Sirolimus-coated balloon angioplasty in maintaining the patency of thrombosed arteriovenous graft: 1-year results of a prospective study

Abstract

Background:

A prospective, pilot study was designed to test the feasibility of using sirolimus-coated balloon (SCB) to treat graft vein junction of thrombosed arteriovenous graft (AVG) following successful pharmacomechanical thrombectomy. The present report provides the 1-year results of this study.

Methods:

This is a 1-year follow-up of a single, prospective, single-arm study that was conducted from 2018 to 2019 in 20 patients who presented to a tertiary institution with thrombosed AVG. The recruited patients received SCB angioplasty at the graft-vein junction following successful endovascular thrombectomy of a thrombosed AVG. One year after recruitment, there were three deaths, one AVG revision, and one AVG explantation among the participants recruited. The outcomes of 15 subjects at 1-year following the index procedure obtained from electronic medical records were re-examined.

Results:

The 1-year access circuit primary patency rate was 40%, while assisted primary and secondary patency rates were 46.7% and 73.3%, respectively. A total of 16 interventions (4 angioplasties, 12 thrombectomies) were performed in 9 patients over the 12 months. Four AVGs were abandoned. The median number of interventions per patient was 1 (0–3) per year. Using Kaplan-Meier analysis, the mean estimated post-intervention access circuit primary patency was 230 (95% CI: 162–300) days, while access circuit assisted primary patency was 253 (95% CI: 187–320) days, and access circuit secondary patency was 292 (95% CI: 230–356) days. Sub-group analysis did not show a significant difference in the mean estimated primary patency between AVG with de novo and recurrent stenosis (245 days, 95% CI: 151–339 vs 210 days, 95% CI: 113–307; p = 0.29).

Conclusions:

SCB may help sustain the patency of thrombosed AVG following successful thrombectomy.

Despite its poorer patency outcomes and the need for repeated interventions when compared to AVF, the use of arteriovenous graft (AVG) in patients with end stage kidney disease (ESKD) requiring hemodialysis (HD) is preferred over central venous catheter (CVC) due to its lower association with vascular access-related events. ¹

Flow limiting stenoses secondary to neointimal hyperplasia at the graft-vein junction and outflow vein stenoses are the main culprit lesions for AVG dysfunction. ² Understanding the pathogenesis of venous neointimal hyperplasia in AVG has allowed interventions aimed at retarding this process. Direct application of anti-proliferative agents such as paclitaxel at the graft-vein junction in dysfunctional AVG has been shown to improve 6-month patency rates post-intervention. ³ Until recently, paclitaxel was used exclusively as the active agent in drug-coated balloon technology. While multiple randomized control trials have demonstrated significant patency benefits in arteriovenous access stenosis treated with paclitaxel-coated balloon,^{4 –7} controversy surfaced after a meta-analysis showed an increased risk of death in patients treated with paclitaxel-based devices for peripheral arterial disease. ⁸

The sirolimus-coated balloon (SCB) has since emerged as the next-generation drug-coated balloon for the treatment of vascular stenosis. Theoretically, the application of sirolimus at the graft-vein junction after successful balloon angioplasty would retard neointimal hyperplasia and improve AVG patency. The present prospective, pilot study was designed to test the feasibility of using SCB to treat graft vein junction of thrombosed AVG following successful pharmacomechanical thrombectomy. The 3- and 6-month results of the study have been published previously. ⁹ This report aims to provide the 1-year results and additional subset analysis.

Material and methods

This is a 1-year follow-up of a single-center, prospective, single-arm study that was conducted from 2018 to 2019 in 20 patients who presented to a tertiary institution with thrombosed AVG (ClinicalTrials.gov Identifier: NCT03666208). The study was conducted according to the Declaration of Helsinki and approved by the institutional ethics committee (CIRB number:2018/2233). All patients provided informed consent. Detailed methodology of the original trial results is available in previous works.^9,10 In short, the recruited patients received SCB angioplasty at the graft-vein junction following successful endovascular thrombectomy of a thrombosed AVG and adequate pre-dilatation of the lesion (defined as <30% residual stenosis). The sirolimus-coated balloon used in this study (MagicTouch, Concept Medical, India) is coated with sirolimus of 1.27 µg/mm² on the balloon’s surface. The primary endpoint was access circuit primary patency rates at 3 months. There were three deaths, one AVG revision, and one AVG explantation during the 12 months. Three participants have died with patent AVG at day-88, 319, and day 341 post-intervention. The revision was due to bleeding pseudoaneurysm in the cannulation zone, while the cause of AVG explantation was due to infection.

The outcome of AVGs and participants of the remaining 15 subjects at 1-year following the index procedure were obtained from electronic medical records. All interventions performed on the index AVG were recorded up to 12 months. The post-intervention primary, assisted primary, and secondary patency rates were defined according to the Society of Interventional Radiology recommendations. ¹¹

Statistical analysis

Continuous variables were summarized as mean ± standard deviations for normally distributed variables or median with interquartile ranges (IQRs 25th percentile, 75th percentile) for non-normally distributed variables, while categorical variables were reported using frequency counts and percentages. Kaplan-Meier survival analyses were used to estimate the mean primary, assisted-primary, and secondary patency rates of the AVGs. The data analyses were performed with STATA (StataCorp 2019, Stata Statistical Software Release 16, StataCorp LLC, College Station, TX) and SPSS version 23 (IBM Corp, Armonk, NY).

Results

Table 1 presents the patient demographics and characteristics. The study population had a mean age of 68 ± 9 years and was predominantly female (73.3%). The etiologies of end-stage renal disease were diabetes mellitus (40%), chronic glomerulonephritis (33.3%), hypertension (20%), and polycystic kidney disease (6.7%). The majority of the AVG were straight in configuration (80%), located in the upper arm (86.7%), with 60% being de novo lesions. The median vintage of AVG was 18 months.

Table 1.

Patient demographics and characteristics.

Demographics or characteristics	n (%) or mean (standard deviation)
Age, years	68 (9)
Gender
Female	11 (73.3)
Vintage of dialysis, months *	31 (12, 127)
Etiology of end-stage renal disease
Diabetes	6 (40)
Hypertension	3 (20)
Chronic glomerulonephritis	5 (33.3)
Polycystic kidney disease	1 (6.7)
Side of AVG
Right	4 (26.7)
Site of AVG
Upper arm	13 (86.7)
Forearm	2 (13.3)
Configuration
Loop	3 (20)
Straight	12 (80)
Anastomosis
Brachioaxillary	11 (73.3)
Brachiobasillic	2 (13.3)
Brachiobrachial	2 (13.3)
Vintage of AVG, months *	18 (13, 29)
De novo lesions	9 (60)
Number of interventions last 12 months
0	11 (73.3)
1–2	3 (20)
3–4	0
⩾5	1 (6.7)

AVG: arteriovenous graft.

^*Results expressed as median (25th, 75th percentile).

The procedural details and outcomes of the AVGs are summarized in Table 2. Cutting balloons were needed to adequately dilate the graft vein junction in four (26.7%) patients. The access circuit primary patency rate at 12 months was 40%, while the assisted primary and secondary patency rates at 12 months were 46.7% and 73.3%, respectively. A total of 16 re-interventions (4 angioplasties, 12 thrombectomies) were performed in 9 patients over the 12 months. Four AVGs were abandoned, three for recurrent thrombosis in short intervals. One patient was deemed medically unfit to undergo further thrombectomy procedure of the AVG; the patient continued dialysis via long-term tunneled dialysis catheter.

Table 2.

The characteristics of AVGs and their outcomes.

Subject	De novo lesions	Number of interventions since AVG creation	AVG diameter, mm	Treatment of graft vein junction			Outcome at 12 months
				Cutting balloon	Diameter of SCB, mm	Largest pre-dilatation balloon, mm	Stenosis episodes	Thrombosis episodes	Abandoned
1	Yes	0	6	Yes	8	7	0	1	No
2	No	5	6	No	8	9	0	0	No
3	No	5	7	No	7	7	1	1	Yes
4	Yes	0	6	Yes	7	7	0	0	No
5	No	3	6	No	7	7	0	3	No
6	Yes	0	6	Yes	7	7	2	0	No
7	No	2	6	No	7	7	1	2	Yes
8	Yes	0	6	No	7	7	0	2	Yes
9	No	1	6	No	7	7	0	1	No
10	Yes	0	6	No	8	7	0	0	No
11	Yes	0	6	No	8	6	0	0	No
12	Yes	0	7	No	8	7	0	0	No
13	No	8	6	No	8	7	0	1	No
14	Yes	0	7	Yes	7	7	0	3	Yes
15	Yes	0	7	No	7	7	0	0	No

Using Kaplan-Meier analysis, the mean estimated post-intervention access circuit primary patency was 230 (95% CI: 162–300) days, while access circuit assisted primary patency was 253 (95% CI: 187–320) days, and access circuit secondary patency was 292 (95% CI: 230–356) days. Sub-group analysis did not show a significant difference in the mean estimated primary patency between AVG with de novo and recurrent stenosis (245 days, 95% CI: 151–339 vs 210 days, 95% CI: 113–307; p = 0.29) (Figure 1).

Figure 1.

Kaplan-Meier survival estimates of access circuit primary patency of AVG with de novo versus recurrent lesions.

Discussion

The expanded polytetrafluoroethylene (ePTFE) AVG has enabled the creation of arteriovenous access since 1970s in patients who would otherwise not be suitable due to poor native vasculature. ¹² It is known that the primary patency rate of AVG was only 58% at 6 months and 33% at 18 months.^13,14 While every effort has been made to preserve AVG patency by endovascular intervention, post-intervention patency rates of successfully salvaged AVG have been far from satisfactory. Endovascular thrombectomy combining thrombectomy or thrombolysis and angioplasty of the underlying stenosis have been an effective treatment to restore the function of thrombosed AVG, with high technical and clinical success rates. ¹⁵ However, the patency outcomes following endovascular thrombectomy is less than ideal because the underlying cellular proliferation involved in the pathogenesis of vascular stenosis is left unaddressed with angioplasty alone. In addition, angioplasty may induce further endothelial injury, causing accelerated neointimal hyperplasia and thus restenosis and compromised blood flow. ¹⁶

As a second-generation drug-coated balloon, SCB has a well-proven safety and efficacy profile in treating coronary artery stenosis. Sirolimus is a cytostatic agent with a wider therapeutic range and safety margin than paclitaxel. Initially used as an immunosuppressive agent, sirolimus also exhibits both anti-inflammatory and anti-angiogenic activities which may be beneficial in reducing the inflammatory response from endothelial injury during angioplasty. ¹⁷ The dose delivered using angioplasty balloon is also much smaller compared to the systemic dose used in immunosuppression, hence the risk of toxicity is low.

Previous results have shown that the SCB may serve as a feasible option in maintaining the short-term patency of AVG following thrombectomy. ⁹ The extended follow-up at 1-year demonstrated sustained intervention-free survival of the AVGs following SCB angioplasty. The 12-month primary patency rate of 40% was higher than the expected post-thrombectomy patency 31% from the literature. ¹¹ Stent-grafts have been shown to effectively improve post-intervention AVG patency in several RCTs.^{18 –20} The current study suggested that post-intervention primary patency following application of SCB may be comparable with the average reported patency of 36% following stent-graft placement at 12 months. ¹¹ Unlike SCB angioplasty which entails no more than conventional percutaneous transluminal angioplasty, there is a learning curve involved in efficient stent deployment, resulting in increased risks of complications, including stent malposition and migration. In addition, stent-graft placement will invariably impede future surgical options for the ipsilateral arm while SCB does not leave a metallic scaffold behind. Hence, it could be an effective alternative to stent-grafts.

The study results also showed that SCB worked equally well in both de novo and recurrent stenosis of venous graft anastomosis. It is known that de novo lesions in arteriovenous fistula usually achieve the best patency rates with the first angioplasty. Some studies on paclitaxel drug-coated balloons (DCB) have suggested that older arteriovenous fistulas (AVF) with recurrent stenosis benefit most from its use.^{4 –6} Although studies that included de novo lesions in AVF suggested that DCB worked in these lesions, the effect was smaller when compared to restenotic lesions. ⁷ This is likely because stenosis in AVF may not always be due to neointimal hyperplasia. AVG stenosis, however has slightly different pathogenesis. Stenosis at the graft-vein junction is almost invariably due to vascular neointimal hyperplasia. ²¹ Turbulent flow caused by hemodynamic stress that activates smooth muscle cells and endothelial cells was thought to be an important initiating event. ²¹ The neointimal hyperplasia could be made worse by the ePTFE material that functions as a foreign body, attracting activated macrophages that produce cytokines to cause cell proliferation and migration. ⁵ Given this continuous and aggressive pathogenic process, it may be reasonable to use SCB even for de novo AVG stenosis, especially in severe stenosis that results in thrombosis.

It is acknowledged that the study has several limitations. The small sample size limits its generalizability and is also underpowered to detect a patency difference in AVG with de novo versus recurrent stenosis following SCB treatment. In addition, there is an absence of a control group to directly compare the efficacy of SCB over conventional plain balloon angioplasty to prolong the patency of AVG. Nevertheless, within the limits of this study, the findings suggest that SCB may help maintain the mid and long-term patency rates of thrombosed AVG and may be an attractive alternative to the stent-graft. However, the study results will need to be verified with a randomized control trial. Future trial designs could also include a non-inferiority study comparing SCB to stent-graft with long-term follow-up, which would help establish the role of SCB in the treatment of thrombosed AVG.