The value of acetazolamide single photon emission computed tomography scans in the preoperative evaluation of asymptomatic critical carotid stenosis.

PURPOSE: Acetazolamide (ACZ)-enhanced single photon emission computed tomography (SPECT) scans can assess both cerebral perfusion and vascular reactivity. Patients with asymptomatic critical carotid artery stenosis were evaluated for cerebral vascular reactivity to determine the effect of extracranial occlusive disease and the effect of carotid endarterectomy (CEA) on intracerebral reactivity. METHODS: In 44 patients with asymptomatic critical carotid artery stenosis, cerebral perfusion and vascular reactivity were assessed before CEA with resting and ACZ-enhanced SPECT scans. All patients had a 70% or greater ipsilateral internal carotid artery stenosis. Preoperative ACZ-enhanced SPECT scans were obtained, usually 5 days before CEA. Postoperative ACZ-enhanced SPECT scans were obtained in 30 patients. RESULTS: Preoperative SPECT scans were asymmetric, revealing focal (n = 19) or global (n = 15) decreased reactivity in 34 patients (77%). Ten patients had symmetric or normal reactivity. After CEA, 23 patients demonstrated an improvement in reactivity ipsilateral to the side of surgery. The remaining seven patients failed to improve after surgery. CONCLUSION: Although all patients had a high-grade internal carotid stenosis, nearly a quarter of the patients had excellent intracerebral collateral flow. Only 71% of patients demonstrated improved intracerebral vasoreactivity after CEA. The lack of improvement in the other patients may have resulted from intracerebral pathology or lack of improvement in the extracranial carotid hemodynamics.

Drag reducing polymers may decrease atherosclerosis by increasing shear in areas normally exposed to low shear stress.

PURPOSE: Drag reducing polymers (DRPs) have been shown to decrease plaque formation. Their mechanism of action is unknown. Atherosclerosis tends to develop in areas of low shear stress. This study investigates whether DRPs increase shear stress in areas normally exposed to low shear stress. METHODS: Six dogs underwent surgical plication of the left half of the aorta. A specially modified 20-MHz Doppler ultrasound probe mounted at a 45-degree angle on a micromanipulator was used to measure blood flow velocity at six 4-mm intervals along both lateral sides of the aorta starting at the aortic wall and then at subsequent 0.1-mm depths moving into the lumen before and after administering DRP. Shear rates were calculated using linear regression and then compared using the paired t test. The blood viscosity remained constant at 0.04 poise during infusions of this amount of DRP. RESULTS: The maximum shear rate occurring during the cardiac cycle on the side of the aortic stenosis (plication) was 9.96 +/- 1.52/sec before the administration of the DRP and 14.27 +/- 2.01/sec after the administration of the DRP (P =.0240). The maximum shear rate on the side of the unstenosed aortic wall was 57.25 +/- 7.93/sec before the administration of the DRP and 44.80 +/- 6.23/sec after the administration of the DRP (P =. 0081). CONCLUSION: One of the ways that DRPs inhibit the development of atherosclerosis appears to be by increasing shear stress in areas normally exposed to low shear stress. Understanding this mechanism may lead to the development of pharmaceutical agents that inhibit the development of atherosclerosis.

Primary aortic mural thrombus: presentation and treatment.

The clinical presentation of patients with acute lower-limb ischemia and primary aortic thrombus prompted this review. Following recognition of the first case in early 1994, relevant patients (n = 6) were kept in a database and were reviewed for presentation, treatment, and follow-up. The median age was 41 and five patients were male. Angiography, computed tomography, and/or magnetic resonance angiography demonstrated one or more aortic sessile or pedunculated thrombus(i) without associated atherosclerotic disease. In two cases, a retropancreatic intraaortic mural thrombus was associated with severe pancreatitis. All other cases presented with acute lower-limb emboli requiring limb salvage embolectomy. Because of significant patient illness, systemic anticoagulation was chosen acutely to prevent recurrent emboli. Interestingly, serial studies demonstrated aortic thrombus resolution. Failure to continue warfarin therapy resulted in recurrent problems (n = 1) unless the instigating event had resolved (n = 3). There were no deaths or amputations. We concluded that surgical embolectomy, when required, with subsequent anticoagulation, results in limb salvage and allows for eventual resolution of the primary aortic thrombus. Long-term anticoagulation is required unless the etiologic process resolves. The literature describes patients with atherosclerosis and overlying thrombus but fails to describe the approach to patients with primary thrombus formation.

Aggressive management of vascular injuries of the thoracic outlet.

PURPOSE: We present a review of 35 patients who underwent an operation for subclavian (n = 18) or axillary (n = 17) vessel injury. In some patients, both an artery and a vein were damaged, resulting in a total of 30 arterial and 16 venous injuries. METHODS: The wounding source included a gunshot (n = 19), a stab wound (n = 9,) and blunt trauma (n = 7). Seven patients had hypotension and were taken immediately to the operating room. Seventeen patients had diminished or absent pulses, whereas 13 patients had normal pulses despite an arterial injury. Associated injuries included nerve injury (n = 15), pneumohemothorax (n = 5), and fractures (n = 7). Angiography in 21 patients demonstrated an intimal flap (n = 8), extravasation (n = 5), a pseudoaneurysm (n = 3), an arteriovenous fistula (n = 2), and occlusion (n = 1). Two angiograms were normal. Arterial repair was accomplished by interposition graft (n = 17), primary repair (n = 9), patch angioplasty (n = 3,) and ligation (n = 1). RESULTS: No functional deficits occurred in patients with an isolated vascular injury. Seven patients with associated brachial plexus injuries experienced severe disability. One arm of a patient was amputated. Two patients died. CONCLUSIONS: The use of angiography helps to confirm and localize injuries. Prompt correction of the vascular injury avoids disability resulting from ischemia. Although the amputation rate is low with vascular repair, the functional disability resulting from associated nerve injuries can be devastating.

Experimental repair of venous valvular insufficiency using a cryopreserved venous valve allograft aided by a distal arteriovenous fistula.

PURPOSE: To evaluate the patency and hemodynamic impact of a cryopreserved allograft venous valve transplanted to the superficial femoral vein (SFV) of a canine insufficiency model aided by a distal arteriovenous fistula (dAVF). METHODS: Eight greyhounds had intravenous hemodynamic parameters measured (venous filling time [VFT], 90% of venous refilling time [VRT90], and simulated ambulatory venous pressure [AVP]) before (T0) and after complete hindlimb venous valvulotomy (T1) to produce venous insufficiency. Simultaneously, a valve-containing vein segment was harvested from the opposite SFV or external jugular vein (n = 1) and cryopreserved. Three weeks later a blood type-matched cryopreserved valve was transplanted to the insufficient SFV aided by a low-flow (n = 4) or high-flow (n = 4) dAVF. The fistula was ligated in 3 to 6 weeks, and venous indexes (T2) were obtained 3 weeks later. Analysis of variances compared the venous indexes at T0, T1, and T2 for statistical significance. Gross and histologic inspection assessed valve integrity. RESULTS: Two valves aided by a low-flow dAVF exhibited thrombosis and scarring. The hemodynamics of the six remaining valves demonstrated normalization of the VRT90, an AVP consistent with insufficiency, and a VFT between normal and total venous insufficiency. The patent valves were normal on gross examination and by histologic examination with signs of normal external healing. CONCLUSIONS: A cryopreserved venous valve allograft transplanted to the SFV of an incompetent hindlimb partially corrects venous hemodynamics. A high-flow arteriovenous fistula most consistently preserves transplant patency.

Cerebral vascular reactivity assessed with acetazolamide single photon emission computer tomography scans before and after carotid endarterectomy.

BACKGROUND AND METHODS: In 64 patients, cerebral perfusion and vascular reactivity were assessed before and after carotid endarterectomy (CEA) using acetazolamide (ACZ)-enhanced single photon emission computer tomography (SPECT). Twenty-five patients were asymptomatic, whereas the remainder were symptomatic. Sixty-one patients had a > or = 70% ipsilateral internal carotid artery stenosis. RESULTS: Fifty SPECT scans revealed decreased vascular reactivity. Twenty-three showed infarcts. Fourteen patients had normal studies. Twenty of the SPECT scans of asymptomatic patients demonstrated poor vascular reactivity. After CEA, 39 patients had improved ipsilateral vasoreactivity. In 12 patients, contralateral improvement was also found. CONCLUSION: ACZ-enhanced SPECT scans, by assessing cerebral perfusion and vascular reactivity, may help to identify patients at risk of stroke should perfusion further diminish. Postoperative studies confirm improvement in vascular reactivity. ACZ-enhanced SPECT scans may provide objective evidence for the selection of patients with a high-grade asymptomatic carotid stenosis for CEA.

Neointimal hyperplasia following argon ion laser carotid endarterectomy: with and without endothelial cell seeding.

BACKGROUND AND OBJECTIVE: This study compares the development of neointimal hyperplasia following conventional and argon ion laser carotid endarterectomy and assesses the potential advantage of endothelial cell seeding. STUDY DESIGN/MATERIALS AND METHODS: Eight dogs underwent conventional endarterectomy in one carotid artery and an argon ion laser endarterectomy in the other. After 42 days, these arteries were harvested and the intimal thickness were compared. Six additional dogs underwent bilateral argon ion laser carotid endarterectomy with endothelial cell seeding on the one side only. These arteries were harvested after 65 days and their mean intimal thickness were compared. RESULTS: At 42 days, the mean intimal thickness in the conventional endarterectomy group was 0.070 +/- 0.007 mm; in the argon ion laser endarterectomy group it was 0.058 +/- 0.001 mm (P = 0.76, NS). At 65 days, the mean intimal thickness in the group without endothelial cell seeding was 0.125 +/- 0.003 mm vs. 0.061 +/- 0.001 mm on the seeded side (P = 0.043). CONCLUSION: Argon ion laser carotid endarterectomy results in no more neointimal hyperplasia than conventional endarterectomy. The neointimal hyperplasia is reduced by endothelial cell seeding.

Comparison of spiral CT scan and arteriography for evaluation of renal and visceral arteries.

Renal and visceral artery images obtained concurrently with spiral CT and conventional arteriography were compared for 32 patients. Indications for imaging were occlusive disease (n = 12), aneurysmal disease (n = 9), and renal or visceral artery disease (n = 11). Conventional arteriography enabled visualization of 64 renal arteries and 15 accessory renal arteries. Lateral aortograms obtained in 15 patients enabled visualization of 14 superior mesenteric (SMA) and 14 celiac arteries. Spiral CT enabled visualization of 60 renal arteries, 12 accessory renal arteries, 27 SMAs, and 22 celiac arteries. Calcification or a disparity in timing of contrast material injection and scanning prevented visualization of the celiac artery in 10 patients and the SMA in four patients. With conventional arteriography as the standard for comparison, spiral CT had a sensitivity of 67% and a specificity of 95% for depiction of at least 75% stenosis in the main renal artery. By means of the Pearson correlation coefficient, significant correlation (p < 0.001) was confirmed between spiral CT and arteriography for evaluation of stenosis of the main renal artery, SMA, and celiac artery. This early experience suggests that spiral CT may be useful in evaluation of renal and visceral arteries and their relationship to aortic disease.

Iliac artery evaluation with two-dimensional time-of-flight MR angiography: update.

PURPOSE: To determine the accuracy of two-dimensional (2D) time-of-flight (TOF) magnetic resonance (MR) angiography, performed with state-of-the-art equipment, in the evaluation of the iliac arteries. MATERIALS AND METHODS: The iliac arteries of 50 patients undergoing conventional arteriography for evaluation of lower-extremity ischemia or abdominal aortic aneurysm were also imaged with axial 2D TOF MR angiography. Blinded interpretations of conventional arteriograms and MR angiograms were compared. MR angiography was performed with a 1.5-T system with 2-mm contiguous axial sections, 60 degrees flip angle, 28-msec repetition time, and 7.4-msec echo time. RESULTS: Blinded interpretations of 2D TOF MR angiograms matched those of conventional arteriograms in 26 of 50 patients (52%). For the diagnosis of obstructive iliac artery lesions, sensitivity and specificity with MR angiography were 85% and 59%, respectively. Two of three saccular iliac artery aneurysms escaped detection with MR angiography. CONCLUSION: Iliac artery evaluation with axial 2D TOF MR angiography is not sufficiently accurate to warrant its use as a replacement for conventional arteriography in patients who lack contraindications to the latter.

Long-term follow-up of the Palmaz stent for iliac occlusive disease.

BACKGROUND: Thirty-eight limbs with iliac occlusive disease were treated with Palmaz stents from 1987 through 1991. METHODS: Indications for stent utilization included dissection induced by percutaneous transluminal balloon angioplasty (PTA) (10), restenosis after PTA (nine), post-PTA residual stenosis (nine), multiple stenoses or occlusion (five), and unfavorable location (five). RESULTS: The ankle/brachial pressure index increased from 0.53 +/- 0.27 to 0.8 +/- 0.26 after stent deployment. The intraluminal pressure gradient decreased from 31.9 +/- 16.3 to 0.9 +/- 2.2 mm Hg after stent deployment. Complications included pseudoaneurysm (one), arteriovenous fistula (one), iliac perforation (one), groin hematoma (two), and occlusion (two). Follow-up arteriogram showed stenosis proximal or distal (n = 4) or within the stents (n = 4). These were treated with PTA or stents. Two patients required an aortobifemoral graft. Nine patients have died. Life table analysis showed a 1-, 3-, and 5-year primary and secondary cumulative patency of 87% +/- 5.9%, 74% +/- 8.2%, and 63% +/- 10% and 91% +/- 5.1%, 91% +/- 5.6%, and 86% +/- 7.6%, respectively. CONCLUSIONS: Palmaz stents, often required to salvage a PTA failure, appear to maintain overall patency at a high level. However, intimal hyperplasia and the progression of atherosclerotic disease may result in a need for additional procedures to obtain this favorable outcome.

Iliac artery MR angiography: comparison of three-dimensional gadolinium-enhanced and two-dimensional time-of-flight techniques.

PURPOSE: To determine if a three-dimensional (3D) gadolinium-enhanced technique is useful for improving the accuracy of magnetic resonance (MR) angiography. MATERIALS AND METHODS: Prospectively, two-dimensional (2D) time-of-flight (TOF) and 3D gadolinium-enhanced MR angiography was performed in 23 patients. Conventional arteriography, the reference standard, was performed in all but five patients. Images from each examination were interpreted blindly by two observers. RESULTS: For common and external iliac arteries, interpretations of dynamic 3D gadolinium-enhanced MR angiograms matched those of conventional arteriograms in 12 of 18 patients (67%) for both observers. Interpretations of 2D TOF images matched those of conventional arteriograms in six (33%) and seven (39%) patients for the two observers respectively. CONCLUSION: Preliminary results suggest that iliac artery MR angiography can be improved by complementing standard 2D TOF acquisitions with 3D gadolinium-enhanced acquisitions.

A comparison of endovascular assisted and conventional in situ bypass grafts.

Thirty-three in situ saphenous vein bypass grafts were performed using a conventional open technique (CI) while 31 in situ bypass grafts were performed using endovascular occlusion of side branches (EAI). Bypass grafts were performed from the femoral to the popliteal (n = 37) or a trifurcation (n = 27) artery for claudication (n = 7), rest pain (n = 14), or tissue loss (n = 43). Wound complications developed in 11 Cl and four EAI limbs. Postoperative hospitalization in CI and EAI patients was, respectively, 8.4 +/- 2.0 days and 4 +/- 1.6 days. Missed arteriovenous fistulas were noted in one CI and 17 EAI limbs postoperatively. At follow-up four (12%) CI and six (19%) EAI grafts were occluded or had undergone revision surgery. Based on life-table analysis CI and EAI cumulative patency rates at 18 months were 79% and 83%, respectively. Although this new technique (EAI bypass grafting) did not reduce operative time, it did decrease the length of surgical incisions and the duration of postoperative hospitalization (p < 0.001, Student's t test). Wound complications occurred less frequently in EAI limbs but the incidence of missed arteriovenous fistulas was significantly higher. These data suggest that EAI and CI patency is comparable. Ultimately long-term patency will be the crucial test for determining the utility of this new technique.

Femorodistal vein grafts: the utility of graft surveillance criteria.

PURPOSE: This retrospective review of femorodistal vein grafts was analyzed to determine the usefulness of various graft surveillance criteria. METHOD: The surveillance schedule involved evaluations at 1 month, every 3 months the first year, and then every 6 months. Salvage intervention or graft occlusion occurring within the next follow-up interval defined surveillance end points. One hundred two grafts (329 surveillance visits) had an ankle/brachial index (ABI). A duplex scanning-determined midgraft peak systolic flow velocity (PSFV) was available for 81 grafts (262 visits). Forty-eight grafts (137 visits) had both a PSFV and entire graft duplex scanning (EGDS) to determine stenosis greater than 50%, whereas 40 grafts (91 visits) had simultaneous ABI and EGDS. RESULTS: When a greater than 15% decrease in ABI denoted an abnormal surveillance study result, a positive predictive value (PPV) of 24.3% and negative predictive value of 94.5% were noted. Similarly, a PSFV cutoff of less than 35 cm/sec demonstrated values of 26.3% and 94.2%, respectively. When an EGDS of greater than 50% stenosis or a PSFV of less than 35 cm/sec were the cutoff criteria, the PPV was 36.7% and negative predictive value 99.1%, whereas characterizing abnormal results further with ABI (> 15% decreases) increased the PPV to 83.3%. CONCLUSION: The combination of an EGDS, midgraft PSFV, and ABI provides optimal follow-up for our patients with a femorodistal vein graft.

Valve transplantation to the canine popliteal vein: the utility of a distal arteriovenous fistula and the hemodynamic result of a single functional valve.

PURPOSE: The objectives were to transplant a venous valve into the popliteal vein of a canine insufficiency model that would maintain long-term function and then to measure the hemodynamic benefit of such a valve. METHODS: Indexes of venous valvular function, including venous filling time, 90% of venous refilling time, and simulated ambulatory venous pressure, were measured in 11 greyhounds before intervention (Control) and after hind limb venous valvulotomy, which produces chronic venous insufficiency. Three weeks later a valve-containing segment of external jugular vein was transplanted to the popliteal vein with (n = 6) or without (n = 5) a distal arteriovenous fistula. The fistula was ligated in 4 to 6 weeks. Repeat venous indexes were obtained an average of 2 weeks after the last operative intervention. Gross anatomic inspection +/- duplex scanning performed before the animals were killed were used to distinguished normal from damaged valves. RESULTS: Four of five simple valve transplants demonstrated scarring and/or thrombosis and the average venous filling time, ambulatory venous pressure, and 90% of venous refilling time were consistent with chronic venous insufficiency and/or obstruction. One fistula failed in the six valve distal arteriovenous fistula transplants, and that valve was incompetent. Analysis of the remaining five functional valves demonstrated venous filling time consistent with chronic venous insufficiency, normalization of ambulatory venous pressure, and 90% of venous refilling time between normal, and chronic venous insufficiency. These valves appeared normal. CONCLUSIONS: Popliteal vein valve transplant anatomy is preserved by an adjunctive distal arteriovenous fistula, and a competent valve transplant improves postexercise venous hemodynamics to approach those of a normal limb.

Aortoiliac occlusive disease: a simple noninvasive method to determine significance.

This study was undertaken to develop a simple, noninvasive method for determining the hemodynamic significance of aortoiliac occlusive disease. In 14 dogs, 10-MHz Doppler recordings from the aorta and femoral arteries were made under normal (control) conditions and after construction of subcritical and critical iliac artery stenoses as documented by the papaverine test. The femoral waveforms obtained were analyzed for configuration (tri-, bi-, or monophasic) and pulsatility index. A ratio of time from ECG QRS peak to peak femoral flow divided by the time to peak aortic flow (TRFA) was calculated as was a ratio of the maximum rate of rise of the femoral Doppler signal to the aortic Doppler signal (MRRFA). These parameters were compared using Fisher's exact test and analysis of variance. TRFA and MRRFA accurately categorized the degree of iliac occlusive disease into statistically distinct groups. A TRFA less than 1.10 or a MRRFA greater than 0.70 identified a normal artery. MRRFA distinguished normal aortoiliac arteries from those with subcritical stenosis with a sensitivity of 0.93, a specificity of 1.00, and an accuracy of 0.96. TRFA distinguished normal aortoiliac arteries from those with a subcritical stenosis with a sensitivity of 0.86, a specificity of 0.86, and an accuracy of 0.86. Both TRFA and MRRFA separated normal aortoiliac arteries from those with a critical stenosis with a sensitivity of 1.00, a specificity of 1.00, and an accuracy of 1.00. The femoral Doppler waveform configuration and pulsatility index could not accurately categorize these differences (P = 0.100).

Early results of endovascular-assisted in situ saphenous vein bypass grafting.

PURPOSE: This study evaluated an endovascular technique for occlusion of arteriovenous fistula when performing saphenous vein in situ bypass grafting. METHODS: In 31 limbs femoropopliteal (17) or femorotibial (14) in situ bypass grafting was performed for claudication/aneurysm (4), rest pain (6), or tissue loss (21). A valvulotome was used for valve lysis. Saphenous vein branches were identified with angioscopy in 16 limbs or with fluoroscopy in the remainder. An electronically steerable endovascular catheter was used to deliver platinum coils into the venous tributaries to occlude them. RESULTS: The maximal number of coils placed in any limb was nine. Most operations were performed with only a groin incision (length = 9.8 +/- 1.6 cm) and a distal incision (length = 16.8 +/- 6.5 cm). Wound complications occurred in four limbs, whereas four limbs developed localized superficial thrombophlebitis. The postoperative ankle-brachial index increased to a mean of 0.91 +/- 0.12. Postoperative duplex imaging revealed a missed arteriovenous fistula in 12 limbs. Two were surgically ligated, whereas the remainder were embolized in the radiology suite. Postoperative length of hospitalization was 4 +/- 2 days in uncomplicated cases. Follow-up revealed five graft occlusions. Occlusion occurred at 12 hours, 2 weeks, and 6, 14, and 15 months after operation. Although there were no perioperative deaths, two patients have died of unrelated causes. CONCLUSION: This endovascular technique of arteriovenous fistula embolization decreased the length of the surgical wounds, and patients were discharged 4 +/- 2 days in uncomplicated cases. The ultimate test of its efficacy, however, will be long-term functional results.

Details of a canine venous insufficiency model.

Continued study of a chronic deep venous insufficiency (CDVI) model allows optimal comparison with the human condition. This study evaluates the model's long-term stability, its lack of observed clinical effect, and a simulated exercise study as a physiologic estimate of normal hindlimb walking. The time to maximal ankle venous pressure after standing (VFT), and to 90% of the venous refilling time after electrical stimulation, quadripedal, or hindlimb walking (VRT90), and the minimal pressure after exercise (AVP) were measured up to 10 months after CDVI model creation. The animals' intravenous resting pressure was obtained after standing stationary on all four limbs. Analysis of variance was used to determine statistical significance where indicated. VFT, AVP, and VRT90 measurements demonstrated values consistent with CDVI in animals studied up to 10 months after model creation and were statistically different from control limb values (p < or = .002, n = 8). Animals studied during quadripedal walking showed no difference in resting pressure, AVP, and VRT90 between model and control limbs (n = 5). There was no statistical difference in AVP or VRT90 measured under conditions of stimulated exercise or bipedal walking; and both conditions produced hemodynamic changes consistent with CDVI (n = 5). This animal model is a reliable long-term CDVI hemodynamic model. The normal venous hemodynamics recorded during quadripedic walking may explain the lack of clinical sequelae observed in this model. Lastly, the method of simulated exercise used in this study is a reliable test that reflects physiologic measurements obtained during bipedal walking.

A prospective, in vivo study of the relationship between blood flow hemodynamics and atherosclerosis in a hyperlipidemic swine model.

PURPOSE: This is the first prospective evaluation of the relationship between shear stress and atherosclerotic plaque formation in a pulsatile, in vivo model. METHODS: A 50% aortic stenosis was created in six Yucatan micropigs by placing a suture line across the left half of the aorta. A specially modified 20 MHz Doppler ultrasound probe mounted at a 45 degree angle on a micromanipulator was used to measure blood flow velocity along the lateral aspects of the aorta. Shear stress at these locations was calculated with linear regression. The pigs were then fed a specially formulated hyperlipidemic diet for 7 months, after which their aortas were harvested. Plaque thickness was measured and correlated with mean, maximum, and minimum shear stress at each site and side on or away from the aortic plication and distance from the level of the aortic plication with general linear modeling. RESULTS: Plaque thickness was negatively correlated with the mean shear stress during the cardiac cycle (correlation coefficient -0.3972 p = 0.0164) and with the maximum shear stress during the cardiac cycle (correlation coefficient -0.4581 p = 0.0050). Plaque thickness was correlated with a multivariate model of mean shear stress, maximum shear stress, and their interactive effects with other parameters in the model with a correlation coefficient of 0.6733, (p = 0.0048). CONCLUSIONS: This is the first prospective, in vivo study with a pulsatile, hyperlipidemic model, which clearly demonstrates that low shear stress is associated with plaque formation.