Multimodality Imaging of Lower Extremity Peripheral Arterial Disease

Abstract

HomeCirculation: Cardiovascular ImagingVol. 5, No. 6Multimodality Imaging of Lower Extremity Peripheral Arterial Disease Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBMultimodality Imaging of Lower Extremity Peripheral Arterial DiseaseCurrent Role and Future Directions Amy W. Pollak, MD, Patrick T. Norton, MD and Christopher M. Kramer, MD Amy W. PollakAmy W. Pollak From the Departments of Medicine (A.W.P., P.T.N., C.M.K) and Radiology (P.T.N., C.M.K.), and the Cardiovascular Imaging Center, University of Virginia Health System, Charlottesville, VA. Search for more papers by this author , Patrick T. NortonPatrick T. Norton From the Departments of Medicine (A.W.P., P.T.N., C.M.K) and Radiology (P.T.N., C.M.K.), and the Cardiovascular Imaging Center, University of Virginia Health System, Charlottesville, VA. Search for more papers by this author and Christopher M. KramerChristopher M. Kramer From the Departments of Medicine (A.W.P., P.T.N., C.M.K) and Radiology (P.T.N., C.M.K.), and the Cardiovascular Imaging Center, University of Virginia Health System, Charlottesville, VA. Search for more papers by this author Originally published1 Nov 2012https://doi.org/10.1161/CIRCIMAGING.111.970814Circulation: Cardiovascular Imaging. 2012;5:797–807IntroductionPatients with symptomatic peripheral arterial disease (PAD) affecting the lower extremities are initially evaluated with an ankle-brachial index (ABI) and segmental pressure measurements. An ABI<0.90 diagnoses significant lower extremity PAD (>50% stenosis) with a sensitivity and specificity of 79% and 96%, respectively.1 An ABI between 0.90 and 1.0 is considered borderline. The localization of the stenosis can be inferred by an abnormal decrease (>20 mm Hg) in segmental lower extremity pressures. In the presence of calcified atherosclerosis, arteries may become stiff and noncompressible, which results in a falsely elevated ABI (often >1.3). For patients with suspected PAD and a normal ABI at rest, it is valuable to obtain postexercise ABI measurements, which if <0.85 are consistent with PAD and is an independent predictor of mortality.2 Imaging is then needed to confirm the location and degree of stenosis before revascularization or if the diagnosis of PAD is uncertain.Characterization of PAD can be performed with noninvasive angiography using computed tomography (CTA) or magnetic resonance angiography (MRA), as well as with duplex ultrasonography (US), depending on patient specific characteristics (Table 1). Advances in both CTA and MRA provide clinicians with the opportunity to obtain a high resolution, 3-dimensional (3-D) road map of the peripheral arterial tree in patients, particularly when planning revascularization strategies. Invasive digital subtraction angiography (DSA) has been the accepted standard for evaluation of lower extremity atherosclerosis. Although DSA is a robust technique for diagnosing significant arterial stenosis or obstruction, it provides a 2-D view of the vessels, which may underestimate the degree of stenosis for tortuous vessels.3 Furthermore, there are inherent risks with arterial access, ionizing radiation, and the use of iodinated contrast media (CM).4Table 1. Comparison Between CTA, MRA, and Duplex US for Diagnosis of PADImaging ModalityStrengthsLimitationsCTARapid acquisitionHeavily calcified vesselsHigh spatial resolutionIn-stent restenosis visualization better than MRAIodinated contrastFaster scan time than MRAIonizing radiation exposureMRABoth gadolinium enhanced and noncontrast approaches availableIn-stent restenosis limited by metal artifact depending on stent type Risk of NSF in patients with advanced renal disease who receive gadoliniumNeither iodinated contrast nor ionizing radiation are usedDuplex USEasily accessibleLess accurate in aortoiliac region because of patient obesity or intestinal gasInexpensiveNo radiationLimited sensitivity for multilevel stenosisNo contrastHeavily calcified vesselsCTA indicates computed tomography angiography; MRA, magnetic resonance angiography; US, ultrasonography; NSF, nephrogenic systemic fibrosis; and PAD, peripheral arterial disease.This review aims to deliver a comprehensive, yet concise review of noninvasive imaging modalities in the diagnosis of lower extremity PAD (defined as involving the aorta, iliacs, and infrainguinal arteries). Furthermore, we will integrate the available evidence and society guidelines, to provide an informed clinical perspective regarding the role of the imaging modalities in patient care. Briefly, future directions in computed tomography (CT) and magnetic resonance imaging (MRI) will be discussed as they pertain to clinical vascular medicine.Computed Tomography AngiographyBackgroundInitially, CTA was limited by the use of single slice scanners that imaged a fraction of the arterial tree at a given time. With the advent of multidetector row CT, images of inflow and runoff vessels became possible using a single acquisition and CM injection. Multidetector row CT has progressed rapidly from 4-detector row scanners to 64-detector rows or more, resulting in increased speed of scanning and greater longitudinal field of view while maintaining near-isotropic imaging voxels. Minimum voxel dimension of 0.4 to 0.6 mm in the z-axis are currently achievable with 64-detector row scanners, optimizing visualization of smaller, distal arteries.Image interpretation is aided by the use of advanced postprocessing techniques. Volume rendered 3-D reconstruction of the arterial tree provides a global overview for the rapid identification of pathology. Maximum intensity projection (MIP) images provide similar views to traditional angiography and are useful for qualitatively assessing the degree of stenosis. The usefulness of both volume rendered and MIP are limited by obscuration of the vessels by bones. Thus, automated software techniques for bone removal have been implemented; however, even the best algorithms require some degree of manual correction. Either automated or interpreter generated centerlines in the vessel of interest can be placed to obtain curved planar reformations, which gives both longitudinal and cross-sectional views of the vessel, which are useful for quantitative measurements; however, this is done at the cost of distortion of nearby anatomic landmarks. Distortion-free detailed interrogation is performed with thin section multiplanar reformats, which can be oriented to any desired plane. Axial 2-D source data are used when other techniques are inconclusive.Vessels with dense calcifications have high attenuation, which may lead to blooming artifact and resultant overestimation of stenosis. To minimize this, a window level of 200 Hounsfield units and width of 1000 Hounsfield units are recommended5 for image analysis. In the setting of heavily calcified vessels or metallic stents, image reconstruction using a sharp kernel (B46) can reduce blooming artifact, although this results in increased background image noise.6 The true lumen in calcified arterial vessels is often best analyzed using a combination of reconstruction techniques and source images. Additionally, dual-energy CTA postprocessing to subtract voxels containing calcium in the vessel wall is a promising technique for evaluating calcified vessels.Technical Considerations for CTA Image AcquisitionImaging of the lower extremity peripheral vasculature is performed from the aortic diaphragmatic hiatus through the toes, to evaluate the entire inflow and runoff (Figure 1). In the upper abdomen, visualization of the splanchnic and renal vasculature relies on good breath holding to prevent obscuration of the vessel by motion artifact. Respiration does not affect imaging beyond the pelvis.Download figureDownload PowerPointFigure 1. Computed tomography angiography of aortoiliac occlusive disease. Bone segmentation was achieved using automated algorithms with manual correction to produce full-volume maximum intensity projection (MIP) image after bone removal (A) and 3-dimensional volume rendered (VR) image with bones included at a partial transparency (B). Aortic occlusion is annotated on the MIP image by the asterisk directly inferior to the renal arteries. Occlusion of the bilateral common iliac arteries and stents, as well as the external iliac arteries, is also shown on both images. Reconstitution of runoff flow occurs at the superficial femoral arteries bilaterally. Collateral pathways are easily visualized on the VR images. The Arc of Riolan (arrow) supplies the inferior mesenteric artery (IMA) territory from the superior mesenteric artery. The superior rectal artery (arrowhead) supplies the internal iliac arteries (black arrows) from the IMA. The internal iliac artery collateralizes with the circumflex femoral arteries supply the runoff vessels.CTA requires the use of iodinated CM for visualization of the vasculature. Injection rates vary depending on protocol, but typically are in the range of 4 to 6 mL/s delivered intravenously by a dual channel power injector. The amount of iodinated media will vary depending upon the patient and protocol used, with ≈100 to 120 mL used for an abdominal CTA with runoffs.5 However, this dose could be reduced to 80 mL in patients with a small body mass index. A saline chaser injected at the completion of the CM media injection results in greater enhancement of the vasculature and a reduction in the CM dose.To maximize the arterial enhancement during the examination, both CM transit time and speed of acquisition need to be considered. CM transit time is the time from the beginning of CM injection to the arrival of CM in the territory of interest, and is derived by either real-time enhancement monitoring (bolus tracking) during the injection of full dose of CM or by a separate smaller test bolus performed before the exam. Historically, with slower scanners, the fastest acquisition speed was chosen and the duration of CM injection was set to the scan time with scan initiation determined by CM transit time. Because of the rapid acquisition of modern scanners, this approach results in under opacification of the peripheral vasculature as a result of insufficient time to fill the vascular tree. Adding 3 to 5 seconds to the CM transit time, reducing the pitch to a moderate value (typically around 0.9) and increasing the CM volume lengthens the scan duration allowing for better filling of the vasculature. However, pathology such as large aneurysms and high-grade stenosis can delay the transit of CM to the distal vasculature, resulting in scanning before its arrival. A second late scan of the calf should be preplanned in the event of inadequate pedal opacification.A new strategy put forth by Fleischmann7 is to fix the duration of scanning the lower extremity vascular to 40 seconds by adjusting the pitch. The volume and rate of CM injection is based on weight and delivered as a bi-phasic injection to provide sustained opacification of the arterial system.Many patients with PAD undergo multiple imaging studies during their lifetime; thus, reducing exposition to radiation is paramount in this population. Reducing tube voltage results in better attenuation of CM because of the energy of the photons being closer to the k-edge of iodine, and reduction in effective dose to the patient. A decrease from 120 to 100 kVp results in a 34% reduction in radiation dose without affecting diagnostic image quality.8 More aggressive dose reduction from 100 to 80 kVp maintains diagnostic image quality with an additional 23% dose savings.9 However, in larger patients, lower kVp protocols are limited by noise. The newest current generation scanners offer an additional dose reduction of 50% while maintaining image quality because of advances in image processing.10 Image domain-based iterative reconstruction algorithms, as compared with filtered back projection image reconstruction of prior generations scanners, reduce noise and maintain image sharpness through modeling techniques made possible by improvement in computer processing power. Alternatively, with a similar radiation dose, iterative reconstruction can be used to increase image sharpness without affecting noise levels, which preliminary studies suggest may improve visualization of the lumen of small stents or calcified vessels.11,12Magnetic Resonance AngiographyCurrently, high performance MR scanners provide striking angiographic images without patient exposure to radiation or the complexity of removing overlying bone from 3-D reconstructed images. MRA can be performed using contrast-enhanced (CE-MRA) approach or noncontrast-enhanced, flow-sensitive techniques that take advantage of the difference in signal properties between static tissue and flowing blood. These noncontrast MRA techniques include time-of-flight (TOF) MRA, ECG-gated partial-Fourier fast spin echo (FSE), and steady-state free precession (SSFP). In addition, phase-contrast MRA shows arterial blood flow direction and velocity.Contrast-Enhanced MRACE-MRA is the workhorse of clinical MRA because of the robustness of the technique (Figure 2). CE-MRA takes advantage of the increase in arterial signal after administration of intravenous paramagnetic gadolinium (Gd)-based contrast agents because of the shortening of T1 relaxation in blood. Fast T1-weighted gradient-echo sequences with low flip angles are used in combination with an intravenous bolus injection, allowing for rapid acquisition of 3-D volumes with high signal-to-noise ratio of vessel to background. Visualization of the vasculature is further enhanced by mask subtraction to remove the background signal, similar to DSA. MRA interpretation relies heavily on full-volume MIP images created from the subtracted datasets, which provide a global prospective. Detailed image analysis is done using subvolume MIPs and multiplanar reformats.Download figureDownload PowerPointFigure 2. Multistation contrast-enhanced magnetic resonance angiography of the abdomen, pelvis, and lower extremities was performed using stepping table acquisition and bi-phasic contrast media injection protocol. Imaging reveals a patent aorto-biiliac surgical bypass graft (extent defined by arrowheads), with bilateral runoff disease consisting of tandem stenoses of the left superficial femoral artery (SFA), punctuated with a focal high-grade stenosis and collateral formation (white arrow) and tandem stenoses with long segment occlusion of the right SFA and popliteal artery with collaterals supplied by the profunda femoral artery (black arrows).In MR, images are derived from Fourier transformation of radiofrequency signals that are acquired during an imaging sequence. K-space is the raw image data as represented in frequency-time domain and is built up in a sequential, step-wise fashion. The data located at the center of k-space contribute most to the contrast agent within an image. Thus, for CE-MRA, acquisition of the center of k-space needs to occur during peak vessel enhancement. The timing of the scan initiation is determined from a test bolus or using bolus-tracking methods, similar to CTA.There are 7 Gd-based contrast agents currently approved by the Food and Drug Administration in the United States, and 2 additional agents used in the European Union (Table 2). They are paramagnetic and classified on the basis of protein binding, linear or cyclic structure, and ionicity.13 In patients with normal renal function, the half-life of Gd-based contrast agents are ≈1.5 hours. All Gd-based contrast agents have been associated with the rare development of nephrogenic systemic fibrosis; however, the majority of cases involve the linear agents: Gadopentetate dimeglumine (Magnevist), Gadodiamide (Omniscan), and Gadoversetamide (Optimark). The current recommendation is to avoid any Gd-based contrast agent in patients with a glomerular filtration rate <30 mL/min per 1.73 m2 given risk of nephrogenic systemic fibrosis.14Table 2. GBCA Used in Contrast-Enhanced MRAGBCAExcretionProtein BindingStructureChargeGadopentetate dimeglumine (Magnevist)KidneysNoneLinearIonicGadobenate disodium (Multihance)Primarily kidneys (small amount in feces)Weak bindingLinearIonicGadodiamide (Omniscan)KidneysNoneLinearNonionicGadoversetamide (Optimark)KidneysNoneLinearNonionicGadoteridol (Prohance)KidneysNoneCyclicNonionicGadoxetate disodium (Eovist)Kidney and feces (enters liver cells)NoneLinearIonicGadofosveset trisodium (Ablavar)KidneysStrong-reversible bindingLinearIonicMRA indicates magnetic resonance angiography; and GBCA, gadolinium-based contrast agent.The older Gd-based contrast agents, move quickly from the vasculature to the interstitial space, which results in limited time for imaging with the maximal spatial resolution. However, a new blood pool contrast agent, gadofosveset trisodium, was recently approved by the Food and Drug Administration for 3-D MRA.15 This agent reversibly binds to albumin, increasing T1 relaxivity in a steady state, which lasts for about an hour. In a small study of 27 patients16 with 334 arterial segments, imaging with 0.03 mmol/kg of gadofosveset trisodium resulted in 100% agreement between degree of stenosis graded with steady-state MRA and DSA.Multistation MRA can be performed to image the aorta and peripheral runoff arteries using the same bolus, typically ≈0.2 mmol/kg body weight, in a bolus-chase technique, wherein the aortoiliac, femoropopliteal, and tibial regions are sequentially imaged. Multistation MRA benefits from a bi-phasic injection protocol with a high initial rate of injection that is reduced for the lower stations, resulting in increased arterial opacification of the distal station. The routine use of parallel imaging, requiring phased array peripheral coils, has reduced typical breath holds requirements for the aortoiliac station from 15 to 20 seconds or less without significant loss of image quality.Despite advances in acquisition speed, venous contamination of the tibial station can be problematic, particularly in the setting of critical limb ischemia, where there is a smaller time difference between arterial and venous enhancement. There are several options for addressing venous contamination,17 including thigh compression cuffs or a semicircular foam pillow behind the knee to impede lower extremity venous return. Another option is to image the calf vessels first with a separate contrast bolus (which eliminates the possibility of venous contamination). However, imaging the tibial station first using time-resolved MRA is the approach most commonly used in practice. Time-resolved MRA uses k-space undersampling with view-sharing to acquire multiple 3-D MRA datasets rapidly every 1 to 2 seconds during the injection of a small amount of contrast agent. Ninety seconds of imaging from the start of injection is sufficient for the calf in most cases. This dataset is subtracted from a mask image and is presented as a cine of MIP images, replicating the appearance of DSA and thus allowing form separation of arterial and venous phases.The use of higher field strength MRI, such as 3.0 Tesla offer a higher signal-to-noise ratio, potential for lower doses of Gd and improved spatial resolution compared with 1.5 Tesla. In a small study of MRA at 3 Tesla, there was excellent agreement between degree of stenosis seen with MRA and conventional angiography,18 and there was nearly isotropic submillimeter 3-D voxels.Noncontrast-Enhanced MRANoncontrast-enhanced MRA (Figure 3) was initially described by Wedeen et al19 and further developed by Miyazaki et al.20 The first such technique was 2-D TOF, a gradient-echo technique that depends on flow-related enhancement. There are several challenges to 2-D TOF, including inadequate signal for vessels deep in the abdomen, flow-related artifacts seen in areas of stenosis, difficulty in obtaining perfectly perpendicular imaging planes, and a lengthy scan time. The diagnostic performance of CE-MRA is >2-D TOF.21 However, 2-D TOF is useful for evaluation of tibial and pedal arterial stenosis or occlusion, particularily if inadequately visualized by contrast angiography and is as accurate as DSA for these vessels.22Download figureDownload PowerPointFigure 3. Noncontrast magnetic resonance angiography (MRA). A, Two-dimensional (2-D) time-of-flight (TOF) MRA is acquired as multiple 2-D slices ideally oriented perpendicular to the vessel. The imaging slice is initially nulled and non-nulled blood is allowed to enter the slice at which time readout is performed. A selective saturation band is placed inferior to the imaging slice to suppress venous flow. Because of the 2-D, slice-by-slice nature of the technique, movement occurring between slice acquisitions appears as step artifact on coronal maximum intensity projection reconstruction. Two-dimensional TOF is susceptible to motion because of the long imaging times, as seen in this patient. B, Three-dimensional fast spin echo at 3 Tesla was acquired in 3 stations, revealing excellent imaging characteristics in this normal volunteer.There has been renewed interest in noncontrast-enhanced MRA given concerns of nephrogenic systemic fibrosis secondary to Gd in patients with advanced renal disease. There are several new techniques23: (1) ECG-gated partial-Fourier FSE, (2) 3-D FSE technique with sampling perfection with application of optimized contrasts using different flip angle evolution (SPACE),24 (3) balanced SSFP with arterial spin labeling, and (4) quiescent-interval single shot MRA using 2-D SSFP.25ECG-gated 3-D FSE exploits the difference in arterial and venous flow velocities during the cardiac cycle. During diastole arteries have lower blood flow, resulting in high signal intensity on T2-W images, whereas in systole, the higher arterial blood flow leads to low signal intensity. Venous blood is bright throughout the cardiac cycle because of the relatively low resting flow. The MRA is then obtained by subtracting systolic from diastolic images. In a small study at 1.5 Tesla,26 ECG-gated 3-D FSE was compared against CE-MRA and found to have a sensitivity and specificity of 85% and 76%, respectively, with an overestimation of stenosis in 22%. The use of parallel imaging and higher field strength may improve acquisition time and image quality.3-D FSE with SPACE24 uses a noncontrast, T1-W dark blood whole body technique with a high degree of reliability in normal volunteers. It has improved k-space sampling efficiency compared with traditional 3-D FSE. Additional studies are needed to evaluate the test characteristics of this technique in patients with PAD.Balanced SSFP with arterial spin labeling results in bright blood images in a single acquisition by first using a nonselective tag pulse to invert the magnetization for the entire imaging plane.23 Then a spatially selective tagging pulse is applied upstream of the arteries of interest and the tagged blood flows into the desired imaging region. In vessels with slow blood flow, the time needed to replace the tagged blood in the imaging plane can approach the T1 recovery time of blood, whereby the tagging effect would be lost.Quiescent-interval single shot MRA using a modified single shot 2-D balanced SSFP pulse sequence25 has been evaluated in a pilot study of normal volunteers and subjects with PAD and can accurately depict peripheral vascular stenosis compared with CE-MRA with a sensitivity and specificity of 87% and 95%, respectively. Quiescent-interval single shot MRA does not use subtractive nonenhanced MRA methods. The sequence requires little optimization for an individual patient, and the peripheral MRA can be performed in <10 minutes. A 2-center study compared CE-MRA and quiescent-interval single shot MRA in 53 patients with known or suspected PAD27 and found a sensitivity and specificity of 85% to 90% and 95% to 97%, respectively, for the detection of significant stenosis.Phase-contrast MRI shows a change in intravascular signal as arterial blood flow velocity deviates from the expected velocity encoded into the pulse sequence. Phase-contrast MRI has a sensitivity and specificity >90% for detection of clinically relevant stenosis (>75%) compared with DSA.28 However, Phase-contrast MRI is typically used only as an adjunct for determining degree of stenosis because of the long image acquisition time.23Duplex USDuplex US uses the combination of gray scale (B-mode) for vessel morphology and color pulsed wave Doppler techniques. It is safe, inexpensive, and can provide functional information about vessel stenosis.4 The primary criteria of intra-arterial peak systolic velocity (PSV) and ratio of PSV between the site of stenosis and adjacent normal vessel are used to gauge the degree of stenosis (Figure 4). A PSV ratio of >2.5 suggests a stenosis of 50% to 74% and if the end-diastolic velocity is >60 cm/s, this suggests a tighter stenosis, in the range of 75% to 99%.29 However, the most commonly used criteria is PSV ratio >2.0, along with PSV>200 cm/s, and aliasing and spectral broadening seen with color Doppler to diagnose a stenosis >50%.30 The addition of color flow imaging improves the diagnostic performance of duplex US for aortoiliac and femoropopliteal arteries.31 If no color Doppler signal is visible in the vessel, arterial occlusion is suggested (Figure 5). The evaluation of infrapopliteal vessels although feasible by US is time consuming and may be technically challenging. Thus, a complete evaluation of the native lower extremity arterial system may not be feasible in many patients.Download figureDownload PowerPointFigure 4. Duplex ultrasonography showing in-stent restenosis of a right superficial femoral artery (SFA) stent (for reference, the proximal SFA velocity is 96 cm/s, giving a peak systolic velocity [PSV] ratio of 278/96 cm/s, or 2.9). EDV indicates end-diastolic velocity; and RI, resistance index.Download figureDownload PowerPointFigure 5. Duplex ultrasonography showing an occluded superficial femoral artery (SFA) stent without any evidence of flow on color or pulsed wave Doppler.For evaluation with duplex US, patient body habitus may limit evaluation of the aortoiliac vessels because of abdominal girth or intestinal gas. The quality of a duplex US can vary depending on the examiner’s experience, and there is only moderate interobserver agreement for duplex evaluation of the aortoiliac and femoropopliteal areas.32 In the presence of sequential stenosis, US is less sensitive for detection of further downstream lesions.Clinical ApplicationsThere are several important considerations when choosing a noninvasive imaging modality to characterize lower extremity PAD: (1) imaging time, (2) need for evaluation of nearby anatomy, such as renovascular disease, (3) whether the indication is for native vessel disease or follow-up after a revascularization (bypass or stenting), and (4) patient specific factors, such as kidney disease, diabetes mellitus, or the presence of implantable metal devices. There are 3 sets of clinical guidelines which apply to the use of noninvasive imaging with lower extremity PAD: the European Society of Cardiology, American Heart Association/American College of Cardiology and Trans-Atlantic Inter-Society Consensus (TASC) II.Rapid image acquisition is a strength of CTA. CTA studies can be performed with scan times <5 minutes compared with MRA (≈20–30 minutes) or duplex US (30–45 minutes). If both legs are evaluated by duplex US, the scan times are quite long (≈1 hour) and may hinder patient flow in the vascular laboratory.Native Vessel DiseaseThe TASC II guidelines4 and European Society of Cardiology33 recommend obtaining a MRA, or CTA, or duplex US for imaging of native vessel disease, depending upon local availability, cost, and experience. American Heart Association/American College of Cardiology30 guidelines support the use of CTA and MRA in the diagnosis of the anatomic location and presence of significant stenosis in patients with native vessel lower extremity PAD. Imaging with either CTA or MRA is well suited for the diagnosis of patients with lower extremity PAD, who have suspected aortic aneurysm/dissection, renovascular disease, trauma, or acute limb ischemia.When choosing an imaging modality for native vessel disease, clinicians should consider whether the patient has underlying renal insufficiency, diabetes mellitus, or implanted metal devices. For patients who have normal renal function and are not diabetic, initial evaluation with either CTA or MRA is reasonable based on their similar excellent diagnostic capabilities. Overall, duplex US is a less sensitive technique for imaging native vessel stenosis compared with CTA or MRA; however, the greatest limitation of duplex US is the time required for evaluation of 2 lower extremities. A brief review of the important recent trials supporting the recommendation for either CTA or MRA in the diagnosis of native vessel disease follows.A recent systematic review of the diagnostic performance of CTA included 20 studies using DSA as the accepted standard.34 This analysis included studies which used 2 to 4 detector row CT scanners as well as the contemporary 16 to 64 detector rows. In this population of nearly 1000 patients, 68% had symptomatic PAD and sensitivity of CTA for stenosis >50% or occlusion was 95% (95% CI, 92%–97%) with a specificity of 96% (95% CI, 93%–97%). Regardless of the location, excellent test characteristics were seen in the aortoiliac (sensitivity 96%, specificity 98%), femoropopliteal (sensitivity 97%, specificity 94%), and tibial arteries (sensitivity 95%, specificity 91%). Traditionally, the diagnostic performance of CTA in tibial disease is lower compared with the aorta-iliac and femoral levels, particularly in the setting of heav