhibitors should be aware of the possibility of enteric infections. A
12 Shorr RI, Daugherty JR, Griffin MR. concurrent use of nonsteroidal anti-inflammatory drugs and oral anticoagulants places elderly persons at high risk for haemorrhagic peptic ulcer disease. Arch Intern Med 1993; 153: 1665e70. 13 Mellemkjaer L, Sorensen HT, Thomassen L, et al. Upper gastrointestinal bleeding among users of NSAIDs: a population-based cohort study in Denmark. Br J Clin Pharmacol 2002; 53: 173e81. 14 Lewis SC, Laporte J, Matthews JN, Rawlins MD, Wiholm B. Doseresponse relationships between individual nonaspirin nonsteroidal antiinflammatory drugs (NANSAIDs) and serious upper gastrointestinal bleeding: a meta-analysis based on individual patient data. Br J Clin Pharmacol 2002; 54: 320. 15 Graham DY. Nonsteroidal anti-inflammatory drugs, Helicobacter pylori, and ulcers: where we stand. Am J Gastroenterol 1996; 91: 2080e6. 16 Tashima K, Fujita A, Umeda M, Takeuchi K. Lack of gastric toxicity of nitric oxide-releasing aspirin, NcX-4016, in the stomach of diabetic rats. Life Sci 2000; 67: 1639e52. 17 Lanas A, Garcia-Rodriguez LA, Arroyo MP, et al. Risk of upper gastrointestinal bleeding associated with selective cyclo-oxygenase2 inhibitors, traditional non-steroidal non-aspirin anti-inflammatory drugs, aspirin and combinations. Gut 2006; 55: 1731e8. 18 Schnitzer TJ, Burmester GR, Mysler E, et al. Comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), reduction in ulcer complications: randomised controlled trial. Lancet 2004; 364: 665e74. 19 Cannon CP, Curtis SP, FitzGerald GA, et al. Cardiovascular outcomes with etoricoxib and diclofenac in patients with osteoarthritis and rheumatoid arthritis in the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) programme: a randomised comparison. Lancet 2006; 368: 1771e81. 20 Dyspepsia: managing dyspepsia in adults in primary care. NICE guideline CG17, 2004. 21 Chan FKL, Wang VWS, Suen BY, et al. Combination of a cyclo-oxygenase-2 inhibitor and a proton-pump inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: a double-blind, randomised trial. The Lancet 2007; 369: 1621e6. 22 Gilard M, Arnaud B, Cornily JC, et al. Influence of omeprazole on the antiplatelet action of clopidogrel associated with aspirin: a randomised, double-blind OCLA (Omeprazole CLopidogrel Aspirin) study. J Am Coll Cardiol 2008; 51: 256e60. 23 Ho MP, Maddox L, Wang L, et al. Risk of adverse outcomes associated with concomitant use of clopidogrel and proton pump inhibitors following acute coronary syndrome. JAMA 2008; 301: 937e44. 24 Bhatt DL. COGENT: a prospective, randomized, placebo-controlled trial of omeprazole in patients receiving aspirin and clopidogrel. Presented at: Transcatheter Cardiovascular Therapeutics, 2009. 25 Chan FK, Sung JJ, Chung SC, et al. Randomised trial of eradication of Helicobacter pylori before non-steroidal anti-inflammatory drug therapy to prevent peptic ulcers. Lancet 1997; 350: 975e9. 26 Hawkey C, Tulassay Z, Szczepanski L, et al. Randomised controlled trial of Helicobacter pylori eradication in patients on non-steroidal antiinflammatory drugs: HELP NSAIDs study. Lancet 1998; 352: 1016e21. 27 Bianchi Porro G, Parente F, Imbesi V, Montrone F, Caruso I. Role of Helicobacter pylori in ulcer healing and recurrence of gastric and duodenal ulcers in longterm NSAID users. Response to omeprazole dual therapy. Gut 1996; 39: 22e6. 28 Graham DY, Agrawal NM, Campbell DR, et al. Ulcer prevention in longterm users of nonsteroidal anti-inflammatory drugs: results of a double-blind, randomized, multicenter, active- and placebo-
REFERENCES 1 Neal KR, Scott HM, Slack RC, Logan RF. Omeprazole as a risk factor for campylobacter gastroenteritis: case-control study. BMJ 1996; 312: 414e5. 2 Cryer B, Bauer DC. Oral bisphosphonates and upper gastrointestinal tract problems: what is the evidence? Mayo Clin Proc 2002; 77: 1031e43. 3 Hernandez-Diaz S, Rodriguez LA. Association between nonsteroidal anti-inflammatory drugs and upper gastrointestinal tract bleeding! perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med 2000; 160: 2093e9. 4 Derry S. Risk of gastrointestinal haemorrhage with long term use of aspirin: meta-analysis. BMJ 2000; 321: 1183e7. 5 Garcia Rodriguez LA, Hernandez-Diaz S, De Abajo J. Association between aspirin and upper gastrointestinal complications: systematic review of epidemilogic studies. Br J Clin Pharmacol 2001; 52: 563e71. 6 Silverstein FE, Goldstein JL, Simon LS, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the class study: a randomized controlled trial. Celecoxib long-term arthritis safety study. JAMA 2000; 284: 1247e55. 7 Bombardier C, Reicin A, Shapiro D, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGoR Study Group. N Engl J Med 2000; 343: 1520e8. 8 Schnitzer TJ, Mysler E, Hochberg MC, et al. comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), reduction in ulcer complications: randomised controlled trial. Lancet 2004; 364: 665e74. 9 Blower AL, Brooks A, Fenn GC, et al. Emergency admissions for upper gastrointestinal disease and their relation to NSAID use. Aliment Pharmacol Ther 1997; 11: 283e91. 10 Laine L, Bombardier C, Hawkey CJ, et al. Stratifying the risk of NSAIDrelated upper gastrointestinal clinical events: results of a doubleblind outcomes study in patients with rheumatoid arthritis. Gastroenterology 2002; 123: 1006e12. 11 Gabriel SE, Jaakkimainen L, Bombardier C. Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. A meta-analysis. Arch Intern Med 1991; 115: 787e96.
MEDICINE 39:3
167
Ó 2010 Published by Elsevier Ltd.
STOMACH
29
30
31
32 33
34
35
36 Titi MA, Seow C, Molloy RG. Nicorandil-induced colonic ulceration: a new cause of colonic ulceration. Report of four cases. Dis Colon Rectum 2008 Oct; 51: 1570e3. 37 King PM, Suttie SA, Jansen JO, Watson AJM. Perforation of the terminal ileum: apossible complication of nicorandil therapy. Surg J R Coll Surg Edinb Ire 2004; 2: 56e7. 38 Matuk R, Crawford J, Abreu MT, Targan SR, Vasilauskas EA, Papadakis KA. The spectrum of gastrointestinal toxicity and effect on disease activity of selective cyclo-oxygenase-2 inhibitors in patients with inflammatory bowel disease. Inflamm Bowel Dis 2004; 10: 352e6. 39 National Institute for Health and clinical Excellence. Guidance on the use of cyclo-oxygenase (cox) II selective inhibitors, celecoxib, rofecoxib, meloxicam and etodolac for oA and RA. Technology Appraisal Guidance No. 27, 2001.
controlled study of misoprostol vs lansoprazole. Arch Intern Med 2002; 162: 169e75. Bresalier RS, Sandler RS, Quan H, et al. cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med 2005; 352: 1092e102. Yeomans N. New data on healing of nonsteroidal anti-inflammatory drug-associated ulcers and erosions. omeprazole NSAID Steering committee. Am J Med 1998; 104: 79Se80. Bjarnason I, Hayllar J, MacPherson AJ, Russell AS. Side effects of nonsteroidal anti-inflammatory drugs on the small and large intestine in humans. Gastroenterology 1993; 104: 1832e47. Bjarnason I, Smith T, Prouse P, Williams P, Smethurst P. NSAID drug induced inflammation in humans. Gastroenterology 1987; 93: 480e9. mez-Rodrı´guez BJ, Romero-Vazquez J, et al. Caunedo-Alvarez A, Go Macroscopic small bowel mucosal injury caused by chronic consume of NSAID assessed by capsule endoscopy. Rev Esp Enferm Dig 2010; 102: 80e5. Chan FK, Scheiman JM, Lanas A, Berger M, Li MC, Goldstein JL. Celecoxib versus Diclofenac and Omeprazole: a randomized controlled trial comparing a composite outcome across the entire GI tract (The Condor Trial). AGA 2010; S20: 110 (Abst). Brown R, Lee A, Welfare M. Nicorandil-induced colonic ulceration. Heart 2008 May; 94: 678.
MEDICINE 39:3
FURTHER READING Fortun PJ, Hawkey CJ. Non-steroidal anti-inflammatory drugs and the small intestine. Curr Opin Gastroenterol 2005; 21: 169e75. Gilman AG, Roll TW, Nies AS, et al. The pharmacological basis of therapeutics. New York: McGraw-Hill, 1992. Rostom A, Dube C, Wells G, et al. Prevention of NSAID-induced gastroduodenal ulcers (cochrane Review). Cochrane Database Syst Rev 2002; (Issue 4).
168
Ó 2010 Published by Elsevier Ltd.
Drug-induced gastrointestinal disorders
principally by inhibiting prostaglandin synthesis. In other cases, impairment of defence mechanisms is less direct; for example, there is a greater incidence of Salmonella and Campylobacter enteritis with proton pump inhibitors because the main physiological role of stomach acid is to sterilize ingested food before passage to the intestine.1 Direct injury to the GI tract e examples include oesophageal damage by potassium preparations, gastroduodenal ulcers caused by cytotoxic drugs, and the ulceration and colitis associated with NSAIDs independent of prostaglandin inhibition. Changes in colonic bacterial flora e widespread use of antibiotics, particularly cephalosporins, is associated with an increased incidence of Clostridium difficile infection and pseudomembranous colitis. However, there may be overlap between the groups (e.g. aspirin exhibits topical toxicity and inhibits prostaglandin-dependent mucosal defences).
Anthony Norman Christopher J Hawkey
Abstract Adverse drug effects on the gastrointestinal (GI) tract can occur as a predictable result of a drug’s mode of action, by direct injury, through compromising GI defences, or as a consequence of changes in colonic bacterial flora. Non-steroidal anti-inflammatory drugs (NSAIDs) are the commonest cause of gastroduodenal injury, owing to inhibition of prostaglandin synthesis, and increase the risk of serious GI complications between 2.5- and 5-fold. There is considerable variation in risk across the recommended dose ranges for individual NSAIDs. Low doses of aspirin are associated with an increased risk of upper GI haemorrhage. COX-2 inhibitors largely spare the GI mucosa from injury, unless they are co-administered with aspirin when the GI safety benefits are abrogated. As the small intestine becomes more accessible to imaging modalities such as capsule endoscopy, small bowel enteropathy is suggested by hypoalbuminaemia and iron deficiency. Drug-induced colitis is an important problem, with antibiotics the commonest drug cause. Drugs can also exacerbate pre-existing inflammatory bowel disease. Strategies to minimize the adverse GI effects of drugs include giving the lowest dose of NSAID for the shortest time, selective COX-2 inhibitors for high-GI/ low-cardiovascular risk patients, and upper GI mucosal protection by co-prescription of NSAIDs with proton pump inhibitors.
Oesophageal disease Heartburn Drugs commonly associated with heartburn and/or oesophageal injury are listed in Table 1. Drugs that relax the lower oesophageal sphincter are particularly likely to cause problems because of the increased acid exposure of the oesophagus (often without mucosal injury). Patients should be managed by drug withdrawal when possible, though a few may require acid suppression. Endoscopy is not usually required if the onset is clearly related to the drug and the condition settles on drug cessation. Candida oesophagitis is a common cause of dyspepsia (heartburn) or odynophagia (painful swallowing) in patients taking corticosteroids, antibiotics or immunosuppressive drugs. The characteristic change on endoscopy is the formation of discrete white plaques.
Keywords adverse events; drugs; gastrointestinal injury; large bowel; NSAIDs; small intestine
Mucosal injury and strictures Tetracyclines, potassium chloride, certain bisphosphonates and quinidine are recognized as toxic to the oesophageal mucosa.
Almost any drug can affect the gastrointestinal (GI) tract. It has been estimated that 40% of adverse drug reactions affect the GI tract, and 4% of admissions to hospital are caused by druginduced disorders. This contribution highlights drug-induced GI disorders that are common, important or often missed. It is useful to consider drug effects on the GI tract in the four categories below. Consequences of pharmacological mode of action e adverse drug effects can occur as a predictable result of a drug’s mode of action. Thus, drugs with anticholinergic effects (e.g. antidepressants) reduce oesophageal sphincter pressure, resulting in reflux and heartburn; they also reduce colonic transit, resulting in constipation. Impairment of GI defences e the best-known examples are non-steroidal anti-inflammatory drugs (NSAIDs), which act
Common drug causes of oesophageal problems Relax lower oesophageal sphincter causing heartburn C Anticholinergic agents (e.g. procyclidine, trihexyphenidyl) C Tricyclic antidepressants C Calcium channel blockers C Nitrates C Phenothiazines Direct mucosal injury C Tetracyclines C Bisphosphonates Associated with strictures C Potassium chloridea C Quinidinea C NSAIDs
Anthony Norman MRCP is SpR in Gastroenterology at University Hospital, Nottingham, UK. Competing interests: none declared. Christopher J Hawkey DM FRCP is Professor of Gastroenterology at University Hospital, Nottingham, UK. Competing interests: none declared.
MEDICINE 39:3
a
Particularly modified-release preparations.
Table 1
162
Ó 2010 Published by Elsevier Ltd.
STOMACH
However, presentation may vary with different drugs. NSAIDs and tetracyclines are a relatively common cause of erosive or pill oesophagitis; potassium chloride and quinidine are more strongly associated with stricture, often in the mid-oesophagus. These problems are more common in elderly patients, particularly those with pre-existing oesophageal disease or massive left atrial enlargement, and are more likely to occur when patients swallow their pills in a reclining position. Sustained-release preparations of potassium and quinidine are particularly likely to cause problems. There is an association between bisphosphonates and reversible oesophagitis, but the risk is low if the medication is administered correctly (i.e. in an upright position).2
low doses for cardiovascular prophylaxis), this drug is now associated with as many ulcer bleeds as all other NSAIDs together. Overall, patients taking NSAIDs probably undergo hospitalization for gastrointestinal problems about once every 50e100 patientyears. In the UK, it is likely that at least 2000 patients per year die as a result of NSAID use (because of GI complications).9 Risk factors Patient-associated risk factors: concomitant risk factors are age and previous gastrointestinal bleeding.3,10e13 Drug-related risk factors: the main drug-related risk factors are dose, choice of NSAID, co-prescription of corticosteroids and coprescription of warfarin. The risk of ulcer complications increases with dose. A meta-analysis by Lewis et al. found significant differences in the odds ratio for upper gastrointestinal bleeding, with an increased risk from ibuprofen (odds ratio 1.7) to ketoprofen (34.9).14 In addition, there was an up to 8-fold variation in risk across the recommended dose ranges for each drug. Indometacin <1200 mg/day was no more harmful than paracetamol, but at doses >1800 mg/day, it had an odds ratio of 4.6. As a general rule, the GI toxicity of NSAIDs correlates with the degree of anti-inflammatory activity, so that those with high analgesic/low anti-inflammatory activity (e.g. ibuprofen) are less ulcerogenic than those with relatively high anti-inflammatory: analgesic properties (e.g. piroxicam).15 The GI risks with aspirin are becoming a major cause of ulcer complications as its use becomes more widespread. Furthermore, the combination of NSAIDs and aspirin has a multiplicative rather than a simple additive effect on risk. Even the low doses of aspirin widely used to prevent cardiovascular events are associated with an increased risk of upper GI haemorrhage, with a daily dose of 75 mg having an odds ratio of around 2, and a daily dose of 300 mg having an odds ratio of almost 4.16 At least some of this is attributable to the anti-haemostatic effects of aspirin. Clopidogrel is also associated with an increased incidence of upper GI haemorrhage. Co-prescription of clopidogrel and aspirin in patients with ischaemic heart disease further increases the risk of gastrointestinal bleeding by a factor of 15.17
Dyspepsia and gastroduodenal ulcers NSAIDs About one-third of patients take NSAIDs experience dyspepsia. In most cases, this occurs in the absence of ulceration. Some studies suggest that NSAID-induced dyspepsia is particularly common in patients infected with Helicobacter pylori. NSAIDs are the commonest cause of gastric and duodenal ulcers (Figure 1). Studies in which patients taking NSAIDs have undergone endoscopy regardless of symptoms have shown a high incidence of ulceration (about 20%). It seems likely, however, that some of these ulcers are relatively trivial or transient, or are superficial erosions mistakenly classified as ulcers. NSAIDs cause ulcers because of inhibition of prostaglandin synthesis. Diagnostic problems e many NSAID-associated ulcers are silent, and NSAIDs often cause dyspepsia without ulceration. Dyspepsia is thus a poor guide when deciding whether to perform endoscopy. Estimation of risk e the use of traditional NSAIDs increases the risk of serious GI complications between 2.5 and 5-fold.3e5 Recent studies have shown that the cumulative incidence of upper GI complications is between 0.92% and 1.4% after 12 months of NSAID use.6e8 With the increasing use of aspirin (particularly in
Cyclo-oxygenase (COX)-2 inhibitors Selective COX-2 inhibitors were heralded as ‘safer NSAIDs’ because they carried the anti-inflammatory benefits of COX-2 inhibition whilst retaining the gastroprotective benefits of COX-1. There have been four large outcome studies of rofecoxib (VIGOR study7), celecoxib (CLASS study6), lumiracoxib (TARGET study18) and etorecoxib (MEDAL programme19). TARGET showed a five-fold reduction in ulcer complications to very low levels that were similar to what is seen in the background population in patients not taking aspirin. All studies have similarly shown reductions in complicated and uncomplicated ulcers or in close surrogate measures to a greater or lesser extent in non-users of aspirin. In all studies aspirin use was associated with a loss of most, if not all, of the benefits of the COX-2 inhibitor. Proton pump inhibitors An important recent development has been the recommendation by NICE that all patients taking an NSAID or a COX-2 inhibitor should be given ulcer prophylaxis with a proton pump inhibitor
Figure 1 An NSAID-induced gastric ulcer with active bleeding. Courtesy of Clinical gastroenterology and hepatology, Mosby, 2005.2
MEDICINE 39:3
163
Ó 2010 Published by Elsevier Ltd.
STOMACH
(PPI).20 This advice reflects the marked fall in the cost of omeprazole (£2.04 for omeprazole 20 mg for 28 days). This is based largely on primary prophylaxis endoscopic studies and secondary prevention studies that have used ulcer bleeding as an endpoint. For those at particularly high risk of GI toxicity, there is evidence to promote the combination of COX-2 inhibitors with PPI therapy as this markedly reduces the incidence of recurrent GI ulcer bleeding.21 There is current concern that at least some PPIs may inhibit the activation of clopidogrel22 and lead to an increased incidence of thrombotic events.23 Giving the PPI in the morning and clopidogrel at night may avoid the problem.24 Prescribers should use the most up-to-date recommendations. H. pylori e the role of H. pylori in drug-induced GI ulceration is complex and there are few systematic trial data. Some studies have shown that eradication of H. pylori before NSAID use reduces the risks of endoscopic ulcers.25 Others have shown no difference26 or that those with continuing H. pylori infection have better rather than worse outcomes when acid-suppressing drugs are used.27
pivampicillin have been associated with peptic ulceration, though the evidence for the association is often informal. Recently, calcium channel antagonists have been implicated in peptic ulcer bleeding. Corticosteroids e it now seems likely that corticosteroids do not increase the risk of ulceration, except in patients taking NSAIDs.
Nausea and vomiting Nausea and vomiting are common adverse drug reactions that may occur as a direct effect of the drug or as a result of gut damage, including ulceration. The more common drug causes are listed in Table 2. Many drugs (e.g. levodopa, opioids, cisplatin, digoxin) induce symptoms by directly stimulating the chemoreceptor trigger zone in the area postrema (located on the dorsal surface of the medulla oblongata at the caudal end of the fourth ventricle). Chemotherapeutic agents probably cause nausea and vomiting by stimulating serotonin release from entero-chromaffin cells within the gut, which activates the vagus. Drugs such as potassium chloride, iron preparations and NSAIDs probably act via direct gastric irritation. In a few patients, these symptoms are accompanied by gastric or duodenal ulceration. Dopamine receptors in the area postrema are the target for anti-emetic drugs like chlorpromazine, haloperidol, domperidone and metoclopramide. Nausea is a challenging symptom to investigate and no drug should be judged blameless until the effect of stopping it has been assessed. Use of several drugs at one time, particularly in elderly patients, can cause nausea, and it is important to simplify drug regimens in such patients.
Recommendations for prophylaxis Patients with one or more risk factors for NSAID associated ulcers (80% of patients) should be advised to take preventive treatment, although this is often neglected. Many patients are given H2-receptor antagonists to protect them against NSAID damage, but there is good evidence to suggest that at conventional dosage these agents prevent only duodenal ulcer. Proton pump inhibitors (PPIs) have been shown to prevent aspirin- and NSAID-induced small upper gastrointestinal injury. There remains some concern over their use in patients treated with clopidogrel, because of abrogation of the anti-platelet effect. Misoprostol is as effective as PPIs but is poorly tolerated.28 COX-2 inhibitors reduce the risk of GI events compared to traditional NSAIDs. This benefit is lost in aspirin users. However, unlike NSAIDs, COX-2 inhibitors do not interfere with aspirin’s ability to inhibit platelet aggregation. Although there is concern about the effects of COX-2 inhibitors (and NSAIDs) on cardiovascular outcomes,29 a combination of a COX-2 inhibitor with aspirin may prove to be a more reliable anti-haemostatic strategy than a combination with a non-selective NSAID. For those at high risk of GI ulceration, addition of a PPI to a COX-2 inhibitor offers even greater GI protection.
Pancreatitis Alcohol is a common cause of pancreatitis. Other drug-related causes are uncommon, but the main culprits are: diuretics (thiazides, furosemide) sulphonamides azathioprine 6-mercaptopurine corticosteroids
Common drug causes of nausea and vomiting Locally irritant C Potassium chloride C Iron preparations C NSAIDs C Theophyllines C Azathioprine (?mechanism) C Metronidazole
Management NSAID-induced ulcers have been shown to heal with all standard anti-ulcer drug regimens. Healing is faster if the NSAID is stopped. Comparative studies show that omeprazole, 20 mg daily, achieves faster healing in gastric and duodenal ulcers than ranitidine, 150 mg twice daily, or misoprostol, 200 mg four times daily.30
Act via the CNS C Levodopa C Bromocriptine C Opioids C Digoxin C Chemotherapeutic agents (e.g. cisplatin, mustine, dacarbazine, cyclophosphamide)
Other causes It should be remembered that dyspepsia occurs with the nonaspirin salicylates (e.g. mesalazine) used in inflammatory bowel disease. For most drugs, the precise mechanism by which dyspepsia arises is unknown, though there are a few exceptions (e.g. erythromycin, which causes epigastric pain and general cramp as a consequence of its prokinetic activity as a motilin receptor agonist). Drugs such as cytotoxic agents, oral gold and
MEDICINE 39:3
Table 2
164
Ó 2010 Published by Elsevier Ltd.
STOMACH
NSAIDs sodium valproate exenatide and liraglutide L-asparaginase (about 7% of recipients).
Common drug causes of constipation and diarrhoea Constipation C Drugs with antimuscarinic effects (e.g. atropine, phenothiazines, tricyclic antidepressants) C Opioids C Mebeverine C Peppermint oil C Aluminium-containing antacids C Sucralfate C Gaviscon C Iron C Laxatives (chronic) Diarrhoea C b-blockers C Misoprostol C Antibiotics C Magnesium-containing antacids C Olsalazine C Mefenamic acid C Iron C Laxatives (acute) C Metformin C Angiotension-converting enzyme inhibitors C Statins
Constipation Opioids and drugs with antimuscarinic effects are the main causes of drug-induced constipation (Table 4). An interesting development is the use of peripheral opioid antagonist, naltrexone, to prevent constipation in patients in palliative care who are receiving opioids. Reduced bowel frequency also occurs with selective serotonin reuptake inhibitors (e.g. fluoxetine) and locally acting antispasmodics (e.g. mebeverine, peppermint oil). Codeine-containing analgesics are available over the counter, and some patients establish a selfdefeating cycle, using them for constipation-induced abdominal pain. Aluminium-containing antacids, sucralfate and Gaviscon also commonly constipate, as does iron (though this can also cause diarrhoea). Drug-induced megacolon from enteric neurotoxicity caused by vincristine or (possibly) laxatives is an uncommon but important cause of drug-induced constipation. Although fibre (dietary and prescribed) is used to treat constipation, it can cause pain, bloating and, occasionally, impaction if excessive amounts are ingested unaccompanied by adequate volumes of liquid. Insoluble dietary fibre is often used to treat constipation but it can cause pain, bloating and occasional impaction if excessive amounts are ingested unaccompanied by adequate volumes of liquid. Soluble fibre (porridge, bananas, isphagula husk) is better tolerated by many patients.
Table 3
Deliberate use of laxatives should always be considered in patients with diarrhoea, particularly when potassium is low. All salicylates can cause diarrhoea, but it is particularly likely with olsalazine, which has been shown to stimulate small intestinal secretion. Azathioprine can occasionally cause diarrhoea, sometimes in the context of toxic shock-like syndrome. This may cause diagnostic confusion when the drug is used for inflammatory bowel disease. It is increasingly recognized that drugs that suppress acid production (H2-antagonists and PPIs) predispose to enteric superinfection.
Melanosis coli Anthranoid laxatives (e.g. senna) are widely used laxatives of natural origin that pass unabsorbed to the large bowel, where they are metabolized to aglycones. These aglycones exert their laxative effect by damaging epithelial cells, which leads directly and indirectly to changes in absorption, secretion and motility. Damaged epithelial cells can be found as apoptotic bodies in the pigmented colonic mucosa, characteristic of pseudomelanosis coli. The belief that laxatives cause enteric neuronal damage leading to chronic megacolon is probably misplaced.
Malabsorption Diarrhoea
Many drugs (e.g. colchicine, methotrexate, laxatives) can cause malabsorption, though this is usually of little clinical significance. Sulfasalazine and phenytoin commonly cause folate malabsorption.
Drug-induced diarrhoea (Table 3) is common. The main causes are: antibiotics, which disturb the colonic flora and tend to select C. difficile (responsible for a growing epidemic of pseudomembranous colitis in hospitalized patients) b-blockers, which act by antagonizing antiperistaltic adrenergic stimulation bile acids, which have a direct irritant action in the colon bisoprostol, which stimulates intestinal secretion and motility. Other causes include the following. Magnesium-containing antacids (available over the counter); in a recent study, they accounted for 4% of patients referred for evaluation of diarrhoea. Generally, this was not detected in the initial history.
MEDICINE 39:3
Small intestinal injury The small intestine may be a more common site for NSAID toxicity than the well-recognized effects on the stomach and duodenum, although these effects may be more occult, and usually arise only after chronic (e.g. 6 months) use. An estimated 70% of NSAID users demonstrate intestinal inflammation when indirectly measured, for example by scintigraphy.31 Small bowel enteropathy, a combination of increased intestinal permeability and mild mucosal inflammation, is suggested by hypoalbuminaemia and iron deficiency. Studies of long-term NSAID users suggest 60e70% have an asymptomatic enteropathy.31,32
165
Ó 2010 Published by Elsevier Ltd.
STOMACH
Capsule endoscopy has allowed the demonstration that NSAIDs cause mucosal erosions in most subjects,33 whilst COX-2 inhibitors have a much reduced or absent effect.7,34 This may contribute to the lower incidence of anaemia seen in patients taking COX-2 inhibitors (Figure 2).
Drugs that cause colitis Antibiotics Amoxicillin Ampicillin Clindamycin Erythromycin Cephalosporins
Colonic ulcers NSAIDs are also recognized to cause sporadic colonic ulcers. Increasing use of nicorandil, which has long been associated with oral and anal ulceration, has led to recognition that it also causes colonic ulceration.35,36 The reversibility of the injury on cessation of therapy is well documented and the association is not in doubt. The mechanism, however, is yet to be established. In terms of clinical manifestations, the patient may develop anaemia, diarrhoea or weight loss, and even ileal perforation has been reported.37
Drugs causing ischaemic colitis Oral contraceptive pill Chemotherapeutic agents e 5-fluorouracil, cisplatin danazol Vasopressin Clindamycin Miscellaneous NSAIDs Methyldopa Penicillamine Gold (oral)
Colitis The most common drug-related causes of colitis are shown in Table 4. Mefenamic acid directly damages epithelial cells; clinical presentation ranges from simple reversible secretory diarrhoea to colitis. Other NSAIDs seldom cause diarrhoea and none has been conclusively shown to be a primary cause of colitis, though they may provoke relapse. Collagenous colitis is a rare condition that may be associated with NSAID use, but this may be a consequence rather than a cause because patients with collagenous colitis exhibit a greater incidence of arthritis, which may require NSAID administration. Antibiotics commonly cause loose stools or diarrhoea, and patients may present with severe pseudo-membranous colitis (see Figure 3) as a result of C. difficile selection.
Table 4
also been implicated in the exacerbation of inflammatory bowel disease.38 The mechanism is unknown. It is increasingly recognized that NSAIDs cause lower gastrointestinal tract problems, including erosions, ulcers and strictures. Presentations include blood and protein loss, perforation and obstruction. Minimizing the adverse gastrointestinal effects of drugs Drugs can get stuck in the oesophagus, particularly when in gelatine capsules; they should be taken with plenty of fluid. NSAID-associated gastroduodenal toxicity varies with drug and dosage; the best advice is to stop the NSAID whenever possible, minimize the dosage, or use ibuprofen, 1200 mg daily, or a COX-2 inhibitor.
Exacerbation of ulcerative colitis There is evidence that analgesic drugs (NSAIDs and paracetamol) provoke relapse of colitis in some patients. COX-2 drugs have
Figure 3 Appearance of pseudomembrane in sigmoid colon after broadspectrum antibiotics (cephalosporin), with proven Clostridium difficile toxin on stool testing. Courtesy of Clinical gastroenterology and hepatology, Mosby, 2005.2
Figure 2 Typical small bowel erosions in the terminal ileum of a healthy volunteer taking NSAIDs, with red halo resulting from hyperaemia and pale centre.
MEDICINE 39:3
166
Ó 2010 Published by Elsevier Ltd.
STOMACH
NICE recommends that COX-2 inhibitors be reserved for patients at ‘high risk’ of GI side-effects; an alternative might be to use COX-2 inhibitors in ‘average-risk’ patients and consider NSAIDs plus a PPI, or even a coxib plus a PPI in high-risk patients, based on individual circumstances.39 Avoiding modified-release preparations may reduce small and large bowel problems associated with NSAIDs and iron preparations. In patients with inflammatory bowel disease, the possibility of drug-induced diarrhoea caused by olsalazine or azathioprine, and disease relapse caused by paracetamol or NSAIDs, requires careful consideration. Diarrhoea caused by b-blockers is common and often not appreciated by prescribers. Prescribers of H2-antagonists and proton pump inhibitors should be aware of the possibility of enteric infections. A
12 Shorr RI, Daugherty JR, Griffin MR. concurrent use of nonsteroidal anti-inflammatory drugs and oral anticoagulants places elderly persons at high risk for haemorrhagic peptic ulcer disease. Arch Intern Med 1993; 153: 1665e70. 13 Mellemkjaer L, Sorensen HT, Thomassen L, et al. Upper gastrointestinal bleeding among users of NSAIDs: a population-based cohort study in Denmark. Br J Clin Pharmacol 2002; 53: 173e81. 14 Lewis SC, Laporte J, Matthews JN, Rawlins MD, Wiholm B. Doseresponse relationships between individual nonaspirin nonsteroidal antiinflammatory drugs (NANSAIDs) and serious upper gastrointestinal bleeding: a meta-analysis based on individual patient data. Br J Clin Pharmacol 2002; 54: 320. 15 Graham DY. Nonsteroidal anti-inflammatory drugs, Helicobacter pylori, and ulcers: where we stand. Am J Gastroenterol 1996; 91: 2080e6. 16 Tashima K, Fujita A, Umeda M, Takeuchi K. Lack of gastric toxicity of nitric oxide-releasing aspirin, NcX-4016, in the stomach of diabetic rats. Life Sci 2000; 67: 1639e52. 17 Lanas A, Garcia-Rodriguez LA, Arroyo MP, et al. Risk of upper gastrointestinal bleeding associated with selective cyclo-oxygenase2 inhibitors, traditional non-steroidal non-aspirin anti-inflammatory drugs, aspirin and combinations. Gut 2006; 55: 1731e8. 18 Schnitzer TJ, Burmester GR, Mysler E, et al. Comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), reduction in ulcer complications: randomised controlled trial. Lancet 2004; 364: 665e74. 19 Cannon CP, Curtis SP, FitzGerald GA, et al. Cardiovascular outcomes with etoricoxib and diclofenac in patients with osteoarthritis and rheumatoid arthritis in the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) programme: a randomised comparison. Lancet 2006; 368: 1771e81. 20 Dyspepsia: managing dyspepsia in adults in primary care. NICE guideline CG17, 2004. 21 Chan FKL, Wang VWS, Suen BY, et al. Combination of a cyclo-oxygenase-2 inhibitor and a proton-pump inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: a double-blind, randomised trial. The Lancet 2007; 369: 1621e6. 22 Gilard M, Arnaud B, Cornily JC, et al. Influence of omeprazole on the antiplatelet action of clopidogrel associated with aspirin: a randomised, double-blind OCLA (Omeprazole CLopidogrel Aspirin) study. J Am Coll Cardiol 2008; 51: 256e60. 23 Ho MP, Maddox L, Wang L, et al. Risk of adverse outcomes associated with concomitant use of clopidogrel and proton pump inhibitors following acute coronary syndrome. JAMA 2008; 301: 937e44. 24 Bhatt DL. COGENT: a prospective, randomized, placebo-controlled trial of omeprazole in patients receiving aspirin and clopidogrel. Presented at: Transcatheter Cardiovascular Therapeutics, 2009. 25 Chan FK, Sung JJ, Chung SC, et al. Randomised trial of eradication of Helicobacter pylori before non-steroidal anti-inflammatory drug therapy to prevent peptic ulcers. Lancet 1997; 350: 975e9. 26 Hawkey C, Tulassay Z, Szczepanski L, et al. Randomised controlled trial of Helicobacter pylori eradication in patients on non-steroidal antiinflammatory drugs: HELP NSAIDs study. Lancet 1998; 352: 1016e21. 27 Bianchi Porro G, Parente F, Imbesi V, Montrone F, Caruso I. Role of Helicobacter pylori in ulcer healing and recurrence of gastric and duodenal ulcers in longterm NSAID users. Response to omeprazole dual therapy. Gut 1996; 39: 22e6. 28 Graham DY, Agrawal NM, Campbell DR, et al. Ulcer prevention in longterm users of nonsteroidal anti-inflammatory drugs: results of a double-blind, randomized, multicenter, active- and placebo-
REFERENCES 1 Neal KR, Scott HM, Slack RC, Logan RF. Omeprazole as a risk factor for campylobacter gastroenteritis: case-control study. BMJ 1996; 312: 414e5. 2 Cryer B, Bauer DC. Oral bisphosphonates and upper gastrointestinal tract problems: what is the evidence? Mayo Clin Proc 2002; 77: 1031e43. 3 Hernandez-Diaz S, Rodriguez LA. Association between nonsteroidal anti-inflammatory drugs and upper gastrointestinal tract bleeding! perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med 2000; 160: 2093e9. 4 Derry S. Risk of gastrointestinal haemorrhage with long term use of aspirin: meta-analysis. BMJ 2000; 321: 1183e7. 5 Garcia Rodriguez LA, Hernandez-Diaz S, De Abajo J. Association between aspirin and upper gastrointestinal complications: systematic review of epidemilogic studies. Br J Clin Pharmacol 2001; 52: 563e71. 6 Silverstein FE, Goldstein JL, Simon LS, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the class study: a randomized controlled trial. Celecoxib long-term arthritis safety study. JAMA 2000; 284: 1247e55. 7 Bombardier C, Reicin A, Shapiro D, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGoR Study Group. N Engl J Med 2000; 343: 1520e8. 8 Schnitzer TJ, Mysler E, Hochberg MC, et al. comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), reduction in ulcer complications: randomised controlled trial. Lancet 2004; 364: 665e74. 9 Blower AL, Brooks A, Fenn GC, et al. Emergency admissions for upper gastrointestinal disease and their relation to NSAID use. Aliment Pharmacol Ther 1997; 11: 283e91. 10 Laine L, Bombardier C, Hawkey CJ, et al. Stratifying the risk of NSAIDrelated upper gastrointestinal clinical events: results of a doubleblind outcomes study in patients with rheumatoid arthritis. Gastroenterology 2002; 123: 1006e12. 11 Gabriel SE, Jaakkimainen L, Bombardier C. Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. A meta-analysis. Arch Intern Med 1991; 115: 787e96.
MEDICINE 39:3
167
Ó 2010 Published by Elsevier Ltd.
STOMACH
29
30
31
32 33
34
35
36 Titi MA, Seow C, Molloy RG. Nicorandil-induced colonic ulceration: a new cause of colonic ulceration. Report of four cases. Dis Colon Rectum 2008 Oct; 51: 1570e3. 37 King PM, Suttie SA, Jansen JO, Watson AJM. Perforation of the terminal ileum: apossible complication of nicorandil therapy. Surg J R Coll Surg Edinb Ire 2004; 2: 56e7. 38 Matuk R, Crawford J, Abreu MT, Targan SR, Vasilauskas EA, Papadakis KA. The spectrum of gastrointestinal toxicity and effect on disease activity of selective cyclo-oxygenase-2 inhibitors in patients with inflammatory bowel disease. Inflamm Bowel Dis 2004; 10: 352e6. 39 National Institute for Health and clinical Excellence. Guidance on the use of cyclo-oxygenase (cox) II selective inhibitors, celecoxib, rofecoxib, meloxicam and etodolac for oA and RA. Technology Appraisal Guidance No. 27, 2001.
controlled study of misoprostol vs lansoprazole. Arch Intern Med 2002; 162: 169e75. Bresalier RS, Sandler RS, Quan H, et al. cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med 2005; 352: 1092e102. Yeomans N. New data on healing of nonsteroidal anti-inflammatory drug-associated ulcers and erosions. omeprazole NSAID Steering committee. Am J Med 1998; 104: 79Se80. Bjarnason I, Hayllar J, MacPherson AJ, Russell AS. Side effects of nonsteroidal anti-inflammatory drugs on the small and large intestine in humans. Gastroenterology 1993; 104: 1832e47. Bjarnason I, Smith T, Prouse P, Williams P, Smethurst P. NSAID drug induced inflammation in humans. Gastroenterology 1987; 93: 480e9. mez-Rodrı´guez BJ, Romero-Vazquez J, et al. Caunedo-Alvarez A, Go Macroscopic small bowel mucosal injury caused by chronic consume of NSAID assessed by capsule endoscopy. Rev Esp Enferm Dig 2010; 102: 80e5. Chan FK, Scheiman JM, Lanas A, Berger M, Li MC, Goldstein JL. Celecoxib versus Diclofenac and Omeprazole: a randomized controlled trial comparing a composite outcome across the entire GI tract (The Condor Trial). AGA 2010; S20: 110 (Abst). Brown R, Lee A, Welfare M. Nicorandil-induced colonic ulceration. Heart 2008 May; 94: 678.
MEDICINE 39:3
FURTHER READING Fortun PJ, Hawkey CJ. Non-steroidal anti-inflammatory drugs and the small intestine. Curr Opin Gastroenterol 2005; 21: 169e75. Gilman AG, Roll TW, Nies AS, et al. The pharmacological basis of therapeutics. New York: McGraw-Hill, 1992. Rostom A, Dube C, Wells G, et al. Prevention of NSAID-induced gastroduodenal ulcers (cochrane Review). Cochrane Database Syst Rev 2002; (Issue 4).
168
Ó 2010 Published by Elsevier Ltd.