Do normal early color-flow duplex surveillance examination results of infrainguinal vein grafts preclude the need late graft revision? Marc A. Passman, MD, Gregory L. Moneta , MD, Mark R. Nehler , MD, Lloyd M. Taylor , Jr., MD, James M. Edwards , MD, Richard A. Yeager, MD, Dona ld B. McConnel l , MD, and John M. Por te r , MD, Portland, Ore.

Purpose: Optimal duration of postoperative duplex surveillance ofinfrainguinal vein grafts is not known. Previous reports have suggested nearly all vein graft stenoses are present within the first postoperative year, and normal duplex examination results during this time eliminate the need for ongoing graft surveillance. To determine whether surveillance may be safely discontinued in patients with normal early postoperative surveillance studies, we reviewed the color-flow surveillance examinations in our patients who underwent infrainguinal reverse vein graft revisions during a 41/2 year period. IVicthods: Clinical and vascular laboratory records were reviewed of all patients who underwent infrainguinal reverse vein bypass grafting followed by subsequent graft revision for a duplex scanning-detected abnormality at our institution between January 1990 and July 1994. Results: Of 447 infrainguinal reverse vein bypasses performed, 36 (8.1%) underwent surgical revision as a result of an abnormal finding during routine duplex surveillance. The initial postoperative duplex examination was obtained within 2 weeks of graft implanta- tion in 23 (64%) patients, between 2 weeks and 3 months in 10 (28%) patients, and between 3 and 6 months in three (8%) patients. Duplex abnormalities prompting revision included 11 (31%) grafts with a mid-graft peak systolic velocity (PSV) _< 45 cm/sec, 23 (64%) grafts with a focal PSV -> 200 cm/sec, one graft with a PSV -> 150 cm/sec but < 200 cm/sec, and one thought to be occluded by duplex but found to be patent by angiography. Abnormal duplex findings were initially detected within 2 weeks of graft implantation in five (14%) patients, between 2 weeks and 3 months in eight (22%) patients, from 3 to 6 months in 12 (33%) patients, from 6 to 12 months in six (17%) patients, and > i year in five (14%) patients. In only 25% of cases were mid-graft PSVs <45 cm/sec or focal velocities _> 200 cm/sec identified on the initial examination; 75% were found during subsequent surveillance. Conclusions: Although most reverse vein graft abnormalities detected by duplex surveil- lance and prompting graft revision appear within the first postoperative year, many are not detected on the initial examination. In our recent experience 31% of duplex abnormalities leading to vein graft revision were first detected more than 6 months after operation. Discontinuation of graft surveillance based on normal early findings will result in thrombosis of some vein grafts that may otherwise be salvaged. (J VAsc SvR~ 1995;22:476-84.)

for

From the Division of Vascular Surgery, Department of Surgery, Oregon Health Sciences University, and Portland Veterans Affairs Hospital, Portland.

Supported in part by NIH Grant RR-00334 and General Clinical Research Centers Branch National Center for Research Re- s o u r c e s .

Presented at the Tenth Annual Meeting of the Western Vascular Society, Phoenix, Ariz., Jan. 15-18, 1995.

Reprint requests: Gregory L. Moneta, MD, 3181 S.W. Sam Jackson Park Road, OP-11, Portland, OR 97201.

24/6/66339

476

Infrainguinal vein graft failure results primarily from stenotic lesions in the vein graft or from progression of disease in the inflow or outflow arteries. 14 An estimated 20% to 37% of all infrain- guinal vein bypass grafts develop graft-threatening lesions within 2 years. 5-7 The ability o f duplex graft surveillance to detect these lesions before they result in graft thrombosis is well documented. 8~13

JOURNAL OF VASCULAR SURGERY Volume 22, Number 4 Passman et al. 4 7 7

The optimal duration of postoperative duplex surveillance of an infrainguinal vein graft is not known. It is generally accepted that most vein graft stenoses develop within 2 years after operation. 2,14,15 However, recently authors have suggested that al- most all vein graft lesions are detectable in the early postoperative period and that less intensive duplex surveillance or even discontinuation of surveillance may be justified, if early studies do not reveal a graft abnormality. 16,17 This approach assumes that early examinations are 100% specific in excluding the discovery of future abnormalities and that vein grafts do not become occluded if results of early duplex surveillance studies are normal. To evaluate whether early, normal, postoperative duplex surveillance ex- amination results preclude the need for late graft revision, we reviewed the duplex surveillance findings in our patients who underwent revision of infraln- guinal vein grafts to maintain patency.

MATERIAL AND M E T H O D S

All patients who underwent both placement of an infrainguinal reverse vein bypass and subsequent graft revision because of abnormal findings on duplex surveillance examinations at either the Oregon Health Sciences University Hospital or Department of Veterans Affairs Medical Center in Portland, Oregon, between January 1990 and July 1994 were identified through our computerized vascular surgery registry. Only patients whose graft revision was prompted by findings on routine duplex surveillance were included in this study.

During this time vein grafts at our institution were monitored with a surveillance program based on color flow duplex examination of the vein graft at 3 monthly intervals for i year and 6 monthly intervals thereafter. All duplex examinations were performed by registered vascular technologists with an Acuson XP-128 color flow duplex scanner (Acuson Inc, Mountain View, Calif.). Attempts were made to visualize and insonate the entire graft, its proximal and distal anastomotic sites, and the adjacent inflow and outflow arteries. When surveillance examina- tions revealed a mid-graft peak systolic velocity (PSV) _< 45 cm/sec or a focal PSV _ 200 cm/sec, angiography was performed in spite of the absence of symptoms or drop in ankle/brachial index (ABI). Although some authors prefer velocity ratios for detecting graft abnormalities, 9,n we prefer both an absolute high- and low-velocity criteria, which has previously been shown to be highly sensitive in detecting > 50% diameter reduction in our vascular laboratory. 18'~9 If the angiogram confirmed the

presence of a stenotic lesion with > 50% diameter reduction, vein graft revision was performed.

Each patient's record was reviewed for the indication for the initial bypass procedure, its proxi- mal and distal anastomotic sites, and the conduit used. The vascular laboratory record was reviewed to determine the time of the initial postoperative duplex surveillance examination, all subsequent duplex find- ings, and spedfically the timing of the duplex finding that led to graft revision. Duplex findings were also correlated with patient symptoms (increasing dau- dication, ischemic rest pain, or new ulceration) and with a change in ABI __ 0.2.

RESULTS

Patients. Between January 1990 and July 1994, 447 infrainguinal bypass operations were performed at our institution. Thirty-six (8.1%) of these grafts underwent surgical revision as a direct result of an abnormal color flow duplex surveillance examination. Thirty-one (86%) men and five (14%) women with a mean age of 66.1 years (range 43 to 86 years) were studied.

Initial bypass. The original indication for infrain- guinal reverse vein bypass was claudication in seven patients, ischemic rest pain in 18 patients, and ulceration or gangrene in 11 patients. The proximal anastomotic site was the common femoral artery in seven (19%) patients, the superficial femoral artery in 14 (39%) patients, and the deep femoral artery in 15 (42%) patients. The distal anastomotic site was the above-knee popliteal artery in three (8%) patients, the below-knee popliteal artery in 24 (67%) patients, and a tibial artery in nine (25%) patients. Thirty-one (86%) of 36 grafts consisted of a single reverse greater saphenous vein segment. Alternative conduits included a single-arm vein segment in one patient, composite arm vein segments in three patients, and composite prosthetic-greater saphenous vein in one patient who subsequently had a duplex abnormality in the venous portion of the graft.

Duplex surveillance. The average time from infrainguinal bypass to first duplex surveillance evalu- ation was 1.3 months (range 1 week to 5.8 months). Ninety-two percent of the initial duplex evaluations were obtained within 3 months of the bypass operation, with most (64%) performed within 2 weeks of graft implantation No intraoperative du- plex studies were performed.

Duplex findings prompting angiography and graft revision included 11 (31%) grafts with a low mid-graft PSV ( <_ 45 cm/sec), 23 (64%) grafts with an increased focal PSV (---200 cm/sec), one graft

4 7 8 Passman et al.

JOURNAL OF VASCULAR SURGERY October 1995

14

12

1 0 . . . . . . . . . . . . . .

6

0 <2 weeks 3 months 6 months 9 months 12 months > 1 year

Fig. 1. Timing of detection for duplex findings that led to surgical revision in 36 patients.

PSV 100-150cm/s 17.4%

'SV 150-200cm/s 13.0%

PSV <lOOcm/s 43.5%

/>200cm/sec 26.1%

Fig. 2. Initial duplex velocity in location of 23 vein grafts that were subsequently revised for PSV - 200 cm/sec.

with a moderate focal increase in PSV ( -> 150 cm/sec, but < 200 cm/sec), and one thought to be occluded by duplex surveillance but found to be patent by angiography.

The average time from infrainguinal bypass to the development of the duplex abnormality leading to graft revision was 5.8 months (range 1 week to 22 months). Duplex findings leading to graft revision were first detected within 2 weeks of graft implan- tation in five (14%) patients, between 2 weeks and 3 months in eight (22%) patients, from 3 to 6 months in 12 (33%) patients, from 6 to 12 months in six (17%) patients, and > 1 year in five (14%) patients (Fig. 1). Only nine (25%) duplex findings leading to graft revision were discovered on the initial duplex examination, and 27 (75%) were found during subsequent surveillance.

For the 23 patients in whom a focal PSV --- 200

cm/sec led to graft revision, the mean PSV obtained immediately before angiography and graft revision were performed was 347 _+ 113 cm/sec (range 202 cm/sec to 642 cm/sec). Six (26%) of these patients had a PSV -> 200 cm/sec on the initial surveillance examination; 74% of the grafts on their initial duplex examination had a PSV < 200 cm/sec in the area that subsequently developed a focal PSV --200 cm/sec (Fig. 2).

Clinical correlation. At the time the duplex abnormality leading to revision was detected, a decrease in ABI of -> 0.2 was found in only 11 (31%) of 36 extremities. Twenty-five (69%) patients were asymptomatic, five (14%) had mild clandication, and six (16%) had either persistent or recurrent ischemic rest pain or ulceration. In all six of the patients with rest pain or ulceration, graft lesions were detected within 3 months of the initial bypass.

JOURNAL OF VASCULAR SURGERY Volume 22, Number 4 Passman et al. 479

Table I. Correlation of duplex surveillance abnormality leading to angiography and operation in 36 patients undergoing revision of an infrainguinal reverse vein graft or its inflow or outflow artery

Location of stenosis by angiography Duplex surveillance

abnormality Proximal Proximal Mid Distal Distal Total prompting revision Inflow anastomosis graft graft graft anastomosis Ou~low Occluded (%)

MID PSV <45 cm/sec 1 1 5 2 2 I1 (30.5) Focal PSV >200 cm/sec 6 5 7 2 1 1 1 23 (63.9) PSV 150-200 cm/sec 1 I (2.7) Pseudoocclusion 1 I (2.7)

Total (%) 1 (2.7) 9 (25.0) 1---6 (27.8) 7 (19.4) 4 (11.1) f (2.7) 3 (8.3) i (2.7) 36

Angiographic correlation. Prerevision anglo- gram confirmed the presence of a stenosis in the graft or an adjacent inflow/outflow artery in aH cases. The stenosis was -> 70% in 31 (86%) of the 36 grafts and _ 60% in all of the grafts.

Table I shows the location of the stenotic lesions identified by angiography and their correlation to the duplex surveillance finding immediately before angi- ography. The most frequently encountered lesion was an isolated intrinsic graft stenosis, accounting for 21 (58%) of the lesions, followed by 10 (28%) anas- tomotic lesions and four (11%) inflow/outflow ab- normalities. One graft occluded by angiography had been previously patent on duplex surveillance with a PSV -> 200 cm/sec at the proximal anastomosis.

Of the 11 grafts that underwent angiography and revision based on duplex findings of a mid-graft PSV ___ 45 cm/sec, six (55%) had an angiographic stenosis at the proximal anastomosis or in the proximal graft. In all six of these grafts the proximal anastomosis was to the deep femoral artery of an anatomically tunnelled reverse vein bypass. Of the remaining five grafts revised based on a mid-graft PSV ___ 45 cm/sec, three had stenoses in the inflow or outflow arteries, and only two had a stenotic lesion in the graft itself.

Late graft stenosis. Beyond the first postopera- tive year, five (14%) vein grafts developed duplex findings that led to graft revision. Four of the five patients were asymptomatic at the time the duplex finding was first detected. Three of these grafts had a focal PSV ---200 cm/sec with stenotic lesions con- firmed by angiography in the proximal graft (one), the mid-graft (one), and at the distal anastomosis (one). In the other two grafts a mid-graft PSV --- 45 cm/sec led to angiography and revision. A stenotic lesion was found in the inflow artery in one of these grafts and at the proximal anastomosis in the other. Between four and six duplex evaluations (average 4.6) were performed on each of these five grafts before the detection of a duplex velocity that sug- gested a stenotic lesion. None of these previous

duplex evaluations had suggested a low-grade steno- sis that progressed with time.

DISCUSSION

In 1973 Szilagyi ct al. 1 reported 377 infralnguinal vein grafts evaluated prospectively with annual angi- ography. In this classic study 33% of all patent bypasses developed some intrinsic structural defect, usually within 2 years of initial operation. The observation that most stenotic lesions showed pro- gression of stenosis leading to graft occlusion estab- fished the association of graft stenosis with graft failure. More recently numerous investigators have determined vein graft stenoscs can be detected noninvasively with duplex scanning and that, as suggested by Szilagyi's work, most will be detected relatively early after operation. 5qa,15 In addition, it appears revision of duplex-detected graft stenoses results in excellent assisted primary patency 2°24 and overall higher assisted primary patency rates than can be obtained when graft revision is prompted only by clinical findings. 2531 As a result postoperative sur- veillance of infrainguinal vein grafts has become standard practice in most vascular surgical centers.

Nevertheless significant questions remain regard- ing postoperative duplex surveillance ofinfrainguinal vein grafts. Among these questions is how long should duplex surveillance be continued after opera- tion? Mills et al. 17 reported that of 135 infrainguinal vein bypasses with a mean follow-up of 1 year, 33% within 3 months of operation had a duplex-detected flow abnormality defined as a peak systolic velocity greater than 150 cm/sec. In grafts with PSVs less than 150 cm/sec the incidence of subsequent duplex- detected stenosis was so low the authors concluded that less intensive surveillance in this group was justified. However, the author concludes in the discussion after this report that, "The point of this study is that normal grafts do not get stenosis; they develop either from early appearing or preexisting areas of abnormality in the conduit." Taylor et al. 16

JOURNAL OF VASCULAR SURGERY 4 8 0 Passman et al. October 1995

Fig. 3. High-grade intrinsic stenosis in proximal region of femoral to below-knee popliteal reverse saphenous vein graft. Lesion was first detected as PSV of 445 cm/sec in the proximal graft 14 months after initial bypass. Previous duplex evaluations at i week and 4, 7, and 10 months had shown PSVs between 96 and 178 cm/sec in this portion of graft.

noted stenosis in 16% of 412 femorodistal grafts. More than 60% of stenoses were detected by 6 months after operation, and no new stenoses were detected beyond 1 year, leading to the conclusion that, "if a policy of strict surveillance during the first year after operation is adhered to, further duplex surveillance is not justified." The implication from both of these studies is that "normal" early duplex surveillance examinations have 100% specificity for excluding a missed early or later developing lesion that threatens graft patency.

In our series of 447 infrainguinal reverse vein bypasses performed during a 41/2-year period, 8.1% underwent revision because of a duplex surveillance abnormality. As in other studies ls,21'27 most vein graft revisions based on surveillance findings were per- formed for stenoses detected during the first postop-

erative year. However, in 75% of our 36 patients the surveillance studies were initially "normal" in the same area that subsequently developed stenosis. Some grafts that subsequently required revision were without a detected focal stenosis (PSV ___ 200 cm/sec) or a mid-graft PSV -< 45 cm/sec on four or more serial examinations during the course of more than i year. Some of the delayed stenoses may represent lesions that were repeatedly missed. This finding does not detract from the observation that many lesions of sufficient severity to ultimately lead to graft throm- bosis 6,9,11 were detected, because surveillance was continued on a regular basis beyond 3 months after operation. In addition, in the absence of angio- graphic verification it is possible studies purporting to find no late graft lesions are missing lesions that are actually present but as of yet have not resulted in graft thrombosis.

In this series 31% of the grafts were revised based on a mid-graft PSV _< 45 cm/sec rather than a high focal PSV. All of these "low flow" grafts had a signif- icant stenosis confirmed by angiography. In nine of the 11 grafts the stenosis was either in the inflow or outflow vessel or in the proximal portion of a graft with the proximal anastomosis to the deep femoral artery. This finding suggests duplex surveillance may lack sensitivity for directly detecting a focal stenosis in anatomically placed deep grafts or in inflow and out- flow vessels. Such stenoses would, however, be ex- pected to have an adverse effect on graft patency. It is clear a mid-graft PSV - 45 cm/sec does not exclude a significant graft stenosis 5'32 and that a mid-graft PSV _< 45 cm/sec does not prove the presence of a high- grade lesion) 3 Nevertheless we continue to regard low mid-graft PSVs as an important finding in our surveillance examinations especially if higher mid- graft velocities were present previously.

Because surveillance duplex examinations involve no patient risk and no perfect arterial substitute has yet been found, it is in our opinion medically justified to continue surveillance of vein grafts as long as the graft is patent and revision is practical. Presently, the cost-versus-benefit relationship for postoperative du- plex surveillance is unknown. Clearly, in this study, if duplex graft surveillance had been discontinued after 1 year of "normal" duplex examination results, five lesions that potentially threatened graft patency would have been missed (Fig. 3). Discontinuation of duplex surveillance based on early normal examina- tion results will likely result in thrombosis of some vein grafts that could otherwise be salvaged with more prolonged surveillance. As long as resources

JOURNAL OF VASCULAR SURGERY Volume 22, Number 4 Passman et aL 481

permit, we will continue to recommend to our patients regular postoperative duplex vein graft surveillance for the life of the vein graft.

R E F E R E N C E S

1. Szilagyi DE, Elliot JP, Hageman JH, Smith RF, Dall'olmo CA. Biologic fate of autogenous vein implants as arterial substitutes: clinical, angiographic, and histopathic observa- tions in femoro-popliteal operations for atherosclerosis. Ann Surg 1973;178:232-46.

2. Whittemore AD, Clowes AW, Couch NP, Mannick JA. Secondary femoropopliteal reconstruction. Ann Surg 1981; 193:35-42.

3. Donaldson MC, Mannick JA, Whittemore AD. Causes of primary graft failure after in situ saphenous vein bypass grafting. J VASC SuRe 1992;15:113-20.

4. Mills JL, Fujitani RM, Taylor SM. The characteristics and anatomic distribution of lesions that cause reversed vein graft failure: a five-year prospective study. J Vasc SURG 1993;17: 195-206.

5. Chang BB, Leather RP, Kaufman JL, Kupinski AM, Leopold PW, Shah DM. Hemodynamic characteristics of failing infraingninal in situ vein bypass. J Vasc SURG i990;12:596- 600.

6. Buth J, Disselhoff B, Sommeling C, Stam L. Color-flow duplex criteria for grading stenosis in infrainguinal vein grafts. J VAse SuRe 1991;14:716-28.

7. Bandyk D, Seabrook G, Moldenhauer P, Lavin J, Edwards J, Cato R, Towne JB. Hemodynamics of vein graft stenosis. J VASC SuR~ 1988;8:688-95.

8. Bandyck DF, Cato RF, Towne JB. A low flow velocity predicts failure of femoropopliteal and femorotibial bypass grafts. Surgery 1984;98:799-809.

9. Grigg MJ, Nicolaides AN, Wolfe JH. Detection and grading of femorodistal vein graft stenosis: duplex velocity measure- ments compared with angiography. J VAse Suit6 1988;8: 661-6.

10. Moody P, DeCossart LM, Douglas HM. Asymptomatic strictures in femoropopliteal vein grafts. Eur J Vasc Surg 1989;3:389-92.

11. Sladen JG, Reid JDS, Cooperberg PL, Harrison PB, Maxwell TM, Riggs MO, Sanders LD. Color flow duplex screening of infrainguinal grafts combining low- and high-velocity criteria. Am J Surg 1989;158:107-12.

12. Londrey GL, Hodgson KJ, Spadone DP, Ramsey DE, Barkmeier LD, Sumner DS. Initial experience with color-flow duplex scanning of infrainguinal bypass grafts. J VASe SURG 1990;12:284-90.

13. Robinson JG, Elliot BM. Does postoperative surveillance with duplex scanning identify the failing distal bypass? Ann Surg i991;5:182-5.

14. Brewster DC, LaSalle AJ, Robison JG, Strayhorn EC, Darling RC. Factors affecting patency of femoropopliteal bypass grafts. Surg Gynecol Obstet 1983;157:437-42.

15. Mills JL, Harris EJ, Taylor LM Jr., Beckett WC, Porter JM. The importance of routine surveillance of distal bypass grafts with duplex scanning: a study of 379 reversed vein grafts. J Vasc SURG 1990;12:379-89.

I6. Taylor PR, Wolfe JHN, Tyrrell MR, Mansfield AO, Nico- laides AN, Houstan RE. Graft stenosis: justification for 1 year surveillance. Br J Surg 1990;77:I125-8.

17. Mills JL, Bandyk DF, Gahtan V, Esses GE. The origin of infrainguinal vein graft stenosis: a prospective study based on duplex surveillance. J Vasc SUXG (in press).

18. Luscombe J, Moneta GL, Cummings CA, Taylor LM Jr, Porter JM. Postoperative surveillance of infrainguinal reverse vein grafts: a hypothesis for improving examination efficiency. J Vase Technol 1993;27:291-4.

19. Moneta GL, Yeager RA, Antonovic Ruza, Hall LD, Caster JD, Cummings CA, Porter JM. Accuracy of lower extremity duplex mapping. J Vasc SURG 1992;15:275-84.

20. Bandyk DF, Schmitt DD, Seabrook GR, Adams MB, Towne JB. Monitoring functional patency of in sire saphenous vein bypasses: the impact of a surveillance protocol and elective revision. J VAse SURG 1989;9:286-96.

21. Green RM, McNamara J, Ouriel K, DeWeese IA. Compari- son ofinfrainguinal graft surveillance techniques. J VAse SuRe 1990;11:207-15.

22. Bergamini TM, Towne JB, Bandyk DF, Seabrook GR, Schmitt DD. Experience with in situ saphenous vein bypass during 1981 to 1989: determinant factors of long-term patency. J VASe SURG 1991;13:137-49.

23. Mattos MA, Van Bemmelen PS, Hodgson KJ, Ranlsey DE, Barkmeier LD, Sumner DS. Does correction of stenoses identified with color duplex scanning improve infrainguinal graft patency? J Vasc SURG 1993;17:54-66.

24. Nehler MR, Moneta GL, Yeager RA, Edwards JM, Taylor LM Jr., Porter JM. Surgical treatment of threatened reversed infrainguinal vein grafts. J VASC SURG 1994;20:558-65.

25. Sladen JG, Gilmour JL. Vein graft stenosis: characteristics and effect of treatment. Am J Surg 1981;141:549-53.

26. O'Mara CS, Flirm WR, Johnson ND, Yao JST. Recognition and surgical management of patent but hemodynamically failed arterial grafts. Ann Surg 1981;193:467-76.

27. Veith FJ, Weiser RK, Gupta SK, Ascer E, et al. Diagnosis and management of failing lower extremity arterial reconstruc- tions prior to graft occlusion. J Cardiovasc Surg 1984;25: 381-4.

28. Turnipseed WD, Acher CW. Postoperative surveillance: an effective means of detecting correctable lesions that threaten graft patency. Arch Surg I985;120:324-8.

29. Barnes RW, Thompson BW, MacDonald CM, Nix ML, et al. Serial noninvasive studies do not herald postoperative failure of femoropopliteal or femorotibial bypass grafts. Ann Surg I989;210:486-94.

30. Idu MM, Blankenstein JD, De Gier P, Truyen E, Buth J. Impact of a color-flow duplex surveillance program on infrainguinal vein graft patency: a five-year experience, l VASe SURG 1993;17:42-53.

31. Laborde AL, Synn AY, Worsey J, Bower TR, et al. A prospective comparison of ankle/brachial indices and color duplex imaging in surveillance of the in situ saphenous vein bypass. J Cardiovasc Surg 1992;33:420-5.

32. DePew RJ, Kazmar J, Moneta GL, Yeager RA, Castillo ES, Cummings C, Edwards JM, Porter JM. Duplex ultrasound detection of pseudo-occluded lower extremity reverse vein grafts. J Vase Tech 1992;16:277-80.

33. Belkin M, Mackey WC, McLaughlin R, Umphrey SE, O'Donnell TP. The variation in vein graft flow velocity with luminal diameter and outflow level. J Vase Svxc i992;15: 991-9.

Submitted Feb. 10, 1995; accepted May 13, 1995.

JOURNAL OF VASCULAR SURGERY 482 Passman et al. October 1995

DISCUSSION

Dr. George Andros (Burbank, Calif.). President Elect Goldstone, members and guests. Dr. Passman and his colleagues from Portland question whether early graft surveillance fails to identify important latent graft defects that may result in late graft occlusion. They hypothesize surveillance should be continued for an extended period of time to detect late appearing lesions.

The senior author of this paper is known in this society not to be without opinions in many areas, and I quote from the Yearbook of Vascular Surgery in 1994. "I do not think that one should be performing leg bypasses at all (sic) unless one is prepared to offer detailed post-operative duplex imaging." That is somewhat at odds with our dealings with insurance carriers in Southern California who will allow these procedures to be done but ultimately do not pay for any kind of surveillance. My first question therefore must be a financial one. How can we recommend extended color duplex graft surveillance when the sources of funds to pay for it are uncertain? The important issue here is detection of the failing graft, the link point between graft failure and graft patency. Dr. Passman has just summarized his data, so I will not repeat it except to say that I was a bit surprised about his comments on the increased difficulties with grafts arising from the deep femoral artery. The most important finding was that 31% of grafts had lesions detected 6 months after their implantation with previous studies having been found to be normal. I should note that the authors perform a complete surveillance of the graft on each occasion it is studied.

I f you consider the standard graphical representation of a life patency table to be a phase display, a graft passes from the patent phase into the occluded phase through the failing graft phase. The failing graft period varies from a very short half-life with prosthetic grafts and homografts to a somewhat longer period of time with autogenous grafts. Additionally, the length of the failing graft period is dependent on the specific lesion that puts the graft at risk.

Looked at from another point of view, a patency curve can be broken down into three component parts, each with a specific slope. Each phase corresponds to the pathophysi- ological process with the early phase accounting for graph occlusions resulting from errors in technique or judg- ment (selecting a bad vein). The middle phase is myo- intimal hyperplasia, and the late phase is progressive atherosclerosis.

I find Dr. Passman's conclusions entirely plausible: stenotic lesions do develop late in the myointimal hyper- plasia phase, and these defects are discoverable by extend- ing graft surveillance. I wonder about the pathologic processes that cause these lesions. Could there be two classes of fibrostenosis, one of which began at the site of preexisting damage to the vein while it was in the venous circulation or was produced by minor trauma during the venous arterialization? Could there be a second class of

myointimal fibrostenosis that appear later, that arises as vein valves fibrose or results from hemodynamic causes? What did you find when you repaired these lesions, and what methods did you use?

The challenge in graft surveillance is obviously to find graft-threatening lesions and repair them before graft occlusion occurs. Your group has converted from graft surveillance at a single mid-graft point to a policy of complete graft inspection. Why the change? Did you find more lesions, and if so, what percentage of them did you repair to achieve the hoped-for objective of improving secondary patency? Your threshold criteria of a peak systolic velocity of 200 cm/sec are lower than advocated by other investigators. Are you finding more lesions that you choose not to repair? Do you think that all scanners would yield the same parameters, and if not, how would you standardize one machine against another? What sort of difference does it make in your surveillance if the graft is deep or superficial? What percentage of your graft occlu- sions occur while the patient is under surveillance?

Dr. Marc A. Passman. The first question concerns our preferred schedule for the duplex surveillance examination.

We prefer an initial postoperative evaluation and then evaluation at three monthly intervals for the first year and then every 6 months thereafter. We have not yet adopted some of the suggestions that there might be some types of stenoses that lend themselves to more frequent surveillance. Based on this study, duplex surveillance should be contin- ued for the entire life of the graft.

The second question concerns payment for duplex surveillance. We try to obtain reimbursement from the primary insurance carriers. When this is not possible, research grants cover the costs or the studies are performed without charge.

Our patients are reasonably compliant with our rec- ommendations. However, being a referral institution, many come from far away, and this does make surveillance difficult in this subpopulation.

The history of graft surveillance at Oregon is one that initially used a mid-graft peak systolic velocity of less than 45 cm/sec as recommended by Bandyk. After 1990, with routine use of color flow for surveillance studies, the higher flow velocity criteria was added. This has worked well for detecting lesions with greater than 50% stenosis by subsequent angiography.

We obtain an angiogram in all cases to confirm the presence of a stenosis. This is especially important in the group that has a low mid-graft velocity where a concomi- tant focal increase in peak systolic velocity was not found.

Another question was directed at the occlusion of grafts while under surveillance. Of the 447 initial bypass opera- tions performed at our institution, 60 revisions were ultimately required. The 36 graft revisions in this study are a subpopulation that had had strict surveillance according

~OURNAL OF VASCULAR SURGERY Volume 22, Number 4 Passrnan et al. 483

to the criteria discussed. The remaining patients include patients who presented with acute graft occlusion, which occurred in eight cases, and grafts with inadequate surveil- lance. At our institution the primary reason for inadequate surveillance is the great distances some patients must travel for follow-up surveillance examinations.

We did not address the pathologic characteristics of the lesions. Angiographically the lesions have the appearance of those felt in other studies to be secondary to intimal hyperplasia.

Regarding cost-effectiveness, it is clear duplex surveil- lance is medically effective. Cost-effectiveness has not been addressed.

Dr. J. Dennis Baker (Los Angeles, Calif.). Across the country there are very wide variations as to what payments are made, especially for postoperative surveillance. Even within the Medicare program there is no uniformity from carrier to carrier.

The specific question I have has to do with ABIs. You said only 30% of the cases had drops in ABIs, but what types of bypasses had these drops?

Dr. Passman. About one-third of the grafts in this study were to a tibial artery. I don't believe there was any difference in the proportion of popliteal or tibial bypass that had a duplex abnormalic leading to revision not associated with a change in ABI. ABI alone is probably inadequate for optimal surveillance.

Dr. William J. Quifiones-Baldrich (Los Angeles, Calif.). In the last 5 or 6 years we have been doing more and more very distal grafts that originate in the popliteal arteries. These are short grafts with limited outflow, and what we find is that in the initial duplex scan their velocities usually range somewhere between 25 and 40 cm/sec.

In our experience when we did angiograms in those grafts where we could not find any specific area of stenoses by duplex scan but had relatively low velocities, no abnormalities were found.

Have you had an opportunity to perform surveillance in these shorter grafts, and does the criteria of 45 cm/sec apply to them?

Dr. Passman. The grafts primarily originated from the common femoral, superficial, or deep femoral artery. No grafts originating from the below-knee popliteal artery were in this study. So I cannot comment on whether our surveillance criteria will fit with that specific sort of graft.

Dr. Gregory L. Moneta (Portland, Ore.). I agree with Dr. Quifiones-Baldrich. Grafts with limited outflow and/or to the pedal vessels often have mid-graft velocities less than 45 cm/sec. We do generally not obtain angiograms in those grafts solely for isolated low-flow velocities.

Dr. Fred A. Weaver (Los Angeles, Calif.). I am interested in your routine use of angiography to confirm the stenoses. We have found that the angiogram often underestimates the graft stenosis, particularly if it is a weblike lesion, to the point where we no longer routinely obtain an angiogram to confirm a physiologic abnormality found by duplex scan.

If you have a discrepancy between the duplex informa- tion and findings on your angiogram, which do you choose to believe, the angiogram or the duplex?

Dr. Passman. You are correct, that in some cases, seemingly minimal lesions on angiography are associated with surprisingly high flow velocities. A web-like lesion is probably better identified with duplex than angiography. If we find a high velocity and any suggestion on the angiogram ofa stenosis we will revise the graft. I f however, a high velocity is not associated with any suspicion of a lesion on a high quality angiogram our tendency is to regard the duplex as a false positive study. Such false positive studies may be more likely in anatomically tunnelled grafts than in-situ bypasses.

Dr. Joseph Mills (Tucson, Ariz.). I just wanted to make a few brief comments and ask two questions. This article actually was a response to a work that Dennis Ban- dyk and I did (J VASC Suv, c 1995;21:16-25). After we presented it last year Dr. Porter told me he did not believe it. The key issue was where do these lesions come from, and when do they occur? Our protocol was prospective. We started with intense early duplex scanning, and to get in our protocol the grafts had to have at least two duplex scans in the first 6 weeks. And actually, approximately three quarters of the grafts had three scans, and we were comfort- able at that point that if we could not find anything wrong on two or three scans, that that graft was in fact normal.

In contrast, this study was retrospective in that only two thirds of the grafts had even one scan early on, but despite those caveats, if you look at the data I have an alternate interpretation. If you look at the number of grafts that required revision for abnormalities that were detected after 1 year, there were only five, and if you subtract the grafts that were revised before then, that gives you an incidence of de novo graft stenosis requiring repair of 1%.

I would suggest that low incidence is not high enough to require the same intensity of surveillance for the entire duration of the graft and what we in fact recommended from our article was that you could stratify the grafts based on early scans. The scans that were abnormal early had a high incidence of progression, and those that were normal early had a low incidence of stenosis requiring repair, which in our series was 2%. So I think your data and ours actually agree.

If you were to carry out the same intensity of surveillance after the first year, I did some calculations, if you do approximately 100 grafts a year, as your group does, and have approximately a 10% revision rate, and approxi- mately 8% of the patients die per year, if you carry that out to the fifth year, you are going to be doing 1700 scans, if you do them every 3 months, which amounts to 7 scans per work day. So basically you would have a lab just scanning distal grafts all day long, which is absolutely impractical.

So I think what I would ask you to do is reinterpret your data. Is there a way we can figure out which grafts require intensive surveillance and which ones we can back offon? What we suggested is that after you got to a year,

JOURNAL OF VASCULAR SURGERY 484 Passman et al. October 1995

if things were normal, probably checking those grafts once or twice a year was sufficient, and you did not need to continue surveillance with the same intensity.

Dr. Passman. In regards to Dr. Mills' calculations and estimations, we had very strict criteria for inclusion in this study and focused on patients who required revision and who had really had high quality surveillance. The 1% calculation (5/447) that you made assumes the same strict criteria for the entire bypass population. You should be more cautious in extrapolating data and in your underlying assumptions.

In terms of which grafts to perform more or less intensive surveillance on, this study was not designed to look at that. This study was designed primarily to look at

whether graft abnormalities by duplex occurred late, that is, beyond 6 months or even beyond 1 year, which I think we adequately showed.

Dr. Moneta. Actually, Dr. Mills' memory has shifted a little bit. When I discussed his paper at the SVS meeting in June 1994 the only thing that I was concerned about was that in the manuscript that I had to review, it stated patients who had normal grafts initially would not require late surveillance. While I disagreed with this and still do, I agree with most everything else Dr. Mills said in his SVS paper. If the subsequent article for final publication in the journal has been revised, I think that reflects well on Dr. Mills' reasoning and probably very well on the review process of the JOURNAL.

O N THE M O V E ? D o n ' t miss a s ingle issue of the journal! To ensu re p r o m p t service w h e n y o u change y o u r address , p lease p h o t o c o p y and comple t e the f o r m below.

Please send your change of address notification at least six weeks before your move to ensure continued service. We regret we cannot guarantee replacement of issues missed due to late notification.

JOURNAL TITLE: Fill in the title of the journal here.

OLD ADDRESS: Affix the address label from a recent issue of the journal here.

NEW ADDRESS: Clearly print your new address here.

Name

Address

City/State/ZIP

COPY AND MAIL THIS FORM TO: Journal Subscript ion Services Mosby-Year Book, Inc. 11830 Westline Industr ia l Dr. St. Louis, MO 63146-3318

OR FAX TO: 314-432-1158

Mosby

OR PHONE: 1-800-453-4351 Outs ide the U.S., call 314-453-4351