Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

Acetylsalicylic acid


See also Benorilate ; Diflunisal ; Lysine acetylsalicylate ; Salicylates, topical ; Salsalate

General information

Over a century after its introduction, acetylsalicylic acid (aspirin) is by far the most commonly used analgesic, sharing its leading position with the relative newcomer paracetamol (acetaminophen), and notwithstanding the fact that other widely used anti-inflammatory drugs, like ibuprofen and naproxen, have in recent years been introduced in over-the-counter versions. Both are also still being prescribed by physicians and are generally used for mild to moderate pain, fever associated with common everyday illnesses, and disorders ranging from head colds and influenza to toothache and headache. Their greatest use is by consumers who obtain them directly at the pharmacy, and in many countries outside pharmacies as well. Perhaps this wide availability and advertising via mass media lead to a lack of appreciation by the lay public that these are medicines with associated adverse effects. Both have at any rate been subject to misuse and excessive use, leading to such problems as chronic salicylate intoxication with aspirin, and severe hepatic damage after overdose with paracetamol. Both aspirin and paracetamol have featured in accidental overdosage (particularly in children) as well as intentional overdosage.

In an investigation of Canadian donors who had not admitted to drug intake, 6–7% of the blood samples taken were found to have detectable concentrations of acetylsalicylic acid and paracetamol [ ]. Such drugs would be potentially capable of causing untoward reactions in the recipients.

To offer some protection against misuse of analgesics, many countries have insisted on the use of packs containing total quantities less than the minimum toxic dose (albeit usually the one obtained for healthy young volunteers and thus disregarding the majority of the population), and supplied in child-resistant packaging. Most important, however, is the need to provide education for the lay public to respect such medicines in general for the good they can do, but more especially for the harm that can arise but which can be avoided. There is a definite role for the prescribing physician, as informing the patient seems to prevent adverse events [ ].

The sale of paracetamol or aspirin in dosage forms in which they are combined with other active ingredients offers considerable risk to the consumer, since the product as sold may not be clearly identified as containing either of these two analgesics. Brand names sometimes obscure the actual composition of older formulations that contain one or both of these analgesics in combination with, for example, a pyrazolone derivative and/or a potentially addictive substance. For instance, in Germany, with the EC harmonization of the Drug Law of 1990, the manufacturers of drugs already marketed before 1978 had the opportunity of exchanging even the active principles without being obliged to undergo a new approval procedure or to abandon their brand name. Combination formulations are still being promoted and sold, and not exclusively in developing countries. Consequently, the patient who is so anxious to allay all his symptoms that he takes several medications concurrently may without knowing it take several doses of aspirin or paracetamol at the same time, perhaps sufficient to cause toxicity. It is essential that product labels clearly state their active ingredients by approved name together with the quantity per dosage form [ ].

The antipyretic analgesics, with the non-steroidal anti-inflammatory drugs (NSAIDs), share a common mechanism of action, namely the inhibition of prostaglandin synthesis from arachidonic acid and their release. More precisely their mode of action is thought to result from inhibition of both the constitutive and the inducible isoenzymes (COX-1 and COX-2) of the cyclo-oxygenase pathway [ ]. However, aspirin and paracetamol are distinguishable from most of the NSAIDs by their ability to inhibit prostaglandin synthesis in the nervous system, and thus the hypothalamic center for body temperature regulation, rather than acting mainly in the periphery.

Endogenous pyrogens (and exogenous pyrogens that have their effects through the endogenous group) induce the hypothalamic vascular endothelium to produce prostaglandins, which activate the thermoregulatory neurons by increasing AMP concentrations. The capacity of the antipyretic analgesics to inhibit hypothalamic prostaglandin synthesis appears to be the basis of their antipyretic action. Neither aspirin nor paracetamol affects the synthesis or release of endogenous pyrogens and neither will lower body temperature if it is normal.

While aspirin significantly inhibits peripheral prostaglandin and thromboxane synthesis, paracetamol is less potent as a synthesis inhibitor than the NSAIDs, except in the brain, and paracetamol has only a weak anti-inflammatory action. It is simple to ascribe the analgesic activity of aspirin to its capacity to inhibit prostaglandin synthesis, with a consequent reduction in inflammatory edema and vasodilatation, since aspirin is most effective in the pain associated with inflammation or injury. However, such a peripheral effect cannot account for the analgesic activity of paracetamol, which is less well understood.

As a prostaglandin synthesis inhibitor, aspirin, like other NSAIDs, is associated with irritation of and damage to the gastrointestinal mucosa. In low doses it can also increase bleeding by inhibiting platelet aggregation; in high doses, prolongation of the prothrombin time will contribute to the bleeding tendency. Intensive treatment can also produce unwanted nervous system effects (salicylism).

Depending on the criteria used, the incidence of aspirin hypersensitivity is variously estimated as being as low as 1% or as high as 50%, the highest frequency being found in asthmatics. The condition is characterized by bronchospasm (asthma), urticaria, angioedema, and vasomotor rhinitis, each occurring alone or in combination, often leading to severe and even life-threatening reactions. There is no clear evidence of an association with tumors, apart from the possible peripheral contribution of aspirin to the development of urinary tract neoplasms in patients with analgesic nephropathy. Indeed, some authors have suggested a role for salicylates in reducing the incidence of colorectal tumors and breast tumors.

The following are absolute contraindications to the use of aspirin:

  • children under 16;

  • people with hypersensitivity to salicylates, NSAIDs, or tartrazine;

  • people with peptic ulceration;

  • people with known coagulopathies, including those induced as part of medical therapy.

The following are relative contraindications to the use of long-term analgesic doses of aspirin:

  • gout, since normal analgesic doses impede the excretion of uric acid (high doses have a uricosuric effect); an additional problem in gout is that salicylates reduce the uricosuric effects of sulfinpyrazone and probenecid;

  • variant angina; a daily dose of 4 g has been found to provoke attacks both at night-time and during the day [ , ], perhaps owing to direct triggering of coronary arterial spasm; blockade of the synthesis of PGI z , which normally protects against vasoconstriction, could be involved;

  • diabetes mellitus, in which aspirin can in theory interfere with the actions of insulin and glucagon sufficiently to derange control;

  • some days before elective surgery (even in coronary artery bypass grafting) or delivery, especially if extradural anesthesia is used [ ], although recent data seem reassuring [ ]; aspirin increases bleeding at dental extraction or perioperatively;

  • in elderly people, who may develop gastrointestinal bleeding;

  • anorectal inflammation (suppositories);

  • pre-existing gastrointestinal disease, liver disease, hypoalbuminemia, hypovolemia, in the third trimester of pregnancy, perioperatively, or in patients with threatening abortion.

Assessing the benefit-to-harm balance of low-dose aspirin in preventing strokes and heart attacks

Although there is clear evidence of benefit of acetylsalicylic acid (aspirin) in secondary prevention of strokes and heart attacks, the question of whether aspirin should also be prescribed for primary prevention in asymptomatic people is still debatable. Trials in primary prevention have given contrasting results [ , ], and aspirin can cause major harms (for example severe gastrointestinal bleeding and hemorrhagic stroke).

Furthermore, despite evidence of the efficacy of aspirin in secondary prevention, its use in patients at high risk of strokes and heart attacks remains suboptimal [ ]. A possible explanation for this underuse may be concern about the relative benefit in relation to the potential risk for serious hemorrhagic events. Accurate evaluation of the benefits and harms of aspirin is therefore warranted.

Two meta-analyses have provided some information. The first examined the benefit and harms of aspirin in subjects without known cardiovascular or cerebrovascular disease (primary prevention) [ ]. The authors selected articles published between 1966 and 2000: five large controlled studies of primary prevention that lasted at least 1 year and nine studies of the effects of aspirin on gastrointestinal bleeding and hemorrhagic stroke. The five randomized, placebo-controlled trials included more than 50 000 patients and the meta-analysis showed that aspirin significantly reduced the risk of the combined outcome (confirmed non-fatal myocardial infarction or death from coronary heart disease) (OR = 0.72; 95% CI = 0.60, 0.87). However, aspirin increased the risk of major gastrointestinal bleeding (OR = 1.7; CI = 1.4, 2.1) significantly, while the small increase found for hemorrhagic stroke (OR = 1.4; CI = 0.9, 2.0) was not statistically significant. All-cause mortality was not significantly affected (OR = 0.93; CI = 0.84; 1.02). Most important was the finding that the net effect of aspirin improved with increasing risk of coronary heart disease. The meta-analysis showed that for 1000 patients with a 5% risk of coronary heart disease events over 5 years, aspirin would prevent 6–20 myocardial infarctions but would cause also 0–2 hemorrhagic strokes and 2–4 major gastrointestinal bleeds. For patients at lower risk (1% over 5 years), aspirin would prevent 1–4 myocardial infarctions but would still cause 0–2 hemorrhagic strokes and 2–4 major gastrointestinal bleeds.

Therefore when deciding to use aspirin in primary prophylaxis, one should take account of the relative utility of the different outcomes that are prevented or caused by aspirin.

The other meta-analysis [ ] compared the benefits of aspirin in secondary prevention with the risk of gastrointestinal bleeding. An earlier analysis of this problem included patients at various levels of risk and doses of aspirin that would currently be regarded as too high [ ], and may therefore have either under-represented the benefit or exaggerated the risk. In another analysis there was no difference in the risk of gastrointestinal bleeding across the whole range of doses used [ ].

The meta-analysis reviewed all randomized, placebo-controlled, secondary prevention trials of at least 3-months duration published from 1970 to 2000. The dosage of aspirin was 50–325 mg/day. Six studies contributed 6300 patients to the analysis (3127 on aspirin and 3173 on placebo). Aspirin reduced all-cause mortality by 18%, the number of strokes by 20%, myocardial infarctions by 30%, and other vascular events by 30%. On the other hand, patients who took aspirin were 2.5 times more likely than those who took placebo to have gastrointestinal tract bleeds. The number of patients needed to be treated (NNT) to prevent one death from any cause was 67 and the NNT to cause one gastrointestinal bleeding event was 100. In other words 1.5 lives can be saved for every gastrointestinal bleed attributed to aspirin. Although the risk of gastrointestinal bleeding was increased by aspirin, the hemorrhagic events were manageable and led to no deaths. On the basis of these data we can conclude that the benefits–harm balance for low-dose aspirin in the secondary prevention of cardiovascular and cerebrovascular events is highly favorable. The same conclusions have been drawn from the systematic overview published by the Antithrombotic Trialists Collaboration Group, which analysed data from 287 studies involving 135 000 patients [ ].

As far as primary prevention of cardiovascular events is concerned, it appears that aspirin can reduce heart attacks and strokes but increases gastrointestinal and intracranial bleeding. The decision to use aspirin in primary prevention should therefore take into account the fact that the net effect of aspirin improves with increasing risk of coronary heart disease as well as the values that patients attach to the main favorable and unfavorable outcomes.

Organs and systems

Cardiovascular

Apart from rare reports of variant angina pectoris and vasculitis theoretically related to thromboxane, aspirin is not associated with adverse effects on the cardiovascular system [ , ], except an increase in circulating plasma volume after large doses.

The effects of aspirin on blood pressure have been investigated in 100 untreated patients with mild hypertension who took aspirin on awakening or before bedtime [ ]. There was no change in blood pressure after dietary recommendations alone or when aspirin was given on awakening. However, there was a highly significant reduction in blood pressure in those who took aspirin before bedtime (reductions of 6 and 4 mmHg in systolic and diastolic blood pressures respectively). As aspirin is given once a day for its cardioprotective effect, giving it in the evening could be of greater benefit if it also results in a reduction in blood pressure.

Respiratory

The effect of aspirin on bronchial musculature is discussed in the section on Immunologic in this monograph.

Salicylates can cause pulmonary edema, particularly in the elderly, especially if they are or have been heavy smokers [ ].

Chronic salicylate toxicity can cause pulmonary injury, leading to respiratory distress. Lung biopsy may show diffuse alveolar damage and fibrosis [ ].

Nervous system

Salicylism is a reaction to very high circulating concentrations of salicylate, characterized by tinnitus, dizziness, confusion, and headache.

Encephalopathy secondary to hyperammonemia has been reported in those rare cases of liver failure that are associated with high doses of aspirin, and this also forms a major feature of Reye’s syndrome (see the section on Liver in this monograph).

One case–control study showed no increased risk of intracerebral hemorrhage in patients using aspirin or other NSAIDs in low dosages as prophylaxis against thrombosis [ ]. However, intracerebral hemorrhage has been reported with aspirin, even in low doses, and in the SALT study [ ] and the Physicians Health Study of 1989 [ ] hemorrhagic stroke and associated deaths occurred with aspirin. In a study in 501 946 Chinese subjects, there was a 1.33-fold (95% CI = 1.13, 1.55) increased risk of major hemorrhagic events during short-term use of low-dose aspirin [ ].

In 208 subjects with intracerebral hemorrhage the 3-month mortality was 33% [ ]. The independent risk factors for death were regular aspirin use at the onset of intracerebral hemorrhage (RR = 2.5; 95% CI = 1.3, 4.6), warfarin use at the onset of intracerebral hemorrhage (RR = 3.2; 95% CI = 1.6, 6.1), and an intracerebral hemorrhage score over 2 on admission (RR = 14; 95% CI = 6.0, 31). Regular aspirin use (median dose 250 mg/day) preceding the onset of intracerebral hemorrhage was significantly associated with hematoma enlargement during the first week after intracerebral hemorrhage.

Sensory systems

Vision

Well-documented acute myopia and increased ocular pressure attributed to aspirin has been described [ ].

Auditory function

With the high concentrations achieved in attempted suicide, tinnitus and hearing loss, leading to deafness, develop within about 5 hours, usually with regression within 48 hours, but permanent damage can occur. Disturbed balance, often with vertigo, can develop, as well as nausea, usually with maintenance of consciousness, even without treatment. It has been postulated that in this state depolarization of the cochlear hair cells occurs, similar to the changes induced by pressure. Tinnitus is also a symptom of salicylism.

Aspirin has been reported to cause damage to the semicircular canals.

  • A 61-year-old man with a monoclonal gammopathy developed severe persistent bilateral vestibular dysfunction after taking a high dose of aspirin (5–6 g/day for 3 days) [ ]. His symptoms (unsteadiness, a broad-based gait, blurred vision, and apparent visual motion when he moved his head and when he walked) persisted for 9 months. Investigations showed a bilateral dynamic deficit of his horizontal semicircular canal.

Metabolism

Aspirin lowers plasma glucose concentrations in C-peptide-positive diabetic subjects and in normoglycemic persons [ ]. This is of no clinical significance.

Fluid balance

NSAIDs can cause fluid retention, but this has rarely been reported with aspirin.

  • Severe fluid retention, possibly due to impaired renal tubular secretion, has been reported in a 29-year-old woman taking aspirin (1.5 g/day for several days) for persistent headache [ ]. During rechallenge with aspirin (0.5 g tds for 3 days) a dynamic renal scintigram showed a substantial fall in tubular filtration. Withdrawal was followed by complete uneventful recovery.

Pulmonary edema is a feature of salicylate intoxication, but this patient was taking a therapeutic dosage.

Hematologic

Thrombocytopenia, agranulocytosis, neutropenia, aplastic anemia, and even pancytopenia have been reported in association with aspirin. The prospect for recovery from the latter is poor, mortality approaching 50%.

Hemolytic anemia can occur in patients with glucose-6-phosphate dehydrogenase deficiency or erythrocyte glutathione peroxidase deficiency [ ]. Whether these reports have anything more than anecdotal value [SEDA-17, 97] is not known.

Simple iron deficiency caused by occult blood loss occurs with a frequency of 1%, and upper gastrointestinal bleeding resulting from regular aspirin ingestion is the reason for hospitalization in about 15 patients per 100 000 aspirin users per year. Aspirin causes bleeding of sufficient severity to lead to iron deficiency anemia in 10–15% of patients taking it continuously for chronic arthritis. Some individuals are particularly at risk because of pregnancy, age, inadequate diet, menorrhagia, gastrectomy, or malabsorption syndromes.

Macrocytic anemia associated with folate deficiency has been described in patients with rheumatoid arthritis [ ] and also in patients who abuse analgesic mixtures containing aspirin [ ].

Effects on coagulation

Aspirin in high doses for several days can reduce prothrombin concentrations and prolong the prothrombin time. This will contribute to bleeding problems initiated by other factors, including aspirin’s local irritant effects on epithelial cells. It is therefore very risky to use aspirin in patients with bleeding disorders. The effect will contribute to increased blood loss at parturition, spontaneous abortion, or menorrhagia, and may be linked to persistent ocular hemorrhage, particularly in older people, with or without associated surgical intervention [ , ].

By virtue of its effects on both cyclo-oxygenase isoenzymes, aspirin inhibits platelet thromboxane A 2 formation. This effect in the platelet is irreversible and will persist for the lifetime of the platelet (that is up to 10 days), since the platelet cannot synthesize new cyclo-oxygenase. It is of clinical significance that the dose of aspirin necessary to inhibit platelet thromboxane A 2 (around 40 mg/day) is much lower than that needed to inactivate the subendothelial prostacyclin (PGI 2 ). Hence, platelet aggregation is inhibited, with some associated dilatation of coronary and cerebral arterioles, at doses that do not interfere with prostacyclin inhibition. It is important, in considering the dosage of aspirin for prophylaxis (see below), to appreciate that prostacyclin is a general inhibitor of platelet aggregation, while aspirin, as a cyclo-oxygenase inhibitor, affects aggregation from a limited number of stimuli, for example ADP, adrenaline, thromboxane A 2 . It is also worth recalling that the vascular endothelium can synthesize new cyclo-oxygenase, so that any effect on prostacyclin synthesis is of limited duration only [SEDA-12, 74] [ ].

Several long-term studies have been carried out since the 1980s to determine the prophylactic usefulness of these effects on clotting. It is now clear that aspirin in dosages of around 300 mg/day can be used successfully for secondary prophylaxis in patients with coronary artery disease, in order to reduce the incidence of severe myocardial infarction, and in patients with cerebrovascular disease to reduce the incidence of transient ischemic attacks and strokes. There is some suggestion that higher doses of aspirin may be required in women. A major drawback has been the high incidence of gastrointestinal adverse effects and particularly bleeding in aspirin-treated groups [ , , , ]. In view of the age group involved, bleeding can have serious implications. In an attempt to avoid this high proportion of ill-effects and yet retain the benefits of prophylactic antithrombotic treatment, a few trials have been conducted using aspirin in a dose of 162 mg (ISIS-2) [ ] and 75 mg (RISC) [ ] in symptomatic coronary heart disease, with good evidence of efficacy. Two studies have been reported in patients with cerebrovascular events, namely the Dutch TIA trial with aspirin 30 versus 283 mg [ ] and the SALT study with aspirin 75 mg [ ]. The former did not show any difference in efficacy between the 30 and 283 mg dose groups, but there was no placebo control. The latter study showed a significant reduction in thrombotic stroke. However, intracerebral hemorrhage has been reported with aspirin, even in low doses, and in this as well as in the Physicians Health Study of 1989 [ ], hemorrhagic stroke and associated deaths occurred with aspirin. On the other hand, the incidence of serious gastrointestinal events was much lower than previously described.

In 711 patients, of whom 320 were taking aspirin at the time of surgical resection for cutaneous head and neck lesions, the incidence of significant postoperative hemorrhage was 1.6% (5 cases) versus none in the control group; the use of aspirin was the only susceptibility factor for significant postoperative hemorrhage [ ].

However, in other studies there was no effect of aspirin on blood loss in patients with hip fractures [ ], after coronary artery bypass surgery [ ], or after tooth extraction [ ].

Relatively few patients developed a prolonged bleeding time while taking aspirin or other NSAIDs and only few had significant intraoperative blood loss. There is variation in the response of patients for unknown reasons and so the recommendation that NSAIDs should be withdrawn before elective surgery awaits confirmation [SEDA-19, 96].

The EIDOS and DoTS classifications of hemorrhage due to salicylates are shown in Figure 1 .

Figure 1, The EIDOS and DoTS descriptions of stress hemorrhage due to salicylates

Gastrointestinal

Gastric erosion and hemorrhage

The gastrointestinal adverse effects of aspirin and the other NSAIDs are the most common. While some argue against a causative relation between aspirin ingestion and chronic gastric ulceration, the current consensus favors such a relation, while admitting that other factors, such as Helicobacter pylori , are likely to play a part. Patients aged over 65 years and women are more at risk, as are those who take aspirin over prolonged periods in a daily dose of about 2 g or more.

However, there is no ambiguity about the association of aspirin with gastritis, gastric erosions, or extensions of existing peptic ulcers, all of which are demonstrable by endoscopy. Even after one or two doses, superficial erosions have been described in over 50% of healthy subjects. This association is now almost universally accepted as the standard basis for comparative testing of NSAIDs and other drugs [ , ]. Whether it is of benefit to use other drugs concomitantly to prevent the effect of gastric acid on the mucosa, and thus reduce the risk of gastric ulceration, is discussed further in this monograph.

Dyspepsia, nausea, and vomiting occur in 2–6% of patients after aspirin ingestion. Patients with rheumatoid arthritis seem to be more sensitive, and the frequency of aspirin-induced dyspepsia in this group is 10–30% [SEDA-9, 129]. However, these symptoms are generally poor predictors of the incidence of mucosal damage [SEDA-18, 90].

The bleeding that occurs is usually triggered by erosions and aggravated by the antithrombotic action of aspirin. While it is reported to occur in up to 100% of regular aspirin takers, bleeding tends to be asymptomatic in young adults, unless it is associated with peptic ulceration, but it is readily detectable by endoscopy and the presence of occult blood in the feces. Hematemesis and melena are less often seen, the odds ratio being 1.5–2.0 in an overview of 21 low-dose aspirin prevention studies [ ]. A degree of resultant iron deficiency anemia is common. Such events are more commonly seen in older people in whom there is a significant proportion of serious bleeding and even deaths. Major gastrointestinal bleeding has an incidence of 15 per 100 000 so-called heavy aspirin users. However, the interpretation of “heavy” and of quantities of aspirin actually taken is to a large extent subjective and very dependent on the questionable accuracy of patient reporting. The risk appears to be greater in women, smokers, and patients concurrently taking other NSAIDs, and is possibly affected by other factors not yet established [ ]. Gastrointestinal perforation can occur without prodromes. Aspirin increases the risk of major upper gastrointestinal bleeding and perforation two- to three-fold in a dose-related manner, but deaths are rare.

Incidence

Of the estimated annual 65 000 upper gastrointestinal emergency admissions in the UK, nearly 20% (including deaths in 3.4%) are attributable to the use of prostaglandin synthesis inhibitors [ ]. As might be expected with an inhibitor of prostaglandin synthesis, the cytoprotective effects of prostaglandin E and prostacyclin (PGI 2 ) are reduced by aspirin, as is the inhibitory action on gastric acid secretion. This effect may be both direct, as is the case with aspirin released in the stomach (or the lower rectum in the case of aspirin suppositories), and indirect following absorption and distribution via the systemic circulation; attempts to reduce the problems by coating and buffering can therefore have only limited success. The indirect type of effect is shown by the fact that these adverse gastric effects can also be exerted by parenteral lysine acetylsalicylate [SEDA-10, 72]. The local effects depend in part on the tablet particle size, solubility, and rate of gastric absorption, while the most important variable appears to be gastric pH. On the other hand, within-day changes in the pharmacokinetics of the analgesic compounds may be involved in the prevalence of gastrointestinal adverse effects.

The estimates of gastrointestinal complication rates from aspirin are generally derived from clinical trials [SEDA-21, 100]. However, the applicability of the results of such trials to the general population may be debatable, as protocols for these studies often are designed precisely to avoid enrollment of patients who are at risk of complications. Indeed differences in benefit-to-harm balance have been found in trials using the same dose of aspirin [ , ]. For this reason, a population-based historical cohort study on frequency of major complications of aspirin used for secondary stroke prevention may be of interest [ ]. The study identified 588 patients who had a first ischemic stroke, transient ischemic attack, or amaurosis fugax during the study period. Of these, 339 patients had taken aspirin for an average of 1.7 years. The mean age of patients who had taken aspirin was 74 years. Complications occurred within 30 days of initiation of treatment in one patient, between 30 days and 6 months in 10 patients, between 6 months and 1 year in seven patients, and between 1 year and 2 years in two patients. Estimated standardized morbidity ratio of gastrointestinal hemorrhage (determined on the basis of 10 observed events and 0.661 expected events, during 576 person-years of observation) was 15 (95% CI = 7, 28). The estimated standardized morbidity ratio of intracerebral hemorrhage (determined on the basis of only one event and 0.59 expected events) was 1.7 (CI = 0.04, 9.4). One patient had a fatal gastrointestinal hemorrhage. Unfortunately these complication rates must be considered estimates, because aspirin therapy was not consistently recorded. However, the rates of complications were similar to those observed in some randomized clinical trials. On the basis of these data and of those of a meta-analysis of 16 trials involving more than 95 000 patients [ ], the overall benefits of aspirin, measured in terms of preventing myocardial infarction and ischemic stroke, clearly outweigh the risks.

Dose-relatedness

The question of whether the risk of gastrointestinal hemorrhage with long-term aspirin is related to dose within the usual therapeutic dosage range [SEDA-12, 100] [ , ] merits attention. In a meta-analysis of the incidence of gastrointestinal hemorrhage associated with long-term aspirin and the effect of dose in 24 randomized, controlled clinical trials including almost 66 000 patients exposed for an average duration of 28 months to a wide range of different doses of aspirin (50–1500 mg/day), gastrointestinal hemorrhage occurred in 2.47% of patients taking aspirin compared with 1.42% taking placebo (OR = 1.68; 95% CI = 1.51, 1.88). In patients taking low doses of aspirin (50–162.5 mg/day; n = 49 927), gastrointestinal hemorrhage occurred in 2.3% compared with 1.45% taking placebo (OR = 1.56; 95% CI = 1.40, 1.81). The pooled OR for gastrointestinal hemorrhage with low-dose aspirin was 1.59 (95% CI = 1.4, 1.81). A meta-regression to test for a linear relation between the daily dose of aspirin and the risk of gastrointestinal hemorrhage gave a pooled OR of 1.015 (95% CI = 0.998, 1.047) per 100 mg dose reduction. The reduction in the incidence of gastrointestinal hemorrhage was estimated to be 1.5% per 100 mg dose reduction, but this was not significant.

In other studies the incidence of upper gastrointestinal hemorrhage has been reported to be similar in patients taking either 75 mg or 325 mg of aspirin per day [ , ].

These data are in apparent contrast with others previously reported [SEDA-21, 100] [ ], which showed that gastrointestinal hemorrhage was related to dose in the usual dosage range. Many reasons may explain these contrasting results, the most important being differences in the definition of the hemorrhagic events, in study design, in the population studied, and in the presence of accessory risk factors [ ].

The trend toward the use of lower doses of aspirin has been driven by the belief that these offer a better safety profile while retaining equivalent therapeutic efficacy. Despite the large number of patients enrolled in randomized clinical trials and included in meta-analyses, there is no firm evidence that dose reduction significantly lowers the risk of gastrointestinal bleeding. Patients and doctors therefore need to consider the trade-off between the benefits and harms of long-term treatment with aspirin. Meanwhile, it seems wise to use the lowest dose of proven efficacy.

A systematic review of 17 epidemiological studies conducted between 1990 and 2001 has provided further data on this topic [ ]. The effect of aspirin dosage was investigated in five studies. There was a greater risk of gastrointestinal complications with aspirin in dosages over 300 mg/day than in dosages of 300 mg/day or less. However, users of low-dose aspirin still had a two-fold increased risk of such complications compared with non-users, with no clear evidence of a dose–response relation at dosages under 300 mg/day, confirming previous findings [ ]. The study also addressed the question of whether the aspirin formulation affects gastrotoxicity. The pooled relative risks of gastrointestinal complications in four studies were 2.4 (95% CI = 1.9, 2.9) for enteric-coated aspirin, 5.3 (3.0, 9.2) for buffered formulations, and 2.6 (2.3, 2.9) for plain aspirin, compared with non-use. These data confirm those from previous studies [SEDA-21, 100] [ ], which negate any protective effect of the most frequently used aspirin formulations. Furthermore, there were higher relative risks, compared with non-use, for gastrointestinal complications in patients who used aspirin regularly (RR = 3.2; CI = 2.6, 5.9) than in patients who used it occasionally (2.1; 1.7, 2.6), and during the first month of use (4.4; 3.2, 6.1) compared with subsequent months (2.6; 2.1, 3.1).

Comparative studies

A comparative study of gastrointestinal blood loss after aspirin 972 mg qds for 4 days versus different doses of piroxicam (20 mg od, 5 mg qds, and 10 mg qds) showed that piroxicam did not increase fecal blood loss, whereas aspirin did. Gastroscopic evidence of irritation was also greater with aspirin [ ].

In a randomized trial comparing ticlopidine (500 mg/day) with aspirin (1300 mg/day) for the prevention of stroke in high-risk patients, the incidence of bleeding was similar in both groups, although more patients treated with aspirin developed peptic ulceration or gastrointestinal hemorrhage [ ].

Susceptibility factors

A study of the susceptibility factors for gastrointestinal perforation, a much less frequent event than bleeding, has confirmed that aspirin and other NSAIDs increase the risk of both upper and lower gastrointestinal perforation (OR 6.7, CI 3.1–14.5 for NSAIDs) [ ]. Gastrointestinal perforation has been associated with other factors, such as coffee consumption, a history of peptic ulcer, and smoking. The combination of NSAIDs, smoking, and alcohol increased the risk of gastrointestinal perforation (OR 10.7, CI 3.8–30) [SEDA-21, 97].

The risks of adverse gastrointestinal effects of aspirin have been studied in relation to susceptibility factors using two major databases the General Practice Research Database in the UK and the Base de Datos para la Investigación Farmacoepidemiológica en Atención Primaria in Spain [ ]. The rates of upper gastrointestinal adverse effects varied depending on age, sex, the use of NSAIDs, and the presence of upper gastrointestinal pain or peptic ulceration. The highest rate was in men aged 80 years and over with complicated ulcers taking NSAIDs (300/1000 person years); the lowest rate was in women not taking NSAIDs and with no other susceptibility factors (0.8/1000 person-years).

Associated effects

Aspirin can also play a role in esophageal bleeding, ulceration, or benign stricture, and it should be considered as a possible cause in patients, particularly the elderly, who present with any of these features. There have also been reports of rectal stricture in the elderly, associated with the use of aspirin suppositories. Effects on both these strictures emphasize the significance of a direct local action of aspirin as well as a systemic action and underlines the relevance of the involvement of oxygen-derived free radicals in the pathogenesis of mucosal lesions in the gastrointestinal tract [ ].

A gastrocolic fistula developed in a 47-year-old woman taking aspirin and prednisone for rheumatoid arthritis [ ]. Other similar case reports have been published [ , ].

Long-term effects

The effects on the stomach of continued exposure to aspirin remain controversial. While in short-term use, gastric mucosal erosions may often be recurrent but transient and comparatively trivial lesions, with longer administration there seems to be an increased risk of progression to ulceration.

Prophylaxis

Enteric-coated aspirin has been associated with gastroduodenal ulcer formation; the enteric coating has been shown to be toxic to the bowel and it is postulated that it is also toxic to the stomach [ ].

Intravenous administration, or the use of enteric-coated formulations or modified-release products all appear to reduce the risk both of bleeding and more particularly of erosions/ulceration. However, because of the indirect effect noted above, such formulations do not eliminate the risk, although they may reduce the incidence of gastric or duodenal ulcer, as may buffered aspirin [ , ].

Considerable attention has been directed toward the efficacy of using synthetic forms of PGE 2 , histamine H 2 receptor antagonists, proton pump inhibitors, or antacids, either to heal peptic ulcers associated with use of prostaglandin inhibitors or more significantly to act prophylactically to protect against ulceration or bleeding associated with aspirin or the NSAIDs. With the exception of PGE 2 analogues, there is no convincing evidence to justify their prophylactic use, as they do not reduce the risk of significant gastrointestinal events. In contrast, their soothing effect on gastrointestinal symptoms may ultimately result in more severe complications [ ]. Since all these agents carry their own potential risks, it is more than questionable whether administration to a patient with normal gastrointestinal mucosa is justified. Generally, use of prostaglandin inhibitors should be limited to the shortest possible duration, thereby minimizing, but not eliminating, the risk of gastrointestinal damage. Only high-risk patients should be eligible for prophylactic drug therapy. Well-known risk factors for the development of mucosal lesions of the gastrointestinal tract are age (over 75 years), a history of peptic ulcer, or gastrointestinal bleeding, and concomitant cardiac disease.

Liver

Aspirin can cause dose-related focal hepatic necrosis that is usually asymptomatic or anicteric. Much of the evidence for hepatotoxicity of aspirin and the salicylates has been shown in children [ , ], usually in patients with connective tissue disorders, taking relatively high long-term dosages for Still’s disease, rheumatoid arthritis, or occasionally systemic lupus erythematosus. Rises in serum transaminases seem to be the most common feature (in up to 50% of patients) and are usually reversible on withdrawal, but they occasionally lead to fatal hepatic necrosis. Severe and even fatal metabolic encephalopathy can also occur, as in Reye’s syndrome (see the section on Reye’s syndrome in this monograph). One can easily overload the young patient’s individual metabolic capacity. The co-existence of hypoalbuminemia may be a particular risk factor; in patients with hypoalbuminemia of 35 g/l or less, close monitoring of the aspartate transaminase is advisable, especially if the concentration of total serum salicylate is 1.1 mmol/l or higher [ ]. Plasma salicylate concentrations in serious cases have usually been in excess of 1.4 mmol/l and liver function tests return rapidly to normal when the drug is withdrawn. Finally, a very small number of cases of chronic active hepatitis have been attributed to aspirin [ ].

Reye’s syndrome

First defined as a distinct syndrome in 1963, Reye’s syndrome came to be regarded some years later as an adverse effect of aspirin. In fact, the position is more complex, and the syndrome still cannot be assigned a specific cause. There is general agreement that the disorder presents a few days after the prodrome of a viral illness. Well over a dozen different viruses have so far been implicated, including influenza A and B, adenovirus, Varicella , and reovirus. Various other factors have also been incriminated, including aflatoxins, certain pesticides, and such antioxidants as butylated hydroxytoluene. Only in the case of aspirin have some epidemiological studies been conducted, and these appeared to show a close correlation with cases of Reye’s syndrome. It was these studies that led to regulatory action against the promotion of salicylate use in children. However, doubt has been thrown on the clarity of the link, and it now seems increasingly likely that while there is some association with aspirin, the etiology is in fact multifactorial, including some genetic predisposition. Studies in Japan did not support the US findings, while studies in Thailand and Canada invoked other factors.

Two characteristic phenomena are present in Reye’s syndrome.

  • 1.

    Damage to mitochondrial structures, with pleomorphism, disorganization of matrix, proliferation of smooth endoplasmic reticulum, and an increase in peroxisomes; mitochondrial enzyme activity is severely reduced, but cytoplasmic enzymes are unaffected. The changes first appear in single cells, but may spread to all hepatocytes. Recovery may be complete by 5–7 days. While these changes are most evident in liver cells, similar effects have been seen in cerebral neurons and skeletal muscle. There appears to be a block in beta-oxidation of fatty acids (inhibition of oxidation of NAB-linked substrates). In vitro aspirin selectively inhibits mitochondrial oxidation of medium- and long-chain fatty acids.

  • 2.

    An acute catabolic state with hypoglycemia, hyperammonemia, raised activities of serum aspartate transaminase and creatine phosphokinase, and increased urinary nitrogen and serum long chain dicarboxylic acid.

Despite our lack of understanding of the syndrome, the decision taken in many countries to advise against the use of salicylates in children under 12 made an impact, in terms of a falling incidence of Reye’s syndrome [SEDA-16, 96; SEDA-17, 97].

In the USA, the incidence of Reye’s syndrome has fallen significantly—from the time that the advice was introduced up to 1999 there were 25 reported cases, but 15 were in adolescents aged 12–17 years, and 8% of cases occurred in patients aged 15 years or over [ ]. In the UK, in view of these findings, the Commission on Safety of Medicines (CSM) amended its original statement and advised that aspirin should be avoided in febrile illnesses or viral infections in patients aged under 16 years. However, the appropriateness of this decision has been challenged [ ]. This is because the incidence of Reye’s syndrome is already low and is falling; furthermore, restricting the use of aspirin leaves paracetamol and ibuprofen as the only available therapeutic alternatives, and their safety is not absolutely guaranteed and might be even worse than that of aspirin.

Pancreas

There are conflicting findings in the literature regarding the possibility that long-term use of aspirin is associated with an increased risk of pancreatic cancer [ ]. New data from a recent study have suggested that extended periods of regular aspirin use appear to be associated with a statistically significant increased risk of pancreatic cancer among women [ ]. However, the results of this study were inconsistent and require confirmation.

Urinary tract

Aspirin is associated with a small but significant risk of hospitalization for acute renal insufficiency [SEDA-19, 95].

When aspirin is used by patients on sodium restriction or with congestive heart failure, there tends to be a reduction in the glomerular filtration rate, with preservation of normal renal plasma flow. Some renal tubular epithelial shedding can also occur.

Severe systemic disease involving the heart, liver, or kidneys seems to predispose the patient to the effects of aspirin and other NSAIDs on renal function [ ].

In 106 elderly in-patients aspirin 100 mg/day for 2 weeks reduced creatinine clearance and uric acid clearance significantly in 70% and 62% of the patients respectively, with mean reductions of 19% and 17% [ ]. After withdrawal of aspirin renal function improved, but 67% of the patients were left with some impairment in creatinine clearance. Those who reacted adversely to aspirin had significantly better pre-study renal function, and lower hemoglobin and serum albumin concentrations.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here

Related Posts