Page 301 - Computational Retinal Image Analysis

CHAPTER

Retinal biomarkers and

cardiovascular disease: 15

A clinical perspective

a
a
Carol Yim-lui Cheung , Posey Po-yin Wong , Tien Yin Wong b
a Department of Ophthalmology and Visual Sciences, The Chinese
University of Hong Kong, Shatin, Hong Kong
b Singapore Eye Research Institute, Singapore National Eye Center, Singapore, Singapore

1 Introduction
Cardiovascular disease (CVD), comprising mainly coronary heart disease and stroke,
remains a major cause of morbidity, disability and mortality worldwide [1, 2].
Current guidelines in the primary prevention of CVD recommend approaches to clas-
sify individuals as high, intermediate, or low risk using CVD risk prediction models
based on the traditional risk factors such as age, gender, race, hypertension, diabetes,
dyslipidemia and history of cigarette smoking. However, the risk prediction equa-
tions derived from cohorts established in the last century (e.g., Framingham risk
score model, Pooled Cohort Equations) may not be useful in contemporary popula-
tions, resulting in undertreatment or overtreatment of CVD risk factors [3–6]. Newer
cardiovascular biomarkers (e.g., C-reactive protein) have only provided modest im-
provements in predictive accuracy [3]. Therefore, new biomarkers that can improve
risk prediction and stratification are needed.
It is recognized that small vessel or microvascular pathology play a major role
in processes leading to the development of CVD events and its risk factors [7–9].
However, the microcirculation has been difficult to access, so robust microvascular
biomarkers have yet to be developed. The retinal vasculature, which is noninvasively
accessible, is a unique biological model to study microvascular abnormalities and
pathology associated with CVD [10–12]. Numerous large population-based stud-
ies have provided evidence that retinal vascular changes are associated with CVD
events (e.g., stroke, heart disease and CVD mortality) independent of traditional risk
factors. These studies support a concept that changes seen in the retinal vascula-
ture (e.g., retinal arteriolar narrowing, arterio-venous nicking) likely reflect simi-
lar changes in the systemic peripheral, cerebral and coronary microcirculation (e.g.,
vasoconstriction, intimal thickening). Thus, studying retinal vascular changes may
provide additional insights into the structure and function of the systemic microcir-
culation that are important in the development of CVD. Furthermore, retinal vascular

Computational Retinal Image Analysis. https://doi.org/10.1016/B978-0-08-102816-2.00016-2 299
© 2019 Elsevier Ltd. All rights reserved.