| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Institute of Diagnostic Radiology (J.K.W., L.M.D., B.M., D.W.), Division of Cardiovascular Surgery (D.M.), and Division of Angiology (M.B.), University Hospital Zurich, Rämistrasse 100, 8091 Zurich, Switzerland; University Institute of Applied Radiophysics, Lausanne, Switzerland (F.R.V.); and Department of Biostatistics, University of Zurich, Switzerland (B.S.). Received September 6, 2002; revision requested November 7; final revision received February 24, 2003; accepted March 28. Address correspondence to D.W. (e-mail: dominik.weishaupt@dmr.usz.ch).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Four-channel multi–detector row CT angiography was performed in 65 consecutive patients with 85 peripheral arterial bypass grafts. Each bypass graft was divided into three segments (proximal anastomosis, course of the graft body, and distal anastomosis), resulting in 255 segments. Two readers evaluated all CT angiograms with regard to image quality and the presence of bypass graft–related abnormalities, including graft stenosis, aneurysmal changes, and arteriovenous fistulas. The results were compared with McNemar test with Bonferroni correction. CT attenuation values were recorded at five different locations from the inflow artery to the outflow artery of the bypass graft. These findings were compared with the findings at duplex ultrasonography (US) in 65 patients and the findings at conventional digital subtraction angiography (DSA) in 27.
RESULTS: Image quality was rated as good or excellent in 250 (98%) and in 252 (99%) of 255 bypass segments, respectively. There was excellent agreement both between readers and between CT angiography and duplex US in the detection of graft stenosis, aneurysmal changes, and arteriovenous fistulas (
= 0.86–0.99). CT angiography and duplex US were compared with conventional DSA, and there was no statistically significant difference (P > .25) in sensitivity or specificity between CT angiography and duplex US for both readers for detection of hemodynamically significant bypass stenosis or occlusion, aneurysmal changes, or arteriovenous fistulas. Mean CT attenuation values ranged from 232 HU in the inflow artery to 281 HU in the outflow artery of the bypass graft.
CONCLUSION: Multi–detector row CT angiography may be an accurate and reliable technique after duplex US in the assessment of peripheral arterial bypass grafts and detection of graft-related complications, including stenosis, aneurysmal changes, and arteriovenous fistulas.
© RSNA, 2003
Index terms: Angiography, comparative studies, 921.12916, 924.12916 • Arteries, grafts and prostheses, 921.452, 924.452 • Arteries, peripheral, 921.452, 924.452 • Computed tomography, (CT) multi–detector row • Ultrasound (US), Doppler studies, 921.12984, 924.12984
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Periodic surveillance of peripheral arterial bypass grafts is considered to be important, since early identification of failing grafts can often avert impending graft failure and improve the secondary bypass graft patency rate (2–4). Because duplex ultrasonography (US) is noninvasive, provides rapid access, and has a high level of accuracy and a low cost, it was advocated as the primary screening technique in the surveillance of grafts and the identification of bypass graft–related complications (5–7).
In most centers, conventional digital subtraction angiography (DSA) is performed as the next diagnostic step when findings at duplex US indicate bypass graft-related abnormalities. This is especially true when a surgical or conventional intervention is intended (8,9). The main drawbacks of conventional DSA are the patient discomfort it causes, its invasive nature, and a complication rate of approximately 1% (10,11).
The recent introduction of multi–detector row computed tomography (CT) has promoted the use of multi–detector row CT angiography as an alternative to conventional DSA in assessment of the vascular system. Multi–detector row CT has substantially improved CT angiography by offering shorter image acquisition times, increased volume coverage, and improved spatial resolution in the assessment of smaller arterial branches and by requiring a smaller dose of contrast medium (12–14).
The purpose of this prospective study was to assess the technical feasibility of multi–detector row CT angiography in the assessment of peripheral arterial bypass grafts and to evaluate its accuracy and reliability in the detection of graft-related complications, including graft stenosis, aneurysmal changes, and arteriovenous fistulas.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
A total of 88 peripheral arterial bypass grafts were implanted in 65 patients for the treatment of peripheral occlusive vascular disease. With the exception of two patients with Buerger disease, the primary cause of arterial occlusive disease was atherosclerosis. Forty-five patients had one peripheral arterial bypass graft, 17 had two peripheral arterial bypass grafts (one in the right leg and one in the left), and three had three peripheral arterial bypass grafts (one in the right leg and two in the left). In three patients with two bypass grafts in the same leg, image analysis of one graft could influence image analysis of the other, so the occluded bypass graft was excluded from image analysis.
A total of 85 bypass grafts were evaluated in 65 patients. The 85 bypass grafts were distributed in the following anatomic locations: distal (eg, below the knee) femoropopliteal artery (n = 33), proximal (eg, above the knee) femoropopliteal artery (n = 19), femorocrural artery (n = 24), femorofemoral artery (n = 2), femoropedal artery (n = 2), popliteopopliteal artery (n = 2), iliacofemoral artery (n = 1), femoroprofundal artery (n = 1), and popliteopedal artery (n = 1). Sixty of the 85 bypass grafts were autologous saphenous vein grafts (42 in situ, 18 inverted), 18 were expanded polytetrafluoroethylene grafts (Vascular Graft; Impra, Tempe, Ariz), six were biografts (Procol; Hancock Jaffe Laboratories, Irvine Calif), and one was a polyester graft (Dacron Gelsoft Plus; Sulzer Vascutek, Renfrewshire, Scotland). The time interval between implantation of the peripheral arterial bypass graft and inclusion into the study ranged from 3 days to 4 years (median, 1 month).
A total of 45 (69%) of 65 patients underwent routine surveillance of peripheral bypass graft patency. Twenty-four patients underwent surveillance 1 month after surgery; eight patients, 6 months after surgery; six patients, 12 months after surgery; two patients, 18 months after surgery; three patients, 36 months after surgery, and two patients, 48 months after surgery. None of these 45 patients had any clinically relevant symptoms. Fifteen (23%) of 65 patients were evaluated for symptoms of peripheral arterial bypass graft stenosis or occlusion (pain or claudication) at 21 days to 40 months after bypass graft surgery, and five (8%) of 65 patients were evaluated because they experienced symptoms of bypass graft stenosis or occlusion within 7 days of surgery.
All 65 patients underwent multi–detector row CT angiography and duplex US within 2 days. Thirty-three patients underwent multi–detector row CT angiography before duplex US, and 32 underwent duplex US before multi–detector row CT angiography. Twenty-seven (42%) of 65 patients with 33 arterial bypass grafts underwent additional conventional DSA. Patients underwent conventional DSA if multi–detector row CT angiographic or duplex US findings indicated the presence of a stenosis or an occlusion and if these findings were considered serious enough to justify performance of conventional DSA for clinical reasons.
Multi–Detector Row CT Angiography
All 65 patients underwent scanning with a four-channel multi–detector row CT scanner (Somatom Volume Zoom; Siemens, Forchheim, Germany). All patients were placed in the supine position so that their feet entered the gantry first. This allowed the portion of the body from the aortic bifurcation to the distal lower limbs to pass through the CT gantry. The patient’s knees and ankles were fixed together with adhesive tape, and a small pillow was placed below the patient’s knees. The patient’s extremities were positioned with the knee and ankle joints in the neutral position.
After an initial anteroposterior scout image (120 kV, 50 mAs) was obtained, the imaging range was planned for each individual. This ensured that the extent of the peripheral arterial bypass graft, including the inflow artery (eg, the vessel proximal to the proximal anastomosis of the arterial bypass graft) and the outflow artery (eg, the vessel distal to the distal anastomosis of the arterial bypass graft), would be imaged. For the purposes of our study, the imaging range included all points between the aortic bifurcation and pedal circulation in all patients, irrespective of the anatomic location of the peripheral bypass graft. The mean imaging coverage was 98 cm (range, 86–105 cm).
For optimal intraluminal contrast enhancement, the delay time between the start of contrast medium administration and the start of imaging was determined for each patient by using a bolus-tracking technique (CARE-Bolus; Siemens). For this purpose, a single, unenhanced low-dose (10-mAs) examination at the level of the common iliac artery was performed. On the basis of this transverse image, a 10–15-mm2 region of interest was placed in the lumen of the common iliac artery by a radiologist (J.K.W.). If two or three peripheral arterial bypass grafts were present at two different sites within the uni- or contralateral leg, the region of interest was placed over the common iliac artery lumen of the leg with the most proximal arterial bypass graft. This region of interest served as a point of reference for the dynamic measurements of contrast enhancement that followed.
Subsequently, 120 mL of iopromidum (Ultravist 300; Schering, Berlin, Germany) was administered via a 20–22-gauge needle, which was placed into a superficial vein in the antecubital fossa. The contrast medium was administered with an automated injector (CT Injector; Ulrich Medical, Ulm-Jungingen, Germany) at a rate of 4 mL/sec. The contrast medium bolus was followed by 30 mL of saline administered at the same rate.
Repetitive low-dose monitoring examinations (120 kV, 10 mAs, 0.5-second scanning time, 1-second interscan delay) were performed 10 seconds after contrast medium injection began. After reaching the preset contrast enhancement level of 100 HU, the multi–detector row CT examination was automatically initiated. Data acquisition was performed in a craniocaudal direction with a nominal section thickness of 2.5 mm, a table feed of 15 mm per rotation, and a gantry rotation of 0.5 second (pitch, 1.5). The x-ray tube potential was 120 kV and the tube current was 390 mA, which led to a tube charge of 130 mAs per gantry rotation ([390 mA x 0.5 second]/1.5 = 130 mAs).
Transverse sections were reconstructed separately for each extremity at a workstation (Volume Zoom Navigator; Siemens). Section thickness was 3.0 mm and the interval was 1.5 mm (1.5 mm overlap), which resulted in a mean of 680 transverse images (range, 574–734) for each extremity. The reconstruction field of view was 25 cm for each extremity. The field of view, matrix size of 512 x 512, and section thickness of 3.0 mm resulted in an interpolated voxel size of 0.36 mm3.
Each multi–detector row CT angiogram was processed in random order into a volume-rendered image and a maximum intensity projection image at a separate workstation (Advantage Windows 4.0; GE Medical Systems, Milwaukee, Wis) by one radiologist (L.M.D.) with 2 years of experience in three-dimensional imaging techniques. This radiologist was not involved in further image analysis and was blinded to patient data and the results from duplex US and conventional DSA. Prior to creating volume-rendered and maximum intensity projection images from CT data sets, segmentation of obscuring bone structures was performed. In the first step of segmentation, a bone model that was based on the CT data set was created. A mean attenuation threshold level of 160 HU was applied to remove most of the nonosseous structures. In addition, supplemental manual cutting (eg, region of interest drawing and scalpel cuts), combined with region growing, was performed to remove all vessels from the bone model. In the second step, the resultant bone model was subtracted from the intact CT data set to obtain a CT data set without bone structures.
For each patient, 36 volume-rendered images and 36 maximum intensity projection images obtained with multi–detector row CT angiography were created perpendicular to the superoinferior axis covering 360° of rotation in 10° increments. Three-dimensional reconstructions were stored on the hard-disk memory of the workstation for subsequent image analysis.
The effective dose delivered during multi-detector row CT angiography was estimated for the regions of the pelvis and hip, since exposure of the extremities minimally contributes to the effective dose. For this purpose, the distance from the pelvic crest to the proximal third of the thighs (including the testicles in men) was measured in each patient by a technologist. The mean measured distance between the pelvis and hip regions was 28 cm for men and 22 cm for women. The dose length product was then calculated by using a normalized weighted CT dose index of 0.10 mGy/mAs. This value is representative of the normalized weighted CT dose index expected with the type of multi–detector row CT scanner set at 120 kV that was used in this study. The dose length product was then converted into effective dose values by means of a conversion factor of 0.019 mSv/mGy · cm, according to the guidelines for quality criteria for CT of the Commission of the European Communities. (15). All dose calculations were performed by a physicist (F.R.V.).
Analysis of Multi–Detector Row CT Angiograms
All multi–detector row CT angiograms were analyzed by two independent radiologists (D.W. and J.K.W.) with 3 and 2 years of experience, respectively, in vascular radiology and the use of an interactive workstation. The reconstructed transverse multi–detector row CT images and the volume-rendered and maximum intensity projection images (as created by L.M.D.) were available for both readers at the workstation. Both readers were allowed to individually adjust window centers and level settings of the multi–detector row CT angiograms for image analysis. A cine mode was available at the workstation for rapid interactive interpretation. Both readers were blinded to patient data, including clinical history, surgical reports, and findings of duplex US and conventional DSA examinations. In patients with one bypass graft in each leg, data reconstruction and image analysis were performed separately.
For the purpose of analysis, each peripheral arterial bypass graft was divided into three segments, including the proximal anastomosis, which included the proximal 1-cm length of the arterial bypass graft body; the course of the body of the graft; and the distal anastomosis, which included the distal 1-cm length of the arterial bypass graft body. Both readers assessed image quality for diagnostic purposes for each of the peripheral arterial bypass graft segments with a four-point Likert scale. Grade 1 indicated nondiagnostic image quality, which meant that diagnostic information was not obtained. Grade 2 indicated adequate image quality, which meant that all clinically relevant diagnostic information was obtained with poor differentiation of the graft lumen because of clip artifacts. Grade 3 indicated good image quality, which meant that all clinically relevant diagnostic information was obtained with good differentiation of the graft lumen. Grade 4 indicated excellent image quality, which meant that all clinically relevant diagnostic information was obtained with excellent differentiation of the graft lumen. Presence of bypass graft stenosis was evaluated for all bypass graft segments.
Arterial bypass graft stenosis was also graded with a four-point scale. Grade 1 indicated a normal peripheral arterial bypass graft or the presence of bypass graft irregularities (<10% luminal narrowing). Grade 2 indicated a mild luminal bypass graft stenosis (10%–49% luminal narrowing). Grade 3 indicated a severe luminal bypass graft stenosis (50%–99% luminal narrowing). Grade 4 indicated occlusion. When two or more stenotic luminal changes were detected in the same segment of the peripheral arterial bypass graft, the most severe change was used for grading.
Evidence of aneurysmal changes in the peripheral arterial bypass graft was separately noted by both readers. An aneurysmal change was diagnosed in the presence of a focal increase in luminal diameter of more than 50% compared with the normal diameter of the remaining peripheral arterial bypass graft.
Finally, both readers were asked to note the presence or absence of arteriovenous fistulas associated with autologous in situ saphenous vein bypass grafts. An arteriovenous fistula was defined as a residual patent vein branch originating from the autologous in situ saphenous vein bypass graft with evidence of early filling of deep and/or superficial venous structures (16).
CT attenuation values were measured at a workstation by a single radiologist (J.K.W.). A circular region of interest was placed over the luminal center of five different arterial segments, including the inflow artery (2 cm proximal to the proximal anastomosis of the arterial bypass graft), the proximal anastomosis, the middle of the body of the arterial bypass graft, the distal anastomosis, and the outflow artery (2 cm distal to the distal anastomosis of the arterial bypass graft). The diameter of the region of interest depended on the size of the vascular segment and ranged from 2 to 7 mm (mean, 4 mm). CT attenuation values were not measured if there was occlusion of segments of the peripheral arterial bypass grafts. In patients with only one peripheral arterial bypass graft (n = 45), additional CT attenuation values were measured at the corresponding level of five arteries in the opposite leg. A region of interest for CT attenuation value measurement was placed in the corresponding arterial segments of the opposite leg only if no hemodynamically significant stenosis (50%–100% luminal narrowing) was present in the corresponding segments. The decision to place a region of interest in the opposite leg was made on the basis of the assessment of the entire multi–detector row CT angiography data set and was made by the radiologist who performed the measurements.
Technical feasibility of all multi–detector row CT angiographic examinations was assessed by both readers. The examinations were technically feasible if they could be performed without any technical complications and without the need to repeat the study.
Duplex US
All duplex US scans in the 65 patients were obtained and interpreted prospectively by a single sonographer (M.B.) with 10 years of experience. A US unit (128/XP 10; Acuson, Mountain View, Calif) equipped with a linear, phased-array high-frequency (5–10-MHz) transducer was used for scanning. Occasionally, additional use of a curved linear medium-frequency (3.5-MHz) transducer was necessary in obese patients and for evaluation of an inflow artery. Doppler velocity waveforms were obtained along the entire course of the peripheral arterial bypass graft, including the arterial bypass graft body, along the proximal and distal anastomosis, along the entire course of the inflow artery vessel, and along the entire course of the outflow artery. Diameter reduction was determined by measuring the diameter of the peripheral arterial bypass graft at the level of the stenosis and at a normal graft segment a few centimeters proximal or distal to the lesion. A focal increase in the angle-adjusted peak systolic velocity in the peripheral arterial bypass graft, combined with a reduction of the luminal diameter of the peripheral arterial bypass graft, indicated arterial bypass graft stenosis. If this was followed by turbulence and the velocity ratio was two or more, the stenosis was considered to be hemodynamically significant (50%–99% luminal narrowing). A diagnosis of occlusion of the peripheral arterial bypass graft was made on the basis of the lack of flow signal in a visualized arterial bypass graft. For quantification of peripheral arterial bypass graft stenosis, the sonographer used the same four-point grading scale used for multi–detector row CT angiography.
Aneurysmal changes in the peripheral arterial bypass graft were considered to be present if there was a focal increase in luminal diameter of more than 50% compared with the normal diameter of the remaining bypass graft. In patients with an autologous in situ saphenous vein graft, the peripheral arterial bypass graft was also evaluated for the presence of arteriovenous fistulas.
Conventional DSA
Intraarterial conventional DSA of the vasculature from the distal aorta to the trifurcation arteries of the lower limb, including the peripheral arterial bypass graft, was performed transfemorally with a 4-F pigtail catheter (AngiOptic; Angiodynamics, Queensbury, NY) by using one of two DSA units (Integris V3000 or V5000; Philips Medical Systems, Best, the Netherlands). The catheter tip was positioned above the aortic bifurcation for evaluation of the pelvic arteries, and multiple images were acquired that encompassed the thighs, knees, and calves. At each station, 20 mL of iopromidum was injected at a rate of 10 mL/sec. The imaging protocol consisted of acquisition of 12 right and 12 left anterior oblique images of the pelvis at a rate of two per second and acquisition of 14 images each in the thigh, knee, and calf at a rate of one or two per second. Images obtained with conventional DSA were printed onto film with customized window width and level settings to allow clear delineation of the enhanced lumen of the vasculature and the arterial bypass grafts.
DSA images were interpreted by the same radiologist who performed the examination. The radiologist, blinded to the results of both duplex US and multi–detector row CT angiography, was asked to note the presence and severity of graft stenosis, the presence of aneurysmal changes of the peripheral arterial bypass graft, and the presence of arteriovenous fistulas by using the same classification as described for evaluation of the angiograms obtained with multi–detector row CT. The results were also reported separately for the proximal anastomosis, the course of the body of the peripheral arterial bypass graft, and the distal anastomosis.
Estimates of the effective dose at conventional DSA were calculated by one physicist (F.R.V.) on the basis of the dose area product quantity corresponding to the acquisition protocol used. The dose area product is displayed by the DSA system itself and is representative of the total energy deposited in the examined volume. The dose area product value displayed by the unit was verified according to the method described by Bochud et al (17). The effective dose was calculated only from the dose area product delivered in the pelvis and hip regions, since exposure of the extremities minimally contributes to the effective dose. Exposure of the pelvis and hip were both calculated with conversion factors of 0.10 and 0.14 mSv/Gy · cm2 for men and women, respectively. These conversion factors were obtained with commercially available software (Organ Doses Calculation software; Rados Technology, Turku, Finland). The difference in conversion factors between men and women is caused by the presence or absence of the gonads in the exposed area. All these conversion factors are in good agreement with data published by Le Heron (18).
Statistical Analysis
Interobserver agreement between both readers who evaluated multi–detector row CT angiograms was calculated by using statistics (including the 95% CIs) with regard to grading of the extent of bypass graft stenosis and determination of the presence of aneurysmal changes and the presence of arteriovenous fistulas of the peripheral arterial bypass grafts. According to Landis and Koch (19), a value of 0 indicated poor agreement, a value of 0.01–0.20 indicated slight agreement, a value of 0.21–0.40 indicated fair agreement, a value of 0.41–0.60 indicated moderate agreement, a value of 0.61–0.80 indicated good agreement, and a value of 0.81–1.00 indicated excellent agreement. statistics were also computed for intermodality agreement between multi–detector row CT angiography and duplex US for both readers. Differences between the three segments of the bypass grafts with regard to grading of bypass graft stenosis were calculated for both readers by using the paired sign test with Bonferroni correction, and P values less than .017 were considered to indicate a statistically significant difference.
Sensitivities and specificities of multi–detector row CT angiography and duplex US compared with conventional DSA for the determination of arterial bypass graft stenosis and the presence of aneurysmal changes or arteriovenous fistulas were calculated in the 27 patients with 33 peripheral arterial bypass grafts (99 arterial bypass graft segments) who underwent additional conventional DSA. For statistical purposes, grade 1 (<10% luminal narrowing) and grade 2 (10%–49% luminal narrowing) arterial bypass graft stenoses were considered hemodynamically insignificant, and grade 3 (50%–99% luminal narrowing) and grade 4 stenoses (occlusion) were considered hemodynamically significant for any imaging modality. The statistical significance of the differences in sensitivities and specificities between multi–detector row CT angiography and duplex US for both readers was assessed with the McNemar test for paired data. Bonferroni correction was used to adjust for multiple comparisons, and a P value of less than .0125 was considered to indicate a statistically significant difference.
The mean, SD, and 95% CIs of the CT attenuation values were calculated for the five different arterial segments of the site of the arterial bypass graft and of the opposite healthy site. Differences between the five different arterial segments with regard to the CT attenuation values and differences between the segments of the graft compared with the opposite leg without an arterial bypass graft were obtained by using a repeated-measures analysis of variance with Greenhouse-Geisser correction and a Bonferroni-Dunn post hoc test. A P value less than .005 was considered to indicate a statistically significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The image quality of multi–detector row CT angiograms was rated as excellent (grade 4) in 202 (79%) of 255 arterial bypass graft segments by reader 1 and in 201 (79%), by reader 2 (Table 1). Reader 1 and reader 2 rated image quality as good (grade 3) in 48 (19%) and 51 (20%) of 255 arterial bypass graft segments, respectively (Fig 1). These readers rated image quality as moderate (grade 2) in five (2%) and three (1%) of 255 arterial bypass graft segments. Multiple vascular clips were present in all arterial bypass graft segments, in which image quality was rated as moderate (grade 2) by both readers, causing metallic artifacts. None of the CT images of the 255 arterial bypass graft segments, however, were rated as nondiagnostic (grade 1).
|
|
The mean, 95% CI, and range of arterial bypass segment attenuations at each of the five localizations are plotted in Figure 2. CT attenuation values were not equivalent at all five localizations where the measurements were performed (P < .001). There was no significant difference in the measured CT attenuation values between the middle of the body of the arterial bypass graft, the distal anastomosis, and the outflow artery (P > .09). The CT attenuation values at those three localizations were, however, significantly higher (P < .001) compared with those measured within the inflow artery and the proximal anastomosis of the bypass graft. Differences in CT attenuation values between the inflow artery and the proximal anastomosis were not statistically significant (P = .02).
|
Multi–Detector Row CT Angiography versus Duplex US
Table 2 demonstrates the breakdown of findings of both readers for multi–detector row CT angiography and duplex US with regard to grading of arterial bypass graft stenosis. On multi–detector row CT angiograms, readers 1 and 2 identified grade 3 bypass graft stenosis in 11 (4%) and 10 (4%) of 255 arterial bypass graft segments, respectively (Fig 3). Both readers detected grade 4 bypass graft stenosis in 39 (15%) of 255 arterial segments. By using multi–detector row CT angiography, there was excellent agreement between both readers for all degrees of arterial stenosis ( = 0.98; 95% CI: 0.96, 1.00) and for diagnosis of hemodynamically significant and insignificant arterial bypass graft stenosis ( = 0.99; 95% CI: 0.97, 1.00). There were statistically significant differences between bypass graft stenosis grading of the proximal and distal segments (P = .001 for reader 1 and P = .0001 for reader 2) and between the middle and distal segments (P = .013 for both readers). Grading of bypass graft stenosis was not significantly different between the proximal anastomosis and the bypass graft body for both readers (P = .63 for reader 1 and P = .25 for reader 2).
|
|
|
values for both readers of 0.89 (95% CI: 0.87, 0.91). There was also excellent agreement between both readers and duplex US with regard to the diagnosis of hemodynamically insignificant versus hemodynamically significant arterial bypass graft stenosis (reader 1, = 0.95; 95% CI: 0.93, 0.97 and reader 2, = 0.94; 95% CI: 0.91, 0.97).
|
|
|
= 1.00).
There was excellent agreement between readers with regard to the detection of arteriovenous fistulas ( = 0.95; 95% CI: 0.90, 1.00). Readers 1 and 2 noted the presence of arteriovenous fistulas in 22 (52%) and 21 (50%) of 42 autologous in situ saphenous vein bypass grafts, respectively, on multi-detector row CT angiograms (Fig 5). Duplex US depicted the presence of arteriovenous fistulas in 19 (45%) of 42 autologous in situ saphenous vein bypass grafts, which resulted in excellent agreement between readers (reader 1, = 0.86; 95% CI: 0.78, 0.94 and reader 2, = 0.90; 95% CI: 0.84, 0.96).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The use of duplex US in the postoperative surveillance of peripheral bypass grafts is well documented (5). Duplex US can demonstrate graft patency and enables detection of complications, including stenosis or occlusion, perigraft fluid collections, arteriovenous fistulas, and pseudoaneurysms (6,7). Because of its noninvasive quality, low cost, and the rapid access it provides, duplex US is considered to be the primary imaging modality for use in postoperative graft surveillance (2).
Many vascular surgeons still consider intraarterial conventional DSA to be a mandatory procedure before performance of a surgical or percutaneous intervention for a bypass graft stenosis depicted with noninvasive modalities (3,8,9). When using duplex US, anatomically placed grafts may be difficult to follow throughout their entire course, which makes it difficult to localize the graft stenosis accurately (22). Landry et al (8) pointed out that use of duplex US alone is not sufficient for planning the surgical treatment of a failed peripheral arterial bypass graft. They found that conventional DSA contributed to preoperative planning in 86 (42%) of 205 patients by revealing more information than duplex US (8). These same authors found conventional DSA to be particularly helpful when a low-flow state of the arterial bypass graft was identified with duplex US or when an inflow lesion was present (8). Moreover, it may be difficult to detect sequential lesions with duplex US because the reduced flow below the proximal anastomosis may reduce the peak systolic flow at the more distal anastomosis of the bypass graft (8). Another potential limitation of duplex US is its difficulty in displaying the tibial and pedal arteries below the distal anastomosis (22).
Because of the invasive nature of conventional DSA and the small but existent complication rate of this procedure, magnetic resonance (MR) angiography has been suggested as a secondary technique for use in the assessment of peripheral arterial bypass grafts (23–25). Limitations of MR angiography for graft assessment relate to the limited spatial resolution of the modality and the fact that vascular clips may simulate graft stenosis (24). In addition, to our knowledge there are no data regarding the use of MR angiography in the detection of graft-related arteriovenous fistulas, which are important to recognize as they are potential complications of peripheral bypass grafts (26).
In this prospective study, we investigated the technical feasibility of multi–detector row CT angiography in the assessment of peripheral arterial bypass grafts and evaluated its accuracy and reliability in the detection of graft-related complications, including graft stenosis, aneurysmal changes, and arteriovenous fistulas. With the imaging protocol used in this study, we achieved opacification greater than 150 HU in 98% of the arterial segments of the peripheral arterial bypass graft. The lower boundary of the 95% CI of arterial enhancement was above 200 HU at all measurement points. There was no statistically significant difference between CT attenuation values of the peripheral arterial bypass graft compared with the CT attenuation values at the corresponding levels of the opposite leg. This excellent opacification of the segments of the arterial bypass graft is also reflected by the overall image quality of multi–detector row CT angiograms, which was assessed as good to excellent by both readers. Image quality was rated as moderate in three arterial bypass graft segments by reader 1 and in five by reader 2. The reduced image quality in these segments was caused by the presence of vascular clip material; however, even the presence of multiple vascular clips close to the peripheral bypass graft did not affect bypass graft evaluation for diagnostic purposes.
The results of our study have demonstrated that multi–detector row CT angiography is well suited for the morphologic assessment of peripheral arterial bypass grafts. Compared with duplex US, multi–detector row CT angiography allows an accurate depiction of graft-related complications, including stenosis, occlusion, aneurysmal changes, and arteriovenous fistulas. When compared with duplex US and conventional DSA, sensitivity and specificity values of more than 95% were achieved by both readers with multi–detector row CT angiography for the diagnosis of arterial bypass graft–related complications. The robustness and reliability of multi–detector row CT angiography are reflected by excellent interobserver agreement. All these findings suggest that multi–detector row CT angiography may be incorporated into a comprehensive graft assessment strategy as a secondary morphologic modality after functional assessment of the bypass graft with duplex US and that it may replace conventional angiography or DSA for this purpose.
Detection of arteriovenous fistulas after placement of in situ saphenous vein bypass grafts is important. Side branches of the saphenous vein, which are not identified and ligated during surgery, may persist as arteriovenous fistulas and cause shunting of blood from the bypass to the venous system. Superficial fistulas may produce local venous hypertension with a painful cutaneous flare, whereas fistulas draining into deep veins may divert flow from the distal part of the graft and result in a loss of pulsatility and thus distal graft thrombosis (27). Arteriovenous fistulas have been reported to occur in up to 75% of in situ saphenous vein bypass grafts (26). Although occurrence of arteriovenous fistulas after in situ placement of a saphenous vein bypass graft increases the proximal flow rate of the bypass, arteriovenous fistulas rarely impair the distal hemodynamics of in situ saphenous vein bypasses without stenosis (26). In combination with a bypass stenosis, however, the simultaneous occurrence of an arteriovenous fistula and a bypass graft stenosis may become hemodynamically significant (26).
In our study, multi–detector row CT angiography was most effective in demonstrating arteriovenous fistulas. In two and three patients, respectively, an arteriovenous fistula was detected with multi– detector row CT angiography but not duplex US. Future studies are warranted to determine if multi–detector row CT angiography allows accurate quantification of arteriovenous fistulas and if it may contribute to an optimization of bypass graft surveillance by allowing better visualization of arteriovenous fistulas than duplex US.
A potential drawback of multi–detector row CT angiography for the assessment of peripheral arterial bypass grafts is the radiation exposure. In our study, an effective dose of 6.92 mSv for men and 5.43 mSv for women was calculated for multi–detector row CT angiography. Compared with the mean effective dose values calculated for conventional DSA, the mean effective dose from multi–detector row CT angiography was lower by a factor of about two for men and about three for women. The lower administered effective dose with multi–detector row CT angiography might be another advantage fostering its use as a compelling alternative to conventional DSA and as a secondary imaging strategy in the assessment of failing peripheral arterial bypass grafts. Because the examination range for multi–detector row CT angiography can be individually adjusted to the extent of the arterial bypass graft, the effective dose may be reduced even further.
We acknowledge limitations of this study. One study limitation relates to the small prevalence of aneurysmal changes and arteriovenous fistulas in the subgroup of patients who underwent additional conventional DSA. This may limit the calculated descriptive statistics for both multi–detector row CT angiography and duplex US versus conventional DSA. Another limitation relates to the fact that only those patients in whom stenosis was suspected or in whom occlusion of bypass graft segments was diagnosed on the basis of duplex US or multi–detector row CT angiography findings underwent additional conventional DSA. This may have resulted in a systematic bias when evaluating conventional DSA for the presence of arterial stenosis or occlusion. Although use of conventional DSA as the standard of reference for the entire study population would have been preferable, ethical concerns regarding the potential risks of conventional DSA led to the exclusion of patients with negative findings at multi–detector row CT angiography or duplex US. Finally, we did not address cost issues of conventional DSA or multi–detector row CT angiography in our study. Further prospective studies are warranted to compare conventional DSA and multi–detector row CT angiography in the evaluation of peripheral arterial bypass grafts in terms of cost-effectiveness, since this analysis may also guide referring physicians to select conventional DSA or multi–detector row CT angiography.
In conclusion, this study has demonstrated in a prospective blinded comparison that multi–detector row CT angiography is feasible, accurate, and reliable in the assessment of peripheral arterial bypass grafts and detection of graft-related complications, including stenosis, aneurysmal changes, and arteriovenous fistulas. Because of its noninvasive nature and lower effective dose, multi–detector row CT angiography may replace conventional DSA as a technique to be used after performance of duplex US to help physicians plan further treatment of peripheral arterial bypass grafts.
|
FOOTNOTES |
|---|
Abbreviation: DSA = digital subtraction angiography
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. H. Lin, C. Bechara, P. Kougias, T. T. Huynh, S. A. LeMaire, and J. S. Coselli Assessment of Aortic Pathology and Peripheral Arterial Disease Using Multidetector Computed Tomographic Angiography Vascular and Endovascular Surgery, December 1, 2009; 42(6): 583 - 598. [Abstract] [PDF]
|
|
|
|
 |
|
 J. E. Lopera, C. K. Trimmer, S. G. Josephs, M. E. Anderson, S. Schuber, R. Li, B. Dolmatch, and B. Toursarkissian Multidetector CT Angiography of Infrainguinal Arterial Bypass RadioGraphics, March 1, 2008; 28(2): 529 - 548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Schernthaner, D. Fleischmann, F. Lomoschitz, A. Stadler, J. Lammer, and C. Loewe Effect of MDCT Angiographic Findings on the Management of Intermittent Claudication Am. J. Roentgenol., November 1, 2007; 189(5): 1215 - 1222. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Burrill, Z. Dabbagh, F. Gollub, and M. Hamady Multidetector computed tomographic angiography of the cardiovascular system Postgrad. Med. J., November 1, 2007; 83(985): 698 - 704. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. T. Johnson and E. K. Fishman IV Contrast Selection for MDCT: Current Thoughts and Practice Am. J. Roentgenol., February 1, 2006; 186(2): 406 - 415. [Abstract] [Full Text] [PDF]< |
