From the Society for Clinical Vascular Surgery

Brachial versus basilic vein dialysis fistulas: Acomparison of maturation and patency ratesKevin Casey, MD, Britt H. Tonnessen, MD, Krishna Mannava, MD, Robert Noll, MD,Samuel R. Money, MD, and W. Charles Sternbergh III, MD, New Orleans, La

Objectives: Although the performance of basilic vein transpositions for dialysis access is well established, the utility andpatency rates of brachial vein transpositions are poorly characterized. The brachial vein is being used increasingly as analternative vein for transposition in an effort to increase the percentage of autogenous fistula utilization. The purpose ofthis study was to review a single-center comparative experience with these fistulas.Methods: A retrospective chart review was performed on 59 patients who received basilic and brachial vein transpositionsbetween January 2000 and December 2006. Patient demographics, comorbidities, mortality, and morbidity wereevaluated. Patency rates were calculated using Kaplan-Meier life-table analysis.Results: Of 59 vein transpositions, there were 42 basilic (71%) and 17 brachial (29%). The 30-day mortality was 0%.Maturation rates were 74% for basilic vein transpositions and 47% for brachial (P � .049). The mean time to maturationwas 11.9 � 8.8 weeks. Primary patency rates at 12 months were 50% for basilic vein transpositions vs 40% for brachial(P � .115). The mean vein size was 4.9 � 0.9 mm. The mean basilic vein transposition diameter of 4.9 � 1.0 mm andbrachial vein transposition diameter of 5.0 � 0.8 mm were not significant (P � .39).Conclusions: Despite a higher rate of initial maturation in basilic vein transpositions, brachial and basilic vein transposi-tions had comparable patency rates at 12 months. These preliminary results require further follow-up and a larger cohortof patients for confirmation. Broader use of the brachial vein transposition for dialysis appears justified and can increase

the overall percentage of autogenous fistula placement. ( J Vasc Surg 2008;47:402-6.)

Dialysis access is a continual challenge for vascular sur-geons and their patients. Maintaining adequate access usuallyrequires more than one procedure in life-long hemodialysispatients. In 1966, Brescia and Cimino1 described the firstautogenous fistula. Which autogenous fistula is the best is thesubject of ongoing discussion and debate, and there is oftendisagreement about which secondary and tertiary procedure isthe best after a failed initial fistula.

The United States has historically lagged behind Eu-rope with respect to the prevalence of autogenous fistulasused for dialysis, with a 24% vs 80% usage, respectively, inthe Dialysis Outcomes and Practice Patterns Study(DOPPS).2 In 1997, the National Kidney Foundation–Dialysis Quality Initiative (NKF-DOQI) proposed guide-lines supporting increased usage of native veins for dialysisaccess.3 These guidelines have since been revised, withgreater emphasis on strategies to increase autogenous fis-tulas. As a result, surgeons have developed innovative tech-niques and methods for improving autogenous access. Infact, the drive for autogenous fistulas has led to a significantincrease in their usage to 46% nationwide in 2007.4

The basilic vein has long been viewed as an acceptableconduit for an autogenous fistula. The basilic vein and brachialartery anastomosis, performed by transposing the vein into a

From the Ochsner Clinic Foundation.Competition of interest: none.Presented at the Thirty-fifth Annual Meeting of the Society for Clinical

Vascular Surgery, Orlando, Fla, Mar 21-24, 2007.Reprint requests: W. Charles Sternbergh III, MD, Department of Vascular

Surgery, Ochsner Clinic Foundation, 1514 Jefferson Highway, NewOrleans, LA 70121 (e-mail: csternbergh@ochsner.org).

0741-5214/$34.00Copyright © 2008 by The Society for Vascular Surgery.

doi:10.1016/j.jvs.2007.10.029

402

subcutaneous tunnel, was first described for dialysis access byDagher et al5 in 1976. Many studies have since evaluated theefficacy of the basilic vein for long-term dialysis access. 6-16

Proponents of the basilic vein transposition (BVT) maintainthat it is a suitable site for access because, like all autogenousfistulas, it has a low incidence of infection, keeps the body freefrom foreign material, and has longer patency rates thanpolytetrafluoroethylene (PTFE).9,11 An additional advantageis that the surgeon does not “burn any bridges” by attemptinga BVT; if it fails, then a subsequent ipsilateral graft can beplaced. The initial use of an upper extremity graft may pre-clude the creation of a BVT afterwards.

The brachial vein, however, has rarely been studied as aconduit for an autogenous fistula.16,17 Although it seemsintuitive that a brachial vein transposition (BrVT) wouldpossess the same advantages as a BVT, this technique mayhave additional limitations. Concerns about vein diameterand length available for mobilization, numerous branches,and greater depth from the skin may discourage surgeonsfrom using this vein for a transposition. The purpose of thisstudy was to review a single institution’s short-term expe-rience with BVT and BrVT.

MATERIALS AND METHODS

We used our institutional operative record/database toidentify 59 consecutive patients who underwent BVT orBrVT at the Ochsner Clinic Foundation between January2000 and December 2006. The study design and protocolwere approved by the Institutional Review Board. Retro-spectively reviewed data were obtained from inpatientcharts, outpatient records, operating room notes, dialysis

records, and phone calls. Patient characteristics collected

JOURNAL OF VASCULAR SURGERYVolume 47, Number 2 Casey et al 403

for the purpose of this study were age, sex, race, and specificcomorbidities. The preoperative vein diameter (as deter-mined by vein mapping), number of prior access procedures,and current hemodialysis status were also evaluated. The post-operative rates for 30-day mortality, complications, matura-tion, primary functional patency, and primary-assisted pa-tency were studied. Maturation was defined as the timeuntil the primary fistula was suitable to allow successfulcannulation. Primary functional patency was defined as afistula that remained patent throughout follow-up and wasused for hemodialysis. Primary-assisted patency was definedas maintained fistula flow as a result of an adjuvant proce-dure or intervention.18

Patient selection. Patients were eligible for a transpo-sition arteriovenous fistula if the veins were a minimum of4.0 mm. When the basilic and brachial veins were compa-rable in size, the basilic vein was selected. Primary radioce-phalic or brachiocephalic fistulas were the first choice if acephalic vein �3.0 mm was present.

Surgical technique. Doppler studies have beenshown to be a reliable and effective evaluation of upperextremity veins.19-22 At our institution, every patient re-ceives preoperative vein mapping in our vascular lab. If nosuitable cephalic vein is identifiable for anastomosis withthe radial or brachial artery, the basilic and brachial veins areevaluated. Either of these veins will be selected preopera-tively if the diameter �4 mm. If the diameters are compa-rable, the basilic vein is selected because of greater ease ofmobilization and the presence of fewer branches. Veindiameter is measured at the mid-humerus and antecubitalregion for both basilic and brachial veins.

Our surgical technique, which is similar to previous de-scriptions of vein transpositions in the literature,5,7,23 is re-viewed here (Fig 1). The preferred method of anesthesia iseither general or interscalene block. It is possible to performthe procedure with conscious sedation and local anesthesia ina compliant patient. The patient’s upper extremity is circum-ferentially prepared to the axilla. A longitudinal incision isbegun superior and medial to the medial epicondyle of thehumerus. This can be placed directly over the brachial arterypulse for a BrVT or more medially if a BVT is planned.

The vein of choice is identified; if vein size is marginal, onecan look for a more suitable vein through this incision. Themedian nerve and other cutaneous nerves are carefully identi-fied and spared. The incision is then extended several centi-meters at a time while sequentially freeing up the anteriorsurface of the vein. The incision will usually extend all the wayto the axilla. Small side branches are ligated with silk ties andclips; larger connections are oversewn with polypropylenemonofilament. Additional length can be gained by mobilizingthe vein below the antecubital crease if necessary.

The vein is then ligated distally, distended with hepa-rinized saline with the anterior surface marked, and a Sera-fin clamp is placed (Fig 2, A). A penetrating towel clamp isused to temporarily approximate the skin edges, and thevein is draped over the skin in the area of the plannedtunnel, which is then marked on the skin (Fig 2, B). A

tunneling device is brought subcutaneously, along the lat-

eral aspect of the arm, distal to proximal. The vein iscarefully oriented and brought through the tunnel on agentle curve, reaching the brachial artery without tension.

The patient is systemically heparinized and the artery isthen mobilized for several centimeters. A standard end-to-side vascular anastomosis is performed. The quality of thethrill is then assessed. A pulsatile fistula may imply a kink ortwist in the vein, constriction from the tunnel, or outflowobstruction. Radial and ulnar pulses are checked, and hep-arin is reversed with protamine. A two-layer closure isperformed with running absorbable suture.

Most patients are discharged home the day of surgeryor the next day. All procedures were performed by one ofthree attending physicians (W. C. S., B. H. T., S. R. M.)with a resident or fellow (Fig 3).

Statistical analysis. The data are expressed as mean �standard deviation. Patient characteristics were comparedusing the �2 test. Primary and primary-assisted patencyraters were evaluated using Kaplan-Meier plots and com-pared using log-rank analysis. A value of P � .05 wasconsidered statistically significant. Statistical analysis wasperformed using SAS 8.2 software (SAS Institute Inc, Cary,NC).

RESULTS

The study comprised 38 men (64%) and 21 women

Fig 1. Anatomy of the upper arm during a basilic vein transposi-tion. Alternatively, the brachial vein could be used as the conduitand anastomosed to the brachial artery.

(36%), and the median age was 55 years (range, 21-82

JOURNAL OF VASCULAR SURGERYFebruary 2008404 Casey et al

years). These 59 patients underwent 59 vein transpositions,of which 42 (71%) were BVTs, and 17 (29%) were BrVTs.In 19 patients (32%), the transposition was the initial accessprocedure (range, 0-3 prior procedures). The mean veinsize was 4.9 � 0.9 mm. The mean BVT diameter of 4.9 �1.0 mm and BrVT diameter of 5.0 � 0.8 mm were not

Fig 2. A, Dissection of the upper arm during a vein transposition.B, The same patient, with the distal vein ligated just before thesubcutaneous tunnel.

Fig 3. A successful matured vein transposition seen in clinic.

significantly different (P � .39).

The median follow-up in our study was 8 months(range, 0-64 months). The fistulas matured in 39 patients(66%) and failed to mature in 12 patients. Three patientswere lost to follow-up before maturation could be assessed.Five patients, two with BVT and three with BrVT, hadfistulas that were still awaiting maturation at 0 to 14 weeksafter the procedure. The maturation rate was 74% for BVT(n � 31) and 47% for BrVT (n � 8). The differencebetween the two groups was statistically significant (P �.049). The mean time to maturation was 11.7 � 8.8 weeksand was not significantly different between the two groups.

In 78% of patients (18 of 23) whose fistulas matured, veindiameters were larger than the mean (4.9 mm) compared with50% (12 of 24) whose diameters were smaller than the mean.This was statistically significant (P � .043). The preoperativevein diameter was not found in 12 patients.

Early primary functional patency rate at 6 months was71% in BVTs and 40% in BrVTs. The 12-month functionalpatency rate was 50% in BVTs and 40% in BrVTs (Fig 4),which was not statistically significant (P � .115). Primary-assisted patency rates (Fig 5) were virtually unchanged andwere also not significant (P � .154).

No preoperative characteristic, including age, sex, race,vein size, number of previous accesses, presence of diabetesmellitus, hypertension, hyperlipidemia, or smoking wasstatistically different between the two groups (Fig 6), andnone of these factors affected maturation rates or patencyrates. The 30-day mortality rate was 0%, although onepatient was lost to follow-up immediately after the proce-dure. Morbidity was 13%. Early complications included awound infection in two patients, hematoma in two patients(one which required evacuation in the operating room),and three cases of vascular steal syndrome, each requiringsubsequent distal revascularization and interval ligationprocedures. Each of these patients required a brachialartery–to–brachial artery bypass. Three different conduitswere used: reversed saphenous vein, contralateral forearm

Fig 4. Primary functional patency. BVT, Basilic vein transposi-tion; BrVT, brachial vein transposition.

vein, and a 6-mm PTFE graft. Finally, one patient pre-

JOURNAL OF VASCULAR SURGERYVolume 47, Number 2 Casey et al 405

sented after surgery with symptomatic ipsilateral edemathat resolved without intervention.

Ten patients required 13 additional procedures, all ofwhich were percutaneous venous angioplasties. Six percutane-ous interventions were performed in BVTs and seven inBrVTs. Five patients had a successful result from the percuta-neous intervention (3 BVTs and 2 BrVTs). However, fiverequired a subsequent operation to achieve dialysis access.

DISCUSSION

Radiocephalic or brachiocephalic arteriovenous fistulasremain the initial dialysis access procedure of choice; how-ever, very often patients do not have veins that are suitablefor creation of a fistula that will mature. Moreover, patientswho require long-term access over many years will un-doubtedly require one or more procedures. Opinions differabout which type of fistula is ideal for subsequent access.

Fig 5. Primary assisted patency. BVT, Basilic vein transposition;BrVT, brachial vein transposition.

0

20

40

60

80

100

%

Diabetes HTN HPL Smoked

Total BVT BrVT

Fig 6. Patient characteristics. HTN, Hypertension; HPL, hyper-lipidemia.

The basilic vein has been recognized as a suitable vein for

creating a native transposed fistula to the brachial artery.Less attention has been given to the brachial vein as aconduit for dialysis access.

At our institution, we will almost exclusively attempt anautogenous fistula in a patient with suitable veins. Eachpatient undergoes preoperative vein mapping in our vascu-lar lab. If a suitable cephalic vein is not identified at the levelof the radial or brachial arteries, we then evaluate thebrachial and basilic veins. For many patients (68% in ourcohort), an upper extremity transposition is not the initialaccess procedure. In appropriately selected patients, how-ever, an upper extremity transposition may be a good firstaccess procedure, particularly if the forearm veins are small.

We will choose either the brachial or basilic vein fortransposition according to the better vein diameter. If veindiameters are comparable, we select the basilic vein owingto the greater ease of mobilization and fewer branches. Ithas also now become our practice to take two measure-ments of these veins, at both the mid-humerus and at theantecubital region. Veins that are too small (�4 mm) ateither region are usually excluded from usage.

Patients with large upper arm girth may not be suitablecandidates for a transposition because the longer tunnelrequires additional length of vein. Patients with prior upperarm grafts or transpositions are poor candidates for a sub-sequent BVT or BrVT because of tunneling issues and thetheoretic risk of increased arm edema. In patients with priorupper arm vein transpositions, we do not perform addi-tional ipsilateral transpositions; however, an arteriovenousgraft can be subsequently placed if needed.

Our experience with BrVT is, to our knowledge, thelargest reported review in the literature. We found that thematuration rate was significantly better for BVTs (74%)than BrVTs (47%); however, at 6 and 12 months, thepatency rates were not significantly different. These earlyresults are comparable with other centers.6-16 These dataindicate that most of our matured upper extremity trans-positions remained functional for the duration of this study.Despite these modest results, we believe that in carefullyselected patients, an upper extremity transposition is a goodautogenous access procedure. Another potential advantageof vein transposition is that this procedure rarely precludesa subsequent ipsilateral upper arm graft.

Our patency rates are somewhat less substantial com-pared with recent and historical studies of PTFEgrafts.9,11,24-26 Cumulative patency rates in these retrospec-tive studies were 46% to �75% at 1 year and beyond;however, these same studies demonstrated a greater needfor secondary procedures to maintain graft patency com-pared with our results. In addition, their patient popula-tions tended to have a greater number of complications,including graft infections, development of pseudoaneu-rysms, and venous outflow stenoses.

The NKF-DOQI initiatives have attempted to promoteincreased use of autogenous veins for fistulas. Autogenousconduits have greater durability, increased patency, and re-quire fewer procedures to maintain patency. Studies have also

shown that prosthetic grafts are more likely to become in-

JOURNAL OF VASCULAR SURGERYFebruary 2008406 Casey et al

fected, require more complex revisions, and are associatedwith a greater number of hospitalizations and increased cost.3

The addition of BrVT in our algorithm for hemodialy-sis access was clearly related to the NFK-DOQI “push” forgreater autogenous accesses. A more difficult question toanswer is whether 1-year patency rates of 40% to 50% oftransposition arteriovenous fistulas are competitive enoughwith results of prosthetic bridge grafts to routinely recom-mend their preferential use. Because BrVT was limited to17 patients, the role of this particular fistula in the hemo-dialysis access algorithm is still uncertain. Longer follow-upwith a more robust sample size may further clarify the roleof BVT and BrVT.

CONCLUSION

Early results from this study demonstrate that BVT hasstatistically greater maturation rates compared with BrVT;however, BVT and BrVT are comparable with respect to earlyprimary functional patency and primary assisted patency rates.Larger studies comparing the two fistula types with prostheticgrafts are needed to evaluate their relative efficacy in thedialysis population. Given the benefits of autogenous fistulas,further investigations of this type will be beneficial.

AUTHOR CONTRIBUTIONS

Conception and design: KC, CS, SM, BTAnalysis and interpretation: KC, BTData collection: KCWriting the article: KC, BTCritical revision of the article: KC, BT, CS, BT, RNFinal approval of the article: KC, BT, CS, KM, RN, SMStatistical analysis: KCObtained funding: Not applicableOverall responsibility: KC

REFERENCES

1. Brescia MJ, Cimino JE, Appel K, Hurwich BJ. Chronic hemodialysisusing venipuncture and a surgically created arteriovenous fistula. N EnglJ Med 1966;275:1089-92.

2. Pisoni RL, Young EW, Dykstra DM, Greenwood RN, Hecking E,Gillespie RL, et al. Vascular access use in Europe and the United States:results from the DOPPS. Kidney Int 2002;61:305-16.

3. The Vascular Access Work Group. NKF-DOQI clinical practice guide-lines for vascular access. National Kidney Foundation—Dialysis Out-comes Quality Initiative. Am J Kidney Dis 1997;30(suppl 3):S150-91.

4. Arteriovenous Fistula First coalition. Fistula First National VascularAccess Improvement Initiative. Available at: www.fistulafirst.org. Ac-cessed May 1, 2007.

5. Dagher F, Gelber R, Ramos E, Sadler J. The use of basilic vein andbrachial artery as an A-V fistula for long term hemodialysis. J Sur Res1976;20:373-6.

6. LoGerfo FW, Menzoian JO, Kumaki DJ, Idelson BA. Transposedbasilic vein-brachial arteriovenous fistula: a reliable secondary access

procedure. Arch Surg 1978;113:1008-10.

7. Hibberd AD. Brachiobasilic fistula with autogenous basilic vein: surgi-cal technique and pilot study. Aust N Z J Surg 1991;61:631-5.

8. Rivers SP, Scher LA, Sheehan E, Lynn R, Veith FJ. Basilic vein trans-position: an underused autologous alternative to prosthetic dialysisangioaccess. J Vasc Surg 1993;18:391-6.

9. Coburn CC, Carney WI. Comparison of basilic vein and polytetrafluo-roethylene for brachial arteriovenous fistula. J Vasc Surg 1994;20:896-904.

10. Butterworth PC, Doughman TM, Wheatley TJ, Nicholson ML. Arte-riovenous fistula using transposed basilic vein. Br J Surg 1998;85:653-4.

11. Matsuura JH, Rosenthal D, Clark M, Frederick WS, Kirby L, ShotwellM, et al. Transposed basilic vein versus polytetrafluoroethylene forbrachial-axillary arteriovenous fistulas. Am J Surg 1998;176:219-21.

12. Murphy GJ, White SA, Knight AJ, Doughman T, Nicholson ML.Long-term results of arteriovenous fistulas using transposed autologousbasilic vein. Br J Surg 2000;87:819-23.

13. Taghizadeh A, Dasgupta P, Khan MS, Taylor J, Koffman G. Long-termoutcomes of brachiobasilic transposition fistula for haemodialysis. Eur JVasc Endovasc Surg 2003;26:670-2.

14. Rao RK, Azin GD, Hood DB, Rowe VL, Kohl RD, Katz SG, et al.Basilic vein transposition fistula: a good option for maintaining hemo-dialysis access site options? J Vasc Surg 2004;39:1043-7.

15. Wolford HY, Hsu J, Rhodes JM, Shortell CK, Davies MG. Outcomeafter autogenous brachial-basilic upper arm transpositions in the post-National Kidney Foundation Dialysis Outcomes Quality Initiative era.J Vasc Surg 2005;42:951-6.

16. El Sayed JF, Mendoza B, Meier GH, LeSar CJ, DeMasi RJ, GlickmanMH, et al. Utility of basilic vein transposition for dialysis access.Vascular 2005;13:268-74.

17. Bazan HA, Schanzer H. Transposition of the brachial vein: a new sourcefor autologous arteriovenous fistulas. J Vasc Surg 2004;40:184-6.

18. Sidawy AN, Gray R, Besarab A, Henry M, Ascher E, Silva M, et al.Recommended standards for reports dealing with arteriovenous hemo-dialysis accesses. J Vasc Surg 2002;35:603-10.

19. Nonnast-Daniel B, Martin RP, Lindert O, Mugge A, Schaeffer J, v dLieth H, et al. Colour Doppler ultrasound assessment of arteriovenoushaemodialysis fistulas. Lancet 1992;339:143-5.

20. Rutherford RB. The value of noninvasive testing before and afterhemodialysis access in the prevention and management of complica-tions. Sem Vasc Surg 1997;10:157-61.

21. Tardoir JH, deBruin HG, Hoeneveld H, Eikelboom BC, Kitslaar P.Duplex ultrasound scanning in the assessment of arteriovenous fistulascreated for hemodialysis access: comparison with digital subtractionangiography. J Vasc Surg 1989;10:122-8.

22. Silva MB, Hobson RW, Pappas PJ, Jamil Z, Araki CT, Goldberg MC, etal. A strategy for increasing use of autogenous hemodialysis accessprocedures: impact of preoperative noninvasive evaluation. J Vasc Surg1998;27:302-8.

23. Stonebridge PA, Edington D, Jenkins AM. ‘Brachial/basilic vein’ trans-position for vascular access. J R Coll Surg Edinb 1995;40:219-20.

24. Jenkins AM, Buist TA, Glover SP. Medium-term follow-up of fortyautogenous vein and forty polytetrafluoroethylene (Gore Tex) grafts forvascular access. Surgery 1980;88:667-72.

25. Kherlakian GM, Roedersheimer LR, Arbaugh JJ, Newman KJ, KingLR. Comparison of autogenous fistula versus expanded polytetrafluo-roethylene graft fistula for angioaccess in hemodialysis. Am J Surg1986;152:238-43.

26. Schuman ES, Gross GF, Hayes JF, Standage BA. Long-term patency ofpolytetrafluoroethylene graft fistulas. Am J Surg 1988;155:644-6.

Submitted May 8, 2007; accepted Oct 17, 2007.