Carotid artery plaque and arterial stiffness

The so-called arterial stiffness-related indices constitute an ever increasing collection of parameters obtained by methods as different as applanation tonometry, ultrasound, MRI and oscillometry and apply to specific locoregional segments and much longer trajectories of the arterial tree. Most of these indices with a few exceptions definitely belong to the field of experimental and (pre)clinical research. One of these exceptions termed pulse wave velocity (PWV) gained special attention and recently entered the field of clinical routine because of its physiology-based characteristics and robust performance in large cardiovascular outcome studies (evidence-based medicine) [1,2]. Evidence is most compelling for the carotid–femoral modification (cf-PWV), and the 2013 European Society of Hypertension/European Society of Cardiology Guidelines for the management of arterial hypertension [3] consider increased cf-PWV (>10 m/s) as a cost-effective indicator of asymptomatic organ damage. Brachial–ankle PWV and Cardio-Ankle Vascular Index are gaining popularity, and validation against cardiovascular endpoints is ongoing [4], but their physiological relevance is disputed by several research groups. Carotid wall thickening (intima–media thickness >0.9 mm) and occurrence of plaque are other accepted markers of asymptomatic organ damage. Some of the most accessible noninvasive methodologies to study arterial wall properties in humans are waveform recordings by tonometry and recordings of the distension of an artery (diastolic to systolic diameter or area) either by ultrasound or MRI. Central blood pressure (cBP), augmentation index normalized at heart rate 75 ([email protected]), forward and backward (sometimes termed reflected) waves (Pf and Pb) derived from radial, brachial, femoral or carotid artery waveforms have been proposed as stiffness-related/elasticity-related parameters with varying degree of evidence for their physiological relevance. Current guidelines do not recommend them for clinical practice but call for more studies. Common distension-deduced indices comprise the beta stiffness index, Young's and Peterson's modulus (parameters increasing with stiffness) and strain and distensibility (parameters decreasing with stiffness). For defining strengths and limitations of the use of local arterial elasticity and stiffness parameters, we refer to recent review articles, for example Patel et al.[5]. Because of an easy noninvasive access and the vicinity of the central circulation, carotid artery biomechanics have been studied extensively. This editorial commentary will focus on the intriguing question whether occurrence of (carotid) plaque confounds or questions the clinical use of cf-PWV and other local and global stiffness-related indices. A recent systematic review by Boesen et al.[6], including 50 studies on the effect of carotid artery atherosclerosis on local strain, distensibility or compliance, Peterson and Young modulus and beta stiffness index with either ultrasound or MRI, demonstrated that local biomechanics are influenced by local atherosclerotic plaques. An important study limitation was that the precise amount by which carotid stiffness increased because of plaque was difficult to estimate. What is the evidence that plaque burden causes the changes in local carotid artery mechanics? It sounds logical but it is still open for discussion. Significant change in distensibility may already have occurred before the formation of overt atheroma fulfilling the classical definition of plaque because of other ongoing processes such as subclinical inflammation, advanced calcification (e.g. in renal patients) and elastin fragmentation. There is a need of longitudinal follow-up to ascertain whether or not adverse progression of local stiffness indices precedes significant plaque formation. Of interest, not all published literature points to the direction of a strong cross-sectional association between local plaque and local strain. Within the constraints of a commentary, we can only discuss a few of the (partially) discordant articles. Paini et al.[7] conducted a study in recent stroke, studying the far wall of the proximal common carotid artery by the Wall Track system. Radial strain, distensibility/compliance, Young's modulus and circumferential wall stress did not differ between patients with and without carotid plaque. However, from a mechanics point of view, they distinguished two patterns of longitudinal axis strain gradients: outward (lower radial strain and distensibility and higher Young's modulus in the plaque versus the adjacent wall) and inward (higher radial strain and distensibility and lower Young's modulus in the plaque) bending strain. In another study, Beaussier et al.[8] failed to show significantly different strain values between simple plaque and adjacent vessel wall after combining echo-tracking with MRI data. The otherwise nice overview by Boesen et al.[6] considered neither one of the cornerstones (PWV) nor other common indices derived from tonometry-based waveforms [9] as it was out of the scope. Does carotid artery atherosclerosis influence PWV and waveform-deduced indices obtained in a clinical setting? There is a paucity of sound articles. Under highly controlled experimental conditions, Feng et al.[10] demonstrated that plaque models in artificial blood vessels give rise to a different morphology of the waveforms. Their findings moved beyond an effect of occurrence of plaque; degree and distribution of simulated calcification and fat content were determinants of the locoregional pressure waveform. Thus, local stiffness indices may bear the potential to act as plaque characteristics. At the time being, it is unclear whether these highly experimental findings will translate into a clinical correlate. Moreover, plaque-induced changes in the waveform, if any, are not a final proof for plaque-induced changes of specific stiffness-related parameters. In this issue of the journal, Grillo et al.[11] claim that measurements of cf-PWV, cBP, [email protected] and Pf and Pb derived from waveform recordings made by the commercial PulsePen tonometer (DiaTecne, Milan, Italy) are only biased to a limited (nonsignificant) degree by carotid artery atherosclerotic plaques. It is novel information. They made their observations in a relatively small group of individuals (limited power) presenting with no significant stenosis (controls), bilateral plaques or right or left monolateral stenosis according to the NASCET criteria. The authors’ reading of the data is that carotid atherosclerosis should not be viewed as an obstacle to accurately measure cf-PWV and other tonometry-derived stiffness-related parameters. The authors’ results are in sharp contrast with those (see above) for local stiffness parameters obtained from changes in diameter distention using ultrasound-based or MRI technology. Regarding clinical perspectives, one should be reminded that both plaque and cf-PWV are well accepted risk indicators, whereas current recommendations are awaiting more sound evidence for the aforementioned local stiffness parameters that are influenced by local plaque formation. Thus, direct clinical implications are restricted to the findings by Grillo et al.[11]. Are there reasons why they could have missed a (small) clinically less-relevant effect of carotid plaque? The authors analyzed the minimal numbers required by the power calculation but still the study might be relatively underpowered. Are the study design (a search for left–right differences) and the methodology (PulsePen device) appropriate to sense the expected small differences in the study parameters induced by carotid plaque? Absence of significant differences between left-sided and right-sided PWV, [email protected], cBP and Pf and Pb in the plaque-free, unilateral and bilateral carotid artery plaque groups constitutes an argument but does not exclude a small effect of plaque on an ipsilateral stiffness index. It is theoretically not excluded that small (maybe below detection level of routine techniques) left–right differences in stiffness parameters are real because of well known left–right anatomical asymmetry. Influence of contralateral carotid plaque and more prevalent plaques along the aorta and the lower limb arteries on cf-PWV is another issue, and the problem can be further complicated by an observation by Bossuyt et al.[12] that the traveled carotid–femoral path assessed by MRI is slightly longer at the right body side. Current expert panel documents prefer the right carotid artery as the reference. Authors used the PulsePen device to perform tonometry. The device has been validated more than a decade ago by Salvi et al.[13]. At first glance, there is no compelling reason to assume that use of competing devices would lead to contrasting results. Do the choice and nature of the stiffness-related parameter and the technology matter? Is the combination of ultrasound and local distension-based parameters better suited to account for presence of local plaque? Grillo et al.[11] did not analyze it excluding head-to-head comparison. Young's and Peterson's modulus, strain and distensibility indices (and beta stiffness index) are intrinsically better descriptors of arterial wall biomechanical behavior than cBP, [email protected] and Pf and Pb according to several experts in the field. However, there is a never ending debate whether all of these parameters explicitly cover stricto sensu the concept and definition of ‘stiffness’ as a biomechanical property. A one-to-one relationship between ‘stiffness’ and cBP, [email protected] and Pf and Pb is questionable as these parameters share variance to a strongly varying degree with hemodynamics, ventricular performance and wave reflections [14]. Theoretically, vessel geometry and morphology (left-sided versus right-sided plaques and plaques at the common, internal or external carotid artery or bifurcation sites) and plaque distribution over the vascular tree [from the carotid arteries over the aorta to the lower body arteries (iliac and femoral)] may modify the effect of plaque on cf-PWV. Plaque is essentially an isolated phenomenon in the young and gradually spreads with aging, reaching a more uniform distribution at older age (in some cases even in middle age). It means that in older individuals, plaque may occur all along the trajectory of cf-PWV and the initial question ‘Does a carotid plaque affect cf-PWV?’ turns into ‘Does plaque burden affect cf-PWV?’. Also that question has not been addressed by Grillo et al.[11] and is not easy to answer by their methodology. In the ASKLEPIOS Study on apparently healthy individuals aged 35–55 years, there were no meaningful differences between left and right carotid artery intima–media thickness values and plaque prevalence [15]. On the contrary, femoral artery atherosclerotic lesions were more prevalent at the right side. What about ‘local PWV’ measurements in the carotid arteries? Over the last decade, cf-PWV, representing an average stiffness over an arterial trajectory, has evolved from a research tool to a standardized, generally accepted clinically useful marker of organ damage and a prognostic factor. In parallel, several research groups developed the concept of ‘local PWV’ interrogating specific sites or segments more prone to the process of arteriosclerosis and atherosclerosis. Conventionally and historically, ‘local PWV’ of thin-walled arteries has been roughly approximated by the Moens–Korteweg equation relating PWV with Young's modulus, wall thickness, radius of the artery under scrutiny and blood density. Assessment of ‘local PWV’ is not yet ready for clinical application but has been extensively studied in vitro and in vivo in experimental settings and in small patients groups starting from sophisticated ultrasound-based technologies [16,17]. By the end of the twentieth century, Brands et al.[18] published a seminal article on a noninvasive method to quantify ‘local PWV’ by means of processed radio frequency-ultrasound signals. Introduction of higher frame rate in ultrasound imaging could allow for a more accurate calculation of ‘piecewise PWV’ at the carotid arteries [19]. Typical study parameters applied to the carotids are PWVs from carotid artery displacement waveforms of the anterior or posterior far walls, from diameter distension waveforms and from acceleration waveforms at systolic foot and dicrotic notch. The a-priori likelihood that local carotid plaque will affect local/segmental PWV is much higher than the likelihood that one or a few local plaques will affect a long trajectory parameter such as cf-PWV. Recently, the concept of segmental or piecewise PWV of the carotid arteries better accounting for real-life three-dimensional geometry and viscoelastic behavior has been refined in experimental models [20]. However, all aforementioned techniques definitely require further evaluation. In conclusion, Grillo et al.[11] reassured the readership that conventional tonometry-based cBP, [email protected], Pf, Pb and cf-PWV are minimally biased by carotid artery plaque. However, it is clear that some ultrasound-based local stiffness-related indices that are not used at the time being in risk stratification are biased by local carotid plaque. There is reasonable doubt whether all of the explored parameters mimic arterial wall characteristics, viscoelastic behavior and stiffness stricto sensu equally well. Do they explore different aspects of arterial wall properties that are less or more sensitive to occurrence of (local) plaque? Are we dealing with conflicts in diagnostic performance between tonometry and ultrasound? Is there a differential behavior of stiffness indices when faced with an isolated (local) plaque in the young and middle-aged individual versus global high plaque burden (carotid and femoral arteries, aorta) in the older individual? Further research is needed, but we are at the dawn of a new era in which young inspired researchers using innovative imaging technology and advanced mathematical modeling will further disentangle the intriguing question raised and partially resolved by Grillo et al.[11]. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.