Keywords

Hypertension has profound effect on the structure and function of the eye. In this chapter, we will focus on hypertensive retinopathy. Only for the sake of completeness, it should be mentioned that hypertension-related changes also include hypertensive optic neuropathy and hypertensive choroidopathy. Moreover, there are ocular diseases where hypertension is a potential risk factor, e.g., age-related macular degeneration and glaucoma. Most importantly, hypertension causes structural and functional vascular changes. We elaborate on the current knowledge of these processes, including innovative findings related to new applied technologies.

1 Vascular Remodeling

Since the pathophysiological concept of vascular remodeling was in detail described in Chap. 5, only few words will be mentioned. Remodeling can be of two types – eutrophic and hypertrophic, but both types result in an increase in the media-to-lumen ratio of small arterioles. In eutrophic remodeling, both the outer and the lumen diameter are reduced, but the media cross-sectional area is not increased because the increase in the wall thickness-to-lumen diameter ratio is caused by a rearrangement of the smooth muscle cells around a narrowed lumen [1]. In contrast, in hypertrophic remodeling, an enhanced growth response resulting in an increased media cross-sectional area was documented. Eutrophic remodeling predominates in patients with essential hypertension, and hypertrophic remodeling was found in patients with severe and long-standing hypertension.

The importance of vascular remodeling is based on a finding that it is one of the early (or even the earliest) processes that occurs in response to increased blood pressure (BP) and leads to hypertensive end-organ damage [2], but also that effective antihypertensive treatment is capable to reverse these vascular adaptive processes [3].

The exceptional position of retinal vasculature is known for a long time. Already in 1939, Keith et al. has stated “because the arterioles are small and are difficult to visualize in the peripheral organs, for example, in the skin, mucous membranes, and voluntary muscle, the retina, as seen through the ophthalmoscope, offers a unique opportunity for observing these small vessels from time to time. Therefore, we think that certain visible changes of the retinal arterioles have been of exceptional value in affording a clearer clinical conception of altered arteriolar function throughout the body.” [4]. Based on the pioneering work of Keith, Wagener and Barker, their four-group grading system with increasing severity (Table 28.1) was widely applied in the last decades for the stratification of risk in hypertensive patients [4]. However, the clinical usefulness, and hence relevance to current clinical practice, has been questioned, because of poor reproducibility (e.g., 20–40 % interobserver variability) and weak association with other target organ damage (TOD) in grade I and II retinopathy, respectively [5].

Table 28.1 Traditional Keith-Wagener-Barker classification and simplified (Wong-Mitchell) classification
Full size table

Subsequently, a simplified three-grade classification system according to the severity of the retinal signs was proposed by Wong and Mitchell [6] (Table 28.1, Fig. 28.1), based on the evidence that certain hypertensive retinopathy signs (e.g., arteriolar narrowing or arteriovenous nicking) are independently associated with cardiovascular (CV) risk. In a small study comprising 50 normal and 50 hypertensive fundi, respectively, inter- and intraobserver reliabilities of the simplified three-grade classification system and the traditional four-grade classification system introduced by Keith, Wagener and Barker were reported to be comparable [7].

Fig. 28.1
figure 1

Funduscopy (grade none, moderate, and malignant)

Full size image

In the ESH/ESC guidelines, it is no longer recommended in general [8]. However, the retinal circulation offers the unique opportunity to visualize repeatedly the body’s microcirculation directly, noninvasively, and safely in vivo. Hence, in the last decade, new and more specific approaches were introduced to overcome these shortcomings and to detect reliable early changes of the retinal circulation.

2 Retinal Photographs/Funduscopy (Static)

In the last two decades, several large-scale, population-based studies assessing retinal photographs were conducted, including patients with and without hypertension. In (most of) these studies, standardized protocols of retinal photographs (45° nonstereoscopic color retinal photograph centered between the optic disk and the macula) were used to define specific signs of retinopathy, but not regarding a prespecified grading system. In part, retinal abnormalities were described based solely on qualitative parameters, such as tortuosity, arteriovenous crossing, caliber, and optic disk, but due to limited clinical usefulness, these data will not be reviewed. As further improvement, the imaging software “Interactive Vessels Analysis” (IVAN) (University of Wisconsin, Madison, WI, USA) has been established. This system conducts semiautomated measurement of retinal arterioles and venules, and hence its ratio (A/V ratio); however, it is not able to evaluate the retinal vascular wall directly.

Based on this approach, these studies have analyzed the relationship between retinal vascular alterations and their association with BP, TOD, and CV events which are (in part) summarized in Table 28.2.

Table 28.2 Large scale, population-based studies (in alphabetical order) assessing associations between retinal vascular caliber (based on retinal photography) and blood pressure, target organ damage, and cardiovascular risk (in chronological order)
Full size table

It has been repeatedly shown that retinal alterations are strongly correlated with past, current, and incident hypertension. In most of these large-scale studies, associations with directly assessed (generalized) arteriolar narrowing or a decreased A/V ratio, indirectly indicative of proposed arteriolar narrowing, and hypertension were reported (for details, see Table 28.2). In contrast, conflicting results according to retinal venules and hypertension were found. For example, in the Blue Mountains Eye Study, venular narrowing was associated with current hypertension [19]. In accordance, in the Rotterdam Study, venular narrowing was found to be predictive of current and incident blood pressure [28], but in the Multi-Ethnic Study of Atherosclerosis, venular widening was associated with incident hypertension [26]. While arteriolar narrowing can easily be harmonized with hypertension, it may be more difficult to explain why wider retinal venular caliber is associated with development of hypertension. A recent meta-analysis, comprising 10,229 patients without prevalent hypertension, diabetes, or CV disease, proposed that 2,599 patients developed new-onset hypertension during follow-up of 2.9–10 years. Both arteriolar narrowing (OR per 20 μm difference 1.29, 95 % CI 1.20–1.39) and venular widening (OR per 20 μm difference 1.14, 95 % CI 1.06–1.23) were independently associated with incident hypertension [37].

Importantly, in a population-based cohort comprising 1,572 children aged 6–8 years, each 10 mmHg increase of systolic BP was associated with arteriolar narrowing by 2.08 μm (95 % CI: 1.38–2.79, p < 0.0001), indicative that effects of elevated BP manifest early in life [36].

Regarding TOD data are even more limited. In 1,439 middle-aged African-Americans participants of the Atherosclerosis Risk in Communities Study A/V ratio was associated with measures of left ventricular hypertrophy, which was partly explained by additional CV risk factors and hypertension [13]. In contrast, in an Italian study comprising 386 untreated and treated hypertensive patients, no intergroup differences in A/V ratio was found between presence and absence of acknowledged TOD like left ventricular hypertrophy, carotid intima-media thickness, or microalbuminuria, hence indicating limited value of A/V ratio for identifying patients with high CV risk based on cardiac and extracardiac TOD [38].

There is also an ambiguous picture of arteriolar and venular diameter and different components of CV events. Regarding incident stroke, associations of both arteriolar narrowing and venular widening were reported in some studies, whereas in the Rotterdam Study, only an association of venular widening was found, but not for arteriolar narrowing [30]. Moreover, in the latter Rotterdam Study, venular widening was also associated with intracerebral hemorrhage [31]. These conflicting results are supported based on meta-analyses performed mainly by the META-EYE study group and published in the last years [18, 39].

These conflicting results of the individual components (with respect to arteriolar and venular diameter) have also to be taken into account, when interpreting reported findings about A/V ratio. An altered A/V ratio can be due to single and concurrent changes and their individual amount, and vice versa nonfindings can be seen, for example, by diverging changes.

3 Global Geometrical and Branching Parameters (Retinal Vascular Network)

The vasculature is a branching system, and alterations from optimal architecture are proposed to impair function and hence increased vascular damage. Thus, interest has gained on further developments in computer-assisted programs enabling the assessment of several quantitative parameters of retinal vascular network. Using these newly developed retinal vascular parameters, analysis of the Singapore Malay Eye Study has shown that a combination of smaller retinal vascular fractal dimensions (D f), proposed to be a global measure of the geometric complexity, and evidence of straighter retinal arterioles indicate poor BP control in treated hypertensive patients [40]. Therefore, retinal alterations can be assumed as pathophysiological markers not only for the severity of hypertension, but also on the effectiveness of drug therapy in hypertension.

Utilizing data from the multiethnic Singapore Prospective Study Program (SP2), the same group has shown that retinal D f was inversely correlated with BP level in all three ethnic groups. Notably, this was the case in patients with uncontrolled as well as untreated hypertension, but not in patients with controlled hypertension [41].

However, by applying again and again several new parameters, and analyzing these new parameters in the same studies, it is still missing the differentiation which is the most promising and reliable parameter to detect early retinal involvement in the clinical course of hypertension.

4 Scanning Laser Doppler Flowmetry (Dynamic)

Funduscopic photographs have the limitation that arteriolar and venular alterations cannot be quantified separately and the vascular wall precisely visualized. Moreover, the term remodeling if assessed in vivo takes two aspects into account, which were interrelated and indistinguishable, namely, morphological changes (i.e., rearrangement of vascular smooth muscle cells) as well as changes of the vascular tone (i.e., endothelial function). To overcome these limitations, one promising approach introduced by our study group about 10 years ago allows the dynamic assessment of both functional and structural parameters by using scanning laser Doppler flowmetry (SLDF) [42]. In brief, SLDF is performed in the juxtapapillary area of the right eye, 2–3 mm temporal superior of the optic nerve at 670 nm (Heidelberg Retina Flowmeter, Heidelberg Engineering, Germany). An arteriole (80–140 μm) of the superficial retinal layer in a retinal sample of 2.56 × 0.64 × 0.30 nm is scanned within 2 s (one systolic and one diastolic phase) and measured every 10 μm of this specific length of the arteriole. The confocal technique of the device ensures that only capillary flow of the superficial layer of 300 μm is measured. The outer arteriole diameter (AD) is measured by reflection images, and the lumen diameter (LD) is measured by perfusion images. Wall-to-lumen ratio (WLR) is calculated using the formula (AD – LD)/LD (Fig. 28.2). Analyses are performed offline with automatic full-field perfusion imaging analysis (AFFPIA) (SLDF Version 4.0 by Welzenbach with improved resolution) [43].

Fig. 28.2
figure 2

Scanning laser Doppler flowmetry (SLDF). (a) Differentiation between retinal arteriole and venule (SLDF live image before measurement). (b) Scanned area – reflection image. (c) Scanned area – perfusion image. (d) Scanned area – corrected and analyzed flow image. (e) Pulse curve run as mean retinal capillary flow (RCF) and time plot.(f) Localization of systolic and diastolic RCF on the image d. (g) Localization of systolic and diastolic RCF on the image c. (h) Calculation of wall-to-lumen ratio (WLR)

Full size image

It is noteworthy to mention that assessing both retinal function and structure by SLDF does not require applying any mydriatic drug, which is not only important from the scientific point of view – local application of tropicamide profoundly affects the retinal perfusion [44], but also for patient management perspective (i.e., no constriction of daily routine).

5 Retinal Capillary Flow

Due to its common origin from the internal carotid artery, the retinal microcirculation is morphologically and functionally related to the cerebral circulation [45].

Further dynamic information (e.g., basal nitric oxide [NO] activity) of the retinal capillaries can be assessed by measuring changes of retinal capillary flow (RCF) due to nonpharmacological and pharmacological tools. Flicker light increases at least in part via a NO-dependent mechanism and represents a nonpharmacological tool to investigate vasodilatory capacity of retinal capillaries. Notably, flicker light exposure has no effects on systemic BP, thereby minimizing potential systemic hemodynamic influences on RCF. Moreover, basal NO activity is assessed by administration of the NO synthase inhibitor NG-monomethyl-L-arginine (L-NMMA). Findings in normo- and hypertensive patients are summarized in Table 28.3.

Table 28.3 Studies of our research group, analyzing hypertensive patients and/or patients with cardiovascular event, using scanning laser Doppler flowmetry to assess early vascular changes (in chronological order)
Full size table

In young hypertensive patients, baseline RCF was similar to normotensive controls. However, assessment of basal NO activity (L-NMMA) revealed an impaired endothelial function in young hypertensive patients, which was improved after treatment with an ARB [42], whereas in elderly hypertensive men treatment with an ARB did not improve RCF at least after short-term treatment of 8 days [46]. Notably, even in hypercholesterolemic patients, treatment with an ARB led to a marked BP reduction, which was associated with an improved vasodilatory capacity (flicker light) [56]. In accordance, it was previously shown that vasodilatory capacity (flicker light) was lower in untreated hypertensive patients compared to normotensive controls. Systolic BP was inversely related to the percent increase of RCF due to flicker light exposure, independent of other CV risk factors [52].

Recently, we have shown that also individual pulsatile pattern of RCF (and structural parameters, see below) of retinal arterioles in systole and diastole can be reliably assessed. By doing so, we could show that with advanced stage of hypertensive disease, namely, patients with treatment-resistant hypertension (TRH), pulsed RCF (difference in RCF between systole and diastole) is exaggerated compared to patients with hypertension stages 1–2 [54].

Moreover, further data of our group reveal that BP and hence pulse pressure (PP) changes have an impact on pulsed RCF. In patients with TRH, we observed a decrease of systolic and pulsed RCF 6 and 12 months after renal denervation (RDN), in parallel to decreases of BP and heart rate. The reduction of pulsed RCF after RDN transfers into less shear stress on the vascular wall and, thereby, suggests an improvement of retinal (and potentially cerebral) microcirculation [55].

The importance of these findings is supported by prospective studies showing that, among others, carotid PP and pulsatility index were each associated with an increased risk for silent subcortical infarcts and with lower memory scores [57]. These data suggest that excessive flow pulsatility damages the microcirculation, clinically detectable by impaired cognitive function. Moreover, in a longitudinal study, PP significantly predicted the incidence of stroke (HR 1.33, 95 % CI 1.16–1.51 for each 10 mmHg of PP), which still remained borderline significant (p = 0.1) after adjustment for classical CV risk factors [58].

6 Structural Parameters of Retinal Arterioles

By using SLDF, also structural parameters of retinal arterioles can be assessed with high reliability [43].

We could show that WLR of retinal arterioles is positively related with systolic and diastolic BP, independent of various other CV risk factors. WLR was higher in never-treated hypertensive patients compared to normotensive controls [48]. Regarding blood pressure control, we found that in patients with poor BP control, WLR is higher than in controlled hypertensive patients [47]. These data are similar to published findings in large-scale population-based studies using retinal funduscopy (see above).

However, in a small cross-sectional study, A/V ratio was not able to discriminate between patients with cerebral event (transient ischemic attack or lacunar cerebral infarct) and normotensive as well as hypertensive patients. In contrast, WLR was significantly higher and could therefore discriminate between patients with cerebrovascular event compared to both normotensive controls and hypertensive patients without cerebrovascular event [49].

In a study comprising patients with wide range of BP values, it was shown that central PP is a strong and independent predictor of WLR (vascular remodeling) beyond “classical” CV risk factors and additional factors that are proposed to have an impact on vascular structure [53]. Such a relationship indicates coupling and intensive cross talk between the micro- and macrovascular changes due to hypertension.

Similar to our RCF analysis, also structural parameters can be assessed according to different heart phases (systole and diastole). In patients with TRH, a stiffer wall of retinal arterioles can be assumed, since wall thickness (WT) remained unchanged between systole and diastole, whereas in patients with hypertension grade 1–2, WT changed dynamically between systole and diastole [54].

Although no data from prospective studies regarding SLDF-assessed WLR and CV events are yet available, indirect, but strong, evidence of the validity for measuring WLR was demonstrated by Rizzoni et al. WLR assessed by SLDF (retinal arterioles in vivo) and media-to-lumen ratio measured with the myograph ex vivo (subcutaneous small arteries taken from a biopsy) showed a close correlation in hypertensive patients, suggesting that SLDF may provide similar information about microcirculation alterations compared to acknowledged prognostic measurement of subcutaneous small arteries, which represent the “gold standard” and prognostically relevant approach to the evaluation of small artery morphology in humans [59]. The absolute values differ due to the different methodologies, e.g., the analysis with myograph takes place ex vivo whereas the SLDF measures the parameters in vivo. The SLDF may underestimate the true internal diameter, since flow diameter does not include any endothelial plasma layer [49].

Nowadays, several other approaches (e.g., adaptive optics and optical coherence tomography) as well as SLDF with another software (e.g., data from Rizzoni et al.) focused on the assessment of WLR of retinal arterioles. It is notable to respect that the methodology of vascular measurements differs between the individual techniques. Therefore, a simple transfer of research findings into clinical practice may not be possible without further validation (see below). An overview of the recent available data is given in Table 28.4.

Table 28.4 Comparison of retinal structural parameters assessed with scanning laser Doppler flowmetry (SLDF), adaptive optics, and optical coherence tomography (OCT)
Full size table

7 Adaptive Optics

Nowadays, available adaptive optics-based fundus cameras are able to assess semiautomatically focal vascular changes (e.g., focal arteriolar narrowing). Moreover, arteriolar morphometry can be applied with a resolution up to near two micrometers, thereby visualizing (among others) vascular wall of retinal arterioles. The feasibility and reproducibility of retinal arterioles imaging was demonstrated in untreated hypertensive patients [62]. Following these pilot investigations, the same authors could show that adaptive optics-based assessment of WLR was positively correlated with mean BP and age which accounted for 43 % of variability of WLR [60]. Although the results on WLR measurements by adaptive optics are close to those reported by SLDF (Table 28.4), it has to be taken into account that no validation of the method in respect to other available techniques is yet provided. Adaptive optics needs to be directly compared with SLDF measurement or even better, with the media-to-lumen ratio of subcutaneous small resistance arteries assessed by the myographic approach. Adaptive optics examinations may be possible without mydriasis in most but not all cases. Only limited data (n = 9) are so far published investigating vascular morphometry before and after locally applied tropicamide. Mean vascular diameter increased only slightly (about 1 %), but data on individual diameters or wall properties are missing [60]. Hence, the effect of locally administered tropicamide cannot be fully excluded.

The major limitation of adaptive optics is that in contrast to SLDF, it cannot measure RCF.

8 Optical Coherence Tomography

The optical coherence tomography (OCT) allows the assessment of retinal circulation in an enhanced resolution within an acceptable time period. However, data about the retinal circulation in arterial hypertension are limited. In an analysis of patients (aged over 50 years), it was shown that mean arteriolar outer and inner diameter did not differ between patients with hypertension (n = 103, defined by use of antihypertensive medication or physician’s diagnosis) and without hypertension (n = 83), but mean arterial wall thickness was significantly larger [61]. This is in line with previous findings (unchanged outer and lumen diameter, but higher wall thickness of retinal arterioles) using SLDF in never-treated hypertensive patients compared to controls [48]. However, OCT-measured WLR was higher than previosuly measured with SLDF, perhaps likely attributable to age differences [48], but no direct comparison has so far been made.

Additional features, which can be assessed using OCT, like retinal nerve fiber layer and its importance in hypertension is not determined yet.

9 Perspective

Exciting new technologies emerged and offered the opportunities to directly visualize vascular remodeling of small retinal arterioles. The clinical perspective is that the physician may be enabled to diagnose early vascular remodeling to hypertension and tailor the antihypertensive strategy for individual patients. The findings may go beyond the retinal arterioles since the changes in the retinal circulation mirror these in cerebrovascular circulation, one of the major targets of hypertensive disease.