Paclitaxel drug-eluting balloons to recurrent in-stent stenoses in autogenous dialysis fistulas: a retrospective study

Introduction

The autogenous arteriovenous fistula (AVF) is the optimum haemodialysis access in end-stage renal failure (1). The commonest problem with the AVF is stenosis, which can prevent the fistula from maturing, can lead to a problematic fistula with inadequate dialysis or can precipitate fistula occlusion (2, 3). Conventional angioplasty is associated with a high recurrence rate of the stenosis and repeated interventions (4). Surgical revision is more invasive and also has a high recurrence rate (5). The advent of successful drug-eluting technology in the treatment of the coronary circulation (6), and subsequent mixed results in the peripheral circulation (7-8-9-10), prompted us to start using drug-eluting balloons (DEB) in recurrent in-stent AVF stenoses when these became available in Australia in 2010.

Methods

Study population

This is a retrospective study based on prospectively collected audit data of endovascular interventions using a DEB on recurrent in-stent stenotic lesions in the AVF circuit of our haemodialysis population.

Since 2003, all patients with a problematic fistula in our service have been referred to the principal author’s Vascular Ultrasound Service for assessment of the fistula problem and intervention planning. Surgical revision for fistula stenosis has declined over the years and since 2008 more than 90% of fistula stenosis interventions have been by endovascular techniques.

Data collection

All data for the study were collected on a prospective audit database held by the principal author. As haemodialysis patients are a captive population, very few are lost to follow-up.

Ethics approval

Approval for the study was obtained from our ethics committee.

Study period

The period of our study was from 13 September 2010, when we used our first DEB in AVF, to 1 January 2014.

Study group

This is a study of the effect of DEB on a particular stenotic lesion in a particular fistula circuit – the index lesion. All index lesions in the study group at the time of the first DEB were in-stent stenoses. The study, therefore, only looks at intervention rates to the index stenosis in a fistula circuit, before and after application of the DEB. Interventions to other parts of the fistula circuit both before and after DEB intervention are not included. All patients who received a DEB to a stenotic lesion in their AVF since the DEB became available in Australia in 2010 were recorded in the prospective audit database.

Inclusion/exclusion criteria for the present study from this group were:

Inclusion

Only the first time that a DEB was used (Index DEB) on the index stenosis qualified for study entry. Second and subsequent DEBs to the index lesion were considered “secondary DEB” interventions. The secondary DEB interventions were used in calculating the Kaplan–Meier curve in Figure 3.

Exclusions

When the DEB was used at the time of the first intervention to a fistula stenosis (i.e. on a de novo, non-recurrent lesion).

When there were multiple overlapping interventions, lesions and stents in a particular fistula making analysis too complicated.

DEB used on the central veins in the fistula circuit (axillo-subclavian-brachiocephalic veins) were excluded because the biggest DEB available was 7 mm, which is of inadequate diameter in the central veins. We used non-standard techniques to deliver drug elution to central venous angioplasty.

Fistula stenosis intervention

Indication for stenosis re-intervention by our group, and for the purposes of this trial, consisted of the following two categories:

1.

A dialysing fistula that is considered to have a significant stenosis. This judgement is based on the following parameters, in isolation or in combination:

-

Dialysis is considered inadequate by the attending renal physician.

-

Physical examination of the fistula suggests a stenosis at the index site.

-

Inadequate fistula volume flow – Qa: As measured on ultrasound, or a declining Qa over time for a given fistula. We consider a Qa <500 mL/min as probably inadequate.

-

“Dialysis numbers” – Qb: A significant stenosis is suggested when the pump speed is <300 mL/min, venous pressure is >140 mmHg and/or arterial pressure >100 mmHg.

-

Minimal diameter of the index lesion <2.7 mm: The 2.7 mm cut-off is derived from our previous study defining what constitutes a significant AVF stenosis (11). For the purposes of this study, the indications for fistula intervention prior to the use of DEB and after their introduction remained identical.

2.

A non-dialysing fistula that is failing to mature and has a stenosis demonstrated on ultrasound of <2.7 mm diameter and has a Qa as measured by ultrasound in the feeding brachial artery of less than 500 mL/min.

Technique of intervention in AVF stenosis by our group is very aggressive, with frequent rupture of the fistula at the site of stenosis, and the frequent use of uncovered nitinol stents. The aggressive angioplasty technique we use is not the subject of this paper, but the stenosis has to be “defeated” at the time of angioplasty; if this is not done, the stenosis will recoil in the early postoperative period (minutes to hours) and there will be a high rate of recurrence. Again, for the purposes of this study, our technique of fistula intervention prior to the use of DEB and after their introduction remained identical.

Drug-eluting balloon

The Medtronic paclitaxel DEB (In.Pact Admiral, Medtronic) was used in all cases. Note that Medtronic has two DEBs on the market, with different manufacturing processes: the In.Pact Admiral, on the 035 wire, used in our series; and the In.Pact Amphirion, on the 018 wire, used in the IN.PACT DEEP DEB BTK trial (12).

To ensure that effective drug delivery to the target lesion occurred, manufacturer’s instructions were carefully followed with the following modifications.

-

We did not remove the red protective sheath in which the balloon is packaged; we left this undisturbed on the DEB and used it to open the valve in the access sheath to avoid losing drug on the valve. This non-standard step was reviewed by the company and approved.

-

Although a 6 F sheath is recommended by Medtronic, we found this a rather “tight fit”, with concerns of drug being lost in the sheath; hence, our standard sheath for DEB was 7 F.

-

Evidence from the use of DEB in the coronary circulation shows that drug is lost during dwell time in the circulation; hence, we endeavoured to deliver the balloon to target and inflate it with minimum delay (13).

-

The DEB is applied as the last step in the procedure. The DEB must cover all areas treated with angioplasty/stenting at the time of intervention plus a 1 cm margin on either side (to avoid the “candy wrapper” effect (14)).

-

The diameter of the DEB was the diameter of the largest angioplasty balloon used + 1 mm.

Follow-up

Our unit has a fistula surveillance programme, whose robustness has increased over the years. During the study period, surveillance was performed by clinical examination and dialysis numbers (Qb) by the dialysis nurses and the patients themselves (about one-third of our patients are on home haemodialysis). In the last 5 years, “flometry” – measurement of fistula flow (Qa) at the time of dialysis – has been introduced to our dialysis centres and periodic ultrasound surveillance has become almost standard. This is particularly true of the problem fistulas with frequent restenosis, and this would include all patients in our study group.

Statistical analysis

Patients who had less than 6 months follow-up after application of the Index DEB at the time of the study analysis were excluded. The data were analysed in two ways.

Firstly, a Kaplan–Meier curve was constructed for the last “disease-free interval” immediately prior to the use of the Index DEB and the first “disease-free interval” immediately after the Index DEB. This allowed us to compare “re-intervention-free percentage at 12 months” before and after the use of the DEB.

Secondly, because the above analysis fails to take into account the often multiple interventions to the index lesion prior and subsequent to the use of the Index DEB, we used a marginal proportional hazards model (Wei, Lin and Weissfield (15)) to estimate the hazard ratio for “DEB present vs. DEB absent” across the multiple recurrences observed for each lesion.

The pre-Index DEB “disease-free survival” was calculated as the mean time between interventions to the index lesion before application of the Index DEB.

The post-Index DEB “disease-free-survival” was calculated as follows:

-

Some lesions did not require any further intervention up to the end of the study, and this time interval was used for the calculation.

-

Some lesions had one or more further interventions with DEB after the Index DEB, and the intervals between these interventions were used for the calculation.

-

Some lesions had further interventions using non-DEB technology after the Index DEB; in this case, only the intervals between Index DEB and first non-DEB re-intervention afterwards were used for the calculation.

-

Some patients died or received a transplant some time after their Index DEB; this was taken as the end of the study.

-

If a fistula occluded during follow-up after the Index DEB, this date was taken as the end of study.

Results

In the 39 months between 13 September 2010 and 31 December 2013, we performed 625 endovascular AVF interventions (Fig. 1). In this period, 86 AVF stenoses were treated with a DEB.

Fig. 1 -

A total of 49 interventions were excluded from the study; 11 were excluded because the DEB was the first intervention to that lesion, i.e. the lesion was not a recurrent stenosis. Six were excluded because the lesion was in the central veins; 18 were excluded because the DEB intervention followed previous DEB intervention; 10 were excluded because the first DEB to the lesion was used after 1 June 2013, giving less than 6 months of follow-up. Four were excluded because the fistula management and lesions were too complex for analysis – multiple, overlapping lesions with multiple interventions both open and endovascular making analysis of the effect of the DEB impossible.

Thirty-seven of the 86 stenoses (in 31 fistulas in 31 patients) were included for the study. There were six fistulas in six patients that had two separate stents with recurrent in-stent stenosis; all six were dialysing fistulas; 28 of the 31 fistulas were currently being used for dialysis and three of the 31 fistulas were being treated for “failure to mature”. The location of the stenosis was in the “useable segment” of the fistula in 13/37, in the juxta-anastomotic segment in 10/37 cases (16), in the “arch vein” (the perpendicular portion of the proximal cephalic vein in the deltopectoral groove, where it enters the subclavian vein) of brachiocephalic fistulas in 8/37, in the radiocephalic fistula outflow at the elbow in 4/37 and in the radial artery in 2/37.

Study group characteristics were: mean patient age: 56.1 years; 16/31 (48%) patients female, 15/31 (52%) male; 24% smokers, 63% diabetics. 54% were forearm fistulas, 46% were upper arm fistulas. The mean age of the fistula at time of study entry was 3.1 years (range 6 months–11 years).

All 37 stenoses included for analysis in the trial contained bare nitinol stents, which had been placed at the time of a previous intervention to the index site.

In the study group, the index stenosis had received between 1 and 11 previous interventions, an average of three interventions prior to the application of the DEB. The bare nitinol stent was placed in the index stenosis mostly at the first or second intervention. The study, therefore, is a comparison of in-stent stenosis before and after the introduction of DEBs.

Figure 2 shows the Kaplan–Meier curve of the “disease-free interval” using only the last pre-DEB interval and the first post-DEB interval. The Kaplan–Meier curve shows a significant difference in time to re-intervention before and after application of the DEB to the index stenosis: before using the DEB, only 19% of the study group stenoses were free of re-intervention at 12 months; after using the DEB, 69% of the study group stenoses were free of re-intervention at 12 months.

Fig. 2 -

Figure 3 shows the Kaplan–Meier curve for “DEB present vs. DEB absent” across the multiple recurrences observed for each lesion. The hazard ratio for “DEB present” vs. “DEB absent” was 0.23 (95% CI 0.14 to 0.36, p<0.001).

Fig. 3 -

Discussion

Our paper addresses in-stent restenosis in the fistula circuit. This restenosis is, in part or in whole, the result of neointimal hyperplasia (NIH), or some variant thereof. Lee et al, in their article on the pathophysiology of haemodialysis access stenosis, identified “venous NIH” as the main culprit in access stenosis. NIH is the blood vessel’s healing response to the trauma of endovascular intervention, and typically occurs over a period of 6 weeks to 6 months following the trauma to the vessel wall. Unless there is ongoing trauma/irritation, the NIH response should have died down by 1 year post-injury in most cases (17). In our study, the stenoses were often multiply recurrent, and a degree of perivenous scarring from repeat angioplasty/repeat needling was clinically evident, and all had received a stent at some stage.

A critical component of NIH is the cellular proliferative stage (9); most experimental studies identify mononuclear leucocytes as the primary inflammatory cell type involved. The rationale of drug elution is to block the NIH response with an antimetabolite such as paclitaxel (11, 18).

Drug-eluting technology, both DEBs and drug-eluting stents, pioneered by the cardiologists (6), has proven itself in clinical practice to be effective in reducing restenosis rates in the coronary circulation. Results with drug-eluting technology in the peripheral circulation have been mixed so far, with favourable 4-year data from the Zilver® PTX® randomized controlled trial (19) (Drug Eluting Stents in femoro-popliteal disease) but negative results with DEBs in infra-popliteal disease in the IN.PACT DEEP randomized trial (12). Katsanos et al (9), in a randomised prospective study investigating angioplasty with paclitaxel-coated balloons vs. plain balloon angioplasty for the treatment of failing native AVFs or prosthetic arteriovenous grafts, concluded that DEB improved patency of venous stenoses of failing vascular access used for dialysis (At 6 months, cumulative target lesion primary patency 70% for the DEB group vs. 25% in Plain Old Balloon Angioplasty group, p<0.001.)

It is important to emphasize that the role of drug elution in the treatment of vascular stenoses is not to obtain a good result; the role of drug elution is to preserve a good result, obtained on the day of intervention, from later restenosis due to NIH. Less obviously, drug elution does not deal with mechanical recoil, which is common after angioplasty (20-21-22) including in the AVF circuit.

We started using DEB in native fistulas when this technology became available in Australia in 2010. Initially, we used the DEBs only on fistula stenoses that had multiple, short spaced recurrences – the “frequent fliers”. These “frequent fliers” represent perhaps 10% of our dialysis population, and the reason for the aggressive NIH response may be partly genetically determined (23). The advent of DEB appeared to have a dramatic effect on this patient population in our practice, with a clinically apparent reduction in their re-intervention rates. Because of these favourable results, we extended the use of DEB in AVF, first to all restenoses (not just the frequent ones), and currently we have started using DEB on those primary stenoses where we anticipate a high chance of early recurrence, such as arch stenoses. It must be emphasised that the patients in the study represent our most difficult patients, with the most aggressive restenosis. They had all received one or more interventions to their lesions using our very aggressive angioplasty techniques outlined above, and all had received a bare uncovered nitinol stent to the index lesion as a result of this. It is a weakness of this study that our indications for intervention and re-intervention in the fistula circuit have our own unit peculiarities, as well as our technique of intervention: however, all these factors were the same before and after the introduction of DEBs.

The statistical analysis of this retrospective series confirms our clinical impression of the significant effect of DEB on reducing fistula stenosis recurrence.

Comparing the results of fistuloplasty between different studies is difficult because of the many variables involved. Mortamais et al (24) looked at the long-term results of fistuloplasty in 147 native fistulas with juxta-anastomotic stenosis, many of which were recurrent; their primary patency rate at 1 year was 47% and at 3 years was 26%, figures in keeping with other reviews on this subject. Patanè et al (10) reported their results using the DEB in 26 consecutive fistulas, but these were mostly de novo stenoses, not recurrent stenoses as in our series. Their target lesion primary patency at 6 months was 96%, at 1 year 81% and at 2 years 52%. The target lesion primary patency in our series at 6 months was 98% and at 18 months was 48%. It is hard to draw any conclusions from these study comparisons.

As a result of the above problems, we have started a randomised control trial, comparing the treatment of recurrent AVF stenoses with standard endovascular treatment + DEB vs. standard endovascular treatment + sham balloon.

Conclusion

This retrospective study would indicate that the DEB significantly reduces re-intervention rates on recurrent in-stent AVF stenoses.

Algorithm for inclusion into the study.

Kaplan-Meier curves for observations from the last pre-DEB intervention and the first post-DEB intervention on the 37 study lesions. DEB = drug-eluting balloon.

The marginal proportional hazards model for drug-eluting balloon present vs. absent across the multiple recurrences observed for each lesion. The hazard ratio for DEB present vs. absent was 0.23 (95% CI 0.14 to 0.36, p<0.001).

• 1. Besarab A.,Allon M.,Robbin ML. Resolved: fistulas are preferred to grafts as initial vascular access for dialysis. Pro J Am Soc Nephrol 2008; 19: 1629-1631
• 2. Clinical practice guidelines for vascular access, update 2006. I. Clinical practice guidelines for vascular access. Am J Kidney Dis 2006; 48: S187-S258
• 3. Beathard GA.,Arnold P.,Jackson J.,Litchfield T. Physician Operators Forum of RMS Lifeline. Kidney Int 2003; 64: 1487-1494
• 4. Rajan DK.,Bunston S.,Misra S.,Pinto R.,Lok CE. Dysfunctional autogenous hemodialysis fistulas: outcomes after angioplasty—are there clinical predictors of patency? Radiology 2004; 232: 508-515
• 5. Lipari G.,Tessitore N.,Poli A. Outcomes of surgical revision of stenosed and thrombosed forearm arteriovenous fistulae for haemodialysis. Nephrol Dial Transplant 2007; 22: 2605-2612
• 6. Stone GW.,Moses JW.,Ellis SG. Safety and efficacy of sirolimus- and paclitaxel-eluting coronary stents. N Engl J Med 2007; 356: 998-1008
• 7. Chan YC.,Cheng SW. Drug-eluting stents and balloons in peripheral arterial disease: evidence so far. Int J Clin Pract 2011; 65: 664-668
• 8. Liistro F.,Angioli P.,Porto I. Paclitaxel-eluting balloon vs. standard angioplasty to reduce recurrent restenosis in diabetic patients with in-stent restenosis of the superficial femoral and proximal popliteal arteries: the DEBATE-ISR study. J Endovasc Ther 2014; 21: 1-8
• 9. Katsanos K.,Karnabatidis D.,Kitrou P.,Spiliopoulos S.,Christeas N.,Siablis D. Paclitaxel-coated balloon angioplasty vs. plain balloon dilation for the treatment of failing dialysis access: 6-month interim results from a prospective randomized controlled trial. J Endovasc Ther 2012; 19: 263-272
• 10. Patanè D.,Giuffrida S.,Morale W. Drug-eluting balloon for the treatment of failing hemodialytic radiocephalic arteriovenous fistulas: our experience in the treatment of juxta-anastomotic stenoses. J Vasc Access 2014; 15: 338-343
• 11. Fahrtash F.,Kairaitis L.,Gruenewald S. Defining a significant stenosis in an autologous radio-cephalic arteriovenous fistula for hemodialysis. Semin Dial 2011; 24: 231-238
• 12. Zeller T.,Baumgartner I.,Scheinert D. IN.PACT DEEP Trial Investigators. Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial. J Am Coll Cardiol 2014; 64: 1568-1576
• 13. Gray WA.,Granada JF. Drug-coated balloons for the prevention of vascular restenosis. Circulation 2010; 121: 2672-2680
• 14. Angiolillo DJ.,Sabatá M.,Alfonso F.,Macaya C. “Candy wrapper” effect after drug-eluting stent implantation: déjà vu or stumbling over the same stone again? Catheter Cardiovasc Interv 2004; 61: 387-391
• 15. Wei L.,Lin D.,Weissfield L. Regression analysis of multivariate incomplete failure time data. J Am Stat Assoc 1989; 84: 1065-1073
• 16. Swinnen J.,Tan K L.,Allen R.,Burgess D.,Mohan V. Juxta-­anastomotic stenting with aggressive angioplasty will salvage the native radiocephalic fistula for dialysis. J Vasc Surg 2015; 61: 436-442
• 17. Lee T.,Roy-Chaudhury P. Advances and new frontiers in the pathophysiology of venous neointimal hyperplasia and dialysis access stenosis. Adv Chronic Kidney Dis 2009; 16: 329-338
• 18. Wöhrle J. Drug-coated balloons for coronary and peripheral interventional procedures. Curr Cardiol Rep 2012; 14: 635-641
• 19. Dake MD.,Ansel GM.,Jaff MR. Sustained safety and effectiveness of paclitaxel-eluting stents for femoropopliteal lesions: 2-year follow-up from the Zilver PTX randomized and single-arm clinical studies. J Am Coll Cardiol 2013; 61: 2417-2427
• 20. Rensing BJ.,Hermans WR.,Strauss BH.,Serruys PW. Regional differences in elastic recoil after percutaneous transluminal coronary angioplasty: a quantitative angiographic study. J Am Coll Cardiol 1991; 17: 34B-38B
• 21. Baumann F.,Fust J.,Engelberger RP. Early recoil after balloon angioplasty of tibial artery obstructions in patients with critical limb ischemia. J Endovasc Ther 2014; 21: 44-51
• 22. Gardiner GA.,Bonn J.,Sullivan KL. Quantification of elastic recoil after balloon angioplasty in the iliac arteries. J Vasc Interv Radiol 2001; 12: 1389-1393
• 23. Lee T.,Wadehra D. Genetic causation of neointimal hyperplasia in hemodialysis vascular access dysfunction. Semin Dial 2012; 25: 65-73
• 24. Mortamais J.,Papillard M.,Girouin N. Endovascular treatment of juxta-anastomotic venous stenoses of forearm radiocephalic fistulas: long-term results and prognostic factors. J Vasc Interv Radiol 2013; 24: 558-564