Initial and long-term outcomes of Percutaneous Transluminal Angioplasty for central venous stenosis or occlusion in chronic hemodialysis patients: Analysis of 363 lesions in single center

Citation: Horita Y, Namura M, Ikeda M, Terai H, Kimura R, et al. (2021) Initial and long-term outcomes of Percutaneous Transluminal Angioplasty for central venous stenosis or occlusion in chronic hemodialysis patients: Analysis of 363 lesions in single center. Arch Clin Nephrol 7(1): 009-017. DOI: https://dx.doi.org/10.17352/acn.000051 https://dx.doi.org/10.17352/acn DOI: 2581-3870 ISSN: C L IN IC A L G R O U P


Introduction
Advances in clinical care have helped improve the survival of patients with end-stage renal disease leading to an increased demand for hemodialysis service. Maintenance of the patency of vascular access is a quintessential element of long-term care of patients undergoing hemodialysis. Patients on chronic hemodialysis often develop central venous stenosis or occlusion in the hemodialysis access limbs. These central venous lesions are usually associated with venous shear stress, trauma caused by previous central venous cannulation and pacemakerleads, and anatomical compression by the clavicle and fi rst rib [1][2][3][4][5][6][7]. In particular, lesions of the left brachiocephalic vein (innominate vein) in the hemodialysis access limb develop both due to organic stenosis as well as compressive stenosis (caused by compression between the sternum and the right brachiocephalic artery) [8,9].

Abstracts
The purpose of this study was to evaluate the initial and long-term outcomes of Percutaneous Transluminal Angioplasty (PTA) for central venous stenosis or occlusion in chronic hemodialysis patients. A total of 363 central venous lesions (277 stenosis and 86 occluded lesions) of 146 patients were enrolled and analyzed retrospectively; these included 130 de novo lesions and 233 restenosis lesions. The procedural success rate in our cohort was 97.0% (352/363 lesions); success rate for stenosis was signifi cantly higher than that for occluded lesions (99.3% vs. 89.5%, P < 0.001). Complications during PTA procedures occurred in nine lesions (2.5%); however, there were no serious complications. A total of 120 stents were implanted for 111 lesions; the rate of stent placement for occluded lesions was signifi cantly higher than that for stenosis lesions (68.8% vs. 18.4%, P < 0.001).
The primary patency of de novo lesions was signifi cantly higher than that of restenosis lesions (P = 0.0004); the assisted patency of de novo lesions at 12 and 36 months were 94.4% and 89.0%, respectively. There was no signifi cant difference of primary patency between balloon angioplasty and stent placement for de novo lesions, however, primary patency after stent placement for restenosis lesions was signifi cantly higher than that after balloon angioplasty (P = 0.001).
The Kidney Disease Outcomes Quality Initiative (KDOQI) clinical practice guidelines for vascular access and the guidelines of the Japanese society for dialysis therapy for vascular access construction and repair in chronic hemodialysis patients recommend percutaneous transluminal angioplasty (PTA) for treatment of central venous lesions as well as for more distal arteriovenous fi stulas [10][11][12]. Standard PTA of central venous lesions entails angioplasty with conventional balloon; in addition, stenting is required if balloon angioplasty alone fails to achieve suffi cient dilatation of the lesion [5][6][7][8][9][10][11][12]. In general, central venous lesions refer to lesions in the cephalic arch vein, subclavian vein, brachiocephalic vein and the iliac vein. The cephalic arch vein refers to the arched portion of the cephalic vein in the axilla before its concourse to subclavian vein.
The purpose of this retrospective study was to evaluate the initial and long-term outcomes of PTA and stent placement for central venous lesions in patients undergoing chronic hemodialysis.

Patients
We enrolled all patients who were referred to our hospital from other hemodialysis clinics or hospitals between 2005 and 2019 because of hemodialysis failure with prolonged posthemodialysis bleeding, elevated venous needle pressure, or swelling of the ipsilateral arm or neck. PTA was performed for central venous lesions after confi rmation of diagnosis using ultrasound, angiography or enhanced Multi-Detector Row Computed Tomography (MDCT) [6,9,11]. Data pertaining to a total of 146 patients with 363 lesions were included in this retrospective analysis. The lesions were categorized into those treated by balloon angioplasty (Balloon group) and stent placement (Stent group) and sub-group analysis was performed disaggregated by de novo and restenosis lesions.
Acute occlusion represented lesions that were occluded within 24 hours, and subacute occlusion represented lesions that were occluded within 1 week from onset of central venous occlusion. The primary patency of central veins was defi ned as the interval between the date of the successful initial procedure and the date of the fi rst repeat intervention or restenosis. The restenosis was defi ned as signifi cant (% diameter stenosis, %DS>50%) stenosis of central vein by the evaluation using ultrasound, enhanced MDCT, or angiography. The assisted patency of central veins was defi ned as the patent central vein that underwent further intervention to improve patency. End points for functional access status included (a) placement of a new access site; (b) abandonment of the access site; (c) ligation of the access sit; and (d) placement of a dialysis catheter.

Venography
Central venography was performed to identify the location and degree of stenosis or occlusion. After puncture of the venous limb of the Arteriovenous Fistula (AVF) or arteriovenous graft (AVG) with 6F or 7F-sheath introducer, postero-anterior and oblique projection images were obtained after manual injection of 10-20 mL of contrast medium. On pre-and post-procedural venograms, the degree of stenosis was measured by comparing with the adjacent normal vein. Reference diameter was defi ned as the average of diameter in proximal and distal normal vein of target lesion. The measurements were performed using the measurement software equipped in the angiography unit (Trinias, SHIMAZU, Japan).

Percutaneous Transluminal Angioplasty (PTA)
PTA was indicated for patients who presented with both clinical symptoms and signifi cant (%DS>50%) stenosis of central vein. The procedures were performed under local anesthesia (1% lidocaine) and 6F or 7F-sheath introducer was inserted into AVF, AVG or femoral vein. After inserting the introducer, the guide wire was negotiated to cross the stenotic or occluded segment of the affected central vein. The 4, 5F-Judkins-Right or Multi-Purpose catheter and 6F-guiding catheter (Mach-1, 55 or 100 cm, Boston Scientifi c, Boston, USA) were sometimes used for Chronic Total Occluded (CTO) lesions which were diffi cult to traverse with the guide wire and required strong back-up. The intravascular ultrasound was used to confi rm the guide wire crossed into true lumen for CTO lesion. Hard-tip guide wire such as 12g Treasure guide wire (ASAHI INTECC, Seto, Aichi, Japan), 15 or 30g Chievalier guide wire (NIPRO, Osaka, Japan) or 40g Astato guide wire (ASAHI INTECC, Seto, Aichi, Japan) were used for tight CTO lesions. For some lesions, a loop snare was placed at the opposite end of the occluded lesion through the femoral vein (bilateral approach), and the pull-though technique for extraction of the guide wire performed to facilitate the passage of the balloon through the tight occluded segment of central venous lesion. All patients received an intravenous injection of heparin (5,000 units) immediately before PTA.
We inserted the balloon (diameter: 5-12 mm) through the introducer and infl ated the balloon in the stenotic or occluded segment with 6-30 atm. for 2 or 3 times. Stenting was performed if balloon angioplasty alone failed to achieve suffi cient dilatation of the lesion. Procedural success was defi ned as post-PTA %DS<30% without disturbance of access fl ow.

Stent placement
The indications for ballooning and stent placement for central venous lesions are as follows [8,9] 1. Balloon angioplasty is indicated for primary stenosis and short CTO lesions.
2. If balloon angioplasty achieves insuffi cient dilation or leads to dissection or acute occlusion of the affected vein, bail-out procedure by stent placement is performed.
3. Elective stent placement is indicated in case of frequent restenosis in the short-term after balloon angioplasty and for long CTO lesions, since it is expected to ensure sustained patency of the vein. between the sternum and the right brachiocephalic artery are impossible to dilate by balloon angioplasty alone and require stent placement. Adequate dilatation of this compressive lesion is diffi cult to achieve with the self-expandable stent. Therefore, balloon-expandable stent is chosen for this compressive lesion [8,9]. Four kinds of self-expandable stents, i.e., SMART or SMART control stent (Cordis, Cardinal Health, CA, USA), Luminnex stent (Bard, BD, Franklin Lakes, USA), Epic stent (Boston Scientifi c, Boston, USA), and two kinds of balloonexpandable stents, i.e., Palmaz stent (Cordis, Cardinal Health, CA, USA) and Express stent (Boston Scientifi c, Boston, USA) were used. All patients who underwent stent implantation were administered aspirin 100 mg/day for at least 1 month after PTA.

Exclusion criteria
Patients with central venous lesions who presented with excessive blood access fl ow (>1,500-2,000 mL/min.) and impaired cardiac function (ejection fraction<30%-35%) were excluded because volume overload immediately after PTA for central venous lesions may induce cardiac failure. Treatment methods appropriate to these conditions should be chosen for surgical treatment of excessive blood access fl ow [11][12][13][14]. In our previous study, we found that the volume overload occurring immediately-after PTA for central venous lesions in patients with normal cardiac function manifested as increased diameter of right atrium and right ventricle, however, these changes were transient and completely recovered on the subsequent day. Thus, PTA can be safely performed for central venous lesions in patients with preserved cardiac function [14]. Lesions with reference diameter >14 mm were excluded because of the lack of availability of stents that are large enough to treat such compressive or stenosis lesions in Japan.
Patients allergic to iodine containing contrast media were excluded from this study; lesions of these patients were treated under guidance of intra-vascular ultrasound or carbon dioxide angiography.

Monitoring of patients
All patients were followed-up at the referring clinic or hospital and were regularly monitored during hemodialysis by nephrologist and the hemodialysis team. Some patients were followed-up at our outpatient clinic to assess the condition of access by ultrasound. In case of any symptoms of restenosis or re-occlusion of the central veins and/or increase in venous needle pressure, the patients were again referred to our hospital and were evaluated using ultrasound, enhanced MDCT or angiography.

Statistical analysis
Data are expressed as mean ±standard deviation. Preand post-PTA comparison of the pressure at the introducertip was performed using the Wilcoxon signed ranks test. The quantitative vascular parameters were compared by Kruskal-Wallis test. Categorical data (such as procedural success) are presented as frequency (percentage) and between-group differences assessed using the Chi-squared test (or Fisher test when necessary). The primary and assisted patency rates of the treated lesions were assessed using Kaplan-Meier analysis and between-group differences assessed using the log-rank tests. P values less than 0.05 were considered indicative of statistical signifi cance.

Left brachiocephalic vein
The left brachiocephalic vein is longer than the right vein, and is sometimes compressed between the sternum and the right brachiocephalic artery. Self-expandable stent for organic lesions. Balloon-expandable stent for compressive lesions.

Right brachiocephalic vein
This lesion is organic stenosis. Both of self-expandable and balloon expandable stents are indicated.   (Table 2) PTA was unsuccessful in 11 lesions (3.0%). The two compressive stenotic lesions of the left brachiocephalic vein with excessive blood fl ow (>1500 mL/min.) were dilated by balloon; however, prolonged acute-recoil occurred after balloon angioplasty. Therefore, treatment of these lesions was modifi ed to surgical treatment for excessive blood fl ow.  large dissection in the left cephalic arch vein was bailed out by stent placement. Stent dislocation occurred in three lesions. An instance of stent slippage to the right brachiocephalic vein after stenting for right subclavian vein was recovered by additional stent placement. A balloon-expandable stent implanted in the right brachiocephalic vein dislocated to the distal part of the right pulmonary artery after insertion of the cuffed catheter immediately after the implantation of this balloon-expandable stent. This migrated stent was monitored without additional treatment. Another instance was stent dislocation on delivery of 8-mm balloon-expandable stent to the compressive stenosis of the left brachiocephalic vein with an internal diameter of 12 mm. The 12-mm balloon was dilated at the bilateral stent-edge to avoid the stent dislocation after 8-mm balloon-expandable stent implantation. However, this balloon-expandable stent slipped and dislocated to the right atrium and was subsequently implanted in the right iliac vein along the long guide wire inserted to the right femoral vein from left access arm.

Initial outcomes and complications
The overall procedural success rate in this cohort was 97.0% (352/363 lesions); the success rate for stenosis lesions (99.3%, 275/277 lesions) was signifi cantly higher than that for occluded lesions (89.5%, 77/86 lesions) (P < 0.001). Among the occluded lesions, the success rate for acute or subacute occluded lesions was 92.3% (12/13 lesions) while that for CTO lesions was 89.0% (65/73 lesions). The success rate of CTO lesions with pacemaker-leads was 66.7% (6/9 lesions). Table  3) A total of 120 stents were implanted for 111 central venous lesions; the rate of stent placement for occluded lesions (69.8%, 60/86 lesions) was signifi cantly higher than that for stenosis lesions (18.4%, 51/277 lesions) (P <0.001). We implanted the 2 stents for 7 lesions and 3 stents for 1 lesion. The rate of implantation of self-expandable stents in the subclavian vein was signifi cantly higher than that in brachiocephalic veins (P = 0.004). The three lesions in the subclavian vein with short stenosis were treated by short balloon-expandable stents.
There were signifi cant differences between the de novo and restenosis lesions with respect to patency rates (P = 0.0004) ( Figure 2). The assisted patency rates at 6, 12, 24, and 36 months were 97.6%, 94.4%, 91.6%, and 89.0%, respectively    We compared the primary patency after balloon angioplasty and stent placement for de novo and restenosis lesions ( Figure   4), and analyzed the lesion location, morphology, and preand post-PTA quantitative vascular parameters in each group (Table 4). Six lesions that were treated with drug-coated balloons were excluded from this analysis. There was no signifi cant difference with respect to primary patency after balloon angioplasty and stent placement for de novo lesions.
For restenosis lesions, the primary patency rate after balloon angioplasty was signifi cantly lower than that after stent placement (P = 0.0010). The primary patency rate after balloon angioplasty for restenosis lesions was signifi cantly lower than that after balloon angioplasty for de novo lesions (P = 0.0025).
The patency rate after stent placement for restenosis lesions was similar to that for de novo lesions (Figure 4). The rates of occluded lesions in stent groups were higher than those in balloon groups for de novo and restenosis lesions (P < 0.001).
The pre-PTA MLDs of de novo and restenosis lesions treated with stents were 0.99±1.58 mm and 1.12±1.31mm, respectively, and these MLDs were signifi cantly smaller than those of lesions treated with balloon angioplasty alone (P < 0.001). Post-PTA, the acute gains in de novo and restenosis lesions treated with stent (6.23±2.22 mm and 5.74±2.19 mm, respectively) were signifi cantly higher than those in lesions treated with balloon angioplasty alone (P < 0.001) ( Table 4).

Discussion
PTA of central venous lesions is recommended for chronic hemodialysis as the fi rst treatment also for the treatment of more distal arteriovenous fi stulas [8][9][10][11][12]. was found to be an independent predictor of venous occlusion [18]. PTA with paclitaxel-coated balloons was performed for six restenosis or re-occluded central venous lesions and the patency rate at 24 months was 83.3% (5/6 lesions). However, the usage of paclitaxel-coated balloon for central venous lesions was not approved in Japan. We had earlier reported the effi cacy of paclitaxel-coated balloon for recurrent stenosis of distal vascular access in chronic hemodialysis patients [19,20].
In addition, several reports have documented better results with paclitaxel-coated balloon compared with conventional balloon angioplasty for central venous lesions [21,22]. We vein. This lesion is strongly compressed between the sternum and the right brachiocephalic artery, we believe that pre-PTA enhanced MDCT is very useful to diagnose this compressive lesion. Use of balloon-expandable stent indicated for this compressive lesion because the radial force of the balloonexpandable stent is stronger than that of the self-expandable stent [8,9]. Kaneko [27,28]. We compared the primary patency between 196 lesions treated by balloon angioplasty and 111 lesions treated by stent placement, disaggregated by de novo and restenosis lesions. The number of lesions included in this study is larger than those in previous studies [15][16][17]24,[27][28][29]. There was no signifi cant difference of primary patency after balloon angioplasty and stent placement for de novo lesions. For restenosis lesions, the primary patency rate after stent placement was signifi cantly higher than that after balloon angioplasty (P = 0.0010), and this result was affected by the larger acute gain in the stent group. Patency rate after balloon angioplasty for restenosis lesion was signifi cantly lower than that after balloon angioplasty for de novo lesions (P = 0.0025), however, each parameter (lesion location, lesion morphology, and quantitative vascular parameters) of de novo and restenosis lesions treated by balloon angioplasty were similar with the exception of lesion length. The cause of lower primary patency rate after balloon angioplasty for restenosis lesions could not be clarifi ed in based on our data. The patency rate after stent placement for de novo lesions was similar to that for restenosis lesions. Our fi ndings indicate that stent placement for restenosis lesions can help improve the patency period.

Limitations
Some limitations of our study should be considered while interpreting the fi ndings. This was a retrospective singlecenter study, which may have introduced an element of bias. In addition, 39 lesions out of 352 successful PTA lesions could not be followed-up as all the enrolled patients were referred from other hemodialysis clinics and hospitals and 16 patients died of underlying diseases which were not related to the procedures.

Conclusions
PTA for central venous stenosis or occlusion are important procedures to maintain the hemodialysis conditions. This therapy was safe procedure with low rate of technical failure. The primary patency rate of PTA for central venous lesions were insuffi cient, however the assisted patency rate by the repeated intervention were high. The patency rate after stent placement for restenosis lesion was signifi cantly higher than that after balloon angioplasty. Stent placement helped improve the patency period of restenosis lesions.

Ethical statement
This retrospective study was approved by the Ethical Committee of Kanazawa Cardiovascular Hospital.

A note
The interventional therapy for central venous lesions was safe procedure with low rate of technical failure. The primary patency rate of PTA for de novo lesions was signifi cantly higher than that for restenosis lesions, and the patency rate after stent placement for restenosis lesion was signifi cantly higher than that after balloon angioplasty.