Link Between Neovascular Eye Disease, CVD Tracked to Retinal LesionsJuly 16, 2022
NEW YORK CITY — Subretinal drusenoid deposits (SDD) had a significant association with underlying cardiovascular disease (CVD), adding a missing link between age-related macular degeneration (AMD) and CVD, a prospective study showed.
Among 47 patients with CVD and AMD, 40 (85%) had SDD with or without subretinal drusen. An analysis aided by artificial intelligence showed that the combination of SDD and HDL-cholesterol <62 mg/dL predicted high-risk CVD with 87% accuracy. The same type of analysis showed that homozygosity for the age-related maculopathy susceptibility gene ARMS2 and underlying CVD predicted presence of SDD with 94% specificity, reported R. Theodore Smith, MD, PhD, of Mount Sinai Health System in New York City.
The data make a compelling case to evaluate patients with SDD for undiagnosed high-risk CVD, he stated in a presentation at the American Society of Retina Specialists meeting. The findings were published simultaneously in Retina.
“In clinical practice, once all this has been formulated and codified, it is possible that we could have these fairly inexpensive retinal cameras available through[out] the medical world so that patients can be screened by them as part of their routine medical care,” Smith told MedPage Today. “Then they can be told immediately whether or not they have these high-risk features. That means they should have further work-up — let’s say an echocardiogram, carotid Doppler — then take it to the next level. Find out what may be hiding.”
Both large, soft drusen and SDDs contribute to development of neovascular AMD and macular atrophy. However, the lesions differ with respect to their systemic associations, locations, genetics, cholesterol content, and prognosis, said Smith. More than a decade ago, research showed an association between SDD and premature mortality, but the precise nature of the relationship remained unclear.
Investigators hypothesized that high-risk CVD — including myocardial infarction (MI), valvular disease, and internal carotid artery stenosis (ICA) — are strongly associated with SDDs but not drusen. They further hypothesized high-risk CVD leads to poor choroidal perfusion.
To investigate the hypothesis, Smith and colleagues prospectively identified 97 patients with SDD, with or without drusen, and 103 with drusen only. Data collected included lipid profile and selected genetic assessment, as well as demographics and medical history.
The 200 patients included 47 individuals with high-risk CVD. Comparison with SDD status showed the following breakdown:
- MI: 19 (16 with SDD)
- Valvular disease: 17 (14)
- ICA: 11 (10)
The findings translated into an odds ratio for high-risk CVD of 9.0 for the association with SDD (95% CI 4.0-22.9, P=0.9-8). A multivariate analysis identified HDL <62 mg/dL as predictive of high-risk CVD in patients with SDD (P=0.0002) and ARMS2 homozygosity as predictive of SDD (P=0.01).
Smith showed selected images that reflected the findings. An older man with myocardial insufficiency had high myopia and confluent SDD with no evidence of drusen. A patient with severe aortic stenosis and hyperopia had no atherosclerosis but a confluence of SDD and no drusen.
A third example represented what Smith called an “experiment in nature.” The patient’s left internal carotid had a 40% stenosis but a clear right internal carotid. The left eye had “loads of confluent subretinal deposits” whereas the right eye had none. Cortical thickness was elevated in the left eye and normal in the right.
“If anyone has an alternate explanation other than perfusion, let me know,”said Smith.
The results are interesting and come from a group of investigators who have studied the association between SDD and CVD for quite awhile, Raj Maturi, MD, of the Midwest Retina Institute in Indianapolis, told MedPage Today via email. The findings involved a fairly modest number of patients and require validation in a larger cohort, noted Maturi, who was not involved in the study.
Maturi asked Smith about the small study population, given the group’s large patient base and long history of investigation. Smith said the size of the study population was limited by the financial resources available to conduct the study.
“We had a couple of fellows doing their best to get patients, and if you want to recruit a patient, it takes a couple of hours to get through all the hoops,” said Smith. “There is a lot of talk about real-world experience; well, this is real life.”
The study was supported in part by Regeneron.
Smith disclosed relationships with MacRegen and Ora.