Cephalic vein grafts for lower extremity revascularization M a r k E. Sesto, M D , T i m o t h y M. Sullivan, M D , N o r m a n R. Hertzer, M D , Leonard P. Krajewski, M D , Patrick J. O'Hara, M D , and E d w i n G. Beven, M D ,
Cleveland, Ohio, and Ft. Lauderdale, Fla. From 1980 to 1989 infrainguinal revascularization was performed with cephalic vein grafts in a consecutive series of 34 patients (35 limbs) whose saphenous veins were either inadequate or already had been harvested for previous coronary (N = 16, 47%) or ipsilateral lower extremity bypass (N = 19, 56%). Surgical indications included ischemic rest pain or focal tissue necrosis in 25 limbs (71%), disabling claudication in six (17%), and popliteal aneurysms or prosthetic femoropopliteal graft infections each in two (6%). Preliminary arteriovenous fistulas were constructed in the arms of 23 patients (68%) to enhance the diameter of their cephalic veins, and 24 (69%) of the 35 infrainguinal procedures in this series were performed with use of cephalic vein alone. The distal popliteal artery was used for the outflow anastomosis in 10 limbs (29%), a tibial vessel was used in 12 (34%), and the peroneal artery was used in 13 (37%). Fourteen graft occlusions (40%) and six amputations (17%) have occurred during follow-up intervals of i to 107 months (mean, 28 months; median, 27 months). At 3 years the cumulative primary patency rate is 40%, the secondary patency rate is 46%, and the limb salvage rate is 82%. Despite their relative inconvenience, cephalic vein grafts appear to be preferable to prosthetic materials for infrainguinal revascularization below the knee. (J VASC SURG 1992;15:543-9.)
The success of arterial bypass in the lower extremity is substantially influenced by the graft material with which it is performed. In their extensive review of the contemporary literature concerning femoropopliteal and distal revascularization, Dalman and Taylor 1 reconfirmed the principle that the late results of either reversed or in situ saphenous vein grafts are clearly superior to those associated with prosthetic bypass to the infrageniculate popliteal artery or the tibioperoneal vessels. According to their collected series, the 3-year primary patency rates (73% to 78%) for saphenous vein grafts constructed to the popliteal segment below the knee are approximately twice as favorable as those (44%) reported for polytetrafluorethylene (PTFE) bypass. These differences are even more impressive for grafts extending to the peroneal or tibial arteries (66% to 74% versus 21%, respectively). Although the advantages of saphenous vein From the Department of Vascular Surgery, The Cleveland Clinic Foundation (Cleveland, Ohio) and Cleveland Clinic Florida (Dr. Sesto) (Fort Lauderdale, Florida) Reprint requests: Norman R. Hertzer, MD, Department of Vascular Surgery, The Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195-5272. 24/1/32638
bypass are firmly established, alternative graft materials occasionally are necessary for limb salvage in patients whose ipsilateral saphenous systems either are inadequate because of their small size or have already been harvested for previous lower extremity or coronary artery revascularization. The contralateral greater saphenous vein can be used as a reversed graft in this setting, but even if it still is available, its use may be contraindicated in patients who predictably will require bypass below the knee in the opposite leg within the foreseeable future. In addition to the ipsilateral lesser saphenous vein, therefore, arm veins may represent the only option for autogenous graft material in selected patients for whom distal popliteal or tibioperoneal revascularization is unavoidable. In 1969 Kakkar 2 demonstrated in cadavers that the tensile strength of the cephalic vein appeared to be adequate for the purpose of arterial replacement. A number of clinical series subsequently have been reported in which cephalic vein grafts were used, often with only modest success, for femoropopliteal or distal revascnlarization.3-~° The largest of these was described by Andros et al.9 in which 88 "full-length" grafts were constructed to outflow vessels below the knee with the cephalic (77/88) and/or the basilic 543
544
Sestoet aL
veins. The 3-year primary patency rate for grafts performed for limb salvage in this series was approximately 63%, a promising figure in comparison to the discouraging intermediate-term results of prosthetic grafts under similar circumstances. The purpose of our report is to assess the outcome of all femoropopliteal or femorotibial-peroneal cephalic vein grafts performed at the Cleveland Clinic during the past decade.
PATIENTS AND METHODS Patient information From 1980 through 1989, 926 distal popliteal or tibioperoneal bypass grafts were constructed at this center. Thirty-five (3.8%) of these required the use of the cephalic vein(s) alone (N = 23; 65%) or in conjunction with composite segments of saphenous vein (N = 9; 26%) or PTFE (N = 3; 9%). This cohort included 26 men (76%) and eight women (24%) with a mean age of 61 years (range, 33 to 81 years). Seventeen (50%) of these patients were hypertensive, and 10 (29%) had diabetes, eight of whom (23%) received insulin. Disabling claudication was the indication for only six cephalic vein grafts (17%). Most of these grafts (N = 25; 71%) were necessary because of ischemic rest pain or focal gangrene, and the remaining four patients (12%) had either popliteal aneurysms or infected femoropopliteal prosthetic grafts. Thirty-one of the 35 procedures in this series have been performed since 1985, and 20 of those during the last 2 years (1988 to 1989) of the study period. Since our preferential adoption of in situ techniques several years ago, the diameter of the greater saphenous vein has become relatively less important in determining its adequacy as an arterial substitute. In our experience, surgical absence of the saphenous vein has been the most common indication for the use of cephalic vein grafts for lower extremity revascularization. The ipsilateral saphenous vein already had been harvested for previous coronary (N = 16; 47%) or femoropopliteal-tibioperoneal bypass (N = 17; 50%) in many of the 35 patients described in this report, and another 11 patients (31%) had undergone autogenous grafts in the contralateral limb.
Surgical management On the basis of preoperative angiograms, tibioperoneal outflow was represented by three vessels in eight limbs (23%), by two vessels in five (14%), and by a single vessel in 22 (63%). The common femoral artery or a segment of the superficial or deep femoral artery immediately adjacent to it was used as the
Journal of VASCULAR SURGERY
inflow source for all 35 cephalic vein grafts in this series, and the targets selected for the distal anastomosis are summarized in Table I. Although the popliteal artery below the knee was appropriate for the distal anastomosis in 10 limbs (29%), direct tibioperoneal revascularization was necessary in the remaining 25 (71%). Sequential bypass to multiple vessels was not used, but the cephalic vein segment was used to cross the knee joint in each of the three patients who required composite synthetic grafts. An operative angiogram was routinely obtained to determine the immediate technical result, and pulse volume recordings (segmental limb pressure measurements with analog waveform analysis) were performed during the same hospitalization after 33 (90%) of the 35 procedures. Approximately 20% of our patients received intravenous heparin (N = 9) or low molecular-weight dextran (N = 7) during the perioperative period, and six (18%) randomly were placed on long-term oral anticoagulation with sodium warfarin (Coumadin). Preliminary arteriovenous fistulas In an attempt to enhance the diameter and the handling characteristics of the cephalic vein, temporary arteriovenous fistulas were constructed in the arm(s) of 23 patients (68%) whose arm veins appeared to be of marginal size on the basis of simple physical examination. After the venous anatomy of the upper extremities had been assessed by operative phlebography, an end-to-side arteriovenous anastomosis was performed at one or more appropriate levels by use of either the radial artery at the wrist or the brachial artery just proximal to the antecubital crease (Fig. 1). These fistulas were permitted to mature for approximately 10 to 14 days before the cephalic veins were harvested for lower extremity grafting. Statistical analysis The follow-up interval for the 34 patients (35 limbs) in this study ranges from 1 to 107 months (mean, 28 months; median, 27 months). Late results concerning patient survival and limb salvage were established during a telephone canvass, but conclusions regarding graft patency are based strictly on the physical findings and objective data from the noninvasive vascular laboratory, which were obtained at the time of the most recent outpatient examination. Elective revisions of patent grafts were considered equivalent to graft occlusions in the determination of secondary patency rates. Paired comparisons were evaluated by use of Fisher's (two-tailed) exact test. Life-table informa-
Volume 15 Number 3 March 1992
Cephalic vein grafCs 545
Fig. 1. A, Preoperative phlebogram demonstrates a diminutive cephalic vein (arrow) in the right forearm; B, Another contrast study shows enlargement of the same vein (arrow) 2 weeks after an arteriovenous fistula had been constructed at the wrist; C, The harvested vein (arrow) after revascularization of the peroneal artery. The proximal segment of this cephalic vein graft was obtained from the contralateral arm. Table I. A summary o f the surgical management used in this series Graft material Cephalic vein Composite cephalic and saphenous veins Composite cephalic vein and prosthetic Total
Below-knee popliteal No. % 8 2 0 10
tion was calculated according to the method described by Cutler and Ederer, 11 and standard errors o f the mean were assessed with the Greenwood formula.
RESULTS N o postoperative deaths occurred in this series. Early thrombosis occurred in two grafts (6%), both o f which were successfully revised. Another imme-
23 6 29
Tibial No.
%
9 2 1 12
25 6 3 34
Peroneal No. % 6 5 2 13
17 14 6 37
Total No.
%
23 9 3 35
65 26 9 100
diate reoperation was necessary to control bleeding from a cephalic vein branch. The mean Doppler ankle/brachial index for all 35 limbs improved from 0.38 before operation to 0.91 after revascularization. Cumulative data concerning late survival, graft patency, and limb salvage are presented in Fig. 2 and in Table II. Four patients have died in their first, second, third, or eighth postoperative year, resulting in an actuarial 3-year survival rate o f 87% + 7%. In
546
Sesto et aL
1.000 .969 • -------(34) ° ~ (34)
100 -'J
~ rrv
Journal of VASCULAR SURGERY
o
~
.931 8 0 (30)
6
80
Table II. Cumulative 3-year results (and standard errors) after cephalic vein bypass
9 (20)
60
Cumulative res#lts Clinical features
1 yr %
2 yr %
3 yr %
97
-+3)
93 (-+5)
87 (-+7)
49 -+9) 68 --8) 82 -+7)
49 (-+9) 58 (-+10) 82 (-+7)
40 (_+11) 44 (-+11) 82 (+7)
(./3 4O
Survival
I >_ O Z UJ a_o~ LL < rr L9
.971
100
~(~5 80
\\
•
I
(35)
-q
40
Patency Primary Secondary
.578 (16) 0
(11o)
Limb salvage .442
.397
(7) 1
1.000
80 60
.675 o (34)
(31)
I
I
o SECONDARY
X ' """''~'~'~
40
O0~ rn
I • PRIMARY
o• , N (35)
60
100 LU (.9 < >
I
824
.824
.824
(35)
(27)
(17)
I
I
I
I
30 DAYS
1
2
3
POSTOPERATIVE YEARS
Fig. 2. Cumulative 3-year survival, patency, and limb salvage rates for this series of 34 patients (35 limbs).
addition to the two postoperative failures, another 14 cephalic vein grafts (40%) have either occluded (N = 11; 31%) for unknown reasons or have been electively revised (N = 3; 9%) to correct incidental lesions that were discovered during follow-up surveillance (Fig. 3). The 3-year primary patency rate currently is 40% _ 11%, the secondary patency rate is 4 4 % _ 11%, and the limb salvage rate is 82% - 7%. It is impractical to evaluate these cumulative results statistically because our series contains too few patients, but selected crude comparisons are summarized in Table III. Variance in the mean follow-up periods reflects the accelerated attrition of occluded grafts within unfavorable groups. These data generally support the importance of several factors (preoperative symptoms, adequacy of outflow, prosthetic requirements) that traditionally have been found to influence the outcome of lower extremity bypass below the knee. Nevertheless, only the difference between failure rates for grafts indicated for limb salvage (52%) and those performed for claudication
(10%) attained statistical significance (p = 0.028; a l l other values p _> 0.2). DISCUSSION
Countless reports have documented the consistently superior performance of saphenous vein grafts for below-knee femoropopliteal or femorotibialperoneal bypass in comparison to a number of prosthetic materials that have been introduced during the past 30 years. The merit of an "all autogenous" policy for these procedures has been assumed to include the use of arm veins under circumstances in which the saphenous system is inadequate or no longer available. Several sources, however, suggest that it may be unrealistic to expect the results of cephalic vein bypass to meet the high standard established by the saphenous vein for lower extremity revascularization. Although the seven cephalic-basilic vein grafts initially performed by Kakkar2 remained patent during the preparation of his original article, their follow-up periods ranged only from 4 to 12 months. Thirteen (81%) of 16 similar grafts in two other early series subsequently failed within 24 months of their construction,4 and cumulative 3- to 5-year patency rates of approximately 30% have since been reported by Whittemore et al.5 and by Schulman and Badhey. 7 Improved patency rates have been described by Graham and Lusby6 (58% at 3 years) and, in probably the most extensive experience with "fulllength" cephalic vein grafts that has yet been pubfished, by Harris et al.8 and Andros et al.9 (63% at 3 years, and 51% at 5 years, respectively). Substantial differences in patient selection (including the number of previous operations in the same limb) and in the quality of arm veins deemed to be acceptable for bypass purposes could easily exist among these series, but it still seems reasonable to conclude that the late
Volume 15 Number 3 March 1992
Cephalic vein grafis 547
Fig. 3. Two examples of dilation and thrombus formation occurring 4 months (A) and 8 months (B) after lower extremity bypass with cephalic vein grafts. results of cephalic vein grafts for lower extremity revascularization have been somewhat inconsistent. Despite this uncertainty, there occasionally is no preferable alternative to the cephalic vein as a bypass conduit in certain patients for whom arterial reconStruction below the knee is absolutely necessary. The contralateral saphenous vein offers obvious advantages as a reversed graft and should be the primary consideration under optimal circumstances. Nevertheless, if the saphenous system was inadequate in size in the ischemic limb, it often is likely to be so on the opposite side as well. Furthermore, the contralateral vein may already have been used during coronary artery surgery or, as was the case in eight (23%) of our patients, for previous femoropopliteal bypass. Finally, many patients also have occlusive arterial disease in the contralateral lower extremity, a feature that might also serve as a future indication for distal revascularization. If the contralateral sapahenous vein is gambled and lost in an unsuccessful bid to salvage
the original jeopardized limb, the ultimate outcome conceivably could be just as unfavorable on the opposite side. The durability of prosthetic grafts extending below the knee generally has been disappointing, especially when the distal anastomosis involves the peroneal artery or a tibial vessel. Although the use of synthetic materials sometimes is unavoidable, our experience suggests that the results of cephalic vein bypass are more favorable than those for prosthetic grafts even though they do not meet the expectations traditionally reserved for saphenous vein grafts. In Table IV, the 3-year cumulative patency rates for the 35 cephalic vein grafts in the present series are compared with those for a group of saphenous vein and PTFE grafts obtained from the computer registry sponsored by the Cleveland Vascular Society. 12 The primary patency rate (40%) for our cephalic vein grafts (25, or 71% of which were constructed to tibioperoneal targets) clearly did not attain the same
548
Journal of VASCULAR SURGERY
Sesto et al.
Table III. Follow-up results for this series of cephalic vein grafts Follow-up (too) Paired comparisons Surgical indication Claudication Severe ischemia Distal anastomosis Below-knee popliteal Tibial or peroneal Graft material Cephalic vein only Composite saphenous vein or prosthetic Preliminary AV fistula Constructed Not constructed Outflow vessels 2-3 1 Coumadin therapy Administered Not administered
Table IV. Comparison of 3-year patency rates for the 35 cephalic vein grafts in this series and for a total of 388 l~mb salvage procedures reported from the Cleveland Vascular Registry~1 3-Year cumulative patency Graft material Present series Cleveland Vascular Registry Saphenous vein Below-knee popliteal Tibioperoneal PTFE Below-knee popliteal Tibioperoneal
Primary (%)
Secondary (%)
35
40
44
120 98
69 43
69 43
101 69
30 18
32 19
No.
overall success as saphenous vein grafts. Perhaps more importantly, it seems to represent reasonable improvement in comparison to the patency rate of below-knee PTFE bypass to either the popliteal or the tibioperoneal level (30% and 18%, respectively). In 1971 Beals 13 first described the use of an adjunctive arteriovenous fistula to mature the cephalic vein before its deployment as an arterial replacement graft. We rediscovered this concept years later because of our experience with hemodialysis angioaccess, and preliminary fistulas were performed in 23 (68%) of the 34 patients in this series. Although a staging interval of 10 to 14 days was sufficient to enhance the diameter of small cephalic veins, the merit of this approach remains largely a matter of speculation. Late failure rates were similar among
Graft occlusions
No.
Range
Mean
No.
%
10 25
1-42 1-40
21 7
1 13
10 52
10 25
2-42 1-40
19 8
2 12
20 52
24 11
1-42 1-41
12 I0
8 6
33 55
23 12
1-42 1-41
13 8
8 6
35 5O
13 22
1-42 1-40
15 9
4 10
31 45
6 29
2-26 1-42
8 12
4 10
67 34
patients who received preliminary fistulas (35%) in comparison with those who did not (50%), and although arteriovenous fistulas theoretically could have contributed to the dilated appearance that subsequently was documented in a few of our grafts (Fig. 3), Andros et al.9 also observed graft ectasia in 17 (30%) of the 56 patients in their series who eventually underwent follow-up angiography. The cephalic vein often is abused in the forearm by frequent venipuncmres and intravenous infusions, but it usually is available in the upper arm where it may be evaluated before operation by phlebography or Doppler ultrasound (duplex) scanning) 4 It is an important resource for femoropopliteal or femorotibial-peroneal bypass in a select group of patients for whom disadvantaged prosthetic grafts otherwise would be required. Under these demanding circumstances, cephalic vein grafts have a specific place in the armamentarium of vascular surgeons. Ming Tan, PhD (Department of Biostatistics and Epidemiology) and Geri Locker (Research Division) performed the statistical analysis. REFERENCES
1. Dalman RL, Taylor LM. Basic data related to infrainguinal revascularization procedures. Ann Vasc Surg 1990;4:309-12. 2. Kakkar VV. The cephalic vein as a peripheral vascular graft. Surg Gynecol Obstet 1969;128:551-6. 3. Vellar ID, Doyle IC. The use of the cephalic and basilic veins as peripheral vascular grafts. Aust N Z J Surg 1970;40: 52-7. 4. Clayson KR, Edwards WH, Allen TK, Dale A. Arm veins for peripheral arterial reconstruction. Arch Surg 1976; 111:127680.
Volume 15 Number 3 March 1992
Cephalic vein grafls 549
5. Whittemore AD, Clowes AW, Couch NP, Mannick JA. Secondary femoropopliteal reconstruction. Arch Surg 1981; 193:35-42. 6. Graham JW, Lusby RJ. Infrapopliteat bypass grafting: use of upper limb vein alone and in autogenous composite grafts. Surgery 1982;91:646-9. 7. Schuhxlan ML, Badhey MR. Late results and angiographic evaluation of arm veins as long bypass grafts. Surgery 1982;92:1032-41. 8. Harris RW, Andros G, Dulawa LB, Oblath RW, Salles-Cunha SX, Apyan R. Successful long-term limb salvage using cephalic vein bypass grafts. Ann Surg 1984;200:785-92. 9. Andros G, Harris RW, Salles-Cunha SX, Dulawa LB, Oblath RW, Apyan RL. Arm veins for arterial revascularization of the leg: arteriographic and clinical observations. J VAsc SURG 1986;4:416-27. 10. Balshi JD, Cantelmo NL, Menzoian JO, LoGerfo FW. The
11. 12. 13. 14.
use of arm veins for infrainguinal bypass in end-stage peripheral vascular disease. Arch Surg 1989;124:1078-81. Cutler SJ, Ederer F. Maximum utilization of the life-table method in analyzing survival. J Chron Dis 1958;8:699-712. Rafferty TD, Avellone JC, Farrell CJ, et at. A metropolitan experience with infralnguinal revascularization. J VASCSFRG 1987;6:365-71. Beats RL. Surgically created arteriovenous fistula to augment the cephalic vein: use as an arterial bypass graft. N Engl J Med 1971;285:29-30. Salles-Cunha SX, Andros G, Harris RW, Dulawa LB, Oblath RW. Preoperative noninvasive assessment of arm veins to be used as bypass grafts in the lower extremities. J VASCSUI
Cleveland, Ohio, and Ft. Lauderdale, Fla. From 1980 to 1989 infrainguinal revascularization was performed with cephalic vein grafts in a consecutive series of 34 patients (35 limbs) whose saphenous veins were either inadequate or already had been harvested for previous coronary (N = 16, 47%) or ipsilateral lower extremity bypass (N = 19, 56%). Surgical indications included ischemic rest pain or focal tissue necrosis in 25 limbs (71%), disabling claudication in six (17%), and popliteal aneurysms or prosthetic femoropopliteal graft infections each in two (6%). Preliminary arteriovenous fistulas were constructed in the arms of 23 patients (68%) to enhance the diameter of their cephalic veins, and 24 (69%) of the 35 infrainguinal procedures in this series were performed with use of cephalic vein alone. The distal popliteal artery was used for the outflow anastomosis in 10 limbs (29%), a tibial vessel was used in 12 (34%), and the peroneal artery was used in 13 (37%). Fourteen graft occlusions (40%) and six amputations (17%) have occurred during follow-up intervals of i to 107 months (mean, 28 months; median, 27 months). At 3 years the cumulative primary patency rate is 40%, the secondary patency rate is 46%, and the limb salvage rate is 82%. Despite their relative inconvenience, cephalic vein grafts appear to be preferable to prosthetic materials for infrainguinal revascularization below the knee. (J VASC SURG 1992;15:543-9.)
The success of arterial bypass in the lower extremity is substantially influenced by the graft material with which it is performed. In their extensive review of the contemporary literature concerning femoropopliteal and distal revascularization, Dalman and Taylor 1 reconfirmed the principle that the late results of either reversed or in situ saphenous vein grafts are clearly superior to those associated with prosthetic bypass to the infrageniculate popliteal artery or the tibioperoneal vessels. According to their collected series, the 3-year primary patency rates (73% to 78%) for saphenous vein grafts constructed to the popliteal segment below the knee are approximately twice as favorable as those (44%) reported for polytetrafluorethylene (PTFE) bypass. These differences are even more impressive for grafts extending to the peroneal or tibial arteries (66% to 74% versus 21%, respectively). Although the advantages of saphenous vein From the Department of Vascular Surgery, The Cleveland Clinic Foundation (Cleveland, Ohio) and Cleveland Clinic Florida (Dr. Sesto) (Fort Lauderdale, Florida) Reprint requests: Norman R. Hertzer, MD, Department of Vascular Surgery, The Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195-5272. 24/1/32638
bypass are firmly established, alternative graft materials occasionally are necessary for limb salvage in patients whose ipsilateral saphenous systems either are inadequate because of their small size or have already been harvested for previous lower extremity or coronary artery revascularization. The contralateral greater saphenous vein can be used as a reversed graft in this setting, but even if it still is available, its use may be contraindicated in patients who predictably will require bypass below the knee in the opposite leg within the foreseeable future. In addition to the ipsilateral lesser saphenous vein, therefore, arm veins may represent the only option for autogenous graft material in selected patients for whom distal popliteal or tibioperoneal revascularization is unavoidable. In 1969 Kakkar 2 demonstrated in cadavers that the tensile strength of the cephalic vein appeared to be adequate for the purpose of arterial replacement. A number of clinical series subsequently have been reported in which cephalic vein grafts were used, often with only modest success, for femoropopliteal or distal revascnlarization.3-~° The largest of these was described by Andros et al.9 in which 88 "full-length" grafts were constructed to outflow vessels below the knee with the cephalic (77/88) and/or the basilic 543
544
Sestoet aL
veins. The 3-year primary patency rate for grafts performed for limb salvage in this series was approximately 63%, a promising figure in comparison to the discouraging intermediate-term results of prosthetic grafts under similar circumstances. The purpose of our report is to assess the outcome of all femoropopliteal or femorotibial-peroneal cephalic vein grafts performed at the Cleveland Clinic during the past decade.
PATIENTS AND METHODS Patient information From 1980 through 1989, 926 distal popliteal or tibioperoneal bypass grafts were constructed at this center. Thirty-five (3.8%) of these required the use of the cephalic vein(s) alone (N = 23; 65%) or in conjunction with composite segments of saphenous vein (N = 9; 26%) or PTFE (N = 3; 9%). This cohort included 26 men (76%) and eight women (24%) with a mean age of 61 years (range, 33 to 81 years). Seventeen (50%) of these patients were hypertensive, and 10 (29%) had diabetes, eight of whom (23%) received insulin. Disabling claudication was the indication for only six cephalic vein grafts (17%). Most of these grafts (N = 25; 71%) were necessary because of ischemic rest pain or focal gangrene, and the remaining four patients (12%) had either popliteal aneurysms or infected femoropopliteal prosthetic grafts. Thirty-one of the 35 procedures in this series have been performed since 1985, and 20 of those during the last 2 years (1988 to 1989) of the study period. Since our preferential adoption of in situ techniques several years ago, the diameter of the greater saphenous vein has become relatively less important in determining its adequacy as an arterial substitute. In our experience, surgical absence of the saphenous vein has been the most common indication for the use of cephalic vein grafts for lower extremity revascularization. The ipsilateral saphenous vein already had been harvested for previous coronary (N = 16; 47%) or femoropopliteal-tibioperoneal bypass (N = 17; 50%) in many of the 35 patients described in this report, and another 11 patients (31%) had undergone autogenous grafts in the contralateral limb.
Surgical management On the basis of preoperative angiograms, tibioperoneal outflow was represented by three vessels in eight limbs (23%), by two vessels in five (14%), and by a single vessel in 22 (63%). The common femoral artery or a segment of the superficial or deep femoral artery immediately adjacent to it was used as the
Journal of VASCULAR SURGERY
inflow source for all 35 cephalic vein grafts in this series, and the targets selected for the distal anastomosis are summarized in Table I. Although the popliteal artery below the knee was appropriate for the distal anastomosis in 10 limbs (29%), direct tibioperoneal revascularization was necessary in the remaining 25 (71%). Sequential bypass to multiple vessels was not used, but the cephalic vein segment was used to cross the knee joint in each of the three patients who required composite synthetic grafts. An operative angiogram was routinely obtained to determine the immediate technical result, and pulse volume recordings (segmental limb pressure measurements with analog waveform analysis) were performed during the same hospitalization after 33 (90%) of the 35 procedures. Approximately 20% of our patients received intravenous heparin (N = 9) or low molecular-weight dextran (N = 7) during the perioperative period, and six (18%) randomly were placed on long-term oral anticoagulation with sodium warfarin (Coumadin). Preliminary arteriovenous fistulas In an attempt to enhance the diameter and the handling characteristics of the cephalic vein, temporary arteriovenous fistulas were constructed in the arm(s) of 23 patients (68%) whose arm veins appeared to be of marginal size on the basis of simple physical examination. After the venous anatomy of the upper extremities had been assessed by operative phlebography, an end-to-side arteriovenous anastomosis was performed at one or more appropriate levels by use of either the radial artery at the wrist or the brachial artery just proximal to the antecubital crease (Fig. 1). These fistulas were permitted to mature for approximately 10 to 14 days before the cephalic veins were harvested for lower extremity grafting. Statistical analysis The follow-up interval for the 34 patients (35 limbs) in this study ranges from 1 to 107 months (mean, 28 months; median, 27 months). Late results concerning patient survival and limb salvage were established during a telephone canvass, but conclusions regarding graft patency are based strictly on the physical findings and objective data from the noninvasive vascular laboratory, which were obtained at the time of the most recent outpatient examination. Elective revisions of patent grafts were considered equivalent to graft occlusions in the determination of secondary patency rates. Paired comparisons were evaluated by use of Fisher's (two-tailed) exact test. Life-table informa-
Volume 15 Number 3 March 1992
Cephalic vein grafCs 545
Fig. 1. A, Preoperative phlebogram demonstrates a diminutive cephalic vein (arrow) in the right forearm; B, Another contrast study shows enlargement of the same vein (arrow) 2 weeks after an arteriovenous fistula had been constructed at the wrist; C, The harvested vein (arrow) after revascularization of the peroneal artery. The proximal segment of this cephalic vein graft was obtained from the contralateral arm. Table I. A summary o f the surgical management used in this series Graft material Cephalic vein Composite cephalic and saphenous veins Composite cephalic vein and prosthetic Total
Below-knee popliteal No. % 8 2 0 10
tion was calculated according to the method described by Cutler and Ederer, 11 and standard errors o f the mean were assessed with the Greenwood formula.
RESULTS N o postoperative deaths occurred in this series. Early thrombosis occurred in two grafts (6%), both o f which were successfully revised. Another imme-
23 6 29
Tibial No.
%
9 2 1 12
25 6 3 34
Peroneal No. % 6 5 2 13
17 14 6 37
Total No.
%
23 9 3 35
65 26 9 100
diate reoperation was necessary to control bleeding from a cephalic vein branch. The mean Doppler ankle/brachial index for all 35 limbs improved from 0.38 before operation to 0.91 after revascularization. Cumulative data concerning late survival, graft patency, and limb salvage are presented in Fig. 2 and in Table II. Four patients have died in their first, second, third, or eighth postoperative year, resulting in an actuarial 3-year survival rate o f 87% + 7%. In
546
Sesto et aL
1.000 .969 • -------(34) ° ~ (34)
100 -'J
~ rrv
Journal of VASCULAR SURGERY
o
~
.931 8 0 (30)
6
80
Table II. Cumulative 3-year results (and standard errors) after cephalic vein bypass
9 (20)
60
Cumulative res#lts Clinical features
1 yr %
2 yr %
3 yr %
97
-+3)
93 (-+5)
87 (-+7)
49 -+9) 68 --8) 82 -+7)
49 (-+9) 58 (-+10) 82 (-+7)
40 (_+11) 44 (-+11) 82 (+7)
(./3 4O
Survival
I >_ O Z UJ a_o~ LL < rr L9
.971
100
~(~5 80
\\
•
I
(35)
-q
40
Patency Primary Secondary
.578 (16) 0
(11o)
Limb salvage .442
.397
(7) 1
1.000
80 60
.675 o (34)
(31)
I
I
o SECONDARY
X ' """''~'~'~
40
O0~ rn
I • PRIMARY
o• , N (35)
60
100 LU (.9 < >
I
824
.824
.824
(35)
(27)
(17)
I
I
I
I
30 DAYS
1
2
3
POSTOPERATIVE YEARS
Fig. 2. Cumulative 3-year survival, patency, and limb salvage rates for this series of 34 patients (35 limbs).
addition to the two postoperative failures, another 14 cephalic vein grafts (40%) have either occluded (N = 11; 31%) for unknown reasons or have been electively revised (N = 3; 9%) to correct incidental lesions that were discovered during follow-up surveillance (Fig. 3). The 3-year primary patency rate currently is 40% _ 11%, the secondary patency rate is 4 4 % _ 11%, and the limb salvage rate is 82% - 7%. It is impractical to evaluate these cumulative results statistically because our series contains too few patients, but selected crude comparisons are summarized in Table III. Variance in the mean follow-up periods reflects the accelerated attrition of occluded grafts within unfavorable groups. These data generally support the importance of several factors (preoperative symptoms, adequacy of outflow, prosthetic requirements) that traditionally have been found to influence the outcome of lower extremity bypass below the knee. Nevertheless, only the difference between failure rates for grafts indicated for limb salvage (52%) and those performed for claudication
(10%) attained statistical significance (p = 0.028; a l l other values p _> 0.2). DISCUSSION
Countless reports have documented the consistently superior performance of saphenous vein grafts for below-knee femoropopliteal or femorotibialperoneal bypass in comparison to a number of prosthetic materials that have been introduced during the past 30 years. The merit of an "all autogenous" policy for these procedures has been assumed to include the use of arm veins under circumstances in which the saphenous system is inadequate or no longer available. Several sources, however, suggest that it may be unrealistic to expect the results of cephalic vein bypass to meet the high standard established by the saphenous vein for lower extremity revascularization. Although the seven cephalic-basilic vein grafts initially performed by Kakkar2 remained patent during the preparation of his original article, their follow-up periods ranged only from 4 to 12 months. Thirteen (81%) of 16 similar grafts in two other early series subsequently failed within 24 months of their construction,4 and cumulative 3- to 5-year patency rates of approximately 30% have since been reported by Whittemore et al.5 and by Schulman and Badhey. 7 Improved patency rates have been described by Graham and Lusby6 (58% at 3 years) and, in probably the most extensive experience with "fulllength" cephalic vein grafts that has yet been pubfished, by Harris et al.8 and Andros et al.9 (63% at 3 years, and 51% at 5 years, respectively). Substantial differences in patient selection (including the number of previous operations in the same limb) and in the quality of arm veins deemed to be acceptable for bypass purposes could easily exist among these series, but it still seems reasonable to conclude that the late
Volume 15 Number 3 March 1992
Cephalic vein grafis 547
Fig. 3. Two examples of dilation and thrombus formation occurring 4 months (A) and 8 months (B) after lower extremity bypass with cephalic vein grafts. results of cephalic vein grafts for lower extremity revascularization have been somewhat inconsistent. Despite this uncertainty, there occasionally is no preferable alternative to the cephalic vein as a bypass conduit in certain patients for whom arterial reconStruction below the knee is absolutely necessary. The contralateral saphenous vein offers obvious advantages as a reversed graft and should be the primary consideration under optimal circumstances. Nevertheless, if the saphenous system was inadequate in size in the ischemic limb, it often is likely to be so on the opposite side as well. Furthermore, the contralateral vein may already have been used during coronary artery surgery or, as was the case in eight (23%) of our patients, for previous femoropopliteal bypass. Finally, many patients also have occlusive arterial disease in the contralateral lower extremity, a feature that might also serve as a future indication for distal revascularization. If the contralateral sapahenous vein is gambled and lost in an unsuccessful bid to salvage
the original jeopardized limb, the ultimate outcome conceivably could be just as unfavorable on the opposite side. The durability of prosthetic grafts extending below the knee generally has been disappointing, especially when the distal anastomosis involves the peroneal artery or a tibial vessel. Although the use of synthetic materials sometimes is unavoidable, our experience suggests that the results of cephalic vein bypass are more favorable than those for prosthetic grafts even though they do not meet the expectations traditionally reserved for saphenous vein grafts. In Table IV, the 3-year cumulative patency rates for the 35 cephalic vein grafts in the present series are compared with those for a group of saphenous vein and PTFE grafts obtained from the computer registry sponsored by the Cleveland Vascular Society. 12 The primary patency rate (40%) for our cephalic vein grafts (25, or 71% of which were constructed to tibioperoneal targets) clearly did not attain the same
548
Journal of VASCULAR SURGERY
Sesto et al.
Table III. Follow-up results for this series of cephalic vein grafts Follow-up (too) Paired comparisons Surgical indication Claudication Severe ischemia Distal anastomosis Below-knee popliteal Tibial or peroneal Graft material Cephalic vein only Composite saphenous vein or prosthetic Preliminary AV fistula Constructed Not constructed Outflow vessels 2-3 1 Coumadin therapy Administered Not administered
Table IV. Comparison of 3-year patency rates for the 35 cephalic vein grafts in this series and for a total of 388 l~mb salvage procedures reported from the Cleveland Vascular Registry~1 3-Year cumulative patency Graft material Present series Cleveland Vascular Registry Saphenous vein Below-knee popliteal Tibioperoneal PTFE Below-knee popliteal Tibioperoneal
Primary (%)
Secondary (%)
35
40
44
120 98
69 43
69 43
101 69
30 18
32 19
No.
overall success as saphenous vein grafts. Perhaps more importantly, it seems to represent reasonable improvement in comparison to the patency rate of below-knee PTFE bypass to either the popliteal or the tibioperoneal level (30% and 18%, respectively). In 1971 Beals 13 first described the use of an adjunctive arteriovenous fistula to mature the cephalic vein before its deployment as an arterial replacement graft. We rediscovered this concept years later because of our experience with hemodialysis angioaccess, and preliminary fistulas were performed in 23 (68%) of the 34 patients in this series. Although a staging interval of 10 to 14 days was sufficient to enhance the diameter of small cephalic veins, the merit of this approach remains largely a matter of speculation. Late failure rates were similar among
Graft occlusions
No.
Range
Mean
No.
%
10 25
1-42 1-40
21 7
1 13
10 52
10 25
2-42 1-40
19 8
2 12
20 52
24 11
1-42 1-41
12 I0
8 6
33 55
23 12
1-42 1-41
13 8
8 6
35 5O
13 22
1-42 1-40
15 9
4 10
31 45
6 29
2-26 1-42
8 12
4 10
67 34
patients who received preliminary fistulas (35%) in comparison with those who did not (50%), and although arteriovenous fistulas theoretically could have contributed to the dilated appearance that subsequently was documented in a few of our grafts (Fig. 3), Andros et al.9 also observed graft ectasia in 17 (30%) of the 56 patients in their series who eventually underwent follow-up angiography. The cephalic vein often is abused in the forearm by frequent venipuncmres and intravenous infusions, but it usually is available in the upper arm where it may be evaluated before operation by phlebography or Doppler ultrasound (duplex) scanning) 4 It is an important resource for femoropopliteal or femorotibial-peroneal bypass in a select group of patients for whom disadvantaged prosthetic grafts otherwise would be required. Under these demanding circumstances, cephalic vein grafts have a specific place in the armamentarium of vascular surgeons. Ming Tan, PhD (Department of Biostatistics and Epidemiology) and Geri Locker (Research Division) performed the statistical analysis. REFERENCES
1. Dalman RL, Taylor LM. Basic data related to infrainguinal revascularization procedures. Ann Vasc Surg 1990;4:309-12. 2. Kakkar VV. The cephalic vein as a peripheral vascular graft. Surg Gynecol Obstet 1969;128:551-6. 3. Vellar ID, Doyle IC. The use of the cephalic and basilic veins as peripheral vascular grafts. Aust N Z J Surg 1970;40: 52-7. 4. Clayson KR, Edwards WH, Allen TK, Dale A. Arm veins for peripheral arterial reconstruction. Arch Surg 1976; 111:127680.
Volume 15 Number 3 March 1992
Cephalic vein grafls 549
5. Whittemore AD, Clowes AW, Couch NP, Mannick JA. Secondary femoropopliteal reconstruction. Arch Surg 1981; 193:35-42. 6. Graham JW, Lusby RJ. Infrapopliteat bypass grafting: use of upper limb vein alone and in autogenous composite grafts. Surgery 1982;91:646-9. 7. Schuhxlan ML, Badhey MR. Late results and angiographic evaluation of arm veins as long bypass grafts. Surgery 1982;92:1032-41. 8. Harris RW, Andros G, Dulawa LB, Oblath RW, Salles-Cunha SX, Apyan R. Successful long-term limb salvage using cephalic vein bypass grafts. Ann Surg 1984;200:785-92. 9. Andros G, Harris RW, Salles-Cunha SX, Dulawa LB, Oblath RW, Apyan RL. Arm veins for arterial revascularization of the leg: arteriographic and clinical observations. J VAsc SURG 1986;4:416-27. 10. Balshi JD, Cantelmo NL, Menzoian JO, LoGerfo FW. The
11. 12. 13. 14.
use of arm veins for infrainguinal bypass in end-stage peripheral vascular disease. Arch Surg 1989;124:1078-81. Cutler SJ, Ederer F. Maximum utilization of the life-table method in analyzing survival. J Chron Dis 1958;8:699-712. Rafferty TD, Avellone JC, Farrell CJ, et at. A metropolitan experience with infralnguinal revascularization. J VASCSFRG 1987;6:365-71. Beats RL. Surgically created arteriovenous fistula to augment the cephalic vein: use as an arterial bypass graft. N Engl J Med 1971;285:29-30. Salles-Cunha SX, Andros G, Harris RW, Dulawa LB, Oblath RW. Preoperative noninvasive assessment of arm veins to be used as bypass grafts in the lower extremities. J VASCSUI
