- In the last 2 decades, various non-invasive quantitative methods have been developed to study the retinal microvascular network in detail.
- The retinal microvascular network offers a powerful tool to study the role of the microcirculation in cardiovascular disease.
- Retinal vascular imaging should be part of the diagnostic work-up of patients at cardiovascular risk.
In a previous issue of the E-Journal-of Cardiology-Practice I reviewed the application of contemporary ophthalmological examinations in the diagnostics and progression of hypertensive disease . In that review I focused on novel non-invasive methods for retinal microcirculatory investigations and their use in assessing microvascular changes in hypertension. In the last 2 decades these new methods have led to a revival of quantitative retinal microvascular analyses as a diagnostic and prognostic tool. Given the recent wide application of retinal microvascular imaging in other forms of cardiovascular disease than hypertension, this short review aims to discuss whether such analyses could have broader prognostic value for examining the progression of cardiovascular diseases and whether they can help in tracking the efficacy of various forms of treatment.
Quantitative non-invasive evaluation of the retinal microcirculation
Ophthalmological inspection, with a focus on the retina, has been a standard procedure in the diagnostic work-up of hypertensive or diabetic patients for many decades. The classical procedure was based on a semiquantitative grading system. However, in the last few decades, new quantitative and non-invasive technologies have become available to evaluate and follow-up changes in the microcirculation of the retina much more precisely. Fluorescent angiography, introduced in the 1960s, offered the possibility of testing the patency and leak of retinal microvessels as well as their flow characteristics. However, the use of fluorescent substances still involved a certain degree of invasiveness. Other - although non-invasive - imaging technologies, such as magnetic resonance imaging, are less suited to the task because of the limited spatial resolution to measure microvascular parameters quantitatively . Around 20 years ago, a major new approach to retinal imaging was introduced by Wong, Hubbard and colleagues which allowed a quantitative analysis of the entire retinal microvascular network . They used a non-mydriatic video camera with advanced software to analyse these networks off-line in terms of microvascular diameters, wall-to-lumen ratios, vessel numbers, branching angles and degree of vessel tortuosity. A major advantage of this approach is its potential to make repeated measurements in the same individuals for follow-up studies. Recently, Wong and colleagues reviewed the results of these studies in a range of patient or population-based studies . Their work suggests that this technology is indeed suited for prognostic studies of the retinal microcirculation in the development or treatment of various cardiovascular diseases.
Around the same time, Vilser and colleagues in Germany introduced the dynamic Retinal Vessel Analyzer or RVA (Imedos, Germany) to study retinal haemodynamics . This non-invasive camera and image acquisition system allows not only the static study of the morphology of the retinal microcirculation but also the study of the dynamic changes in vessel diameter upon stimulation with flicker light. Although such changes have been interpreted as indicators of the regulation of vascular tone, the underlying physiological mechanisms still need further investigation. A recent review paper gives an excellent overview of the applications of the DRVA technique . Another potentially important new approach to non-invasive quantitative retinal microvascular imaging was the introduction of the scanning laser Doppler flowmetry (SLDF) by Harazny and colleagues from Heidelberg, Germany in 2007 . This technique combines confocal and laser Doppler measurements and used an advanced software program to assess both structural and functional flow-related microvascular parameters. However, the clinical application of this approach has been limited due to a lack of commercial development of the necessary equipment.
Recently, a commercially available adaptive optics (AO) retinal imaging system, which was originally developed for astronomic applications, has been successfully used in a range of clinical studies. This AO system was developed in France (rtx-1 Adaptive Optics Camera; Imagine Eyes) and allows acquisition of high-quality morphological images of retinal microvessels with a resolution of up to 1 micrometer. This system is now widely used throughout the world in ophthalmological centres collaborating with cardiovascular specialists. Rizzoni and colleagues  recently reviewed the initial result of this non-invasive method to evaluate retinal microvascular structure and function. A potential advantage of this method is that it also allows the evaluation of endothelial function with a flicker light stimulus, although more studies are required to confirm the physiological basis of this application. Figure 1 shows an example of a morphological analysis of an arteriole in the human retina.
Figure 1. Evaluation of retinal arteriolar morphology with adaptive optics. a & b) An example of images of a retinal arteriole obtained with an adaptive optics camera (rtx-1 Adaptive Optics Camera; Imagine Eyes). c) Measurement of morphological parameters using a dedicated software. D: diameter; L: length.
Copyright © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins.  Koch E, et al. Morphometric analysis of small arteries in the human retina using adaptive optics imaging: relationship with blood pressure and focal vascular changes. J Hypertens. 2014;32:890-8.
Finally, several imaging modalities have been applied recently to study retinal flow distribution patterns and microvascular network characteristics [9,10,11]. These modalities are based on recent major steps forward in computer-based quantitative image analysis technology. Such advanced software programs allow detailed studies of the topology of the retinal microvascular architecture in terms of microvascular length, diameter, number and branching pattern. Optical coherence tomography angiography (OCTA) is of particular interest since it allows a three-dimensional representation of retinal flow patterns . In addition, OCTA even allows assessment of the architecture of retinal capillary networks. Recent studies have used OCTA to compare changes in retinal capillary networks with parameters of adverse cardiac remodelling . Although few clinical data are available with these novel imaging methods, they may offer excellent tools for future detailed follow-up studies on the topology of retinal microvascular networks in cardiovascular patients.
Previous research in animal models of hypertension has already highlighted the importance of changes in microvascular network architecture in the development and maintenance of hypertension . Recent clinical observations on the frequent occurrence of hypertension and other cardiovascular diseases in cancer patients treated with anti-angiogenic drugs – drugs that influence microvascular growth and network characteristics – add further evidence to the pivotal role of microvascular structure in cardiovascular disease.
In my previous contribution to this journal  I briefly reviewed the evidence for the retinal microcirculation as a marker and prognostic tool in hypertension and I referred to 2 comprehensive reviews on this topic [15,16]. After this first short review which focused mostly on the results obtained with the non-mydriatic camera observations in hypertensive patients, many new experimental and review papers were published on the basis of DRVA, AO or OCTA methods in patients with other forms of cardiometabolic disease. Gradually the question of whether retinal microvascular investigations should become standard procedure in the follow-up of patients with cardiometabolic disease seems more and more justified. The retinal microvascular bed is no longer only regarded as a window to the brain, but also to the heart . Even in the absence of local eye disease, quantitative analysis of the retinal microcirculation enables quantification of systemic cardiovascular risk. In addition, retinal microvessel biomarkers can be used in large scale population-based follow-up studies or clinical trials. Most of these studies thus far focused on diagnostic features and cardiovascular risk assessment. More research needs to be done on the underlying pathophysiological mechanisms involved as well as the effect of other interventions (i.e., pharmacological) to treat cardiovascular diseases.
The prognostic value of retinal arteriolar narrowing for the development of hypertension has now been firmly established in several cohort studies in different parts of the world including diverse ethnicities [6, 12]. A meta-analysis of most of these studies showed that each reduction of the retinal arteriolar diameter by 3 micrometers is associated with a 10 mmHg rise in blood pressure . Similar meta-analyses showed independent association of retinal arteriolar diameters with body mass index, metabolic syndrome, higher LDL cholesterol, lower HDL cholesterol and higher triglycerides . In addition to these metabolic parameters, several large cohort studies in diabetic patients have shown association with changes in retinal microvascular structure and the severity of both type 1 and type 2 diabetes, although the pattern of the retinal vessel phenotype associated with diabetes remains inconclusive . The recent analysis of one of the largest cohort studies thus far – the Maastricht Study – reported an association of diabetes with wider retinal arterioles . The loss of myogenic constriction of retinal arterioles due to hyperglycaemia has been postulated as the mechanism of arteriolar widening. On the other hand, long-term hyperglycaemia may lead to structural remodelling and arteriolar narrowing, in particular in cases of a combination of diabetes and hypertensive disease.
Non-invasive methods have not only been applied in studies of the association of retinal microvessels with cardiometabolic risk factors, but also in cardiovascular disease outcome studies. The first series of population-based studies using the non-mydriatic camera method found an association of both retinal arteriolar narrowing and venular widening with coronary artery disease, heart failure and stroke . However, not all results were consistent with respect to age and sex. Retinal microcirculation analysis at a young age, even in its early stages, has a stronger prognostic power than measurements in older persons . It seems likely that cardiovascular disease in the elderly is determined by more risk factors than microvascular alterations. On the other hand, the stronger predictive value of retinal microcirculation changes in women than in men is in agreement with the general observation of the more important role of microvascular dysfunction in cardiovascular disease in women than in men .
One of the longest running population-based cardiovascular disease outcome studies is the Atherosclerosis Risk in Communities (ARIC) Study, which has now collected follow-up data on the retinal microcirculation for some 20 years. These data confirm the excellent prognostic value of the retinal microcirculation in the long run . In addition to cardiovascular mortality data, recent population-based studies also suggest an important prognostic potential of the retinal microvascular biomarkers for the development of dementia in later stages of life . Although the exact role of (micro)vascular pathologies in the development of various forms of dementia is still not known, these studies suggest more mechanistic research at the level of the microcirculation is warranted.
The availability of various quantitative, non-invasive methods to investigate the retinal microvascular network has provided a powerful tool to study the role of the microcirculation in cardiovascular disease. The prognostic value of retinal microvascular biomarkers for the progression of different forms of cardiovascular disease has been well established in the past 2 decades. For future clinical research the following questions seem particularly relevant: 1) which pathophysiological mechanisms underlie changes in the retinal microvascular network in the different forms of cardiovascular disease. Most studies thus far have been of an observational nature: more mechanistic investigations are needed to allow future interventions targeted at the (micro)vascular level; 2) what is the effect of (pharmacological) interventions to treat cardiovascular diseases on the retinal microcirculation and 3) can future therapies be designed that act primarily on the (retinal) microcirculation to prevent the progression of cardiovascular disease.
The eye is not only our window through which we look out at the world but has turned out to be an inward window as well, one that allows us to observe the progression of our cardiovascular health. The retinal microvascular network contains physiological biomarkers to assess and follow-up our cardiovascular health. Various novel non-invasive imaging modalities now allow quantitative measurements of the retinal microvascular network. Such methods allow future studies on the progression and treatment of cardiovascular and metabolic diseases at the level of the microcirculation.