Retinal vessel pathology and ocular disease burden in patients with cardiovascular disease: the Australian Heart Eye Study

Abstract

Background: A series of population-based studies across the world over the last half century have confirmed that retinal microvascular signs can predict clinical coronary heart disease. There is also increasing recognition that coronary microvascular dysfunction may play a role in coronary heart disease. Existing studies, however, lack a substantial proportion of individuals with coronary artery disease (CAD) within their respective study samples, and rarely assess CAD using objective indices of disease extent and severity as quantified by coronary angiography. Extant literature also focuses more on the role of retinopathy and retinal vessel calibre, whereas this thesis also turns its attention to methods of quantifying the retinal vascular architecture using newer retinal vessel geometric measures like Df, curvature tortuosity, and branching angle. The rationale for the use of these retinal vessel geometric measures is that they add to the growing array of non-invasive research tools to probe the microcirculation. In this way, they serve as an excellent surrogate for the systemic microcirculation, and could even be used as a prediction tool for cardiovascular disease.1 There is growing interest in the means by which coronary microvascular dysfunction play a role in cardiovascular disease, particularly in women. People with symptoms of acute coronary syndrome who undergo coronary catheterisation and angiography but are found to have minimal quantitative evidence for CAD (cardiac syndrome X) also pose a diagnostic dilemma. This group has been assumed to have coronary microvascular dysfunction, but confirming this diagnosis has been difficult due to the lack of non-invasive modalities to image the coronary microcirculation. The retina provides exactly such a means of visualising the microcirculation, affording an in vivo window into the structure and function of the human circulation and its role in cardiovascular disease pathology. Thus, this thesis attempts to explore this unique link between retinal pathology with CAD and other cardiovascular disease, as well as to examine the association between common ocular conditions and CAD. Objectives: (i) To determine the prevalence of retinal vessel measures (retinal arteriolar narrowing and venular widening), as well as common ocular conditions, in a unique clinical sample of patients with or at high risk of CAD; (ii) to assess the associations between CAD, metabolic syndrome, and hypertension with retinal microvascular signs/ocular conditions — including retinal vessel measures, age-related macular degeneration (AMD), epiretinal membrane (ERM), and cataract. The end goal was to identify new modifiable risk factors for CAD and other systemic vascular disease, as well as retinal and other ocular pathology. Methods: The Australian Heart Eye Study (AHES) is a cross-sectional observational study that surveyed 1680 participants who presented to a tertiary referral hospital for the evaluation of potential CAD by coronary angiography. The objective of the study was to evaluate the associations of retinal microvascular signs with angiographically confirmed CAD. The candidate received training in Singapore in order to utilise a semi-automated computer-assisted program developed by the National University of Singapore and Singapore Eye Research Institute known as Singapore ‘I’ Vessel Assessment (SIVA version 1.0) to quantitatively assess a range of retinal vascular geometric measures from digital fundus photographs. With the assistance of another trained grader, the candidate graded the AHES photographs using SIVA by following a standardised protocol, masked to patient characteristics.2, 3 Fractal dimension (Df) was calculated from skeletonised line tracings using a box-counting method, which divided each photograph into a series of squares of various side lengths.4 Df was defined as the gradient of logarithms of the number of boxes and the size of those boxes.5, 6 The more complex the branching pattern, the greater the Df. Curvature tortuosity was derived from the integral of the curvature square along the path of the vessel, normalised by the total path length.7 This took into account bowing and points of inflection,8 in contrast with simple tortuosity, which fails to distinguish between increased length due to bowing and that due to multiple points of inflection.9 The straighter the vessel, the lower the tortuosity value.8 Retinal arteriolar tortuosity and retinal venular tortuosity were thus a measure of the average tortuosity of the arterioles and venules in the eye, respectively. Retinal vascular branching angle was defined as the first angle subtended between two daughter vessels at each vascular bifurcation.8, 10 Retinal arteriolar branching angle and retinal venular branching angle quantify the average branching angles of arterioles and venules of the eye, respectively.4 Retinal vessel calibre measures were also obtained using retinal grading software.11, 12 Average retinal arteriolar and venular calibres were calculated using the Knudtson-Hubbard formula and presented as central retinal arteriolar equivalent (CRAE) or central venular equivalent (CRVE), respectively.13 A combined retinal score was constructed to attempt to reflect the joint effect of multiple retinal vessel parameters on CAD using those variables that were most strongly significant in multivariate analysis — Df, arteriolar curvature tortuosity, and retinal arteriolar calibre. Each of these variables were considered in binary terms (above or below the median), giving a total of eight possible combinations of these variables. Those combinations with all three variables above their respective medians were assigned a combined retinal score of 0, while those combinations with all three variables below their respective medians were assigned a combined score of 2. All other combinations were assigned a score of 1. AMD is the leading cause of blindness and low vision in older adults.14 The presence of early and late AMD was determined using the Wisconsin AMD Grading System.15 Early AMD prevalence was defined as the absence of late AMD and presence of either (i) large (0.125 mm diameter) indistinct soft or reticular drusen, or (ii) both large distinct soft drusen and retinal pigmentary abnormalities (hyperpigmentation or hypopigmentation). Similarly, late AMD prevalence was defined as the presence of either neovascular or atrophic AMD in that eye. Neovascular AMD was defined as presence of serous or haemorrhagic detachment of the retinal pigment epithelium (RPE) or sensory retina, presence of subretinal or sub-RPE haemorrhage, or subretinal fibrosis. Atrophic AMD was defined as a discrete area, at least 175 µm in diameter, of retinal depigmentation characterised by a sharp border and presence of visible choroidal vessels.16 “Any AMD” prevalence was defined as the presence of either early or late AMD.17 The classification and grading system for ERM was the same as in the baseline Blue Mountains Eye Study (BMES-1),18 adopted from Klein et al.19 Two types of ERMs were identified: a more severe form, termed preretinal macular fibrosis (PMF), in which superficial retinal folds and traction lines were identified; and a less severe form termed cellophane macular reflex (CMR), without visible retinal folds. Eyes with both CMR and PMF present were classified as having PMF. As quantitative data on cataract was not available from the AHES, cataract surgery prevalence, as obtained from a detailed questionnaire, was used instead as a marker variable for cataract. CAD was quantified using objective scoring systems based on the severity and extent of coronary artery stenosis, as assessed from coronary angiography. The coronary artery segments were defined using the Syntax system, which divides the arterial tree into 16 segments, based on the modified American Heart Association (AHA) classification.20 For each segment, the severity of obstruction was documented using several grades: normal, 1-25%, 25-50%, 50-74%, 75-99% and 100% (occluded). Each lesion that was visually scored as greater than 50% luminal obstruction in a vessel that was ≥1.5mm diameter was further analysed using quantitative coronary analysis (QCA). The specific parameter used to quantify CAD are described in further detail in the Methodology chapter (2.1). Metabolic syndrome was defined as per the Third Report of the National Cholesterol Education Program (NCEP) Adult Treatment Panel,26 please see Chapter 2.1 for further details. Results: Retinal vessel calibre Persons with metabolic syndrome (compared to without) had narrower retinal arteriolar calibre in multivariate analysis (mean difference 4.3 µm, p<0.0001). Similarly, those with hypertension (compared to without) had narrower retinal arteriolar calibre in multivariate analysis (mean arteriolar calibre difference 2.1 µm, p=0.02). This association was present among persons both with and without CAD (mean difference 5.0 µm, p=0.04). Stratification by sex indicated that women with hypertension had narrower retinal arterioles compared to normotensive women (multivariable-adjusted p=0.04). Retinal vascular geometric measures Retinal vascular Df and curvature tortuosity decreased with increasing age; women had significantly lower Df than men (p<0.003). Straighter retinal vessels were associated with CAD extent and Gensini scores in multivariate analysis (p<0.02). In sex-stratified multivariate analysis, straighter arterioles and narrower venular branching angles were associated with greater odds of overall CAD in men, while straighter venules were associated with CAD in women. Accounting for media opacity by sub-group analysis in pseudophakic patients, the combined retinal score was associated with stenosis greater than 50% in any coronary artery segment (vessel score) and obstructive coronary stenosis in all three main coronary arteries (segment score) (p=0.01). Lower Df and narrower arteriolar branching angle were associated with CAD vessel score (p<0.03). Other ocular conditions Prevalence of early and late AMD was 5.8% (n = 86) and 1.4% (n = 21), respectively. After multivariable adjustment, patients with obstructive coronary stenosis in all three main coronary arteries (segment score) had almost three-fold higher likelihood of early AMD, OR 2.67 (95% CI 1.24-5.78). CAD was not associated with late AMD. There were no significant associations between ERM or cataract surgery with CAD. However, prevalence of severe ERM — PMF — was significantly higher than the corresponding age-standardised prevalence in the BMES-1 (p<0.001). Overall prevalence of ERM was 7.0% (n = 115), with that of CMR and PMF each being 3.5%. Prevalence of cataract surgery was 13.1% (n = 218) in the AHES, with a mean age of 67.1 years. The prevalence of cataract surgery in this clinic-based cohort was significantly greater (p<0.0001) and the mean age was significantly lower (p≤0.0005) than that of the population-based cohort, the BMES-1. Conclusion: This thesis represents the largest clinic-based cohort that examined the associations between quantitatively-assessed CAD with and without coronary artery stenosis, and a wide spectrum of retinal microvascular signs and ocular diseases. These include retinal arteriolar narrowing, venular widening, Df, curvature tortuosity, branching angle, AMD, ERM, and cataract. No studies have as yet examined the relationship of retinal microvascular signs specifically to the presence, absence, extent, or severity of angiographically-assessed CAD. This data will assist in determining how we may infer retinal microvascular signs from differences in coronary vasculopathy. With regards to retinal vessel calibre, the candidate found that metabolic syndrome was independently associated with narrower retinal arterioles among those at high risk of CAD. As well, hypertension was independently associated with narrower retinal arterioles in those with and without CAD, both confirming and augmenting extant literature on this subject. In terms of retinal vascular geometric measures, this thesis provided some evidence to suggest that Df and curvature tortuosity are associated with CAD extent and severity, after accounting for the impact of media opacity. A sparser retinal microvascular network (smaller Df) was associated with older age and female sex. In regards to key ocular diseases in the AHES, severity of coronary stenosis and the presence of stenotic lesions were independently associated with AMD, specifically early AMD, in our cohort of patients presenting for coronary angiography to Westmead Hospital. While cardiovascular disease (specifically severity and extent of CAD) was not associated with ERM, this thesis presents evidence to suggest that there may be a greater prevalence of severe ERM (PMF) in a high cardiovascular risk cohort (specifically, the AHES), relative to population-based studies. This thesis also proposes that patients with cardiovascular risk factors have significantly younger age of onset and greater prevalence of cataract surgery than that of the general population, although severity and extent of CAD was not shown to be associated with prevalent cataract surgery. Overall, the findings of this thesis establish independent links between cardiovascular and ophthalmic pathology, that is, between macrovascular disease such as CAD, and both retinal vessel calibre and newer retinal vascular geometric microvascular signs. These findings help strengthen the ongoing hypothesis that non-invasive retinal and other ophthalmic imaging could be a useful adjunct to coronary angiography and other more conventional means of assessing cardiovascular disease.