From the Eastern Vascular Society Early surgical thrombectomy improves salvage of thrombosed vascular accesses Nirvana Sadaghianloo, MD, a,b Elixène Jean-Baptiste, MD, PhD, a,b Hacène Gaid, MD, c Mohamed S. Islam, MD, a,d Christophe Robino, MD, c Serge Declemy, MD, a,b Alan Dardik, MD, PhD, e and Réda Hassen-Khodja, MD, a,b Nice, France; Monte Carlo, Principality of Monaco; and New Haven, Conn Objective: The timing and urgency of salvage attempts for acutely thrombosed hemodialysis vascular accesses remain poorly defined. We examined the outcome of early surgical thrombectomy after acute access thrombosis to assess the influence of expedited timing on access salvage. Methods: Between January 2007 and October 2012, 114 surgical thrombectomy attempts were performed on 82 patients to salvage 89 accesses. The time between the diagnosis of thrombosis and admission to the operative suite (T1), the time between diagnosis and the following dialysis session (T2), and clinical and biologic parameters were collected prospec- tively. Data were retrospectively compared between the early (T1 <6 hours) and later (T1 >6 hours) treatment groups. The main outcome measure was technical success. Kaplan-Meier survival analysis was used to estimate functional patency rates. Results: Mean patient follow-up was 22 6 18 months. The mean time from referral to procedure (T1) was 5.7 6 4.5 hours. The mean time T1 was 3.6 6 1.2 hours in the early group and 10.3 6 5.4 hours in the later group. The mean time to dialysis (T2) was 14.3 6 6.5 hours in the early group and 23.9 6 9.4 hours in the later group. Thrombectomy performed #6 hours after diagnosis (T1 <6 hours) had significantly higher technical success of 86% compared with 69% for thrombectomy performed later (T1 >6 hours; P [ .04). The two groups did not differ significantly in patient comorbidities, type of access, or adjunctive procedures performed (P $ .1). At 12 months, the primary patency rate for all index cases, including technical failures, was 55% 6 7.1% in the early group and 33% 6 9.7% in the later group (P [ .13). The secondary patency rate was 67% 6 6.8% in the early group and 50% 6 9.9% in the later group (P [ .05). Conclusions: After acute access thrombosis, early surgical thrombectomy was associated with higher technical success and potentially improved midterm patency. (J Vasc Surg 2014;59:1377-84.) Acute thrombosis is a significant complication of he- modialysis vascular access that leads to morbidity, hos- pitalization, salvage procedures, potentially new access placement, and associated costs. 1,2 The National Kidney Foundation-Kidney Disease Outcomes Quality Initiative (NKF-KDOQI) guidelines recommend attempting access salvage as early as possible after thrombosis and also state that thrombectomy can be effective after several days. 3 Owing to the lack of published evidence, this practical recommendation remains misunderstood by many. The ur- gency of treatment of acute access thrombosis is often debated, with some authors suggesting that delays in treatment can be associated with an increased need for placement of short-term dialysis catheters, with their latent short-term and long-term complications. 4 Therefore, we examined the outcome of early surgical thrombectomy af- ter acute access thrombosis to assess the influence of expe- dited timing on access salvage. METHODS The clinical data of patients who underwent surgical thrombectomy of their hemodialysis vascular access in the Department of Vascular Surgery at the University of Nice between January 2007 and October 2012 were prospec- tively collected in a secured computerized database and retrospectively analyzed. The study was conducted accord- ing to the Declaration of Helsinki, and our institutional Ethics Committee approved the protocol. Patients were immediately referred to a vascular sur- geon by the treating nephrologist after the diagnosis of thrombosis. Early-access thrombosis (eg, <30 days after the original access creation) was not included in this study because emergent access salvage was attempted for all autologous and prosthetic functional vascular accesses with acute thrombosis. Preinterventional urgent hemodial- ysis was undertaken using a short-term jugular or femoral catheter only in cases of life-threatening clinical deteriora- tion such as serum potassium >6.5 mmol/L or acute pul- monary edema. Patients with acute access thrombosis >30 days after the original procedure were taken to the From the University of Nice Sophia Antipolis, Nice a ; the Department of Vascular Surgery, University Hospital of Nice, Nice b ; the Department of Nephrology-Hemodialysis, Princess Grace Hospital Centre, Monte Carlo c ; the Department of Hemodialysis, University Hospital of Nice, Nice d ; and the Department of Surgery, Yale University School of Medi- cine, New Haven. e Author conflict of interest: none. Presented at the Twenty-seventh Annual Meeting of the Eastern Vascular Society, White Sulphur Springs, WVa, September 19-22, 2013. Additional material for this article may be found online at www.jvascsurg.org. Reprint requests: Réda Hassen-Khodja, MD, Department of Vascular Sur- gery, University Hospital of Nice, 5 Rue Pierre Dévoluy, 06000 Nice, France (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214/$36.00 Copyright Ó 2014 by the Society for Vascular Surgery. http://dx.doi.org/10.1016/j.jvs.2013.11.092 1377
Mohamed S. Islam, MD,a,d Christophe Robino, MD,c Serge Declemy, MD,a,b Alan Dardik, MD, PhD,e andRéda Hassen-Khodja, MD,a,b Nice, France; Monte Carlo, Principality of Monaco; and New Haven, Conn
Objective: The timing and urgency of salvage attempts for acutely thrombosed hemodialysis vascular accesses remainpoorly defined. We examined the outcome of early surgical thrombectomy after acute access thrombosis to assess theinfluence of expedited timing on access salvage.Methods: Between January 2007 and October 2012, 114 surgical thrombectomy attempts were performed on 82 patientsto salvage 89 accesses. The time between the diagnosis of thrombosis and admission to the operative suite (T1), the timebetween diagnosis and the following dialysis session (T2), and clinical and biologic parameters were collected prospec-tively. Data were retrospectively compared between the early (T1 <6 hours) and later (T1 >6 hours) treatment groups.The main outcome measure was technical success. Kaplan-Meier survival analysis was used to estimate functional patencyrates.Results: Mean patient follow-up was 22 6 18 months. The mean time from referral to procedure (T1) was 5.7 6
4.5 hours. The mean time T1 was 3.6 6 1.2 hours in the early group and 10.3 6 5.4 hours in the later group. The meantime to dialysis (T2) was 14.3 6 6.5 hours in the early group and 23.9 6 9.4 hours in the later group. Thrombectomyperformed #6 hours after diagnosis (T1 <6 hours) had significantly higher technical success of 86% compared with 69%for thrombectomy performed later (T1 >6 hours; P [ .04). The two groups did not differ significantly in patientcomorbidities, type of access, or adjunctive procedures performed (P$ .1). At 12 months, the primary patency rate for allindex cases, including technical failures, was 55% 6 7.1% in the early group and 33% 6 9.7% in the later group (P [ .13).The secondary patency rate was 67% 6 6.8% in the early group and 50% 6 9.9% in the later group (P [ .05).Conclusions: After acute access thrombosis, early surgical thrombectomy was associated with higher technical success andpotentially improved midterm patency. (J Vasc Surg 2014;59:1377-84.)
Acute thrombosis is a significant complication of he-modialysis vascular access that leads to morbidity, hos-pitalization, salvage procedures, potentially new accessplacement, and associated costs.1,2 The National KidneyFoundation-Kidney Disease Outcomes Quality Initiative(NKF-KDOQI) guidelines recommend attempting accesssalvage as early as possible after thrombosis and also statethat thrombectomy can be effective after several days.3
Owing to the lack of published evidence, this practicalrecommendation remains misunderstood by many. The ur-gency of treatment of acute access thrombosis is oftendebated, with some authors suggesting that delays in
the University of Nice Sophia Antipolis, Nicea; the Department ofascular Surgery, University Hospital of Nice, Niceb; the DepartmentNephrology-Hemodialysis, Princess Grace Hospital Centre, Monte
arloc; the Department of Hemodialysis, University Hospital of Nice,iced; and the Department of Surgery, Yale University School of Medi-ne, New Haven.e
or conflict of interest: none.ented at the Twenty-seventh Annual Meeting of the Eastern Vascularciety, White Sulphur Springs, WVa, September 19-22, 2013.itional material for this article may be found online at www.jvascsurg.org.rint requests: Réda Hassen-Khodja, MD, Department of Vascular Sur-ry, University Hospital of Nice, 5 Rue Pierre Dévoluy, 06000 Nice,ance (e-mail: [email protected]).editors and reviewers of this article have no relevant financial relationshipsdisclose per the JVS policy that requires reviewers to decline review of anyanuscript for which they may have a conflict of interest.-5214/$36.00yright � 2014 by the Society for Vascular Surgery.://dx.doi.org/10.1016/j.jvs.2013.11.092
treatment can be associated with an increased need forplacement of short-term dialysis catheters, with their latentshort-term and long-term complications.4 Therefore, weexamined the outcome of early surgical thrombectomy af-ter acute access thrombosis to assess the influence of expe-dited timing on access salvage.
METHODS
The clinical data of patients who underwent surgicalthrombectomy of their hemodialysis vascular access in theDepartment of Vascular Surgery at the University of Nicebetween January 2007 and October 2012 were prospec-tively collected in a secured computerized database andretrospectively analyzed. The study was conducted accord-ing to the Declaration of Helsinki, and our institutionalEthics Committee approved the protocol.
Patients were immediately referred to a vascular sur-geon by the treating nephrologist after the diagnosis ofthrombosis. Early-access thrombosis (eg, <30 days afterthe original access creation) was not included in this studybecause emergent access salvage was attempted for allautologous and prosthetic functional vascular accesseswith acute thrombosis. Preinterventional urgent hemodial-ysis was undertaken using a short-term jugular or femoralcatheter only in cases of life-threatening clinical deteriora-tion such as serum potassium >6.5 mmol/L or acute pul-monary edema. Patients with acute access thrombosis>30 days after the original procedure were taken to the
JOURNAL OF VASCULAR SURGERY1378 Sadaghianloo et al May 2014
operative suite as early as possible and were included in thisstudy.
The surgical thrombectomy was performed in theoperative suite by an experienced vascular surgeon withthe patient under local anesthesia and intravenous sedation.Patients were brought to the operating room as early aspossible; our service has two rooms for elective cases andone for emergent cases.
The surgical technique consisted of milking thethrombus from the thrombosed access (manual clot re-moval) as described by Cull et al,5 adapted for prostheticgrafts, as well as fistulas, as originally described. A fistulo-gram was performed when technically feasible to confirmthe completion of thrombectomy and to find a cause of ac-cess failure. Stenoses were treated in the same procedurewith balloon angioplasty and stent placement or surgicalrevision, including total or partial conduit replacementand venous or prosthetic patch placement. Failed throm-bectomies led to access abandonment and placement of ashort-term catheter for the following dialysis, with a newvascular access created during the same procedure orplanned within a short time. Procedures were performedas outpatient treatments if possible, unless timing of hemo-dialysis was facilitated by an overnight admission.
The treating nephrologist and the vascular surgeonboth monitored the access during follow-up. In our prac-tice, nephrologists carefully follow-up the patients withclinical examination as well as flow measurement duringthe dialysis session. Doppler ultrasound imaging is per-formed if stenosis or thrombosis is suspected.
Demographic data and cardiovascular risk factors wererecorded for each patient, as were the number, location,and type of any current or previous access. The time be-tween the diagnosis of vascular access thrombosis and theadmission to the operative suite was recorded as T1, andthe time between the diagnosis and the following dialysissession was recorded as T2. Information regarding surgicalintervention and access outcome, such as complicationsand functional patency, were also recorded. Data wereretrospectively analyzed with comparison between the early(T1 <6 hours) and later (T1 >6 hours) treatment groups.
The main outcome measure was technical success,defined as the re-establishment of a functional accessthat was used for the following dialysis session. Any surgi-cal replacement of the access puncture zone was recordedas a technical failure. The secondary outcome was func-tional patency rate. An access was defined as functionalwhen able to deliver a flow rate of 350 to 400 mL/minwithout access recirculation to maintain a treatment timeof <4 hours.6
Primary patency was defined as the interval from thetime of the initial thrombectomy procedure (index proce-dure) until any intervention designed to maintain or re-establish patency, or until access rethrombosis. Secondarypatency was defined as the interval from the time of the in-dex procedure until access abandonment, rethrombosis, orperformance of interventions designed to maintain or re-establish functionality in a thrombosed access.
Statistical analysis was performed with StatView 5.0software (SAS Institute Inc, Cary, NC). All results werecalculated on an intention-to-treat-basis. Nominal variablesare expressed as a number and a percentage of patients oraccess. Differences of categoric variables were assessed bythe two-tailed Fisher exact test. Continuous variables areexpressed as the mean 6 standard deviation. Differencesbetween continuous variables were assessed by a two-tailed unpaired t-test or by nonparametric tests for non-Gaussian variables. Variables analyzed with univariable testswith a value of P < .10 were analyzed using logistic regres-sion analysis and reported as odds ratios (ORs) with 95%confidence intervals (CIs). Kaplan-Meier analysis wasused to calculate patency curves and survival after indexprocedures, with curves compared using log-rank andBreslow-Gehan-Wilcoxon analyses. The Cox proportionalhazards regression model was used to control potentiallyconfounding factors for censored variables. Differenceswere expressed as hazard ratios with 95% CIs. A value ofP < .05 was considered significant.
RESULTS
During the study period, 82 patients underwent 114surgical thrombectomy procedures to salvage a throm-bosed vascular access. Seven patients had two or morevascular access sites, leading to 89 index procedures on89 accesses. Patient demographics and access characteris-tics are summarized in Table I. Mean patient follow-upwas 22 6 18 months. Two patients were lost to follow-up at 1 month.
The mean time between the diagnosis of vascular ac-cess thrombosis and the admission to the operative suite(T1) was 5.7 6 4.5 hours. The mean time T1 was 3.6 61.2 hours in the early group and 10.3 6 5.4 hours in thelater treatment group. Fig 1 shows the distribution oftime to procedure (T1) in the cohort and shows an earlygroup with a normal distribution and a later group witheven distribution. Only 20 to 30 minutes were requiredfor transport of most patients from their place of referralto our center. Four patients required emergent hemodialy-sis with a short-term catheter before the surgical procedurefor life-threatening elevations in serum potassium (n ¼ 3)or acute pulmonary edema (n ¼ 1).
Stenosis was identified as the etiology for vascularaccess thrombosis in all but one procedure; stenoses weretypically in the venous outflow, although they occurredat all sites. The remaining access was thought to be throm-bosed due to severe cardiac dysfunction with a concomi-tant low-flow state. Table II lists the adjunctiveprocedures performed after the thrombectomy and themean procedure times. More surgical revisions were per-formed in the later group, although the difference wasnot statistically significant (P ¼ .09).
The overall technical success rate was 81% (92 of 114;Table III). A short-term catheter was necessary after 20technical failures (18%). Catheter placement was avoidedin the two remaining failed procedures: another func-tional vascular access was created in one patient in
Table I. Patient demographics and risk factors
Variablesa Total (N ¼ 82) T1 <6 hours (n ¼ 54) T1 >6 hours (n ¼ 28) P
T1, Mean time from referral to procedure.aContinuous data are shown as mean 6 standard deviation and categoricdata as number (%).
JOURNAL OF VASCULAR SURGERYVolume 59, Number 5 Sadaghianloo et al 1379
anticipation after a diagnosis of vascular access dysfunc-tion, and in the other patient, puncture of a newly createdbrachiocephalic fistula was permitted because of a mature
venous line (above a radiocephalic fistula). Three failuresoccurred #12 hours after a successful thrombectomyfrom unknown etiology. The technical success rate was87% in accesses without a history of previous throm-bosis compared with 70% in accesses with a history ofprevious thrombosis. Thrombectomy attempts per-formed #6 hours (T1 <6 hours) after diagnosis hadsignificantly higher technical success of 86% comparedwith 69% for procedures performed later (T1$6 hours; Table III). These variables were both signif-icant when analyzed with multivariable analysis (OR,3.11; 95% CI, 1.16-8.34; P ¼ .02; and OR, 2.96;95% CI, 1.10-7.98; P ¼ .03, respectively). Table IIIcompares technical success rates between the early andlater treatment group within different subgroups ofthe cohort.
Table III. Technical success rates, with subgroup analysis based on T1 by Fisher exact test
T1, Mean time from referral to procedure.aContinuous data are shown as mean 6 standard deviation and categoric data as number (%).
Fig 2. Kaplan-Meier analysis shows cumulative patency after thrombectomy attempts for all index cases. SE, Standard error.
JOURNAL OF VASCULAR SURGERY1380 Sadaghianloo et al May 2014
The mean interval between the time of diagnosisof thrombosed access and the following dialysis session(T2) was 17.3 6 8.8 hours. The mean time T2 was14.3 6 6.5 hours in the early and 23.9 6 9.4 hours inthe later treatment groups. Delayed dialysis was not identi-fied as a cause of complication after any of the 114procedures.
The overall survival rate was 82% 6 5% at 12 months.Two patients with severe comorbidities died of multisystemorgan failure (2%) within the postoperative period (#30days). Primary patency was 47% at 1 year and 35% at 2 yearsafter all index procedures, whereas secondary patency was61% at 1 year and 46% at 2 years (Fig 2). After successfulthrombectomy, primary patency was 56% at 1 year and
Fig 3. Cumulative (A) primary patency and (B) secondary patency by Kaplan-Meier analysis, with comparison ofcurves with Breslow-Gehan-Wilcoxon analysis, is shown after thrombectomy attempts for all index cases. Subgroupanalysis depended on time to procedure T1 (<6 hours or >6 hours). SE, Standard error.
JOURNAL OF VASCULAR SURGERYVolume 59, Number 5 Sadaghianloo et al 1381
41% at 2 years, whereas secondary patency was 74% at 1 yearand 56% at 2 years (data not shown). At 1 year, the primarypatency rate for all index procedures was 55% in the earlygroup and 33% in the later group (log-rank, P ¼ .26;Breslow-Gehan-Wilcoxon, P ¼ .13; Fig 3, A). The second-ary patency rate was 67% in the early group and 50% in thelater group (Fig 3, B); the comparison of the two curveswith the Breslow-Gehan-Wilcoxon test shows improved sec-ondary patency (P ¼ .05; log-rank, P ¼ .16).
To determine the effect of access type on outcome,outcome after thrombectomy of autologous accesses wascompared with prosthetic accesses. Technical success rateswere similar between autologous and prosthetic accesses(81% vs 80%, respectively; P > .99). For prosthetic ac-cesses, there was a trend toward higher technical successrate in the early treatment group (P ¼ .09; Table III).Fig 4 shows the patency curves of autologous (A) andprosthetic grafts (B), including all index cases. Among
Fig 4. Kaplan-Meier analysis shows cumulative patency after thrombectomy attempts and subgroup analysis of (A)autologous fistulas vs (B) prosthetic access grafts for index cases. C, Secondary patency is shown for autologous vsprosthetic accesses, with exclusion of technical failures. SE, Standard error.
JOURNAL OF VASCULAR SURGERY1382 Sadaghianloo et al May 2014
JOURNAL OF VASCULAR SURGERYVolume 59, Number 5 Sadaghianloo et al 1383
autologous accesses, primary patency was 57% and second-ary patency was 77% at 1 year after successful thrombec-tomy compared with 55% and 67% among prostheticaccesses, respectively. Although autologous accesses wereassociated with a trend toward higher primary patency(P ¼ .09, data not shown), they had higher secondarypatency rates compared with prosthetic accesses (log-rank, P ¼ .01; Fig 4, C). Similarly, multivariable Coxanalysis showed that autologous access was significantlyassociated with increased patency (hazard ratio, 0.51;95% CI, 0.27-0.96; P ¼ .01). The duration T1 was notassociated with patency in the subgroup analysis betweenautologous and prosthetic access (P > .2; SupplementaryFig, online only).
DISCUSSION
In a practice where early thrombectomy is the rule,thrombectomy performed #6 hours after diagnosis hadsignificantly better outcome than thrombectomy per-formed >6 hours (Table III). We believe that this is thefirst report demonstrating that time to performance ofthe procedure can improve access salvage. These resultsconfirm the NKF-DOQI paradigm, suggesting thatvascular access thrombosis is an emergency for the access.3
Our overall technical success rate was 81%, which is inthe range of previously published data for open thrombec-tomies for autologous and prosthetic accesses.5,7-12 Onecould have expected higher success rates; however, 39%of our thrombectomy cohort presented with a history ofprevious thrombosis. Our results with this high-risk groupconfirmed that reintervention after a previous thrombosiswas less likely to be successful, as previously suggested.13
In addition, as a tertiary center that manages complicatedaccess cases, our high proportion of complex cases, suchas reoperations and unusual access sites, makes compari-son with published data on homogeneous cohorts moredifficult.
Lack of power in the subgroup analysis of our cohort didnot allow us to prove that time to thrombectomy attempt in-fluences long-term patency. However, studies on lower limbveins have shown that prolonged thrombosis is responsiblefor venous wall inflammation, characterized by cell and ma-trix proliferation as well as by endothelial dysfunction. Thick-ening, fibrosis, and loss of venous wall compliance may favorrethrombosis.14,15 In thrombosed autologous fistulas, thisinflammatory process appears early after thrombosis andmay explain low patency after thrombectomy.16 In arterio-venous prosthetic grafts, >90% of failures were attributedto intimal hyperplasia of the venous anastomosis.13,17,18
Furthermore, the prosthetic graft develops neointimal hy-perplasia after implantation and is no longer an inert mate-rial.19 However, our data do not clarify how much timepasses between the onset of thrombosis and subsequentvenous wall injury. We only found that autologous accesseshad higher post-thrombectomy patency rates than pros-thetic grafts, confirming previously published data.13,20,21
Short-term catheter placement and use are by no meanstrivial. Consequences for the patient range from discomfort
and pain to infection, central vein thrombosis, and eventu-ally, death.22,23 Although it was impossible to predict howmany unnecessary short-term catheters (and their latentcomplications) would have been avoided with early throm-bectomy, we believe that our usage of 18% is a very lownumber. Rates of catheter placement after access thrombosisare rarely published, but a recent meta-analysis reported anaverage of 26% based on three prospective studies.9
We acknowledge some limitations to our study. Thiswas a single-center, nonrandomized, noncontrolled study.Our cohort was heterogeneous in access type, makingcomparison with previous success and patency data diffi-cult. This study was not designed to prove the efficiencyof our surgical thrombectomy technique but rather the in-fluence of its prompt implementation on access salvage.Indisputable evidence in favor of any thrombectomy tech-nique is still lacking.7,13 Our early thrombectomy policywas also limited by objective data regarding the exact onsetof access thrombosis, as limited by detection of the refer-ring physician.
Finally, larger studies will be needed to tell whetherautologous and prosthetic access should be treated thesame or differently regarding time to thrombectomyattempt. The small number of patients in this study in-creases the likelihood of a type II error and increases theportion of the survival curves with a standard error of>10%.
CONCLUSIONS
Our study tends to confirm that early surgical throm-bectomy, attempted as soon as possible after referral for ac-cess thrombosis, is associated with higher technical successand potentially better midterm patency. Improved salvagemay allow reduced numbers of access catheters to beplaced, thus reducing patient morbidity and concomitantcosts of care.
AUTHOR CONTRIBUTIONS
Conception and design: NS, EJ, HG, MI, CR, SD, AD,RH
Analysis and interpretation: NS, SD, ADData collection: NS, EJ, HG, MI, CR, SDWriting the article: NS, AD, RHCritical revision of the article: NS, EJ, HG, MI, CR, SD,
AD, RHFinal approval of the article: NS, EJ, HG, MI, CR, SD,
AD, RHStatistical analysis: NSObtained funding: Not applicableOverall responsibility: RH
REFERENCES
1. Brenner L, Singh AK, Campbell D, Frei F, Winkelmayer WC. Associ-ations between demographic factors and provider structures on costand length of stay for hemodialysis patients with vascular access failure.Clin J Am Soc Nephrol 2006;1:455-61.
2. Sawant A, Mills PK, Dhingra H. Increased length of stay and costsassociated with inpatient management of vascular access failures. SeminDial 2013;26:106-10.
JOURNAL OF VASCULAR SURGERY1384 Sadaghianloo et al May 2014
3. National Kidney Foundation (NKF) Kidney Disease Outcomes QualityInitiative (KDOQI). Clinical practice guidelines for vascular access2006. Available at: http://www.kidney.org/professionals/KDOQI/guideline_upHD_PD_VA/index.htm. Accessed August 2013.
4. Hakim R, Himmelfarb J. Hemodialysis access failure: a call to action.Kidney Int 1998;54:1029-40.
5. Cull DL, Washer JD, Carsten CG, Keahey G, Johnson B. Descriptionand outcomes of a simple surgical technique to treat thrombosedautogenous accesses. J Vasc Surg 2012;56:861-5.
6. Sidawy AN, Gray R, Besarab A, Henry M, Ascher E, Silva M Jr, et al.Recommended standards for reports dealing with arteriovenous he-modialysis accesses. J Vasc Surg 2002;35:603-10.
7. Tordoir JH, Bode AS, Peppelenbosch N, van der Sande FM, deHaan MW. Surgical or endovascular repair of thrombosed dialysisvascular access: is there any evidence? J Vasc Surg 2009;50:953-6.
8. Lipari G, Tessitore N, Poli A, Bedogna V, Impedovo A, Lupo A, et al.Outcomes of surgical revision of stenosed and thrombosed forearmarteriovenous fistulae for haemodialysis. Nephrol Dial Transplant2007;22:2605-12.
9. Kuhan G, Antoniou GA, Nikam M, Mitra S, Farquharson F,Brittenden J, et al. A meta-analysis of randomized trials comparingsurgery versus endovascular therapy for thrombosed arteriovenous fis-tulas and grafts in hemodialysis. Cardiovasc Intervent Radiol 2013;36:699-705.
10. Ponikvar R. Surgical salvage of thrombosed arteriovenous fistulas andgrafts. Ther Apher Dial 2005;9:245-9.
11. Palmer RM, Cull DL, Kalbaugh C, Carsten CG, Taylor SM,Snyder BA, et al. Is surgical thrombectomy to salvage failed autogenousarteriovenous fistulae worthwhile? Am Surg 2006;72:1231-3.
12. Marston WA, Criado E, Jaques PF, Mauro MA, Burnham SJ,Keagy BA. Prospective randomized comparison of surgical versusendovascular management of thrombosed dialysis access grafts. J VascSurg 1997;26:373-80; discussion: 380-1.
13. Sidawy AN, Spergel LM, Besarab A, Allon M, Jennings WC,Padberg FT Jr, et al. The Society for Vascular Surgery: clinical practiceguidelines for the surgical placement and maintenance of arteriovenoushemodialysis access. J Vasc Surg 2008;48:2-25S.
14. Henke PK, Wakefield T. Thrombus resolution and vein wall injury:dependence on chemokines and leukocytes. Thromb Res 2009;123(Suppl 4):S72-8.
15. Wakefield TW, Myers DD, Henke PK. Mechanisms of venousthrombosis and resolution. Arterioscler Thromb Vasc Biol 2008;28:387-91.
16. Chang CJ, Ko YS, Ko PJ, Hsu LA, Chen CF, Yang CW, et al.Thrombosed arteriovenous fistula for hemodialysis access is character-ized by a marked inflammatory activity. Kidney Int 2005;68:1312-9.
17. Munda R, First MR, Alexander JW, Linnemann CC Jr, Fidler JP,Kittur D. Polytetrafluoroethylene graft survival in hemodialysis. JAMA1983;249:219-22.
18. Swedberg SH, Brown BG, Sigley R, Wight TN, Gordon D,Nicholls SC. Intimal fibromuscular hyperplasia at the venous anasto-mosis of PTFE grafts in hemodialysis patients. Clinical, immunocyto-chemical, light and electron microscopic assessment. Circulation1989;80:1726-36.
19. Roy-Chaudhury P, Kelly BS, Miller MA, Reaves A, Armstrong J,Nanayakkara N, et al. Venous neointimal hyperplasia in polytetra-fluoroethylene dialysis grafts. Kidney Int 2001;59:2325-34.
21. Ito Y, Sato T, Okada R, Nakamura N, Kimura K, Takahashi R, et al.Comparison of clinical effectiveness between surgical and endovasculartreatment for thrombotic obstruction in hemodialysis access. J VascAccess 2011;12:63-6.
22. Oguzkurt L, Tercan F, Torun D, Yildirim T, Zumrutdal A, Kizilkilic O.Impact of short-term hemodialysis catheters on the central veins: acatheter venographic study. Eur J Radiol 2004;52:293-9.
23. Vats HS. Complications of catheters: tunneled and nontunneled. AdvChronic Kidney Dis 2012;19:188-94.
Submitted Sep 11, 2013; accepted Nov 29, 2013.
Additional material for this article may be found onlineat www.jvascsurg.org.
Supplementary Fig (online only). Kaplan-Meier and Breslow-Gehan-Wilcoxon analysis show cumulative patencyafter thrombectomy attempts and subgroup analysis of autologous vs prosthetic accesses depending on time to pro-cedure T1 (<6 hours or >6 hours) for all index cases. A, Primary and (B) secondary patency are shown for autologousfistulas. SE, Standard error.
JOURNAL OF VASCULAR SURGERYVolume 59, Number 5 Sadaghianloo et al 1384.e1
Supplementary Fig (online only). Continued. C, Primary and (D) secondary patency are shown for prosthetic grafts.SE, Standard error.
JOURNAL OF VASCULAR SURGERY1384.e2 Sadaghianloo et al May 2014
