Prospective Validation of Threshold Criteria for Intervention in Infrainguinal Vein Grafts Undergoing Duplex Surveillance

Prospective Validation of Threshold Criteria for Intervention in Infrainguinal Vein Grafts Undergoing Duplex Surveillance Alex Westerband, MD, Joseph L. Mills, MD, Sherry Kistler, RN, Scott S. Berman, MD, Glenn C. Hunter, MD, and John M. Marek, MD, Tucson, Arizona Although color flow duplex surveillance (CFDS) of infrainguinal vein grafts has gained wide acceptance, definitive criteria mandating graft revision remain to be established. We prospectively evaluated 101 infrainguinal vein grafts undergoing CFDS in order to validate threshold duplex criteria for intervention which were derived from our previous experience and that reported by others. Complete CFDS of the bypass conduit and adjacent inflow and outflow arteries and Doppler-derived ankle brachial indices (ABI) were obtained every 3 months × 4 and every 6 months thereafter. The following threshold criteria mandating further evaluation and intervention to prevent graft occlusion were applied: high-velocity criteria (HVC) defined as peak systolic velocity (PSV) >300 cm/sec and velocity ratio (Vr) >3.5; low-velocity criteria (LVC) defined as PSV <45 cm/sec; an ABI decrease >0.15. Fifty-one grafts had normal serial CFDS and ABI; none subsequently occluded or required revision. Stenosis was detected by CFDS in 43 grafts (PSV > 180 cm/sec, Vr > 1.5). Within this subgroup, 54% of grafts subsequently required revision (20/43) or occluded (3/43). All grafts in this subgroup with stenoses progressed to PSV > 300 or Vr > 3.5 prior to revision or occlusion. Ten lesions (23%) regressed spontaneously without intervention (mean PSV 252 cm/sec, mean Vr 3.2); 10 lesions (23%) are stable, nonprogressive, and remain under surveillance. Two grafts were abnormal by LVC; one was successfully revised, the other occluded prior to intervention. Five grafts had normal CFDS and ABI decrease >0.15. Four were revised (three inflow lesions, one outflow lesion) and one occluded (missed lesion by CFDS). Only five graft occlusions occurred in the entire series: three grafts met HVC and occluded prior to intervention; one developed an ABI drop of 0.4 due to graft stenosis missed by CFDS and uncovered following thrombolysis, and the other graft met LVC and occluded prior to intervention. Infrainguinal vein grafts with normal serial CFDS and ABI are at minimal risk of spontaneous graft occlusion. When CFDS is abnormal (PSV > 180 cm/sec, Vr > 1.5), over 50% of grafts will ultimately require revision or progress to occlusion. Grafts with such lesions can be safely monitored by CFDS until progression to lesions meeting HVC occurs with minimal risk of graft occlusion. A decrease in ABI >0.15 with normal CFDS mandates arteriography to identify inflow and outflow lesions or a missed graft stenosis. The present study prospectively validates threshold intervention criteria for graft lesions meeting HVC (PSV > 300 cm/sec, Vr > 3.5), LVC (PSV < 45 cm/sec throughout graft) or an ABI decrease >0.15. (Ann Vasc Surg 1997;11:44-48.)

From the Section of Vascular Surgery, University of Arizona Health Sciences Center, Tucson, AZ. Presented at Twenty-first Annual Meeting of The Peripheral Vascular Surgery Society, Chicago, IL, June 8, 1996. Supported by a grant from the American Heart Association, Grant A2GS-44-96. Correspondence to: Joseph L. Mills, MD, Associate Professor of Surgery, Room 5406, University of Arizona Health Sciences Center, Tucson AZ 85724, USA. 44

Intermediate and late failures following infrainguinal arterial bypass surgery using autogenous venous conduits are most often preceded by the development of intrinsic graft lesions. Indeed, it has been known since the 1960s that autogenous vein grafts implanted into the arterial circulation are prone to the development of intrinsic structural defects which may require repair to prevent graft occlusion.1-4 Before the advent of duplex graft sur-

Vol. 11, No. 1, 1997

Intervention in infrainguinal vein grafts 45

Fig. 1. A Duplex scan-derived peak systolic velocity of 650 cm/sec (HVC) obtained at the site of a proximal juxtaanastomotic vein graft stenosis. B Arteriogram confirms the presence of a high-grade proximal vein graft lesion.

veillance, however, there was no clinically convenient method to identify and localize these lesions in a timely fashion in order to intervene prior to graft thrombosis. Routine Doppler-derived ankle pressure measurement was insufficiently sensitive and serial arteriographic studies were too invasive and expensive for widespread application. There is now convincing data to support postoperative duplex surveillance of both in situ and reverse vein conduits following infrainguinal revascularization. This noninvasive method has been found to be highly accurate and sensitive for early detection of intrinsic graft abnormalities and to follow the progression of established infrainguinal vein graft stenosis.5-12 Multiple series have shown improved assisted primary patency rates for grafts subjected to such surveillance.12-18 However, the precise noninvasive criteria which should be applied as a threshold for intervention are in need of further delineation. Based on our previous work as well as a review of the literature, we derived threshold duplex criteria mandating intervention to prevent graft occlusion.11,19 The purpose of the present study was to prospectively evaluate and validate these threshold duplex criteria.

duplex surveillance have been previously described.11,13,19 Ankle brachial indices were also obtained at each visit. All grafts were studied using an Ultramark IX Duplex Scanner (Advanced Technology Laboratories, Bothell, WA). The study population consisted of 101 consecutive infrainguinal vein grafts in 98 patients subjected to CFDS at The University of Arizona Health Sciences Center beginning in August 1994. Based on our literature review and previous experience, the following threshold criteria mandating further graft evaluation and intervention to prevent graft occlusion were applied: (1) high-velocity criteria (HVC) defined as a peak systolic velocity (PSV) >300 cm/sec at the site of the stenosis and/or a velocity ratio (Vr) >3.5; (2) low-velocity criteria (LVC) defined as peak systolic velocity in a normal caliber segment of the graft and throughout the graft <45 cm/sec; and (3) a decrease in ankle brachial index >0.15. The purpose of the present study was to apply these criteria prospectively to a series of consecutive vein grafts to determine the diagnostic yield and impact on vein graft patency.

METHODS

One hundred and one infrainguinal vein grafts have been prospectively followed by CFDS and ABIs for the past 2 years. Patients who were lost to follow-up less than a year after the operation and those with prosthetic grafts were excluded. The mean patient age was 68.3 years, (range 31-87 years). Thirty-nine patients (40%) were women

Color-flow duplex scanning (CFDS) of the entire bypass conduit and adjacent inflow and outflow arteries and Doppler-derived ankle brachial indices (ABI) were obtained every 3 months for 1 year and every 6 months thereafter. The precise methods of

RESULTS

46

Westerband et al.

Annals of Vascular Surgery

Table I. Distribution of inflow and outflow vessels in a series of 101 infrainguinal veins grafts Inflow vessels

Common femoral artery Superficial femoral artery Profunda femoris Above-knee popliteal artery Below-knee popliteal artery

Outflow vessels

50 (49.5%)

Above-knee popliteal artery

14 (13.8%)

39 (38.6%) 8 (7.9%)

Below-knee popliteal artery Tibio-peroneal trunk

35 (34.7%) 4 (4%)

1 (1%)

Aterior tibial artery

15 (14.9%)

3 (3%)

Dorsalis pedis artery Posterior tibial artery Peroneal artery

5 (5%) 9 (8.9%) 19 (18.8%)

and 59 (60%) were men. Risk factors included hypertension in 77 patients (78.5%), cardiac disease in 50 (51%), hyperlipidemia in 47 (47.9%), diabetes in 33 (33.6%), cerebrovascular disease in 22 (22.4%), chronic renal failure in 17 (17.3%), and a known hypercoagulable state in two (2%). Fifteen patients (15.3%) had already undergone a leg bypass operation; 77 patients had a previous history of smoking. Indications for reconstruction were limb salvage (70), claudication (29), and a popliteal aneurysm in two patients. In situ saphenous vein was used in 27 patients, a reverse conduit in 61 patients, spliced veins in 13. The distribution of inflow and outflow vessels is shown in Table I. Fifty-one grafts in this series had normal serial CFDS and ankle brachial index determination throughout the study protocol. None of these grafts subsequently occluded or required revision during the follow-up. A stenosis was detected by CFDS in 43 grafts. The presence of the stenosis was defined as a PSV >180 cm/sec with a velocity ratio >1.5. Within this subgroup, 54% of grafts subsequently progressed to previously identified threshold criteria and required revision (20/43) or occluded prior to a planned revision (3/43). All grafts in this subgroup which occluded or were revised had progressed to a peak systolic velocity of >300 cm/sec or a velocity ratio of >3.5 prior to the event (Fig. 1). Patients who were revised were operated upon within 2 weeks after identification of high-grade stenosis. Ten lesions (23%) regressed spontaneously without any required intervention. The mean peak systolic velocity for these grafts was 253 cm/sec with a mean velocity ratio of 3.2. It should be noted that these values were below the established threshold for intervention. Ten lesions (23%) remain stable, are nonprogressive, and remain under duplex surveillance.

Two grafts had no stenosis identified by HVC but were abnormal by LVC. One graft was successfully revised (revision of a proximal vein graft after a left superficial femoral artery to peroneal artery bypass), and the other was occluded prior to any intervention. Five additional patients had normal CFDS but developed a decrease in ankle brachial index by more than 0.15, (mean 0.52). In each of these patients, significant graft-threatening abnormalities were identified despite the presence of a normal CFDS. Four were revised following diagnostic arteriography which identified three significant inflow lesions, and one outflow lesion; the fifth graft occluded. When this latter graft was subsequently subjected to thrombolytic therapy, an intrinsic graft stenosis was identified which had been missed by CFDS. This was the only intrinsic conduit lesion in the entire series documented to have been missed by serial surveillance. The other four grafts which had normal CFDS with decreased ABI had inflow or outflow lesions, but no intrinsic vein conduit lesions. A total of five graft occlusions have occurred in the entire series. Three of these grafts met HVC criteria, but occluded either prior to planned intervention or after the patient refused intervention. The occlusions occurred within 2 weeks and were confirmed by repeat duplex scan. In another instance, the ankle brachial index had dropped by 0.4 due to a graft stenosis which had been missed by CFDS. This case was alluded to above, and the lesion was uncovered following thrombolysis. The final graft which occluded met LVC for intervention, but due to a clinical decision of the attending surgeon, no further investigation was undertaken, and the graft had already occluded by the time the patient presented for follow-up 3 weeks later.

Vol. 11, No. 1, 1997

Intervention in infrainguinal vein grafts 47

DISCUSSION

this series, of ten lesions which regressed spontaneously, the mean peak systolic velocity was 250 cm/ sec and the velocity ratio was only 3.2. Nevertheless, our data do suggest that when the peak systolic velocity progresses to >300 cm/sec and a velocity ratio >3.5, the graft is at significant risk for failure. In addition, our observations confirm that the other factors which are a critical component of vein graft surveillance include the distal or global graft flow velocity (GGFV) and determination of the ankle brachial index. If a previously normal GGFV falls to less than 45 cm/sec, this satisfies low-graft-velocity criteria. In certain situations such as a large caliber conduit, the GGFV may remain low from the onset; however, if a low-velocity state develops during follow-up, this finding should not be ignored. However, if the GGFV is normal or unchanged, but the ABI has decreased by more than 0.15, it indicates hemodynamic deterioration, and further evaluation of the graft should be carried out. This caveat is critical since an isolated ABI drop (in the absence of missed intragraft lesions) has been found to correlate in our series with the presence of inflow or outflow lesions not necessarily detectable by duplex scanning. Therefore, such grafts should be investigated by arteriography to rule out a missed graft lesion by CFDS as well as possible high-grade inflow or outflow graft stenoses which would need to be addressed to normalize graft hemodynamics and prevent graft occlusion. While it seems true that inflow and outflow lesions are less threatening to intermediate graft patency than intrinsic graft lesions, hemodynamically significant lesions in the presence of a patent graft should be investigated.30 When these criteria were applied to our series of 101 consecutive grafts, only five graft occlusions occurred in the entire series. All five of these grafts met criteria for intervention due to HVC, LVC, or hemodynamic deterioration but either because of patient refusal, delay in performing a planned surgical revision, or a clinical decision to follow the graft for an additional period of time, graft occlusion occurred. Thus, all five grafts which thrombosed in this series did so because of protocol violations. Three of these grafts met HVC criteria and could have been revised prior to graft occlusion. The other two grafts either met low-flow criteria by a decrease of the global graft velocity or had hemodynamic deterioration (ABI decrease), which should have been investigated. Also in this series, only one lesion initially meeting high velocity criteria subsequently regressed without repair. Thus it would appear that attempts to adjust the velocity criteria any further would result in an unacceptable incidence of spontaneous graft occlusion.

Duplex surveillance is based upon several underlying premises which deserve critical appraisal. These major premises are reasonably well supported by the literature, and include the observations that vein graft failure most often results from the development of intrinsic graft stenosis, and that highgrade stenoses lead to graft thrombosis if not revised.4,20 In addition, it was recognized prior to the advent of duplex scanning that prophylactic revision of patent but failing grafts yields results superior to those obtained following thrombectomy or thrombolysis and revision of occluded vein grafts. 2,21-23 However, numerous studies have shown that vein graft stenosis is often clinically silent and not readily detectable prior to graft occlusion by history, physical examination, or simple noninvasive measurements.24,25 The lack of correlation between ankle brachial index measurement alone and graft patency has been noted by Barnes and associates26 and underscores this point. The introduction of duplex scanning in clinical practice has been associated with an improved cumulative patency rate of infrainguinal vein grafts since more vein graft stenoses and low-flow states can be accurately identified, graded, and monitored for progression by appropriate surveillance.12-18 There is no prospective randomized trial to establish the following premise, but the 5-year secondary patency rate following graft revision for a stenotic but patent graft can be as high as 80%-90%,13 whereas the successful 1- to 2-year outcome following thrombolytic therapy or thrombectomy and revision is in the order of 20%-50%.27,28 Thus, the justifications for graft surveillance appear to be sound. Our previous work suggested, based on a retrospective analysis, that when graft lesions reached a peak systolic velocity greater than 300 cm/sec or a velocity ratio greater than 3.5, the graft stenosis was unlikely to regress and there was a significant risk for graft occlusion if no intervention was undertaken.11 Sladen et al. previously described nearly identical criteria.6 However, precise criteria mandating graft intervention have not yet been established or prospectively validated. Recent experimental work by Papanicolaou and associates29 would indicate that systolic flow limitation begins to occur in stenotic lower extremity vein grafts when the peak systolic velocity exceeds >250 cm/sec. In clinical practice, we have seen minimal complications when following grafts with peak systolic velocities between 250 and 300 cm/sec if the distal graft flow velocity is still normal and the ankle-brachial index has not fallen significantly. In

48

Westerband et al.

We conclude that infrainguinal vein grafts with normal serial CFDS (PSV between 45 and 180 cm/ sec) and normal ankle brachial index are at minimal risk for spontaneous graft occlusion. When CFDS reveals a focal, abnormally high PSV > 180 cm/sec or a Vr > 1.5, over half of such grafts will ultimately require revision or progress to occlusion. However, grafts with such lesions of intermediate hemodynamic significance can be safely monitored by CFDS every 4-6 weeks until progression to lesions meeting HVC occurs. Using this protocol, there is minimal risk of spontaneous graft occlusion. A major caveat is that if the ankle brachial index drops by greater than 0.15, or the global graft flow velocity falls to less than 45 cm/sec, further evaluation with arteriography should be carried out to make sure that there are no significant inflow or outflow lesions as well as to ascertain whether or not there was a missed graft lesion by CFDS. The present study appears to prospectively validate threshold intervention criteria for graft lesions meeting HVC (PSV > 300 cm/sec, Vr > 3.5), LVC (PSV < 45 cm/sec throughout the graft) and/or an ABI decrease exceeding 0.15.

Annals of Vascular Surgery

12.

13.

14.

15.

16.

17.

18.

19.

20.

REFERENCES 1. Szilagyi DE, Elliott JP, Hageman J, et al. Biologic fate of autogenous vein implants as arterial substitutes: Clinical, angiographic, and histo-pathologic observations in femoro-popliteal operations for atherosclerosis. Ann Surg 1973;178:232-246. 2. Whittemore AD, Clowes AW, Couch NP, et al. Secondary femoropopliteal reconstruction. Ann Surg 1981;193:35-42. 3. Griggs MJ, Nicolaides AN, Wolfe JHN. Femorodistal vein bypass graft stenosis. Br J Surg 1988;75:737-740. 4. Moody P, de Cossart LM, Douglas HM, et al. Asymptomatic strictures in femoro-popliteal vein grafts. Eur J Vasc Surg 1989;3:389-392. 5. Grigg MJ, Nicolaides AN, Wolfe JHN. Detection and grading of femorodistal vein graft stenoses: Duplex velocity measurements compared with angiography. J Vasc Surg 1988;8: 661-666. 6. Sladen JG, Reid JDS, Cooperberg PL, et al. Color flow duplex screening of infrainguinal grafts combining low-and highvelocity criteria. Am J Surg 1989;158:107-112. 7. Mills JL, Harris EJ, Taylor LM Jr, et al. The importance of routine surveillance of distal bypass grafts with duplex scanning: A study of 379 reverse vein grafts. J Vasc Surg 1990; 12:379-389. 8. Buth J, Disselhoff B, Sommeling C, et al. Color-flow duplex criteria for grading stenosis in infrainguinal vein grafts. J Vasc Surg 1991;14(6):716-728. 9. Mills JL, Fujitani RM, Taylor SM, et al. The characteristics and anatomic distribution of lesions that cause reversed vein graft failure: A five-year prospective study. J Vasc Surg 1993;17(1):195-206. 10. Wilson WG, Davies AH, Carrie IC, et al. The value of predischarge duplex scanning in infrainguinal graft surveillance. Eur J Endovasc Surg 1995;10:237-242. 11. Mills JL, Bandyk DF, Gahtan V, et al. The origin of infrain-

21.

22.

23.

24. 25.

26.

27.

28.

29.

30.

guinal vein graft stenosis: A prospective study based on duplex surveillance. J Vasc Surg 1995;21:16-25. Caps MT, Cantwell-Gab K, Bergelin RO, et al. Vein graft lesions: Time of onset and rate of progression. J Vasc Surg 1995;22:466-475. Bandyk DF, Schmitt DD, Seabrook GR, et al. Monitoring functional patency of in-situ saphenous vein bypasses: the impact of a surveillance protocol and elective revision. J Vasc Surg 1989;9:286-296. Taylor PR, Wolfe JHN, Tyrrell MR, et al. Graft stenosis— justification for 1-year surveillance. Br J Surg 1990;77(10): 1125-1128. Moody P, Gould DA, Harris PL. Vein graft surveillance improves patency in femoro-popliteal bypass. Eur J Vasc Surg 1990;4:117-121. Idu MM, Blankenstein JD, de Gier P, et al. Impact of a colorflow duplex surveillance program on infrainguinal vein graft patency: A five-year experience. J Vasc Surg 1993;17(1):42-53. Mattos MA, van Bemmelen PS, Hodgson KJ, et al. Does correction of stenoses identified with color duplex scanning improve infrainguinal graft patency? J Vasc Surg 1993;17:54-66. Bergamini TM, George SM, Massey HT, et al. Intensive surveillance of femoropopliteal-tibial autogenous vein bypasses improves long-term graft patency and limb salvage. Ann Surg 1995;221(5):507-518. Gahtan V, Payne LP, Roper LD, Mills JL, et al. Duplex criteria for predicting progression of vein graft lesions: Which stenoses can be followed? J Vasc Tech 1995;19(4):211-215. Bandyk DF, Bergamini TM, Towne JB, et al. Durability of vein graft revision: the outcome of secondary procedures. J Vasc Surg 1991;13:200-210. Veith FJ, Weiser RK, Gupta SK, et al. Diagnosis and management of failing lower extremity arterial reconstructions prior to graft occlusion. J Cardiovasc Surg 1984;25:381-384. Cohen JR, Mannick JA, Couch NP, et al. Recognition and management of impending vein-graft failure. Importance for long-term patency. Arch Surg 1986;121:758-759. Berkowitz HD, Greenstein SM. Improved patency in reverse femoro-infrapopliteal autogenous vein grafts by early detection and treatment of the failing graft. J Vasc Surg 1987;5: 755-761. Wolfe JHN, Lea Thomas M, Jamieson CW, et al. Early diagnosis of femoro-distal graft stenoses. Br J Surg 1987;74:268-270. Bandyk DF, Kaebnick HW, Stewart GW, et al. Durability of the in situ saphenous vein arterial bypass: a comparison of primary and secondary patency. J Vasc Surg 1987;5:256-268. Barnes RW, Thompson BW, MacDonald CM, et al. Serial noninvasive studies do not herald postoperative failure of femoropopliteal or femorotibial bypass grafts. Ann Surg 1989;210:486-494. Bergamini TM, Towne JB, Bandyk DF, et al. Experience with in situ saphenous vein bypasses during 1981 to 1989. Determinant factors of long-term patency. J Vasc Surg 1991; 13:137-149. Donaldson MC, Mannick JA, Whittemore AD. Causes of primary graft failure after in situ saphenous vein bypass grafting. J Vasc Surg 1992;15:113-120. Papanicolaou G, Zierler RE, Beach KW, et al. Hemodynamic parameters of failing infrainguinal bypass grafts. Am J Surg 1995;169:238-244. Taylor SM, Mills JL, Fujitani RM, et al. Does arterial inflow failure cause distal vein graft thrombosis? A prospective analysis of 450 infrainguinal vascular reconstructions. Ann Vasc Surg 1994;8(1):92-98.