Remodeling of the Radial Artery in Response to a Chronic Increase in Shear Stress
Abstract Chronic changes in large artery blood flow rates induce corresponding adjustments in arterial diameter, but little is known about structural adaptations of the vessel wall in humans. We used a high-resolution echo-tracking system to measure radial artery internal diameter, wall thickness, and mean blood flow on both arms of 11 patients with end-stage renal disease. Measurements were performed on the wrist side of the arteriovenous fistula. The contralateral radial artery was investigated as control. Wall cross-sectional area, circumferential wall stress, and mean wall shear stress were calculated. Results indicate a sixfold increase in blood flow on the side of the arteriovenous fistula compared with the control side, with a 1.4-fold increase in internal diameter. The diameter enlargement was sufficient to normalize wall shear stress. Changes in diameter were not associated with arterial wall hypertrophy because wall cross-sectional area was not increased and rather suggest a “remodeling” of the arterial wall. For the same level of blood pressure, circumferential wall stress was increased on the side of the arteriovenous fistula. These results suggest that the structural adaptations of the arterial wall to a chronic increase in blood flow normalize wall shear stress and overcome stretch-induced changes in the particular circumstance of arteriovenous fistula.
Hemodynamic factors are important regulators of vessel wall structure. In large vessels, chronic changes in intraluminal BP are associated with changes in wall thickness, to normalize circumferential wall stress,1 2 and blood flow regulates vessel diameter through changes in wall shear stress.3 4 5 6 These mechanisms, well demonstrated in animal models, have been more recently observed in humans during noninvasive studies with ultrasound methodology. Thickened arterial walls were observed at the site of large and medium-sized arteries7 8 in untreated hypertensive subjects, and dilatation of large arteries was demonstrated in normal subjects in response to a large increase in blood flow velocity9 and in dialysis patients with an AVF.10 Hemodynamic factors that regulate vessel wall structure are most often intricate. In addition to BP, blood flow velocity appears to be particularly important for regulating wall thickening. Zarins et al3 showed that AVF in primates led to increased wall mass, whereas Langille and O’Donnell5 found no change in medial mass with decreased flow in rabbits. Recently, Dobrin and coworkers11 12 reported that intimal thickening in autogenous vein grafts was best correlated with low flow velocity, a correlate of low shear stress, whereas medial thickening was best correlated with deformation in the circumferential direction. Other studies suggested that intimal thickening of veins and prosthetic grafts was correlated with circumferential wall tension.13
In human native vessels these phenomena have not been studied in great detail because of the lack of appropriate tools to measure arterial diameter and wall thickness in vivo. Recent advances in ultrasound methodology have made it possible to measure noninvasively and with a high accuracy the internal diameter and wall thickness of superficial arteries in humans.14 15 We designed the present study to further investigate the relationships between chronic changes in blood flow and the geometry of peripheral conduit arteries in humans. For this purpose, we studied the forearm circulation of patients with ESRD bearing an AVF for hemodialysis using a high-resolution echo-tracking system for internal diameter and wall thickness measurements coupled with a continuous Doppler system for blood flow velocity measurement. The objectives of the study were (1) to describe the geometry and hemodynamics of the radial artery feeding the AVF, (2) to compare these parameters with the contralateral radial artery taken as control, and (3) to compare the arterial parameters of patients with ESRD with parameters of a group of subjects matched by sex, age, and BP to determine whether the difference between cases and control subjects may be ascribed to the renal disease or hemodynamic pattern of the fistula.
The study included 11 patients (10 men) with stable ESRD and 22 control subjects matched to the cases by sex, age (±3 years), and systolic BP (±5 mm Hg). Causes of renal failure were chronic glomerulonephritis in 5 patients, nephroangiosclerosis in 3, interstitial nephritis in 2, and polycystic renal disease in 1. ESRD patients were treated by hemodialysis (mean duration of dialysis, 7.5±6.3 years) with a well-functioning AVF created at least 3 months previously. Median predialysis BP for the 6 months preceding their inclusion was less than 160/95 mm Hg (134±23/77±12 mm Hg). Inclusion criteria included absence of present or history of cardiovascular complications (acute myocardial infarction, valvular heart disease, cerebral vascular disease, or decompensated heart failure) and normal BP during the 6 months preceding the study.
Among the 11 patients (mean age, 56±15 years), 9 received antihypertensive therapy, including calcium antagonists alone (n=2) or in combination with angiotensin-converting enzyme inhibitors (n=5). Two patients received a combination of three medications (calcium antagonist plus angiotensin-converting enzyme inhibitor plus β-blocking agent).
The hemodynamic study was performed 24 hours after the midweek hemodialysis. Patients were dialyzed three times per week on AN 69 membranes (Hospal); the duration of hemodialysis was 4 to 5 hours. Dialysate was delivered by a system including bicarbonate delivery, adjustable sodium concentration, and controlled ultrafiltration.
Control subjects were matched as two controls for one case. In this control group, subjects had never received any antihypertensive treatment. No subject had signs or symptoms of hypertensive complications, and none had valvular heart disease, major arrhythmia, carotid artery stenosis, or diseases of noncardiovascular nature. The study was approved by the Ethics Committee of the Broussais Hospital, and all patients gave written informed consent.
The investigation was performed in a controlled environment at 22±2°C. Each patient was studied in the supine position after at least 10 minutes of rest. Systolic and diastolic pressures were determined automatically every 5 minutes at the non-AVF arm with a Dinamap 845 oscillometric BP recorder. Systolic and diastolic pressures were determined as the average of five measurements. Mean BP was determined according to the usual definition (diastolic BP plus one third pulse pressure).
During this period, measurements of radial artery parameters were obtained on both arms after the forearm was extended and secured comfortably on a splint. On the AVF side, measurements were performed on the radial artery at the wrist, a site at which the artery was close to the anastomosis but not included within the fistula. At that site, the radial artery was exposed to high blood flow because of the low-resistance shunt induced by the proximity of the fistula. Patients in whom high turbulent blood flow made blood flow velocity and diameter measurements impossible or artery calcifications made wall thickness measurements impossible were excluded from the study.
Measurements of Arterial Parameters
Internal Diameter and Wall Thickness
The ultrasound system used in the present study (NIUS 02, SMH) has been previously described14 15 16 and validated for measurement of radial artery internal diameter and its systolic-diastolic variations and measurement of radial artery wall thickness in humans.7 14 17 A high-resolution pulse echo-tracking device was used to acquire backscattered radiofrequency data from the radial artery at the wrist. The probe consisted of a 10-MHz strongly focused piezoelectric transducer (6-mm diameter, 11-mm focal length) operated in the pulse-echo mode. The −10 dB beam width is 0.3 mm at the focal point, and the depth of field at −10 dB is 5 mm. A stereotactic arm permits motion of the transducer in x, y, and z coordinates with micrometric steps for placement of the probe perpendicular to the arterial axis in its largest cross-sectional dimension. The transducer is positioned so that its focal zone is located in the center of the artery, and the backscattered echoes from both the anterior and posterior walls can be visualized. A typical radiofrequency signal is then displayed on a computer monitor interfaced to the transducer system. Arterial diameter and posterior wall thickness are measured when a “double peak” radiofrequency ultrasound signal of the anterior and posterior walls is obtained. These signals are visible only as the ultrasound beam crosses the axis (center) of the vessel. They are characterized by a high-amplitude signal followed by a relatively silent acoustic zone and then a second high-amplitude signal. For measurement of internal diameter and intima-media wall thickness of the posterior wall, electronic trackers are positioned as follows: on the outer radiofrequency line of the second peak of the anterior wall and on the inner radiofrequency lines of the first and second peaks of the posterior wall (Fig 1⇓). Their movements are electronically tracked for 10 seconds, sampled at 100 MHz over 8 bits, and stored at a 50-Hz repetition frequency on a 120-megabyte hard disk for further data processing. Data are determined by computing the analytic signal according to methods previously described and based on Fourier transformation. Finally, diameter and wall thickness are expressed in millimeters by multiplying time-of-flight measurements by the approximate speed of sound in tissue (1.54 mm/μs). All data processing was performed with software developed by Asulab installed in a 486, 25-MHz AT computer. The pulse length of this 10-MHz ultrasound system was 0.1 microseconds at 6 dB and corresponded with a practical axial resolution of 0.16 mm for absolute internal diameter or wall thickness measurements and 0.0025 mm for these parameters during systolic-diastolic changes.14 15
Systolic-diastolic variations in internal diameter and posterior wall thickness were recorded. For data analysis, each patient was described by the average of recordings of mean wall thickness (millimeters), calculated by integrating the time course of the systolic-diastolic variations, and diastolic internal diameter (millimeters), over several cardiac cycles. Radius-thickness ratio was calculated as Di/(h×2), where Di is internal diameter, and h is mean wall thickness. Because wall thickness is influenced by the changes in distending BP (hence internal diameter), WCSA is more appropriate for detecting arterial wall hypertrophy. Because of the incompressibility of the arterial mass,17 WCSA (millimeters squared) can be calculated as π(Re2−Ri2), where Re and Ri are the values of internal and external radii, respectively. Mean circumferential wall stress (kilopascals) is calculated as MBP·2h/Di, where MBP is mean BP.
Blood Flow Velocity
Radial artery blood flow velocity was measured by continuous-wave Doppler (8-MHz transducer at a 60° angle, Doptek 2002, Deltex), distal to the 10-MHz probe, at the site of diameter measurements. No significant interference between the two waves was noted throughout the experiment. Resting blood flow (milliliters per minute) is the product of time-averaged mean velocity and arterial lumen cross section. Mean shear stress (dynes per centimeter squared) was approximated with a variation of the Hagen-Poiseuille equation18 : τ=4ηQ/πRi3, where η is the viscosity of blood (0.035 poise), Q is the velocity of blood flow (milliliters per second), and Ri is the vessel radius (centimeters).
Data are expressed as mean±SD. Differences in arterial parameters between the AVF side and control side were analyzed with paired Student’s t test. The relationship between continuous variables was evaluated by linear regression. Statistical significance was assumed at a value of P<.05.
AVF Side Versus Contralateral Side
The Table⇓ compares the radial artery parameters on the AVF side and contralateral side. On the AVF side, internal diastolic diameter was significantly increased compared with the contralateral side, whereas wall thickness and WCSA were not significantly different. Circumferential wall stress, calculated with the same value of mean BP (96±14 mm Hg) on both sides, was significantly increased on the AVF side. A sixfold increase in mean blood flow was observed on the AVF side compared with the contralateral side. These changes in internal diameter and blood flow were observed along with a normalization in mean wall shear stress.
Comparison With Control Subjects
To determine whether the geometric modifications of the radial artery in ESRD patients were related to the renal disease or chronic blood flow changes, we analyzed the relationship between intima-media thickening and BP in ESRD patients and a group of 22 control subjects matched to the cases by sex, age, and BP (136±22/81±9 mm Hg). In this group, the mean internal diameter was 2.494±0.386 mm and the mean wall thickness was 254±56 μm. Fig 2⇓ shows the negative relationship between the radius-thickness ratio and systolic BP in this group of 22 control subjects (y=−0.037x+10.1; r=.752). For the following analysis, we referred to this regression line and its 95% upper and lower confidence intervals. For the radial artery of the AVF side, the radius-thickness ratio was above the upper confidence interval line in 9 of 11 patients, indicating an inappropriate thickening of the arterial wall to the increased internal diameter for a given level of BP. By contrast, for the radial artery of the normal forearm, 9 of 11 patients with ESRD had values of the radius-thickness ratio inside the 95% confidence interval (Fig 3⇓).
The major findings of the present study are the following: (1) A sixfold increase in blood flow at the site of the radial artery feeding an AVF is associated with a 1.4-fold increase in its internal diameter; (2) wall shear stress is maintained constant in the radial artery, whatever the blood flow; (3) the structural adaptation of the radial artery to a chronic increase in blood flow does not result in vascular hypertrophy but rather in remodeling, as WCSA was not increased; and (4) the remodeling of the radial artery is accompanied by an increase in circumferential stress.
Consideration of Methods
To study vascular geometry in vivo, it is necessary to measure internal diameter and wall thickness. Only recently has ultrasound methodology made possible these noninvasive measurements in humans. In the present study, we used a high-resolution echo-tracking system previously validated for its ability to measure the wall thickness of a medium-sized muscular artery such as the radial artery14 and for the good reproducibility of these measurements in a clinical setting.7 The coupling of this apparatus with a continuous-wave Doppler system made it possible for us to calculate blood flow variations along the cardiac cycle and estimate wall shear stress. For newtonian fluids, shear stress equals the local viscosity times the local wall shear rate, which can be derived from the measures or estimates shape of the instantaneous velocity across the lumen.18 With our apparatus, it was impossible to determine the velocity gradient near the wall. Thus, we estimated mean wall shear stress by the Hagen-Poiseuille equation. Values obtained with this approach have been described to underestimate the true value.18 However, comparison of two different arteries in the same patient with the same device minimizes the methodological reserves for the interpretation of the results.
Exposure of an artery to pressure induces deformation in each of the three directions of space, resulting in an increase in circumference and length as well as radial deformation, the latter inducing a decrease in wall thickness. In opposition to these deformations, the vessel wall generates increased stress in these three directions. Stresses in the circumferential and longitudinal directions are tensile, whereas stress in the radial direction is compressive.2 In clinical studies performed in humans, circumferential stress is usually calculated with Lamé’s equation (ςθ=MBP·2h/Di). In the present study, internal diameter (Di) and wall thickness (h) were precisely measured at the site of each artery, but mean BP (MBP) was estimated from the systolic and diastolic BPs measured at the arm. Because the protocol was noninvasive, it was impossible to measure mean BP intra-arterially at the site of diameter measurement. However, in a previous study, we demonstrated that mean BP was not significantly different between the site of the AVF and the rest of the vasculature.19 Thus, we considered that the calculation of circumferential wall stress with the present methodology was reliable.
Interpretation of Findings
In vitro and in vivo studies have demonstrated that an increase in arterial blood flow and/or blood viscosity elicited increased shear stress and endothelium-dependent vasodilation.3 4 In the human radial artery, it was recently demonstrated that nitric oxide but not prostacyclin is essential for flow-mediated arterial dilation in vivo under acute conditions.9 The present study indicates that a chronic increase in arterial blood flow induces a sustained and striking increase in arterial diameter. This result confirms previous studies that have documented that the increase in arterial diameter started shortly after the creation of an AVF.10
In the present study, the large increase in radial artery diameter (+38% on the AVF side compared with the control side) suggests that structural modifications of the arterial wall contributed to the diameter enlargement in addition to a sustained phenomenon of flow-dependent dilation. Indeed, in response to an equivalent increase in blood flow occurring under acute conditions (reactive hyperemia), Joannides et al,9 using the NIUS apparatus, previously observed a 10% increase in radial artery diameter.
Despite high blood flow velocity on the AVF side, wall shear stress remained unchanged, probably because of diameter enlargement. Indeed, wall shear stress on the AVF side was not significantly different from control values. We must admit that the lack of statistical significance could be due to the wide SD values on both sides, reflecting a high intersubject variability. This high variability is mainly due to the variability of blood flow velocity, thus mean blood flow, as reflected by the high SD values shown in the Table⇑. Other researchers have reported similar SD values for mean blood flow.9 Another argument favoring the consistency of our finding is that the difference in wall shear stress between sides would be fourfold to fivefold if no diameter changes were observed, indicating that shears are undoubtedly reduced greatly by arterial expansion.
The original finding of the present study is the demonstration that the geometric adaptation of the arterial wall to a chronic increase in blood flow does not result in wall hypertrophy, despite the sustained increase in circumferential wall stress. Indeed, one of the consequences of the increased radial artery diameter is an increased parietal tension. According to Lamé’s equation, which states that circumferential wall stress is directly proportional to intra-arterial pressure and arterial radius and inversely proportional to wall thickness, arterial wall hypertrophy should develop to normalize wall stress. Thus, if normal wall stress is to be maintained, with respect to an increased diameter and similar pressure, a constant relationship between pressure and the radius-thickness ratio should be observed. This was the case in nonuremic subjects and ESRD patients on the normal arm side (Fig 3⇑) but not in ESRD patients on the AVF side (Fig 2⇑).
The observed modifications of radial artery geometry in ESRD patients on the AVF side cannot be defined as true arterial wall hypertrophy because WCSA remained unchanged; rather, they must be defined as remodeling. The term remodeling has formerly been used for describing geometric changes of the small arteries in response to high BP.20 21 These changes are characterized by reduced internal and external diameters with an unchanged WCSA, indicating a rearrangement of the wall material around a smaller lumen. In the present study, remodeling may be used in a broader sense: an unchanged WCSA, whatever the lumen. Indeed, the radial artery on the AVF side had increased internal and external diameters, whereas WCSA was unchanged.
The reasons for this response are not clear. The influence of uremia per se is unlikely, because the relationship between BP and the radius-thickness ratio on the control artery of ESRD patients is superimposable on that of control subjects, indicating an adequate response of the arterial wall to the distending pressure in ESRD patients. The second possibility is that conduit arteries have a limited capacity to respond adequately to a combined flow and wall tension overload. In vein grafts subjected to separate mechanical factors, such as circumferential stretching and changes in blood flow velocity, Dobrin et al11 demonstrated that changes in shear stress influence intimal thickening, whereas medial thickening responds to changes in circumferential wall stress. Intimal thickening occurs in response to low flow velocity and decreased shear stress, whereas medial thickening occurs in response to increased parietal tension.
Furthermore, Kraiss et al22 demonstrated that an elevation in shear stress inhibited smooth muscle cell proliferation and neointimal thickening in polytetrafluoroethylene graft, and Rekhter et al23 demonstrated that cell proliferation associated with neovascularization was one of the major mechanisms of intimal hyperplastic lesions in the anastomosis region of the polytetrafluoroethylene graft material vein in human AVFs. However, Masuda et al24 suggested that when an artery is enlarged in response to increased flow, the intima may also thicken in addition to the media to provide wall thickness for restoration of normal wall tensile stress. Whatever the respective effects of shear stress and circumferential wall stress on intimal and medial thickening, the present results indicate that both stimuli, occurring simultaneously, result in unchanged intima-media thickening. Whether the increased shear stress inhibits the media thickening because of increased wall stress remains to be determined. This interpretation is limited by the current impossibility of noninvasive discrimination between the two layers of radial artery wall thickness. Further studies are needed for determination of the histological structure of high- and low-flow arteries in humans.
Selected Abbreviations and Acronyms
|ESRD||=||end-stage renal disease|
|WCSA||=||wall cross-sectional area|
This study was partly supported by funds from the Assistance Publique-Hôpitaux de Paris (appel d’offres “Biologie du vieillissement” projet 94.17.17) and with grants from Institut National de la Santé et de la Recherche Médicale (INSERM).
Reprint requests to Prof Michel Safar, Service de Médecine Interne, Hôpital Broussais, 96, rue Didot, 75674 Paris Cedex 14, France.
Wolinsky H, Glagov S. A lamellar unit of aortic medial structure and function in mammals. Circ Res. 1967;20:99-111.
Dobrin PB. Vascular mechanics. In: Shepherd JT, Abboud FM, eds. Handbook of Physiology. Section 2, Volume 3: The Cardiovascular System. Washington, DC: American Physiological Society; 1983:65-102.
Kamiya A, Togawa T. Adaptative regulation of wall shear stress to flow change in the canine carotid artery. Am J Physiol. 1980;239:H14-H21.
Langille BL, O’Donnell F. Reductions in arterial diameter produced by chronic decreases in blood flow are endothelium-dependent. Science. 1986;231:405-407.
Langille BL, Bendeck MP, Keeley FW. Adaptations of carotid arteries of young and mature rabbits to reduced carotid blood flow. Am J Physiol. 1989;256(4, pt 2):H931-H939.
Girerd X, Mourad JJ, Copie X, Moulin C, Acar C, Safar M, Laurent S. Noninvasive detection of an increased vascular mass in untreated hypertensive patients. Am J Hypertens. 1994;7:1076-1084.
Roman MJ, Saba PS, Pini R, Spitzer M, Pickering TG, Rosen S, Aldermann MH, Devereux RB. Parallel cardiac and vascular adaptation in hypertension. Circulation. 1992;86:1909-1918.
Joannides R, Haefeli WE, Linder L, Richard V, Bakkali EH, Thuillez C, Lüscher TF. Nitric oxide is responsible for flow-dependent dilatation of human peripheral conduit arteries in vivo. Circulation. 1995;91:1314-1319.
Dobrin PB. Mechanical factors associated with the development of intimal and medial thickening in vein grafts subjected to arterial pressure: a model of arteries exposed to hypertension. Hypertension. 1995;26:38-43.
Pomposelli F, Schoen F, Cohen R, O’Leary D, Johnson R. Conformational stress and anastomotic hyperplasia. J Vasc Surg. 1987;51:126-136.
Girerd XJ, Acar C, Mourad JJ, Boutouyrie P, Safar ME, Laurent S. Incompressibility of the human arterial wall: an in vitro ultrasound study. J Hypertens. 1992;10(suppl 6):S133-S136.
Hoeks APG, Samijo SK, Brands PJ, Reneman RS. Noninvasive determination of shear-rate distribution across the arterial lumen. Hypertension. 1995;26:26-33.
Safar M, London G, Bedrossian J, Baglin A, Meyer P, Milliez P. The role of kidneys in hypertension in chronic renal failure. In: Proceedings of the 5th International Congress on Nephrology. Basel, Switzerland: Karger; 1974;3:91-98.
Heagerty AM, Aalkjær C, Bund SJ, Korsgaard N, Mulvany MJ. Small artery structure in hypertension: dual processes of remodeling and growth. Hypertension. 1993;21:391-397.
Kraiss LW, Kirkman TR, Kohler TR, Zierler B, Clowes AW. Shear stress regulates smooth muscle proliferation and neointimal thickening in porous polytetrafluoroethylene grafts. Arterioscler Thromb. 1991;11:1844-1852.
Rekhter M, Nicholls S, Ferguson M, Gordon D. Cell proliferation in human arteriovenous fistulas used for hemodialysis. Arterioscler Thromb. 1993;13:609-617.
Masuda H, Bassionny H, Glagov S, Zarins CK. Artery wall restructuring in response to increased flow. Surg Forum. 1989;40:285-286.