Introduction

Arterial hypertension is one of the major causes of cardiovascular diseases worldwide; in patients with hypertension the risk of coronary heart disease, congestive heart failure, ischemic and hemorrhagic stroke, renal failure and peripheral arterial disease is more than doubled [1]. There is a general agreement on the fact that essential hypertension is associated with the presence of structural alterations in the microcirculation [2–6], that represent that part of the vascular tree in which the major part of energy dissipates in order to overcome resistance.

The human vasculature is composed of large vessels (diameter >300 mm) and of small vessels (diameter <300 mm). The microcirculation includes the small arteries (diameter ranging from 100 and 300 mm), the arterioles (diameter <100 mm) and the capillary network (diameter around 7 mm). Small arteries probably contribute for about 30-50% of precapillary blood pressure drop, although an additional 30% drop occurs at the arteriolar level [6, 7]. An increase in the vascular wall thickness, together with a reduction in the internal diameter, may play a major role in the increased peripheral resistance observed in hypertension.

Small resistance artery remodeling

Peripheral resistance is determined by two independent components: active properties (mainly contraction) and structural properties (lumen diameter, wall thickness, number and re-arrangement of smooth muscle cells) of small arteries. The increase in resistance to flow that accompanies hypertension may be due partly to an increase in the vasoconstrictor tone, caused by activation of neurohumoral functional mechanisms, and partly to the increase in the wall thickness (mainly tunica media thickness) of the small arteries in relation to the lumen (M/L ratio) [4]. The increase in the M/L ratio of the small resistance arteries per se may lead to greater vasoconstriction for any degree of shortening of smooth muscle cells, thus enhancing the effect of any vasoconstrictor stimulus. This phenomenon represents a critical element in the maintenance and progressive aggravation of hypertension [4].

The increase in M/L ratio of human subcutaneous small resistance arteries can be caused by two different types of remodeling. The first one is the hypertrophic remodeling: the increase in the M/L ratio and in the cross-sectional area is due to hypertrophy and/or hyperplasia of the vascular smooth muscle cells. An increased M/L ratio may also be due to eutrophic remodeling, that consist in a rearrangement of the same amount of wall material around a smaller lumen without an increase in the cross-sectional area. In patients with essential hypertension, an eutrophic remodeling was generally observed; on the contrary in patients with secondary forms of hypertension a hypertrophic vascular remodeling was detected [8].

A significant correlation between coronary flow reserve and subcutaneous small resistance artery remodeling has been detected in hypertensive patients [9], suggesting that structural alterations in small resistance arteries may be present at the same time and with similar characteristics in different vascular beds. An increased M/L ratio of subcutaneous small resistance arteries is also present in diabetic and in obese patients [10–12]. It seems that the simultaneous presence of several cardiovascular risk factors may have a synergistic and deleterious effect on the microcirculation.

A relevant prognostic role of structural alterations in the microcirculation has been demonstrated independently from the blood pressure values [13] and also during antihypertensive treatment [14]. In fact, an increased M/L ratio of subcutaneous small resistance arteries is associated with a reduced event-free survival. The M/L ratio of small arteries and the pulse pressure are both major predictors of cardio-cerebrovascular events. Therefore, microvascular structure evaluation represents an important element also in terms of cardiovascular risk stratification [15-16].

Microvascular rarefaction

The vascular resistance of a tissue is determined also by the absolute number of perfused microvessels. In fact, another mechanism involved in the increase of vascular resistance is the reduction of microvascular density (rarefaction), which mainly affects the smaller vessels (less than 100 mm) such as arterioles and capillaries [17].

Rarefaction may be functional, when the microvessels are temporarily non-perfused or recruited, or anatomical, when the vessels are permanently absent. There is little information available about the mechanisms involved in the control of capillary density in the different vascular districts. Numerous studies have shown the presence of microvascular rarefaction in animal models [18]. In humans, several studies demonstrated a reduction of the number of capillaries in the skin of the dorsum of the finger of patients with essential hypertension [19-20]. It is not currently known whether capillary rarefaction may possess a prognostic significance per se, but there is evidence suggesting a correlation between capillary rarefaction and the M/L ratio of subcutaneous small arteries [21].

Methods for the evaluation of microvascular structure 

The methods for the evaluation of microvascular structural alterations available in humans are relatively few. Some of them are described in Table 1. One of the first methods proposed many years ago has been the evaluation of minimum vascular resistance in the forearm by plethysmography. Minimum vascular resistance, calculated from maximum post-ischemic flow and mean blood pressure was demonstrated to be correlated with the M/L ratio of subcutaneous small resistance arteries [22].

Table 1. Main methods and measurements to assess microcirculation in humans

AVR: arteriolar to venular ratio; SLDF: scanner laser Doppler flowmetry; AO: adaptive optics; M//L ratio: media to lumen ratio; W/L ratio: wall to lumen ratio;

A critical element inherent in this approach is the need to obtain true maximum vasodilatation, which is achieved via ischemia, muscular effort and heat. In condition of maximum post ischemic blood flow it is possible to calculate the minimum vascular resistance. Forearm minimum vascular resistance represents, therefore, an indirect index of microvascular structural alterations that may be assessed non-invasively with minimum discomfort.

In more details, the plethysmographic technique consists in a complete occlusion of the brachial artery of the dominant arm through the inflation of a sphygmomanometer up to 300 mmHg for about 13 minutes and a subsequent dynamic exercise (i.e 20-30 hand grips against a fixed resistance). The arterial occlusion is rapidly removed while venous occlusion is maintained (around 60 mmHg of pressure) [23]. If venous backward flow is absent, the increased volume of the forearm evaluated through a plethysmographic sensor around the forearm provides an estimation of forearm arterial flow. The ratio between mean blood pressure, evaluated either with invasive or non-invasive methods, and maximum arterial post-ischemic flow represents the forearm minimum vascular resistance.

Micromyography

The gold standard technique for the evaluation of structural alterations of small resistance arteries in humans is wire or pressure micromyography [24-26]. Wire micromyography was developed by Mulvany and Halpern in the 1970s [27-28]. This technique was used for the evaluation of the morphology and function of small arteries obtained from biopsies of subcutaneous tissue from the gluteal or the anterior abdominal regions (taken during elective surgery), in normotensive individuals as well as in hypertensive patients [22, 24, 29].

Small resistance artery segments, few millimeters long, were dissected free of periadventitial fat tissue, and then cannulated with 40 mm diameter steel wires, resulting in a ring preparation mounted on a wire micromyography. Mechanical stretch may be applied through a micrometric screw, while a force transducer records the passive tension that developed. The vessel may also be maintained on a constant stretch and stimulated to contract by adding various substances to the organ bath, such as norepinephrine, potassium, serotonin, thus allowing the measurements of the active tension developed. It is important that the vessel is not damaged during the mounting procedure, because even a minimal damage may lead to alterations in the contractile responses and errors during the measurement of morphological parameters.

Subsequently, the vessel, in relaxed condition, is transferred on the stage of an immersion lens microscope, and through a micrometric ocular (magnitude about 600x), the total wall thickness, the adventitia, media and intima thicknesses as well as the internal diameter are measured. The most useful parameter obtained with micromyographic approaches is represented by M/L ratio, since it is independent from the vessel’s dimensions [30-31].

Similar to the wire micromyography is the perfusion-pressure micromyography: isolated vessels are mounted in a pressurized myograph chamber and slipped into two glass microcannulae, connected to a perfusion system that allows a constant intraluminal pressure of 60 mmHg [31]. Morphology of the vessels is evaluated by computer-assisted video analyzers. Recently, reference values for M/L ratio in subcutaneous small arteries have been published, according to age and sex [32].

No difference between the two techniques was observed, providing that the vessels were analyzed under similar conditions [24, 30, 33]. Pressure micromyography allows a better evaluation of functional responses, although the precision of the morphological assessment can be slightly lower [30]. The evaluation of vascular myogenic tone and mechanical properties need a stepwise increase in intraluminal pressure, their estimation therefore a pressure micromyographic approach [34‑36].

Finally, pressure perfusion system allows to transfer small interfering RNA in vessels and plasmid or adenoviral vectors, which mediate gene delivery for evaluate vascular phenotype following gene modulation [37-38]. Micromyographic approaches were extensively used in the last decades, but the technique is limited by the local invasiveness of the procedure, requiring tissue biopsies. 

Videomicroscopy/capillaroscopy

Capillaroscopy is a reliable tool for the evaluation of capillary density in the cutaneous/subcutaneous vascular bed. This non-invasive method permits to study morphological and functional characteristics of the distal microcirculation. Currently the optical probe capillaroscopy is used; a polarized light at variable magnifications allows direct visualization of the capillaries on a monitor. The examination is performed with the patient in a sitting position, with the palm of the non-dominant hand resting on the observation plane. The nailfold area or the dorsum of the IV finger is analyzed by applying a drop of diaphanous oil (usually cedar oil or paraffin) to minimize light reflections and improve resolution without interfering with the optical properties [39].

After a basal evaluation the microcirculation is studied under conditions of maximum perfusion (obtained with venous congestion through the inflation of a sphygmomanometer applied to the base of the finger up to 60 mmHg for 2 minutes) to recruit those vessels not perfused in basal conditions and, therefore, to evaluate the total density of vessels of capillary district. The same procedure is used also to evaluate capillary density in the forearm skin. Capillary morphology and capillary density may be evaluated by in traditional capillaroscopy, while dynamic capillaroscopy explores also capillary flow velocity [40]. In combination with intravenous administration of fluorescent dyes, such as sodium fluoresceine or indocyanide green (fluorescence videomicroscopy or fluorescence angiography), capillaroscopy can be used to evaluate the heterogeneity of capillary flow distribution, or to disclose structures that cannot be seen with traditional capillaroscopy (such as capillary aneurisms) and to follow the transcapillary diffusion of markers (capillary permeability) [41]. Capillary density, defined as the number of capillaries for unit of skin area is clearly reduced in essential hypertension [40]. This technique is influenced by skin transparency, manicure procedures, and it is partly operator-dependent.

Retinal district

Retinal microcirculation represents a microvascular district that can be directly and easily observed with relatively simple approaches, such as a slit lamp or an ophthalmoscope [42]. Vasculature of the eye shares anatomic, physiological and embryological features with the heart and the brain [42-43], thus it can be assumed that also pathological changes occurring in these districts are similar [44].

Several studies have analyzed the relationship between retinal arteriolar signs and their associations with systemic vascular disease [45]. Evidences about the association between retinal microvasculature signs (i.e.: microaneurysms, cotton wool spots, etc.) and cardiovascular outcomes have been previously reported [46]. Wong et al investigated the association between markers of microvasculature damage in retinal district and the incidence of coronary artery disease [47]: the authors have used the ratio between arteriolar and venular external diameters (AVR) to assess retinal arteriolar narrowing, demonstrating the presence of lower values in hypertensive patients compared to normotensive controls [47]. Results were confirmed by a meta-analysis performed by Ding et al [48]: retinal arteriolar narrowing and venular widening was independently associated with an increased risk of hypertension [47].

However, the prognostic role of AVR is not well established, as a correlation between AVR and incidence of cardiovascular events was detected only in women [13]. Also, its potential role in stratifying hypertensive patients according to organ damage has been questioned: there is no relationship between quartiles of AVR and left ventricular mass, carotid artery intima-media thickness or urinary albumin excretion [49]. A meta-analysis of relationship between retinal vessel caliber (AVR, central retina artery equivalent, central retina vein equivalent) and future stroke events revealed that wider retinal venular caliber predicted stroke, whereas the caliber of retinal arterioles was not associated with stroke [1].

Other authors have also proposed a topological assessment of retinal microvasculature. Hughes et al. [50] have demonstrated the possibility to quantify topological changes in retinal vascular architecture through a specific software: they found an association between the presence of essential hypertension and an increased arteriolar length-to-diameter ratio. In hypertensive patients compared with normotensive controls also changes in arteriolar topology, indicative of rarefaction were observed, including a reduction in the number of terminal branches [50]. These differences were even more pronounced in patients with malignant hypertension. Moreover, antihypertensive drugs seem to beneficially affect some of the above-mentioned parameters [51].

Recent improvements in optical imaging techniques enabled the generation of high-resolution, non-invasive capillary perfusion maps with a retinal function images with consequent investigation of segmentation results and fractal analysis [52]. In particular, calculation of fractal dimensions of retinal microvasculature has been investigated as an index of complexity of the retinal network and, for this index, prognostic relevance was demonstrated. In a cohort of more than 3000 individuals, those with suboptimal retinal vascular fractal dimensions (lowest and highest quartiles) had 40-50% higher 14-year mortality for coronary heart disease than those with optimal fractal dimensions, independently from main baseline features as age and other risk factors [53].

Is should be however mentioned that in the majority of the studies performed in the last decades, the assessment of retinal vascular alterations was performed with funduscameras or slit-lamps and retinal photographs obtained were interpreted by ophthalmologists and quantified according the Keith-Wegener-Barker grading system; however some concerns have been raised in relation to this largely used approach [54-55].

In fact, criticism with respect to the reproducibility of grade I and grade II retinopathy has been raised, since even experienced investigators showed an high inter-observer and intra-observer variability [56-57]. For these reasons according to 2013 European Society of Hypertension/European Society of Cardiology Guidelines fundus oculi examination should be considered only in difficult- to-control or resistant hypertensive patients and is not recommended in mild-to-moderate hypertensive patients without diabetes [1].

Potential assessment of endothelial function has been proposed for retinal microvascular network: flicker light-induced retinal vasodilation evaluated by dynamic vessel analysis has been hypothesized to be an indicator of endothelial dysfunction [58]. Even if a potential clinical relevance has been shown in previous reports [59], further validating studies are needed [58].

Optical coherence tomography (OCT)

Intensity graph-assisted measurements using spectral- domain optical coherence tomography (OCT) may provide objective measurements of retinal vessels lumen diameters and wall thicknesses [60]. In a previous report, OCT has been proposed for the investigation of retinal vessels: Schuster et colleagues [61] have demonstrated a relationship between mean arterial blood pressure and OCT-based AVR. Moreover Muraoka et al. [62] have evaluated outer and inner diameters in four large retinal arteries and veins in 238 subjects without ocular disease, identifying a circular region around the optic disc, then performing an OCT scanning. Arterial and venous diameters were similar to those obtained with other techniques, such as scanned laser Doppler flowmetry (SLDF), even if a direct comparison between two methods was not performed. Additionally, the authors found that wall thicknesses of arteries and veins were significantly thickened as a consequence of aging and hypertension. However no adjustment for patients’ risk profile (i.e. diabetes or dyslipidemia) was performed. Recent improvements in OCT technology have allowed the potential exploration of the choroid. Variations in the choroidal thickness have been associated to systolic blood pressure values in healthy subjects [63]. Choroidal thickness, measured with spectral-domain OCT, was decreased in patients with arterial hypertension. This was attributed to arteriolar sclerosis and vascular contraction, caused by a high intravascular pressure in the choroid [64]. On the contrary, Gök et al. [65] have reported that arterial hypertension do not influence subfoveal choroidal thickness in comparison to healthy controls. OCT may be instrumental in order to document hypertensive retinopathy and choroidopathy severity, however further studies are needed to better understand the impact of OCT assessments on patients’ cardiovascular risk profile and on outcomes [66].

Scanning laser Doppler flowmetry (SLDF)

AAbout ten years ago, Harazny, Michelson, and Schmieder proposed an interesting and new noninvasive approach for the assessment of structural abnormalities in the retinal vascular district. A quantification of the wall-to-lumen ratio (W/L ratio) of retinal arterioles was obtained using scanning laser doppler flowmetry (SLDF) (Heidelberg Retina Flowmeter, Heidelberg Engineering, Heidelberg, Germany) and appropriate software analysis [67-68].

This technique performs a confocal measurement of the external diameter of retinal arteriole and an estimation of the internal diameter with a laser Doppler approach. Arterioles with dimensions between 80 and 140 μm in the superficial retinal layer may be investigated. Measurements of morphological parameters are performed offline with an automatic full-field perfusion imaging analysis: the outer arteriole diameter is measured in reflection images, and the lumen diameter is measured in perfusion images. The software automatically compares the two images taken in the same retinal area (Figure 1). Advantages of this approach are related to the easy repeatability with very little discomfort for the patients, and, according to the Authors to a relatively low inter- and intraobserver coefficient of variation of the measure (less than 10%) [69]. However, the reproducibility of the non-invasive measurement of the W/L ratio with the concerned technical approach in real life situations is probably less satisfactory, due to the possible problems in a correct estimation of the internal diameter with the Doppler approach. In addition, the Heidelberg Retina Flowmeter in no more present in the market, thus limiting the scientific interest and possible clinical development of this approach.

Figure 1. Reflection image and flow image in the retinal vascular bed assessed by Scanner laser Doppler flowmetry in four different regions of the eye (A, B, C, D)

There are however several evidences supporting the clinical relevance of the evaluation of the W/L ratio of retinal arterioles. As previously mentioned, in hypertension an increase in the M/L ratio of small resistance arteries seems to be present in different districts [9, 70]. Microvascular morphological changes observed peripheral district (e.g. subcutaneous tissue) might reflect similar changes in coronary or cerebral circulation [70]. An increased W/L ratio of retinal arterioles has been reported in patients with hypertension [68], with the history of a cerebrovascular event [67] or with other clinical signs of target organ damage, such as albuminuria, expression of the microvascular damage at the kidney level [71]. W/L ratio of retinal arterioles was also demonstrated to be correlated with peripheral, central, and 24-hour blood pressures [72-73].

When W/L ratio and AVR of retinal vessels were compared, inconsistencies were detected: W/L ratio of retinal arterioles was progressively higher in normotensive individuals, treated hypertensive individuals and hypertensive patients with a history of a cerebrovascular event, but this was not the case of AVR values; in addition, W/L ratio but not AVR values paralleled those observed for carotid intima-media thickness [74]. The W/L ratio of retinal arterioles assessed with SLDF was compared with the M/L ratio of subcutaneous small resistance arteries evaluated by wire micromyography. A rather good agreement between the two techniques, with a Pearson’s correlation index above 0.76 was shown [75].

Adaptive optics (AO)

Recently, a novel and extremely promising approach became commercially available: direct measurement of W/L ratio of retinal arterioles using an adaptive optics (AO) imaging system (AO camera, Rtx-1, Imagine Eyes, Orsay, France) [76-77]. The implementation of AO in a fuduscamera lead to a quite higher quality of images through the correction of wavefront aberrations: in fact such a technique has been widely employed in ground-based telescopes [78] to measure and correct atmospheric aberrations, allowing the formation of high-quality images of astronomical objects [76-77, 79‑80]. In the retinal microvasculature this methods permits to investigate vessels from 20 μm to more than 150 μm of diameter [79-80]. A beam of light enters the eye, and a small amount is reflected back out of the eye and into the optical system: wavefront aberrations in the reflected image are sensed by a suitable image sensor in the system, and corrected for by a deformable mirror [76, 81]. The achieved image resolution is of the order of 1 μm. The fundus illumination is obtained by a temporally low-coherent light-emitting diode flashed flood source at 840 nm; the frequency of images acquiring is of 10 Hz for a total of 4 seconds. The final image provided by the instrument is the average of a sequence of such images [76] (Figure 2). In most cases pupil dilation is not needed, anyhow topical tropicamide can be applied to improve images acquiring.

Figure 2. Retinal arteriolar vessel assessed by adaptive optics (AO)

The system, thus, provides images of a quality and resolution never previously obtained [81-82]. The vessel walls are clearly visible in most circumstances, provided that the eye fixation is correct and that the ocular media are clear. Studies in populations of normotensive subjects and hypertensive patients highlighted the potential role of AO in assessing age- related remodeling of the retinal arterioles; in fact W/L ratio of retinal arterioles has shown a linearly relationship with age in previous reports [77, 83-84]. Interestingly, it has been found a dependence of the W/L ratio on the wall thickness, but not on the internal vessel’s diameter.

Rosenbaum et al have observed a correlation between age-corrected W/L ratio of retinal arterioles and blood pressure values [77]. Blood pressure and age both independently increase W/L ratio by thickening the arteriolar wall. Dealing with anti-hypertensive treatment, they have observed that even a short-term reduction in blood pressure obtained by antihypertensive drugs induced a decrease in W/L ratio of retinal arterioles due to an increase in the diameter rather than to wall thickness changes; by contrast, no modifications were observed in subjects with negligible reduction in blood pressure values during treatment. In treated and controlled hypertensives under monotherapy W/L ratio normalization was observed, to be ascribed to combined wall thickness decrease and lumen dilatation independently of antihypertensive pharmacological classes, suggesting that adequate control of blood pressure may provide protection form microvascular alterations.

Recently, De Ciuceis C et al. have compared M/L ratio of subcutaneous small resistance arteries assessed by wire micromyography and W/L ratio of retinal arterioles assessed by AO and SLDF in a population of hypertensive and/or obese patients [85]. The correlation between the M/L ratio of subcutaneous small arteries and the W/L ratio of retinal arterioles evaluated by AO (r=0.84, r² =0.64, p<0.001) was closer that that observed between the M/L ratio of subcutaneous small arteries and the W/L ratio of retinal arterioles assessed by SLDF (r=0.48, r² =0.2649, p<0.05, slopes of the relations: p<0.01 AO vs. SLDF). Receiver operating characteristic curves of AO and SLDF exploring the ability of W/L ratio to discriminate patients with a M/L ratio above or below the cutpoint of 0.05 resulted significantly different in term of sensitivity and specificity, clearly in favor of adaptive optics (p <0.05). When intraobserver and interobserver coefficient of variation have been analyzed, smaller values were observed for AO compared to SLDF, confirming the very good reproducibility of AO [86].

Conclusions

There is convincing evidence of a strong relationship between alterations in the microcirculation and cardiovascular outcomes. Detection of such alterations may be clinically useful for a better stratification of cardiovascular risk and possibly for guiding antihypertensive treatment, through a precise assessment of possible regression of microvascular damage. New technologies made it simpler to assess the microvascular structural changes and non-invasive techniques, especially in the retinal vascular bed, may allow us to perform such evaluations in almost all hypertensive patients. However, at present we have no direct evidence of a prognostic value of non-invasive measures of microvascular structure.

Adaptive optics is a promising approach for non-invasive evaluation of microvascular structure. Nevertheless, future studies are needed in order to establish a possible prognostic relevance of this method and its usefulness for monitoring the effectiveness of treatment.