Long-term aspirin use in older patients increases risk for major bleeding

JoAnn E. Manson, MD, DrPH

Decisions about the use of aspirin in either the primary or secondary prevention of CVD require an individualized benefit-risk assessment, which includes consideration of the patient’s risk for bleeding. The bleeding risk with aspirin increases dramatically with age (approximately doubling each decade), but these risks have not been given adequate attention. Gastrointestinal bleeding is a major potential complication of aspirin use that may alter the benefit-risk ratio of treatment and warrants consideration of proton-pump inhibitors or other preventive medication in selected patients.

The Oxford Vascular Study report fills an important gap by quantifying the bleeding risk on aspirin in individuals aged 75 and older. It highlights the very high risk for major upper gastrointestinal bleeds (> fourfold increase) and the alarmingly high risk (> tenfold increase) of disabling or fatal upper gastrointestinal bleeds in these older patients. Based on these statistics, they estimate the number needed to treat for 5 years to prevent one major upper gastrointestinal bleed would be less than 25 if patients aged 75 and older were routinely co-administered proton-pump inhibitors with aspirin.

Limitations of the study should be considered, however. This is an observational study, not a randomized clinical trial, and confounding bias (although likely to be minimal) cannot be excluded. But randomized trials to date are inadequate and have often excluded older patients. We need an improved understanding of the balance of benefits and risks of aspirin use in these older patients who are being prescribed aspirin in clinical settings, but have been underrepresented in most previous trials. We need to know whether routine co-prescription of proton-pump inhibitors would actually improve the benefit-risk ratio of aspirin use in this patient population, or whether individualized decision-making based on bleeding risk factors would be preferable. It is important to consider that proton-pump inhibitors are not free of risk and have been linked to C. difficile and other infections, reduced absorption of key nutrients and possible adverse effects on bone health and cognition.

Large-scale randomized trials will report results soon on the balance of benefits and risks of aspirin in older individuals. One such trial is ASPREE. These trials will help to inform clinical decision-making in the elderly. It will be important for professional societies responsible for developing clinical guidelines on aspirin use in the elderly to take the Oxford study’s findings into account and regularly update the guidelines as the results from ASPREE and other trials become available. These committees should also seriously consider the benefits and risks of long-term proton-pump inhibitor use in these patients and evaluate whether routine vs. selected use based on bleeding risk factors would be indicated.

Regarding the use of aspirin in primary prevention of atherosclerotic CVD, the 2016 U.S. Preventive Services Task Force guidelines did not recommend initiation of aspirin for primary prevention in patients aged 70 and older. The Task Force gave a Grade B recommendation (moderate certainty for net benefit) for the use of low-dose aspirin (81 mg per day) in adults age 50 to 59 years who have high baseline atherosclerotic CVD risk (10-year risk ≥ 10%), no increased risk of bleeding (eg, no recent bleeding, no recent history of gastrointestinal ulcers or use of medications that increase bleeding risk such as anticoagulant/antiplatelet agents) and meeting other specific criteria (Guirguis-Blake JM, et al. Ann Intern Med. 2016;doi:10.7326/M15-2113). For adults age 60 to 69 years meeting the above criteria, the Task Force gave a Grade C recommendation (not routinely recommended; individualize the decision); and for all adults less than 50 or 70 years or older, they considered the evidence to be insufficient (Grade I).

Until more detailed benefit-risk assessments have been conducted to evaluate a strategy of routine co-administration of proton-pump inhibitors in patients above a specific age cut-point, I would favor a strategy of individualized decision making based on the patient’s bleeding risk score. Several publications addressing risk factors for aspirin-related bleeding are available to assist with this risk assessment (Whitlock EP, et al. Ann Intern Med. 2016;doi:10.7326/M15-2112; Hernández-Díaz S, et al. BMC Medicine. 2006;doi:10.1186/1741-7015-4-22).

Although several risk factors for bleeding have been identified, including age, smoking, diabetes and hypertension (in addition to those highlighted above by the USPSTF), clinicians have received minimal guidance in calculating a bleeding risk score. Recognizing this need, we have developed a free mobile app called AspirinGuide, available for iPhone, iPad and Android devices (and also accessible at www.aspiringuide.org), that can be used to calculate both the patient’s 10-year atherosclerotic CVD risk and a personalized bleeding risk score (Mora S, et al. JAMA. 2016:doi:10.1001/jama.2016.8362). This clinical decision support tool helps clinicians weigh the probability of benefit from aspirin for CVD prevention against the potential harms. Patients who are at particularly high risk for upper gastrointestinal bleeding based on their bleeding risk score are likely to receive net benefit from proton-pump inhibitors, and reassessment would be advisable at regular intervals, such as every 3 to 5 years.