5 The Influence of Shear Stress on Restenosis

The Influence of Shear Stress on Restenosis

Attila Thury, ^MD , ^{P h D} , Jolanda J. Wentzel, ^{P h D} , Frank J. H. Gijsen, ^{P h D} ,

Johan C. H. Schuurbiers, ^B sc, Rob Krams, ^MD , ^{P h D} ,

Pim J. de Feyter, ^MD , ^{P h D} , Patrick W. Serruys, ^MD , ^{P h D} , and Cornelis J. Slager, ^{P h D}

C

ONTENTS

S

HEAR

S

TRESS AND

V

ASCULAR

(P

ATHO

) B

IOLOGY

S

HEAR

S

TRESS IN

P

ATIENTS

S

HEAR

S

TRESS AND

R

ESTENOSIS

A

FTER

B

ALLOON

A

NGIOPLASTY

S

HEAR

S

TRESS AND

R

ESTENOSIS

A

FTER

S

TENTING

C

^ONCLUSIONS

R

^EFERENCES

5

SHEAR STRESS AND VASCULAR (PATHO) BIOLOGY Hemodynamic Forces Acting on the Vessel Wall

Wall shear stress (WSS) is the (tangential) drag force acting on the luminal wall, induced by blood flow, normalized to wall area. As WSS is defined as force/area, its dimension equals that of pressure, i.e., N/m

²

or Pa. An older, frequently used unit for shear stress, i.e., dyne/cm

²

relates to Pa according to 1 Pa = 10 dyne/cm

²

. WSS on the endothelium (Fig. 1) can be calculated from the local shear rate (s

^–1

) times blood vis- cosity ( µ) (Pa/s). Shear rate is the spatial blood velocity gradient ([m/s]/m). Especially near the vessel wall, generally large velocity gradients between adjacent fluid layers exist and the shear stress is at its highest value. In a simple straight tube the Hagen- Poiseuille formula (WSS = 4µQ/πR

³

, with µ viscosity, R tube radius and Q flow) can be applied for steady laminar viscous flow. A normal WSS range of 0.68 ± 0.2 Pa was derived from Doppler based velocity measurements in angiographically normal coronary

From: Contemporary Cardiology: Essentials of Restenosis: For the Interventional Cardiologist Edited by: H. J. Duckers, E. G. Nabel, and P. W. Serruys © Humana Press Inc., Totowa, NJ

59

arteries of 21 patients (1). Considering the fact that the flow in the investigated branches varied from 6 to 123.2 mL/min, while blood viscosity was assumed constant (3.5 mPa/s), the surprisingly narrow range of shear stress values clearly demonstrates the efficacy of the regulation of coronary lumen size by WSS.

From a mechanical point of view, it is important to realize that WSS, when increased above normal levels, has an extremely small magnitude and therefore is not likely to have direct mechanical consequences that would impair the endothelium by wear out or erosion.

As an example, normal arterial WSS is less than the shear stress (~2–3 Pa) acting on the adhesive strip of a Post-it™ note pulled down by its weight when stuck to a vertical surface.

Arterial Adaptation to Shear Stress

Arteries sensitively respond both in short- and long-term to WSS changes in order to keep the local WSS in a narrow range. Therefore, it is generally accepted that a negative feedback loop exists between WSS and the vessel lumen dimensions with WSS as the controlling variable (2). The shear stress imposed on the endothelium by the movement of blood deforms the endothelial cells (ECs) to a microscopically small amount, but stim- ulates shear stress sensing elements in the ECs. These sensors induce the activation of second messenger systems (Fig. 2), which ultimately results in a biological response (3).

It has been well established (4) that the vessel dilates under the influence of an acute

increment in flow, thereby controlling the WSS in an artery (flow-dependent vasodilata-

tion). A variety of vasoactive substances are produced by the endothelium under the influ-

ence of WSS (5,6). The best-studied factor is endothelium-derived relaxing factor (7) that

appeared to be nitric oxide (8) and which is produced in a WSS dependent way (9). The

Fig. 1. Top: flow velocity profile in axial cross-section. Over a very small distance (bottom) velocity increases linearly with distance and wall shear rate (velocity difference/distance) approaches v_d/d.

endothelium also produces, as a response to alterations of WSS, prostacyclin and endothelin-1 (5,10). After sustained periods of WSS alterations the ECs accommodate to the new environment through several genes (11), including those of intercellular cell adhesion molecule-1, vascular cell adhesion molecule-1, and platelet derived growth factor (PDGF)- β. It has been shown that the regulation of some of these genes is dependent on shear responsive elements (12). More sustained stimulation with WSS remodels the organization of F-actin microfilaments and aligns the ECs to the stream- lines of the flow (12). Furthermore, integrins in the cell membrane tend to cluster after chronic WSS increments, which ultimately enable the EC to adapt its shape and pheno- type to local flow conditions (9). The mechanisms described previously involve many other aspects, which exceed the scope of this chapter. However, it is clear that acute and chronic adaptation of the arterial wall, are WSS (flow) dependent processes.

Role of Shear Stress in Localization of Atherosclerosis

The systemic nature of atherosclerotic risk factors cannot explain the observation that

atherosclerosis occurs predominantly at certain locations in the arterial tree including

Fig. 2. Endothelial responses to shear stress. The endothelium rapidly responds to sudden changes in WSS, with changes in membrane potential and in intracellular calcium concentration. These induce the cell-signaling cascades within the endothelial cells including activation of the mitogen-activated protein kinase (MAP kinase) signaling cascade. There are also changes within the cytoskeleton of the cell and the cell membrane, both of which are likely to facilitate the release of nitric oxide and other vasodilators, including prostacyclin. These immediate changes are followed within a few hours by changes in the regulation of a subset of genes. Intercellular adhesion molecule-I.

bifurcations and near the junction sites of side branches (13). As these predilection sites are associated with deviations of the normal velocity field, it has been postulated that flow-induced shear stress, acting on the ECs, plays an important role in plaque local- ization and plaque growth (14,15). Indeed, low or oscillating WSS results in cellular (e.g., monocyte) adhesion onto the vascular wall, lipid accumulation and oxidation, release of vasoactive substances, induction of growth factors, and smooth muscle cell proliferatory characteristics (16,17). It has not unequivocally been shown whether spatial gradients in WSS or low oscillatory WSS is the main modulator of atherogenesis, although most evidences point to the latter (13,18,19). For example, the carotid arteries are often unilaterally involved in atherosclerosis, with the affected region displaying lower mean WSS (15). In diabetics the common carotid arteries have a lower mean WSS than that found in normoglycemic individuals, with unilateral lesions again asso- ciated with localized areas of low WSS (20).

SHEAR STRESS IN PATIENTS

Computational Fluid Dynamics: In Vivo Application in Human Coronary Arteries

In vivo accurate measurement of the wall shear rate is difficult and at this moment it can- not be applied to map WSS distributions in human coronary arteries. Therefore, to deter- mine the local wall WSS in these arteries, computational fluid dynamics is applied (21). As a first step, a computational grid fitting in a three dimensions (3D) reconstructed lumen of human coronary arteries is generated (Fig. 3). Next, measured flow rate and patient specific blood viscosity data are transformed to the appropriate boundary condi- tions. The computational grid and the input conditions are subsequently fed into a numerical code to solve the Navier-Stokes equations and the continuity equation. This procedure renders the local velocity and WSS distribution at any point in the computa- tional grid of the reconstructed human coronary artery. This method has been applied for various relevant hemodynamical problems (21–26).

Atherosclerosis, Vascular Remodeling, and Shear Stress

Vascular remodeling has been observed in physiological conditions in response to changes in WSS, where it is aimed at restoring the original values of WSS (27). WSS acting on ECs is a major regulator of remodeling in developing blood vessels (28) and in blood vessels affected by atherosclerotic lesions (14). This is of importance, as it has been shown that the endothelium plays an essential role in vascular remodeling (29).

During early atherosclerosis, the control of lumen dimensions by WSS is still operable (30). Hence, Glagov et al. (31) suggested that WSS control might explain his observa- tions on lumen preservation by compensatory enlargement of arteries during early plaque accumulation. Only if the plaque area exceeded 40% of the intima-bounded area was lumen narrowing observed (31). A consequence of lumen preservation by compensa- tory enlargement is that plaque buildup remains clinically and angiographically unno- ticed until this compensation gradually fails or acute coronary syndromes occur.

Prolonged plaque accumulation at persisting low WSS predilection locations explains a negative relation between wall thickness (WT) and WSS (21).

Until recently, there has been no data on the relation between WT and WSS when lumen

narrowing and loss of compensatory remodeling commence. The authors investigated (32)

angiographically normal arteries (stenosis <50%) of 14 patients with ANGiography and ivUS (ANGUS) to provide 3D lumen and wall geometry (33). WSS of 25 segments was calculated by computational fluid dynamics. For segments of preserved lumens axially averaged WT and WSS were negatively related and positive vascular remodel- ing was observed. Narrowed segments showed no relation between WT and WSS or vascular remodeling. It was concluded (32) that in a patient with coronary arteries, the often-reported negative WT-WSS relationship appears restricted to lumen preservation and positive vascular remodeling. Its disappearance with lumen narrowing suggests a growing importance of non-WSS-related plaque progression.

Recently, Stone et al. (34) studied six native arteries of patients by 3D reconstruction and intravascular flow profiling at baseline and after 6-mo follow-up. Regions of abnor- mally low baseline WSS exhibited a significant increase in plaque thickness and enlargement of the outer vessel wall, such that lumen radius remained unchanged (out- ward remodeling). Regions of physiological WSS showed little change. Regions with increased WSS exhibited outward remodeling with normalization of WSS (34).

Fig. 3. (A) After lumen meshing, computational flow dynamics allows (B) detailed velocity determi- nation at any cross-section. From this (C) the local velocity profile and wall WSS (colored band) are derived. Ultimately, (D) WSS is calculated at any location of the lumen wall as shown in color-code.

Note, that in general at the outer curve of this human coronary example WT is smallest and WSS high- est, whereas the reverse can be noticed at the inner vessel curve. (Please see insert for color version.)

SHEAR STRESS AND RESTENOSIS AFTER BALLOON ANGIOPLASTY Vascular Changes After Balloon Angioplasty

Early balloon injury denudation models provided important insights into the mech- anisms of restenosis after angioplasty. Kohler and Jawien (35) examined the effects of blood flow on intimal hyperplasia after balloon catheter injury of the rat common carotid artery. By ligation of the ipsilateral carotid artery they obtained 35% decrease in blood flow and corresponding mean blood velocity. They found that early (2–4 wk after the injury) neointimal hyperplasia (NIH) increased when flow was reduced, and related it to alteration of smooth muscle cell migration (35). Thus, it is readily sugges- tive that evolution of the healing reaction is sensitive to flow and thereby to WSS (36).

Immediately after balloon angioplasty the endothelium is damaged and also after regeneration the endothelial layer may be dysfunctional (29). Balloon dilatation causes plaque fractures, breaks, dissection clefts and cracks extending from the lumen to resulting in injury to the intima (37). Owing to these spatial changes of the intimal sur- face, major changes in local WSS take place. In addition, localized medial dissection, stretching of the plaque-free wall and combination of plaque stretching and compres- sion, result in geometrical transitions that are also responsible for alterations in wall shear and tensile stresses. As blood flow is determined by dimensional adaptations to WSS and wall tensile stress (WS, force acting perpendicularly to a surface, normalized to area), a reasonable hypothetical paradigm has been set forth by Glagov (36). This describes the sequence of adaptive functional responses underlying outcomes after interventional injury of arteries; after modification of WSS and WS, these regulate the healing tissue responses (cell proliferation and matrix biosynthesis) until baseline (stable) levels of these stresses are re-established. If these levels are established, it will result in arrest of intimal thickening thus maintaining lumen patency. If it fails, persist- ent intimal proliferation or negative remodeling (considered to be the predominant cause of restenosis after angioplasty [38,39]) will cause restenosis. This hypothesis is descriptive, however, the causative factor for distinguishing between these two processes is not clarified. Below, arguments and findings further describing this hypothesis are given.

Low Shear Stress Related to Restenosis

Segmental coronary blood flow may remain impaired after angioplasty (40) because of microvascular dysfunction, infarction of some of the tissue downstream, or the pres- ence of persistent competing collateral circulation. Increased incidence of restenosis in patients has been repeatedly associated with reduced flow (41,42) and abnormally low flow reserve (43). The mechanisms by which these hemodynamic forces may modulate the vascular smooth muscle cells (VSMC) proliferative and migratory responses in vivo include induction of mitogenic cytokines, such as PDGF. PDGF contributes to these responses, as both PDGF and its receptor ( β-type) are upregulated by injury (44).

Other mediators, such as matrix metalloproteinases are also involved (45).

As WSS and WS have been implicated as regulators of vascular remodeling during

physiological conditions, the authors evaluated their role in vascular remodeling after

balloon angioplasty in a well-accepted pig model (46). It was argued that if WSS and/or

WS are important regulators in vascular remodeling after balloon angioplasty then they

should comply with the following criteria. First, WSS or WS will change after balloon angioplasty, second, the change of WSS or WS induced by balloon angioplasty should be predictive of vascular remodeling, and third, vascular remodeling should arrest when WSS and WS have been returned to “normal values.”

The results showed (46) that a significant decrease in WSS and increase in WS after balloon angioplasty was reached and these changes measured at baseline were predic- tive for vascular remodeling at 6 wk follow-up. Moreover, data showed that the WSS and WS were returned to the reference values. As all data in respect to the values of ref- erence segments were normalized, results were fully independent of the assumed inflow conditions for the WSS calculation. Hence, from these results it can be concluded that in the Yucatan atherosclerotic pig model, vascular remodeling after gentle balloon angioplasty is controlled by a WSS and a WS negative feedback mechanism, aiming at keeping WSS and WS constant (Fig. 4).

Because the integrity of the endothelial layer is damaged directly after balloon angio- plasty pathways other than the endothelium might get activated. Sterpetti et al. (47) reported that VSMCs also sense WSS and that increasing WSS directly inhibits whereas decreasing WSS facilitates VSMC proliferation. Recent studies indicate that when cul- tured VSMCs are in the synthetic phenotype, they respond to WSS in an analogous manner to the endothelium, altering their production of growth factors (48). In the bal- loon catheter–injured and de-endothelialized rat carotid artery, the VSMCs that migrate through the internal elastic lamina to form the neointima rapidly change to the syn- thetic phenotype, and they maintain phenotypic modulation until at least 2 wk after injury (49). Thus, it is possible that juxtaluminal synthetic VSMCs could respond to abnormal shear forces in a manner similar to the endothelium and hence potentially influence inward remodeling.

Ward et al. (41) examined how increases or decreases in blood flowthrough balloon catheter–injured rat carotid arteries affected vessel morphometry, cell migration, and levels of promigratory mRNAs. After 28 d, the luminal area in vessels with low blood flow was significantly smaller than in those with normal and high blood flow, predom- inantly because of accentuated inward remodeling. Low flow also enhanced VSMC migration 4 d after injury by 90% above normal and high flows. They concluded (41) that low blood flow might promote restenosis after angioplasty because of its adverse effect on vascular remodeling, and its association with the augmented expression of multiple genes central to cell migration and restenosis.

However, other than negative remodeling, additional potential mechanisms for

restenosis after angioplasty should be mentioned. Bassiouny et al. (45) found that

platelet activation after experimental arterial injury as measured by thromboxane B

₂

levels was greater with injured arteries subjected to reduced flow compared with nor-

mal and increased flow conditions. This suggests that platelet adhesion to regions of

injury is enhanced in a low-flow and low-shear environment. Activated platelets release

many growth factors, including PDGF, basic fibroblast growth factor, and transforming

growth factor. These factors play an important role in regulating VSMC migration and

proliferation (50). Conclusively, studies point at low WSS leading to inward remodel-

ing after balloon angioplasty ultimately resulting in restenosis. However, the reason

why the physiological feedback mechanism (i.e., process to restore normal range of

WSS) passes the point of the normal lumen settings and fails by further lumen loss is

not yet clear.

Flow Eddies After Balloon Angioplasty and Restenosis

It has been suggested that apart from uniform sluggish laminar flow, irregular flow patterns including eddies that result from high average flow velocities and irregular lumen may also predict restenosis after angioplasty (51). Similarly, after balloon angio- plasty, mild residual stenosis with complex ultrastructure of intimal flaps and cracks may be optimal for development of local small oscillating flow patterns, which might arise within the “normal” range of brisk blood flow.

The authors published an appealing finding (52) when average WSS was calculated from the gross blood velocity using a Doppler wire proximal to, in and distal to the lesion after angiographically successful balloon dilatation. The Hagen-Poiseuille for- mula was applied and volumetric blood flow and lumen radius were derived from Doppler velocities and videodensitometric cross-sectional areas. Postprocedural proxi- mal and in-lesion WSS values were higher in vessels that developed restenosis (n = 72;

1.2 ± 0.6 and 3.6 ± 2.4 N/m

²

, respectively) than those without (n = 130, 1.1 ± 0.5 N/m

²

and 2.5 ± 1.4, respectively, p < 0.05). In-lesion WSS was predictive of restenosis,

whereas WSS of the proximal segment was associated with an increased rate of target

lesion revascularization (odds ratio = 2.33, p < 0.005). Moreover, proximal high WSS

was the only independent predictor when entered with known predictors as diameter

stenosis and coronary flow reserve (odds ratio was 2.15, p < 0.05). This paradoxical

finding raised the following arguments as theoretical explanations (52). Although high

average WSS values for the entire cross-section were calculated in the patient group, no

certainty about the spatial distribution of the WSS could be investigated. Indeed, it is

possible that next to a high average lumen WSS, in the adjacent wall regions of flaps

and cracks low WSS zones exist. These zones of low WSS can also coexist with flow

separation zones inducing strong secondary and oscillatory flows (53). It might be that

especially these regions are prone to the evolvement of restenosis: possibly the patients

that developed restenosis had more severe vessel injury after angioplasty, with more

flow separation and more regions of secondary flow and localized low WSS. One might

Fig. 4. Proposed mechanism of the influence of geometry on control mechanisms during the follow- up period after PT(C)A.

also speculate that high flow velocity across residual stenosis results in low WSS or flow separation downstream at the immediate distal side of the lesion (not the site of the distal flow velocity measurement), known to be a preferential part for plaque pro- gression (54).

Role of Collateral Circulation

Collateral arteries potentially offer an important alternative source of blood supply when the original vessel fails to provide sufficient blood (55). The mechanisms of col- lateral vessel growth are vasculogenesis (initial migration of angioblasts to discrete locations and further differentiation) and subsequent arteriogenesis (transformation of pre-existing collateral arterioles into functional arteries) (55). The process of arterio- genesis is mediated through an increase in WSS (56). In the event of a hemodynami- cally relevant stenosis of a main feeding artery, a pressure gradient is created and pre-existing collateral pathways are recruited. This results in an increased flow velocity and therefore increased WSS in the pre-existent collateral arteries, which leads to a marked activation of the endothelium and upregulation of cell adhesion molecules.

Subsequently, several morphological changes and vascular remodeling occur (56) ulti- mately resulting in matured muscular arteries.

Werner et al. (57) studied the functional and anatomical time-dependent regression of human collaterals after recanalization of chronic total occlusions. Collateral function was assessed by intracoronary Doppler and pressure recordings before and after recanalization, and again after 5 mo (57). Immediately after restoration of antegrade flow, collateral function is attenuated, with a further regression during follow-up.

However, the complete collateral recovery observed in patients with reocclusion during follow-up suggests that collaterals remain recruitable during recurrence of an occlusion and do not disappear completely after recanalization. The mechanisms involved in col- lateral regression and recovery are probably flow-dependent changes of the collateral vascular tone (58). This was also documented in the study of Werner et al (57) by the increase of collateral resistance index, which reflects the functional (flow-mediated) state of the collateral vessel. On the other hand, a study after stenting showed (59) that there is no relation between a well-developed collateral supply and the risk of reocclu- sion in recanalized chronic total occlusions; at that time this was rather determined by the stented segment length.

SHEAR STRESS AND RESTENOSIS AFTER STENTING Vascular Changes Caused by Stent Implantation

C

HANGES IN

G

EOMETRY BY

S

TENT

I

MPLANTATION

R

ESULT IN

C

HANGES OF

F

LOW

D

YNAMICS

Before the advent of polymer coated and drug-eluting stents (DES) intrastent restenosis occurred in 20–30% of the cases following a bare metallic stent implantation (60). NIH, made up of VSMC proliferation and extracellular matrix, is considered to be a major mechanism (61). Three distinct phenomena lead to NIH:

1. The expansion of the stent wires at the time of implantation can result in severe vascu- lar trauma (overstretch injury), which has been quantitatively related to the extent of NIH (62).

2. Stent implantation induces complex interactions between blood and vessel wall com- ponents, and the metal structure of the prosthesis. The materials and the roughness of stent surface can affect the adsorption of plasma proteins, which generates an inflam- matory response (63).

3. Stent implantation alters the coronary flow and induces changes in pressure distribu- tion and local flow velocity. These changes result in the local modification of the mechanical stresses exerted on the ECs.

One of the main effects on vessel hemodynamics caused by the stent is the longitu- dinal straightening effect. The authors evaluated the regional changes in 3-D geometry and WSS induced by stent placement in coronary arteries of pigs (22). The previously quoted ANGUS technique (33) was applied after implantation of Wallstents in seven coronary arteries of five pigs. This 3-D geometry was used to calculate locally the cur- vature, whereas the WSS distribution was obtained by computational fluid dynamics.

Local changes in WSS were obtained at the entrance and exit of the stent. After stent implantation, the curvature increased by 121% at the entrance and by 100% at the exit of the stent, resulting in local changes in WSS (22). In general, at the entrance of the stent, increased local maxima in WSS were generated, whereas at the exit both local maxima and minima in WSS were observed. Additionally, the WSS at the entrance and exit of the stent correlated with the local curvature (r = 0.30–0.84, p < 0.05). Thus, it was postulated (22) that nonuniform distribution of in-stent restenosis (i.e., higher rate of edge restenosis) may arise from changes in WSS (resulting from altered 3-D vessel geometry) caused by stent implantation.

Similar mechanism was also demonstrated in the clinical setting (64). Gyongyosi et al. (64) showed that prestent vessel angulation and the changes in vessel angulation after stent implantation were significant independent variables predicting angiographic severity of restenosis.

E

^{FFECT OF}

S

^{TENT AND}

S

^TENT

’

^S

S

^TRUT

D

^ESIGN

However, changes in a given cross-section also occur, at deployment; a stent stretches a vessel, imposing a cross-sectional polygonal luminal shape that depends on the stent design, with each strut serving to generate a vertex (65). Garasic et al. (65) found that stents designed with 12 struts per cross-section had twofold less neointimal area than identical stents with only eight struts per cross-section. Postdeployment luminal geo- metry dictated by stent design, determined neointimal thickness (NIT) independently of the extent of arterial injury. Moreover, this group could predict the extent of NIH with mathematical modeling based on local geometrical changes (65).

Arterial injury has been repeatedly proven to be directly related to NIH formation (62,66) and been attributed to different degrees of deep vessel laceration and in some cases medial rupture (62). Using various implantation conditions Gunn et al. (67) estab- lished a new injury score, incorporating both stretch and deep injury, which correlated with neointimal area (r = 0.60, p < 0.001). A variety of design parameters, including cell geometry, strut thickness, acute recoil and surface characteristics (including electro- mechanical characteristics) have an important effect on clinical outcomes.

However, the placement of a stent in an artery not only results in change or injury to

a vessel segment as a whole, but also affects the details of flow adjacent to the artery

wall (68). A typical stent strut is approx 15 µm in thickness, and stents are deployed

into arteries at least 3 mm in diameter. Thus, the direct flow effects of stent strut protru-

sion into the lumen are confined to a region close to the artery wall. However, it is

important to realize that the stent struts are covered with thrombus within a few hours or days following implantation. Thus, a study on suggested backward and forward blood flow over stent struts is only relevant to the acute stages of implantation. However, there is some evidence that the stent strut geometry is still represented in the neointimal devel- opment patterns even weeks after implantation (69). The effects of stent strut spacing on blood flow patterns and platelet deposition adjacent to the artery wall have also been studied using computational flow dynamics techniques (70,71). They found that with smaller stent wire spacings less than six wire diameters, the stagnation regions from adjacent struts merged together for the entire cardiac cycle, creating one single recircu- lation region (70). For wire spacings larger than six wire diameters, the stagnation regions were split for at least some portion of the cardiac cycle, with the flow reattach- ing in between. Whether these flow-related design characteristics are associated with less restenosis is unknown.

Relation of Shear Stress and NIH

I

NTIMAL

P

ROLIFERATION IN

E

NDOVASCULAR

G

RAFTS

The flow-related phenomena of buildup of NIH after implantation of an endovascular graft have been the subject of many studies. The first experiment (72), investigating the influence of WSS on neointimal formation after intervention was performed by Kohler et al. in 1991. He showed in endothelialized grafts placed in baboons that increasing the WSS in the graft reduced the neointimal formation at follow-up (72). With reduced flow, the NIT was increased. Similar observations were reported by Salam et al. (73) in polytetrafluoroethylene bypass grafts implanted in dogs. At the low WSS regions the neointima was significantly thicker than at the high WSS regions (73). As a possible mechanism, Kraiss et al. (74) have reported upregulated expression of PDGF when flowthrough prosthetic grafts is reduced in a baboon model. Furthermore, increasing the WSS even 2 mo after graft implantation still induced a regression of the neointima (75). WSS related neointimal formation has been shown at the edges of anastomosis as well (76). At these edges, eddies in the flow as well as low WSS regions were observed, which were related to the neointimal growth (76). These findings suggest that WSS is involved in the generation of neointimal formation.

R

E

-

ENDOTHELIALIZATION OF

S

TENT

S

URFACE

Perhaps the most important aspect of EC behavior regarding stents is their ability to

regrow over the denuded artery wall and stent surface. The first experiments suggesting

WSS to be related to in-stent neointimal formation through affecting re-endothelialization

refer to in vitro experiments. Sprague et al. (77) found that the re-endothelialization of

stent material appeared to be WSS dependent, such that the low WSS regions are delayed

in endothelialization compared with high WSS regions (77). As re-endothelialization is

assumed to play a key role in controlling neointimal formation after stent implantation

(78), this observation suggests that also in-stent neointimal formation might be related

to WSS. Furthermore, in resemblance to the neointimal formation observed in vascular

bypass grafts or anastomosis, it is not unlikely that WSS has similar influence on the

vascular repair process after stent implantation, thereby influencing the neointimal for-

mation. For the enhancement of re-endothelialization, many studies were reported (78)

and the biocompatibility of the generally used polymer stent coating is still a crucial

issue even in the era of DES.

G

LOBAL

E

FFECTS OF

S

HEAR

S

TRESS ON THE

D

EVELOPMENT OF

NIH

Neointima formation is often observed at specific locations in the stented segment (79). Localizing factors that have been studied include plaque burden (80) and wall (tensile) stress (81). Until now the role of WSS in neointimal formation within coro- nary stents has been hardly recognized and referred mostly to in vitro experiments. In order to study the role of WSS in neointimal formation, the authors studied 14 coronary arteries of patients 6 mo after Wallstent implantation (23). Figure 5 shows an example of a 3D reconstruction (33) of a stented right coronary artery of a patient. The NIT and WSS of the same patient are shown in (Fig. 6). From these figures it can be observed that neointimal growth was least at relatively high WSS locations and smaller than at low WSS locations (p < 0.05). The average relationship between NIT and WSS for all 14 patients was: NIT = (0.59 ± 0.24) ( (0.08 ± 0.10) × WSS mm, (p < 0.05). These data show similarity to the results obtained from animal studies, where in-stent neointimal formation and that in bypass grafts (72,73,82) was found to be related to low WSS.

When considering the geometrical distribution of NIH, one must account for pre- existing eccentric lesions, one of the possible confounding factors, which may point to existence and progression of atherosclerosis at the low WSS side (21). However, previ- ous studies only report on the relationship between cross-sectional area of the plaque burden and neointima formation (80,83). No data exist concerning, whether the asym- metries in neointimal formation may be explained by the eccentricity of the persistent plaque burden. Recent investigation (34) failed to show a relation between baseline WSS values and subsequent NIH. Clearly, this subject warrants further studies.

O

SCILLATORY

F

LOWS

: S

TEP

-U

P

Implantation of a stent in a coronary artery often results in “step-up” of the lumen as interventional cardiologists call it. Step-up is generally considered to be a favorable pattern after stent-deployment. Nevertheless, its optimal degree is unknown. The authors investigated (24) this phenomenon after 3D arterial reconstruction (33) in a patient who had underwent coronary angioplasty of an occluded proximal and midleft anterior descending coronary artery. Although the stent was well apposed and deployed as indicated by intravascular ultrasound, the patient presented with worsening anginal symptoms 4 mo later. Both the angiogram and IVUS showed focal in-stent restenosis at the proximal edge of the proximal stent (78% on quantitative coronary analysis) and mild diffuse NIH through the entire length of the stent. The result of computational fluid dynamics provided the WSS at the wall as a function of time over the cardiac cycle. The ANGUS procedure was repeated when the patient presented at 4 mo, and NIH was determined from this 3D reconstruction (Fig. 7F). As found previously (23), NIH was highest near the locations where average SS was low (Fig. 7E–G). Subsequently, the tem- poral WSS pattern in the region of the step-up (Fig. 7C) was evaluated. This analysis showed (24) that the WSS vectors were either permanently or temporarily retrogradely directed near the “corner” of the step-up. This indicates the existence of a region of flow separation (Fig. 7D). In Fig. 8C, locations showing retrograde axial velocities are presented in black at five time-points as indicated in the Doppler recordings (Fig. 8B).

At locations that temporarily experience retrograde velocities, WSS alters direction

periodically. Interestingly, those locations of oscillating WSS were nearest to the area

of highest NIH (Fig. 8A). This case warrants (24) that such a marked diameter increase

might translate into severe lumen obstruction in the long-term. The mechanisms behind

this observation are diverse, but the results suggest that oscillating WSS may play a role

in in-stent neointimal proliferation.

S

^{TUDY IN}

DES

DES represent the current most effective tool for the prevention of restenosis after PCI (84). As described previously, in uncoated (nondrug-eluting) Wallstents a negative relationship between NIT and WSS was observed by the author’s group (23). This rela- tionship was assigned to the restored WSS related response of the endothelium that limits intimal growth in high WSS regions (75). However, some neointimal growth was observed at all locations of the stented segment.

In case of DES, some incomplete apposition of the stent struts has been observed (85) in sirolimus-eluting stents (SES) at follow-up. This could be attributed to the com- bination of inhibition of VSMC proliferation and clearance of postprocedure throm- botic material or apoptotic cells from the damaged vessel wall (85). In addition, a temporary sirolimus-stimulated production of prostacyclin (86), which is a potent vasodilator, may also have played a role. The absence of late thrombotic events in the patients of the RAVEL trial (87) and the results from animal experiments (88) indicate that the endothelium in the stented segment may be restored.

Fig. 5. (A) Angiogram of a stented right coronary artery (B) 3D ANGUS reconstruction of the right coronary artery showing vessel lumen and wall.

Drug release rates from coated stents approximate classic first-order kinetics, where a fixed fraction of elutable drug is released per unit time (89). The kinetics of elution follow the Fick`s law of diffusion, but on the lumen side blood flow, i.e., actual shear rates near the surface of each stent strut will greatly influence the dissolution of the drug. This latter relation gave rise to the initiation of the study (25) to investigate NIT distribution in SES. The authors obtained the 3-D geometry of arteries in 6 patients at 6 mo follow-up after implantation of a SES. NIT was defined as the distance between the extrapolated lumen stent surface and the (neo) intimal tissue border. WSS and NIT data were axially averaged to study their relationship. NIT was −0.03 ± 0.03 mm and shallow pits were observed between the stent struts in all patients. In-stent WSS was 0.46 ± 0.06 Pa and blood viscosity was 3.1 ± 0.4 mPa s. Combining the data of all patients, a significant inverse correlation was found between NIT and WSS: NIT = 0.21 mm −0.24 * WSS (r

²

= 0.24, p < 0.001). In this study, the negative relationship between NIT and WSS implied that deeper pits between the stent struts were present in the high WSS regions along the outside curve of arterial segments.

The presence of the shallow pits in the high WSS regions triggered the investigation of the hemodynamics at follow-up more closely. It was found that in the deeper pits in the high WSS zone, the vessel wall experiences very low WSS values and flow recircu- lation zones may even develop (Fig. 9), which usually are associated with intimal hyperplasia (9,24). However, endothelial cells in the high WSS region surrounding the pits may create a larger WSS protected environment by local diffusion or transport (90) of WSS induced production of particularly nitric oxide. This will inhibit VSMC prolif- eration and may explain the persistence of shallow pits between the stent struts in the high WSS regions. This phenomenon was further evaluated long-term with repeated intravascular ultrasound and clinical follow-up (91). SESs did not show further increase in malapposition or change in vessel dimensions and were not associated with late thrombosis (91). In the low WSS regions along the inner bend, the endothelium lacks sufficient WSS stimulation, which may explain the filling of the gaps between the stent struts. In addition, the conventional risk factor including previous plaque burden (80), being usually present at low WSS locations (21), may have also contributed to this gap filling. The findings suggest that further exploration of local hemodynamics might help

Fig. 6. 2D-map of neointimal thickness in the human coronary artery shown left. Corresponding 2D- map of WSS in the human coronary artery shown at the right. Left to right: vessel circumference.

Bottom to top: proximal to distal axial vessel location.

to clarify the mechanism of incomplete stent apposition or edge restenosis occurring after DES implantation.

The possibility of edge effects, analogous to those observed with radioactive stents and after intravascular brachytherapy, might limit the effectiveness of DES implanta- tion. In the initial trials with SES, no edge effect was reported. In the SIRIUS trial (92) (a multicenter study of the SIRolImUS-eluting Bx-velocity stent in the treatment of patients with de novo coronary artery lesions), which evaluated SES in a more complex population, a higher rate of significant stenosis was observed at the proximal edge of the SES than at either the stented region or its distal edge (92). In contrast, a significant reduction in late lumen loss beyond the distal stent edge was recently observed with paclitaxel-eluting stents (93). The difference between proximal and distal edges draws attention to the fact that the elution process is dependent on flow characteristics, for example, shear rate.

Fig. 7. (A) Lateral angiographic view of the left anterior descending coronary artery after stent place- ment. Open arrow indicates location of step-up. (B) 3D (ANGUS) reconstruction of the coronary artery shown in A, clearly showing the step-up phenomenon at the proximal edge of the stent (open arrow). (C) Segment in which detailed analysis of the temporal WSS variations is performed. (D) Cartoon showing the existence of a region with retrograde velocities and flow separation. (E) Averaged WSS over the cardiac cycle color-coded at the surface of the stented region of the 3D reconstruction.

(F) Neointimal thickness color-coded at the lumen surface of the stented region. (G) In-stent average neointimal thickness per cross-section vs the WSS averaged over the cardiac cycle and per cross- section showing a nonlinear inverse relationship (NIH= 0.3 + 0.2 × WSS⁻¹[mm]; r²= 0.34, p < 0.01).

(Please see insert for color version.)

Therapeutic Implications G

LOBAL

E

FFECT OF

S

HEAR

S

TRESS

: O

VERSIZING OF

S

TENTS

Kuntz et al. (94) introduced the concept “the bigger the better” in balloon angio- plasty: the “bigger” the lumen size postprocedure, the “better” the expected reduced restenosis rate. Serruys et al. (95) showed that it also holds true for coronary stent implantation. Recently, Sick et al. (96) carried out a retrospective study on a large num- ber of patients. They classified lesions into four subgroups according to the degree of residual stenosis after stent implantation: group 1, gross oversizing < −15%; group 2, slight oversizing –15% to <0%; group 3, mild residual 0% to <15%; group 4, moderate residual stenosis of 15% to <30%. In 1882, stenoses with angiographic follow-up, gross oversizing of stents led to a significantly (p < 0.001) higher increase of percent stenosis associated with a higher restenosis rate (group 1: 34.7% vs groups 2, 3, and 4: 32.5%, 28.2%, and 29.6%, respectively). A multiple regression analysis was performed with conventional risk factors. Optimal results regarding stent thrombosis and restenosis were achieved with mild residual stenoses between 0% and 15% after stent implantation.

Fig. 8. (A) Neointimal hyperplasia, which is color-coded at the 3-dimensionally reconstructed lumen at baseline. The perspective view of Figure 8A and 8C differs from Fig. 7. (B) Doppler measure- ments used for the time-dependent flow calculations. (C) At 5 time-points during the cardiac cycle, locations (black) in the stent experience retrograde axial velocities. As can be observed in this view, the size of this area changes over the cycle, which implies that locations with temporary retrograde velocity experience oscillating WSS. (Please see insert for color version.)

This is in good accordance with the earlier-described finding of the step-up phenome- non and further encourages the acknowledgment of low WSS as a causative mechanism for enhanced neointimal growth. One might speculate that in case of stent oversizing by 15% (as group 1 in the study of Sick et al. [96]), which approximately lowers WSS (proportional to 1/diameter

³

) by 39%, might have been indeed associated with the increased NIH in the long-term. Similarly, undersizing by 15% enhances WSS by approx 63%, thus providing a strong inhibitory effect for neointimal growth. Thus, it might be concluded that oversizing of stents is no longer necessary with current stan- dards of antiplatelet treatment.

E

NHANCEMENT OF

S

^HEAR

S

^TRESS

: F

^LOW

D

^IVIDER

In line with the earlier-described research on Wallstents, a new therapeutic experi- mental device was developed to increase WSS aiming at reduction of neointimal for- mation. This device (97) (Endoart, Lausanne, Switzerland) consists of a small cylinder (1 mm diameter) placed in the middle of a (stented) arterial segment and divides the flow, thereby increasing WSS while keeping the pressure drop below 1 mmHg. The authors performed (97) an animal study to evaluate the possible therapeutic application of the flow divider. In this investigation, the iliac arteries of nine rabbits were stented.

Fig. 9. Two-dimensional axial cross-section of a slightly curved stented segment (B). The results of a computation of velocity and WSS distribution in this segment at follow-up show regions with low WSS (A) that coincide with the shallow pits in B. Along the inner curve, the region with lower WSS, the pits are virtually absent and WSS distribution is much more homogeneous (C). In some of the pits along the outside curve B, flow reversal can be observed (inset and D). The areas containing reversed axial velocity are indicated by the shaded boxed regions in B.

One of the stented arteries was randomly equipped with the flow divider, whereas the other artery served as a control. In the arteries with flow divider, CFD computations indi- cated that the WSS in the stented segment increased by 100%. The resulting neointimal formation in the stented arteries with the flow divider was compared with the stented seg- ments without the device. Histological analysis showed (97) that the presence of the flow divider reduced neointimal formation by approx 40% (Fig. 10). This large effect identifies WSS as a major independent modulating factor in restenosis and merits further research on methods to, at least temporarily, create a local increase of WSS to suppress NIH.

E

NHANCEMENT OF

S

HEAR

S

TRESS

: P

HYSICAL

E

XERCISE

The notion that regular aerobic exercise reduces cardiovascular morbidity and mortal- ity in the general population as well as in patients with coronary artery disease is strongly supported by evidence derived from epidemiological studies (98). More recently, exer- cise-induced increases in blood flow and WSS have been observed to enhance vascular function and structure (99). Quantitative angiographic studies (100,101) have revealed that exercise programs reduce the progression of coronary artery disease and even showed superiority over percutaneous intervention in a certain patient subset in a recent study (102). These observations provided the basis for the hypothesis that exercise may also induce changes in lumen diameter in patients after coronary angioplasty. Indeed, patients randomized to a 12-wk intervention program consisting of daily exercise after balloon angioplasty enjoyed a significantly lower rate of restenosis than patients in the control group (103). A recent study of Indolfi et al. (104) showed a significant reduction of both NIH after stenting and negative remodeling 14 d after balloon injury in trained compared with sedentary rats.

L

-NMMA administration eliminated the benefits of physical training on vessel wall after balloon dilation, which signifies that increase in WSS related eNOS expression and activity might contribute to the potential beneficial effects of exercise.

Moreover, they demonstrated (104) that exercise training produced accelerated re- endothelialization of the balloon injured arterial segments compared with sedentary. It can partly be attributed to a very recent finding of Laufs et al. (105) who demonstrated that physical activity increases the production and circulating numbers of endothelial pro- genitor cells (EPCs). Concerning the general beneficial effect of exercise training on car- diovascular events, a new study (106) revealed a possible mechanism: in the apoE-deficient mice, exercise training reduced neointimal growth and even stabilized vas- cular lesions after injury. This was based on the reduction of the number of Mac-3–posi- tive, oxidized LDL-containing macrophages in the vessel wall and the increased content of collagen fibers. Plasminogen activator inhibitor-1, tissue factor, and fibrinogen were all significantly reduced in the lesions of trained mice (106). These observations further encourage population-based studies to establish the optimal duration and intensity of exercise, which should, preferably be individually recommended for each patient.

P

ROGRAMMABLE

DES

A next-generation coronary drug delivery system (Conor MedStent™) has recently

been proposed as being engineered specifically to allow programmable control over

spatial and temporal release kinetics of different drugs and to enhance drug-loading

capacity (107). This stent has relatively large reservoirs (holes) being able to elute any

drug (micro to macro molecules), oligonucleotides, microspheres, cDNA toward the

wall and the lumen in a coronary artery. However, the transport of certain materials by

blood and uptake into the vessel wall are complex processes (108) and found to be

dependent on the nature of the flow in the vessel and blood rheological properties including the red cell concentration (109). One of the promising perspectives is to address the treatment of distal vulnerable plaque or diffuse disease by long time drug delivery to the lumen distal of the stent. However, because of the complex process of endothelialization and drug diffusion from the possibly tissue covered reservoirs the feasibility of drug delivery over a long time period is yet unknown. Extensive studies are needed to test the theory and efficacy of such a therapy with various agents.

S

TENTS

S

EEDED

W

ITH

EPCs

In order to enhance rapid re-endothelialization, early attempts were reported

(110,111) using ECs. However, difficulty of harvesting autologous ECs from patients

has hampered clinical usage of this method (112). Recently, EPCs have been used for

fabrication of new cell-seeded stents (113). Circulating EPCs have been demonstrated

in the peripheral blood and shown to differentiate into a functional EC type, along with

the ability to traffic to damaged vasculature (114). Moreover, impaired adhesion and

reduced numbers of circulating EPCs in patients with diffuse in-stent restenosis was

detected (115). Another problem to solve is how to achieve a stent surface properly

Fig. 10. Cross-sectional histological specimens of iliac arteries of rabbits after placement of a flow divider in the center of a stent that increases WSS at the stent surface. This resulted in a reduced intimal hyperplasia as described in the text.

seeded with a sufficient number of ECs/EPCs, capable of withstanding balloon trauma at implantation. This issue has been recently addressed with the newly designed Genous

^™

Bio-engineered Surface (Orbus Medical Technologies, Inc. FL). This utilizes antibodies to capture the patient’s own circulating EPCs.

Blood flow most certainly plays a crucial rule when attachment of EPCs to antibodies and their differentiation into mature EC lining are considered. Intuitively, two opposing effects of the magnitude of WSS at a certain location of the stented surface take place.

Low WSS by increasing near-wall residence time of any particle or cell favors EPC cap- ture. A proof for this was elegantly presented in a recent report (116) by establishing a representative quantitative correlation between monocyte deposition and residence time. On the other hand, low WSS may stimulate thrombosis possibly covering the antibodies. High shear rate causing short residence time and increased detachment force might prevent EPCs to be arrested by antibodies (117). Maintenance of the EC monolayer is also dependent on physiologically relevant (high) WSS conditions (118,119). EPCs are more sensitive to WSS than matured ECs (120) and moreover, WSS was shown to stimulate EPC proliferation and their alignment in the direction of flow (120). One might consider these mechanisms to be hardly optimally distributed at all locations of the stent struts. Further studies are encouraged to investigate the delicate WSS related processes of EPC capturing and proliferation.

CONCLUSIONS

WSS is of paramount importance in vascular biology because of its multiple effects on EC. WSS explains the local and eccentric nature of atherosclerosis and vascular remodeling. With advanced technologies it became feasible to investigate these mecha- nisms and its effect after vascular intervention. Low and oscillating WSS have been associated with restenosis after balloon angioplasty and stent implantation. The process of neointima growth has many WSS related facets. The manipulation of WSS might have future clinical implications as has been already shown in the case of the flow divider.

REFERENCES

1. Doriot PA, Dorsaz PA, Dorsaz L, et al. In-vivo measurements of wall shear stress in human coronary arteries. Coron Artery Dis 2000;11:495–502.

2. Kamiya A, Togawa T. Adaptive regulation of wall shear stress to flow change in the canine carotid artery. Am J Physiol 1980;239:H14–H21.

3. Fisher AB, Chien S, Barakat AI, et al. Endothelial cellular response to altered shear stress. Am J Physiol Lung Cell Mol Physiol 2001;281:L529–L533.

4. Lie M, Sejersted OM, Kiil F. Local regulation of vascular cross section during changes in femoral arterial blood flow in dogs. Circ Res 1970;27:727–737.

5. Davies PF, Tripathi SC. Mechanical stress mechanisms and the cell. An endothelial paradigm. Circ Res 1993;72:239–245.

6. McIntire LV. 1992 ALZA Distinguished Lecture: bioengineering and vascular biology. Ann Biomed Eng 1994;22:2–13.

7. Furchgott RF, Zawadzki JV. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 27 1980;288(5789):373–376.

8. Furchgott RF, Jothianandan D. Endothelium-dependent and -independent vasodilation involving cyclic GMP: relaxation induced by nitric oxide, carbon monoxide and light. Blood Vessels 1991;28:52–61.

9. Malek AM, Alper SL, Izumo S. Hemodynamic shear stress and its role in atherosclerosis. JAMA 1999;282:2035–2042.

10. Redmond EM, Cahill PA, J.V. S. Flow-mediated regulation of endothelium receptors in cocultered vascular smooth muscle cells: an endothelium-dependent effect. Vasc Res 1997;34:425–435.

11. Gimbrone MA Jr, Nagel T, Topper JN. Biomechanical activation: an emerging paradigm in endothe- lial adhesion biology. J Clin Invest 1997;100:S61–S65.

12. Resnick N, Yahav H, Khachigian LM, et al. Endothelial gene regulation by laminar shear stress. Anal Quantitative Cardiol 1997;13:155–164.

13. VanderLaan PA, Reardon CA, Getz GS. Site specificity of atherosclerosis: site-selective responses to atherosclerotic modulators. Arterioscler Thromb Vasc Biol 2004;24:12–22.

14. Zarins CK, Giddens DP, Bharadvaj BK, et al. Carotid bifurcation atherosclerosis. Quantitative corre- lation of plaque localization with flow velocity profiles and wall shear stress. Circ Res 1983;53:502–514.

15. Gnasso A, Irace C, Carallo C, et al. In vivo association between low wall shear stress and plaque in subjects with asymmetrical carotid atherosclerosis. Stroke 1997;28:993–998.

16. Khachigian LM, Anderson KR, Halnon NJ, et al. Egr-1 is activated in endothelial cells exposed to fluid shear stress and interacts with a novel shear-stress-response element in the PDGF A-chain pro- moter. Arterioscler Thromb Vasc Biol 1997;17:2280–2286.

17. Hagiwara H, Mitsumata M, Yamane T, et al. Laminar shear stress induced GRO mRNA and protein expression in endothelial cells. Circulation 1998;98:2584–2590.

18. Friedman MH, Deters OJ, Bargeron CB, et al. Shear-dependent thickening of the human arterial intima. Atherosclerosis 1986;60:161–171.

19. Giddens DP, Zarins CK, Glagov S. The role of fluid mechanics in the localization and detection of atherosclerosis. J Biomech Eng 1993;115:588–594.

20. Irace C, Carallo C, Crescenzo A, et al. NIDDM is associated with lower wall shear stress of the com- mon carotid artery. Diabetes 1999;48:193–197.

21. Krams R, Wentzel J, Oomen J, et al. Evaluation of endothelial shear stress and 3D geometry as fac- tors determining the development of atherosclerosis and remodeling in human coronary arteries in vivo. Combining 3D reconstruction from angiography and IVUS (ANGUS) with computational fluid dynamics. Arterioscler Thromb Vasc Biol 1997;17:2061–2065.

22. Wentzel JJ, Whelan MD, van der Giessen WJ, et al. Coronary stent implantation changes 3-D vessel geometry and 3-D shear stress distribution. J Biomech 2000;33:1287–1295.

23. Wentzel JJ, Krams R, Schuurbiers JC, et al. Relationship between neointimal thickness and shear stress after Wallstent implantation in human coronary arteries. Circulation 2001;103:1740–1745.

24. Thury A, Wentzel JJ, Vinke RV, et al. Images in cardiovascular medicine. Focal in-stent restenosis near step-up: roles of low and oscillating shear stress? Circulation 11 2002;105:E185–E187.

25. Gijsen FJ, Oortman RM, Wentzel JJ, et al. Usefulness of shear stress pattern in predicting neointima distribution in sirolimus-eluting stents in coronary arteries. Am J Cardiol 2003;92:1325–1328.

26. Gijsen FJ, Allanic E, van de Vosse FN, et al. The influence of the non-Newtonian properties of blood on the flow in large arteries: unsteady flow in a 90 degrees curved tube. J Biomech 1999;32:705–713.

27. Zarins CK, Bomberger RA, Glagov S. Local effects of stenoses: increased flow velocity inhibits atherogenesis. Circulation 1981;64:II221–II227.

28. Cho A, Mitchell L, Koopmans D, et al. Effects of changes in blood flow rate on cell death and cell proliferation in carotid arteries of immature rabbits. Circ Res 1997;81:328–337.

29. Shimokawa H, Flavahan NA, Vanhoutte PM. Natural course of the impairment of endothelium- dependent relaxations after balloon endothelium removal in porcine coronary arteries. Possible dys- function of a pertussis toxin-sensitive G protein. Circ Res 1989;65:740–753.

30. Zarins CK, Zatina MA, Giddens DP, et al. Shear stress regulation of artery lumen diameter in exper- imental atherogenesis. J Vasc Surg 1987;5:413–420.

31. Glagov S, Weisenberg E, Zarins CK, et al. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 1987;316:1371–1375.

32. Wentzel JJ, Janssen E, Vos J, et al. Extension of increased atherosclerotic wall thickness into high shear stress regions is associated with loss of compensatory remodeling. Circulation 2003;108:17–23.

33. Slager CJ, Wentzel JJ, Schuurbiers JC, et al. True 3-dimensional reconstruction of coronary arteries in patients by fusion of angiography and IVUS (ANGUS) and its quantitative validation. Circulation 2000;102:511–516.

34. Stone PH, Coskun AU, Kinlay S, et al. Effect of endothelial shear stress on the progression of coro- nary artery disease, vascular remodeling, and in-stent restenosis in humans: in vivo 6-month follow- up study. Circulation 29 2003;108:438–444.

35. Kohler TR, Jawien A. Flow affects development of intimal hyperplasia after arterial injury in rats.

Arterioscler Thromb 1992;12:963–971.

36. Glagov S. Intimal hyperplasia, vascular modeling, and the restenosis problem. Circulation 1994;89:2888–2891.

37. Waller BF. Coronary luminal shape and the arc of disease-free wall: morphologic observations and clinical relevance. J Am Coll Cardiol 1985;6:1100–1101.

38. Mintz GS, Popma JJ, Pichard AD, et al. Arterial remodeling after coronary angioplasty: a serial intravascular ultrasound study. Circulation 1996;94:35–43.

39. Pasterkamp G, de Kleijn DP, Borst C. Arterial remodeling in atherosclerosis, restenosis and after alteration of blood flow: potential mechanisms and clinical implications. Cardiovasc Res 2000;45:843–852.

40. Uren NG, Crake T, Lefroy DC, et al. Delayed recovery of coronary resistive vessel function after coronary angioplasty. J Am Coll Cardiol 1993;21:612–621.

41. Ward MR, Tsao PS, Agrotis A, et al. Low blood flow after angioplasty augments mechanisms of restenosis: inward vessel remodeling, cell migration, and activity of genes regulating migration.

Arterioscler Thromb Vasc Biol 2001;21:208–213.

42. Stankovic G, Manginas A, Voudris V, et al. Prediction of restenosis after coronary angioplasty by use of a new index: TIMI frame count/minimal luminal diameter ratio. Circulation 2000;101:962–968.

43. Serruys PW, di Mario C, Piek J, et al. Prognostic value of intracoronary flow velocity and diameter stenosis in assessing the short- and long-term outcomes of coronary balloon angioplasty: the DEBATE Study (Doppler Endpoints Balloon Angioplasty Trial Europe). Circulation 18 1997;96:

3369–3377.

44. Majesky MW, Reidy MA, Bowen-Pope DF, et al. PDGF ligand and receptor gene expression during repair of arterial injury. J Cell Biol 1990;111:2149–2158.

45. Bassiouny HS, Song RH, Hong XF, et al. Flow regulation of 72-kD collagenase IV (MMP-2) after experimental arterial injury. Circulation 1998;98:157–163.

46. Wentzel JJ, Kloet J, Andhyiswara I, et al. Shear-stress and wall-stress regulation of vascular remod- eling after balloon angioplasty: effect of matrix metalloproteinase inhibition. Circulation 2001;

104:91–96.

47. Sterpetti AV, Cucina A, D’Angelo LS, et al. Shear stress modulates the proliferation rate, protein syn- thesis, and mitogenic activity of arterial smooth muscle cells. Surgery 1993;113:691–699.

48. Ueba H, Kawakami M, Yaginuma T. Shear stress as an inhibitor of vascular smooth muscle cell pro- liferation. Role of transforming growth factor-beta 1 and tissue- type plasminogen activator.

Arterioscler Thromb Vasc Biol 1997;17:1512–1516.

49. Thyberg J, Blomgren K, Hedin U, et al. Phenotypic modulation of smooth muscle cells during the formation of neointimal thickenings in the rat carotid artery after balloon injury: an electron- microscopic and stereological study. Cell Tissue Res 1995;281:421–433.

50. Jawien A, Bowen-Pope DF, Lindner V, et al. Platelet-derived growth factor promotes smooth muscle migration and intimal thickening in a rat model of balloon angioplasty. J Clin Invest 1992;89:

507–511.

51. Kinlay S, Grewal J, Manuelin D, et al. Coronary flow velocity and disturbed flow predict adverse clinical outcome after coronary angioplasty. Arterioscler Thromb Vasc Biol 2002;22:1334–1340.

52. Thury A, van Langenhove G, Carlier SG, et al. High shear stress after successful balloon angioplasty is associated with restenosis and target lesion revascularization. Am Heart J 2002;144:136–143.

53. Asakura T, Karino T. Flow patterns and spatial distribution of atherosclerotic lesions in human coro- nary arteries. Circ Res 1990;66:1045–1066.

54. Smedby O. Do plaques grow upstream or downstream?: an angiographic study in the femoral artery.

Arterioscler Thromb Vasc Biol 1997;17:912–918.

55. Seiler C. The human coronary collateral circulation. Heart 2003;89:1352–1357.

56. van Royen N, Piek JJ, Buschmann I, et al. Stimulation of arteriogenesis; a new concept for the treat- ment of arterial occlusive disease. Cardiovasc Res 2001;49:543–553.

57. Werner GS, Emig U, Mutschke O, et al. Regression of collateral function after recanalization of chronic total coronary occlusions: a serial assessment by intracoronary pressure and Doppler record- ings. Circulation 2003;108:2877–2882.

58. Watanabe N, Yonekura S, Williams AG, Jr., et al. Regression and recovery of well-developed coro- nary collateral function in canine hearts after aorta-coronary bypass. J Thorac Cardiovasc Surg 1989;97:286–296.

59. Werner GS, Bahrmann P, Mutschke O, et al. Determinants of target vessel failure in chronic total coronary occlusions after stent implantation. The influence of collateral function and coronary hemo- dynamics. J Am Coll Cardiol 2003;42:219–225.

60. Holmes DR Jr, Hirshfeld J Jr, Faxon D, et al. ACC Expert Consensus document on coronary artery stents. Document of the American College of Cardiology. J Am Coll Cardiol 1998;32:1471–1482.

61. Virmani R, Farb A. Pathology of in-stent restenosis. Curr Opin Lipidol 1999;10:499–506.

62. Schwartz RS, Huber KC, Murphy JG, et al. Restenosis and the proportional neointimal response to coronary artery injury: results in a porcine model. J Am Coll Cardiol 1992;19:267–274.

63. Hehrlein C, Zimmermann M, Metz J, et al. Influence of surface texture and charge on the biocompat- ibility of endovascular stents. Coron Artery Dis 1995;6:581–586.

64. Gyongyosi M, Yang P, Khorsand A, et al. Longitudinal straightening effect of stents is an additional predictor for major adverse cardiac events. Austrian Wiktor Stent Study Group and European Paragon Stent Investigators. J Am Coll Cardiol 2000;35:1580–1589.

65. Garasic JM, Edelman ER, Squire JC, et al. Stent and artery geometry determine intimal thickening independent of arterial injury. Circulation 22 2000;101:812–818.

66. Rogers C, Edelman ER. Endovascular stent design dictates experimental restenosis and thrombosis.

Circulation 15 1995;91:2995–3001.

67. Gunn J, Arnold N, Chan KH, et al. Coronary artery stretch versus deep injury in the development of in-stent neointima. Heart 2002;88:401–405.

68. Peacock J, Hankins S, Jones T, et al. Flow instabilities induced by coronary artery stents: assessment with an in vitro pulse duplicator. J Biomech 1995;28:17–26.

69. Tominaga R, Kambic HE, Emoto H, et al. Effects of design geometry of intravascular endoprostheses on stenosis rate in normal rabbits. Am Heart J 1992;123:21–28.

70. Berry JL, Santamarina A, Moore JE Jr, et al. Experimental and computational flow evaluation of coronary stents. Ann Biomed Eng 2000;28:386–398.

71. Robaina S, Jayachandran B, He Y, et al. Platelet adhesion to simulated stented surfaces. J Endovasc Ther 2003;10:978–986.

72. Kohler TR, Kirkman TR, Kraiss LW, et al. Increased blood flow inhibits neointimal hyperplasia in endothelialized vascular grafts. Circ Res 1991;69:1557–1565.

73. Salam T, Lumsden A, Suggs W, et al. Low shear stress promotes intimal hyperplasia thickening.

J vasc invest 1996;2:12–22.

74. Kraiss LW, Geary RL, Mattsson EJ, et al. Acute reductions in blood flow and shear stress induce platelet-derived growth factor-A expression in baboon prosthetic grafts. Circ Res 1996;79:45–53.

75. Mattsson EJ, Kohler TR, Vergel SM, et al. Increased blood flow induces regression of intimal hyper- plasia. Arterioscler Thromb Vasc Biol 1997;17:2245–2249.

76. Liu SQ. Prevention of focal intimal hyperplasia in rat vein grafts by using a tissue engineering approach. Atherosclerosis 1998;140:365–377.

77. Sprague EA, Luo J, Palmaz JC. Human aortic endothelial cell migration onto stent surfaces under static and flow conditions. J Vasc Interv Radiol 1997;8:83–92.

78. Van Belle E, Bauters C, Asahara T, et al. Endothelial regrowth after arterial injury: from vascular repair to therapeutics. Cardiovasc Res 1998;38:54–68.

79. Fontaine AB, Spigos DG, Eaton G, et al. Stent-induced intimal hyperplasia: are there fundamental differences between flexible and rigid stent designs? J Vasc Interv Radiol 1994;5:739–744.

80. Prati F, Di Mario C, Moussa I, et al. In-stent neointimal proliferation correlates with the amount of residual plaque burden outside the stent: an intravascular ultrasound study. Circulation 1999;99:

1011–1014.

81. Vorwerk D, Redha F, Neuerburg J, et al. Neointima formation following arterial placement of self- expanding stents of different radial force: experimental results. Cardiovasc Intervent Radiol 1994;17:27–32.

82. Richter GM, Palmaz JC, Noeldge G, et al. Relationship between blood flow, thrombus, and neoin- tima in stents. J Vasc Interv Radiol 1999;10:598–604.

83. de Smet BJ, Kuntz RE, van der Helm YJ, et al. Relationship between plaque mass and neointimal hyperplasia after stent placement in Yucatan micropigs. Radiology 1997;203:484–488.

84. Lemos PA, Serruys PW, Sousa JE. Drug-eluting stents: cost versus clinical benefit. Circulation 2003;107:3003–3007.