Effect of Drug-Coated Balloons on Recurrent Venous Hypertension Due to Central Venous Lesions in Patients with Hemodialysis Vascular Access

Abstract

Objectives

This study aimed to evaluate the efficacy of drug-coated balloons (DCBs) for symptomatic and recurrent central venous lesions (including in-stent restenosis and occlusion) in patients undergoing hemodialysis.

Methods

Twenty-seven patients treated with DCBs for symptomatic central venous lesions between June 2021 and September 2025 were included. We examined the overall efficacy of DCBs, the outcomes of repeated use, and factors affecting patency.

Results

Target lesion patency at 6 months was significantly improved following initial DCB use (77% vs. 30% prior to DCB use, p <0.0001). The efficacy of repeated DCB use was not significantly different between the first and second applications. In a subgroup analysis, there was a trend toward better patency rates in patients treated with peripheral cutting balloons (PCBs) for pre-dilatation compared to those treated with conventional balloons, although the difference was not statistically significant.

Conclusions

DCBs demonstrated satisfactory efficacy for recurrent central venous lesions, which was maintained with repeated use. Additionally, pre-dilatation using PCBs showed a trend toward improved patency rates. (This is a translation of Jpn J Vasc Surg 2026; 35: 23–28.)

Introduction

Central venous lesions ipsilateral to the arteriovenous (AV) access have been reported in 17%–26% of patients on hemodialysis with access failure.¹⁾ Upper-extremity venous stasis symptoms caused by central venous lesions can severely impair patients’ quality of life. Furthermore, these lesions can hamper the continuation of hemodialysis, as difficult or unsuccessful cannulation can lead to hematoma formation or infection.

Although balloon angioplasty and bare metal stent (BMS) placement are effective treatment options for such lesions that achieve acceptable patency rates,^2,3) restenosis and in-stent restenosis remain major long-term clinical challenges. In recent years, paclitaxel-based drug-coated balloons (DCBs) have emerged as novel modalities for preventing restenosis, particularly early restenosis, and their efficacy has been increasingly established.^4–6) Following approval for health insurance reimbursement in 2021, DCBs have been widely adopted in Japan. However, their efficacy for central venous lesions remains controversial,^5,7) and to date, no studies from Japan have specifically assessed their use for this indication. Building on our previous report on the outcomes of BMS placement for central venous lesions,³⁾ the present study evaluated the efficacy of DCBs for symptomatic, recurrent central venous lesions in patients undergoing hemodialysis, including the effects of repeated DCB treatment and the factors associated with patency.

Materials and Methods

Study design and patient population

Between June 2021 and September 2025, 79 DCB procedures were performed in 30 patients for symptomatic central venous lesions, including in-stent restenosis and chronic occlusion. Twenty-seven patients who underwent initial DCB treatment for restenosis were included in this study (Table 1). All patients presented with symptomatic venous hypertension, including upper-extremity edema and superficial venous engorgement. Patients with asymptomatic central venous lesions or isolated elevation of venous pressure without clinical symptoms were excluded. Central veins were defined as the subclavian vein and more proximal veins. Lesions involving the axillary vein or cephalic arch were excluded.

Table 1. Baseline characteristics of the patients (N = 27).

	No. (%)
Age, years, mean ± SD	69.7 ± 11.1
Male	18 (67)
Female	9 (33)
HD duration (years), mean ± SD	13.4 ± 9.9
Comorbidities
Hypertension	20 (74)
Dyslipidemia	9 (33)
Diabetes mellitus	9 (33)
Cerebrovascular disease	11 (41)
Coronary artery disease	8 (30)
Peripheral arterial disease	3 (11)
Congestive heart failure	3 (11)
Hepatic disease	3 (11)
Smoker	9 (33)
Obesity (BMI >25)	4 (15)
Cause of HD
Chronic glomerulonephritis	7 (26)
Diabetic nephropathy	6 (22)
Hypertensive nephropathy	4 (15)
Polycystic kidney disease	2 (7)
Others	5 (19)
Unknown	3 (11)

SD, standard deviation

Device indications and off-label use

The instructions for use of DCBs state that the efficacy and safety of treatment proximal to the axillary–subclavian vein junction have not been established. Therefore, treatment in this region requires careful consideration of the risks and benefits, but it is not considered strictly off-label. Although DCB treatment for in-stent restenosis and chronic occlusive lesions represents off-label use, previous research has demonstrated its efficacy in lower-extremity arterial disease.⁸⁾ The chronic occlusions included in this study were identified on the basis of the clinical history and successful guidewire crossing.

Endovascular procedure

Percutaneous transluminal angioplasty (PTA) was performed through a 7-French sheath inserted into the forearm or upper-arm vascular access. Diagnostic angiography was used to identify the target lesion, followed by lesion crossing with a 0.018-inch guidewire (Fuga, Boston Scientific, Marlborough, MA, USA). For chronic occlusions, a guidewire with a tip load of ≥30 g was used as necessary. Severe stenoses were subjected to pre-dilatation using a Peripheral Cutting Balloon (PCB; Boston Scientific), either alone or in combination with a conventional balloon.

The IN.PACT AV drug-coated balloon (Medtronic Vascular, Santa Clara, CA, USA), currently the only DCB reimbursed for vascular-access interventions in Japan, was selected according to the reference vessel diameter or the diameter of a previously implanted BMS. Balloon inflation was maintained for 3 min. Final angiography confirmed satisfactory luminal expansion before sheath removal and manual hemostasis.

Post-procedural management and re-intervention strategy

No routine post-procedural antithrombotic therapy was administered. Patients underwent surveillance every 3 months. If vascular access occlusion or recurrent upper-extremity edema developed, angiography, thrombectomy, or repeated PTA was performed as clinically indicated. For repeated interventions, DCB treatment was continued when the patency interval after the previous procedure was prolonged compared with that of earlier treatments. If the patency interval was unchanged or shortened, then conventional balloon angioplasty or BMS placement was considered.

Outcome measures and data collection

Technical success was defined as residual stenosis <30%, whereas clinical success was defined as the successful completion of at least 1 hemodialysis session without procedure-related complications. The patients’ demographics, comorbidities, procedural characteristics, patency outcomes, and post-procedural complications were collected from our institutional database.

Definitions and statistical analysis

Patency was defined as described by Gray et al.⁹⁾ Primary patency was defined as the interval from the index intervention to the first re-intervention. Assisted primary patency was defined as the interval until thrombosis or re-intervention. Target lesion patency was defined as the time to repeat intervention for the treated central venous lesion.

Continuous variables were expressed as the mean ± standard deviation. Kaplan–Meier survival analysis was used to estimate the patency rates.

Ethical approval

Informed consent was obtained through the institutional surgical consent process, which informed patients that anonymized clinical data could be used for research while allowing them to opt out of this usage. The study protocol was approved by the Institutional Review Board (IRB) of Fukuoka City Hospital (Approval No. 242). Independent IRB approval (Approval No. 138) was also obtained for the off-label use of BMSs in the central venous lesions included in this study.

Results

Procedural outcomes

Procedural outcomes are summarized in Table 2. Among the 27 included patients, 23 presented with stenotic lesions, whereas 4 presented with chronic occlusive lesions. In-stent restenosis was identified in 7 patients, and it involved the Epic Self-Expanding Nitinol Stent System (Boston Scientific) in 6 patients and the E-Luminexx Vascular Stent (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) in 1 patient. Two patients presented with in-stent occlusion, both involving the Epic stent.

Table 2. Procedural details (N = 27).

	No. (%)
Procedure time (min), mean ± SD	27 ± 9.5
Blood loss (g), mean ± SD	10 ± 2.5
Central venous lesion
Stenosis	16 (59)
In-stent restenosis	7 (26)
Occlusion	2 (7)
In-stent occlusion	2 (7)
Observation time (days)	770 ± 490
Balloons used before DCB
PCB (6 mm) + conventional balloon	12 (44)
Conventional balloon	15 (56)
Pre-dilatation
PCB (6 mm)	14 (52)
PCB (6 mm) + conventional balloon	6 (22)
Conventional balloon	7 (26)
DCB size
10 × 40 mm	22 (81)
9 × 60 mm	2 (7)
8 × 60 mm	2 (7)
8 × 80 mm	1 (4)
Location
Subclavian vein	23 (85)
Innominate vein	4 (15)
Initial success (technical and clinical)	27 (100)

SD, standard deviation

The technical and clinical success rates were 100%, and all patients experienced improvement in their subjective symptoms following PTA. Representative clinical case images are presented in Fig. 1. One patient who developed severe elastic recoil following DCB application and required bailout BMS placement was excluded from the study.

Fig. 1. Representative case of DCB use for symptomatic central venous occlusion. Angiography shows right subclavian vein obstruction (A). After recanalization of the occluded lesion, the lesion was dilated with an 8 × 60-mm DCB (B). Completion angiography is shown in (C). The patient’s right upper extremity was swollen before treatment (D). After treatment, upper extremity swelling resolved (E).

Patient outcomes

The primary, assisted primary, target lesion, and secondary patency rates following initial DCB treatment are presented in Fig. 2. The 6-month target lesion patency rate increased significantly following DCB introduction, improving from 30% before DCB treatment to 77% after initial DCB application (p <0.0001; Fig. 3).

Fig. 2. Kaplan–Meier curve of primary patency after using DCB. The dotted line represents primary patency (PP), the solid line represents target lesion patency (TLP), the dashed line represents assisted primary patency (APP), and the long-dashed line represents secondary patency (SP).

Fig. 3. Kaplan–Meier curves of target lesion patency before and after DCB use.

Repeated DCB treatment

During follow-up, 19 of the 27 patients required re-intervention. In 1 patient, a conventional balloon was selected because the initial DCB treatment was considered ineffective. However, following a further reduction in the patency interval, subsequent re-interventions were performed using DCBs. The remaining 18 patients underwent a second DCB implantation.

As presented in Fig. 4, the target lesion patency rate following the second DCB implantation was comparable to that observed after the initial DCB application, with no statistically significant difference between the 2 treatment episodes.

Fig. 4. Kaplan–Meier curves of target lesion patency after first and second use of DCB.

Subgroup analysis

Subgroup analysis demonstrated a trend toward improved target lesion patency rates in patients who underwent pre-dilatation with PCB compared with those treated without PCB pre-dilatation, although this difference did not reach statistical significance (6-month target lesion patency rate: 89% vs. 43%, p = 0.086; Fig. 5). No statistically significant differences in patency rates were observed according to the lesion occlusion status, prior stent placement, lesion location, or DCB diameter.

Fig. 5. Kaplan–Meier curves of target lesion patency in the group using PCBs for pre-dilatation versus the group using conventional balloon (CB).

Discussion

Although the efficacy of DCBs for failing AV fistulas is widely accepted, their role in the treatment of central venous lesions remains controversial. Most meta-analyses evaluating DCBs excluded central venous lesions from their analyses,^6,10,11) and the 2 meta-analyses that included subgroup analyses of these lesions reported conflicting conclusions.^5,7) Moreover, the total number of patients included in these studies was limited to 68 and 30 patients, respectively, making it difficult to draw definitive conclusions regarding the efficacy of DCBs in this setting. Although our study also included a relatively small cohort of 27 patients, target lesion patency rate improved significantly following DCB treatment compared with that in the pre-DCB period, suggesting that DCBs represent an effective therapeutic option for recurrent central venous lesions.

Regarding repeated DCB treatment, Kitrou et al.¹²⁾ reported no significant difference in patency rates between the first and second DCB applications in a cohort of 38 patients, including 8 patients with central venous lesions. Consistent with these findings, we observed comparable target lesion patency rates following the first and second DCB implantations, indicating that repeated DCB application might continue to provide a therapeutic benefit rather than reverting to conventional balloon angioplasty alone. Nevertheless, the possibility of selection bias should be acknowledged, as our treatment strategy reserved conventional balloon angioplasty or BMS placement for lesions considered unresponsive to the initial DCB treatment.

Successful DCB therapy depends on adequate lesion preparation before drug delivery. Because DCB use is contraindicated in the presence of substantial residual recoil (≥30% residual stenosis) after pre-dilatation, we preferentially used PCBs, which have been reported to reduce elastic recoil.¹³⁾ In our practice, PCBs were frequently used either alone or in combination with conventional balloons, although they were often undersized relative to the reference vessel diameter. In the present subgroup analysis, patients treated with PCB pre-dilatation demonstrated a higher target lesion patency rate than those treated with conventional balloons alone, although the difference did not reach statistical significance. PCBs facilitate vessel expansion at relatively low inflation pressures while minimizing vessel injury. In the coronary circulation, PCB pre-dilatation before stent implantation has been associated with improved long-term patency.¹⁴⁾ Similarly, a previous study reporting favorable outcomes of balloon angioplasty for central venous occlusions²⁾ suggested that PCB pre-dilatation might have contributed to these results. Furthermore, improved patency has also been reported following the combined use of PCBs and DCBs for noncentral AV fistula stenoses.¹⁵⁾ Although our findings did not reach statistical significance, they suggest that PCB pre-dilatation might enhance the effectiveness of DCB placement for central venous lesions. However, given the substantial cost associated with both PCBs and DCBs, further studies should also evaluate the cost-effectiveness of this treatment strategy.

Based on our findings, we propose a practical treatment algorithm for recurrent central venous lesions. For symptomatic stenotic or chronic lesions involving the access extremity, conventional balloon angioplasty should be considered as the initial treatment. If satisfactory luminal expansion is achieved, then no additional intervention is required. Conversely, lesions displaying marked elastic recoil or requiring particularly difficult lesion crossing might be better managed with immediate BMS placement and subsequent post-dilatation. Patients should then undergo routine surveillance, with repeat balloon angioplasty performed in cases of restenosis or in-stent restenosis. DCBs might be particularly valuable for lesions exhibiting early recurrent restenosis. If DCB treatment fails to meaningfully prolong patency, placement of an additional BMS can be considered. In our practice, we limit treatment to a maximum of 2 BMS implantations to minimize the luminal compromise associated with multiple overlapping stents. Beyond this stage, repeated balloon angioplasty, including DCB treatment when appropriate, remains the preferred strategy as long as it continues to provide clinically acceptable access function. Although covered stents might theoretically represent an alternative option, we generally avoided their use because of the potential risks of stent migration and occlusion of important venous side branches. Overall, the incorporation of DCBs into the treatment algorithm expands the therapeutic options available for managing recurrent central venous lesions, and this strategy might contribute to prolonged vascular access patency in selected patients.

Conclusion

DCB treatment was associated with improved target lesion patency compared with that achieved after conventional balloon angioplasty in patients with recurrent central venous lesions, and similar efficacy was observed with repeated DCB use. PCB pre-dilatation was also associated with a trend toward improved patency. Larger prospective studies are warranted to confirm these findings and further define the role of DCB-based treatment strategies for central venous lesions.

Declarations

Author contributions

Study conception: Daihiko Eguchi

Data collection: Daihiko Eguchi

Data analysis: Daihiko Eguchi

Investigation: Daihiko Eguchi

Manuscript preparation: Daihiko Eguchi

Critical review and revision: all authors

Final approval of the article: all authors

Accountability for all aspects of the work: all authors.

Disclosure statement

The authors declare no potential conflicts of interest with respect to the research, authorship, or publication of this article.