extremity arterial Accuracy of lower duplex mapping

Transcription

1 Accuracy of lower duplex mapping extremity arterial Gregory L. Moneta, MD, Richard A. Yeager, MD, Ruza Antonovic, MD, Lee Do Hall, MD, John D. Caster, RN, RVT, Cary A. Cummings, RN, RVT, and John M. Porter, MD, Portland, Ore. We performed lower extremity arterial duplex mapping from the aortic bifurcation to the ankle in 150 consecutive patients evaluated for aortic and lower extremity arterial reconstruction and compared lower extremity arterial duplex mapping in a blinded fashion to angiography. On the basis of history, physical examination, and four-cuff segmental Doppler pressures individual lower extremities were classified as normal, isolated aortoiliac disease, infrainguinal disease, and multilevel inflow and outflow disease. For vessels proximal to the tibial arteries, lower extremity arterial duplex mapping was analyzed for its ability to insonate individual arterial segments, detect a 50% or greater stenosis, and distinguish stenosis from occlusion. In the tibial arteries lower extremity aj~erial duplex mapping was evaluated for its ability to visualize tibial vessels and to predict interruption of tibial artery patency from origin to ankle. Lower extremity arterial duplex mapping visualized 99% of arterial segments proximal to the tibial vessels, with overall sensitivities for detecting a 50% or greater lesion ranging from 89% in the iliac vessels to 67% at the popliteal artery. Stenosis was successfully distinguished from occlusion in 98% of cases. In the tibial vessels lower extremity arterial duplex mapping was better at visualizing anterior tibial and posterior tibial artery segments (94% and 96%) than peroneal artery segments (83%), (p < 0.001). Overall sensitivities for predicting interruption of tibial artery patency were 90% for the anterior tibial, 90% for the posterior tibial, and 82% for the peroneal Clinical disease category did not influence in a major way the accuracy of lower extremity arterial duplex mapping in either above-knee or bdow-knee vessels. (J VASC St3RG 1992;15: ) Angiography is generally considered to be essential to provide detailed information necessary to formulate optimal plans for operative and/or endovascular treatment of lower extremity atherosclerotic occlusive disease. Although duplex scanning of lower extremity arteries has been proposed as an alternative to angiography, the technique has been systematically evaluated in only a small number of patients in a few centers. Many important questions remain unanswered concerning the potential clinical usefulness of lower extremity arterial duplex mapping (LEADM). These include the technical feasibility of From the Departments of Surgery and Radiology, Oregon Health Sciences University, and Veterans Affairs Medical Center, Portland. Supported by a grant from the Research Advisory Group, Department of Veterans Affairs. Presented at the Forty-fifth Annual Meeting of the Society for Vascular Surgery, Boston, Mass., June 4-5, Reprint requests: Gregory L. Moneta, MD, Department of Surgery, Division of Vascular Surgery (OP11), Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Rd., Portland, OR /6/33162 the examination, that is, how often can the various lower extremity arteries actually be insonated with sufficient clarity to permit the acquisition of clinically relevant information? Is it possible to reliably distinguish occlusion from high-grade stenosis? Will the technique be more accurate in patients with mild rather than advanced arterial occlusive disease? Finally what is the role if any of duplex scanning in the evaluation of the tibial arteries? Can the technique be used to identify tibial artery disease and thereby predict the presence of arterial continuity or discontinuity from the level of the popliteal artery to the anne? This report describes a blinded, prospective study designed to address these and other questions relevant to the possibility of eventually using LEADM as an adjunct to, or possibly a replacement for, angiography in the planning of operative and catheter-based procedures in the treatment of lower extremity arterial occlusive disease. PATIENTS AND METHODS All patients admitted over an 8-month period to the vascular surgery service of the Portland, Oregon, 275

2 276 Moneta Journal of VASCULAR SURGERY Table I. Criteria for clinical classification of lower extremity arterial occlusive disease in individual lower extremities Group Clinical disease pattern Criteria A No significant arterial Palpable pedal pulse; occlusive disease ABI* -> 0.9 B Aortoiliac (inflow) dis- Diminished groin ease only pulse; HTI** < 0.9; HTI _<ABI C Infrainguinal (outflow) Normal groin pulse; disease onlyf HTI > 0.9 ABI < 0.9 D Multilevel inflow and Diminished groin outflow diseasef pulse; HTI < 0.9 HTI _>ABI *ABI, Ankle-brachial index (highest ankle Doppler derived systolic pressure/highest brachial artery Doppler derived systolic pressure). ~*HTI, High thigh index. (high-thigh Doppler derived systolic pressure/highest brachial artery Doppler derived systolic pressure). fall patients with incompressible tibial arteries were classified as group C or D depending on inflow status. Veterans Affairs Hospital for elective treatment of infrarenal arterial disease were eligible for the study. All such patients underwent routine history and physical examinations as well as standard four-cuff lower extremity segmental Doppler pressures, in the manner described by Bridges and Barnes) All segmental Doppler pressures were performed by one of two registered vascular technologists. With use of the classification system outlined in Table I, each patient's lower extremities were classified by an attending vascular surgeon as being hemodynamically normal, (i.e., no apparent significant occlusive disease), or as having aortoiliac occlusive disease, femoropopliteal-tibial occlusive disease, or multilevel inflow and outflow occlusive disease. Angiography Patients with appropriate indications for lower extremity arterial reconstruction (including those with infrarenal aortic aneurysms) underwent standard diagnostic angiography, with use of transfcmoral, transaxillary, and, on occasion, ttanslumbar approaches as appropriate. Both digital subtraction, biplanar, and standard cut-film techniques were used. In patients with aortic aneurysms and normal segmental Doppler pressures, angiographic studies were frequently limited to the aorta, iliac artery, common femoral artery, and the proximal deep femoral artery and superficial femoral artery (SFA). Patients with significant renal insufficiency and unilateral lower extremity arterial symptoms often underwent selective studies of all or part of their symptomatic extremity only. All other patients underwent standard aortography with bilateral visualization of run-off vessels to the ankle and foot. All angiograms were read by one of two attending radiologists (R.A. and L.D.H.) or an attending vascular surgeon (G.L.M.) without knowledge of the results of any duplex studies. Individual arterial lesions in the common iliac, external iliac, and common femoral artery, and the proximal, middle, and distal segments of the SFA, above-knee and below-knee popliteal artery segments, and the origin of the deep femoral artery were classified as occlusion, 0% to 49% stenosis, or 50% to 99% stenosis. For nonoccluding lesions the degree of stenosis for individual lesions was based on the width of the contrast column immediately proximal to the lesion. When necessary, caliper measurements were used to classify the lesion into the appropriate category. Each infrapopliteal vessel was evaluated for angiographic occlusion in its proximal, middle, and distal segment. lnfrapopliteal arteries were then assessed for whether they were continuously patent or occluded at some point from their origin to the ankle. In this study luminal irregularity of individual arterial segments was not considered. Therefore vessels without significant focal stenosis, even if obviously diseased, were classified as less than 50% stenosis. In a similar manner, for purposes of data analysis, infrapopliteal arteries with subtotal occlusive lesions were considered patent as long as there was not total interruption of the contrast column from the tibial or peroneal artery origin to the ankle. No attempt was made in this study to grade individual infrapopliteal arteries or segments beyond the binary classification system of patent or occluded. Duplex examinations During the period of the study, 150 consecutive patients who underwent lower extremity angiography in preparation for elective infrarenal arterial reconstruction also underwent LEADM. Patients gave informed consent for the duplex examination, and the study was approved by the Human Subjects Committee of the Portland, Oregon, Veterans Affairs Hospital. All duplex examinations were performed by one of two registered vascular technologists with extensive experience in duplex techniques. Both technologists underwent a preparatory 4-month training period specifically devoted to LEADM before the beginning of formal data acquisition. This training period consisted of daily

3 Volume 15 Number 2 February I992 Lower extremity arterial duplex mapping 277 LEADM examinations of hospitalized patients undergoing vascular surgery, with periodic angiographic review in consultation with the physiciandirector of the study (G.L.M.). Lower extremity arterial duplex mapping was performed after an overnight fast and within 5 days before or after the angiogram and before any surgical or interventional radiologic procedure. With the exception of the popliteal artery, which was examined with the patient in a lateral or prone position, studies were conducted with subjects supine and the head of the bed elevated approximately 20 to 30 degrees. All examinations were performed with an Acuson-128 (Acuson ]inc., Mountain View, Calif.) color-flow duplex scanner. Beginning at the aortic bifurcation, 2 or 3 MHz probes were used to examine the common and external iliac vessels. A 5 MHz transducer was used for examining the infrainguinal arteries. In vessels proximal to the tibial vessels color flow was used I:o initially identify the vessel. Although a marked shift, in the color spectrum may suggest the presence of an arterial lesion, definitive determination of stenosis in this study was based on previously published duplex velocity criteria for lower extremity arterial stenosis. In particular, a focal increase in peaksystolic velocity (PSV) greater than 100% of that in the immediate preceeding arterial segment combined with loss of a previously present end-systolic reverse flow component was considered indicative of a 50% or greater stenosis at that site when the angle of insonation was 70 degrees or less (Fig. 1). 2 In addition, in the common iliac arteries any velocity recording with a PSV exceeding 200 cm/sec at an angle of insonation less than 70% was interpreted as indicating the presence of a 50% or greater stenosis. 3 Data from vessels requiring Doppler angles greater than 70 degrees were considered unreliable, and such vessels were classified as nonvisualized. 4's Therefore with the exception of an isolated peak-systolic iliac velocity greater than 200 cm/sec, no specific velocity cutoffs were used to grade individual arterial lesions, from the popliteal artery cephald, as less than or greater than 50% stenosis. The identification of stenotic lesions was dependent on meticulous pointby-point examination of the vessel with the Doppler portion of the duplex scanner in an effort to identify significant step-ups in PSV. The highest PSVs originating from the common and external iliac artery, the common femoral artery, origin of the deep femoral artery, proximal, middle and distal thirds of the SFA, and the above-knee and below-knee popliteal arteries were recorded. In any area suspicious for stenosis, PSVs were also recorded immediately before, within, and just distal to the stenosis. A vessel was considered occluded when it was identified with the B-mode component of the duplex scanner but failed to fill with color on color-flow examination and had no flow signal when insonated with the pulsed Doppler. Final interpretation of the duplex studies was performed by the vascular technologists in conjunction with the physician-director of the study. Duplex examinations were performed and interpreted without knowledge of the patients angiographic studies. Infrapopliteal arteries were initially examined with color-flow imaging beginning at the level of the infrageniculate popliteal artery and proceeding distally (Fig. 2). Occasionally, in patients with large legs or popliteal artery occlusions, the examination was facilitated by starting more distally in the leg and moving cephald. On duplex scanning an infrapopliteal vessel was considered occluded if at the point of examination no color flow was present within the vessel and no pulsatile arterial flow was present with the pulsed Doppler. Inffapopliteal arteries were classified as continuously patent when flow could be demonstrated by color flow and Doppler within the vessel from origin to ankle. Vessels were also classified continuously patent even if the vessel could not be visualized for a short distance provided flow could be identified in the artery immediately proximal an~-*distal to the nonvisualized segment, and there was no decrease in PSV. With use of the techniques described above, complete bilateral lower extremity arterial duplex examinations were performed in almost all patients in tess than 90 minutes. Data analysis The ability of LEADM to visualize individual arterial segments was calculated as the percentage of angiographically visualized segments that were also visualized with LEADM. In addition, comparison was made with angiographic controls to determine sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values of LEADM to detect a 50% or greater stenosis and/or occlusion in the common iliac, external iliac, common femoral, origin of the deep femoral, proximal, mid, and distal thirds of the SFA, and the above-knee and below-knee popliteal artery. For purposes of data an@is all SFA segments were considered together in analysis of LEADM in the SFA, and all popliteal artery segments were similarly grouped in analysis of LEADM applied to the popliteal artery. The ability of LEADM to distinguish a 50% or greater stenosis from

4 Journal of VASCULAR SURGERY 278 Moneta Fig. 1. Angiogram and color-flow duplex examination from a patient with severe aortoiliac disease. Multiple foci of severe stenosis (A, B, and C) were identified by duplex scanning through detection of focal step-ups in PSV. occlusion was also determined for the iliac, SFA, and popliteal arteries. For the infrapopliteal arteries, LEADM was evaluated in comparison with angiography for its sensitivity, specificity, PPV, and NPV in determining interruption of arterial patency from the level of the popliteal artery to the ankle. Data were analyzed for the entire patient group and according to the clinical determination of the distribution of atherosclerotic occlusive lesions. RESULTS Patients/distribution of occlusive disease Of the 150 patients entered into the stiady, the indication for angiography was abdominal aortic aneurysm in 24 patients, short distance claudication in 30 patients, and limb-threatening ischemia (ischemic rest pain, ulceration, or gangrene) in 96 patients. All but six patients were men, and the median age was years. Arterial segments visualized satisfactorily with angiography were available for comparison with duplex scanning from 286 limbs. Fourteen limbs either were not studied angiographically because of previous amputation or need to limit contrast volume or had angiographic studies technically unsuitable for satisfactory data analysis. According to the classification system presented in Table I, 80 limbs were without significant occlusive disease (group A), 44 limbs were clinically classified as having aortoiliac disease (group B), 117 limbs were classified as having infrainguinal occlusive disease (group C), and 45 limbs were classified as having multilevel inflow and outflow occlusive disease (group D). The distribution of angiographic

5 Volume I5 Number 2 February 199'2 Lower extremity arterial duplex mapping 279 C! Fig. 2. Angiogram and color-flow duplex examination from a patient with occlusion of the SFA and proximal popliteal artery. Duplex scanning identified the site of popliteal artery reconstitution (A), flow in the distal popliteal artery (B), and demonstrated continuous patency of the anterior tibial (C) and posterior tibial (D and E) arteries. The angiographically patent peroneal artery was erroneously judged occluded by the duplex examination. lesions for the four clinically determined patterns of arterial disease are shown in Tables II and III, respectively. The median ankle-brachial indexes were group A, ; group B, 0.51 _+ 0.12; group C, ; and group D, 0.41 _+ 0.16, excluding limbs with obviously incompressible tibial vessels. Ninety-one percent of all angiographically visualized infrapopliteal artery segments were successfully visualized with duplex scanning. In the infrapopliteal arteries, visualization of arterial segments was more successful in the anterior and posterior tibial arteries (94% and 96%) than in the peroneal artery (83%), p < 0.001, chi-square analysis. Visualization o f arterial segments by duplex sc:mning Table IV summarizes the number and distribution of arterial segments visualized angiographically and thus are available for comparison with duplex scanning, Overall, 99% of arterial segments visualized angiographically proximal to the tibial arteries were successfully identified with duplex scanning. Accuracy o f duplex scanning Table V depicts the sensitivity, specificity, PPV, and NPV for LEADM in the detection of a 50% or greater lesion in each of the arterial segments proximal to the tibial vessels, as a function of the clinical classification of disease. The values are similar in the three common clinical patterns of lower extremity atherosclerosis (groups B, C, and D), and

6 280 2~loneta Journal of VASCULAR SURGERY Table II. Angiographic status ( < 50% stenosis/_> 50% to 99% stenosis/occlusion) of arteries proximal to the tibial vessels according to clinically determined disease pattern Popliteal Superficial femoral (above-knee and Clinical Common External Common Deep femoral (proximal, mid, below-knee group* iliac iliac femoral (origin) and distal thirds) segments) A 71/4/0 73/3/1 67/0/1 57/3/3 122/4/1 65/3/2 B 14/10/20 24/5/15 29/1/9 30/2/3 71/7/6 44/2/0 C 100/6/1 92/7/0 92/7/0 79/24/0 126/42/ /8/32 D 27/7/9 20/13/12 30/3/12 33/7/5 34/21/73 69/1/12 *See text for delineation of clinical groups. Table III. Angiographic status (patent/occluded) of angiographically visualized tibial artery segments (proximal, mid, distal thirds)* Clinical Anterior Posterior group ~ tibial tibial Peroneal A 75/24 91/8 76/16 B 48/14 53/8 53/12 C 176/91 176/79 177/70 D 79/33 87/23 90/22 *Columns represent total number of patent/occluded segments for each tibial artery. tsee text for delineation of clinical groups. Table IV. Visualization of arterial segments by duplex scanning in 286 lower extremities studied by angiography Artery No. of segments satisfactorily visualized by angiography Percent of angiographic visualized segments visualized by duplex scanning Common iliac External ifiac Common femoral Deep femoral (origin) Superficial femoral* Poplitealt Anterior tibial* Posterior tibial* Peroneal* *Proximal, mid, and distal segments of the vessel were analyzed individually. Table represents total number of individual segments for each artery. tabove-knee and below-knee popllteal segments were analyzed individually. Table represents total of above-knee and below-knee popliteal segments analyzed. they indicate that the accuracy of LEADM in detecting a 50% or greater stenotic lesion in the iliofemoral-popliteal system is not substantially affected by clinical disease pattern. Furthermore, LEADM appears to be quite accurate in detecting high-grade iliac, SFA, and popliteal lesions in limbs clinically thought to be without significant occlusive disease (sensitivity of 100%, 67%, and 100%, respectively for iliac, SFA, and popliteal lesions in patients in group A, Tables II and V). Table VI lists the sensitivities, specificities, PPVs, and NPVs for LEADM in visualized vessels for determining interruption of infrapopliteal arteries at some point from the arterial origin to the ankle. The results indicate that with the exception of the peroneal artery the accuracy of infrapopliteal artery mapping by duplex scanning is substantially unaffected by the clinically determined pattern of atherosclerosis. Distinguishing stenosis from occlusion in vessels frequently suitable for transluminal angioplasty (iliacs, SFA, popliteal) In the iliac, SFA, and popliteal arteries, LEADM indicated the presence of a 50% to 99% stenosis in 87 arterial segments that were angiographically equal to or greater than 50% to 99% stenotic or occluded (28 iliac, 53 SFA, 6 popliteal). Lower extremity arterial duplex mapping suggested occlusion in 252 arterial segments that were angiographically equal to or greater than 50% to 99% stenotic or occluded (43 iliac, 176 SFA, 33 popliteal). In 98% of cases LEADM successfully distinguished stenosis from occlusion. Most errors involved classifying occluded lesions as stenotic and likely reflected the detection of high velocities in a collateral vessel. Five of the eight errors involved classifying an occluded SFA segment as stenotic, and in one case an occluded iliac segment was classified as stenotic. DISCUSSION It has long been the goal of the vascular laboratory to noninvasively characterize and quantify arterial occlusive disease. Over the last 20 years techniques have evolved from simple measurement of ankle blood pressures that globally document and crudely

7 Volume 15 Number 2 February 1992 Lower extremity arterial duplex mapping 281 Table V. Sensitivities, specificities, PPV, and NPV for duplex scanning in detecting a >-_ 50% stenosis or occlusion in lower extremity arteries proximal to the tibial vessels Clinical Common Deep Superficial group* Itiac femoral femoral femoral PopliteaI A 100/ $ 40/98 67/99 100/99 100/100 67/95 80/98 80/100 B 94/100 82/ /97 77/99 - $ 100/92 100/94 83/100 91/96 C 71/99 71/98 89/95 92/98 67/99 83/98 71/98 80/97 99/91 96/92 D 92/98 80/100 83/97 78/97 69/96 97/94 100/91 91/94 98/67 75/94 Total 89/99 76/99 83/97 87/98 67/99 94/97 93/96 83/97 97/89 93/94 *See test for delineation of clinical groups.?sensitivity/specificity PPV/NPV. ;Analysis not performed if fewer than five high-grade lesions in a category. Table VI. Sensitivities, specificities, PPV, and NPV of duplex scanning for identifying interruption of tibial artery patency Chnical group * Anterior tibial Posterior tibial PeroneM A 80/85t 50/ $ B 100/ / /72 100/ /100 55/100 C 73/96 100/88 92/76 92/85 76/100 73/94 D 73/89 63/92 60/61 85/81 71/89 40/78 Total all groups 90/93 90/92 82/74 84/88 77/97 62/90 *See text for delineation of clinical groups.?sensitivity/specificity PPV/NPV. SAnalysis not performed for categories with fewer than five arteries with segmental or total occlusion. quantify overall lower extremity arterial occlusive disease, to the use of segmental Doppler pressures and pulse volume recordings that partially localize but poorly characterize occlusive lesions, to duplex scanning with color-flow imaging that has the potential to precisely characterize occlusive lesions with respect to both location and severity. The ultimate goal of such testing is to eventually provide a noninvasively derived "map" of lower extremity atherosclerotic lesions that is both clinically relevant and sufficiently accurate to permit safe and efficient conduct of arterial reconstructions without antecedent angiography. Preoperative arteriography may not be required for all arterial reconstructions. Carotid endarterectomy based on duplex findings alone is now widely performed and appears safe. 6,7 Patients with in situ vein grafts in whom vein graft stenosis develops have, in many cases, very focal lesions that may be amenable to identification with duplex scanning. 8 Safe graft revision without confirmatory angiography may therefore be possible. Two recent papers have also shown duplex scanning to be quite accurate in predicting the suitability of arterial lesions for endovascular interventions?,9 Suggestions that duplex scanning may substitute for angiography in the planning of surgical arterial reconstructions are, however, probably premature, x To date no data exist to confirm the accuracy of LEADM in a large preoperative patient group. In addition, knowledge of popliteal artery outflow is frequently important in planning surgical reconstruction, and the ability of LEADM to insonate and accurately assess the infrapopliteal arteries has not been established previously. The objective of this study was to assess in a blinded fashion the feasibility and accuracy of LEADM in comparison with angiography in a large group of patients with varying patterns of clinically determined lower extremity atherosclerosis and to evaluate the ability of LEADM to determine run-off

8 282 Moneta Journal of VASCULAR SURGERY status of the infrapopliteal arteries from their origin to the ankle. We think these are essential preliminary steps in establishing the overall clinical usefulness of LEADM in evaluating patients with lower extremity arterial occlusive disease. A number of interesting findings were observed. Our results confirm the findings of previous smaller studies regarding the accuracy of LEADM in detecting clinically significant occlusive lesions proximal to the tibia] arteries. TM Our overall sensitivities, specificities, and PPVs in detecting such lesions (Table V) are similar to those previously published. Despite the fact that our vascular technologists had limited experience with LEADM before the initiation of this study, clinically useful studies were obtained in almost all patients. Finally, it appears the accuracy of LEADM is relatively unaffected by the clinically determined pattern of lower extremity atherosclerosis and that the technique can well predict the status of anterior and posterior tibia] artery runoff (Tables IV, V, and VI). These findings suggest the potential clinical applications of LEADM may be extensive. Lower extremity arterial duplex mapping appears to be quite useful in detecting hemodynamically significant lesions proximal to the groin. The ability of LEADM to identify significant iliac lesions as well as to distinguish stenosis from occlusion may in the future make LEADM the preferred noninvasive method for assessing the adequacy of arterial inflow to the groin. It must be acknowledged, however, that the ability of LEADM to predict resting and/or exercise induced arterial pressure gradients has not been established and that knowledge of such gradients is frequently important in the planning of surgical procedures.12'l~ Perhaps of more importance is the ability of LEADM to determine that arteries proximal to the tibia] vessels are unlikely to have a greater than 50% stenosis. Overall specificities and NPVs with respect to detecting a greater than 50% stenosis exceed 90% for all vessels except the SFA (NPV, 89%). Thus it appears that if LEADM does not identify a flowlimiting lesion proximal to the tibial vessels, it is very unlikely one would actually be found on angiography. This does not appear to be the case for determining the absence of significant arterial lesions based on the combination of segmental pressures and physical examination (Table II). This suggests that duplex scanning may be the optima] method for serially monitoring individual arterial' segments in studies devoted to assessing the natural history of lower extremity atherosclerosis. The results of tibia] artery scanning must, how- ever, be viewed with more caution. Although LEADM of the tibial arteries appears reasonably good in predicting popliteal artery outflow with respect to the anterior and posterior tibia] arteries (sensitivities and specificities for detecting occlusion within the anterior and posterior tibia] arteries exceeding 90%), the results are not as good for the peroneal artery. The results for the peroneal artery are even more disappointing when one considers that 17% of the peroneal artery segments visualized angiographicauy could not be visualized by colorflow duplex scanning. Although we feel this study establishes the feasibility of tibia] artery duplex scanning it would certainly be premature to suggest that tibia] artery duplex mapping is currently sufficient to plan a surgical procedure where an infrapoplitea] anastomosis is likely. More work is certainly needed to design a test of the ability of tibial artery duplex mapping to quantify tibia] artery lesions beyond the binary classification of patent/occluded and to assess for luminal irregularity. Will future lower extremity arterial surgery under certain conditions be based on duplex scanning without routine preoperative angiography? We suggest the answer is a qualified yes. Although investigators of extremity duplex scanning have, as yet, not attempted to evaluate "quality" of arteries and their suitability for anastomosis, such a project is ongoing in our laboratory and presumably in others. Although we have not previously relied on color flow to quantify stenosis, it may be that color will be important in determining the luminal irregularity of individual arterial segments and thus their potential suitability for a surgical anastomosis. In addition, the relative importance of the angiographic appearance of an artery compared with its hemodynamic performance is unclear. No experienced surgeon can deny many vessels are far more diseased at operation than suggested by preoperative arteriography. Fortunately, unpleasant surprises have been tempered by advances in surgical technique and equipment making it possible to routinely sew to even heavily calcified vessels. We believe that LEADM as a preoperative planning tool will, in many cases, be limited by unique circumstances surrounding particular operations. For instance, a patient with limited autogenous conduit who would be best served by a distal deep femoral origin of a vein graft would not currently be a candidate for LEADM alone because the ability of LEADM to evaluate the deep femora] artery beyond its origin is unproven. On the other hand, patients with iliac artery occlusions and normal runoff deter-

9 Volume 15 Number 2 February 1992 Lower extremity arterial duplex reaapping 283 mined with LEADM probably could currently be operated on without angiography with a high expectation of success and a low chance of encountering a significant unexpected problem. Certainly carefully performed prospective trials determining the circumstances under which LEADM can serve as a substitute for angiography are likely to be reasonable sometime in the near future. REFERENCES 1. Bridges RA, Barnes RW. Segmental limb pressures. In: Kempczinski RF, Yao JST, eds. Practical noninvasive vascular diagnosis. 2nd ed. Chicago: Year Book Medical Publishers, 1987: Jager KA, Phillips DJ, Martin ILL, et al. Noninvasive mapping of lower limb arterial lesions. Ultrasound Med Biol I985; 11: Cossman DV, Ellison JE, Wagner WH, et al. Comparison of contrast arteriography to arterial mapping with color-flow duplex imaging in the lower extremities. J Vasc SURG 1989;10: Rizzo RJ, Sandager G, Astleford P, et al. Mesenteric flow velocity variations as a function of angle of insonation. J VASC SURG 1990;11: Beach KW, Strandness DE Jr. Carotid artery velocity waveform analysis. In: Bernstein EF, ed. Noninvasive diagnostic techniques in vascular surgery. St. Louis: CV Mosby, 1985: Walsh J, Markowitz I, Kerstein MD. Carotid endarterectomy for amaurosis fugax without angiography. Am J Surg I986; 152: Moore WS, Ziomeks S, Quinones-Baldrich WI, Machleder HI, Busuttili RW, Baker JD. Can clinical evaluation and noninvasive testing substitute for arteriography in the evaluation of carotid artery disease? Ann Surg 1988;208: Bandyk DF, Schmitt DD, Seabrook GR, Adams MB, Towne JB. Monitoring fimctional patency of insitu saphenous vein bypass: the impact of a surveillance protocol and elective revision. J VASC SUr, G 1989;9: Edwards JM, Caldwell DM, Goldman ML, Strandness DE Jr. The role of duplex scanning in the selection of patients for transltuninal angioplasty. J VAsc SUe, G 1991;13: Kohler TR, Andros G, Porter ~M, et al. Can duplex scanning replace arteriography for lower extremity arterial disease? Ann Vase Surg 1990;4: Kohler TR, Nance DR, Cramer MM, Vandenburghe N, Strandness DE Jr. Duplex scanning for diagnosis of aortoiliac and femoropopliteal disease: a prospective study. Circulation 1987;6: Langsfeld M, Nepute J, Hershey FB, et al. The use of deep duplex scanning to predict hcmodynamicatly significant aortoiliac stenoses. J VAse SURG 1988;7: Kohler TR, Nicholls SC, Zierler RE, Beach KW, Schubart PJ, Strandness DE Jr. Assessment of pressure gradient by Doppler ultrasound: experimental and clinical observations. J VASC SURG 1987;6: Submitted June 10, 1991; accepted Aug. 15, DISCUSS][ON Dr. Brian Thiele (Hershey, Pa.). This is a carefully conducted study, the largest of its typc to dare, which prospectiveiy evaluates duplex scanning techniques for accuracy in identifying mild lesions of less than 50% diameter-reducing stenosis, severe lesions in the range of 50% to 99% stenosis, and complete occlusions in the iliac, femoral, and popliteal arteries. Color duplex ultrasonography was also performed to evaluate the infrageniculate arteries to document patent or occluded vessels. The results, as vce have heard, were comparcd with arteriographic studies that were analyzed for the same criteria by evaluators blinded to the results of those obtained with the duplex studies. This study is an important additional step in the gradual progression of the application of duplex ultrasonography in the evaluation of patients with peripheral vascular disease. This report has several very encouraging features that I think are worth emphasizing. First, adequate duplex scanning could be performed not only in all patients, but in almost all of the segments evaluated. The training the technologists undertook to complete these studies was not of enormous duration and demonstrates that with current instrumentation the whole of the lower extremity arterial system is aocessible to these examinations. Second, calcification, which is relatively common, particularly in the aortoiliac segment and the leg vessels and interferes with signal quality, did not appear to be a major practical problem in obtaining the excellent results presented for evaluation of the aortoiliac segment. Third, the evaluation of the infrageniculate arteries with real-time color ultrasonography correlated very well with the arteriographic studies apart from the patency of the peroneal vessels and clearly demonstrates the facility of this examination in evaluating the distal segment of the arterial tree, which in many institutions is not optimally visualized with arteriography unless special precautions are taken. The authors also enthusiastically project the concept of lower extremity reconstructive surgery being performed without arteriography by use of further development of the types of studies they reported here today. In this setting, however, I would caution that it is important to differentiate between getting away with something and doing what is right. I would like to direct the following comments and questions to the authors and specifically to Dr. Moneta. One of the basic observations regarding lower extremity ischemia that has been made as a result of vascular

10 284 Moneta Journal of VASCULAR SURGERY laboratory studies is that arterial structure and function are not uniformly related to one another. Severe morphologic change is frequently tolerated because of good collateral development. It seems to me that these studies reside somewhere between classic morphologic studies and physiologic studies. For example, can we, as a resuk of the duplex criteria, clearly determine that a lesion is responsible for producing a pressure gradient at rest or under the stimulus of exercise? I believe the answer at the moment is no. The second comment and question relates to the use of the traditional standard of arteriography for the purposes of comparison. In evaluating the aortoiliac segment in particular, many feel more comfortable with biplane views in assessing the degree of stenosis, and I ask the authors if there was a subset of patients in whom biplane views were available for comparison with the duplex studies. I am perplexed by the relatively poor results obtained from the popliteal artery segment and the peroneal artery, and I wonder if the authors could speculate on the reason for the inability to identify these vessels well. Finally, I ask the authors, in view of the fact that they are aggressively pursuing additional studies in this area, how they believe these studies will interact with or replace the more traditional vascular laboratory examinations such as the measurement of pressure gradients. Dr. Gregory Moneta. I agree it is unknown what duplex criteria correlate with resting or papaverine or exercise-induced pressure gradients. Previous studies addressing this issue are conflicting, and at this point we really cannot tell what the pressure gradient is across an individual stenosis based on duplex studies alone. I also agree that the use of angiography as a gold standard with which to compare duplex scanning is suboptimal. But we have to admit that single physician interpretation of angiograms is what we all do every day to make clinical decisions. Although a single viewer anglogram interpretation may not be the scientifically most sound approach, it is probably the most clinically relevant. We do have biplanar angiography in our hospital. In fact, it is used routinely whenever there is a question of an iliac or deep femoral lesion. However, we did not separate patients in terms of whether they had biplane or single plane angiograms in determining the accuracy of duplex scanning in the iliac system. I also think the results of the popliteal artery segments could be better. It would seem this should be a fairly easy arterial segment to examine. I suspect, however, that we do not always examine the artery in an optimal fashion. This part of the examination, I believe, should be done with the patient either in the prone or lateral position. In fact it is frequently performed with the patient supine and bending their knee shghdy and the technician trying to examine it from a posterior approach. Another complicating factor may be the presence of prominent geniculate vessels that can be confused with small but patent popliteal arteries. As for the peroneal artery, it is relatively easy to understand why duplex scanning may not work as well for this vessel as compared with the other tibial arteries. The artery is quite deep and travels parallel to the skin and has many fascial borders surrounding it that make it difficuk to insonate. I believe better rates of visualization of the peroneal artery will likely require increased technician experience and additional improvements in the engineering of these machines that allow better B-mode imaging of small, deep vessels. The final question that you asked was whether duplex scanning can be used in lieu of segmental pressures or pulse volume recordings. I think that as this technology gets better, and more technicians are trained in its use, pulse volume recorders and four-cuffpressure devices will start to gather dust in a lot of laboratories. The duplex scanner detects many lesions that are simply missed with four-cuff pressures and clinical evaluation alone. In fact, we have recently compared four-cuff pressures to duplex scanning with regard to their respective ability to localize arterial lesions to a particular arterial segment. At every applicable level duplex scanning is far superior.