reconstructions: Fate of repaired and

Transcription

1 Intraoperative duplex scanning of arterial reconstructions: Fate of repaired and unrepaired defects Dennis F. Bandyk, MD, Joseph L. Mills, MD, Vivian Gahtan, MD, and Glenn E. Esses, MD, Tampa, Fla. Purpose: Because unrecognized lesions can cause an arterial reconstruction to fail, duplex ultrasonography was evaluated as an intraoperative aid to assess technical adequacy and provide criteria for which lesions should be repaired immediately versus safely followed. Methods: Since 1990 intraoperative color duplex scanning (7 to 10 MHz linear array probe, pulsed-wave Doppler test spectrum analysis) was used to assess the frequency and severity of residual lesions in 368 patients after carotid endarterectomy (n = 210), infrainguinal vein bypass (n = 135), or visceral/renal reconstruction (n = 23). Duplex scan results were categorized as normal or abnormal, with immediate repair of lesions demonstrating both lumen reduction and severe focal flow abnormalities (peak systolic velocity [Vp] > 150 to 180 cm/sec; velocity ratio [Vr] > 2.4). Arteriography was also performed in 81% of lower limb bypass procedures. Results: Duplex scanning identified technical (residual plaque, stricture) or intrinsic defects (platelet thrombus, distal thrombosis) requiring revision in 37 (10%) of the reconstructions. Infrainguinal bypass had the highest incidence of corrected defects (14%) and adverse events (3%). No adverse events occurred in patients with normal duplex scan results or after carotid endarterectomy. Overall, 76% of identified defects were corrected (carotid, 17 of 24; infrainguinal bypass, 19 of 24; visceral bypass, 1 of 1). Uurepaired flow defects (Vp = 150 to 190 cm/sec; Vr = 1.8 to 2.5) led to one graft occlusion and three early revisions. Postoperative duplex scanning demonstrated residual stenosis in seven of 12 patients with unrepaired defects, two of 36 patients with repaired defects, and five of 312 patients with normal scan results (p < 0.001). Conclusion: Based on the types of lesions corrected and the low ( < 0.5%) complication rate after a normal or modified arterial reconstruction, duplex scanning was found to be a valuable intraoperative aid. Uurepaired defects require close surveillance for progression. (J VASC SURG 1994;20: ) Despite careful operative technique, vascular defects can reside within an arterial reconstruction and escape detection after blood flow is restored. Failure to recognize technical errors (intimal flap, plaque dissection, suture stricture, graft entrapment) or intrinsic abnormalities (mural thrombus, platelet aggregation, obstructed flow) can lead to vessel thrombosis or distal embolization. For carotid and From the Division of Vascular Surgery, Department of Surgery, University of South Florida College of Medicine, Tampa. Presented at the Eighteenth Annual Meeting of The Southern Association for Vascular Surgery, Scottsdale, Ariz., Jan , Reprint requests: Dennis Bandyk, MD, Harbourside Medical Tower, No. 730, 4 Columbia Dr., Tampa, FL Copyright 1994 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter /94/$ /6] visceral reconstructions the sequelae of unrecognized technical errors can be catastrophic. Although monitoring arterial reconstructions for technical adequacy is an accepted principle of vascular surgery, no diagnostic method has achieved universal acceptance. Most surgeons rely on arteriography to obtain an image of critical arterial segments for technical errors and on Doppler flow detection techniques to confirm the presence of unobstructed blood flow. These methods, though helpful, are not without diagnostic error. Routine completion angiography after infrainguinal vein bypass is still associated with a 3% to 7% early graft thrombosis rate) s More sophisticated monitoring techniques such as ultrasonography and angioscopy are capable of detecting a spectrum of lesions, but grading an identified lesion for its propensity to cause early repair failure versus remain stable, or regress is difficult. Surveillance studies have

2 Volume 20, Number 3 Bandyk et al. 427 shown many arterial reconstructions are imperfect with the incidence of residual anatomic defects in the range of 15% to 30%. 3,4 Of importance, most early failures and postoperative complications reside in the cohort with abnormal intraoperative studies. Because the goal of arterial surgical procedures is to correct or bypass a lesion and construct a neoarterial segment void of anatomic and hemodynamic abnormalities, intraoperative monitoring techniques should evaluate these characteristics to appraise technical precision. The refinement of duplex technology to include color Doppler imaging has provided vascular surgeons with a powerful diagnostic method for intraoperative application. Color duplex systems that display color-coded flow data within a high-resolution, real-time B-mode image facilitates vessel imaging and the detection of focal flow abnormalities compared with conventional gray-scale duplex scanning. Our experience with intraoperative ultrasound techniques has been favorable and prompted recommendation for its application in "difficult" arterial reconstructions such as carotid endarterectomy, infrainguinal in situ saphenous vein arterial bypass, and visceral/renal reconstructions. 46 Beginning in 1989 the primary author (D.F.B.) has used color-flow duplex scanning initially to evaluate carotid bifurcation endarterectomy sites at operation and subsequently to scan infrainguinal vein bypass and visceral artery reconstructions with the intent to develop threshold criteria for intervention when residual lesions were identified that would be applicable for a variety of arterial repairs. In this study we report 3 years of clinical experience with intraoperative duplex scanning, highlighting the spectrum of vessel and flow abnormalities that were identified and corrected, and the outcome of reconstructions with unrepaired defects. PATIENTS AND METHODS From March 1990 through November 1993, 368 patients (281 men, 87 women) undergoing elective carotid endarterectomy (n ), infrainguinal vein (saphenous, cephalic, basilic) arterial bypass (n = 135), or visceral (n = 10)/renal (n = 13) bypass were evaluated by color-flow duplex ultrasonography after blood flow was restored to assess technical adequacy. Because the frequency and nature of vascular defects could be related to surgical technique, the reconstructions examined had the following characteristics. After carotid endarterectomy, the bifurcation was reconstructed with saphenous/internal jugular vein patch angioplasty in 188 (89%) of carotid artery procedures. Ninety-two (68%) of the infrainguinal vein bypasses were constructed with the in situ or nonreversed translocated saphenous vein grafting techniques. The remaining infrainguinal bypasses used reversed saphenous (n -- 33) or upper extremity veins (n -- 10). The site of distal anastomosis was the above-knee popliteal artery in 16 cases, below-knee popliteal artery in 42 cases, or a tibial or pedal artery in 77 cases. All visceral/renal bypasses were constructed with reversed or nonreversed (valves excised) greater saphenous veins. Arterial repairs were visually inspected and patency confirmed by pulse palpation or audible continuous-wave Doppler flowmeter before duplex scanning. Papaverine (30 mg) was injected into infrainguinal vein bypasses to augment flow before (3 to 5 minutes) graft blood flow velocities were recorded. Completion arteriography was also performed in 110 (81%) of the infrainguinal bypasses to assess the distal anastomosis and runoff vessels. The entire arterial reconstruction was scanned for anatomic and flow abnormalities with any inadequately imaged segments noted. Identified vascular defects were further assessed by arteriography or reexploration of the reconstruction if judged to represent a hemodynamically significant flow abnormality/anatomic lesion (> 50% diameter reduction) by the operating surgeon (D.F.B., LL.M.). Based on previous experience with pulsed-wave Doppler spectral analysis and conventional duplex scanning, it was our policy to revise carotid artery repairs with focal flow abnormalities identified by spectral broadening and peak systolic velocity (Vp) in excess of 150 cm/sec. 4,s Criteria for revision of infrainguinal vein grafts were similar, but because of the smaller diameter vessels encountered with bypass grafting to infrageniculate arteries, a higher ( > 180 cm/sec) Vp threshold was adopted. Instrumentation and scanning technique. Intraoperative duplex scans were performed with one of three color duplex ultrasound systems (ATL Ultramark 9, Advanced Technology Laboratories, Bothell, Wash.; Accuson 128XP, Accuson Corporation, Smyrna, Ga.; QAD-1, Quantum Medical Systems, Issaquah, Wash.). All studies were conducted with a linear array probe (7 to 10 MHz) with or without an acoustic stand-off. The intraoperative study was scheduled through the vascular laboratory, and a registered vascular technologist accompanied the scanner to the operating room to assist in the examination. A sterile plastic sleeve filled with acoustic gel was used to cover the transducer, which was positioned directly over the arterial repair or

3 428 Bandyk et al. September 1994 Table I. Categories of duplex-identified residual lesions Stenosis category Peak systolic velocity (cm/sec) Velocity ratio (Vr) Clinical relevance Normal/mild < Normal flow pattern Moderate Residual flow abnormal, angiography? Severe > Repair defect High-grade > 300 > 4.0 Pulse deficit evident on cfinical exam, repair defect bypass graft with sterile saline solution in the wound or on the skin for acoustic coupling. Scanning was performed in both longitudinal and transverse planes with the technologist providing optimal instrument settings for color-flow Doppler imaging and sample volume placement for Doppler-angle corrected (60 degrees or less relative to vessel axis) velocity spectrum analysis. Representative centerstream velocity spectra were recorded at anastomotic sites, endarterectomy endpoints, vascular clamp occlusion sites, and along the length of vein conduits. Areas of colorcoded flow abnormality (increased velocity, aliasing, color-flow jet) were examined in detail by image enlargement and centerstream spectral analysis proximal to, at, and distal to the site of maximum flow disturbance. Vein graft lesions were further characterized by measurement of velocity ratio (Vr), where gr = gpa r lesion/rvpproximal. Lesions with flow patterns and velocity spectra of a stenosis were classified by criteria similar to the duplex classification of grading stenosis in the extracranial and peripheral arterial circulations (Table I). Lesions were considered for immediate revision if duplex scanning demonstrated both an abnormality on B-mode image, indicating an anatomic lesion, and abnormal velocity spectra of a severe or high-grade stenosis. Focal residual atherosclerotic plaque, or intimal flaps were subjected to site revision, if associated with a significant centerstream flow abnormality (high Vp or Vr across lesion). Data analysis. Residual lesions identified and corrected were tabulated for each type of arterial reconstruction studied. Adverse patient outcome data (stroke related to internal carotid artery [ICA], graft thrombosis, repair site revision, abnormal postoperative duplex scan) were recorded for the first postoperative month and analyzed relative to the duplex scan findings at surgery (i.e., normal scan results, corrected duplex-identified defect, or unrepaired defect). Postoperative duplex scans were performed on all patients after infrainguinal vein bypass at 1 week, 6 weeks, and 3 months; and at 2 to 3 months after carotid endarterectomy. Visceral/renal bypasses were evaluated at 2 to 3 months after the operation by duplex scanning, arteriography, or perfusion scans to assess patency and graft/renal function. Differences between groups were compared by chi squared analysis. RESULTS Duplex scanning identified technical or intrinsic conduit defects judged sufficiently severe to warrant immediate revision in 37 (10%) of the 368 reconstructions. Only three of the lesions identified and corrected had velocity spectra of a high-grade stenosis. No additional significant defects requiring revision were identified by arteriography alone. In two patients undergoing femorotibial bypass grafting, intraoperative duplex scanning was unable to adequately visualize the distal anastomosis because of acoustic shadowing produced by wall calcification of the runoff tibial artery. Adequate images of all carotid and visceral artery repairs were obtained, and no supplemental diagnostic arteriography was performed in repairs selected for revision based on the duplex findings. Normal duplex examinations were typically completed within 10 to 15 minutes. Identification of a residual lesion prompted a more detailed examination, occasionally including the use of arteriography for confirmation, and repeat graft scanning after revision. The three duplex scanners used in this study were found to be equally effective for intraoperative assessment. The frequency and type of defect varied with type of procedure (Table II). All lesions judged to warrant repair had an abnormality found at exploration, and the velocity at the revision site reverted to normal, or a mild/moderate residual flow disturbance persisted after repair. Infrainguinal vein bypasses had the highest incidence of abnormal intraoperative scans (18%), corrected defects (14%), and 30-day adverse events (3%). The incidence of abnormal duplex scan results was similar with in situ (15%) and reversed (22%) saphenous vein grafting methods. The only instance of graft thrombosis occurred after in situ femoral posterior tibial saphenous vein bypass that required intraoperative modification because of inadequate vein length and demonstrated a residual lesion (Vp = 160 cm/sec, Vr = 2.0) in the aboveknee vein segment. The duplex flow abnormality was

4 Volume 20, Number 3 Bandyk et al. 429 Table II. Incidence, location, and nature of residual lesions identified by intraoperative duplex scanning Scan category Residual lesions Type of procedure Normal Abnormal Corrected~Identified 30-Day adverse events Carotid endarterectomy (n = 210) 186 Infrainguinal bypass (n = 135) 111 Visceral/renal bypass (n = 23) (11%) Residual plaque - 12/16 Stroke - 0 CCA-5 Stenosis - 3/5 Thrombosis - 0 ICA-7 Platelet thrombus - 2/2 ECA-t2 Kinked ICA - 0/1 24 (18%) Retained valve - 8/8 Thrombosis - 1 vein - 21 Stenosis - 8/13 Revision - 3 anas. - 3 Platelet thrombus - 3/3 1 (4%) Distal renal art. thrombus 1 - immediate nephrectomy attributed to small vein diameter and not repaired, but at reoperation several hours after the primary procedure, platelet thrombus was found at the site, and a successful graft revision procedure accomplished. Three infrainguinal vein bypasses with an unrepaired defects including two with conduit modifications during the primary procedure required a secondary revision within 30 days for a duplexidentified lesion and low graft blood flow. At 30-days the assisted primary patency of the infrainguinal vein bypass group was 100%. No adverse events developed after carotid endarterectomy despite immediate correction of 12 lesions involving either the internal or common carotid arteries. Residual atherosclerotic plaque was the most common defect identified and typically was located in the distal external carotid artery (ECA). In four patients the operating surgeon decided not to remove the residual ECA plaque. Three additional patients with residual duplex flow abnormalities (Vp = 140 to 150 cm/sec, moderate spectral broadening) in the ICA were not revised because of adequate lumen caliber ( < 50% diameter reduction) on longitudinal and transverse color-flow imaging. The cause of the flow abnormality was suture stenosis in two cases and a tortuous ICA distal to the vein patch angioplasty in one patient. Postoperative duplex scans of these unrepaired defects demonstrated a similar or less severe flow abnormality. Duplex scans obtained at 2 to 3 months after the procedure identified no ICA occlusions or high-grade (> 75% diameter reduction) stenoses in the ICA or common carotid artery (CCA). ECAs with both repaired and unrepaired defects remained patent. One of 23 visceral/renal bypasses was abnormal by intraoperative duplex scanning. Low flow (Vp < 50 cm/sec, absent antegrade diastolic flow) was present in the graft, and imaging of the distal renal artery showed thrombus within the lumen. Exploration of the repair confirmed the presence of chronic thrombus, and a nephrectomy was performed. At 30 days all visceral/renal repairs were patent and functional based on results of clinical examination, duplex scanning, renal perfusion scanning, or, in one case of bilateral hypogastric artery aortorenal bypasses, arteriography. Duplex scanning identified anatomic and hemodynamic abnormalities caused by platelet aggregation within the reconstruction in two (1%) carotid artery repairs and three (2.7%) saphenous vein infrainguinal bypasses. As previously stated the lesion was recognized in one bypass but was erroneously judged to be insignificant, and progression to thrombosis occurred. In the preocclusive phase color-flow duplex scanning of platelet aggregation demonstrated an irregular, poorly visualized flow lumen, color map aliasing, and peak velocities ranging from 200 cm/sec to greater than 400 cm/sec. The three instances of high-grade residual stenosis were the result ofplatelet aggregation within the repair. One vein bypass with platelet aggregation within the thigh graft segment had similar abnormalities by duplex scanning at downstream valve sites and at the distal anastomosis. These defects were confirmed with arteriography and successfully treated with urokinase infusion. After interposition grafting of the abnormal vein graft segment was performed, the duplex scan normalized distally, and graft patency was sustained. In the 360 (98%) patients with postoperative duplex scans, residual stenosis was identified in seven of 12 patients with unrepaired defects, two of 36 patients with repaired defects, and five of 312 patients with normal scans (p < 0.001). Five patients with residual lesions (CCA, one, ICA, one, ECA, one, vein graft, two) showed improvement in the velocity spectra at the site of intraoperative abnormality. Of note, three of five unrepaired vein graft lesions (Vp, 140 to 180 cm/sec,vr, 2.2 to 2.8) required subse-

5 430 Bandyk et al. September 1994 quent correction because ofgraft thrombosis (n = 1) or progression to a more severe lesion with decrease in overall graft blood flow velocity. No patient with a normal duplex scanning result required a secondary procedure during a minimum 2-month follow-up interval. DISCUSSION A spectrum of complex arterial reconstructions was successfully imaged and interrogated for focal abnormalities at operation with color duplex ultrasonography. Residual defects were identified in 13% of the arterial reconstructions, and the operating surgeon elected to repair 76% of the duplexidentified lesions. We believe the low (two of 352) incidence of adverse events in reconstructions with normal and upgraded duplex scan results attests to the efficacy of intraoperative duplex monitoring. Both technical errors and intrinsic vessel abnormalities were identified, including the recognition of platelet aggregation within the repair during its preocclusive phase. Prompt discovery and treatment of this lesion was a major reason for the low thrombosis rate associated with carotid endarterectomy (0%) and infrainguinal bypass (0.7%). Platelet aggregation on damaged or diseased luminal surfaces has been documented as an important cause of stroke and early vein graft thrombosis. 1,s Duplex scanning can identify the focal flow disturbance produced by the accumulation of platelet-fibrin aggregates. Angiographic diagnosis of platelet thrombus can be subtle, but the duplex signature of this lesion was velocity spectra of a high-grade stenosis. The incidence of residual lesions correlated with the complexity of the arterial reconstruction, being highest for infrainguinai vein bypass, less for carotid endarterectomy, and least after visceral/renal bypass. Arteriography did not identify any additional lesions after infrainguinal vein bypass but remains a useful method to confirm the presence of a lesion, assess the outcome of graft revision procedures, and image distal runoff vessels. Based on the patient outcome data and postoperative duplex results, intraoperative duplex scanning affords the best oppornmity to construct a technically precise arterial reconstruction and thus minimize procedure-related complications. An algorithm based on the severity of duplexidentified flow disturbance and abnormalities seen on color-flow Doppler imaging can provide the necessary information to assess technical adequacy and implement revision (Fig. 1). The threshold criteria for intervention should be based on quantitative velocity spectral analysis and varies little with the type of reconstruction. Based on the experience acquired in this and previous studies, we recommend immediate revision of lesions demonstrating focal severe spectral broadening, a Vp greater than 180 cm/sec, and a Vr of 2.5 or greater. After infrainguinal bypass, only high-grade residual lesions were associated with low (< 40 cm/sec) velocity in normal-sized (4 mm diameter) vein graft segments. This study provides preliminary data regarding the fate of unrepaired lesions, but the number of lesions followed was small (n = 12), and an expanded, more rigorous study needs to be conducted. The observation that only four of 12 unrepaired defects regressed in severity by duplex scanning and three of five unrepaired lesions in vein grafts required early revision suggests that a more aggressive policy regarding immediate repair may be warranted. Pulsed Doppler spectral analysis, though accurate in screening for residual defects, can be associated with false-positive study results in the setting of spasm and hyperemic flow. Elevated duplex-derived blood flow velocities were recorded in small ( < 3.5 mm) diameter vein conduits and distal to anastomotic sites to the tibial/renal arteries in response to the hyperemic flow of successful limb/renal revascularization. After lower limb bypass grafting is performed, we inject papaverine into the graft before duplex imaging to augment blood flow and relieve vessel spasm. Thus the measurement of peak systolic velocities in the range of 125 to 160 cm/sec was by itself not diagnostic of a significant residual lesion, and several small diameter grafts with this hemodynamic finding have not required subsequent revision. Similarly, Vp of 150 to 200 cm/sec can be recorded from tibial arteries distal to the bypass. In the setting of a high Vp but no apparent lumen abnormality by B-mode imaging or lumen narrowing by color Doppler, the measurement of a Vr less than 2 across the lesion and completion angiography are useful methods to exclude an important residual lesion. A normal velocity ratio indicates the high Vp is the result of hyperemic flow rather than focal stenosis. The intraoperative application of duplex scanning is possible in most vascular surgeons' practices. Availability of the instrumentation is not a major problem, because duplex scanning has become the standard in a number of clinical applications and thus should be available in all hospitals where vascular surgical procedures are performed. Although all vascular surgeons may have some familiarity with duplex technology and interpret noninvasive vascular laboratory studies, intraoperative application re-

6 Volume 20, Number 3 Bandyk et al. 431 FLOW DISTURBANCE B-MODE U/S IMAGE REQUIRED ACTION None/Mild Moderate Severe... No defect No defect...i Abnormal l No defect Abnormal None Arteriogram Immediate Revision Fig. 1. Diagnostic algorithm with color-flow duplex scanning for intraoperative assessment. Recommended action is based on velocity spectral changes graded as mild, moderate, or severe residual stenosis and vessel lumen abnormalities on real-time B-mode imaging. quires the surgeon to develop hands-on experience with the instrument and be familiar with the pitfalls of color-flow Doppler imaging, including recognition of artifacts, aliasing, and improper Doppler angle assignment of sample volume placement for pulsed-wave Doppler spectnam analysis. Knowledge of sonographic vascular anatomy, good hand-eye coordination, and the ability to mentally visualize the three-dimensional aspects of the arterial repair are other essential aspects of intraoperative scanning. These skills can be acquired, usually with the assistance of vascular laboratory personnel. InitiaUy, intraoperative duplex scanning should be used to evaluate less complex repairs such as carotid endarterectomy. An image of this reconstruction is easily obtained, less artifacts are encountered, and the velocity spectra of a normal repair mimics that of an undiseased carotid bifurcation. With experience more complex reconstructions can be evaluated. Distinct advantages of duplex scanning compared with modalities such as arteriography and angioscopy are the ability to evaluate blood flow hemodynamics and repeat the examination, multiple times if necessary, to assess for lesion for progression and its resolution after correction. When a residual lesion is detected the surgeon must determine its significance and decide regarding the safety and benefit of reexploration. For severe le- sions such as those associated with a pulse deficit or a greater than 60% stenosis on arteriography, revision is recommended because the likelihood of thrombosis is high. For less severe lesions the surgeon should consider the functional status of the patient and whether technical problems were encountered during the procedure. The incidence of duplex-identified flow abnormalities is expected to be higher in complex reconstruction, such as when endarterectomy endpoints required tacking sutures, a small diameter vein was used, or an anastomosis was to a calcified artery. In such instances the surgeon must balance the ability to repair the defect with the likelihood that the lesion will lead to a significant problem. Reports on the fate of unrepaired lesions encountered after carotid endarterectomy indicate a favorable prognosis for lesions with less than 30% diameter reduction. 7 Sawchuck et al., s using intraoperative B-mode ultrasonography, found that 16 of 19 carotid arteries with minor ( < 1 mm in ICA; < 3 mm in CCA) flaps were normal at the time of the first postoperative duplex scan. Using duplex scanning Baker et al. 9 found a similar operative stroke incidence in repairs with normal (1.6%) scans or minor defects (1.9%). The "acceptibility" of a residual lesion depends on the sensitivity of the monitoring technique and the surgeon's confidence in its diagnostic accuracy. We have found that intraoperative duplex scanning will

7 432 Bandyk et al. September 1994 detect a spectrum of abnormalities, but only the most severe require correction. The relation between technical precision and durability of arterial reconstructions warrants further study. Duplex surveillance studies have indicated a higher failure rate in reconstructions with repaired and unrepaired defects. Both Baker et al. 9 and Kinney et al.~0 reported a greater incidence of ICA occlusion/restenosis (p < 0.007) in carotid artery repairs with unrepaired defects. Similarly, after infrainguinal bypass grafting the failure rate of modified vein grafts was increased compared with that of grafts not found to have an intraoperative defect. ~ Our infrainguinal vein graft surveillance studies indicate that most postimplantation occlusive lesions have their origin as unrepaired defects or develop in grafts with abnormal or spliced vein segments. Thus an added benefit of intraoperative duplex scanning is its ability to identify the "high-risk" arterial reconstructions, which are prone to develop myointimal lesions. Arterial repairs with normal intraoperative duplex scans appear to be at low risk for failure and frequent surveillance is not necessary. By contrast, careful duplex surveillance of arterial reconstructions with a residual defect is recommended to assess whether the lesion regresses or progress on serial scans, indicating the need for correction. REFERENCES 1. Bandyk DF, Towne JB, Schmitt DD, et al. Therapeutic options for acute thrombosed in situ saphenous vein arterial bypass grafts. J VASC SURG 1990;11: Mills JL, Fujitani RM, Taylor SM. Contribution of routine intraoperative completion arteriography to early infrainguinal bypass patency. Am J Surg 1992;164:506-11, 3. Miller A, Maracaccio EJ, Tannenbaum GA, et al. Comparison of angioscopy and angiography for monitoring infrainguinal bypass vein grafts: results of a prospective randomized trial. J VAsc SURG 1993;17: Bandyk DF, Jorgensen RA, Towne JB. Intraoperative assessment of in situ saphenous vein arterial grafts using pulsed Doppler spectral analysis. Arch Surg 1986;121: Bandyk DF, Kaebnick HW, Adams MB, et al. Turbulence occurring after carotid bifurcation endarterectomy: a harbinger of residual and recurrent carotid stenosis. J VAsc SUV, G 1988;7: Bandyk DF, Govostis DM. Intraoperative color flow imaging of "difficult arterial reconstructions." Video J Color Flow Imaging 1991;1: Courbier R, Ferdani M, Reggi M. Fate of unrepaired defects after carotid endarterectomy. In: Berstein EF, ed. Vascular diagnosis. 4th ed. St. Louis: Mosby, 1993: Sawchuck AP, Flanigan DP, Machi J, et al. The fate of unrepaired minor technical defects by intraoperative ultrasonography during carotid endarterectomy. J VAse SuRe 1989;9: Baker WH, Koustas G, Burke K, et al. Intraoperative duplex scanning and late carotid artery stenosis. J VASC SURe 1994;19: Kinney EV, Seabrook GR, Kinney LY, et al. The importance ofintraoperative detection of residual flow abnormalities after carotid endarterectomy. J VAsc SURG 1993;17: Bergamini TM, Towne JB, Bandyk DF, et al. Durability of the in situ bypass following modification of abnormal vein segment. J Surg Res 1993;54: Submitted Feb. 3, 1994; accepted May 25, DISCUSSION Dr. Steven J. Burnham (Chapel Hill, N.C.). I would be interested to know about the time involved for this addition to each of the types of procedure. I am also interested in the time commitment for the vascular technologist and for the duplex scanner. Along the same lines, perhaps you could tell who holds the transducer, who interprets the data from the duplex scan, and tell us about the nature of the interaction between the surgeon and the vascular technologist relative to the completeness of the examination. Second, what advice do you offer to those of us contemplating purchase of a color-flow duplex scanner for the operating room? And how long will it take us to become proficient? How can we best train our residents and our fellows to do this well? Can they master the technique from watching you in the operating room, or do they need additional tutoring in the noninvasive vascular laboratory? Dr. G. Patrick Clagett (Dallas, Texas). You implicate platelet thrombus formation, and I am always a little skeptical when people blame things on platelets. What was your antithrombotic regimen in these patients? Were they receiving perioperative aspirin? What was your heparin regimen and policy with regard to protamine reversal? Could you please clarify that? Dr. Ronald J. Stoney (San Francisco, Calif.). There are a few concerns with this technology. One is that it takes 8 or 10 minutes to do it, at least in our own hands. Second, it sometimes requires the translocation or transportation of your duplex scanning device to the operating room if you do not have a dedicated unit there. We found that different heads create some problems. What type of head and what Megahertz head do you use for interrogating the various

8 1OURNAL OF VASCULAR SURGERY Volume 20, Number 3 Bandyk et al 433 arteries that you evaluated in the neck, abdomen, and extremity. We are bothered by one thing, and that is what I call minor defects with minimal to no velocity changes. We rend to ignore them, but because we are looking at the outcome of an endarterectomy, we are quite concerned about the appearance of flaps, particularly flaps that are distally based, or residual disease even in vessels with high flow like a renal or a visceral artery. What would you accept in the carotid artery after endarterectomy? We have opened vessels for fairly minimal to no findings. It is a very sensitive technology, and if you are not careful, you can kind easily overinterpret disease of no significance when obtaining an image of a repaired vessel. Dr. P. Kevin Zirkle (Knoxville, Tenn.). It seems that many of the severe lesions can be identified with continuous-wave Doppler scanning and a surgeon that is used to using this. It seems that many of the normal reconstructions can be identified that way, and I just wonder what percent of these lesions really require this degree of technology. It is certainly an elegant way to do it, but it adds a tremendous amount of cost. Are we talking about 1% or 2% gain here? Or 10% or 20%? Dr. Dennis F. Bandyk. From a practical standpoint you do need dedicated support from the vascular laboratory. When the intraoperative scan result is normal, the scanner and the technologist are only in the operating room for about 15 minutes. When the scan result is abnormal repeat imaging is recommended and this circumstance can disrupt the subsequent vascular laboratory schedule. There is a learning curve to performing intraoperative duplex scanning, My recommendation is to get hands-on experience with the help from your technologists. In the operating room, begin by evaluating simple arterial repairs such as carotid endarterectomy. This procedure is performed relatively frequently, and the number of artifacts is very low. As surgeons we attempt to construct a carotid artery repair that mimics a normal bifurcation, and this endpoint can be recognized easily by duplex scanning. Our vascular fellows spend a half day per week in the vascular laboratory during the research year obtaining hands-on experience with duplex scanning. I believe the vascular surgeons of the furore will be very comfortable using this technology in the operating room. Regarding the development of platelet thrombus that was recognized by duplex scanning, all the patients had received antiplatelet drugs. This lesion tended to develop in abnormal veins: veins where a side branch was occluded or when we were trying to make an abnormal vein work. In the case ofendarterectomy, typically a deep endarterectomy plane was developed. When platelet aggregation occurred it was a progressive lesion. If the scan was performed 5 minutes after blood flow was established, a velocity of 200 cm/sec was recorded at the lesion, when 15 minutes lapsed, it increased to 300. In fact, all high-grade lesions identified were the result of platelet aggregates accumulation. Dr. Stoney, we used 7 to 10 MHz linear array probes. We will be seeing improved probes for intraoperative use as ultrasound instrument manufacturers recognize this duplex application, and surgeons take the scanner into the operating room with greater frequency. We have also seen intimal remnants particularly in the B-mode imaging, but the use ofhemodynamic criteria (i.e., spectral analysis) has permitted us to classify lesions and select those that should be repaired immediately or be safely followed. If a severe focal flow disturbance is identified adjacent to a residual intimal flap, the lesion should be repaired. Dr. Zirkle's comments regarding the usefulness of continuous-wave Doppler flow analysis, a simpler technique, are appreciated, but with this instrument and audible interpretation of the waveform, I cannot distinguish a lesion with a peak velocity of 150 cm/sec from one with a velocity of 190 cm/sec. Duplex scanning gives the user more objective hemodynamic criteria. Most of the lesions we repaired had velocities in the range of 200 to 250 cm/sec. Very few of them are associated with a pulse deficit or a high-pitched velocity spectra, which would indicate a critical stenosis. Duplex scanning permits identification and correction of serious residual lesions, and the observation of less severe lesions. Continuous wave Doppler flow analysis is not sufficiently sensitive to permit this degree of lesion classification.