Focus on small artery stiffness

Nebivolol is a third generation selective β-blocker with vasodilatory properties. In contrast to carvedilol, it is a selective β1-blocker devoid of α-blocking properties 1. Its vasodilatory action is via facilitation of nitric oxide (NO) release 2,3. In this issue of the journal, the study by Arosio et al. 4 is an attempt to improve our knowledge on the haemodynamic effects and merit of a drug which may be of value in improving endothelial function and, consequently, in prevention of cardiovascular events. Here, the values and limitations of the methods used in their study are discussed. The design of the study is an adapted randomized, cross-over design. Nebivolol and atenolol, but not placebo, were randomized. This makes the differences between the effects of nebivolol and atenolol well-controlled, but not their absolute effects (comparison with placebo). A more optimal design would be a double blind, three-way, cross-over design randomizing the two drugs and placebo, and making use of the double dummy technique to blind both investigator and subject for the drugs used. Another factor which limits the conclusions of the study is the use of a single oral dose. Because, after a single dose, the full effect may not be present, the authors used atenolol 100 mg instead of 50 mg, the conventional once daily dose. However, the β1-blocking effect of nebivolol 5 mg once daily is, on average, 50% higher after 7 days of treatment than after a single oral dose 1, but the authors did not adapt the dose of nebivolol. In addition, after a single oral dose, the effect can rapidly change and differences between the drugs can be due to differences in time to peak effect. The present study well standardized the resting period, ambient temperature and start of measurements related to time of day and time after drug administration. Arosio et al. 4 demonstrate an increased vasodilatory response of cutaneous microcirculation on acetylcholine iontoforesis with nebivolol using well-established methods (laser Doppler and iontoforesis) and confirm earlier reports that an oral dose of nebivolol 5 mg can facilitate NO release 3. The use of time to peak divided by total time (PT/TT ratio) of the pressure curve obtained from a digital artery as a distensibility index of small arteries is more difficult to interpret. This raises two major questions: (i) does small artery stiffness have the same meaning as large artery stiffness? (ii) Is PT/TT a valuable measure of small artery stiffness? Large artery stiffness has been studied for many years. An increased stiffness leads to an increased pulse pressure. Both a high pulse pressure 5–7 and high arterial stiffness 8 have been recognized as risk factors for cardiovascular events. It is hypothesized that the increased cardiovascular risk results from an increased stress of the vessel wall. Arterial stiffness increases this stress 9,10, thereby increasing not only the risk on intimal lesion (an early feature in atherogenesis), but also the risk on plaque rupture, the major cause of acute cardiovascular events. Another important wall property influenced by arterial stiffness is total arterial compliance. This buffering capacity of the arterial tree is an important determinant of the afterload on the heart 11. The majority of total arterial compliance comes from compliance of elastic central arteries such as the aorta and carotid arteries. In addition, pulse wave velocity increases in stiff arteries, together with reflection points closer to the heart, leading to early wave reflections. Consequently, reflected waves arrive during systole in the ascending aorta and boost endsystolic pressure 12. This boosting of systolic pressure by early reflected waves is expressed by the augmentation index. Early reflected waves boost the systolic pressure in the ascending aorta, but not in the brachial artery 12, showing that the effect of arterial stiffness may vary between vascular territories. In patients with cardiovascular risk factors, but without clinically overt atherosclerosis, arterial stiffness is increased 13–17. This suggests that arterial stiffness may help to identify those patients who are at risk for atherosclerotic disease. However, the increase in arterial stiffness may depend on the vascular territory. In the early stages of metabolic disease, such as diabetes 16 and mild non-familiar hypercholesterolemia 17, aortic and carotid artery stiffness were not altered, while femoral artery stiffness was increased, indicating that the effect of a risk factor may differ between large artery territories. Except for the coronary arteries, atherosclerotic events predominantly result from large, more central arteries, such as the aorta, carotid arteries and the femoral arteries. In the coronary arteries, not only the circumferential wall stress induced by the blood pressure is important, but also the bending and stretching of the arteries due to cardiac contractions. Atherosclerotic events originating from small arteries are rare or absent. Is small artery stiffness important? This is not fully clear because small artery stiffness has not been studied well. It can be hypothesized that small arteries may help to unload stiffening central arteries. What are the arguments to support this hypothesis? Central arteries are most prone to arterial stiffening with age. Medium-sized muscular arteries, such as the brachial artery, increase compliance with ageing 18, suggesting that these arteries are able to compensate for the loss of compliance in more central arteries. As small arteries are also muscular arteries, they may behave similarly to medium-sized arteries. A second argument comes from the fact that a majority of first line antihypertensive drugs decrease arterial stiffness through an effect on small arteries 19. Indeed, the decrease in arterial stiffness of central arteries by (vasodilatory) antihypertensive drugs is achieved only in part, or not at all, by a direct action of the drug on central arteries. This is accomplished: (i) passively by a decrease in blood pressure (decrease in mean arterial pressure) resulting in an unloading of the more central arteries; (ii) by an increase in the buffering capacity of the muscular, more peripheral arteries; and (iii) by a decrease in haemodynamically important pulse wave reflections through a slower pulse wave velocity in the unloaded arteries and through moving sites of changes in impedance (reflection sites) to more distal sites in the arterial tree. This shows the important influence that dilation of small arteries and resistance vessels (and possibly also small artery stiffness) may have on arterial stiffness and pulse pressure of more central arteries. How to study small artery stiffness? Distensibility of small arteries is not well studied. An attempt has been made with the HDI/PulseWave*CR-2000 Research CardioVascular Profiling System (Hypertension Diagnostics Inc., Eagan, Minnesota, USA), which analyses the diastolic pulse contour using an adapted Windkessel model. In this model, it is claimed that C1 reflects the compliance of large arteries and, C2, the compliance of small arteries 20. Because this model cannot be checked by direct measurements of small artery stiffness, it remains inconclusive. Another approach is that used by Pancera et al. 21, who used the PT/TT ratio as an index of small artery distensibility. It is likely that a change in elasticity will steepen the ascending slope of the pressure wave, but not per se change the time to peak. It is assumed that the time to peak will depend on the viscoelastic properties and, in particular, on the viscosity of the arterial wall, which may be linked to arterial stiffness. Similar to the left ventricular ejection time used in systolic time intervals, PT is corrected for the heart rate by dividing by TT. Whether this correction for TT is accurate is not clear. Consequently, although a part of the measurement is related to viscoelastic properties, it is not clear whether the use of this ratio accurately reflects a distensibility index. In addition, PT/TT is not only determined by arterial viscoelastic properties. As digital arteries are very close to arterial endings, wave reflections from these arterial endings may influence PT and consequently the ratio. This is most likely to occur when vasoconstriction is present, such as in the present study during the handgrip test. Important pulse wave reflections in digital arteries during vasoconstriction or in disease states are suggested by the fact that, after smoking a cigarette, the pulse pressure at the digital artery measured with Finapres increases much more than the pulse pressure at the brachial artery measured with the Dinamap 22. In a group of patients with uncomplicated diabetes mellitus, brachial artery pulse pressure (Dinamap) was similar to that in a matched group of healthy subjects, but digital artery pulse pressure (Finapres) was much higher, suggesting vascular abnormalities in the finger arteries with considerable wave reflections 22. This suggests that, similar to the augmentation index in central arteries, the PT/TT ratio in digital arteries cannot distinguish between arterial stiffness (a wall property) and early wave reflections. Time to peak is also dependent on cardiac contractility and ejection time. The latter depends not only on cardiac contractility, but also on cardiac afterload. Nebivolol 5 mg has β1-blockade with low afterload 23. This demonstrates the limitations of the use of this ratio. Similar to large arteries, efforts should be made to measure small artery stiffness more directly, avoiding models as was advised by the Consensus Conference on Arterial Stiffness 24. The small artery tract is too short to measure pulse wave velocity accurately. Local arterial distensibility and compliance require an accurate measurement of change in volume or diameter of the artery and change in pressure 24. The change in volume can be measured using techniques such as photoplethysmography, and the change in pressure with techniques such as Finapres. It is doubtful whether the current techniques are accurate enough to measure small artery stiffness adequately.