Infrapopliteal prosthetic graft patency by use of the distal adjunctive arteriovenous fistula Herbert Dardik, MD, Silvia M. Berry, MSc, KVT, Alan Dardik, BS, Fred Wolodiger, MD, Joseph Pecoraro, MD, Ibrahim M. Ibrahim, MD, Mark Kahn, MD, and Barry Sussman, MD, Englewood,N.J. From November 1979 through December 1989, 210 distal arteriovenous fistulas were constructed as adjuncts to tibial and peroneal vascular reconstructive procedures in 203 patients threatened with limb loss. Two-year cumulative patency rates were calculated by grouping patients on the basis o f changing indications in sequential time periods: group 1 (n = 61): 1979to 1983, 18%;group2 (n = 80): 1983to 1986,33%;group3 (n = 69): 1986 to 1989, 44%. Although the therapeutic results observed in these groups are not statistically comparable, they show a perceptible trend. Postoperative arteriography showed that flow is prograde in the distal vessels beyond the distal arteriovenous fistula. Graft surveillance by duplex ultrasonography also confirmed that flow in the distal arteries is prograde and that "steal" does not occur. Peak systolic velocity (174 -+ 38 cm/sec) and mean velocity (92 + 23) flow rates are increased in grafts with patent distal arteriovenous fistulas compared to those bypasses with closed distal arteriovenous fistulas (p < 0.01). There were no differences in the flow measurements for the arteries beyond the distal anastomoses and distal arteriovenous fistulas, confirming the prograde nature of the distal flow. In 22 patients analysis of graft and fistula patency by duplex sonography showed that one fourth of all grafts were patent without fistulas at 1 and 2 years after operation. Alternatively, 68% o f patent grafts at I year had patent fistulas and 58% had patent fistulas at 2 years. We conclude that the distal arteriovenous fistula will increase graft flow and simultaneously prevent distal arterial overload without causing "steal." This technique should be considered whenever a prosthetic graft is necessary for crural reconstruction and only in selected instances o f revascularization with autologous veins. (J VAsc Sult¢
1991;13:685-91.)
The excellent clinical results obtained by the use of the autologous saphenous vein for lower limb revascularization has diminished the need for the use of prosthetic arterial substitutes. 1"2 Nevertheless, there are occasions where autologous tissue may not be available or suitable so that a prosthetic graft may be preferable or the only alternative to amputation. 3-s The results of prosthetic graft bypasses to the tibia] and peroneal arteries have been shown to be decidedly inferior to those obtained with autologous veins. 4-7 To circumvent these problems, the major surgical approaches to augment graft patency rates have been (1) construction of a sequential component to the bypass graft and (2) creation of an From the Vascular Surgical Service,EnglewoodHospital, Englewood. Presented at the Forty-fourthAnnual Meeting of the Societyfor Vascular Surgery, Los Angeles, Calif., June 4-6, 1990. Reprint requests: Herbert Dardik, MD, 375 Engle St., Englewood, NJ 07631. 24/1/26927
arteriovenous fistula (dAVF) in relationship to the distal anastomosis of the prosthetic graft. The former, if dual runoff vessels are not present, may be neither possible nor effective if they overlap similar anatomic patterns of distribution. The latter may also not be possible since dAVF function has a critical dependence on the configuration and quality of the deep venous system, s-l° Since 1979 we have been actively involved in these adjunctive distal procedures to enhance prosthetic and, at times, autologous graft patency rates. This article summarizes our experience and conclusions relating to the indications for and results of using these procedurcs. MATERIAL A N D M E T H O D S During the 10-year period, from November 1979 through December 1989, 210 dAVFs were constructed in 203 patients. The group consisted of 127 men and 76 women aged 31 to 85 years (mean 68 years). One hundred six (52%) patients had diabetes. Prior vascular reconstructive procedures had been 685
lournJof VASCULAR SURGERY
686 Dardik et al.
Fig. 1. Duplex scan of a distal anastomosis of an umbilical vein graft (HUV) to the peroneal artery (D, distal; P, proximal) with a dAVF (Fistula).
performed in the involved limb in 130 (63%) instances, 15 of these involved two or more procedures. Group 1, previously reported in 1983, 8 consisted of 61 dAVFs (58 patients) and provided the foundation for understanding the critical elements of technique and appreciating the status of the deep venous system. Group 2 patients had 80 dAVFs created between March 1983 and January 1986. This time period was synchronous with the transition in our institution from the use of prosthetics to autologous vein as the preferred graft material. Group 3 represents those patients undergoing bypasses and dAVFs (n = 69) from January 1986 through December 1989, a time period when our policy for autologous tissue as the conduit of choice was operative and use ofprosthetics and dAVFs was restricted. Nine patients in this series underwent dAVF construction as an adjunct to autologous vein bypass surgery, six of whom had failed prior autologous vein bypasses. The specific indications for creating an adjunctive dAVF reported earlier have now been simplified to include all bypasses performed to the crural arteries where a prosthetic is used. 8 Relative contraindications to performing a dAVF include virtual absence of arterial runoff, small venae comitantes, and extensive sclerosis or other sequelae of deep venous thrombosis. Construction of the dAVF was usually by a side-to-side technique. Infrequently, variations were used such as placing the fistula remote from the distal
graft anastomosis or using only the cephalad segment. Most of the dAVFs were performed as primary adjuncts, that is, synchronous construction of both the dAVF and the bypass. Secondary dAVFs were those performed as a separate procedure along with thrombectomy within the early postoperative period of a previously placed graft. Prosthetic graft materials consisted of glutaraldehyde-tanned human umbilical cord vein and, in one instance, thin-walled polytetrafluoroethylene. Autologous veins were usually from the saphenous vein systems, greater and lesser, and some from arm veins. These were used in various configurations, singly or spliced. Completion arteriography was performed routinely and occasionally complemented with flow studies by use of the Statham-Gould electromagnetic flowmeter (model SP-2204). Confirmation of prograde flow in the distal artery and cephalad flow in the upper venous component of the fistula was additionally obtained with an intraoperative Doppler (Bach Simpson Ltd, model 348). After discharge from the hospital follow-up was performed at 3-week intervals for the first 6 months and more frequently as required for wound care and other conditions. Subsequent surveillance was performed at 3-month intervals and extended where appropriate. Noninvasive testing consisted of duplex scanning and pulse volume recordings with anklebrachial pressure measurements and waveform analysis. The former was performed with the patients in a supine position by use of a mechanical 8 MHz
Volume 13 Number 5 May 1991
Graft patency with the distal arteriovenous fistula
687
Fig. 2. Characteristic flow signals of a patent fistula (left) and graft (right).
100
i
.......
v n~ Q. u_
!
80
60
26.2
\11 s2.6
U
6•2.6 %4
40 Cure. Graft Cure. Fistula 20 Fist. Pat. = 68% of Pat. Grafts I
months
6
I
Fist.Pat.= 58 % of Pat. Grafts I
12
24
Fig. 3. Cumulative graft and fistula patency rates for 22 patients with duplex follow-up studies. One fourth of all the grafts were patent without fistula patency at 1 and 2 years. Of those grafts patent at 1 year, 68% still had patent fistulas. At 2 years, fistula patency occurred in 58% of the grafts patent at that time.
sector scanner with 6.5 M H z angle corrected Doppler. The entire length of the graft was imaged from the proximal to the distal anastomosis. The dAVF and the artery beyond the distal anastomosis were similarly evaluated. For each site, peak systolic velocity (V) and mean velocity (9) were measured. A resistivity index (RI) was calculated: RI=
Peak systolic velocity-Diastolic velocity Peak systolic velocity
With open dAVFs, diastolic flow is high, consistent with a low resistance. The converse is true with a closed dAVF. Finally, direction o f flow was noted at all sites. Postoperative standard contrast or digital
subtraction arteriography was performed as required, and particularly when duplex scanning indicated flow changes. Actuarial life tables were constructed by the Kaplan-Meier method to calculate cumulative graft patency rates.' 1 RESULTS Most dAVFs performed in this study were of the primary type. Secondary dAVFs are rarely performed ever since the indications for this procedure have been liberalized. Furthermore, there has been a decrease in the total number of dAVFs performed because of our commitment to using autologous veins for bypass material wherever possible. A typical
688
Journal of VASCULAR SURGERY
Dardik et al.
Fig. 4. Color Doppler scan of a distal anastomosis and dAVF depicts continuous prograde flow (red) and retrograde flow (blue) in venous component of the fistula. The red areas in the latter site are tributaries with reverse flow up to the valves. Table I. Hemodynamic parameters obtained by duplex scanning of grafts, distal arteries, and fistulas with patent or closed dAVFs Grafts
V V RI
Distal artery
Patent dA VF
Closed dA VF
Patent dA VF
Closed via VF
Fistula
174 ± 38 9 2 +_ 2 3 * 0.5 + 0 . 1 "
1 2 5 _+ 61 2 5 _+ 8 0 . 9 _+ 0.1
134 ± 39 52 ± 23 0.7 + 0.2
1 5 5 _+ 19 4 2 _ + 16 0 . 8 +_ 0.1
333 + 76 143 _+ 3 4 0.5 _+ 0 . 2
v, Peak systolicvelocity;V, mean velocity.;RI, resistivity.,index. ~p < o.01 distal anastomosis and dAVF as visualized by duplex scanning is shown in Fig. 1. Visualization of the peroneal arteries as demonstrated here was facilitated by the segmental absence of the fibula, a consequence of the surgical exposure. Doppler velocity waveforms of grafts with patent fistulas differed from those with occluded fistulas. Flow signals characteristic of a patent fistula and graft are shown in Fig. 2. Diastolic flow is increased with a patent fistula, signifying a lower resistance. Decreased or absent diastolic flow and flow reversal are indicative of a closed dAVF. Two-year cumulative graft patency rates for groups 1, 2, and 3 were 18%, 33%, and 44%, respectively. These results are statistically insignificant from one another but do show a trend of improvement from one period to another. Cumulative graft and fistula patency rates for 22 patients in whom postoperative duplex examinations were performed are shown in Fig. 3. At the end of 12 and 24 months graft patency exceeds fistula patency by 24.7% and 26.2%, respectively, with the fistula being open in 68% of those
grafts patent at 12 months and 58% of those patent at 24 months. Comparative results (mean _+ SD) of peak systolic velocity, mean velocity, and resistivity index measured from Doppler signals in the grafts with patent and occluded dAVFs are presented in Table I. There was a significant (p < 0.01) difference between the two groups (open or closed dAVF) for graft mean velocity and resistivity index. There were no significant differences in the flow measurements or resistance calculations for the arteries beyond the distal anastomosis. Direction of flow was ascertained, and in all cases graft and distal artery flow was caudad; flow direction in the upper venous component of the fistula was cephalad. Additionally color Doppler studies clearly showed prograde flow into the distal arteries with absolutely no steal (Fig. 4). Postoperative arteriograms by standard contrast or digital subtraction technique also demonstrated excellent distal perfusion even with high flows into the venous circuit via the fistula (Fig. 5). Postoperative
Volume 13 Number 5 May 1991
Graft patency with the distal arteriovenousfistula
689
Table II. Results (mean _+ SD) of preoperative and postoperative ankle pressures and ankle/brachial indexes in patients with patent and occluded dAVFs Patent dA VF Preoperative AP ABI Postoperative AP ABI
47.1 0.31 136.7 0.90
+ 36.6 + 0.22 + 43.6 +- 0.10
Occluded dA VF 31.3 0.23 126.8 0.96
+- 24.6 _+ 0.20 + 24.5 -+ 0.00
AP, ankle pressure; ABI, ankle brachial index.
ankle pressures and ankle/brachial indexes for both the patent and closed dAVF groups (patent grafts) were significantly improved from preoperative levels, p < 0.01, (Table II). There were no differences between these two groups either before or after operation. DISCUSSION
Graft failure is often associated with high resistance in the distal flow circuit. Methods that have been suggested for maintaining graft patency under these circumstances include construction of a sequential component anticoagulant therapy,~2-~ improving rheologic characteristics, ~ and venous arterialization. ~6 The concept and theoretic basis for using dAVFs to maintain both graft patency and distal perfusion is based on the need to reduce overload on a fixed distal arterial circuit and simultaneously keep graft flow above the critical thrombotic threshold level. ~6~8 Although most of the graft flow is directed into the low-resistance, high-capacitance venous system, blood will still perfuse the distal arterial circuit. This increment, albeit limited as a function of the runoff capacitance, will nonetheless reverse the ischemic state. Considerable concern has been expressed regarding the possibility for dAVFs to steal blood from the distal circulation. ~9"2° Recommendations to avoid this situation include placement of the fistula distal or proximal to the distal anastomosis and limiting the size of the fistula. 2°-~2 We have documented directly by completion arteriography, postoperative routine contrast and digital subtraction studies, and directional duplex scanning that flow is prograde in the distal arterial circuit in the presence of a functional dAVF. Although peak systolic velocity in the arteries distal to a patent fistula was 17% lower than in those with no fistula, this difference was not statistically significant. Mean velocity flows in the distal arteries of both groups were also similar, confirming that distal arterial flows are comparable whether or not a dAVF is present. Peak systolic velocities in grafts with
Fig. 5. Postoperative digital subtraction arteriogram of a femoral peroneal bypass with dAVF. In addition to the excellent cephalad perfusion of the venous system, distal prograde flow into the peroneal artery is obvious. patent fistulas were 29% higher than in those grafts with closed fistulas, but this difference was not statistically different. On the other hand, mean velocity, a more accurate value for determining volume flow, was significantly higher (74%) in those grafts with patent dAVFs (p < 0.01). These data indicate that total flow into the venous circuit and distal arteries is sustained by the increased volume and velocity produced in those grafts constructed with an adjunctive dAVF. A low peak velocity in grafts with occluded fistulas is predictive for graft failure. Nevertheless, some patients continue to maintain graft patency even under the most adverse conditions. Perhaps, there may be a role for graft "acclimatization" during the period o f d A V F function. The possibility that the dAVF may never have been necessary from the very
690
) ournaJ ot VAY;CULAR StJRGERY
Dardik et al.
b e g i n n i n g is also true. A p r e l i m i n a r y s t u d y o f graft a n d fistula p a t e n c y b y d u p l e x s o n o g r a p h y s h o w e d t h a t o n e f o u r t h o f all the grafts w e r e p a t e n t w i t h o u t fistulas at 1 a n d 2 years after o p e r a t i o n . Fistula p a t e n c y was d e m o n s t r a t e d in 6 8 % o f those grafts still p a t e n t at 1 year a n d in 5 8 % o f those p a t e n t at 2 years. P r e v e n t i n g closure o f the fistula represents a challenge, the s o l u t i o n o f w h i c h will certainly enhance graft patency. O n the basis o f o u r clinical experience, w e have c o n c l u d e d t h a t c o n s t r u c t i o n o f a distal crural bypass w i t h i n a d e q u a t e arterial r u n o f f will o f t e n fail, particularly i f p r o s t h e t i c m a t e r i a l is used. I n these circumstances c o n s i d e r a t i o n s h o u l d be given to a d d i n g an adjunctive d A V F to increase the velocity o f b l o o d flow in the graft a n d t h e r e b y p o s s i b l y p r e v e n t t h r o m b o s i s . I n c r e a s e d peak systolic a n d m e a n velocity rates, a n d l o w e r resistivity indexes confirm t h a t graft flow increases significantly a n d that steal does n o t o c c u r d e s p i t e diversion o f m o s t o f this increased graft flow into the v e n o u s circuit. I m p r o v e d case selection w i t h specific criteria a n d precise surgical t e c h n i q u e will yield b e t t e r p a t e n c y rates a n d increase the l i k e l i h o o d for l i m b salvage. L o n g - t e r m p a t e n c y continues as a m a j o r challenge because o f progressive distal arterial disease a n d m y o i n t i m a l hyperplasia, particularly at the distal anastomosis. F u r t h e r research is essential to assess the c o n t r i b u t i o n o f the d A V F to graft p a t e n c y results a n d the clinical i m p a c t o f e n h a n c e d flow o f b l o o d .
7.
8. 9. 10. 11. 12.
13.
14. 15. 16. 17.
REFERENCES
1. Kent KC, Whittemore AD, Mannick JA. Short-term midterm results of an all-autogenous tissue policy for infrainguinal reconstruction. J VASCSURG 1989;9:107-14. 2. Taylor LM Jr, Phinney ES, Porter MJ. Present status of reversed vein bypass for lower extremity revascularization. J VASCSURG 1986;3:288-97. 3. Dennis JW, Littooy FN, Greisler HP, Baker WH. Secondary vascular procedures with polytetrafluoroethylene grafts for lower extremity ischemia in a male veteran population. J VASC SUNG 1988;8:137-42. 4. Hobson RW, Lynch TG, Jamil Z, et al. Results of revascularization and amputation in severe lower extremity ischemia: a five year clinical experience. J VAse SURG 1985;2: 174-85. 5. Dardik H, Miller N, Dardik A, et al. A decade of experience with the glutaraldehyde-tanned human umbilical cord vein graft for revascularization of the lower limb. J VAse SURG 1988;7:336-46. 6. Veith FJ, Gupta SK, Ascer E, et al. Six-year prospective
DISCUSSION Dr. F r a n k Veith (New York, N.Y.). The concept of
using an arteriovenous fistula, or AVF, in the treatment of ischemic limbs has been around since 1902. It has
18.
19. 20. 21.
22.
multicenter randomized comparison of autologous saphenous vein and expanded polytetrafluoroethylene grafts in infrainguinal arterial reconstructions. J VASC SURG 1986;3: 10414. Charlesworth PM, Brewster DA, Darling RC, Robinson JG, Hallet JW. The fate of polytetrafluoroethylene grafts in lower limb bypass surgery: a six-year follow-up. Brit J Surg 1985; 72:896-9. Dardik H, Sussman B, Ibrahim IM, et al. Distal arteriovenous fistula as an adjunct to maintaining arterial and graft patency for limb salvage. Surgery 1983;94:478-86. Dardik H. The distal arteriovenous fistula: a useful adjunct or a passing fancy? Eur J Vasc Surg 1988;2:67-9. Dardik H. Arterial reconstruction in the lower extremity. New York: McGraw-Hill, 1985. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958;53:457-81. Ascer E, Veith FJ, Morin L, et al. Components of outflow resistance and their correlation with graft patency in lower extremity arterial reconstructions. J VASCSURG 1984; 1:81728. Bandyk DF, Schmitt DD, Seabrook GR, Adams MB, Towne JB. Monitoring functional patency of in situ saphenous vein bypasses: the impact of a surveillance protocol and elective revision. J VASCSURG 1989;9:286-96. Flinn WR, Rohrer MJ, Yao JST, McCarthy WJ, Fahey VA, Bergan JJ. Improved long-term patency of infragenicular polytetrafluoroethylene grafts. J VASCSURG 1988;7:685-90. Rutherford R, Jones DN, Bergentz SE, et al. Factors affecting the patency of infrainguinal bypass. J VASe SUR6 1988;8: 236-46. Johansen K, Bernstein EF. Revascularization of the ischemic canine hindlimb by arteriovenous reversal. Ann Surg 1979; 190:243-53. Dardik H. The use of an adjunctive arteriovenous fistula in distal extremity bypass grafts with outflow obstruction. In: Kempczinski RF, ed. The ischemic leg. Chicago: Year Book Medical Publishers, Inc, 1985:463-74. Sauvage LR, Wallor MW, Berge K, et al. Current arterial prosthesis - experimental evaluation by implantation in the carotid and circumflex coronary artery of the dog. Arch Surg 1979;114:687-91. McGovern PJ, Jain KM, Kerr JC, Swan KG, Rocko JM. Haemodynamics of an anastomotic arteriovenous fistula. Am J Surg 1985;149:368-70. Paty PSK, Shah DM, SaifiJ, et al. Remote distal arteriovenous fstula to improve infrapopliteal bypass patency. J VAse SUR~; 1990;11:171-8. Harris PL. Adjuvant arteriovenous fistula at the distal anastomosis of a femorotibial bypass graft. In: Greenhalgh RM, ed. Vascular surgical techniques: an atlas. London, England: WB Saunders, 1989:306-9. Maurer PC, Sommoggy SV, Dorrler J, Esch U. Femoro-distal PTFE bypasses with integrated AV-fistulas for limb salvage. Sixteenth Annual Symposium on Current Critical Problems and New Horizons in Vascular Surgery, Montefiore Medical Center. November 17, 1989, New York, New York.
fascinated some great names in surgery including Carrell and Halsted. An extensive literature exists, and the AVF concept has gained popularity and then been rejected three times during the last 90 years. It is therefore not surprising
Volume 13 Number 5 May 1991
that with the increasing interest in limb salvage surgery over the last decade there has been renewed interest in use of the venous tree as a means to overcome arterial ischemia. Drs. Johansen and Bernstein began the modern reexamination of this topic in the laboratory, and Dr. Dardik has led the reexamination of the AVF in patients. His paper claims improved graft patency by using a distal adjunctive AVF. The paper is provocative. It clearly shows that AVFs probably do little harm and do not steal from the arterial outflow bed. However, the paper does not prove that AVFs increase prosthetic graft patency. The problem is that prosthetic grafts with AVFs sometimes can work with very bad outflow. In our randomized prospective study prosthetic tibial grafts without AVFs had a 38% 2-year patency-not greatly different from Dr. Dardik's 2-year patency with A V F s - a n d patency with sodium warfarin (Coumadin), and no AVF has been even better. Despite these skeptical observations, I would like to make some positive points regarding adjunctive AVFs. We have been interested in them to improve graft patency in otherwise impossible situations since 1978. We have used a variety of adjunctive fistulas over the last 12 years in 21 patients, most of whom had just had acute thrombosis of a PTFE tibial bypass. These fistulas have been of three types: the Dardik common ostium in five, the Shah distal AVF in four, and 12 AVFs of a new type introduced by Dr. Ascer and other members of our group. Nine of our 21 grafts with AVFs have remained patent over 6 months, and six are patent over 1 year. All recent patients are also on Coumadin. Some of these grafts are patent in remarkably bad circumstances. Thus despite my skepticism, I believe the dAVF may have value for the rare patient requiring a very distal prosthetic graft, particularly after such a graft thromboses. I would like to ask Dr. Dardik three questions. Have you seen venous stasis changes in the foot as we have in our saphenous turn down patients? Do you think retrograde venous perfusion ever has any nutritive value as we and some others believe? And, do you anticoagulate your AVF patients ? Dr. Diraj Shah (Albany, N.Y.). Dr. Fry, members and guests. I, too, congratulate Dr. Dardik for his excellent presentation of a 10-year experience with adjunctive dAVF in conjunction with tibial artery bypass. Distal arteriovenous fistula increases the blood flow through the graft, possibly above its thrombotic threshold thereby improving the patency of the graft. Initial results of this technique support this hypothesis. We have also used a remote dAVF to improve graft patency in synthetic tibial artery bypasses. Since 1987 we have performed a remote dAVF in 22 patients. These were highly selected patients in whom multiple previous vein and synthetic graft bypass had failed. They also had diagnosed hypercoagulable states. One-year graft patency was 77%, and limb salvage was achieved in these patients. We do not have many patients going beyond 2 years, and I think long-term patency may still be a problem as it was noted in this paper.
GraB patency with the distal arteriovenousfistula 691
Although our techniques are somewhat different from those of Dr. Dardik's, nonetheless, the basic concept of increasing graft blood flow by providing a concomitant low resistance vascular bed is the same. I have one question. So far, this technique has been used in a disadvantaged group of patients as a last resort. Since at least the initial result is hopeful, should we use this technique more liberally, especially in patients with vein grafts to limited outflow tracts? In that context, did you see any difference in performance of nine dAVFs done with autogenous vein in your series compared to synthetic grafts? Dr. Dardik. Thank you, Dr. Veith and Dr. Shah. The challenge of the dAVF procedure has opened up new questions, more than we can provide answers for. First and foremost is the question of steal, and I hope that I provided some evidence today to indicate that this does not occur in the clinical setting of limb-threatening ischemia. The actual results of performing synthetic graft reconstructions using dAVFs will, I believe, improve with time. There is a steep learning curve, and many who attempt this procedure for the first time will fail purely on the basis of technique. Our data show the step-wise improvement with time based on improved technical skills and patient selection. Dr. Veith, you asked about complications relating to the venous circuit. We have not seen any other than a case referred to me with a functioning posterior tibial bypass and dAVF suffering from massive swelling of the leg from the ankle to the knee with open sores and weeping of lymph. We discovered that the popliteal vein had been inadvertantly ligated during an earlier operation. We corrected that by placement of an interposition graft in the venous circuit, and for the ensuing 2 years that we followed the patient there was a remarkable reversal of the swelling and pain. In fact, the fistula was maintaining both the arterial inflow and the venous return. With regard to the question about retrograde nutritional flow via the veins, the famous physiologist, Bernard Zweifach, showed many years ago that, in the frog, retrograde nutritional flow can occur. Subsequently, Dr. Bernstein showed in the experimental laboratory that such flow can be staged in the dog, and by their method of arteriovenous reversal ischemia can be prevented. In our clinical experience we found no durable patency with grafts that depended solely on the venous outflow circuit for patency. There must be some arterial runoff to reverse ischemia, but nevertheless we suspect that there is a secondary role for retrograde venous perfusion that may occur with time. You also asked about the role of anticoagulants. Our data are very soft in this regard. We do put all of our patients on Coumadin after the lead of the Chicago group, but most important, and this is also in response to Dr. Shah's question, I think the real challenge at this time is to be able to prevent the development ofintimal hyperplasia. This process is significant in causing intermediate and late fistula closure and ultimately closure of the graft. Correction of this problem will obviously enhance graft and fistula patency and limb salvage results.
1991;13:685-91.)
The excellent clinical results obtained by the use of the autologous saphenous vein for lower limb revascularization has diminished the need for the use of prosthetic arterial substitutes. 1"2 Nevertheless, there are occasions where autologous tissue may not be available or suitable so that a prosthetic graft may be preferable or the only alternative to amputation. 3-s The results of prosthetic graft bypasses to the tibia] and peroneal arteries have been shown to be decidedly inferior to those obtained with autologous veins. 4-7 To circumvent these problems, the major surgical approaches to augment graft patency rates have been (1) construction of a sequential component to the bypass graft and (2) creation of an From the Vascular Surgical Service,EnglewoodHospital, Englewood. Presented at the Forty-fourthAnnual Meeting of the Societyfor Vascular Surgery, Los Angeles, Calif., June 4-6, 1990. Reprint requests: Herbert Dardik, MD, 375 Engle St., Englewood, NJ 07631. 24/1/26927
arteriovenous fistula (dAVF) in relationship to the distal anastomosis of the prosthetic graft. The former, if dual runoff vessels are not present, may be neither possible nor effective if they overlap similar anatomic patterns of distribution. The latter may also not be possible since dAVF function has a critical dependence on the configuration and quality of the deep venous system, s-l° Since 1979 we have been actively involved in these adjunctive distal procedures to enhance prosthetic and, at times, autologous graft patency rates. This article summarizes our experience and conclusions relating to the indications for and results of using these procedurcs. MATERIAL A N D M E T H O D S During the 10-year period, from November 1979 through December 1989, 210 dAVFs were constructed in 203 patients. The group consisted of 127 men and 76 women aged 31 to 85 years (mean 68 years). One hundred six (52%) patients had diabetes. Prior vascular reconstructive procedures had been 685
lournJof VASCULAR SURGERY
686 Dardik et al.
Fig. 1. Duplex scan of a distal anastomosis of an umbilical vein graft (HUV) to the peroneal artery (D, distal; P, proximal) with a dAVF (Fistula).
performed in the involved limb in 130 (63%) instances, 15 of these involved two or more procedures. Group 1, previously reported in 1983, 8 consisted of 61 dAVFs (58 patients) and provided the foundation for understanding the critical elements of technique and appreciating the status of the deep venous system. Group 2 patients had 80 dAVFs created between March 1983 and January 1986. This time period was synchronous with the transition in our institution from the use of prosthetics to autologous vein as the preferred graft material. Group 3 represents those patients undergoing bypasses and dAVFs (n = 69) from January 1986 through December 1989, a time period when our policy for autologous tissue as the conduit of choice was operative and use ofprosthetics and dAVFs was restricted. Nine patients in this series underwent dAVF construction as an adjunct to autologous vein bypass surgery, six of whom had failed prior autologous vein bypasses. The specific indications for creating an adjunctive dAVF reported earlier have now been simplified to include all bypasses performed to the crural arteries where a prosthetic is used. 8 Relative contraindications to performing a dAVF include virtual absence of arterial runoff, small venae comitantes, and extensive sclerosis or other sequelae of deep venous thrombosis. Construction of the dAVF was usually by a side-to-side technique. Infrequently, variations were used such as placing the fistula remote from the distal
graft anastomosis or using only the cephalad segment. Most of the dAVFs were performed as primary adjuncts, that is, synchronous construction of both the dAVF and the bypass. Secondary dAVFs were those performed as a separate procedure along with thrombectomy within the early postoperative period of a previously placed graft. Prosthetic graft materials consisted of glutaraldehyde-tanned human umbilical cord vein and, in one instance, thin-walled polytetrafluoroethylene. Autologous veins were usually from the saphenous vein systems, greater and lesser, and some from arm veins. These were used in various configurations, singly or spliced. Completion arteriography was performed routinely and occasionally complemented with flow studies by use of the Statham-Gould electromagnetic flowmeter (model SP-2204). Confirmation of prograde flow in the distal artery and cephalad flow in the upper venous component of the fistula was additionally obtained with an intraoperative Doppler (Bach Simpson Ltd, model 348). After discharge from the hospital follow-up was performed at 3-week intervals for the first 6 months and more frequently as required for wound care and other conditions. Subsequent surveillance was performed at 3-month intervals and extended where appropriate. Noninvasive testing consisted of duplex scanning and pulse volume recordings with anklebrachial pressure measurements and waveform analysis. The former was performed with the patients in a supine position by use of a mechanical 8 MHz
Volume 13 Number 5 May 1991
Graft patency with the distal arteriovenous fistula
687
Fig. 2. Characteristic flow signals of a patent fistula (left) and graft (right).
100
i
.......
v n~ Q. u_
!
80
60
26.2
\11 s2.6
U
6•2.6 %4
40 Cure. Graft Cure. Fistula 20 Fist. Pat. = 68% of Pat. Grafts I
months
6
I
Fist.Pat.= 58 % of Pat. Grafts I
12
24
Fig. 3. Cumulative graft and fistula patency rates for 22 patients with duplex follow-up studies. One fourth of all the grafts were patent without fistula patency at 1 and 2 years. Of those grafts patent at 1 year, 68% still had patent fistulas. At 2 years, fistula patency occurred in 58% of the grafts patent at that time.
sector scanner with 6.5 M H z angle corrected Doppler. The entire length of the graft was imaged from the proximal to the distal anastomosis. The dAVF and the artery beyond the distal anastomosis were similarly evaluated. For each site, peak systolic velocity (V) and mean velocity (9) were measured. A resistivity index (RI) was calculated: RI=
Peak systolic velocity-Diastolic velocity Peak systolic velocity
With open dAVFs, diastolic flow is high, consistent with a low resistance. The converse is true with a closed dAVF. Finally, direction o f flow was noted at all sites. Postoperative standard contrast or digital
subtraction arteriography was performed as required, and particularly when duplex scanning indicated flow changes. Actuarial life tables were constructed by the Kaplan-Meier method to calculate cumulative graft patency rates.' 1 RESULTS Most dAVFs performed in this study were of the primary type. Secondary dAVFs are rarely performed ever since the indications for this procedure have been liberalized. Furthermore, there has been a decrease in the total number of dAVFs performed because of our commitment to using autologous veins for bypass material wherever possible. A typical
688
Journal of VASCULAR SURGERY
Dardik et al.
Fig. 4. Color Doppler scan of a distal anastomosis and dAVF depicts continuous prograde flow (red) and retrograde flow (blue) in venous component of the fistula. The red areas in the latter site are tributaries with reverse flow up to the valves. Table I. Hemodynamic parameters obtained by duplex scanning of grafts, distal arteries, and fistulas with patent or closed dAVFs Grafts
V V RI
Distal artery
Patent dA VF
Closed dA VF
Patent dA VF
Closed via VF
Fistula
174 ± 38 9 2 +_ 2 3 * 0.5 + 0 . 1 "
1 2 5 _+ 61 2 5 _+ 8 0 . 9 _+ 0.1
134 ± 39 52 ± 23 0.7 + 0.2
1 5 5 _+ 19 4 2 _ + 16 0 . 8 +_ 0.1
333 + 76 143 _+ 3 4 0.5 _+ 0 . 2
v, Peak systolicvelocity;V, mean velocity.;RI, resistivity.,index. ~p < o.01 distal anastomosis and dAVF as visualized by duplex scanning is shown in Fig. 1. Visualization of the peroneal arteries as demonstrated here was facilitated by the segmental absence of the fibula, a consequence of the surgical exposure. Doppler velocity waveforms of grafts with patent fistulas differed from those with occluded fistulas. Flow signals characteristic of a patent fistula and graft are shown in Fig. 2. Diastolic flow is increased with a patent fistula, signifying a lower resistance. Decreased or absent diastolic flow and flow reversal are indicative of a closed dAVF. Two-year cumulative graft patency rates for groups 1, 2, and 3 were 18%, 33%, and 44%, respectively. These results are statistically insignificant from one another but do show a trend of improvement from one period to another. Cumulative graft and fistula patency rates for 22 patients in whom postoperative duplex examinations were performed are shown in Fig. 3. At the end of 12 and 24 months graft patency exceeds fistula patency by 24.7% and 26.2%, respectively, with the fistula being open in 68% of those
grafts patent at 12 months and 58% of those patent at 24 months. Comparative results (mean _+ SD) of peak systolic velocity, mean velocity, and resistivity index measured from Doppler signals in the grafts with patent and occluded dAVFs are presented in Table I. There was a significant (p < 0.01) difference between the two groups (open or closed dAVF) for graft mean velocity and resistivity index. There were no significant differences in the flow measurements or resistance calculations for the arteries beyond the distal anastomosis. Direction of flow was ascertained, and in all cases graft and distal artery flow was caudad; flow direction in the upper venous component of the fistula was cephalad. Additionally color Doppler studies clearly showed prograde flow into the distal arteries with absolutely no steal (Fig. 4). Postoperative arteriograms by standard contrast or digital subtraction technique also demonstrated excellent distal perfusion even with high flows into the venous circuit via the fistula (Fig. 5). Postoperative
Volume 13 Number 5 May 1991
Graft patency with the distal arteriovenousfistula
689
Table II. Results (mean _+ SD) of preoperative and postoperative ankle pressures and ankle/brachial indexes in patients with patent and occluded dAVFs Patent dA VF Preoperative AP ABI Postoperative AP ABI
47.1 0.31 136.7 0.90
+ 36.6 + 0.22 + 43.6 +- 0.10
Occluded dA VF 31.3 0.23 126.8 0.96
+- 24.6 _+ 0.20 + 24.5 -+ 0.00
AP, ankle pressure; ABI, ankle brachial index.
ankle pressures and ankle/brachial indexes for both the patent and closed dAVF groups (patent grafts) were significantly improved from preoperative levels, p < 0.01, (Table II). There were no differences between these two groups either before or after operation. DISCUSSION
Graft failure is often associated with high resistance in the distal flow circuit. Methods that have been suggested for maintaining graft patency under these circumstances include construction of a sequential component anticoagulant therapy,~2-~ improving rheologic characteristics, ~ and venous arterialization. ~6 The concept and theoretic basis for using dAVFs to maintain both graft patency and distal perfusion is based on the need to reduce overload on a fixed distal arterial circuit and simultaneously keep graft flow above the critical thrombotic threshold level. ~6~8 Although most of the graft flow is directed into the low-resistance, high-capacitance venous system, blood will still perfuse the distal arterial circuit. This increment, albeit limited as a function of the runoff capacitance, will nonetheless reverse the ischemic state. Considerable concern has been expressed regarding the possibility for dAVFs to steal blood from the distal circulation. ~9"2° Recommendations to avoid this situation include placement of the fistula distal or proximal to the distal anastomosis and limiting the size of the fistula. 2°-~2 We have documented directly by completion arteriography, postoperative routine contrast and digital subtraction studies, and directional duplex scanning that flow is prograde in the distal arterial circuit in the presence of a functional dAVF. Although peak systolic velocity in the arteries distal to a patent fistula was 17% lower than in those with no fistula, this difference was not statistically significant. Mean velocity flows in the distal arteries of both groups were also similar, confirming that distal arterial flows are comparable whether or not a dAVF is present. Peak systolic velocities in grafts with
Fig. 5. Postoperative digital subtraction arteriogram of a femoral peroneal bypass with dAVF. In addition to the excellent cephalad perfusion of the venous system, distal prograde flow into the peroneal artery is obvious. patent fistulas were 29% higher than in those grafts with closed fistulas, but this difference was not statistically different. On the other hand, mean velocity, a more accurate value for determining volume flow, was significantly higher (74%) in those grafts with patent dAVFs (p < 0.01). These data indicate that total flow into the venous circuit and distal arteries is sustained by the increased volume and velocity produced in those grafts constructed with an adjunctive dAVF. A low peak velocity in grafts with occluded fistulas is predictive for graft failure. Nevertheless, some patients continue to maintain graft patency even under the most adverse conditions. Perhaps, there may be a role for graft "acclimatization" during the period o f d A V F function. The possibility that the dAVF may never have been necessary from the very
690
) ournaJ ot VAY;CULAR StJRGERY
Dardik et al.
b e g i n n i n g is also true. A p r e l i m i n a r y s t u d y o f graft a n d fistula p a t e n c y b y d u p l e x s o n o g r a p h y s h o w e d t h a t o n e f o u r t h o f all the grafts w e r e p a t e n t w i t h o u t fistulas at 1 a n d 2 years after o p e r a t i o n . Fistula p a t e n c y was d e m o n s t r a t e d in 6 8 % o f those grafts still p a t e n t at 1 year a n d in 5 8 % o f those p a t e n t at 2 years. P r e v e n t i n g closure o f the fistula represents a challenge, the s o l u t i o n o f w h i c h will certainly enhance graft patency. O n the basis o f o u r clinical experience, w e have c o n c l u d e d t h a t c o n s t r u c t i o n o f a distal crural bypass w i t h i n a d e q u a t e arterial r u n o f f will o f t e n fail, particularly i f p r o s t h e t i c m a t e r i a l is used. I n these circumstances c o n s i d e r a t i o n s h o u l d be given to a d d i n g an adjunctive d A V F to increase the velocity o f b l o o d flow in the graft a n d t h e r e b y p o s s i b l y p r e v e n t t h r o m b o s i s . I n c r e a s e d peak systolic a n d m e a n velocity rates, a n d l o w e r resistivity indexes confirm t h a t graft flow increases significantly a n d that steal does n o t o c c u r d e s p i t e diversion o f m o s t o f this increased graft flow into the v e n o u s circuit. I m p r o v e d case selection w i t h specific criteria a n d precise surgical t e c h n i q u e will yield b e t t e r p a t e n c y rates a n d increase the l i k e l i h o o d for l i m b salvage. L o n g - t e r m p a t e n c y continues as a m a j o r challenge because o f progressive distal arterial disease a n d m y o i n t i m a l hyperplasia, particularly at the distal anastomosis. F u r t h e r research is essential to assess the c o n t r i b u t i o n o f the d A V F to graft p a t e n c y results a n d the clinical i m p a c t o f e n h a n c e d flow o f b l o o d .
7.
8. 9. 10. 11. 12.
13.
14. 15. 16. 17.
REFERENCES
1. Kent KC, Whittemore AD, Mannick JA. Short-term midterm results of an all-autogenous tissue policy for infrainguinal reconstruction. J VASCSURG 1989;9:107-14. 2. Taylor LM Jr, Phinney ES, Porter MJ. Present status of reversed vein bypass for lower extremity revascularization. J VASCSURG 1986;3:288-97. 3. Dennis JW, Littooy FN, Greisler HP, Baker WH. Secondary vascular procedures with polytetrafluoroethylene grafts for lower extremity ischemia in a male veteran population. J VASC SUNG 1988;8:137-42. 4. Hobson RW, Lynch TG, Jamil Z, et al. Results of revascularization and amputation in severe lower extremity ischemia: a five year clinical experience. J VAse SURG 1985;2: 174-85. 5. Dardik H, Miller N, Dardik A, et al. A decade of experience with the glutaraldehyde-tanned human umbilical cord vein graft for revascularization of the lower limb. J VAse SURG 1988;7:336-46. 6. Veith FJ, Gupta SK, Ascer E, et al. Six-year prospective
DISCUSSION Dr. F r a n k Veith (New York, N.Y.). The concept of
using an arteriovenous fistula, or AVF, in the treatment of ischemic limbs has been around since 1902. It has
18.
19. 20. 21.
22.
multicenter randomized comparison of autologous saphenous vein and expanded polytetrafluoroethylene grafts in infrainguinal arterial reconstructions. J VASC SURG 1986;3: 10414. Charlesworth PM, Brewster DA, Darling RC, Robinson JG, Hallet JW. The fate of polytetrafluoroethylene grafts in lower limb bypass surgery: a six-year follow-up. Brit J Surg 1985; 72:896-9. Dardik H, Sussman B, Ibrahim IM, et al. Distal arteriovenous fistula as an adjunct to maintaining arterial and graft patency for limb salvage. Surgery 1983;94:478-86. Dardik H. The distal arteriovenous fistula: a useful adjunct or a passing fancy? Eur J Vasc Surg 1988;2:67-9. Dardik H. Arterial reconstruction in the lower extremity. New York: McGraw-Hill, 1985. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958;53:457-81. Ascer E, Veith FJ, Morin L, et al. Components of outflow resistance and their correlation with graft patency in lower extremity arterial reconstructions. J VASCSURG 1984; 1:81728. Bandyk DF, Schmitt DD, Seabrook GR, Adams MB, Towne JB. Monitoring functional patency of in situ saphenous vein bypasses: the impact of a surveillance protocol and elective revision. J VASCSURG 1989;9:286-96. Flinn WR, Rohrer MJ, Yao JST, McCarthy WJ, Fahey VA, Bergan JJ. Improved long-term patency of infragenicular polytetrafluoroethylene grafts. J VASCSURG 1988;7:685-90. Rutherford R, Jones DN, Bergentz SE, et al. Factors affecting the patency of infrainguinal bypass. J VASe SUR6 1988;8: 236-46. Johansen K, Bernstein EF. Revascularization of the ischemic canine hindlimb by arteriovenous reversal. Ann Surg 1979; 190:243-53. Dardik H. The use of an adjunctive arteriovenous fistula in distal extremity bypass grafts with outflow obstruction. In: Kempczinski RF, ed. The ischemic leg. Chicago: Year Book Medical Publishers, Inc, 1985:463-74. Sauvage LR, Wallor MW, Berge K, et al. Current arterial prosthesis - experimental evaluation by implantation in the carotid and circumflex coronary artery of the dog. Arch Surg 1979;114:687-91. McGovern PJ, Jain KM, Kerr JC, Swan KG, Rocko JM. Haemodynamics of an anastomotic arteriovenous fistula. Am J Surg 1985;149:368-70. Paty PSK, Shah DM, SaifiJ, et al. Remote distal arteriovenous fstula to improve infrapopliteal bypass patency. J VAse SUR~; 1990;11:171-8. Harris PL. Adjuvant arteriovenous fistula at the distal anastomosis of a femorotibial bypass graft. In: Greenhalgh RM, ed. Vascular surgical techniques: an atlas. London, England: WB Saunders, 1989:306-9. Maurer PC, Sommoggy SV, Dorrler J, Esch U. Femoro-distal PTFE bypasses with integrated AV-fistulas for limb salvage. Sixteenth Annual Symposium on Current Critical Problems and New Horizons in Vascular Surgery, Montefiore Medical Center. November 17, 1989, New York, New York.
fascinated some great names in surgery including Carrell and Halsted. An extensive literature exists, and the AVF concept has gained popularity and then been rejected three times during the last 90 years. It is therefore not surprising
Volume 13 Number 5 May 1991
that with the increasing interest in limb salvage surgery over the last decade there has been renewed interest in use of the venous tree as a means to overcome arterial ischemia. Drs. Johansen and Bernstein began the modern reexamination of this topic in the laboratory, and Dr. Dardik has led the reexamination of the AVF in patients. His paper claims improved graft patency by using a distal adjunctive AVF. The paper is provocative. It clearly shows that AVFs probably do little harm and do not steal from the arterial outflow bed. However, the paper does not prove that AVFs increase prosthetic graft patency. The problem is that prosthetic grafts with AVFs sometimes can work with very bad outflow. In our randomized prospective study prosthetic tibial grafts without AVFs had a 38% 2-year patency-not greatly different from Dr. Dardik's 2-year patency with A V F s - a n d patency with sodium warfarin (Coumadin), and no AVF has been even better. Despite these skeptical observations, I would like to make some positive points regarding adjunctive AVFs. We have been interested in them to improve graft patency in otherwise impossible situations since 1978. We have used a variety of adjunctive fistulas over the last 12 years in 21 patients, most of whom had just had acute thrombosis of a PTFE tibial bypass. These fistulas have been of three types: the Dardik common ostium in five, the Shah distal AVF in four, and 12 AVFs of a new type introduced by Dr. Ascer and other members of our group. Nine of our 21 grafts with AVFs have remained patent over 6 months, and six are patent over 1 year. All recent patients are also on Coumadin. Some of these grafts are patent in remarkably bad circumstances. Thus despite my skepticism, I believe the dAVF may have value for the rare patient requiring a very distal prosthetic graft, particularly after such a graft thromboses. I would like to ask Dr. Dardik three questions. Have you seen venous stasis changes in the foot as we have in our saphenous turn down patients? Do you think retrograde venous perfusion ever has any nutritive value as we and some others believe? And, do you anticoagulate your AVF patients ? Dr. Diraj Shah (Albany, N.Y.). Dr. Fry, members and guests. I, too, congratulate Dr. Dardik for his excellent presentation of a 10-year experience with adjunctive dAVF in conjunction with tibial artery bypass. Distal arteriovenous fistula increases the blood flow through the graft, possibly above its thrombotic threshold thereby improving the patency of the graft. Initial results of this technique support this hypothesis. We have also used a remote dAVF to improve graft patency in synthetic tibial artery bypasses. Since 1987 we have performed a remote dAVF in 22 patients. These were highly selected patients in whom multiple previous vein and synthetic graft bypass had failed. They also had diagnosed hypercoagulable states. One-year graft patency was 77%, and limb salvage was achieved in these patients. We do not have many patients going beyond 2 years, and I think long-term patency may still be a problem as it was noted in this paper.
GraB patency with the distal arteriovenousfistula 691
Although our techniques are somewhat different from those of Dr. Dardik's, nonetheless, the basic concept of increasing graft blood flow by providing a concomitant low resistance vascular bed is the same. I have one question. So far, this technique has been used in a disadvantaged group of patients as a last resort. Since at least the initial result is hopeful, should we use this technique more liberally, especially in patients with vein grafts to limited outflow tracts? In that context, did you see any difference in performance of nine dAVFs done with autogenous vein in your series compared to synthetic grafts? Dr. Dardik. Thank you, Dr. Veith and Dr. Shah. The challenge of the dAVF procedure has opened up new questions, more than we can provide answers for. First and foremost is the question of steal, and I hope that I provided some evidence today to indicate that this does not occur in the clinical setting of limb-threatening ischemia. The actual results of performing synthetic graft reconstructions using dAVFs will, I believe, improve with time. There is a steep learning curve, and many who attempt this procedure for the first time will fail purely on the basis of technique. Our data show the step-wise improvement with time based on improved technical skills and patient selection. Dr. Veith, you asked about complications relating to the venous circuit. We have not seen any other than a case referred to me with a functioning posterior tibial bypass and dAVF suffering from massive swelling of the leg from the ankle to the knee with open sores and weeping of lymph. We discovered that the popliteal vein had been inadvertantly ligated during an earlier operation. We corrected that by placement of an interposition graft in the venous circuit, and for the ensuing 2 years that we followed the patient there was a remarkable reversal of the swelling and pain. In fact, the fistula was maintaining both the arterial inflow and the venous return. With regard to the question about retrograde nutritional flow via the veins, the famous physiologist, Bernard Zweifach, showed many years ago that, in the frog, retrograde nutritional flow can occur. Subsequently, Dr. Bernstein showed in the experimental laboratory that such flow can be staged in the dog, and by their method of arteriovenous reversal ischemia can be prevented. In our clinical experience we found no durable patency with grafts that depended solely on the venous outflow circuit for patency. There must be some arterial runoff to reverse ischemia, but nevertheless we suspect that there is a secondary role for retrograde venous perfusion that may occur with time. You also asked about the role of anticoagulants. Our data are very soft in this regard. We do put all of our patients on Coumadin after the lead of the Chicago group, but most important, and this is also in response to Dr. Shah's question, I think the real challenge at this time is to be able to prevent the development ofintimal hyperplasia. This process is significant in causing intermediate and late fistula closure and ultimately closure of the graft. Correction of this problem will obviously enhance graft and fistula patency and limb salvage results.
