Improving the gastrointestinal tolerability of aspirin in older people.

Interventions to reduce mortality and disability in older people are vital. Aspirin is cheap and effective and known to prevent cardiovascular and cerebrovascular disease, many cancers, and Alzheimer dementia. The widespread use of aspirin in older people is limited by its gastrointestinal side effects. Understanding age-related changes in gastrointestinal physiology that could put older people at risk of the side effects of aspirin may direct strategies to improve tolerance and hence lead to greater numbers of older people being able to take this effective intervention.

Introduction

As the population ages, clinical and cost effective strategies to increase lifespan whilst reducing disability are becoming a major public health agenda. The cyclooxygenase inhibitor aspirin has the potential to be such a strategy, as it is a simple, cheap intervention that is increasingly recognized as preventing a wide range of diseases associated with significant morbidity, and as a result will ensure healthy life into older age.

Evidence for the benefits of aspirin is now so overwhelming that strategies to deliver it to as many older people as possible are a health promotion priority. Many cancers, strokes, myocardial infarctions, and cases of Alzheimer dementia might be prevented if more older people could benefit from the risk reductions of aspirin.

Evidence supporting the benefits of aspirin

Aspirin is one of the most frequently prescribed medications for older people (Scomerville et al 1986). It is taken to control inflammation in arthritis, but more importantly has been shown in meta-analyses to reduce events in those at risk from cardiovascular or cerebrovascular disease (Antiplatelet Trialists Collaborative 2002; Hayden et al 2002). There is also emerging evidence that aspirin use protects against Alzheimer dementia and a wide range of cancers including breast, esophageal, prostate, colorectal, and lung carcinomas (Dubois et al 1998; Smith et al 2000; Bardou et al 2004).

Mathematical modeling confirms that the beneficial effects of aspirin are potentially so great that encouraging low-dose aspirin use in all 50-year-old subjects would reduce disability and double the chances of living a healthy life into old age (Morgan 2003). Many studies are now suggesting that the benefits of long-term low-dose aspirin outweigh the risks (Eidelman et al 2003). In the elderly, these benefits are to an extent similar to, and often greater than, that observed in younger age groups (Dornbrook et al 2003). However, the studies necessary to acquire such data in older age groups will require huge numbers such as that seen in the ongoing Aspirin in Reducing Events in the Elderly (ASPREE) study with 15000 subjects aged over 70 being followed for 5 years (Nelson et al 2003).

Despite the acknowledged benefits of aspirin, a cross-sectional population survey found only 7.1% were taking aspirin as a primary preventative measure (Trinder et al 2003). Furthermore, of those prescribed aspirin for secondary prevention, 8% were no longer taking it at 6-month follow up, presumably because of side effects (Eagle et al 2004).

Gastrointestinal adverse effects of aspirin

The widespread use of aspirin by older people has historically been limited as many develop abdominal side effects. Almost 50% of those prescribed aspirin for secondary prevention report gastrointestinal symptoms after just 2 weeks of use (Laheij et al 2001; Niv et al 2005) and almost one-third of aspirin users have endoscopically visible lesions within one hour of ingestion (Hawkey et al 1991; Cole et al 1999). Symptoms are recognized as a poor predictor for gastrointestinal lesions with 48% of asymptomatic aspirin users having lesions visible at endoscopy.

Aspirin can lead to adverse gastrointestinal effects ranging from dyspepsia with endoscopically normal gastric mucosa, asymptomatic and symptomatic lesions such as erosions and ulcers, and complications of ulcers including bleeding and perforation. Although these gastrointestinal effects are dose dependant, even lower doses of aspirin are being increasingly recognized as a cause of gastrointestinal bleeding (Stack et al 2002).

It is controversial, however, whether simply being old makes you more susceptible to aspirin-induced gastrointestinal damage or whether comorbidity, comedications, and past history are more important predictors of toxicity than age and perhaps more relevant to therapeutic decision making in this population (Solomon and Gurwtiz 1997). Risk factors for aspirin-induced gastrointestinal complications are shown in Table 1.

Table 1

Risk factors for aspirin-induced gastrointestinal complications

• Advancing age

• Female sex

• History of peptic ulcer disease

• Type and dose of NSAID

• Duration of use

• Use of combinations of NSAIDs

• Concomitant use of drugs such as steroids or anticoagulants

Adapted from Aalykke et al (2001); Gallerani et al (2004).

Developing strategies to improve tolerability of aspirin

Gastrointestinal side effects of aspirin occur more frequently in older people (Aalykke 2001). Therefore strategies to improve tolerability might be directed in two ways: at those specific physiological abnormalities that identify individuals who are less able to tolerate aspirin irrespective of their age, and at age-related changes in gastrointestinal physiology that might predict why older people tolerate aspirin less well compared with younger age groups (Table 2). This review will focus primarily on this second strategy.

Table 2

Potential strategies to improve tolerability of aspirin in older people

Helicobacter pylori eradication
Coprescription:
–With PPI
–With prostaglandin analog
Improve mucosal protective mechanisms
–Reverse the age-associated decline in mucus thickness
–Improve secretion of mucosal protective molecules, eg,TFF2
Reduce contact time by reversing the age-associated decline in gastric emptying

Abbreviations: PPI, proton pump inhibitor;TTF2, trefoil factor family 2.

Strategies to improve tolerability: the eradication of Helicobacter pylori

Many changes in gastrointestinal physiology once thought to be primary effects of aging have been reexamined since the discovery of the microorganism Helicobacter pylori (Kateralis et al 1993; Feldman et al 1996). Infection with H. pylori itself induces changes in gastrointestinal physiology, which is of relevance when it is appreciated that in the Western world infection rates increase with age, with up to 80% of 80-year-old subjects infected (Marshall 1994).

Both aspirin use and H. pylori infection cause peptic ulcers, but whether the incidence is greater when both are present is unclear (Voutilainen et al 2001). H. pylori and aspirin are independent risk factors for ulceration in all age groups (Lanas et al 2002), however, studies specifically involving older people suggest that there may be a synergistic effect on risk (Ng et al 2000; Seinela and Ahvenainen 2000).

Low doses of aspirin induced endoscopically visible upper gastrointestinal mucosal damage more frequently in H. pylori positive subjects (50%) compared with 16% of H. pylori negative volunteers (Feldman et al 2001). Furthermore, eradication of H. pylori reduces damage caused by low doses of aspirin and recurrence of ulcers during aspirin use (McCarthy 1998) and improves adaptability of the gastrointestinal tract to aspirin (Konteruk et al 1997, 1998). Unfortunately, H. pylori eradication will not always improve aspirin tolerability, as gastrointestinal symptoms, ulcers, and their complications are associated with aspirin use in those with, and without, H. pylori infection (Seinela and Ahvenainen 2000).

Despite this, it would seem pragmatic to recommend eradication of H. pylori infection prior to commencing long-term aspirin treatment. Whether or not this approach is appropriate in all older people, irrespective of symptoms, or only in those with gastrointestinal symptoms or a history of peptic ulceration (NICE 2005) requires large studies in older people and should include evaluation of aspirin tolerance after H. pylori eradication (Figure 1).

Figure 1

Proposed algorithm to improve gastrointestinal tolerance of aspirin in older people.

Abbreviations: PPI, proton pump inhibitor.

Structural changes with age in the upper gastrointestinal tract could affect tolerance of aspirin

Gastrointestinal transit time, in particular gastric emptying, is slower in older people (Brogna et al 1999), which potentially increases exposure of the gastric mucosa to ingested drugs. Such direct toxic effects provide the rationale for the use of enteric-coated aspirin. Theoretically, influencing the time of exposure to, or the formulation of, aspirin preparations, or reversing the age-associated decline in gastrointestinal transit time, may influence the ability of older people to tolerate aspirin.

Atrophy of the gastric mucosa incidence increases with age (James 2000) partly because of the increased prevalence of H. pylori in older people. Gastric atrophy results in smaller volumes of less acidic gastric juice in the stomach lumen. This reduced ability to dilute ingested drugs will potentially increase the risk of direct gastrointestinal toxic side effects. Further study is required to determine whether widespread H. pylori eradication programmes would decrease gastric atrophy prevalence in older people and allow better tolerance of aspirin.

Improving tolerance of aspirin by addressing age-related changes in gastrointestinal physiology

There is limited research examining age-related changes in the human upper gastrointestinal tract that might explain why older people tolerate aspirin less well. What is currently known about changes in physiology in the older stomach and duodenum is summarized in Figure 2. An understanding of these changes in gastric physiology with age could direct interventions that lead to improved tolerance.

Figure 2

Age-related changes in gastrointestinal physiology

In the human gastrointestinal tract there is a balance between aggressive factors (gastric acid and pepsin) and mucosal protective mechanisms (mucus and bicarbonate). Current evidence does not suggest that the increase in dyspeptic symptoms or ulceration in older people taking aspirin is related to an age-related increase in the aggressive factors: gastric acid or pepsin. Gastric acid secretion may be reduced in older people due to the increase in gastric atrophy, and pepsin output is also lower (Feldman et al 1996). However, age-related deficiencies in the ability of the mucosa to protect and repair itself have been documented, and any additional depletion due to medication such as aspirin will further increase mucosal vulnerability (Guslandi et al 1999).

Gastrointestinal mucosal protective molecules such as prostaglandins decline with advancing age (Cryer, Lee, et al 1992; Cryer, Redfern, et al 1992; Goto et al 1992; Lee and Feldman 1994). Prostaglandins stimulate protective mechanisms such as mucus and bicarbonate whilst aspirin inhibits prostaglandin production and causes gastric damage (Sababi et al 1995). The lower levels of prostaglandins in the gastric mucosa of older people makes them more susceptible to damage by a further reduction in prostaglandin synthesis caused by aspirin.

The first line of mucosal protection from exogenous toxins and luminal acid and pepsin is the mucus gel layer. Studies have shown both a quantitative reduction in mucus production with age and impaired quality of the mucus (Corfield et al 1993; Farinati et al 1993; Newton, Jordan et al 2000), and as a result an increased susceptibility to damage by aspirin (Corfield et al 1993).

The ability of the gastric mucosa to protect itself by repelling toxins is independently decreased in association with H. pylori induced gastritis and NSAID use (Goddard et al 1987; Spychal et al 1990) and also with aging (Hackelsberger et al 1998). It is not clear whether these factors work synergistically.

There is also an age-related decline in the ability of the gastrointestinal mucosa to neutralize luminal acid by bicarbonate secretion (Kim et al 1990; Lee 1997; Feldman et al 1998; Guslandi et al 1999). The older stomach is also less able to repair itself after damage (Lee et al 1998; Liu et al 1998), which when taken with the antithrombotic effects of aspirin may account for gastrointestinal side effects. Control of the repair processes is poorly understood but does have the potential to be manipulated (Majumdar et al 1997). Trefoil proteins are a family of mucosal repair proteins thought to be important in gastrointestinal protection (May and Westley 1997). Trefoil factor family 1 (TFF1)and TFF2 are expressed in the stomach and TFF2 and TFF3 are expressed in the duodenum. TFF1 is intimately associated with gastric mucus (Newton, Allen, et al 2000) and gastric concentrations of TFF2 show a circadian variation (Semple et al 2001) increasing dramatically at night. Recent work by our group has shown that the nocturnal peak of mucosally protective TFF2 is lower and earlier in older people (Johns et al 2005), and manipulating the nocturnal increase of cytoprotective TFF2 or encouraging patients to take aspirin at night when the mucosal protective mechanisms are optimal may improve tolerability.

In animals, a reduction in basal gastric blood flow (Lee 1996) and an attenuation of gastric blood flow in response to injury has been observed with age (Gronbech and Lacy 1994; Miyake et al 1996), though this is controversial (Taha 1993). If there are reductions in gastric blood flow in older people, coprescription of the vasodilator nitric oxide with aspirin (NO-NSAID) may have potential (Wallace et al 1995).

Strategies known to reduce gastrointestinal complications of aspirin in older people

Studies carried out specifically in older age groups have shown in both acute and chronic users of aspirin, that concomitant use of a proton pump inhibitor (PPI) reduces the risk of gastrointestinal bleeding (odds ratio [OR] 1.12 chronic, 1.05, acute) (Pilotto et al 2003) and risk of peptic ulcer disease (Pilotto et al 2004).

This benefit from PPI coprescription with aspirin is greater than with either the prostaglandin analog misoprostol (OR 1.91) and H2 blockers (OR 2.26). This has prompted the recommendation that PPI cotreatment is advisable in symptomatic older patients who need treatment with aspirin. However, whether a similar approach is appropriate for primary prevention of complications in asymptomatic older people requiring aspirin needs study.

In addition, although the studies confirm that both PPI and misoprostol are effective at reducing the risk of ulcer recurrence (Goldstein et al 2004), this does not guide clinical practice when confronted with a patient who simply develops gastrointestinal symptoms (rather than complications) after commencing aspirin.

Some have suggested that in high-risk patients with aspirin-associated peptic ulcer disease, conversion to other antiplatelet therapies such as clopidogrel may be appropriate. However, studies suggest that complication rates with clopidogrel are no better than continuation of aspirin alone (Ng et al 2004), with studies confirming that aspirin plus PPI is superior to clopidogrel in the prevention of recurrence (Chan et al 2005).

Patients taking long-term, low-dose aspirin who have had ulcer complications respond to acid suppressive treatments such as a PPI after eradication of H. pylori (Lai et al 2002), but eradication alone may be superior to the use of a PPI (Chan et al 2001). It should be remembered that coprescription of acid suppressive treatments with aspirin to improve tolerability in older people is unlikely to be the whole answer as physiologically there is no age-related increase in acid.

Future directions

Damage to the gastrointestinal mucosa is related to aspirin dose (Moore et al 1989, 1991) and lower doses of aspirin have fewer side effects (Serrano et al 2002). Therefore, it may be that some aspirin is better than no aspirin, and studies in older age groups would determine whether smaller doses of aspirin could be tolerated and give some, if not all, of the benefits obtained with larger doses in younger age groups. In the past, treatment of those who developed adverse side effects was terminated. The evidence for the use of aspirin is now so overwhelming that we need to consider how to give some aspirin to as many people as possible.

There are huge implications for allowing greater numbers of older people to benefit from taking aspirin for prevention of cardiovascular and cerebrovascular disease, Alzheimer dementia, and cancer. Defining the underlying age-related changes in physiology that make the older gastrointestinal tract susceptible to damage will identify targets for therapy. Reducing deficiencies in mucosal repair proteins, such as growth factors and trefoil proteins, may improve the reparative process. Delineating the pathways of action of these molecules in older people is therefore important. Conceptually simple strategies such as reducing the prevalence of gastric atrophy or reversing the age-associated decline in gastric emptying may also be effective.

Conclusion

Aging of the stomach and duodenum is an important but poorly understood area of gerontology. In older people, the major causes of mortality and morbidity are cardiovascular and cerebrovascular diseases, gastrointestinal cancers, and dementia. Understanding how aspirin reduces risk in these diseases and whether or not older people are intrinsically at risk of side effects because of age-related changes in gastrointestinal physiology should allow greater numbers of older people to benefit from the risk reductions associated with taking aspirin. Research in this area could be translated directly into the clinical setting and potentially make a real impact upon the quality of life of older people.