Evidence suggests no harmful effects of aspirin on AMD progression

Take-home message: Observational data from the AREDS2 study suggests that aspirin use is not associated with progression of age-related macular degeneration or development of geographic atrophy or neovascularization.

By Nancy Groves; Reviewed by Emily Y. Chew, MD

Bethesda, MD-Longitudinal assessment of the progression of age-related macular degeneration (AMD) in the Age-Related Eye Disease Study2 (AREDS2) demonstrated no harmful association with aspirin use, according to Emily Y. Chew, MD.

This finding emerged from analysis of AREDS2 data from 2006 to 2012 using the statistical matching technique of propensity scoring.

Although studies have explored the role of aspirin use in AMD for more than 25 years, several recent reports motivated the AREDS2 investigators to perform their own study, said Dr. Chew, deputy director of the Division of Epidemiology and Applications and the deputy clinical director of the National Eye Institute, National Institutes of Health, Bethesda, MD.

According to Dr. Chew, a 1988 study suggested that aspirin increases the risk of hemorrhage in patients with AMD, but that theory was countered by another study a year later finding no difference.

More recently, at least one clinic-based study showed that aspirin had a protective effect, whereas another suggested a harmful effect. However, large population-based studies on aspirin use have generated the most attention.

A report from the European Eye Study, published by de Jong et al in Ophthalmology in 2012, generated significant interest for its conclusion that daily aspirin use would worsen macular degeneration and increase the risk of progression to the neovascular form of the disease. This cross-sectional study included 4,691 patients over age 65; daily aspirin use was 17%.

Results showed that patients on aspirin had a greater risk of intermediate and large drusen and neovascular AMD, a finding that concerned many patients.

A second population-based study by Klein et al. also appeared in 2012 reporting recent findings from the ongoing Beaver Dam Eye Study, with 15 years of follow-up. Aspirin use was defined as two times a week for more than 3 months. The risk of neovascular AMD more than doubled in patients who had been taking aspirin regularly 10 years prior to an observed incidence.

AREDS2 Research Group

Dr. Chew reported the results on behalf of the AREDS2 Research Group.

The AREDS2 Research Group analyzed a group of 2,442 participants in the larger AREDS2 clinical trial with no baseline geographic atrophy or neovascular AMD.

Propensity scoring was used to estimate the probability of a participant using aspirin given covariates, Dr. Chew explained.

This statistical method reduces bias from confounding variables and matches aspirin and non-aspirin users so that they are comparable in other aspects. Logistic regression was used to calculate propensity scores for each patient, based on aspirin use at baseline, and participants were matched by propensity score. Hazard ratios were then calculated. The variables used to create the propensity score included age, gender, race, education, smoking, hypertension, diabetes, and angina.

Aspirin users were more likely to be older, male, non-white, less educated, be smokers, and have diabetes, hypertension, angina, and other cardiovascular diseases.

However, these differences were not statistically significant when participants were matched by propensity score. A difference in age of as much as 30 years emerged in persons with the same propensity scoring, so age was included in the model, Dr. Chew said.

Investigators were also concerned about death as a competing risk. Results for neovascular AMD without death as a competing risk showed that age was significant (p < 0.0001) but aspirin use (propensity score) was not (p = 0.3049); the hazard ratio was 0.60. With an adjustment for death as competing risk, the hazard ratio for aspirin use was 0.52 (p = 0.1834). Age was significant in these two analyses as well as all others.

Results for development of geographic atrophy showed a hazard ratio of 1.04 without death as a competing risk and when the data were adjusted for death. The hazard ratios for aspirin use were 1.31 (p = 0.5688) and 1.13 (p = 0.8032), respectively.

“Longitudinal assessment shows that there is no harmful association with aspirin use,” Dr. Chew said.

She added that randomized, controlled clinical trials of aspirin use generally show a protective though not statistically significant effect. Among population-based studies, only the Beaver Dam Eye Study and the European Eye Study are in contradiction of the AREDS2 findings.

“The totality of evidence-especially the information from the randomized clinical trials of aspirin-suggests that there are no harmful effects of aspirin associated with the progression of AMD,” Dr. Chew said.

“We suggest that individuals with AMD should consider aspirin therapy when medically indicated,” she added. “Especially now when we have good treatment for neovascular AMD, it’s much more important for these patients to take the aspirin that is required for their cardiovascular disease and other medical conditions that are of concern to them.”

Emily Y. Chew, MD

P: 301/496-6583

E: echew@nei.nih.gov

This article was adapted from Dr. Chew’s presentation during the 2014 meeting of the American Academy of Ophthalmology. Dr. Chew reported no financial interests or relationships.