Drug-induced gastrointestinal bleeding. Report from the Boston Collaborative Drug Surveillance Program, Boston University Medical Center

Diagnosis of gastrointestinal lesions during treatment with non-steroidal anti-inflammatory drugs

Probably all non-steroidal anti-inflammatory drugs (NSAIDs) increase the risk of gastrointestinal (GI) mucosal injury. The most frequent lesions are gastric erosions in the pre-pyloric region. Non-steroidal anti-inflammatory drugs also increase the incidence of peptic ulcers, although it is not yet clear whether more gastric ulcers or more duodenal ulcers are produced. Gastrointestinal symptoms are prevalent during NSAID treatment, but there is no correlation between symptoms and lesions. Clinical presentation may therefore vary considerably: many symptomatic patients will have no lesions, whereas others will present with complications but no symptoms. Blood loss is the most frequent complication of NSAID-induced GI lesions. Bleeding is usually 'silent' and occult. Overt haemorrhage, though rare, is more frequent in patients taking NSAIDs. Other complications, such as penetration and perforation, may also occur. Endoscopy is the diagnostic method of choice, as this method can detect even superficial mucosal lesions. However, because of the lack of correlation between symptoms and lesions, it is difficult to ascertain which patients are at risk and thus to avoid unnecessary diagnostic procedures. Future efforts should therefore be directed towards the prevention of NSAID-induced GI lesions.

Comparison of evidence on harms of medical interventions in randomized and nonrandomized studies

BACKGROUND

Information on major harms of medical interventions comes primarily from epidemiologic studies performed after licensing and marketing. Comparison with data from large-scale randomized trials is occasionally feasible. We compared evidence from randomized trials with that from epidemiologic studies to determine whether they give different estimates of risk for important harms of medical interventions.

METHODS

We targeted well-defined, specific harms of various medical interventions for which data were already available from large-scale randomized trials (> 4000 subjects). Nonrandomized studies involving at least 4000 subjects addressing these same harms were retrieved through a search of MEDLINE. We compared the relative risks and absolute risk differences for specific harms in the randomized and nonrandomized studies.

RESULTS

Eligible nonrandomized studies were found for 15 harms for which data were available from randomized trials addressing the same harms. Comparisons of relative risks between the study types were feasible for 13 of the 15 topics, and of absolute risk differences for 8 topics. The estimated increase in relative risk differed more than 2-fold between the randomized and nonrandomized studies for 7 (54%) of the 13 topics; the estimated increase in absolute risk differed more than 2-fold for 5 (62%) of the 8 topics. There was no clear predilection for randomized or nonrandomized studies to estimate greater relative risks, but usually (75% [6/8]) the randomized trials estimated larger absolute excess risks of harm than the nonrandomized studies did.

INTERPRETATION

Nonrandomized studies are often conservative in estimating absolute risks of harms. It would be useful to compare and scrutinize the evidence on harms obtained from both randomized and nonrandomized studies.

Risk of hospitalization for upper gastrointestinal tract bleeding

OBJECTIVE

This study evaluates the hospitalization risk for upper gastrointestinal bleeding (UGIB) with reference to the clinical characteristics of patients and drugs taken before admission.

METHODS

This study is based on the GIFA (Italian Group for the Pharmacosurveillance in the Elderly) database. Cases with an ICD-9 code of esophagus, stomach or duodenum bleeding, or acute esophago-gastroduodenal disease associated with anemia have been classified as UGIB. Sex, age, year of observation, drugs taken at home, comorbidity, smoking, alcohol, and use of gastroprotectants have been also taken into account. Statistical analysis has been conducted using multivariate logistic regression models.

RESULTS

32,388 patients have been enrolled, 940 of which presented UGIB. Age, comorbidity, use of smoke and alcohol, hospitalization duration, and mortality during hospitalization were significantly higher in UGIB than nonUGIB patients. Increased UGIB risk has been found in patients taking NSAIDs (both when aspirin was included or excluded), acetaminophen, constipating agents, iron, ethacrynic acid, propranolol. Reduced UGIB risk has been found in patients taking nitrates.

CONCLUSIONS

UGIB risk appears to correlate with clinical characteristics of the patient: it increases with age, comorbidity, and smoke and alcohol consumption. Among drugs, NSAIDs are associated with the highest UGIB risk, while nitrates with a reduction of risk.

Interaction between nonsteroidal anti-inflammatory drug intake and calcium-channel blocker-based antihypertensive treatment in the Syst-Eur trial

OBJECTIVE

To assess the relationship between chronic intake of nonsteroidal anti-inflammatory drugs (NSAID) and outcome, in particular (gastrointestinal) bleeding and to investigate whether the effect of chronic NSAID intake was similar in untreated and treated elderly hypertensives.

METHODS

Eligible patients (> or = 60 years, with systolic blood pressure 160-219 mm Hg and diastolic blood pressure < 95 mm Hg) were randomised to active treatment or placebo. Active treatment consisted of nitrendipine, with the possible addition of enalapril, hydrochlorothiazide, or both, titrated or combined to reduce the sitting systolic blood pressure by at least 20 mm Hg to below 150 mm Hg. Patients never taking NSAIDs (n = 2882) were compared with patients on chronic NSAID intake (n = 861), defined as reporting NSAID intake on at least 50% of the patient forms.

RESULTS

There was a tendency towards lower mortality (relative hazard rate (95% confidence interval (CI), 0.77 (0.56-1.06)) and higher incidence of bleeding (1.13 (0.63-2.05) with chronic NSAID intake. Although there was no significant interaction between calcium-channel blocker (CCB)-based treatment and chronic NSAID intake for any of the end points, chronic NSAID intake tended to be associated with a lower incidence of bleeding on active treatment as compared to placebo (P-value of the interaction term = 0.07).

CONCLUSION

The effect of chronic NSAID intake on outcome was similar in patients on active treatment based on a dihydropyridine CCB or on placebo. However, chronic NSAID intake might have a less deleterious effect on bleeding on active treatment as compared to placebo.

Antihypertensive drugs and the risk of gastrointestinal bleeding

PURPOSE

Calcium channel blockers have been reported to increase the risk of gastrointestinal bleeding. We tested this hypothesis, and also assessed whether beta blockers decrease this risk.

SUBJECTS AND METHODS

A nested case-control design within a population-based cohort of all 34,074 new users of beta blockers, angiotensin-converting enzyme (ACE) inhibitors, or calcium channel blockers in Saskatchewan, from 1990 to 1993 and followed up to March 1995, was used. We identified all 311 subjects hospitalized because of gastrointestinal bleeding during this period, each of whom was matched to 10 randomly selected controls.

RESULTS

The rate of hospitalization for gastrointestinal bleeding was 3.0 per 1,000 per year. The adjusted rate ratio of gastrointestinal bleeding for current use of calcium channel blockers was 1.1 (95% confidence interval [CI] 0.8 to 1.4) and 0.66 (95% CI 0.44 to 0.98) for beta blockers compared with no current use of anti-hypertensive drugs. The adjusted rate ratio for ACE inhibitor use was 1.0 (95% CI 0.7 to 1.3) while that for diuretic use was 1.4 (95% CI 1.0 to 2.0).

CONCLUSIONS

The use of calcium channel blockers does not appear to increase the risk of gastrointestinal bleeding in the first five years of treatment, while beta blockers may prevent this adverse event. The unexpected elevated risk associated with the use of diuretics needs to be investigated further.

Upper gastrointestinal bleeding in relation to previous use of analgesics and non-steroidal anti-inflammatory drugs. Catalan Countries Study on Upper Gastrointestinal Bleeding

To assess the risk of upper gastrointestinal bleeding associated with the use of individual non-narcotic analgesics and non-steroidal anti-inflammatory drugs (NSAIDs), a multicentre study of 875 cases of upper gastrointestinal bleeding and 2682 hospital controls was done. With control for confounding factors, the overall odds ratio estimate for aspirin taken at least once during the week before the first symptom was 7.2 (95% confidence interval 5.4-9.6). Non-aspirin NSAIDs associated with upper gastrointestinal bleeding were diclofenac (7.9 [4.3-14.6]), indomethacin (4.9 [2.0-12.2]), naproxen (6.5 [2.2-19.6]), and piroxicam (19.1 [8.2-44.3]). Paracetamol, propyphenazone, and dipyrone did not increase the risk. A previous history of gastrointestinal bleeding or peptic ulcer did not greatly affect odds ratio estimates, which differed according to sex and were higher for younger than for older patients. However, the incidence of upper gastrointestinal bleeding was higher among the elderly.

Anaemia in rheumatoid arthritis: the role of iron, vitamin B12, and folic acid deficiency, and erythropoietin responsiveness

Thirty six patients with rheumatoid arthritis (RA) (25 with anaemia) were studied to establish the role of iron, vitamin B12, and folic acid deficiency, erythropoietin responsiveness, and iron absorption in the diagnosis and pathogenesis of anaemia in RA. Iron deficiency, assessed by stainable bone marrow iron content, occurred in 13/25 (52%), vitamin B12 deficiency in 7/24 (29%), and folic acid deficiency in 5/24 (21%) of the anaemic patients. Only 8/25 (32%) had just one type of anaemia. The iron deficiency of anaemia of chronic disease (ACD) was distinguished by ferritin concentration, which was higher in that group. Mean cell volume (MCV) and mean cell haemoglobin (MCH) were lower in both anaemic groups, but most pronounced in iron deficient patients. Folic acid, and especially vitamin B12 deficiency, masked iron deficiency by increasing the MCV and MCH. Iron absorption tended to be highest in iron deficiency and lowest in ACD, suggesting that decreased iron absorption is not a cause of ACD in RA. No specific causes were found for vitamin B12 or folic acid deficiency. Haemoglobin concentration was negatively correlated with erythrocyte sedimentation rate in the group with ACD. Erythropoietin response was lower in ACD than in iron deficient patients. It was concluded that generally more than one type of anaemia is present simultaneously in anaemic patients with RA. The diagnosis of each type may be masked by another. Studies on pathogenesis of the anaemia are difficult as deficiencies generally coexist with ACD. Disease activity and, possibly, erythropoietin responsiveness are major factors in ACD pathogenesis.

Anaemia in rheumatoid arthritis: pathogenesis, diagnosis and treatment

The pathogenesis, diagnosis and treatment of the anaemia of chronic disorders (ACD) in rheumatoid arthritis (RA) were reviewed. Causes of anaemia other than ACD frequently present in RA. Decreased iron absorption was shown to be the result of active RA rather than a cause of ACD or iron deficiency. It has been hypothesized that bone marrow iron availability decreases due to decreased iron release by the mononuclear phagocyte system or that the anaemia in ACD is due to ineffective erythropoiesis; these remain controversial theories. Studies considering a decreased erythropoietin responsiveness have not produced consistent results. Erythroid colony growth is suppressed in vitro by interleukins and tumour necrosis factor but their role in vivo in ACD is unknown. The diagnosis of ACD is made by exclusion. Iron deficiency is detected by transferrin, ferritin, and cellular indices after adaptation of their normal values. Treatment of the anaemia consists merely of antirheumatic treatment. Iron administration is counterproductive since iron chelators or exogenous erythropoietin administration might increase erythropoiesis.

[A fatality caused by heparin overdose]

Because of suspected pulmonary embolism 550.000 I.U. of heparin were administered instead of 55.000 I.U. of heparin within 24 h to a 62 year old patient. The man died because of cerebral hemorrhage the following day. Autopsy findings are reported and difficulties in medical expertise are discussed.

Ulcerogenic drugs and upper gastrointestinal bleeding

Aspirin and other NSAIDs are drugs for which the causal association with major gastrointestinal bleeding has not been adequately or conclusively demonstrated, although a certain degree of correlation is very likely. For aspirin ingestion in particular the increased risk is confined to patients taking the drug at heavy and regular dosages (less than 1% of users), and can be reduced further by the use of enteric-coated formulations. For non-aspirin NSAIDs, the relative risk of GI bleeding after repeated and prolonged exposure (in comparison to controls) has been quantified between 1.5 and 2.7, which is a small but significant figure, and it is increased by the age of the patients, by the duration of treatment and by the dose of drug. No consistent causal relationship can be found between major GI bleeding (or other major peptic ulcer complications) and steroids or other 'ulcerogenic' drugs. The therapy of drug-induced (or drug-associated) GI bleeding is probably not different from the usual treatment of upper GI haemorrhage. As far as the treatment of drug-associated gastroduodenal mucosal damage is concerned, it appears that with mucoprotective agents or H2 antagonists the healing rates of peptic ulcers is slower than observed in non-drug-associated disease. Prophylactic treatment with prostaglandins has only been proposed; and prophylactic treatment with H2 antagonists has been disappointing.

Clinical spectrum of the upper gastrointestinal effects of nonsteroidal anti-inflammatory drugs. Natural history, symptomatology, and significance

Nonsteroidal anti-inflammatory drugs (NSAIDs) relieve rheumatic pain and are in extensive use. Symptomatic complications of NSAIDs requiring the discontinuation of their use occur in 2 to 10 percent of patients with rheumatic diseases in sharp contrast to the common asymptomatic problems of gastroduodenal erosions, ulcerations, and bleeding, with resulting anemia in more than 40 percent of these patients. Opinions concerning the clinical significance of these complications are not uniform. The natural history of the effects of NSAIDs on the gastroduodenal mucosa reveals a sequence of initial subepithelial hemorrhage over a 24-hour period followed by gastroduodenal erosions and ulcerations in the next two weeks. From one week to three months, gastroduodenal erosions and ulcerations disappear in about half of the patients as an adaptation to continuing NSAID ingestion occurs. Hemorrhage may occur at any time in most patients and in a small minority (1 percent) it is massive. Non-aspirin NSAIDs (NANSAIDs) exhibit significantly fewer complications than do aspirin. These complications, however, demand considerable clinical attention and are ordered in a constant hierarchy, suggesting variable risks of complications among agents. NSAIDs are a blessing for those who have chronic pain, but that blessing does not prevent significant asymptomatic complications in the same patients. Prophylaxis for high-risk groups, such as women over the age of 65 years, should be subjected to study.

What's new in steroid and nonsteroid drug effects on gastroduodenal mucosa?

A survey is given on the damaging effect of acetylsalicylic acid, other nonsteroidal antiinflammatory drugs and corticosteroids on the gastroduodenal mucosa. The results of blood loss studies and endoscopic investigations are reviewed. Also, the histologic aspects of such damage are discussed. Modern concepts of the pathophysiology of these lesions stress the cytoprotective role of endogenous prostaglandins. Epidemiologic data strongly support an association between frequent and heavy intake of acetylsalicylic acid and gastric ulcer as well as gastrointestinal bleeding, whereas the association with duodenal ulcers is far less clearly established. Conclusive evidence is currently unavailable proving the superiority of other nonsteroidal antiinflammatory drugs in this regard. The ulcerogenic potency of corticosteroids at least in the small or medium dose range probably has been overstated in the past. Intensive ulcer therapy making use of H2 receptor antagonists often allows healing of small ulcers with a diameter up to 1 cm despite continued treatment with low dose corticosteroids or nonsteroidal antiinflammatory drugs, whereas continuation of these drugs is associated with very poor healing in ulcers larger than this size. The danger of perforation has to be taken into consideration.

Anti-rheumatic and analgesic drug usage and acute gastro-intestinal bleeding in elderly patients

The records of 1878 general medical in-patients (aged 75 years and over) at Hereford have been examined for associations between the commonly used antirheumatic and analgesic drugs and acute gastro-intestinal bleeding. One third of 93 in-patients admitted because of acute bleeding had been taking ibuprofen, indomethacin or aspirin. The usage of these drugs was significantly greater than in the control population who comprised the remaining 1785 elderly medical in-patients. It is suggested that very elderly patients taking any of these three drugs form a vulnerable high risk group who are liable to develop acute gastro-intestinal bleeding.

A controlled endoscopic study comparing the toxic effects of sulindac, naproxen, aspirin, and placebo on the gastric mucosa of health volunteers

Sixty volunteers were endoscopically evaluated to compare gastric mucosal injury following oral administration of sulindac, naproxen, aspirin, or placebo for two consecutive seven-day periods. A single-blind technique was utilized wherein the endoscopist was unaware which drug each volunteer had received. The following dosages were employed for the two study periods: sulindac, 150 and 200 mg, b.i.d., naproxen, 250 and 375 mg, b.i.d., and aspirin, 650 and 975 mg, q.i.d. The only subject who developed a frank ulcer with mucosal bleeding was in the sulindac group, however volunteers taking sulindac demonstrated statistically less significant mucosal injury on endoscopic examination than those receiving naproxen or aspirin.

Gastrointestinal blood loss of oxaprozin and aspirin with placebo control

The objective of this study was to compare the effects of oxaprozin (4,5-diphenyl-2-oxazolepropionic acid), a nonsteroidal, antiinflammatory compound, and aspirin in a double-blind, placebo-controlled study to estimate gastrointestinal bleeding. The determination of fecal blood loss was made quantitatively by the use of the radioactive (51Cr) technique. During the first week, subjects were controlled with and without placebo. At the end of the second week, the subjects were divided and randomly assigned to one of three groups; 10 received 1200 mg oxaprozin (600 mg twice daily), 11 received 3900 mg aspirin (975 mg four times a day), and the remaining 8 subjects received placebo for two weeks. During the last two weeks, all received placebo again. A statistical analysis of variance showed that there were no statistical differences between the groups during the first and last two weeks of placebo therapy. During the active treatment period, weeks 3 and 4, there were statistically significant differences among the three groups. The mean blood loss during week 3 was significantly greater for the aspirin group, 8.8 ml/day, than the oxaprozin group, 3.3 ml/day (P less than 0.05), and the placebo group, 1.4 ml/day (P less than 0.001). The smaller difference between oxaprozin and placebo was also significant (P less than 0.05). During the fourth week, the mean daily blood loss among oxaprozin patients had decreased to 2.3 ml/day, and no statistically significant difference from placebo (1.1 ml/day) was found.

Antiprostaglandin synthetase activity of nonsteroidal anti-inflammatory drugs and gastrointestinal micro-bleeding: a comparison of flurbiprofen with benoxaprofen

By using 51Cr-labelled erythrocytes and 141Ce-labelled microspheres to correct for daily variations in faecal output, gastrointestinal microbleeding was measured in hospital patients with rheumatoid arthritis before and after administration of 2 propionic acid derivatives. These were flurbiprofen and benoxaprofen, respectively a potent and a weak prostaglandin synthetase inhibitor. Increased faecal blood loss occurred consistently with flurbiprofen and not with benoxaprofen. Our results provide indirect evidence for the cytoprotective effect of prostaglandins on the gastrointestinal mucosa in man. The quantitative correlation between faecal occult blood loss as measured by the above techniques and the Haemoccult slide test is also discussed.

Gastrointestinal bleeding in patients with spinal cord trauma. Effects of steroids, cimetidine, and mini-dose heparin

The frequency and degree of gastrointestinal (GI) bleeding were examined in 131 patients with spinal cord injuries. All patients were randomly assigned to either high- or low-dose steroid regimens and some form of GI prophylaxis. The latter consisted of antacids alone or antacids supplemented with cimetidine when this medication became available. Segments of the population were treated with mini-dose or full-dose heparin. The incidence and degree of GI bleeding did not appear to be affected by steroid dose level, regimen of prophylaxis, or mini-dose heparin. Only full heparinization was found to significantly increase bleeding. These results place in question the benefits of adding cimetidine to antacids as a prophylactic atreatment in patients with no history of ulcer.

Interactions of aspirin with acetaminophen and caffeine in rat stomach: pharmacokinetics of absorption and accumulation in gastric mucosa

To study the pharmacokinetic interactions between aspirin (250 mg/kg) and simultaneously administered oral acetaminophen (125 mg/kg) or caffeine (50 mg/kg) in male rats, noninterfering GLC assays for these drugs were developed. Acetaminophen and caffeine both retarded the appearance of salicylate in plasma. During the elimination phase, acetaminophen enhanced plasms salicylate levels whereas caffeine did not. Aspirin reduced the plasms levels of both acetaminophen and caffeine during absorption and elimination. Regardless of whether the drugs had been administered separately or in combination, higher concentrations of salicylate, acetaminophen, and caffeine were found in the glandular part of the stomach compared to the nonglandular part (rumen). In both parts, the absorption of acetaminophen increased in the presence of aspirin. Simultaneous administration of aspirin with caffeine did not influence the absorption of either drug in the glandular and ruminal parts. The inhibitory action of acetaminophen and the potentiating action of caffeine on the erosive activity of aspirin are not due to any effects of these drugs on salicylate accumulation in glandular tissue.

Prevention of aspirin-induced faecal blood loss by prostaglandin E2

Prostaglandins have been shown in animal laboratory studies to be capable of protecting the gastrointestinal tract against injury by exogenous agents. This study was conducted to determine if prostaglandin E2 (PGE2), which is native to the human gastric mucosa, could influence the increase in faecal blood loss associated with the ingestion of aspirin (ASA). A randomised double-blind study was performed on 27 healthy men. Faecal blood loss was measured by the 51Cr labelled red cell technique. ASA (600 mg four times daily) caused a significant increase in faecal blood loss. PGE2 (1 mg four times daily) had no effect on faecal blood loss when administered alone. When given in addition to ASA it resulted in a faecal blood loss not significantly different from control. No significant alteration in intestinal transit occurred. It is concluded that PGE2 protects man from the gastrointestinal injury associated with ASA.

Gastrointestinal toxicity of minor analgesics

1 The evidence for the risk of gastric erosions from aspirin is fragmentary. 2 Occult gastric bleeding following aspirin is poorly studied and the skewed distribution is unexplained; platelet factors may be relevant. 3 Overt gastric bleeding may follow aspirin; the risk is probably about one episode per two million doses. 4 There is epidemiological, clinical, experimental and histopathological evidence for an association between chronic aspirin use and chronic gastric ulcer. 5 An alternative to the Davenport hypothesis is proposed to explain the gastric action of aspirin and the non-steroidal anti-inflammatory agents. 6 Paracetamol is probably bland in its gastric actions.