Gastrointestinal ADRs

Gastrointestinal ADRs

INTRODUCTION

Disturbances of gastrointestinal function are common events that can be attributed to the ingestion of a wide range of drug classes. In the 1970s, it was reported that some 20%–40% of adverse drug reactions (ADRs) in hospital monitoring were gastrointestinal in origin (Hurwitz and Wade, 1969). More recent estimates of the incidence of ADRs in hospitalised patients (Bates, Leape and Petrycki, 1993; Bowman, Carlstedt and Black, 1994; Lazarou, Pomeranz and Corey, 1998; Bates et al., 1995a,b) or of subjects admitted to hospi-tal due to an ADR (Col, Fanale and Kronholm, 1990; Einarson, 1993; Nelson and Talbert, 1996; Lazarou, Pomeranz and Corey, 1998; Roughead et al., 1998; Pouyanne et al., 2000) provide only limited informa-tion specifically about gastrointestinal events.

A study in over 4000 hospitalised patients in the United States found 247 ADRs among 207 admissions (Bates et al., 1997). The examination of each by organ system affected showed that 18% of events were of gastrointestinal origin, predominantly nausea, vomit-ing and antibiotic-associated diarrhoea. An almost identical rate was reported in an observational study of 1024 patients in an internal medicine ward in the United States (Bowman, Carlstedt and Black, 1994). The gastrointestinal system was the organ system affected in 17.8% of drug-related adverse events.

The findings from a prospective study in France showed that gastrointestinal events were the most frequent cause for admission to hospital for an ADR (Pouyanne et al., 2000). Of 100 admissions, 27 were gastrointestinal, including 13 cases of gastrointesti-nal haemorrhage caused by anticoagulant drugs and 9 caused by the ingestion of non-steroidal anti-inflammatory drugs (NSAIDs).

The extent of drug-related hospital admissions in Australia was reviewed from Australian studies published between 1988 and 1996 (Roughead et al., 1998). Fourteen studies were included in the analy-sis although the diagnosis associated with the drug-related admissions was available from only five reports. Among the conditions commonly identified was gastrointestinal bleeding, which usually was asso-ciated with either warfarin or NSAID therapy.

Many drugs causing gastrointestinal disorders have been recognised (Bateman and Aziz, 1998).

Well-established unwanted effects of drugs include changes in gastrointestinal motility, altered gastric emptying, disturbances of nutrient absorption, antimicrobial-associated colitis and pseudomembra-nous colitis. Furthermore, drug-induced lesions are documented for all sections of the gastrointesti-nal tract. These encompass a wide range of pathophysiological processes including inflammation, the formation of strictures, haemorrhage, ulceration and perforation. Others consist of symptoms such as nausea and vomiting (Quigley, Hasler and Parkman, 2001), diarrhoea (Fine and Schiller, 1999) or consti-pation (Locke, Pemberton and Phillips, 2000) in the absence of underlying pathology.

The medical literature on gastrointestinal ADRs is dominated by reports concerning the NSAIDs. Effects have been documented over many years, but it has been during the 1990s that the risk factors for upper gastrointestinal problems have been systematically examined. Over the same period, the small and large bowel toxicities of the NSAIDs have also become clearly identified.

In this chapter, we summarise some of the impor-tant literature and reviews from the 1990s concerning the adverse effects of NSAIDs on the gastrointesti-nal tract. We also review the medical literature of the 1990s to identify adverse gastrointestinal effects with other medications detected using a variety of pharma-covigilance techniques.

The oesophagus, despite its physiological defence mechanisms, is prone to injury induced by a wide variety of agents. Medication-induced oesophageal injury or ‘pill oesophagitis’ was first described in the 1970s (Pemberton, 1970). In most cases, direct oesophageal toxicity is the cause, and the condition is generally fully reversible on the withdrawal of treatment (Doman and Ginsberg, 1981; Kikendall, 1999a). Pill oesophagitis is often underdiagnosed; in many instances, it is incorrectly believed to be gastro-oesophageal reflux disease (Doman and Ginsberg, 1981; Bonavina et al., 1987). Almost 1000 reports in the medical literature of pill oesophagitis attributable to about 100 different medications have been exten-sively reviewed (Kikendall, 1999a,b). Drugs most frequently implicated in pill oesophagitis (reports of ≥10 cases) include antibiotics (doxycycline, tetra-cycline hydrochloride and other unspecified tetra-cyclines, oxytetracycline, pivmecillinam), potassium chloride, alendronate, ferrous sulphate and ferrous succinate, quinidine, naproxen, aspirin, emepronium bromide, pinaverium bromide and alprenolol (Bott, Prakash and McCallum, 1987; Baehr and McDonald, 1998; Kikendall, 1999a,b; Graham, 2000).

In the upper gastrointestinal tract, NSAIDs are causally associated with peptic ulceration along with associated complications such as bleeding and perfo-ration. Non-steroidal anti-inflammatory drugs also cause upper gastrointestinal haemorrhage as may the selective serotonin re-uptake inhibitors. Studies in volunteers have shown that alendronate, one of the bisphosphonate class of drugs, may cause acute gastric mucosal damage and gastric ulceration.

Non-steroidal anti-inflammatory drugs can also cause a low-grade enteropathy in the small intestine. Additionally, in both small and large intestine, they have been associated with the formation of strictures, bleeding and perforation.

Recently, an association between a rotavirus vaccine and intussusception in children has been reported, and fibrosing colonopathy has been linked with the use of pancreatin supplements in children and adults with cystic fibrosis. The possibility that measles–mumps–rubella (MMR) vaccination may be a causal factor in the development of inflamma-tory bowel disease is currently a matter of some controversy.

Numerous drugs have been reported to have caused obstruction of the gastrointestinal tract (Iredale, 1993). Acute colonic pseudo-obstruction is characterised by massive colonic dilation with a clinical and radiolog-ical appearance of mechanical obstruction but in the absence of primary colonic pathology. Although the underlying pathogenetic mechanisms are unknown, it is commonly associated with surgery, trauma, metabolic imbalance, neurological disease and seri-ous systemic illness. Anecdotal case reports in the 1980s and 1990s have associated various drugs with colonic pseudo-obstruction including cloni-dine (Maganini and Pollitt, 1983; Stieger, Cantieni and Frutiger, 1997), imipramine (Sood and Kumar, 1996), amitriptyline (McMahon, 1989), amitriptyline with concomitant lithium (Fava and Galizia, 1995), nimodipine (Fahy, 1996), tocolytic therapy compris-ing intravenous magnesium and nifedipine (Pecha and Danilewitz, 1996), interleukin-2 (Post, Falk and Bukowski, 1991), diltiazem (Mantzoros, Prabhu and Sowers, 1994; Fauville et al., 1995), morphine (Murthy, Ion and Winstanley, 1998), fludarabine (Campbell et al., 2000), and enteral activated charcoal alone (Brubacher, Levine and Hoffman, 1996) and together with sorbitol and papaveretum (Longdon and Henderson, 1992) when given for the management of theophylline overdose.