Prognostic application of arterial stiffness: task forces

Abstract

Epidemiologic and clinical studies have shown that increased pulse pressure is an independent cardiovascular risk factor in general population. Pulse pressure is determined by combined effects of cardiac factors (stroke volume) and the arterial stiffness. Arterial stiffness can be more directly evaluated by several measurements including the measure of pulse wave velocity (PWV). Aortic PWV, a marker of aortic stiffness, has been shown to be a strong independent predictor of cardiovascular and all-cause mortality in patients with end-stage renal disease (ESRD) on hemodialysis as well as in patients with essential hypertension and older subjects over 80 years. Local arterial stiffness assessment, namely carotid distensibility was also shown to predict cardiovascular risk, both in ESRD patients and in renal transplant recipients. Furthermore, it has been shown in a therapeutic trial that the lack of aortic PWV attenuation despite significant drug-induced reduction in mean blood pressure was a significant predictor of cardiovascular death in subjects with ESRD. These results support the hypothesis that measurement of aortic PWV could then help, not only in risk assessment strategies, but also in risk reduction strategies by monitoring arterial stiffness under different pharmacologic regimens. The drug-related reduction of aortic PWV could then give prognostic information, additionally to blood pressure reduction.

Aortic stiffness measurements could serve as an important tool in identifying ESRD patients at higher risk of cardiovascular disease. The ability to identify these patients would lead to better risk stratification and earlier and more cost-effective preventive therapy.

Hypertension, pulse pressure, stiffness, morbidity, mortality

Epidemiologic studies have emphasized the close relationship between the elevation of blood pressure (BP) and the incidence of cardiovascular (CV) diseases.^1,2 In the past, clinical hypertension was classified principally on the basis of diastolic BP (DBP), and the hypertension was attributed to a reduction in the caliber or number of small arteries secondary to adaptive changes in the structure and function of the heart. Recent prospective studies have directed attention to systolic BP (SBP) as a better guide for CV mortality,^3,–5 and to increased pulse pressure (PP) as an independent CV risk factor.^6,,,,,,–13 These studies have focused the attention on the implication of large arteries stiffness and on the way it determines the level of systolic and pulse pressures and CV risk.^{14,,,,,,,,,,,,,,,,,,–33}

The viscoelastic properties of large arteries can be described in terms of compliance, distensibility, or stiffness of the aorta or an individual artery.^14,,–17 These parameters are usually determined from systolic–diastolic changes in diameter coupled with the measure of the local pulse pressure.^{14,,,,–19,24,25,,,–37} Alternative possibilities could be used, the most common methods being based on the study of pulse wave velocity (PWV).^14,15,21

Pulse pressure, arterial stiffness, and cardiovascular risk

Pulse pressure as an independent predictor of cardiovascular risk

From a methodologic viewpoint, the concept that PP, per se, plays an independent role in addition to SBP, DBP, and mean BP is difficult to demonstrate. Early reports from the Framingham and Chicago^4,6 studies suggested that PP was not better than SBP in predicting coronary heart disease.

The relative risk associated to DBP, SBP, and PP on 5-year mortality was evaluated in the Hypertension Detection and Follow-up Program. After adjustment for other confounding factors, the PP was shown to be a significant predictor of overall mortality. In a model that included only untreated patients PP remained significant even if the model included SBP. The independent role of PP as a determinant of CV events during antihypertensive therapy was also demonstrated by Alderman et al.¹³

In a French study of normotensive and untreated hypertensive adults, the pulsatile component index of BP was associated with left ventricular (LV) hypertrophy.⁸ During a 10-year follow-up, the pulsatile component index was associated with an increased risk of death from coronary artery disease in women. In another prospective evaluation of hypertensive patients, those in the highest tertile of PP before initiation of antihypertensive therapy had an increased risk of myocardial infarction. Multivariate analyses revealed that PP as a categoric variable was an independent predictor of myocardial infarction.⁹

Millar and Lever from the Medical Research Council (MRC) trial³⁸ and Blacher et al³⁹ from the European Working Party on High Blood Pressure in the Elderly (EWPHE), Systolic Hypertension in Europe trial (Syst-Eur), and Systolic Hypertension in China trial (Syst-China) have shown that brachial PP was a stronger predictor than SBP for myocardial infarction in hypertensive subjects. The best predictor function of all possible linear combinations of SBP and DBP was shown to be similar to that of PP alone, indicating a causal association and not only a statistical artifact due to correlation between PP and SBP. The highest predictive risk was observed when the increase in PP was associated with high SBP in the presence of progressive decrease in DBP.³⁹

Finally, in the Framingham study including 1924 subjects between 50 and 79 years at baseline, coronary risk increased with low DBP at any level of SBP >120 mm Hg.⁴⁰ In a large population of normotensive and hypertensive men followed for 20 years, Benetos et al¹¹ confirmed that increased PP was a strong predictor of myocardial infarction, both normotensive and hypertensive ranges, especially in men more than 55 years. From the studies of Benetos et al,^11,41 it appeared that PP was a risk factor, even in individuals with a mean BP of <107 mm Hg, that is, pressure corresponding to 140/90 mm Hg.

Pulse pressure and end-organ damage

A number of publications support the view that high PP is associated with end-organ damage.

Heart, coronary circulation, and large arteries

Pulse pressure has been shown to be an independent predictor of CV mortality with recurrent myocardial infarction, congestive heart disease, and LV dysfunction. Increased PP was associated with development of LV hypertrophy independently of mean BP.^{12,42,,,,–47} Central aortic pressure was shown to be a major determinant of aortic dilation in Marfan syndrome⁴⁸ and PP was the principal pressure parameter associated with reverse remodeling during long-term antihypertensive therapy.⁴⁹

Carotid–cerebral circulation

Some observations suggest that PP predicts thrombotic and hemorrhagic stroke.^50,51 Conversely, a significant association has been observed between SBP and PP and carotid artery stenosis.⁵¹

Kidney function

It has been recognized that increased PP is a typical finding observed in patients with end-stage renal disease (ESRD),^19,30,52 and more recently, increased PP and arterial stiffness have been found linked to plasma creatinine and the presence of microalbuminuria in subjects with mild and moderate renal insufficiency.^33,53

Taken together, these findings suggest that brachial PP should be considered as an independent predictor of CV risk or alternatively as a surrogate marker of arterial disease.

Arterial stiffness and pulse wave velocity

Pulse pressure is influenced by LV ejection (stroke volume and ejection time) and arterial stiffness (principally that of the aorta and large central arteries). Because LV ejection remains stable or even decreases with age, arterial stiffness is the principal factor responsible for increased PP during aging, in patients with isolated systolic hypertension or in chronic renal diseases. As a result the question has arisen as to whether the arterial stiffness or pulse wave velocity (PWV), a classic marker of arterial stiffness, might be a marker of vascular disease and a risk factor for CV or overall mortality. Several studies have shown that arterial stiffness is a BP-independent factor associated with end-organ alterations such as LV hypertrophy^19,25 and arterial intima–media thickening.¹⁹

In a prospective study of a large cohort of 6992 normotensive men and women aged 45 to 64 years, Liao et al⁵⁴ showed that the cumulative incidence of hypertension was related to higher arterial stiffness, which can contribute as an independent factor to the pathogenesis of hypertension. Another study by De Simone et al⁵⁵ showed an association between stroke volume-to-PP ratio and CV risk in arterial hypertension.

Blacher and colleagues³¹ applied logistic regression and Cox analyses to the characteristics of subjects with ESRD. After adjustment for all risk factors (age, pre-existing CV disease, BP, anemia, LV hypertrophy, and so on), the odds ratio for PWV (>1227 cm/sec) was 4.4 (confidence interval [CI] 2.3 to 8.5) for all-cause mortality and 5 (CI 2.3 to 10.9) for CV mortality (Figs. 1 and 2). The association between mortality and aortic PWV was confirmed further by Shoji et al⁵⁶ in diabetic patients with renal failure. In a follow-up study by Guérin et al⁵⁷ it has been shown that improvement of aortic distensibility (decrease in aortic PWV) in response to antihypertensive treatment in ESRD patients was associated with decreased mortality and improved survival. The PWV is a parameter integrating arterial geometry (h = wall thickness; r = arterial radius; ρ = density) and intrinsic elastic properties (E) described by the Moens-Korteweg equation, PWV² = Eh/2rρ. Blacher and colleagues³² have shown that the principal predictor of CV and all-cause mortality in ESRD was the elastic modulus E. Finally, in a follow-up study of a cohort of ESRD patients, London et al⁵⁸ have shown that the increased effect of wave reflections on central arterial pressure was, independently of aortic stiffness, another independent factor associated with poor survival.

Probability of survival (all-cause mortality) in the end-stage renal disease patients according to the level of aortic pulse wave velocity (PWV) divided in tertiles. (χ² = 57.97; P < .0001.)

Figure 1.

Probability of event-free survival (cardiovascular) according to the level of aortic PWV divided in tertiles. (χ² = 52.37; P < .0001.) Abbreviation as in Fig. 1.

Figure 2.

In a follow-up study in subjects with essential hypertension, Laurent et al⁵⁹ have shown that aortic PWV was also associated with CV morbidity and mortality. Calculation of CV risk using Framingham equations was done in a study that included 710 subjects with hypertension.³³ The risk of CV complications consistently increased in parallel with the increase in aortic PWV. Furthermore, at any given age, aortic PWV was the strongest predictor of CV mortality. The odds ratio of being in the group of high risk of CV mortality (>5%/10 years) for patients with PWV >13.5 m/sec was 7.1 (95% CI 4.5 to 11.3). These findings are in agreement with a longitudinal study that showed that the ratio between stroke volume and PP, an indirect surrogate marker of arterial stiffness, was an independent predictor of CV risk.⁶⁰These findings suggest that increased aortic PWV should be considered as an independent predictor of CV risk and a surrogate marker of vascular disease.

Structural remodeling also occurs in smaller arteries and branch point. The changes between mechanical properties of conduit and resistive arteries influence the wave reflections and can contribute to augmentation of late SBP in the root of the aorta. Computer analysis of the diastolic decay of the PP wave was proposed for the analysis of mechanical properties (stiffness) of small arteries.⁶⁰ Studies with this method have identified an increased resistive artery stiffness with aging, hypertension, smoking, and diabetes.⁶¹ Preliminary studies suggest that this alteration increases the risk for CV events.⁶²