Vascular Surgery - part 2 pps

As the resistance R of a blood vessel segment is defined as the ratio of the pressure gradient across and the flow through the segment AP/Q, it is clear that resistance is in-versely pro

106 Part I Imaging Techniques

ventional spin echo, segmented gradient-echo sequences,

segmented turbo-FLASH, 3-D TOP, spiral acquisitions,

and echo planar imaging (13,24-26) MRA visualization

of proximal coronary vessels correlates well (>95%) with

that of conventional angiography (13,27) Visualization

of proximal coronary vessels is far superior to distal vessel

imaging and severe stenoses are more accurately identified

(27) MR imaging can detect a high proportion of severe

stenoses but only a low proportion of moderate stenoses

The sensitivity and specificity of coronary MRA for

detecting severe stenosis are 85% and 80% respectively

A moderate decrease in blood flow results in a significant

decrease of sensitivity to 38% (26)

The advantages of coronary arterial imaging with

MR have been mostly noted in the visualization of

anom-alous coronary vessels (28) Although conventional

an-giography can show anomalous vessels, the position of

the vessel relative to the aorta and adjacent organs can be

difficult to appreciate MRA can clearly demonstrate the

passage of the anomalous vessels anterior or posterior to

the aorta and their spatial relationship to nerves, venous

and other parenchymal structures, making it a useful

preoperative imaging tool (28,29)

Overall, coronary MRA for identification of

coro-nary stenoses is not generally accepted with the currently

existing technology Further refinement of imagingtechniques is necessary before coronary MRA willachieve widespread acceptance

Aortic Arch and Thoracic and Abdominal Aorta

MRA can delineate the aortic arch and its branches with ahigh degree of resolution (Fig 7.4) Aortic dissections can

be reliably diagnosed and classified as either type A(involving the ascending aorta) or type B (distal to the leftsubclavian artery) by MRA MRA accurately demon-strates the relationship of branch arteries to true and falselumen anatomy as well as defining the proximal and distalextents of the dissection flap (Fig 7.5)

Non-nephrotoxic contrast agents such as gadolinium(Gd) have enhanced the accuracy of imaging the aorticarch and aortic branch vessels (renal and visceral abdom-inal arteries) 3-D TOP MRA is used for evaluation prior

to thoracoabdominal as well as infrarenal aortic, renal,and visceral reconstructions The use of contrast enhancesthe resolution of the signals, improving detection ofbranch disease A prospective study of 63 patientswith suspected visceral aortic disease showed that usingbreath-hold ultrafast 3-D Gd-enhanced MRA techniques

FIGURE 7.4 Dissection and

occlu-sion of left common carotid arteryseen by arteriography (A) and MRA(B) Anomalous aortic arch (bovine)shown by contrast arteriography(C) and MRA (D) (Reproduced by per-mission from J vase Surg 1997; 25(1):147.)

Chapter 7 Magnetic Resonance Angiography 107

FICURE7.6 (A) MRA shows a left ternal iliac artery aneurysm (B) In-traoperativeangiogram confirmsthe finding which is then success-fully treated percutaneously by

in-an endovascular approach thatincluded coiling of the aneurysmand covering the inflow to theaneurysm using a commerciallyavailable stent graft (C) The sizing

of the stent graft was designedfrom the MRA images

combined with 2-D TOP, MR could accurately identify

and grade all (n = 51) renal, celiac, superior mesenteric,

and inferior mesenteric artery stenosis or occlusions The

combined MRA imaging techniques have 100%

sensitivi-ty and specificisensitivi-ty when compared with conventional

an-giography (30) MRA correctly predicts cross-clamping

site in 87%, proximal anastomotic site in 95%, need forrenal revascularization in 91 %, and the use of bifurcatedgraft in 75% of abdominal aortic aneurysm patients.MRA can also be successfully used as the sole imagingmodality for aortic or iliac endoprosthetic devices (Figs.7.6 and 7.7) In a prospective study of 96 consecutive pa-

108 Part I Imaging Techniques

FIGURE 7.7 (A-C) infrarenal abdominal aneurysm treated based on preoperative MRA (D) Intraoperative

angiograms confirming the MRA findings (E) Completion arteriogram after successful endografting

tients, data were collected using Gd-enhanced MRA

pre-operatively in place of conventional imaging for patients

with renal insufficiency or history of contrast allergy (31)

A total of 14 patients had their endograft designed solely

on Gd-enhanced MRA The frequency of intraoperative

access failure, the need for proximal or distal extensions,

the rate of conversion to open procedures, as well the

inci-dence of endoleaks were equal in both the MRA-designed

and control groups

Renal Artery stenosis

MRA has been advocated for evaluation of renal arteries

for the past decade Initial techniques were limited due to

motion artifact and limited spatial resolution Earlier

TOP MRA, when compared with conventional

angiogra-phy, had 91% sensitivity, with a 94% negative predictive

value Overall diagnostic accuracy of these techniques

was good(81%)(32); however, the detection of accessory

renal artery was poor (14) Images and diagnostic

accura-cy have improved greatly with the use of Gd-enhanced

MRA (Fig 7.8) Sensitivities of 50% to 70% have been

reported in the identification of accessory renal arteries

(33) Use of breath-hold ultra-fast 3-D Gd-enhanced

tech-niques has increased diagnostic yield of accessory renal

arteries to between 89% and 100% (34) This is primarilydue to increased spatial resolution and larger field ofview with these recent techniques Re formating the 3-Dvolume acquisition of the vascular anatomy can provideuseful preoperative information about aberrant arteries,degree of stenosis, aneurysms, and associated aortic dis-sections In contrast, conventional angiography relies onoblique imaging planes to delineate a profile of the steno-sis, making ostial lesions more difficult to be accuratelystudied, particularly in the setting where the total amount

of potentially nephrotoxic contrast volume is restricted.Contrast-enhanced MRA techniques are not associatedwith contrast nephropathy and can be used safely inpatients with renal insufficiency

Peripheral Circulation

Lack of filling distal to serial stenoses or occlusions andthe presence of bony cortex hinder the ability of conven-tional angiography to detect small and diseased distalrunoff vessels MRA avoids the complications of arterialpuncture, eliminates the risk of contrast-induced renalfailure, and has been shown to have a greater sensitivitythan contrast angiography for identifying distal runoffvessels in patients with severe peripheral arterial occlusive

Chapter? Magnetic Resonance Angiography 109

FIGURE 7.8 (A) MRA demonstrating right renal stenosis (B) cross-sectional view confirms the renal stenosis (C)

Arrows demonstrate celiac stenosis, SMA stenosis, and an aortic ulcer visualized by MRA (D) MRA demonstratingnormal aortoiliac arterial anatomy with normal visceral and renal branches (E) Superimposed venous, arterial andparenchymal imaging information acquired by MRl/MRA/MRV

disease (35) Recent refinements of magnetic resonance

angiography have replaced conventional angiography in

some centers

In studies of the aorta, iliac, and femoral inflow, MRA

is highly concordant with conventional contrast

angiog-raphy MRA has a sensitivity of 99.6%, a specificity of

100 %, a positive predictive value of 100%, and a negative

predictive value of 98.5% in detecting patent segments,

occluded segments, and hemodynamically significant

stenoses of aortic, pelvic, and proximal femoral inflow

vessels (4) The degree of arterial stenosis is measured with

high accuracy by MRA compared with conventional

angiography (36) Furthermore, MRA provides better

in-formation about spatial relationship of blood flow and

plaque morphology than conventional angiography (15)

This is mostly the result of sophisticated software

process-ing of MRI/MRA data, providprocess-ing enhanced views that

may include 3-D reconstructions in multiple longitudinal

projections and rotational views in addition to the 2-D

cross-sectional and axial views

MRA can be used as the sole preoperative imaging

modality for successful open vascular or endovascular

in-terventions (Figs 7.9 to 7.11) In one such study, outpatient

MRA of the juxtarenal aorta imaged 80 consecutive

pa-tients with ischemic rest pain or tissue loss through the foot

(4) Intraoperative pressure measurements of proximalvessels and post-bypass arteriography were performed.Graft patency and limb salvage was evaluated using lifetable analysis All patients underwent reconstructive pro-cedures based on MRA alone (11 aortobifemoral and 67infrainguinal procedures) The intraoperative findings andintraoperative completion arteriography confirmed the ac-curacy of inflow and outflow imaging by preoperativeMRA The limb salvage rate was 84 % with all -month pa-tency rate of 78% for infrainguinal reconstruction based

on MRA alone, and was no different from that of a controlgroup whose operations were planned with conventionalcontrast angiography (37)

MRA can detect angiographically occult distal runoffvessels In studies of lower extremity ischemia patients

in which MRA and conventional angiography were pared, the detection of distal runoff vessels was superiorwith MRA Operative exploration and intraoperative an-giograms confirmed the preoperative evaluation by MRA(4) A subsequent investigation of the adequacy of theseoccult runoff vessels for use in limb salvage bypass proce-dures showed no significant differences in primary graftpatency rate between bypasses planned using convention-

com-al angiography to those done to angiographiccom-ally occultrunoff vessels detected only by MRA (38)

110 Part I Imaging Techniques

FIGURE 7.9 (A) MRA showing normal femoral arterial segments (B) MRA demonstrating a short-segment stenosisand a more distal segmental occlusion of right superficial femoral artery (SFA) The left SFA shows mild diffuse dis-ease

FIGURE 7.10 The use of bolus chase techniques can facilitate rapid imaging of the distal runoff where the (A)popliteal, (B) inf rapopliteal, and (C) foot vessels are accurately visualized

MRA can enhance the clinical accuracy when

per-formed in addition to conventional angiography In a

blinded prospective study in six USA hospitals, MRA was

compared to contrast angiography to evaluate severe

lower limb atherosclerotic occlusive disease in candidates

for percutaneous or surgical intervention (39) Sensitivity

in distinguishing patent segments from occluded

seg-ments was 83% with contrast angiography and 85% in

MRA However, the inclusion of MRA preoperative

plan-ning resulted in a change of treatment plan for 13 % of

pa-tients and provided superior overall diagnostic accuracy

(86%) The improved accuracy related mostly to the

in-creased sensitivity of MRA in identifying patent runoff

vessels (48%) when compared with conventional

angio-graphy (24 %) (40) MRA is most useful in the detection ofpatent runoff vessels of the distal segments The detection

of patent runoff vessels by MR which are not identified

by conventional angiography can lead to improved limbsalvage in 13% to 22% of cases (39-40)

A meta-analysis of 34 studies indicated that MRA ishighly accurate for assessment of lower extremity arteries(41) Techniques using 3-D Gd-enhanced MRA appear to

be superior to 2-D methods and to contrast angiography.The superiority of MR techniques over traditional imag-ing techniques is due to characteristics of blood flow indiseased vessels and the sensitivity of MR for detection ofslow flow (2cm/s) Images from contrast angiographymay not show distal vessels owing to multiple dilutions

Chapter 7 Magnetic Resonance Angiography 111

FIGURE 7.11 (A) Distal runoff as

vi-sualized by conventional raphy demonstrating a diseasedposteriortibial artery (B) MRA re-veals that the anterior tibial andperoneal arteries are also patent.(C) intraoperativearteriogramafter bypass performed to an an-giographically occult dorsalispedis artery visualized preopera-tively by MRA, but not by preoper-ative contrast arteriography.(Reproduced by permission from JVase Surg 1996; 23:483-489.)

angiog-and reconstitution of the contrast material as the bolus 1

passes distally (Fig 7.11)

MRA can also be used as a sole preoperative imaging

modality prior to endovascular procedures (42) A

total of 119 consecutive patients underwent MRA for

symptomatic leg ischemia Intraoperative road-map

arteriography was performed in patients that underwent 2

endovascular procedures and compared to preoperative

MRA images There were no false positive or negative

stud-ies with MRA A reduction in cost was also noted owing to

the elimination of preoperative diagnostic arteriography

New Developments

Research in MR techniques continues to improve

success-ful clinical applications Bolus chase techniques involve

the movement of the scanner table in a stepwise manner to

allow sequential imaging of a bolus during arterial transit 4

(43) Using conventional angiograms as a reference

standard, manual bolus chase has been demonstrated to

have high sensitivity (93-94%) and specificity (97-98%)

MR incompatibility—risk for device displacement.Some recent endovascular devices that use stainlesssteel in covered stents for aortic aneurysm treatmentrepresent a contraindication for the use of MR imag-ing MR is also contraindicated in patients with pace-makers or retinal or intracranial metallic objects.Image degradation of horizontal vasculature Thickslices in coronal reconstructions of 2-D images (thatare obtained perpendicular to the long axis of thebody) result in a string of diamond appearance ofhorizontal vessels Thin slices and better imageresolution reduce these artifacts

Lengthy period of data acquisition: Improvements inreal-time MRA and bolus chase techniques decreasethe length of time required for peripheral MRAstudies

Existing MRA techniques have a number of related artifacts, due to signal loss or intravoxel de-phasing, resulting in overestimation of the degree andlength of arterial stenosis or signal dropout artifact.Pulsatile arterial flow can also result in ghostingartifacts in peripheral arterial evaluation Contrastagents reduce these effects

flow-112 Parti Imaging Techniques

Conclusion

The time-honored method of contrast angiography is

as-sociated with inherent risks and limitations

Develop-ments in noninvasive modalities offer potential benefits in

diagnostic accuracy and reduction of costs and morbidity

MRA represents an evolving technology that offers

promise as a noninvasive adjunct for vascular imaging

In-dividual centers must validate their MR data and

interpre-tation against conventional arteriography techniques

The preoperative workup and eventual therapeutic plan

can in many cases be successfully accomplished with the

sole or adjunctive use of MR imaging in the treatment of

vascular patients

References

1 Hessel SJ, Adams DF, Abrams HL Complications of

an-giography Radiology 1981; 138:273

2 Sjejado WJ, Toniolo G Adverse reactions to contrast

media: a report from the Committee on Safety of

Con-trast Media of the International Society of Radiology

Radiology 1980; 137:299

3 D'Elia JA, Gleason RE, Alday M Nephrotoxicity form

angiographic contrast material—a prospective study Am

JMedl976;72:719

4 Carpenter JP, Owen RS, et al Magnetic resonance

angio-graphy of the aorta, iliac, and femoral arteries Surgery

1994; 116(1): 17-23

5 Velazquez OC, Baum RA, Carpenter, JP Magnetic

reso-nance angiography of lower—extremity arterial disease

SurgClin North Am 1998; 78: 519-537

6 Yin D, Baum RA, et al The cost-effectiveness of magnetic

resonance angiography in symptomatic peripheral

vascu-lar disease Radiology 1995; 194: 757

7 Kent KC, Kuntz KM, et al Perioperative imaging

strate-gies for carotid endarterectomy: an analysis of morbidity

and cost-effectiveness in symptomatic patients JAMA

1995; 274: 888-893

8 Turnipseed WD, Kennell TW, et al Combined use of

duplex imaging and magnetic resonance angiography

for evaluation of patients with symptomatic ipsilateral

high-grade carotid stenosis J Vase Surg 1993; 17:

832-839; discussion 839-840

9 Polak JF, Kalina P, et al Carotid endarterectomy:

preop-erative evaluation of candidates with combined Doppler

sonography and MR angiography Radiology 1993; 186:

333-338

10 Schiebler ML, Listerud J, et al MR arteriography of

the pelvis and lower extremities Magnetic Resonance

Quarterly 1993; 9(3): 152

11 Keller P Time of flight magnetic resonance angiography

Neuroimaging Clin N Am 1992; 4: 639-656

12 Dumoulin CL Phase Contrast MR angiography

techniques MagnReson Imaging Clin N Am 1995; 3:

399-411

13 Edelman RR, Mattle HP, et al Extracranial carotid

arteries: evaluation with "black blood" MR

angiogra-phy Radiology 1990; 177:45-50

14 Velazquez OC, Baum RA, Carpenter JP: Magnetic

reso-nance imaging and angiography, Chapter 15 Rutherford

Vascular Surgery, 5th edn

15 Yucel EK, Anderson CM, et al Magnetic resonanceangiography: update on applications for extracranialarteries Circulation 1999; 100:2284-2301

16 Mitt RL Jr, Broderick M, et al Blinded-reader son of magnetic resonance angiography and duplex ul-trasonography for carotid artery bifurcation stenosis.Stroke 1994; 25(1): 4-10

compari-17 Pan XM, Saloner D, et al Assessment of carotid arterystenosis by ultrasonography, conventional angiography,and magnetic resonance angiography: correlation with

ex vivo measurement of plaque stenosis J Vase Surg1995; 21: 82-88

18 Kuntz KM, Skillamn JJ, et al Carotid endarterectomy

in asymptomatic patients: is contrast angiographynecessary? A morbidity analysis J Vase Surg 1995;22: 706-714

19 DeMarco JK, Nesbit GM, et al Prospective evaluation

of extracranial carotid stenosis: MR angiograph withmaximum-intensity projections and multiplanar refor-mation compared with conventional angiography AJR1994;163:1205-1212

20 Culebras A, Kase CS, et al Practice guidelines for the use

of imaging in transient ischemic attacks and acute stroke:

a report of the Stroke Council, American Heart tion Stroke 1997; 28:1480-1497

Associa-21 Dodge JT Jr, Brown BG, et al Lumen diameter of normalcoronary arteries: influence of age, sex, anatomic varia-tion, and left ventricular hypertrophy or dilation Circu-lation 1992; 86:232-246

22 Wang Y, Riederer SJ, Ehman RL Respiratory motion

of the heart: kinetics and the implications for the spatialresolution in coronary imaging Magn Reson Med 1995;33:713-719

23 McDonald IG The shape and movements of the humanleft ventricle during systole: a study by cineangiographyand by cineradiography of epicardila markers Am JCardiol 1970; 26:221-230

24 Meyer CH, Hu BS, et al Fast spiral coronary arteryimaging Magn Reson Med 1992; 28:202-213

25 Wang Y, Winchester PA, et al Contrast-enhanced pheral MR angiography form the abdominal aorta to thepedal arteries: combined dynamic two-dimensional andbolus-chase three-dimensional acquisitions InvestigRadiolo 2001; 36(3): 170-177

peri-26 Watanuki A, Yoshino H, et al Quantitative evaluation

of coronary stenosis by coronary magnetic resonanceangiography Heart Vessels 2000; 15(4): 159-166

27 Pennell DJ, Bogren HG, et al Assessment of coronaryartery stenosis by magnetic resonance imaging Heart1996; 75(2): 127-133

28 Post JC, Van Rossum AC, et al Magnetic resonance giography of anomalous coronary arteries: a new goldstandard for delineating the proximal course? Circula-tion 1995; 92: 3163-3171

an-29 Li D, Paschal CB, et al Coronary arteries: dimensional MR imaging with fat saturation andmagnetization transfer contrast Radiology 1993; 187:401-406

three-30 Siegelman ES, Gilfeather M, et al Breath-hold ultrafastthree-dimensional gadolinium-enhance MR angiography

of the renovascular system AJR 1997; 168:1035

31 Neschis DG, Velazquez OC, et al The role of magneticresonance angiography for endoprosthetic design J VaseSurg 2001; 33(3): 488-494

Chapter 7 Magnetic Resonance Angiography 113

32 Hertz SM, Baum RA, et al Magnetic resonance

angio-graphic imaging of angioplasty and atherectomy sites

J Cardiovasc Surg (Torino) 1994; 35(1): 1-6

3 3 Prince MR, Anzai Y, et al MRA contrast bolus timing

with ultrasound bubbles J Magnetic Reson Imag 1999;

10:389-394

34 Hertz SM, Holland GA, et al Evaluation of renal artery

stenosis by magnetic resonance angiography Am J Surg

1994; 168:140-143

35 Carpenter JP, Owen RS, et al Magnetic resonance

angio-graphy of peripheral runoff vessels J Vase Surg 1992;

16(6): 807-813 Comment in: J Vase Surg 1993; 17:

1136-1137

3 6 Owen RS, Carpenter JP, et al Magnetic resonance

imag-ing of angiographically occult runoff vessels in peripheral

arterial occlusive disease N Engl J Med 1992; 326:

1577-1581

37 Carpenter JP, Baum RA, et al Peripheral vascular surgery

with magnetic resonance angiography as the sole

preop-erative imaging modality J Vase Surg 1994; 20: 861-869

3 8 Carpenter JP, Golden MA, et al The fate of bypass grafts

to angiographically occult runoff vessels detected bymagnetic resonance angiography J Vase Surg 1996; 23:483-489

39 Baum RA, Rutter CM, et al: Multicenter trial to evaluatevascular magnetic resonance angiography of the lowerextremity JAMA 1995; 274: 875-880

40 Owen RS, Carpenter JP, et al Magnetic resonance ing of angiographically occult runoff vessels in peripheralarterial occlusive disease N Engl J Med 1992; 326:1577

imag-41 Koelemay, MJW, Lijmer JG, et al Magnetic resonanceangiography for the evaluation of lower extremity dis-ease: a meta-analysis JAMA 2001; 285:1338-1345

42 Levy MM, Baum RA, Carpenter JP Endovascularsurgery based solely on noninvasive preproceduralimaging J Vase Surg 1998; 28:995-1003

43 Prince MR, Yucel EK, et al Dynamic enhanced three-dimensional abdominal MR arteriogra-phy.JMagn Reson Imaging 1993; 3: 877-881

gadolinium-This page intentionally left blank

P A R T I I

Basic Cardiovascular Problems

This page intentionally left blank

C H A P T E R 8

Hemodynamics of vascular Disease: Applications to

Diagnosis and Treatment

David S Sumner

The surgeon faced with diagnosis and treatment of

vascu-lar disease must make decisions based on an assessment

of hemodynamic and rheologic factors Fluid dynamics is

exceedingly complex, even under optimally controlled

conditions; therefore, no practical formulas capable of

predicting outcomes have been devised It is possible,

however, to use some generally recognized principles to

formulate guidelines of value to the surgeon Although

many of these principles are intuitively evident, others are

less so and require some insight into the physical behavior

of fluids in motion Moreover, flow disturbances not only

affect the immediate supply of blood to the peripheral

tis-sues, but also directly interact with the wall of the conduit,

playing a role—now appreciated as quite important—in

the development of atherosclerotic plaques, platelet

depo-sition, and proliferation of fibromuscular tissues, all of

which may influence the outcome of any reconstructive

procedure

Normal Blood Flow

The fundamental principle governing blood flow is that

developed by Bernoulli:

P! + Vipf i + f>gh 1 = P 2 + ViP^f + PS^2 + ^ eat (8.1)

This equation simply states that the total fluid energy

(P + l /2pv 2 + pg^) must be greater upstream than

down-stream if blood is to move against a resistance, the energy

"lost" in the transition being dissipated in the form of

heat Pressure (?)—ordinarily the largest component oftotal fluid energy—may be segregated into dynamic pres-sure, derived largely from the contraction of the left

ventricle, and hydrostatic pressure (-pgh), which is

equivalent to the weight of a column of blood extendingfrom the point of measurement to the heart In this ex-pression, p is the density of blood (about 1.056 g/cm3); g is

the acceleration due to gravity (980cm/s2); and h is the

distance in centimeters above the heart Gravitational tential energy (+pg^) has the same dimensions as hydro-static pressure but has the opposite sign It represents theenergy imparted to blood by virtue of its elevation relative

po-to the surface of the earth Since, in most circumstances,gravitational potential energy is numerically equivalent tohydrostatic pressure, the two cancel out There are, how-ever; situations in which the two differ—especially on thevenous side of the circulation Finally, kinetic energy, theenergy imparted to blood by its motion, is proportional tothe product of its density and the square of its velocity

Viscous Energy "Losses"

Heat is generated by the interaction of contiguous cles of fluid in motion In a long, straight, rigid, cylindricaltube with perfectly steady laminar flow, viscosity ac-counts for all of the energy losses Poiseuille's law definesthe relation between the pressure (energy) gradient andflow under these strict conditions:

parti-117

118 Part II Basic Cardiovascular Problems

where T| represents the coefficient of viscosity measured in

poise and r the inside radius of the vessel This equation

states that, given a constant flow, the pressure gradient is

directly related to the length of the segment (L) and to the

viscosity of blood but is inversely related to the fourth

power of the radius The radius, therefore, has a profound

influence on energy losses

Of the many factors that determine the viscosity of

blood, hematocrit is the most important, the viscosity at a

hematocrit of 50% being roughly twice that at 35% (1)

Thus, in situations where laminar flow predominates, the

hematocrit may have a significant effect on pressure

gradi-ent or blood flow A further complicating feature is the

fact that the viscosity of blood, unlike that of water, varies

with shear rate (change in velocity between adjacent

lami-nae of blood, -dv/dr) (2) Viscosity increases markedly as

shear rates drop below 10/s; above this level, the viscosity

is essentially constant Although the mean shear rate (8/3

x v/r) in all blood vessels is well above this critical level, it

may fall below the critical value during those phases of the

pulse cycle in which the velocity decreases These

"non-Newtonian" characteristics of blood are probably not too

important, producing changes of only 1% or 2% in the

pressure gradient

When flow is laminar; the velocity profile across the

lumen of the vessel assumes a parabolic configuration

(Fig 8.1) At the wall, blood is essentially stationary;

maximal velocities are in the center of the tube; and the

mean velocity is exactly half the maximum In real life,

however; profiles approaching parabolic are found only

in the smaller or medium-sized blood vessels and then

only during peak systole Depending on the length, shape,

and curvature of the vessel and on the phase of the pulse

FIGURE 8.1 Velocity profiles Parabolic profiles occur

only during Ideal conditions Because of entrance

effects and flow disturbances, profiles are often

blunted (Reproduced by permission from Sumner DS

Hemodynamics and pathophysiology of arterial disease, in:

Rutherford RB, ed vascular surgery, 5th edn Philadelphia:

WBSaunders,2000.)

cycle, the profile may be blunted or severely skewed Sincethe adjacent particles of blood are flowing at nearly thesame velocity when the profile is blunt, there is little vis-cous interaction except near the wall; consequently,Poiseuille's law does not hold under these conditions

inertia I Energy "Losses"

Because velocity is a vector quantity, force is required toovercome inertia every time there is a change in the direc-tion of flow Directional changes occur in every curve, atevery bifurcation or branch point, and whenever thelumen of the vessel narrows or expands With each pulsecycle, blood accelerates during systole, decelerates andoften reverses during diastole, moves toward the wall asthe vessel expands, and moves toward the center of thelumen as the vessel contracts All motion that deviatesfrom the long axis of the vessel is inefficient in terms ofmoving blood toward its goal The energy thus "lost" tofriction is proportional to the product of the density ofblood and the square of the change in velocity:

In this chapter, these losses are called inertial losses

to distinguish them from those covered by Poiseuille'sequation

Resistance

As the relative contributions of viscosity and inertia varygreatly, it is impossible to characterize blood flow evenunder normal conditions with a simple formula; however,

a general equation incorporating the foregoing concepts is

as follows (3):

where &v represents a constant related to viscosity and k { ,

a constant related to inertial losses These constants varywith many factors, including the viscosity and density ofblood, the dimensions and configuration of the vessel, re-flection of pulses from the periphery, and heart rate, andare really unique to only a single situation In all cases, theenergy losses will exceed—often by a large amount—those predicted by Poiseuille's law

The equation of continuity states that in the absence

of intervening branches or tributaries, flow (Q) of an compressible fluid (such as blood and water) is constant

in-in all portions of a contin-inuous vessel Velocity, however,may differ from point to point, depending on the cross-

sectional area (A = nr 2 ):

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 119

It is interesting to note that the substitution of

equa-8.5 in equation 8.4 gives:

tion 8.5 in equation 8.4 gives:

after the constants have been appropriately modified As

the resistance (R) of a blood vessel segment is defined as

the ratio of the pressure gradient across and the flow

through the segment (AP/Q), it is clear that resistance is

in-versely proportional to the fourth power of the radius:

This formula also shows that resistance is not

con-stant but increases as flow increases (3) Therefore, unlike

an electrical wire, which has a rather constant resistance

over a wide range of currents, the resistance of a segment

of blood vessel can be defined only under precise

condi-tions of flow, pulse rate, and other factors

Nonetheless, resistance is a very useful concept in

thinking about blood flow Analogous to electrical

cir-cuits, the resistances of blood vessels in series are roughly

additive:

Arterial Stenoses

The presence of a stenotic lesion in an artery adds dously to the complexities of blood flow Approaching astenosis, the particles of blood— both microscopic and ul-tramicroscopic— must accelerate and change directions tosqueeze through an orifice narrower than that of the unin-volved vessel upstream (Fig 8.2) A pressure drop occurs

tremen-at this point as potential energy is transformed intokinetic energy Within the stenosis, the increase in velocity

is determined by the reduction in cross-sectional area Atthe exit, blood emerges at this same high velocity, forming

a jet, which disintegrates into disturbed or turbulent flow

as the mean velocity decreases to accommodate the largercross-sectional area Once again, an energy transforma-tion occurs— this time from kinetic back to potentialenergy The efficiency of these transformations deter-mines to a large extent the energy gradient across a steno-sis (Fig 8.3)

Inertial losses are greatest at the exit, where flow ismost disturbed (4,5) Expressed in terms of pressure gra-dient, these losses are proportional to the square of the dif-ference between the velocity of blood within the stenosis

(v s ) and that in the distal vessel (v d ):

and the reciprocals of those in parallel are likewise

additive:

where RTis the total resistance

Finally, although we can never say what the actual

re-sistance of a blood vessel or graft is without measuring

flow and pressure gradients under defined conditions, we

can calculate its minimal resistance using Poiseuille's law:

It must be emphasized that its actual resistance will

al-ways exceed this value

Reynolds Number

Fluids in motion behave similarly when they have

the same Reynolds number (Re), a dimensionless number

that depends on velocity, diameter (2r), and the ratio of

density to viscosity (p/T|):

The shape of the exit determines the severity of theflow disorganization, an abrupt orifice causing more dis-turbance than one that gradually expands (see Fig 8.2)

Laminar flow tends to break down into turbulence

when Reynolds numbers exceed 2000 Although this

breakdown normally occurs only during peak systole in

the aortic arch, flow may become unstable in other vessels

when stenoses are present—even with Reynolds numbers

in the hundreds Under these circumstances, inertial

energy losses are magnified

FIGURE 8.2 Flow patterns through axisymmetricalstenoses Disturbances of flow are greater when the

orifice is abrupt (upperpanel) than they are when the orifice is smooth and tapered (lowerpanel), velocities

and shear rates are low in areas of flow separationwhere, near the wall, the direction of flow may bereversed

120 Part n Basic Cardiovascular Problems

FIGURE 8.3 Relation between percentage of diameter

reduction and resistance of a 1 -cm-long "abrupt"

ax-isymmetrical stenosis in an artery with a diameter of

0.5 cm Total resistance increases rapidly, becoming

infinite at 100% stenosis (total occlusion) Resistance

due to inertial factors (kinetic fraction) exceeds that

due to viscosity when the diameter stenosis is

be-tween 25% and 85%, constituting over 70% of the total

resistance when the diameter stenosis is between

50% and 70% This iterative computer model is based

on equations 8.2,8.8, and 8.12

Reflecting the shape of the orifice, the constant, k, varies

from about 0.2 (gradual) to 1.0 (abrupt) (6) At the

en-trance, a similar relation exists, but flow disturbances and

inertial losses are less severe It is also true that

asymmetri-cal stenoses offer more resistance than axisymmetriasymmetri-cal

stenoses with the same reduction in cross-sectional area

(7) In part, this may account for the surprisingly high

re-sistance associated with iliac arteries, which do not

ap-pear to be significantly obstructed in the anteroposterior

arteriographic projection but are narrowed in the lateral

projection

Velocity profiles are quite blunt at the entrance to a

stenosis The distance (Le) required to regain a parabolic

profile is a function of the radius of the stenosis and the

Reynolds number (Le = 0.16rRe) Unless the stenosis is

quite long, a fully developed parabolic profile is never

es-tablished Although there is little viscous interaction

be-tween adjacent laminae in the blunt region of the velocity

profile, shear rates (dv/dr) near the wall are increased,

and viscous losses actually exceed those predicted by

Poiseuille's law

Arteriosclerosis is a diffuse process, and tandem

le-sions occurring in the same stretch of artery are not

un-common Because energy losses are greatest at the

entrance and exit, two separate lesions will offer more

re-sistance than a single lesion having a length equal to thecombined lengths of the two separate lesions—assuming,

of course, that the diameters are the same The resistances

of lesions in series are, however; not strictly additive (8,9)

In other words, the total resistance offered by two cal lesions would be less than double their individual re-sistances Although there are several reasons for this, thedecrease in peak systolic flow and velocity probably ac-counts for most of the disparity Since resistance is a func-tion of velocity, any reduction in velocity would result in adecreased resistance in each of the stenoses

identi-Pulsatile flow introduces other complexities (4,10) Ifflow reversal persists during a portion of the cardiac cycle,the entrance temporarily becomes the exit, and the exit,the entrance (Usually, however, flow reversal is not main-tained in the presence of significant arterial stenosis.) As innormal vessels, the periodic acceleration and decelerationaugment inertial losses Consequently, the resistance of alesion tends to increase with increasing pulse rate

To summarize, the energy-depleting effects of a sis are inversely proportional to the fourth power of itsradius (or the square of its cross-sectional area), are direct-

steno-ly proportional to the velocity and to the square of thechanges in velocity that occur at the entrance and exit, aremore dependent on inertial than viscous effects, areusually greatest at the entrance and exit, and are influenced

by the shape and symmetry of the stenotic orifices (5,11).Since resistance is a function of flow, and flow, in turn, is afunction of resistance, the resistance of a stenosis may varyconsiderably under different physiologic conditions

Effect on Pressure and Flow

Arterial stenoses must always be considered as part of alarger vascular circuit, consisting not only of the vesselsproximal and distal to the stenosis but also of any collat-eral vessels that bypass the stenotic region (12) To beginwith the most simple case, the resistances distal and prox-imal to the stenosis are considered to be constant, andcollaterals are considered to be absent Under these condi-tions, advancing stenosis causes a reduction in flow and anequivalent increase in the pressure gradient (Fig 8.4) (3).Changes in pressure and flow ordinarily become percep-tible only after the cross-sectional area has been reduced

by about 75%, which, in an axisymmetrically stenosedvessel, is equivalent to a 50% diameter reduction (13,14).Beyond this point, which is known as the point of "criticalstenosis," the stenosis is said to be "hemodynamicallysignificant." With decreasing peripheral resistance, thecurves are shifted to the left, and critical stenosis occurswith less diameter reduction Thus a lesion that does notcompromise blood flow in an artery feeding a high-resistance peripheral vascular bed may do so in an arterysupplying a low-resistance bed (15)

Gradual dilation of the peripheral arterioles is one oftwo mechanisms by which the body attempts to compen-sate for the increased resistance imposed by a stenosis(16-18) Until the arterioles become maximally dilated,

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 121

FIGURE 8.4 Effect of increasing diameter reduction on

flow through and pressure drop across an abrupt

ax-isymmetrical stenosis in a circuit with a fixed

periph-eral resistance Same model as in Figure 8.3

flow through the stenosis remains undiminished despite

its decreasing diameter The pressure gradient, however,

will increase more precipitously (19,20) After the ability

to dilate has been exhausted, further reductions in lumen

area will cause a rapid fall in both pressure and flow

(Fig 8.5)

The development of collaterals is the second major

compensatory mechanism Provided the collaterals are

large enough, the resistance of the vascular segment

con-taining the stenosis may remain unchanged, and

peripher-al pressure and flow will not be adversely affected Under

these circumstances, there will be no pressure drop across

the stenosis, but flow through the stenosis will be severely

curtailed Collaterals capable of such efficiency are the

ex-ception rather than the rule; in most clinical situations,

therefore, the segmental resistance is increased despite

ample time for the collaterals to mature (18,21) As a

re-sult, there is usually some decrease in pressure and some

drop in flow across the stenotic lesion, although one of the

two may be more affected than the other This is simply a

reflection of the fact that both pressure and flow are

mani-festations of total fluid energy

Estimating the resistance of a lesion by measuring

only pressure gradient or only flow—as some have done—

is likely to provide misleading information Both must

be measured Even then, the results pertain only to the

specific conditions existing at the time

FIGURE 8.5 Effect of compensatory peripheral larvasodilation on flow through and pressure dropacross a stenosis Segmental resistance refers to thecombined resistances of the stenosis and the parallelcollateral bed (Reproduced by permission from sumner

arterio-DS Correlation of lesion configuration with functional nificance, in: Bond Me, Insull W Jr., et al., eds Clinical diagno-sis of atherosclerosis: quantitative methods of evaluation.New York: Springer-Verlag, 1983.)

sig-(v 0 ) is determined solely by the relative radii of the

stenotic (r s ) and unobstructed segments (r 0 ):

Diameter stenosis (%) = (l - ^v 0 /v s ] x 100

The actual velocity of blood in the stenotic region,however, is determined not only by the relative radii butalso by the flow As a result, velocity increases with pro-gressive narrowing of the lumen until the stenosis be-comes quite severe and then drops off very rapidly as thelumen approaches total occlusion (Fig 8.6) (22,23).Because the Doppler flow detector can measure veloc-ity percutaneously, it has been used noninvasively to esti-mate the degree of stenosis It is evident that this approach

is strictly valid only if the mean velocities in the stenoticand unobstructed segments are compared Nonetheless,when a vascular bed (such as that containing the carotidartery) is well defined and peripheral autoregulationmaintains flow at normal levels, velocities above certainarbitrary values have proved to be useful in estimating thedegree of stenosis, albeit within broad limits

Effect on velocity

Unlike the pressure gradient and flow, which are functions

of resistance, the ratio of the mean velocity of flow

through a stenosis (v ) to that in the unobstructed vessel

Effect on Pulse wave Contours

A stenosis in an otherwise compliant vessel acts like a pass filter in an electrical circuit, attenuating the high-frequency harmonies of the flow or pressure wave (Fig

low-122 Part II Basic Cardiovascular Problems

8.7) (24) This tends to change the contour of the pulse

dis-tal to the stenosis, making it more rounded than that

above the stenosis The upslope becomes less steep, the

peak becomes more rounded, and the downslope bows

FIGURE 8.6 Effect of increasing diameter reduction on

velocity of flow through a stenosis Velocity increases

even though flow actually decreases until a critical

point is reached (Same model as in Figure 8.3.)

away from the baseline (25,26) Reversed flow nents are less evident and often disappear entirely (27).Fluctuations around the mean value are decreased, a factthat serves as the basis for the calculation of pulsatility in-dices (all of which, in one way or another, compare thetotal excursion of the pulse to its mean value) (28,29).Thus, decreases in the pulsatility index over an arterialsegment not only predict the presence of a stenosis butalso correlate with its severity (30,31) In contrast, reflec-tions originating from the stenosis may increase the excur-sion of the pulse wave above a lesion and thereforeincrease the pulsatility index (32-34) This finding mayalso have diagnostic value

compo-Effect on Shear Rate and Atherogenesis

Shear rate (D = -dv/dr) is the rate at which the velocity

of flow changes between concentric laminae of blood.Although the thin layer of blood in contact with theinner wall of a vessel is static, the adjacent layers are inmotion, creating a shear rate at the wall (Dw) and a corre-sponding shear stress (iw) on the endothelial surface Both

are directly proportional to the mean velocity of flow (v) and inversely proportional to the inner radius (r) of the

vessel:

Dw= 4 - tw= 4 n

FIGURES.? Effect of stenosis in a

compliant artery on the contour ofpressure and flow pulses Faucetrepresents the variable resistance

of the peripheral vascular bed

Mean pressure (dashed line) is duced, but mean flow (dashed line)

re-is unchanged (Reproduced by mission from Sumner DS Correlation oflesion configuration with functionalsignificance, in Bond Me, Insull W Jr., etal., eds Clinical diagnosis of atheroscle-rosis: quantitative methods of evalua-tion New York: Springer-verlag, 1983.)

per-Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 123

Thus, at any instant in the pulse cycle, shear rate and

shear stress increase as the mean velocity increases or the

radius decreases, and they decrease as the velocity

de-creases or the radius inde-creases

As the jet of blood emerges from the exit of the

stenosis, it diverges, coming in contact with the wall

downstream (see Fig 8.2) This creates an area of

flow separation, extending from the end of the lesion

to the point of reattachment Within the region of

separation, flow is very sluggish and may even be reversed

Shear rates are therefore correspondingly low and

may also be reversed During the cardiac cycle, shear

rates can alternate between forward and reversed

orienta-tions (10) The longitudinal extent of the zone of flow

separation varies with Reynolds number and the shape

of the orifice When Reynolds number is low and the

orifice angle is gradual, there may be little or no flow

separation (7)

The physiologic and pathophysiologic importance of

shear rate and shear stress is now well established Low

shear rates permit the accumulation of platelets and other

substances that interact with the vascular wall to foster

the development of atherosclerotic plaques, intimal

thick-ening, and fibromuscular hyperplasia (35,36) This

ex-plains the preferential location of plaques in the carotid

bulb opposite the flow divider and the frequency with

which atherosclerotic plaques form at the bifurcations of

the terminal aorta, the common femoral artery, and

popliteal artery—all areas in which geometry promotes

flow separation and decreased shear rates (37,38) Once a

plaque has formed, further extension may be promoted by

the area of stagnant or reversed flow that develops

imme-diately beyond the stenosis Distal to a stenosis, altered

shear stresses (39) and vibrations (40) generated in the

ar-terial wall by disturbed or turbulent flow may be

respon-sible for poststenotic dilation

Within the stenosis, shear rates may be quite high

and may exceed values demonstrated to cause endothelial

injury, but there is little evidence that this is conducive

to atherogenesis (41,42) In fact, the endothelium seems

to sense the increased shear and transmits this

infor-mation to the muscular elements of the arterial wall;

dilation occurs, and shear rates return toward prestenotic

levels This has the effect of ameliorating the severity of

the stenosis and may be responsible for some of the

re-ported arteriographic observations suggesting plaque

resolution (43,44) Other investigators, however, have

observed a positive correlation between shear rate and

platelet and fibrin deposition on damaged endothelial

surfaces and suggest that increased shear rates may

be conducive to arterial thrombosis under certain

circumstances (45)

Thus stenoses not only affect pressure and peripheral

perfusion but may also have local effects that are equally

important Research in this area promises to enhance the

understanding of atherogenesis and should provide

infor-mation of practical value to the surgeon involved in the

management of this disease

Stenosis as Part of a Larger Arterial Circuit

As mentioned previously, the stenotic artery and its erals may be considered as a unit, an arterial segment, inother words, with its own relatively "fixed" resistance (Fig.8.8) This segment is in series with a peripheral vascularbed, the resistance of which varies extensively in response

collat-to stress and other stimuli Included in this peripheral bedare the arteries distal to the most distal collateral inflow site,the arterioles, capillaries, venules, and veins Because oftheir small diameters, their muscular walls, and their copi-ous innervation, most of the peripheral resistance is con-centrated in the arterioles It is the arterioles, therefore, thatlargely control changes in peripheral resistance

Although the actual hemodynamic features of such

a complex circuit cannot be depicted by simple formulas,simple formulas analogous to Ohm's law facilitate ourunderstanding of the physiology (3,12) Blood flow(QT) through the peripheral vascular bed is determinednot only by the pressure gradient existing between thecentral arteries (PJ and the central veins (Pv) but also bythe total resistance of the circuit, which is the sum of thesegmental resistance (-Rseg) and the peripheral resistance

causes arteriolar dilation and a reduction in R , flow is

markedly increased—often by as much as five to ten timesbaseline levels (Fig 8.9) (20,46,47) In the presence of a

proximal arterial obstruction, R is almost always

in-creased, despite the development of collaterals As long asthe autoregulatory capacity of the peripheral arterioles

has not been exceeded, R decreases enough to

compen-sate for the increased proximal resistance, total resistance

is unchanged, and peripheral blood flow is maintained atnormal levels (see Fig 8.5) During exercise, however, fur-

ther reduction in R is limited; consequently, the fall in

total resistance is not sufficient to augment flow to thelevels required to sustain the increased demands ofthe muscles, and claudication is experienced (Fig 8.9)

(20,46-49) In the worst situation, R is so high that

ar-teriolar dilation is unable to reduce the total resistance tonormal levels, even at rest When this situation occurs, pe-ripheral perfusion fails to sustain normal metabolic activ-ities, and rest pain or gangrene may ensue (50-52).The pressure gradient across a stenotic segment isdetermined by its resistance and the magnitude of theflow:

P a -P c j=Q.R s e or P d = P a - Q K S e g (8.16)

where Pd is the arterial pressure distal to the stenosis but

proximal to the peripheral bed Normally, R is so low

124 Part II Basic Cardiovascular Problems

FIGURE 8.8 (Upperpanel)

compo-nents of a vascular circuit ing an arterial stenosis or occlusion

contain-RIGHT (Lowerpanel) An electrical

ana-ATRIUM logue, in which the battery

repre-sents the left ventricle and theground potential represents theright atrium (Reproduced by permis-sion from Sumner DS Hemodynamics ofabnormal blood flow, in: Wilson SE, Veith

FJ, et al eds vascular surgery, principlesand practice New York: McGraw-Hill,

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 125

that the gradient is only a few mmHg [Actually, because

of reflected waves, the systolic pressure in the distal artery

may exceed that in the proximal artery but the mean

pres-sure will always be somewhat less (53-55).] Even though

flow is increased many-fold with exercise, the product,

QxR remains low in normal limbs, and the peripheral

pressure drop is insignificant (see Fig 8.9) If there is any

concomitant rise in the arterial perfusion pressure, the

dis-tal pressure may even increase somewhat

Because compensatory peripheral arteriolar dilation

maintains resting blood flow at normal levels, any

in-crease in segmental resistance causes a similar inin-crease in

the pressure gradient across the segment and, provided

that the central pressure remains constant, a decrease in

peripheral arterial pressure (see Fig 8.9) Exercise, by

augmenting blood flow, causes the peripheral pressure to

drop even further, not infrequently to the point where it

can no longer be measured (20,54,56-58) Following the

cessation of exercise, blood flow decreases as the

metabol-ic debt incurred by the exercising muscles is repaid In

nor-mal limbs this debt is mininor-mal, and flow rapidly falls to

pre-exercise levels, but in diseased limbs—especially those

with the most severely compromised circulation—many

minutes may be required before the debt is repaid and flow

returns to baseline (19,54,56-60) As long as flow is

in-creased, the peripheral pressure remains dein-creased, rising

gradually in the postexercise period to pre-exercise levels

as flow returns to normal resting values

The situation becomes more complex when there are

multiple levels of obstruction (3,60,61) In such cases, the

physiologic effects are not simply due to the sum of the

segmental resistances but involve steal phenomena as

well Since the proximal arterial segment supplies not only

the vascular bed fed by the distal segment but also a more

proximal bed, exercise will cause some of the blood

des-tined for the distal tissues to be diverted into the more

proximal bed For example, consider a series of

obstruc-tions involving the aortoiliac and superficial femoral

seg-ments The arteries comprising the aortoiliac segment

feed the tissues of the buttocks, thighs, and calf, while the

superficial femoral segment mainly supplies the calf and

foot During exercise, the arterioles in all these muscles

are dilated, blood flow through the iliac segment is greatly

increased, and the pressure in the common femoral

artery falls Since the common femoral artery supplies

the superficial femoral segment, the expected increase in

flow through this segment will not develop despite a

profound reduction in the resistance of the arterioles in

the calf In fact, flow may actually fall below resting values

in the more peripheral tissues, such as those of the

foot (47,57,62) After exercise, the flow debt to the

but-tock and thigh muscles is the first to be repaid As flow

through the aortoiliac segment subsides, the common

femoral pressure rises, and flow through the superficial

femoral segment increases, allowing repayment of the

metabolic debt incurred by the calf muscles During the

postexercise period, the pressure in the distal arteries

remains severely depressed until the flow through the

su-perficial femoral segment reaches its peak and begins tofall (57,60,63)

Because blood flow is difficult to measure sively and because there is a wide range of normal restingvalues and an even wider range of normal exercise values,physiologic assessment, in clinical practice, is usually lim-ited to the measurement of peripheral pressures (57) Un-like flow, normal values for pressure can be assumed to beclose to the central arterial pressure Moreover, pressure,which represents potential energy, reflects more accurate-

noninva-ly than flow the capacity of the circulation to accomplishits work

Collaterals and Segmental Resistance

As mentioned earlier, collateral development is rarely ficient to maintain normal segmental resistance when themajor artery of the segment is severely stenosed or occlud-

suf-ed (18,21) Since collateral resistance parallels that of thediseased artery and since the resistance of each collateral isinversely proportional to the fourth power of its radius, itwould take 16 collaterals with a diameter of 0.25cm or

625 collaterals with a diameter of 0.1 cm to have a tance as low as that of an unobstructed vessel with adiameter of 0.5 cm The former would have a total cross-sectional area of 3.1cm2, and the latter, a total cross-sectional area of 19.6cm2—4 and 25 times, respectively,that of the unobstructed vessel (0.8cm2) Clearly, a fewlarge collaterals are likely to be far more efficient than alarge number of small collaterals

resis-Collaterals, basically, are arteries whose primaryfunction is to supply nutrients to the tissues throughwhich they pass When recruited to serve as conduitsaround an arterial obstruction, they dilate in response tothe increased shear stress imposed by the augmentedblood flow but retain their primary function (64,65).Thus their effective resistance must exceed that suggested

by their lengths and diameters since only a portion ofthe blood they carry reenters the major arterial system(12,59) Moreover, during exercise, their effective resis-tance may rise as more blood is siphoned off to supply themuscular tissues through which they pass

Thus, it may be very difficult to evaluate the capacity

of the collateral channels visible on an arteriogram.Segmental resistance, like that of the lesion itself, is bestevaluated by physiologic tests (66)

Bypass Grafts

Because increased segmental resistance is responsible forall the physiologic effects of arterial occlusive disease, themost direct treatment involves reduction of this resis-tance If the lesion is well defined and short enough, re-duction can be accomplished by endarterectomy or byendovascular dilation and stenting, but in the majority ofcases, insertion of a bypass graft is the best approach Inessence, the bypass graft serves as another collateral chan-

126 Part n Basic Cardiovascular Problems

nel, acting in parallel with the diseased arteries and the

ex-isting collateral system The resistance of the graft is

deter-mined not only by its length and diameter but also by the

configuration of the proximal and distal anastomoses

Resistance of the Graft

Poiseuille's law can be used to calculate the minimal

resis-tance of a prosthetic graft This calculation, ofcourse,

neglects energy losses due to inertia, which occur at the

en-trance and exit and at each curve These losses can be quite

significant (67,68) Moreover, pulsatile flow also

increas-es the lossincreas-es over those expected for steady laminar flow

As shown in Table 8.1, a 20-cm-long aortofemoral graft

with a diameter of 7mm should be capable of sustaining

flows of 3000mL/mm with a minimal pressure drop; but

a 5-mm graft would offer an appreciable resistance,

even discounting inertial factors Similarly, 40-cm-long

femoropopliteal grafts with diameters of 4 mm or greater

should function satisfactorily when called on to transmit

flows of up to 500mL/mm, but grafts with diameters less

than 4mm would offer an unacceptably high resistance

Long grafts (80cm) from the femoral to tibial arteries are

ordinarily used for the treatment of ischemic symptoms;

resting flow rates are not high, and pressure drops of

lOmmHg may be acceptable Still, long segments of such

grafts with either distal or proximal diameters less than

3 mm are inefficient blood conduits

After implantation, prosthetic grafts develop a

pseudointima that further reduces the effective internal

diameter Although a 0.5-mm layer, applied

circumferen-tially, would have little influence on the pressure gradient

across a large graft, it might adversely affect the function

of a graft with borderline dimensions Since high ties are conducive to the formation of a thin, tightly ad-herent pseudointima, graft diameters should be no largerthan necessary to ensure satisfactory flow dynamics If thediameter of the graft is too large, clots tend to form on theinner walls as the flow stream attempts to mold itself to thediameter of the recipient vessel These clots are loosely at-tached and may form an embolus, causing graft failure Asindicated by equation 8.5, given the same mean flow rate,the velocity in a 7-mm graft would be double that in a 10-

veloci-mm graft Because there is little difference in the

function-al capacity of these two grafts in the iliac region, thesmaller diameter is preferred

Saphenous veins used for femoropopliteal andfemorotibial bypasses contain valves that reduce the cross-sectional area by about 60% (69,70) Although the length

of the obstruction so created is quite short, the intact valvesare capable of causing additional inertial losses Studieshave shown that resistance to flow, even in the reversedsaphenous vein, is decreased by valve bisection (71,72).Autogenous vein grafts are subject to narrowingcaused by intimal hyperplasia, the development of whichhas been shown to be associated with low shear rates(35,73) Low shear rates cause smooth muscle cells to be-come secretory and enhance platelet adherence (73) Highshear rates, on the other hand, foster continued patencyand lessen the tendency for the intima to become hyper-plastic The protective effect of high shear has been attrib-uted to suppression of the release of endothelin-1, apeptide found in endothelial cells that acts as a vasocon-strictor and a mitogen for smooth muscle cells (74)

TABLE 8.1 Pressure gradients across grafts (mmHg)

5000.2(0.2)0.7(0.91.4(1.7)2.9(3.6)

15000.5(0.9)0.2(3.9)4.1(7.2)8.6(15.0)

30001.1(2.7)4.5(11.1)8.3(20.6)17.1(42.8)

Femoropopliteal length=40 cm

500.3(0.3)0.6(0.6)1.4(1.4)4.4(4.5)

Femorotibial length -80 cm'

1500.8(0.9)1.7(1.8)4.2(4.3)13.2(13.7)

3001.7(1.8)3.4(3.78.4(9.0)26.4(28.4)

Values are viscous only, equation 8.2; or viscous + kinetic (in parentheses), equation 8.6; T) = 0.035 poise; p = 1.056 g/cm 3

* Evenly tapered grafts, largest diameter to smallest.

5002.8(3.1)5.7(6.4)13.9(15.7)44.0(49.5)

6-4

5-3

4-2

501.3(1.3)3.5(3.5)13.0(13.1)

1002.6(2.6)6.9(7.0)26.0(26.3)

1503.9(4.0)10.4(10.5)39.0(39.7)

2005.2(5.4)13.8(14.2)52.0(53.3)

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 127Distribution of Flow in Parallel Graft and

Stenotic Artery

Surgeons occasionally express concern over the

possibili-ty that continued patency of a stenotic artery might lead to

thrombosis of a parallel graft To allay this fear; they

either avoid end-to-side anastomoses or ligate the stenotic

artery Theoretical considerations strongly suggest that

such concerns are not valid, provided that the arterial

seg-ment is sufficiently diseased to merit bypass grafting As

shown in Figure 8.10, even when the preoperative

pres-sure gradient across a stenosed artery is only lOmmHg,

over 90% of the flow will be diverted into the graft The

choice of an end-to-side anastomosis should, therefore, be

based on other considerations

vein Grafts with Double Lumens

Not uncommonly, saphenous veins bifurcate into two

separate and parallel channels that rejoin after a variable

distance to reconstitute a single lumen When this

situa-tion is encountered, the surgeon must decide whether or

not to include both channels in the graft

Since both of the duplicated channels will have a

lumen diameter less than that of the "parent" vein, it is

clear that each will offer more resistance than an equal

length of undivided vein If the channels are of the same

size, their combined resistance will be greater than that of

an equal length of undivided vein (unless their individual

diameters exceed 84% of the diameter of the undivided

vein) Thus, in most cases, the combined resistance of the

two parallel channels exceeds that of the undivided vein

Obviously, the adverse hemodynamic effects are

propor-tional to the relative lengths of the divided and undivided

parts, in other words, at a given flow rate, the pressure dient across a bifurcated graft increases as the length ofthe divided segment increases

gra-As shown in Figure 8.11, at the same flow rate to thethigh and calf muscles, the distal (popliteal) pressure andthe flow rate through a bifurcated femoropopliteal graft arehigher when both channels are preserved than they arewhen one channel has been ligated Although the differ-ences are small at rest, they become appreciable during ex-ercise Both configurations, however, represent a markedimprovement over the nonbypassed situation The argu-ment that preserving both channels jeopardizes the sur-vival of the graft by decreasing flow velocity through thebifurcated segment is not valid Even when both channelsare functional, the velocity in each exceeds that in the undi-vided part of the vein From this analysis, one must con-clude that preservation of two equal-sized channels isdesirable but certainly not mandatory On the other hand,

if one of the channels is distinctly larger than the other, there

is little to be gained by preserving the smaller of the two

Limiting the reconstruction to a femoral-(blind)popliteal bypass usually secures only a modest increase in

FIGURE 8.10 Relative flow through

bypass graft and stenotic artery Asthe preoperative pressure dropacross the artery increases (indi-cating increasingly severe steno-sis), the percentage of flowdiverted to the graft increases.Lumen of the graft is equal tothat of the unobstructed artery.(Reproduced by permission fromStrandness DE Jr., Sumner DS Hemo-dvnamics for Surgeons New York:Crune&Stratton,l975.)

128 Part II B asic Cardiovascular Problems

FIGURE 8.11 Resting and exercise

flow and flow velocity through a

40-cm femoropopliteal bypass graftwith a 20-cm divided segment Thediameter of the undivided graft is

5 mm and that of each of the

divid-ed segments is 3 mm Arteries are asfollows: common femoral (CF),superficial femoral (SF), profundafemoris (PF), popliteal (P), and thighcollateral (TO Arrows indicatedirection of flow Thigh and calfresistances are autoregulated tomaintain resting flows of 200 andiOOmL/mm respectively, computermodel is based on equations 8.2,8.5,8.8, and 8.9

ankle and calf perfusion pressure (Fig 8.12) If

below-knee resistances are quite high, the patient may derive

little benefit from this procedure Although both

femorotibial and femoropopliteal-tibial grafts yield

sig-nificant and virtually equivalent increases in ankle

pres-sure and are capable of relieving foot ischemia, the latter

has the advantage of providing a greater increase in

popliteal and tibial pressure Thus sequential grafts are

better equipped to cope with the demands of calf muscle

exercise (Fig 8.13)

Flow rates in femorotibial grafts should theoretically

be lower than those in the proximal segment of sequential

grafts but higher than those in the distal segment (see Figs

8.12 and 8.13) (76-78) The proximal segment of a

se-quential graft contributes blood not only to the calf but

also, in a retrograde fashion, to the thigh Having no

di-rect communication with the popliteal artery,

femorotib-ial grafts supply more blood in a retrograde direction to

the proximal tissues of the calf than distal segments of

se-quential grafts do Because flow velocities ate a function of

flow rates, distal segments of sequential grafts may be

more susceptible than proximal segments to failure (79).

On the other hand, a femorotibial graft may be more

like-ly to fail than the proximal segment of a sequential graft

Outflow Resistance

Failure of infrainguinal bypass grafts has been correlatedwith high outflow resistance (80-82) Since outflow resis-

tance, which is roughly analogous to R in equation 8.15,

is in series with graft resistance, blood flow throughthe graft is inversely proportional to the sum of the tworesistances

Although various methods for estimating outflow sistance have been described, all measure the pressuregenerated in the distal graft while saline is being infusedinto the graft at a known rate Outflow resistance is simplythe ratio of the pressure and the flow rate of saline Mea-sured in this way, outflow resistance reflects both the

re-"true" resistance of the peripheral vascular bed and theresistance of the collateral arteries At low infusion rates,the pressure developed in the graft does not exceed that at

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 129

FIGURE 8.12 Resting flow through

a 40-cm femoral-(blind) poplitealbypass graft, a 60-cm femorotibialgraft, and a 40-cm proximal, 20-cmdistal, sequential femoropopliteal-tibial graft Diameter of the graft is

5 mm throughout Symbols not cluded in Figure 8.11 are calf collat-eral (CO and tibial arteries (7).Resting flows to the thigh muscle,calf muscle, and distal leg andfoot are 200,70, and 30 mL/mm,respectively

in-the proximal end of in-the collaterals; consequently,

collater-al flow competes with flow from the graft to supply the

pe-ripheral vascular bed On the other hand, at high infusion

rates, the pressure developed in the graft is sufficiently

high to reverse flow in the collaterals, which then become

a part of the outflow system of the graft (see Fig 8.11) It

turns out, therefore, that the apparent outflow resistance

varies with the rate at which saline is being infused, being

deceptively high at low rates of infusion and deceptively

low at high infusion rates (Table 8.2) (83) Thus, to

accu-rately reflect outflow resistance, measurement should be

made at pressures similar to those expected when the graft

is functioning

Although clamping the recipient artery proximal

to the distal anastomosis decreases the size of thecollateral bed and makes the measurements more reflec-tive of the "true" peripheral resistance, it will notaffect those collaterals that enter below the anastomosis.Nevertheless, this maneuver does appear to improve theability of outflow resistance to identify those grafts des-tined to fail (81) The fact that saline, which has a viscosi-

ty much less than that of blood, is used as the infusateintroduces another confounding variable One wouldexpect the resistance measured with saline to be con-siderably less than that actually existing when the graft isfunctioning

130 Part n Basic Cardiovascular Problems

FIGURE 8.13 Exercise flow throughfemoral-(blind) popliteal, femoro-tibial, and sequential femo-ropopliteal-tibial grafts Exerciseflows to the thigh muscle, calfmuscle, and distal leg and foot are400,140, and 30 ml/mm,

respectively

Crossover Crafts

Femoral-femoral, axillary-axillary,

subclavian-subclavian, axillary-femoral, and other similar grafts all

depend for their proper function on the ability of the donor

artery to supply an increased blood flow without

sustain-ing an appreciably increased pressure drop Since the drop

in pressure across any arterial segment is a function of the

product of its resistance and the flow rate (equation 8.16),

the resistance of the donor artery must be relatively low

When the donor artery is disease free, there ordinarily is no

problem; but when the donor artery contains

atheroscle-rotic plaques (as many do), a steal phenomenon maydevelop (Table 8.3) (84,85) Questions regarding theresistance of the donor artery are best resolved byhemodynamic measurements Arteriography may be de-ceiving For example, before performing a femoral-femoral bypass, the surgeon who is concerned about thecapacity of the donor vessel should measure the commonfemoral artery pressure on the donor side with the flowrate at least double the resting value This is most easily ac-complished pharmacologically by the administration ofpapaverine If the operation is being performed to relieveclaudication, there should be relatively little pressure

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 131 TABLE 8.2 Relation of apparent outflow resistance to "true" peripheral resistance

Flow Rates (mL/min)

Collateral*

+84.5+68.9+53.4+37.9+6.8-24.2-86.3-148.4

Input Pressure(mmHg)65.771.477.182.794.1105.5128.2150.9

Apparent Outflow Resistance( mmHg/mL/min )2.631.431.030.830.630.530.430.38

Apparent/TrueResistance Ratio4.382.381.711.381.050.880.710.63

Based on diagram in Figure 8.11, assuming constant resistances (mmHg/mL/min): true peripheral = 0.6, collateral = 0.35; thigh muscle = 0.475; dafemoris = 0.017.

profun-indicates antegrade; - profun-indicates retrograde collateral flow

TABLE 8.3 Theoretic effect of femoral-femoral graft (data from reference 84)

No Stenosis of Donor Iliac Stenotic Donor Iliac

Before Graft After Graft Before Graft After Graft Before Graft After Graft Before Graft After Graft

Donor

Iliac flow (mL/min) 250 476 1266 2282 250 311 645 730 Common femoral pressure 99 98 95 91 80 75* 48 42* (mm Hg)

Common femoral flow 250 248 1266 1211 250 235* 645 554* (mL/min)

Recipient

Iliac collateral flow (mL/min)

Common femoral pressure

—

18 97 246 228

426 32 426

—

84 87

1155 1071

250 60 250

—

157 75 233 76

426 32 426

—

369 41 545 176

Aortic pressure =100 mmHg, graft resistance = 0.004 mmHg/mL/min.

* Pressure and flow drops indicative of a "steal"

drop, but if the purpose is to alleviate ischemia, a

some-what larger pressure drop may be permissible In other

words, the pressure delivered to the recipient common

femoral artery should be high enough to ensure adequate

perfusion of the target tissues One must also consider the

effect of the reduced pressure on the donor limb In most

cases this will be minimal, but when stenoses or occlusions

of the thigh or calf arteries are present, the fall in pressure

may be sufficient to induce symptoms in a previously

asymptomatic limb or worsen those in a previously

symp-tomatic limb

Anastomotic Configuration

To reduce energy losses due to flow disturbances, the

tran-sition from graft to host vessel should be as smooth as

pos-sible (86,87) End-to-end anastomoses, therefore, most

closely approximate the ideal End-to-side or side-to-end

anastomoses always result in alterations in flow direction

(Fig 8.14) Tailoring the anastomosis to enter the ent artery or leave the donor artery at an acute angle willminimize but can never eliminate flow disturbances Al-though decreasing the angle will reduce flow disturbances

recipi-in the antegrade limb of a recipient artery, it will ate those in the retrograde limb, where flow vectors are al-most completely reversed (88) Other energy-depletingpitfalls to be avoided include marked disparity betweenthe diameters of the graft and the artery to which it is con-nected, and slit-like configurations of the orifice betweenthe two conduits (89) The latter occurs when the graftlumen is stretched to accommodate an excessively long in-cision in the artery

accentu-Despite these theoretical considerations, in practicethere is usually little difference in the pressure gradientsacross anastomoses, regardless of their angle or configu-ration (provided, of course, that the anastomoses havebeen carefully constructed and that there are no stenoses)(90) There may, however, be important differences that

132 Part n Basic Cardiovascular Problems

determine the longevity of graft function (91,92)

When-ever there are flow disturbances, regions of flow

separa-tion are always present (88,93) The "floor" of an

end-to-side anastomosis (in the recipient vessel opposite

the anastomosis), the "toe" of the anastomosis (on the

near wall just beyond the suture line), and the "heel" (on

the near wall proximal to the junction) appear to be

prominent sites of flow separation where shear is low and

shear stress fluctuates (94,95) Because low shear and

os-cillatory shear stresses are conducive to platelet adhesion,

intimal hyperplasia, and atherosclerosis (36,88,93,96),

the ultimate success of an arterial reconstruction may

de-pend on how closely the surgeon adheres to recognized

he-modynamic principles in constructing the anastomosis

Geometric considerations make it impractical to

re-duce the graft-host vessel angle of a conventional

end-to-side anastomosis much below 30% without unduly

extending the length of the suture line (Disregarding the

additional few millimeters of anastomotic length

as-sociated with the change from a circular to elliptical

cross-section that occurs when a larger graft is joined to a

smaller artery, the minimum length of a 30% anastomosis

would be twice the diameter of the graft, while that of a

10% anastomosis would be almost six times the graft

FLOW SEPARATION

FIGURE 8.14 Flow patterns at end-to-side and

side-to-end anastomoses Note areas of flow separation Flow

in some areas may reverse and travel

circumferential-ly to reach the recipient artery or graft (Reproduced

by permission from Sumner DS Hemodynamics of abnormal

blood flow, in: Wilson SE, Veith FJ, et al., eds Vascular

surgery, principles and practice New York: McGraw-Hill,

1987.)

diameter.) The Taylor patch, which uses a vein patch to tend the toe of the anastomosis, makes construction of a10% anastomotic angle possible (97) Finite elementanalysis has confirmed that wall shear stress gradients atthe critical toe and heel regions are significantly less withthe Taylor patch than they are with the standard an-astomosis, especially during exercise (98) These samecomputational methods have been used to design an "op-timized" end-to-side anastomotic configuration thatgreatly reduces wall shear stress compared to the standardand Taylor patch configurations (98) The optimizedanastomosis has a smooth transitional curve at the heeland toe, an anastomotic angle of 10 % to 15 %, and a 1.6:1graft-to-artery diameter ratio Because the dimensionsand location of recipient arteries vary, fashioning thehoods of autogenous grafts or fabricating prosthetic cuffsthat meet the ideal specifications may not be possible

ex-Bifurcation Grafts

When Y grafts used for aortobiiliac and aortobifemoralbypasses have secondary limbs with diameters that areone-half that of the primary tube, each of the secondarylimbs has 16 times the resistance of the primary tube, and,

in parallel, they have eight times the resistance of the mary tube (Fig 8.15) Flow velocity is doubled, and

pri-FIGURE 8.15 Hemodynamic attributes of bifurcation

grafts, (r,, radius of primary tube; r2, radius of ondary limbs; A,,, cross-sectional area of primary tube;and A2, cross-sectional area of secondary tube.) (Re-produced by permission from Strandness DE Jr Sumner DS.Hemodynamics for surgeons New York Crune and Stratton,

sec-1975.)

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 133almost 50% of the incident pulsatile energy is reflected.

The reflected energy may contribute to weakening of the

proximal suture line in a severely diseased friable aorta,

leading to the development of false aneurysms and

aor-toenteric fistulas (99) Clearly, this is not the optimum

configuration (100)

No geometric configuration will satisfy all

require-ments (12) For example, to maintain a constant flow

ve-locity across the bifurcation, the ratio of the diameter of

the secondary tube to that of the primary tube must be

0.71; to maintain the same pressure gradient, the diameter

ratio must be 0.84; and to achieve minimal pulse

reflec-tion, the diameter ratio must be 0.76 (see Fig 8.15) In

animals and in human infants, the ratio is about 0.74 to

0.76, suggesting that the body attempts to minimize

re-flections at bifurcations The 16 x 9mm, 14 x 8mm, and

12 x 7mm grafts that are now commercially available

have diameter ratios of 0.56, 0.57, and 0.58 respectively

While these ratios represent some improvement over the

0.5 ratio of the older grafts, they still result in increased

flow velocity, an increased pressure gradient, and

relative-ly little decrease in the amount of energy reflected (30%

vs 50%) Thus the hemodynamically optimum

bifurca-tion graft has yet to be manufactured

The angle between the limbs of a bifurcation graft is

also of hemodynamic importance Flow disturbances are

minimized when the angle is narrow and are exaggerated

when the limbs are widely separated (Fig 8.16) The latter

configuration generates regions of flow separation along

the walls opposite the flow divider, encouraging the

depo-sition of thrombus By keeping the primary limb short and

using longer secondary limbs, the surgeon can reduce the

angle

FIGURE 8.16 Effect of angle between limbs of

bifurca-tion graft on flow disturbances When the limbs are

widely separated, areas of flow separation (indicated

by shading) develop (Reproduced by permission from

Malan E, Longo T Principles of qualitative hemodynamics in

vascular surgery, in: Haimovici H, ed Vascular surgery,

princi-ples and techniques, 2nd ed East Norwalk, CT:

Appleton-Century-Crofts, 1984.)

Conclusion

Understanding the symptoms of arterial occlusive disease,interpreting the results of physiologic tests, and planningeffective surgical therapy are all facilitated by a basicknowledge of hemodynamic and rheologic principles.When predicting the effects of a stenosis, a graft, or otherchanges in the vascular circuit, one must consider all as-pects of the circuit, including collateral input, peripheralresistance, autoregulation, direction of flow, steal phe-nomena, and inertial factors; otherwise, "armchair"conclusions are apt to be erroneous This chapter hasconcentrated on "generic solutions" to various problemscommonly encountered in vascular surgery and has basedthese solutions primarily on models; consequently, the ab-solute values may differ somewhat from those encoun-tered in real life Each situation is different and requirescareful physiologic assessment, by either noninvasive orinvasive measurement of both pressure and flow It ishoped that this chapter will stimulate others to make thesemeasurements and that the information presented will aid

4 Young DF, Tsai FY Flow characteristics of models ofarterial stenosis II Unsteady flow J Biomech 1973; 6:547

5 Berguer R, Hwang NHC Critical arterial stenosis: atheoretical and experimental solution Ann Surg 1974;180:39

6 Daugherty HI, Franzini JE Steady flow of ible fluids in pipes In: Fluid mechanics with engineeringapplications, 4th edn New York: McGraw-Hill, 1965:191

incompress-7 Young DF, Tsai FY Flow characteristics in models of terial stenoses I Steady flow J Biomech 1973; 6: 395

ar-8 Flanigan DP, Tullis JP, et al Multiple subcritical arterialstenoses: effect on poststenotic pressure and flow AnnSurg 1977; 186: 663

9 Karayannaeos PE, Talukder N, et al The role of multiplenoncritical arterial stenoses in the pathogenesis of is-chemia J Thorac Cardiovasc Surg 1977,73:458

10 Cheng LC, Clark ME, Robertson JM Numerical tions of oscillating flow in the vicinity of square wall ob-stacles in plane conduits J Biomech 1972; 5:467

calcula-11 Byar D, Fiddian RV, et al The fallacy of applyingPoiseuille equation to segmented arterial stenosis AmHeart] 1965; 70:216

12 Strandness DE Jr Sumner DS Hemodynamics for geons New York: Grune & Stratton, 1975

sur-13 May AG, Van deBerg L, et al Critical arterial stenosis.Surgery 1963; 54:250

134 Part II Basic Cardiovascular Problems

14 Moore WS, Malone JM Effect of flow rate and vessel

calibre on critical arterial stenosis J Surg Res 1979: 26:1

15 Moore WS, Hall AD Unrecognized aorto-iliac stenosis

A physiologic approach to the diagnosis Arch Surg

1971:103:633

16 Jones RD, Berne RM Intrinsic regulation of skeletal

muscle blood flow Circ Res 1964; 14:126

17 Kjellmer I On the competition between metabolic

va-sodilation and neurogenic vasoconstruction in skeletal

muscle Acta Physiol Scand 1965; 63:450

18 Ludbrook J Collateral artery resistance in the human

lower limb J Surg Res 1966; 6:423

19 Shepherd Jr Physiology of the circulation in human limbs

in health and disease Philadelphia: WB Saunders, 1963

20 Wolf EA Jr SumnerDS, Strandness DE Jr Correlation

between nutritive blood flow and pressure in limbs of

pa-tients with intermittent claudication Surg Forum 1972;

23:238

21 Edwards EA Scope and limitations of collateral

circula-tion Arch Surg 1984; 119: 761

22 Spencer MP, Reid JM Quantitation of carotid stenosis

with continuous-wave (c-w) Doppler ultrasound Stroke

1979:10:326

23 Russell JB, Miles RD, et al Effect of arterial stenosis on

Doppler frequency spectrum Proc 32nd Annu Conf Eng

MedBioll979;21:45

24 Keitzer WE Fry WJ, et al Hemodynamic mechanism for

pulse changes seen in occlusive vascular disease Surgery

1965; 57:163

25 Strandness DE Jr., Bell JW Peripheral vascular disease

Diagnosis and objective evaluation using a mercury

strain gauge Ann Surg 1965; 161 (Suppl): 1

26 Darling RC, Raines JK, et al Quantitative segmental

pulse and volume recorder: a clinical tool Surgery 1973;

72: 873

27 Jager KA, Phillips DJ, et al Noninvasive mapping of

lower limb arterial lesions Ultrasound Med Biol 1985; 1

1:515

28 Woodcock JE, Gosling RG, Fitzgerald DE A new

non-invasive technique for assessment of superficial femoral

artery obstruction Br J Surg 1972: 59:226

29 Johnston KW, Matuzzo BC, Cobbold RSC Doppler

methods for quantitative measurement and localization

of peripheral arterial occlusive disease by analysis of the

blood velocity waveform Ultrasound Med Biol 1978; 4:

209

30 Evans DH, Barrie WW, et al The relationship between

ultrasonic pulsatility index and proximal arterial

stenoses in a canine model Circ Res 1980; 46:470

31 Baird RN, Bird DR, et al Upstream stenosis, its diagnosis

by Doppler signals from the femoral artery Arch Surg

1980; 115:1316

32 Rittenhouse EA, Maxiner W, et al Directional arterial

flow velocity: a sensitive index of changes in peripheral

vascular resistance Surgery 1976; 79:359

33 Farrar DJ, Malindzak GS Jr., Johnson G Jr Large vessel

impedance in peripheral atherosclerosis Circulation

1977; 56 (Suppl 2}: 171

34 Skidmore R, Woodcock JP Physiological interpretation

of Doppler-shift waveforms II Validation of the Laplace

transform method for characterization of the common

femoral blood-velocity/time waveform Ultrasound Med

Biol 1980: 6: 219

35 Berguer R, Higgins RF, Reddy DJ Intimal hyperplasia

An experimental study Arch Surg 1980; 115: 332

36 Zarins CK, Giddens DP, et al Carotid bifurcation osclerosis: quantitative correlation of plaque localizationwith flow velocity profiles and wall shear stress Circ Res1983:53:502

ather-37 Ku DN, Giddens DP, et al Pulsatile flow and rosis in the human carotid bifurcation Positive correla-tion between plaque location and low and oscillatingshear stress Arteriosclerosis 1985; 5:293

atheroscle-3 8 Sharp WV, Donovan DL, et al Arterial occlusive disease:

a function of vessel bifurcation angle Surgery 1982; 91:680

39 Ojha M, Johnston KW, Cobbold RSC Evidence of a sible link between poststenotic dilation and wall shearstress J Vase Surg 1990; 1 1:127

pos-40 Boughner DR, Roach MR Effect of low frequency tion on the arterial wall Circ Res 1971; 29:136

vibra-41 Zarins CK, Bomberger RA, Glagov S Local effects ofstenoses: increased flow velocity in hi bits atherogenesis.Circulation 1981; 64 (Suppl 2): 221

42 Vaishnav RM, Patel DJ, et al Determination of the localerosion stress of the canine endothelium using a let im-pingement method ASME J Biomech Eng 1983; 105:77

43 Zarins CK, Zatina MA, et al Shear stress regulation ofartery lumen diameter in experimental atherogenesis JVase Surg 1987; 5:413

44 Glagov S, Weisenberg F, et al Compensatory ment of human atherosclerotic coronary arteries N EnglJMed 1987; 316:1371

enlarge-45 Ouriel K, Donayre C, et al The hemodynamics of bus formation in arteries J Vase Surg 1991; 14: 757

throm-46 Pentecost BL The effect of exercise on the external iliacvein blood flow and local oxygen consumption in normalsubjects, and in those with occlusive arterial disease ClinSci 1964:27:437

47 Lassen NA, Kampp M Calf muscle blood flow duringwalking studied by the Xel33 method in normals and inpatients with intermittent claudication Scand J Clin LabInvest 1965; 17:447

48 Folse R Alterations in femoral blood flow and resistanceduring rhythmic exercise and sustained muscular con-tractions in patients with arteriosclerosis Surg GynecolObstet 1965; 121: 767

49 Hauser CJ, Shoemaker WC Use of transcutaneous Po2

regional perfusion index to quantify tissue perfusion inperipheral vascular disease Ann Surg 1983; 197: 337

50 Clyne CAC, Ryan J, et al Oxygen tension on the skin ofischemic legs Am J Surg 1982; 143.315

51 Tonnesen KH, Noer I, et al Classification of peripheralocclusive arterial disease based on symptoms, signs, anddistal blood pressure measurements Acta Chir Scand1980; 146:101

52 Ramsey DE, Manke DA, Sumner DS Toe bloodpressure—a valuable adjunct to ankle pressure measure-ment for assessing peripheral arterial disease J Cardio-vasc Surg 1983; 24:43

53 Remington JW, Wood EH Formation of peripheral pulsecontour in man J Appl Physiol 1956; 9:433

54 Yao ST Haemodynamic studies in peripheral arterial ease BrJ Surg 1970; 57: 761

dis-55 Westerhof N, Sipkema P et al Forward and backwardwaves in the arterial system Cardiovasc Res 1972; 6: 648

Chapter 8 Hemodynamics of Vascular Disease: Applications to Diagnosis and Treatment 135

56 StrandnessDEJr, Bell JW An evaluation of the

hemody-namic response of the claudicating extremity to exercise

SurgGynecolObstetl964; 119:1237

57 Sumner DS, Strandness DE Jr The relationship

between calf blood flow and ankle blood pressure in

pa-tients with intermittent claudication Surgery 1969; 65:

763

58 Lewis JD,PapathanaiouC,etal Simultaneous flow and

pressure measurements in intermittent claudication Br J

Surgl972;59:418

59 Sumner DS, Strandness PE Jr The effect of exercise on

re-sistance to blood flow in limbs with an occluded

superfi-dal femoral artery Vase Surg 1970; 4:229

60 Angelides NS, Nicolaides AN, et al The mechanism

of calf claudication: studies of simultaneous clearance

of 99-Tc from the calf and thigh Br J Surg 1978; 65:

204

61 Angelides NS, Nicolaides AN Simultaneous isotope

clearance from the muscles of the calf and thigh Br J Surg

1980; 67: 220

62 Allwood MJ Redistribution of blood flow in limbs

with obstruction of a main artery Clin Sci 1962; 22:

279

63 Sumner DS Hemodynamics of abnormal blood flow In:

Wilson SE, Veith FJ, et al., eds Vascular surgery,

princi-ples and practice New York: McGraw-Hill, 1987

64 Rosenthal SL, Guyton AC Hemodynamics of collateral

vasodilation following femoral artery occlusion in

anes-thetized dogs Circ Res 1968:23:239

65 Conrad MC, Anderson JL 111, GarrettJB Jr Chronic

col-lateral growth after femoral artery occlusion in the dog J

ApplPhysioll971;31:550

66 FlaniganDP, RyanTJ, etal Aortofemoral or

femoropopliteal revascularization? A prospective

evalu-ation of the papaverine test J Vase Surg 1984; 1:215

67 Schultz RD, Hokanson DE, Strandness DE Jr

Pressure-flow relations of the end-side anastomosis

Surgery 1967; 62: 319

68 Sanders RJ, Kempczinski RF, et al The significance of

graft diameter Surgery 1980; 88: 856

69 Whitney DG,Kuhn EM, EstesJW Valvular occlusion of

the arterialized saphenous vein Am Surg 1976; 42: 879

70 McCaughan JJ, Walsh DB, et al In vitro observations of

greater saphenous vein valves during pulsatile and

non-pulsatile flow and following lysis J Vase Surg 1984:1:

356

71 Walsh DB, Downing S, et al Valvular obstruction of

blood flow through saphenous veins.J Surg Res 1987; 42:

39

72 Ku DN, Klafta JM, et al The contributions of valves to

saphenous vein graft resistance.J Vase Surg 1987; 6: 2

74

73 Okadone K, Yukizane T, et al Ultrastructural evidence of

the effects of shear stress variation on intimal thickening

in dogs with arterially transplanted autotogous grafts J

Cardiovasc Surg 1990; 31: 719

74 Sharefkin JB, Diamond SL, et al Fluid flow decreases

preproendothel in mRNA levels and suppresses

endothelin-1 peptide release in cultured human

endothe-lial cells J Vase Surg 1991; 14:1

75 Brewster DC, Charlesworth PM, et al Isolated popliteal

segment v tibial bypass Comparison of hemodynamic

and clinical results Arch Surg 1984:119: 775

76 Jarrett F, Perca A, et al Hemodynamics of sequentialbypass grafts in peripheral arterial occlusions SurgGynecol Obstet 1980; 150: 377

77 Jarrett F, Berkoff HA, et al Femorotibial bypass graftswith sequential techniques Arch Surg 1983; 116: 709

78 Hadcock MM, Ubatuba J, et al Hemodynamics of quential grafts Am J Surg 1983; 146:170

se-79 Flinn WR, Flanigan DP, et al Sequential femotal-tibialbypass for severe limb ischemia Surgery 1980; 88: 357

8 0 Ascer E, Veith FJ, et al Components of outflow tance and their correlation with graft patency in lowerextremity arterial reconstructions J Vase Surg 1984:1:817

resis-81 LaMorte WW, Menzoian JO, et al A new method for theprediction of peripheral vascular resistance from the pre-operative angiogram J Vase Surg 1985; 2: 703

82 Ascer E, Veith FJ, et at Intraoperative outflow resistance

as a predictor of late patency of femoropopliteal andinfrapopliteat arterial bypasses J Vase Surg 1987; 5:820

83 Bliss BP Peripheral resistance in the leg in arterial sive disease Cardiovasc Res 1971; 5: 337

occlu-84 Sumner DS, Strandness DE Jr The hemodynamics of thefemorofemoral shunt Surg Gynecol Obstet 1972; 134:629

85 Shin CS, Chaudhry AG Hemodvnamics of anatomical bypass following restriction of inflow' andoutflow in the donor artery in dogs World J Surg 1980;4: 71 7

extra-86 Malan E, Noseda G, Longt T Approach to fluid dynamicproblems in reconstructive vascular surgery Surgery1969; 66: 994

87 Malan E, Longo T Principles of qualitative ics in vascular surgery In: Haimovici H Vascularsurgery, 2nd edn Norwalk, CT: Appleton-Century-Crofts, 1984

hemodynam-8 hemodynam-8 Crawshaw HM, Quist WC, et al Flow disturbance at thedistal end-to-side anastomosis Effect of patency of theproximal outflow segment and angle of anastomosis.Arch Surg 1980; 115:1280

89 Klimach 0, Chapman BLW, et al An investigation intohow the geometry of an end-to-side arterial anastoinosisaffects its function Br J Surg 1984; 71:43

90 Lye CR, Sumner DS, Strandness DE Jr The ics of the retrograde crosspubic anastoniosis Surg Forum1975; 26:298

hemodynam-91 Bond MG, Hosteller JR, et al Intimal changes in ovenous bypass grafts Effect of varying the angle of im-plantation at the proximal anastomosis and of producingstenosis in the distal runoff artery J Thorac CardiovascSurg 1976; 71: 907

arteri-92 LoGerfo FW, Quist WC, et al Downstream anastomotichyperplasia A mechanism of failure in Dacron arterialgrafts Ann Surg 1983; 197:479

93 LoGerfo FW, Soncrant T, et al Boundary layer tion in models of side-to-end arterial anastomoses ArchSurg 1979; 114:1369

separa-94 Bassiouny HS, White S, et al Anastomotic intimal plasia: mechanical injury or flow induced J Vase Surg1992; 15: 708

hyper-95 Ojha M, Ethier CR, et al Steady and pulsatile flow fields

in an end-to-side arterial anastomosis model J Vase Surg1990; 12: 747

136 Part El Basic Cardiovascular Problems

96 McMillan DE Blood flow and the location of

atheroscle-rotic plaques Stroke 1985; 16:582.

97 Taylor RS, Loh A, McFarland RJ, Cox M, Chester JF

In-proved techniques for PTFE bypass grafting: long-term

results using anastomotic vein patches Br J Surg 1991;

79: 348

9 8 Lei M, Archie JP, Kleinstreuer C Computational design

of a bypass graft that minimizes wall shear stress

gradi-ents in the region of the distal anastomosis J Vase Surg

1997; 25:637.

99 Newman DL, Gosling RG, et al Pressure amplitude

in-crease on unmatching the aorto-iliac unction of the dog.

Cardiovasc Res 1973; 7: 6.

100 Buxton BF, Wukasch DC, et al Practical considerations

in fabric vascular grafts Introduction of a new bifurcated

graft Am J Surg 1973; 125:288.

Bibliography

Archie JP Jr Presidential address: A brief history of arterial blood flow—from Harvey and Newton to computational analysis J Vase Surg 2001; 34:398.

Nichols WW, O'Rourke ME McDonald's blood flow in ies Philadelphia: Lea & Febiger, 1990.

arter-Milnor WR Hemodynamics, 2nd edn Baltimore: Williams & Wilkins, 1989.

Strandness DE Jr., Sumner DS Hemodynamics for surgeons New York: Grune & Stratton, 1975.

Patel DJ, Vaishnav RN Basic hemodynamics and its role in ease processes Baltimore: University Park Press, 1980 Sumner DS Essential hemodynamic principles In Rutherford

dis-RB, ed Vascular surgery, 5th edn Philadelphia: WB Saunders,2000.

C H A P T E R 9

Atherosclerosis: Biological and Surgical Considerations

Bauer E.Sumpio

Historical Perspective

The word atherosclerosis is derived from the Greek—

meaning both softening (athere) and hardening (skleros)—

and refers to a complex disease process affecting the major

blood vessels of the body It is a disease that has plagued

humans for centuries There is evidence that ancient

Egyp-tians suffered from atherosclerosis much the same way as

we do now Paleopathologists have used sophisticated

his-tological techniques to study the blood vessels of Egyptian

mummies dating to 1400 BC Peripheral arteries were

har-vested from limbs that had escaped the mutilation that

usually accompanied embalming Patches of

atheroma-tous plaques lined along the length of the aorta, the

com-mon carotid, and the iliac vessels The smaller tributaries

of the vessels of the lower limbs were like calcified tubes

Histologically, these ancient diseased vessels

demonstrat-ed endothelial and muscular degeneration with focal

areas of increased fibrosis and calcification (Fig 9.1)

(1,2)

The study of atherosclerosis spans centuries, but the

most significant findings have been made only within the

last 150 years (Table 9.1) Although the ancient Greek

physician Galen reported many vascular anomalies such

as aortic and peripheral arterial aneurysms, there is no

ev-idence that he described atherosclerotic lesions (3) Even

as late as the sixteenth century, when the infamous

anatomist Andreas Vesalius carefully characterized

aneurysms, there was still no concept of the

atherosclerot-ic lesion and its signifatherosclerot-icance (4) Despite the contribution

of William Harvey and Daniel Sennet to the

understand-ing of the anatomy and physiology of the circulatory

sys-tem, there was still no recognition of the atheroscleroticdisease process (5) It was not until the mid-seventeenthcentury that a process that resulted in degeneration of thearteries with advancing age was recognized In 1755, theSwiss physiologist Albrecht von Haller reported on pro-gressive atherosclerotic changes in the blood vessels of theelderly (6) Later, in 1761, the Italian physician andpathologist Giovanni Battista Morgagni heralded the idea

of using microscopic evaluation of tissues to correlate ease with histology His work, and that of his pupil Anto-nio Scarpa, correlated a lesion they described as similar to

dis-an ulcerated plaque to dis-aneurysm formation (7) Thus,atheromatous lesions became the focus of study—first, as

a precursor to aneurysm formation, and then, as a rate pathologic entity

sepa-The earliest evidence of understanding sis comes from the research of a surgeon, Joseph Hodg-son, in London He proposed that inflammation was theunderlying cause of these plaque formations and hypoth-esized that the process was linked to the intimal layer ofblood vessels In his monograph (1852), the Viennesepathologist Carl Rokitansky included accurate descrip-tions of atherosclerotic lesions Rokitansky was one of thefirst to observe and document that there were both throm-bogenic and calcific components to atherosclerotic lesions(8) Eventually, the proposals of Hodgson and Rokitan-sky were clarified by the pioneering observations andstudies done by Rudolf Virchow (Fig 9.2) Virchow con-cluded that atherosclerotic lesions were located in the inti-mal layer and described the process of plaque formationthat was initiated by the formation of a coagulum which

atherosclero-he called thrombus By studying microscopic sections of

137

138 Part II Basic Cardiovascular Problems

diseased vessels, Virchow generated a theory of

athero-sclerosis that involved connective tissue proliferation

stimulated by intimal deposits, resulting in further vessel

wall degeneration (9,10)

The studies of Alexander Ignatovski and Nikolai

An-itschkov in the early 1900s demonstrated that

atheroscle-rotic changes could be induced in animals by a diet rich in

cholesterol (11) This led to the important discovery in

1910 by German chemist Adolf Windaus that human

ath-erosclerotic lesions contained cholesterol Further research

has focused not only on understanding the atherosclerotic

process but also on trying to intervene to retard and reverse

the clinical manifestations of this disease

Epidemiology

The word atherosclerosis should not be used

interchange-ably with arteriosclerosis, a word introduced by French

pathologist Jean Lobstein in 1829 Arteriosclerosis refers

to generalized hardening and thickening of arterieswhereas atherosclerosis is more specific to the process re-sulting in lipid accumulation within the intimal layer ofblood vessels In arteriosclerosis, the increase in vesselwall thickness is due to an increased amount of basementmaterial and plasma protein deposition (12) Althoughoften associated with hypertension, arteriosclerosis isnot necessarily pathologic and may simply represent be-nign changes that occur as a result of the aging process.Interestingly, however, arteriosclerosis is pronounced

in patients with hypertension and diabetes mellitus—

FIGURE 9.1 Frozen section of tibial artery from

Egypt-ian mummy Lipid deposition can be seen in an

atheromatous lesion

FIGURE 9.2 Rudolf Virchow (1821-1902) He made

significant contributions to the understanding ofatherosclerosis and vascular disease

TABLE 9.1 Historical evolution of the understanding of vascular disease

Andreas Vesalius and Gabriel Fallappio 1500s

Willam Harvey 1628

Daniel Sennet 1628

Albrecht von Haller 1755

Giovanni Batista Morgagni 1761

Nikolai Anitschkov and Ludwig Aschoff 1933

Described aortic and peripheral aneurysms Described cardiovascular system as a circuit Described arteries as comprised of two concentric layers Described progressive changes within arterial walls Described microscopic changes occurring within atheromas Correlated ulcerated atheromatous lesions with aneurysmal development Proposed inflammation as a cause of atherosclerosis

Coined the term arteriosclerosis Detailed descriptions of early and mature atheromatous plaques Described the process of thrombosis and embolism

Experimentally induced atherosclerosis in rabbits Discovered cholesterol within atherosclerotic lesions Provided summaries of early experimentation and reults regarding the research of atherosclerosis

Chapter 9 Atherosclerosis: Biological and Surgical Considerations 139

diseases that are both also associated with a higher risk of

atherosclerosis

Atherosclerosis-related cardiovascular disease is the

most common cause of morbidity and mortality in the

United States Atherosclerosis resulting in myocardial

in-farction, stroke, and gangrene of the extremities is

respon-sible for approximately 50% of all mortality As will be

discussed later, atherosclerosis has a predilection for

spe-cific anatomic sites at the ostia and bifurcations of the

aorta, iliac, and femoral arteries Atherosclerosis remains

the leading disorder affecting lower limb circulation

Infrapopliteal arteries are commonly affected,

contribut-ing to end-organ disease (i.e., ischemia and gangrene)

Patients with comparable degrees of atherosclerotic

dis-ease, anatomically, may, nonetheless, present with varying

degrees of clinical symptoms Symptomatology depends

on several different factors other than the presence or

ab-sence of atherosclerosis (13) For example, the rate of

disease progression, the severity of the decrease in blood

flow, the presence or absence of collateral circulation, and

the presence of thrombus or embolism causing acute

va-sospasm or occlusion are all factors affecting presentation

The majority of patients with peripheral arterial

dis-ease tend to exhibit a stable course over a 5-year period

However, 15-20% of these patients will eventually

devel-op tissue loss or rest pain requiring vascular surgery

Moreover, amputation will ultimately be required in 1 %

of patients per year The Framingham study allowed close

evaluation of a defined cohort over the span of 30 years A

comparison of incidences in angina, TIA, and calf

claudi-cation is shown (Fig 9.3) In comparison to angina,

pe-ripheral artery disease increases in prevalence throughout

life and even exceeds anginal symptoms if the patients live

over the age of 75 (14,15) In addition to heart and

pe-ripheral vascular disease, cerebrovascular disease is also a

major consequence of the atherosclerotic process Stroke,

with an incidence of 500,000 cases yearly, is the third

lead-ing cause of death in the US In one study, the annual

stroke rate was determined to be 1.3 % per year in patients

with up to 75% carotid stenosis The rate of stroke is

near-ly tripled in patients with higher-grade lesions (Table 9.2)

(16) Thus, the results of untreated or poorly treated

ath-erosclerotic disease has significant medical consequences

Normal Anatomy

The vascular system is derived from the mesoderm andoriginates as aortic arches which bridge to connect the em-bryonic dorsal aorta to the aortic sac Some branches ofthe dorsal aorta remain as either intercostal arteries orlumbar intersegmental arteries The fifth pair of lumbarintersegmental arteries become the common iliac arteries

By the fourth week of development, the aortic archestransform and develop into their adult derivatives Ofnote, the third pair of arches becomes the common carotidarteries and the pulmonary arteries arise from the sxthpair

The earliest vascular primordia are endothelial cellclusters called blood islands These rests of cells arise onthe yolk sac between the splancnic mesoderm and endo-derm The blood island cells differentiate and separateinto peripherally located endothelial cells and centralblood cells Mesenchymal cells then migrate into thesubendothelial space and differentiate into smooth mus-cle cells Development of the extracellular matrix thenprogresses as smooth muscle cells and fibroblasts secreteangiogenic factors such as fibroblast growth factor (FGF)and vascular endothelial cell growth factor (VEGF).These signaling substances promote the generation of newbranches that extend from the preexisting main vessels(Fig 9.4) (17)

Arteries are made up of three distinct concentric ers (Fig 9.5) The innermost layer, the intima, is com-posed of endothelial cells The media, the next layer,contains smooth muscle cells in various configurationsand is separated from the intima by the internal elasticlamina, a network of alveolar and elastic tissue The out-ermost layer, the adventitia, is a meshwork of collagen,elastic, and fibrous tissue that, along with the media, pro-vides a strong physical support The media and adventitiaare separated by the external elastic lamina The intimaand inner portion of the media receive blood supply di-rectly from luminal blood In contrast, there is a complexnetwork of small vessels called the vasa vasorum that sup-ply the adventitia and outer media (18)

lay-The proximal vessels are subjected to a high-pressuresystem and this is reflected in the structure and composi-tion of these vessels as represented in Figure 9.6 Theseproximal, large elastic arteries serve to smooth the flow ofblood through systole and diastole The media of thesearteries have thick, highly organized layers of elasticfibers arranged circumferentially that expand and recoil

TABLE 9.2 The risk of transient ischemic attacks (TlAs)and stroke in patients with asymptomatic carotidstenosis

FIGURE 9.3 Symptoms of atherosclerotic disease in

the Framingham study

<50% (mild)50-75% (moderate)

>75% (severe)

1.03.07.2

1.31.33.3

140 Part II Basic Cardiovascular Problems

FIGURE 9.4 Differentiation of vessels in the embryo.

The process proceeds from endothelium

differentia-tion to full development of veins and arteries

through each cardiac cycle In contrast, distal blood

ves-sels tend to be more muscular in structure The media of

these arteries are comprised mainly of smooth muscle cells

with few intermixed unorganized elastic fiber layers

Muscular arteries are highly contractile and under the

direct control of the autonomic nervous system

Arteries consist of two major cell types:

1 Endotheiial cells line the luminal surface serve to

con-trol vascular tone and secrete matrix substances such

as elastin, collagen, and proteoglycans The lial cell layer serves to protect against thrombosis byproviding a selective barrier between circulatingblood and interstitial fluid

endothe-2 Smooth muscle cells are contained deeper within thearterial wall, constituting 40% to 50% of the medialvolume in large elastic arteries and 80% to 85% insmaller muscular arteries Smooth muscle cells main-tain vascular tone of the arterial wall and secreteextracellular matrix proteins such as elastin, colla-gen, and glycosaminoglycans In addition, smoothmuscle cells have been found to contain receptorsfor lipoproteins and growth factors and synthe-size prostaglandins to mechanically regulate bloodflow

In vivo, endothelial cells and smooth muscle cellsusually exist in a quiescent state The endothelium, viacontact inhibition, exists as an obligate monolayer.Smooth muscle cells, however, have been shown to have aturnover rate of 0.06% per day These two cell types existtogether with a complex network of signals between themmodulating each other's function For example, end-othelial cells secrete products which influence smoothmuscle cell function (19) Vasodilating substances such

as prostacyclin, prostaglandin E2, and dependent relaxing factor (EDRF) are secreted by func-tional endothelium in response to local thrombotic events(20) This may explain the observation that coronaryvessels with intimal lesions causing less than 40% luminalstenosis become dilated in response to changes in bloodflow Only after the intimal lesion occupies greater than40% of the lumen does the blood flow decrease (Fig 9.6)(21) The increase in vessel wall size is dependent not onendothelial cell proliferation, but on accumulation ofsmooth muscle cells and associated matrix proteins with-

endothelial-in the endothelial-intima Several stimulators have been elucidatedand, as will be discussed in a later section, platelet-derivedgrowth factor (PDGF) has been found to be one of themost potent

Theories of Atherosclerosis Monoclonal Hypothesis

This theory is borne from the observation by Benditt andBenditt that individual cells from plaques of heterozygotefemales for the X-linked glucose-6-phosphate dehydroge-nase (G-6PD) gene usually only exhibit one G-6PDisotype (Fig 9.7) (22) This suggests that the cells of a par-ticular plaque are derived from a single progenitorsmooth muscle cell, and, although, some smooth musclecells may infiltrate the intima, the bulk of the cells foundwithin a plaque are likely a result of monoclonal prolifer-ation of modified smooth muscle cells Another study hascorroborated the monoclonal behavior of human plaquecells using LDH as a marker LDH isoenzyme analysis

Chapter 9 Atherosclerosis: Biological and Surgical Considerations 141

FIGURE 9.5 (A) Cross-section of an arterial wall (B) Normal muscular artery (C) Normal elastic artery.

FIGURE 9.6 Diagrammatic representation of the

pos-sible sequence of changes occurring In an

atheroscle-rotic artery leading, eventually, to lumen narrowing

carried out on the blood vessel and plaque separately

revealed a shift in isoenzyme pattern (Fig 9.8) This shift

represents an alteration of smooth muscle type,

distin-guishing plaque smooth muscle cells from intimal smooth

muscle cells (23) This finding adds support to the

mono-clonal hypothesis, but does not explain other aspects of

the atherosclerotic process

intimal Cell Mass Hypothesis

This hypothesis comes from the observation that small

ac-cumulations of smooth muscle cells can be found in

chil-dren where atherosclerosis later develops It is uncertain

how these rests develop, but they may be primordial rests

of stem cells that are susceptible to atherogenesis These

accumulations of smooth muscle cells within the vessels of

FIGURE 9.7 Zymogram of samples from: (1) blood, (2)

normal tissue, and (3-6) atheromatous plaques ples from the different plaques demonstrate expres-sion of the Type A or Type B forms of the enzyme, both

Sam-of which are found In the blood and normal tissue

FIGURE 9.8 Representative LDH isoenzyme blot

stain-ing for: (A) media and (B) plaque

142 Part n Basic Cardiovascular Problems

children can be found worldwide regardless of the

preva-lence of atherosclerosis This suggests that the eventual

development of atherosclerosis is determined by extrinsic

factors such as increased cholesterol levels, cigarette

smoking, etc The Pathobiological Determinants of

Ath-erosclerosis in Youth (PDAY) research group examined

this phenomenon Coronary arteries, aortas, and other

pertinent tissues from persons 15 to 34 years of age were

collected and studied The researchers reported that aortic

fatty streak lesions were prevalent in almost all

individu-als by the age of 15 and that raised, fibrous plaques were

present in some by the age of 20 Consistent with clinical

observations, the coronary arteries of young males were

found to have a significantly increased number of raised

plaques as compared to their female counterparts (24)

Encrustation Hypothesis

The encrustation hypothesis proposes that repeated

cycles of thrombosis and healing serve as the source of

plaque progression As thrombosis is known to be a late

component of the atherosclerotic lesion, this theory does

not explain the initiation of plaque formation (25)

Lipid Hypothesis

An alternative hypothesis postulates that increased levels

of LDL results in abnormal lipid accumulation in smooth

muscle cells and macrophages as it passes through the

ves-sel wall (26,27) As LDL is oxidized, endothelial cells

become damaged and the atherogenic events mentioned

previously proceed to form plaque Oxidized lipoproteins

have been found to cause cell injury regardless of how the

oxidation occurs (28-31) Lipoprotein oxidation results

in the development of several toxic products that include

7-p-hydroperoxycholesterol, 7-ketocholesterol, lysoPC,

oxidized fatty acids, and epoxysterols (32,33) The exact

mechanism by which cell death occurs is not yet known

One theory for the development of atherosclerosis caused

by oxidized LDL is illustrated (Fig 9.9) (34) VLDL and

LDL accumulate in the intima Increased lipoprotein

lev-els and binding to connective tissue elements increases the

residence time of the lipoproteins in the intima, thereby

in-creasing the probability of undergoing oxidation (35,36)

Once oxidized, the modified lipoproteins stimulate the

entry of monocytes and lymphocytes into the intima

Ox-idized LDL also promotes migration and proliferation of

smooth muscle cells, contributing to the genesis of an

ath-erosclerotic plaque Evidence supporting this hypothesis

has recently emerged Specifically, it has been shown that

the lipid found in plaques comes directly from the blood

and there is substantial evidence that links

hypercholes-terolemia with an increased propensity to develop

athero-sclerotic lesions (37,38) Increased serum levels of LDL

lead to increased interstitial levels of LDL which bind to

proteoglycans This accumulation of LDL increases the

propensity for lipoprotein oxidation to occur which has

been shown to cause increased PDGF expression by

smooth muscle cells, increased death of proliferatingsmooth muscle cells, impaired endothelium healing, andmonocyte proliferation via upregulation of monocytechemoattractant protein-1 (MCP-1) secretion from en-dothelial cells (27)

Reaction-to-lnjury Hypothesis

The process of atherosclerosis is a chronic and insidiousone usually occurring over several decades Several theo-

FIGURE 9.9 Schematic of a hypothetical sequence in

which lipoprotein oxidation causes atherosclerosis

Chapter 9 Atherosclerosis: Biological and Surgical Considerations 143

ries have been postulated to explain how the process

begins One such theory arises from research that has

found that endothelial cell dysfunction leads to

athero-sclerosis Endothelial cell dysfunction results in increased

vascular permeability, increased leukocyte adherence,

and functional imbalances in pro- and antithrombotic

factors, growth modulators, and vasoactive substances

(Fig 9.10) This initial dysfunction of endothelial cells

FIGURE 9.10 Endothelial dysfunction in response to

injury

also triggers progression of atherosclerosis Leukocyteswhich accumulate at the site of injury release more growthfactors which induce migration of vascular smooth mus-cle cells into the intima This reaction-to-injury hypothe-sis also postulates that platelets which are present in areas

of denuded endothelium secrete potent mitogenic factors,thereby stimulating smooth muscle proliferation Thishypothesis incorporates three important processes thatare involved in atherogenesis:

1 focal intimal migration, proliferation, and tion of various cells such as macrophages and smoothmuscle cells;

accumula-2 increased production of extracellular matrix; and

3 lipid aggregation

These processes are set into motion when the vessel dothelium is exposed to some sort of injury Continuousexposure to endothelial injury elicits a chronic focal in-flammatory response that results in the development of anatherosclerotic plaque (Fig 9.11) Indeed, all of the theo-ries mentioned attempt to explain atherosclerosis, but, atbest, help only to explain particular aspects of a very com-plex process (39-41)

en-" FIGURE 9.11 Reaction-to-injury hypothesis.

Of note, each of the stages is potentially versible if the injurious agent(s) are removed

re-144 Part II Basic Cardiovascular Problems

Morphology and Hemodynamics

It should be reiterated that arterial blood vessels are

sub-jected to major hemodynamic forces which impact on the

endothelial cell lining The endothelial cell monolayer is

an active participant in the complex interactions that

occur between the luminal blood and vessel wall In fact, it

is the biologic response of the endothelium to

hemody-namic forces that is pivotal in the process of

atherosclero-sis The arterial blood vessel is subjected primarily to two

major hemodynamic forces: shear stress and cyclic strain

(Fig 9.12) As blood moves along the endothelium, a

tan-gential drag force is produced called shear stress (42,43)

The magnitude of the shear stress is directly proportional

to blood viscosity and inversely proportional to the radius

of the blood vessel cubed Research has shown that high

shear stress is inversely proportional to the distribution of

early intimal lesions That is, areas affected by increased

shear stress were protected and had fewer intimal lesions

compared with areas of low or oscillatory levels of shear

stress (Fig 9.13) (44-46) This finding has led to a vast

amount of investigation trying to characterize the effects

of hemodynamics on vascular biology Results of these

studies demonstrate that endothelial cells respond to

shear stress in several different ways (Table 9.3) For

ex-ample, endothelial cells have been found to change

align-ment in the direction of flow when subjected to shear

stress Additionally, reorganization of endothelial cell

F-actin contained within the cytoskeleton allows

morpho-logic changes to occur under the influence of shear stress

(Fig 9.14) (47-50) As illustrated in this figure, prominent

actin microfilament bundles are localized and aligned in

areas of high shear stress In areas where the shear stress is

low and flow is nonlaminar, the actin monofilament

bun-dles remain dense and nonaligned It has been shown that

shear stress inhibits endothelial cell migration and eration (51) Lastly, shear stress affects the biologic func-tion of endothelial cells, providing evidence of its role

prolif-in protectprolif-ing vessels from atherosclerosis Shear stressincreases prostacyclin secretion, which acts as a potentvasodilating and anti-platelet-aggregating substance(52,53) Similarly, secretion of tissue plasminogen activa-tor, a potent thrombolytic, and nitric oxide, a potentmediator of vasomotor tone and smooth muscle prolifer-ation, is enhanced with higher levels of shear stress(54,55) These findings imply a possible mechanism thatmay explain the finding of increased atherogenicity inareas of low shear stress

Cyclic strain refers to the repetitive, circumferentialpulsatile pressure distention conferred to the vessel wall

As with shear stress, endothelial cells react in specific ways

to cyclic strain Cultured endothelial cells have beenshown to proliferate and exhibit morphologic changes inresponse to cyclic strain (Table 9.3) The morphologicchanges occur secondary to actin rearrangement withinthe cytoskeleton resulting in an organized cellular align-ment perpendicular to the force vector (56,57) Severalmacromolecules have been found to be stimulated bycyclic strain As with cells that are subjected to shearstress, endothelial cells undergoing cyclic strain exhibit in-creased levels of prostacyclin and tPA In addition, en-dothelial nitric oxide synthase and, subsequently, nitricoxide levels are also increased (58-60) Moreover, cyclicstrain has been shown to stimulate expression of cellularadhesion molecules such as ICAM-1 (61) Studies havealso shown that second messenger systems such asthe adenylate cyclase-cAMP and diacylglycerol-IP3pathways become activated by cyclic strain (62) Al-

FICURE 9.12 Schematic of hemodynamic forces

gen-erated during systole The shear stress force vector is

parallel to blood flow and is unidirectional In

con-trast, cyclic strain force vectors are multiplanar and

multidirectional field in the area of the carotid bifurcation.FIGURE 9.13 Diagrammatic representation of the flow

Chapter 9 Atherosclerosis: Biological and Surgical Considerations 145

TABLE 9.3 Hemodynamic effects on cell function

Endothelial cells elongate and align themselves

in the direction of flowStimulation of nitric oxide, PGI2 and tPA secretionActivation of DAG/IP3 pathways, integrins, andMAP kinases

Stimulation of smooth muscle cell proliferationEndothelial cells align themselves perpendicular tothe force vector

Stimulation of nitric oxide, PGI2 and tPA secretionActivation of DAG/IP3 pathways, MAP kinases andTGF-p

FIGURE 9.14 Morphologic changes inactin microfilaments of rabbit aortaunder the effects of shear stress (A) Tho-racic aorta (B) Low shear stress (C) Highshear stress (D) Electron photomicro-graph of endothelial cells subjected tohigh stress Note the prominent actinmicrofilament bundle

though not clearly defined as yet, these findings may

provide clues to the mechanism or mechanisms by which

endothelial cell responses are mediated when affected by

atherosclerosis

In the support of the critical role of hemodynamic

forces, it should be noted that atherosclerotic lesions do 2

not occur randomly within the vasculature (63) Michael

DeBakey and co-workers (64) divided arterial plaque

dis-tribution into five categories They noted that the coro- 3

nary arteries, the major branches of the aortic arch, the

abdominal aorta, and the major visceral and lower

ex-tremity branches were particularly susceptible to

athero-sclerosis Plaque localization at these sites accounts for the 4

majority of clinical manifestations associated with this

disease (Fig 9.15) (65)

1 Category 1 includes the coronary arteries which

contain many branch points that are subjected to

me-chanical torsions during each heartbeat rotic lesions are commonly found at the bifurcations

Atheroscle-of major vessels such as where the left main coronaryartery splits into the left anterior descending and leftcircumflex arteries

Category 2 includes major branches of the aorticarch The carotid arteries are especially prone to ath-erosclerotic disease

The third category consists of the visceral branches ofthe abdominal aorta Susceptible category 3 arteriesinclude the celiac axis, the superior and inferiormesenteric arteries, and the renal arteries

Category 4 vessels include the distal abdominal aortaand its ileofemoral branches Most patients with ath-erosclerotic disease fall into this category Additional-

ly, patients with symptomatic plaques in the terminalaorta and lower extremities had the highest probabil-ity of having atherosclerotic disease elsewhere (64)