Beraprost sodium versus clopidogrel for preventing vascular thromboembolic events of arteriovenous fistula in uraemic patients: a retrospective study with a mean 3-year follow-up

Abstract

Objective

This study aimed to evaluate the efﬁcacy of beraprost sodium (BPS) or clopidogrel (CL) using vascular thromboembolic events (VTEs) of arteriovenous fistula as a primary endpoint in patients with end-stage renal disease (ESRD) undergoing arteriovenous fistula surgery.

Methods

We performed a multicentre, retrospective cohort study from August 2012 to August 2016. We studied patients with ESRD who underwent arteriovenous fistula surgery and received peroral administration of 40 µg BPS, three times per day, for 1 month, or 75 mg CL (initial dose of 300 mg), one time per day, for 1 month. The time to first on-study VTE was the primary endpoint.

Results

The BPS-treated cohort had a significantly delayed time to first VTE compared with the CL-treated cohort (hazard ratio 0.33, 95% confidence interval 0.18–0.56). An increased incidence of VTEs was detected in the 1-month follow-up, with rates of 2.4% and 8.7% for BPS and CL, respectively. This difference persisted over time, with rates of 8.0% and 18.1% at the final follow-up, respectively.

Conclusion

CL-treated patients with ESRD have a greater risk of VTEs compared with BPS-treated patients. CL-treated patients also tend to experience a VTE within the first month after cessation of oral administration.

Keywords

Beraprost sodium clopidogrel arteriovenous fistula vascular thromboembolic event uraemia end-stage renal disease

Introduction

Uraemia is a potent risk factor for inducing vascular thromboembolic events (VTEs). VTEs manifest as endovascular thrombosis following interventions for arteriovenous fistula, and only inadequate antithrombotic medications pose a greater risk. Uraemia independently increases the risk of thrombosis by four to eight fold, implicating renal failure-specific prothrombotic mediators.^1–4 The estimated incidence of end-stage renal disease (ESRD) in China in 2017 was 300,00 patients, of whom almost 60% eventually died, which is partially related to common delays in diagnosis. Intervention for arteriovenous fistula, which is regarded as a lifeline of patients with ESRD, establishes permanent vascular access for these patients.^5,6 In recent years, the use of artificial blood vessels and semi-permanent tube surgery in patients with ESRD has rapidly developed.^7,8 However, this method has short-term effects, increased complications, and is expensive. Therefore, there has been an increase in these patients’ mental and economic burden, which can have serious effects on their quality of life.^9,10 Previous studies have shown that the failure rate of intervention for arteriovenous fistula is 20% to 54%, mainly due to thrombosis, non-maturation, and mature obstruction of arteriovenous fistula.^11,12 The incidence of thrombus in arteriovenous fistula is more frequent within 30 days after surgery.⁷ Previous studies have provided proof-of-concept that vascular access dysfunction is a major cause of morbidity and hospitalization in patients with uraemia, and this can be attributed to venous stenosis caused by neointimal hyperplasia after vascular access surgery.¹¹ Thrombosis caused by stenosis of an internal fistula in the forearm is common in the 3- to 4-cm venous end of the internal fistula.^8,9 Although several studies have advocated a characteristic of pharmacological agents, including anti-platelet drugs in preventing thrombosis, these studies have provided inconsistent findings.¹³ Recently, pharmacological agents have been recommended to be used to decrease neointimal hyperplasia for improving vascular access patency in patients with uraemia on haemodialysis.^13,14 Among these agents, beraprost sodium (BPS), which is an oral compound synthesized as a prostacyclin (prostaglandin I₂) derivative, shows strong anti-platelet activity (reduction in platelet aggregation and blockade of platelet adhesion). BPS is used to maintain vascular access patency because it inhibits neointimal formation after vascular surgery.^15–17 However, whether BPS can maintain vascular access patency in patients with uraemia on haemodialysis is unknown.^16,18 Clopidogrel (CL), which is an antithrombotic medication, is successfully used in preventing arteriovenous fistula failure caused by thrombosis. However, patients with uraemia on CL have increased vascular risk factors.^19,20 Previous randomized, controlled trials that compared CL with a placebo showed that patients with uraemia on CL have reduced rates of arteriovenous fistula thrombosis.^11,21

To the best of our knowledge, direct comparison between BPS and CL has infrequently been described in the published literature.^8,11 Conventional antithrombotic medications have reduced effectiveness in patients with uraemia and thrombotic complications or the occurrence of VTEs. BPS and CL are promising options to reduce these uraemia-related thrombotic complications or occurrence of VTEs. However, several unanswered questions remain. This study aimed to evaluate the efﬁcacy of BPS or CL, with the time to the first VTE of arteriovenous fistula as the primary endpoint, in patients with uraemia who underwent arteriovenous fistula surgery in an Asian population. We ultimately hope to provide a reference for clinical treatment.

Materials and methods

Setting

We followed the reporting guidelines for observational studies in the STROBE guidelines for this retrospective, comparative study. The study was approved by the institutional ethics review boards (the First Affiliated Hospital of Sun Yat-sen University, Jinshan Hospital, Fudan University, and the First Affiliated Hospital of Chongqing Medical University). An exemption for informed consent was obtained from these institutional ethics review boards.

Study population

We retrospectively collected data of 367 patients with ESRD who were diagnosed as having ESRD in line with standard criteria.²¹ The patients were admitted to three hospitals from August 2012 to August 2016. The criteria that were used to decide on creation of arteriovenous fistula were based on previous literature.¹⁴ The type of drug treatment that a patient received depended on power analysis before the study. The inclusion criteria were as follows: age of 20 to 95 years; either sex; patients with ESRD requiring maintenance haemodialysis; access to the cephalic vein fistula via end-to-side anastomosis of the radial artery, which was performed according to acceptable surgical techniques as previously described;²² and a diameter of the fistula anastomosis of 7 to 10 mm. Further, the patients underwent dialysis three times/week for 3 to 4 hours using heparin or low-molecular-weight heparin anticoagulation. After dialysis, proper compression (no bleeding and palpable tremor) of the puncture point was performed for 15 to 20 minutes. Preoperative evaluation (physical with duplex scanning) was performed in all patients. Within 1 month before surgery, no anticoagulant or vasodilator was used in any patient. All of the patients underwent arteriovenous fistula surgery under general anaesthesia. The patients received peroral administration of 40 µg BPS (Beijing Ted Pharmaceutical Co., Ltd., China, drug specification: 20 µg) three times per day^16,18 for 1 month or 75 mg CL (initial dose of 300 mg; Hangzhou Sanofi Pharmaceutical Co., Ltd., Hangzhou, China, drug specification: 75 mg) once per day for 1 month.^19,20 BPS or CL was applied on the day before surgery. The main exclusion criteria were as follows: incomplete data; previous use of the study drug (BPS or CL) or any anticoagulant drug that affects coagulation function during 3 months before surgery; non-healed or planned surgery; combined severe infection; severe liver disease; chemotherapy for other diseases; cardio-cerebrovascular diseases; circulatory system diseases; malignant tumours; further deterioration of the kidneys during anticoagulation drug treatment or during the follow-up period; modification, discontinuation or interruption of BPS or CL treatment; lactating and pregnant patients; life expectancy < 1 year; co-occurring mental illness; and an American Society of Anesthesiologists score of IV or V.

Definitions of main descriptive variables

Duplex scanning of VTEs, which was conﬁrmed by a chief physician of the Ultrasound Department, was carried out within 1 week, every 2 weeks thereafter, and at the final follow-up. VTEs were deﬁned as an interruption of blood flow due to thrombus or anastomotic stenosis, which was confirmed by duplex scanning. An independent review committee that was masked to the study group adjudicated whether the VTEs were due to thrombus or anastomotic stenosis (or attributable to investigation of suspected VTEs) to compare the incidence of VTEs between groups. Primary patency, strong tremors, and a strong vascular murmur at the anastomotic location were defined as occurring from the interval from the time of access placement to any intervention that was required to maintain or re-establish patency. Patency was assessed by duplex scanning. If the internal diameter of the venous vessels was ≥5 mm, it could meet the requirements of dialysis. In contrast, if the internal diameter was < 5 mm, the fistula tended to be blocked. The success rate of the internal fistula = patency number/total number × 100%.

Statistical analysis

Continuous data are presented as the mean and standard deviation or as the median and quartiles. Comparisons between the two groups were performed using Pearson’s chi-square test (Fisher’s exact test was used when necessary) or the Mann–Whitney U test for continuous variables. The time to first on-study VTE of vascular access was compared using Kaplan–Meier analysis and the log-rank test. The analyses were based on the data up to the primary analysis cut-off. All P values were two sided and a P value < 0.05 was used as the threshold for statistical significance. IBM SPSS Statistics for Windows, version 24.0 (IBM Corp., Armonk, NY, USA) was used to conduct all analyses.

Results

Patients’ characteristics

After the exclusion criteria were applied, 252 cases of uraemia were identified among 367 consecutive patients with uraemia who underwent arteriovenous fistula surgery and received peroral administration of 40 µg BPS or 75 mg CL (initial dose of 300 mg) (Figure 1). These cases included 125 patients with a mean age of 43.7 ± 7.24 years who received peroral administration of 40 µg BPS. There were 127 patients with a mean age of 44.1 ± 8.11 years who received peroral administration of 75 mg CL (initial dose of 300 mg). The patients’ demographics are shown in Tables 1 and 2. There were no significant differences in the proportion of patients with medical diseases (i.e., hypertension, diabetes mellitus, or peripheral vascular diseases) between the groups after initiation of drug prevention (Table 1). There were also no significant differences in laboratory parameters between the groups (Table 2). The mean duration at the cut-off date of the primary analysis was 36.2 months (interquartile range: 36–38) and 36.8 months (interquartile range: 36–38) for the BPS- and CL-treated cohorts, respectively.

Figure 1.

Flow diagram showing methods for identifying studies to assess the efficacy of BPS or CL using VTEs of arteriovenous fistula as the primary endpoint. Patients with ESRD who underwent arteriovenous fistula surgery in an Asian population were studied. BPS: beraprost sodium; CL: clopidogrel; VTEs: vascular thromboembolic events; ESRD: end-stage renal disease.

Table 1.

Comparison of patients’ demographics between the groups.

Variable	BPS (n = 125)	CL (n = 127)	P value
Age (years)			0.70^a
20–39	37	35
40–59	47	48
60–79	26	28
80–95	15	16
Sex (n), M/F	64/61	65/62	1.00^b
Current smoker (n)	33	37	0.63^b
Systolic BP (mmHg)	153.12 ± 21.32	152.76 ± 25.63	0.26^c
Diastolic BP (mmHg)	90.17 ± 15.46	89.72 ± 16.84	0.10^c
Causes of end-stage renal disease (n)			0.82^a
Diabetes mellitus	34	36
Chronic glomerulonephritis	49	44
Hypertension	22	25
Others	20	22
Access flow rate (mL/minute)	962 ± 537	954 ± 671	0.29^c
eGFR (mL/minute/1.73 m²)	12.21 ± 2.71	12.18 ± 2.95	0.53^c
BUN (mg/dL)	41.81 ± 13.57	42.10 ± 12.16	0.13^c
Hb (g/dL)	9.91 ± 1.61	10.15 ± 2.18	0.29^c
Creatinine (mg/dL)	11.17 ± 2.35	10.93 ± 2.69	0.41^c
Dialysis duration with respective access (months)	14.61 ± 6.92	14.57 ± 7.01	0.37^c
Urea kinetics, single-pool Kt/V^#	1.82 ± 0.53	1.83 ± 0.49	0.35^c
Type of dialysis (n)			0.66^b
Haemodialysis	123	124
Peritoneal dialysis	2	3
Haemodialysis duration (months)	31.46 ± 7.36	31.24 ± 6.53	0.14^c
ASA level			0.80^a
1	45	44
2	49	50
3	31	33
Ambulatory status (n)			0.45^a
Normal walking	113	111
Walking with assistive devices	12	16
Completely restricted walking	0	0
Diabetes mellitus
Duration (years)	21 (13–25)	20 (12–26)	0.11^c
Type 2 (n)	19	21	0.77^b
Insulin use (n)	11	14	0.57^b
HMG-CoA reductase inhibitors (n)	32	35	0.73^b
ACEIs or ARBs (n)	27	29	0.81^b
Beta-blockers (n)	23	26	0.68^b
Calcium channel blockers (n)	17	19	0.76^b
Aspirin (n)	37	39	0.85^b
Erythropoietin, Unit/kg/week	142 ± 71	143 ± 69	0.43^c
BMI (kg/m²)	25.23 ± 3.71	25.15 ± 4.42	0.17^c
BMD	−2.34 ± 0.26	−2.35 ± 0.11	0.28^c
Diuretics (n)	43	44	0.97^b
Personal history of VTEs (n)	11/125	12/127	0.86^b
Family history of VTEs (n)	13/125	15/127	0.72^b

Values are mean ± standard deviation or median and interquartile range. ^aAnalysed using the Mann–Whitney test; ^banalysed using the chi-square test; ^canalysed using the independent samples t-test. ^#K: dialyzer blood urea clearance in L/h; t: dialysis session length in hours; V: urea distribution volume in litres. BPS: beraprost sodium; CL: clopidogrel; M: male; F: female; BP: blood pressure; eGFR: estimated glomerular filtration rate; BUN: blood urea nitrogen; Hb: haemoglobin; ASA: American Society of Anesthesiologists; ACEIs: angiotensin-converting enzyme inhibitors; ARBs: angiotensin II receptor blockers; HMG-CoA: 3-hydroxy-3-methylglutaryl-coenzyme A; BMI: body mass index; BMD: bone mineral density; VTEs: vascular thromboembolic events.

Table 2.

Comparison of laboratory data between the two groups.

Laboratory data	BPS (n = 125)	CL (n = 127)	P value
Platelets, ×1000/mm³	173 ± 52	168 ± 67	0.12^a
White blood cells, /mm³	7583 ± 1458	7402 ± 1772	0.35^a
Haemoglobin, g/dL	11.12 ± 0.43	10.81 ± 0.26	0.17^a
Serum albumin, g/dL	3.15 ± 0.26	2.99 ± 0.75	0.22^a
Phosphate, mg/dL	4.92 ± 0.47	5.10 ± 0.93	0.37^a
Uric acid, mg/dL	9.21 ± 2.04	8.92 ± 1.52	0.41^a
Total cholesterol, mg/dL	132 ± 46	136 ± 35	0.20^a
Triglycerides, mg/dL	100 (76–144)	99 (74–147)	0.37^a
Ferritin, ng/mL	170 (111–347)	172 (122–352)	0.45^a
Haemoglobin A1c, %	5.11 ± 1.22	4.97 ± 1.84	0.72^a
Low-density lipoprotein cholesterol, mg/dL	74 ± 16	76 ± 13	0.24^a
High-density lipoprotein cholesterol, mg/dL	42 ± 11	44 ± 16	0.28^a
Total calcium, mg/dL	126 ± 23	128 ± 16	0.36^a
Parathyroid hormone, pg/mL	134 (100–186)	137 (97–192)	0.51^a
Bicarbonate, mEq/L	26.40 ± 2.73	25.92 ± 2.25	0.33^a
Transferrin saturation, %	20 (19–27)	22 (18–25)	0.76^a

Values are mean ± standard deviation or median and interquartile range. ^aAnalysed using the independent samples t-test. BPS: beraprost sodium; CL: clopidogrel.

Incidence and risk of VTEs

The incidence rate of the ﬁrst on-study VTEs in the BPS-treated cohort was significantly lower than that in the CL-treated cohort (8.0% [10/125] versus 18.1% [23/127], P = 0.02) (Table 3). The mean time to the first on-study VTE was 2.7 months (95% confidence interval [CI] 14.3–45.3) and 1.5 months (9.9–45.6 months) for the BPS- and CL-treated cohorts (hazard ratio [HR] 0.31, 95% CI 0.24–0.67; P = 0.001). There was an appreciably delayed time to first on-study VTE in the BPS-treated cohort compared with the CL-treated cohort (HR 0.33, 95% CI 0.18–0.56; P = 0.001).

Table 3.

Comparison of the incidence of VTE between the two groups.

Variable	BPS (n = 125)	CL (n = 127)	P value
Total incidence of VTEs	10/125	23/127	0.02
Incidence of VTEs
Within the first month	3/125	11/127	0.03
Within the second month	4/125	4/127	0.98
From the third month to the final follow-up	3/125	8/127	0.13

Values were analysed using the chi-square test. BPS: beraprost sodium; CL: clopidogrel; VTEs: vascular thromboembolic events.

The BPS-treated cohort had a lower occurrence of ﬁrst on-study VTEs compared with the CL-treated cohort at the final follow-up (10 versus 23, P = 0.02). Within the first month after the end of CL therapy, 11 (11/23) patients suffered from a VTE. Within the second month after the end of CL therapy, four (4/23) patients suffered from a VTE and eight (8/23) patients suffered from a VTE from the third month to the final follow-up. VTEs mainly occurred within the first month after cessation of oral administration. There was a time distribution characteristic (tended to suffer from a VTE within the first month) in the occurrence of VTEs after the end of CL therapy. In contrast, the occurrence of VTEs and the time after the end of BPS therapy failed to show obvious distribution characteristics. Within the first month after the end of BPS therapy, three (3/10) patients had a VTE. Within the second month after the end of BPS therapy, four (4/10) patients had a VTE and three (3/10) patients had a VTE from the third month to the final follow-up. BPS reduced the risk of a ﬁrst on-study VTE by 47% compared with CL (HR 0.22, 95% CI 0.15–0.29; P = 0.001) (Figure 2). The risk of a ﬁrst on-study VTE in the CL-treated cohort was approximately three times higher than that in the BPS-treated cohort (HR 3.14 [17.53/5.58], P = 0.001) (Table 4).

Figure 2.

Kaplan–Meier estimates of VTEs of arteriovenous fistula between the groups. At the final follow-up, the incidence of VTEs was higher in the CL group than in the BPS group (P = 0.001). VTEs: vascular thromboembolic events; BPS: beraprost sodium; CL: clopidogrel.

Table 4.

VTE risk ratio between the two groups.

Variable	BPS (n = 125)	CL (n = 127)	P value
HR (95% CI) for total VTEs	5.58	17.53	0.001
	(3.25–14.37)	(2.54–22.14)
HR (95% CI) for VTEs
Within the first month	3.33	8.27	0.003
	(1.21–6.67)	(1.49–15.11)
Within the second month	32.54	36.76	0.151
	(5.43–154.86)	(6.24–97.44)
From the third month	36.64	37.32	0.221
to the final follow-up	(4.65–83.31)	(3.12–88.34)

VTEs: vascular thromboembolic events; BPS: beraprost sodium; CL: clopidogrel; HR: hazard ratio; CI: confidence interval.

Discussion

In our study involving patients with ESRD, the most important finding was that CL-treated patients with uraemia had an increased risk of VTEs. Additionally, CL-treated patients had more VTEs within the first month after cessation of oral administration compared with BPS-treated patients.

A recently published study on preventing VTEs in patients with uraemia suggested that factors, such as advanced age, hypertension, anaemia, arteriosclerosis, abnormal lipid metabolism, and accumulation of uraemic toxins in the body, could result in a patient’s poor vascular elasticity and vascular endovascular stenosis.²³ During surgery, traction of the arteries can lead to postoperative vasospasm, decreased blood flow, and endovascular damage.^24,25 These factors contribute to formation of thrombus after fistula formation, which results in failure of internal fistula function.¹⁴ A recent study reported by Kim et al.²⁶ showed that BPS could improve the success rate of arteriovenous fistula in older patients and the efficacy was improved. Furthermore, previous studies have shown that several vascular anti-proliferative agents (including dipyridamole, statins, and fish oil) that affect the patency rate of arteriovenous fistula have the potential for benefits in anti-neointimal hyperplasia patients on haemodialysis. Additionally, the proportion of patients with diabetic nephropathy who undergo maintenance haemodialysis is increasing every year.²⁷ Patients with diabetic nephropathy are more prone to developing thrombosis and other complications because of their poor blood flow, difficulty in establishing arteriovenous fistula, tendency to merge metabolic disorders, malnutrition, vascular sclerosis, increased release of von Willebrand factor, platelet aggregation, or vascular endothelial cell injury.^27–30

BPS has the advantages of a stable structure, long half-life, and convenient oral administration. BPS is mainly used in clinical treatment of lower extremity arteriosclerosis occlusion diseases, pulmonary hypertension, diabetic peripheral neuropathy, and diabetic nephropathy.^17,28 The curative effect of BPS is distinct.²⁸ Prostacyclin can be combined with prostacyclin receptors on platelets and vascular smooth muscle and activates adenylate cyclase. Prostacyclin increases intracellular adenosine monophosphate concentrations, and inhibits the influx of calcium and production of thromboxane A₂, which can play a role in anti-platelet aggregation and dilating blood vessels.^18,26–28 Prostaglandin E₁ is produced by vascular endothelial cells, and can regulate calcium activity of vascular smooth muscle and prevent noradrenaline release from sympathetic nerve terminals to dilate blood vessels. Prostaglandin E₁ also has the effects of inhibiting platelet aggregation, protecting vascular endothelial cells, and preventing atherosclerosis.^31,32 Additionally, prostaglandin E₁ has an obvious targeting distribution to damaged vessels. This can improve vasospasm of the radial artery, increase blood flow of an internal fistula, and promote maturation of an internal fistula.^31,32

Thrombosis is the major cause of failure of arteriovenous fistulas in patients on haemodialysis.^8,14 Formation of vortices induced by vascular intimal injury at the anastomotic site promotes adhesion of monocyte macrophages, which exacerbates the vicious cycle of fibrosis and atherosclerosis.^5,14 In turn, formation of haematomas and local infection make the local injury site difficult to repair, thereby having a detrimental effect on the patency of blood vessels.²⁶ CL, which is an inactive prodrug, is converted into active components by the cytochrome P₄₅₀3A4 enzyme. The cytochrome P₄₅₀3A4 enzyme can selectively inhibit binding of adenosine diphosphate with its platelet membrane receptor. Ultimately, CL prevents fibrinogen from binding to glycoprotein receptors.¹¹ In addition to adenosine diphosphate, CL can inhibit platelet aggregation induced by other agonists by blocking expansion of platelet activation caused by adenosine diphosphate.^11,20 Additionally, CL has a certain anti-inflammatory effect and can effectively reduce thrombosis.²¹

BPS compared with CL in prevention of VTEs in patients with uraemia and ESRD has been the focus of several previous studies.^8,14,25 However, potential disadvantages in the majority of these studies are the relatively small sample size and/or short-term follow-up. Consequently, making conclusions about the relative superiority of one drug over the other using the time to first on-study VTE as the primary outcome might be incorrect. Chue et al.³² showed that, in patients with ESRD, BPS was inferior to CL in delaying the time to first on-study VTE (HR 1.43; P = 0.001). However, Coleman et al.¹³ further assessed the finding from Chue et al’s.³² study and showed that the rate of VTEs was 9% (n = 108) for BPS versus 11% (n = 110) for CL. BPS delayed the first on-study VTE by 19% versus 11% for CL (HR, 0.33; 95% CI, 0.21–0.76; P = 0.022). In our study, the VTE rate in the CL-treated cohort was considerably higher than that in the BPS-treated cohort, regardless of whether the evaluation was performed during drug use, after the cessation of drug use, or during the follow-up period. The incidence of VTEs in the BPS- and CL-treated cohorts was 8.0% (10/125) and 18.1% (23/127), respectively, at the final follow-up. These values were not higher than 19.2% (10-month follow-up) and 22.7% (16-month follow-up) in other studies.^14,30

Our study has certain limitations that should be considered in the interpretation of our results. First, this study was retrospective in nature with the inherent limitations of retrospective studies. These limitations include the quality of imaging at different centres and variable follow-up of patients. However, we collected data accurately via chart review and electronic records. Second, this study had a small sample size, it was a double-centre study, and the population was largely limited to patients with uraemia who underwent arteriovenous fistula for the first time. Therefore, our findings might not be applicable to general populations of patients with uraemia, such as those who received arteriovenous fistula for the second time. Third, we had no data on important risk factors, such as smoking status, blood pressure, and lipoproteins. Fourth, the analyses did not eliminate imperceptible VTEs that occurred before the endpoint, although the study was powered to detect differences in VTEs.

At the time of the cut-off date for this analysis, we conclude, as expected, that BPS significantly reduced the risk of VTEs compared with CL in patients with uraemia, regardless of whether the patients were followed for 3 years or less. This finding provides further evidence of the superior ability of BPS to prevent VTEs versus CL. Further investigation of the long-term use of BPS and CL is necessary.

Footnotes

Declaration of conflicting interest

The authors declare that there is no conflict of interest.

Funding

Funding for this research was received from the Shanghai Municipal Health and Family Planning Commission Fund Project (201640057) and the Natural Science Foundation of China (81770876, 81270011, 81472125).

References

Kai

Hamada

Morioka

et al . Arteriovenous fistulas at the cervicomedullary junction presenting with subarachnoid hemorrhage: Six case reports with special reference to the angiographic pattern of venous drainage. AJNR Am J Neuroradiol 2005; 26: 1949–1954.

Faiyaz

Abreo

Zaman

et al . Salvage of poorly developed arteriovenous fistulae with percutaneous ligation of accessory veins. Am J Kidney Dis 2002; 39: 824–827.

Almasi-Sperling

Galiano

Lang

et al . Timing of first arteriovenous fistula cannulation in children on hemodialysis. Pediatr Nephrol 2016; 31: 1647–1657.

Chou

Tseng

Shih

et al . Influence of intravenous drug abuse on native arteriovenous fistula thrombosis in chronic hemodialysis patients. Ther Apher Dial 2008; 12: 152–156.

Usta

Elkrinawi

Salehi-Gilani

et al . Risk factors predicting the successful function and use of autogenous arteriovenous fistulae for hemodialysis. Thorac Cardiovasc Surg 2013; 61: 438–444.

Persic

Ponikvar

Buturovic-Ponikvar

Preoperative ultrasonographic mapping of blood vessels before arteriovenous fistula construction in elderly patients with end-stage renal disease.

Ther Apher Dial 2009; 13: 334–339.

Park

Song

Kim

et al . Fabrication of artificial arteriovenous fistula and analysis of flow field and shear stress by using mu-PIV technology. J Mech Sci Technol 2016; 30: 5503–5511.

Chanliau

Charasse

Rose

et al . Clinical evaluation of an expert system for arteriovenous fistula assessment. Int J Artif Organs 2014; 37: 809–815.

Afsar

Elsurer

The primary arteriovenous fistula failure-a comparison between diabetic and non-diabetic patients: glycemic control matters.

Int Urol Nephrol 2012; 44: 575–581.

10.

Glassock

RJ.

Uremia (end-stage renal disease): how cost-effective are preventive strategies?

J Ren Nutr 2010; 20(5 Suppl): S131–S134.

11.

Dember

Beck

Allon

et al . Effect of clopidogrel on early failure of arteriovenous fistulas for hemodialysis - a randomized controlled trial. JAMA 2008; 299: 2164–2171.

12.

Vora

Stanislawski

Grunwald

et al . Association between chronic kidney disease and rates of transfusion and progression to end-stage renal disease in patients undergoing transradial versus transfemoral cardiac catheterization-an analysis from the veterans affairs Clinical Assessment Reporting and Tracking (CART) program. J Am Heart Assoc 2017; 6: e004819.

13.

Coleman

Tuttle

Teevan

et al . Antiplatelet agents for the prevention of arteriovenous fistula and graft thrombosis: a meta analysis. Int J Clin Pract 2010; 64: 1239–1244.

14.

Vachharajani

Wise

Klingensmith

et al . Vascular access creation and maintenance in the USA. J Vasc Access 2015; 16(Suppl 9): S1–S4.

15.

Matsumoto

Iwasa

Kyuragi

et al.

The efficacy of oral beraprost sodium, a prostaglandin I-2 analogue, for treating intermittent claudication in patients with arteriosclerosis obliterans. Int Angiol 2010; 29: 49–54.

16.

Murakami

Watanabe

Furukawa

et al . The prostacyclin analogue beraprost sodium prevents occlusion of bypass grafts in patients with lower extremity arterial occlusive disease: A 20-year retrospective study. Ann Vasc Surg 2005; 19: 838–842.

17.

Mohler

3rd Hiatt

Olin

et al . Treatment of intermittent claudication with beraprost sodium, an orally active prostaglandin I-2 analogue - a double-blinded, randomized, controlled trial. J Am Coll Cardiol 2003; 41: 1679–1686.

18.

Shimamura

Miyakawa

Doi

et al . The pharmacokinetics of beraprost sodium following single oral administration to subjects with impaired kidney function. J Clin Pharmacol 2017; 57: 524–535.

19.

Chemla

Morsy

A European perspective on the Dialysis Access Consortium (DAC) study regarding the effects of clopidogrel on early failure of arteriovenous fistulas for hemodialysis.

J Vasc Access 2008; 9: 229–230.

20.

Gilpin

Review of an article: effects of clopidogrel on early failure of arteriovenous fistulas for hemodialysis by Dember, LM, Beck, GJ, & Allon, M., et al. JAMA 2008; 299(18): 2164–71. J Vasc Nurs 2009; 27: 78.

21.

Ismail

Abushouk

Bekhet

et al . Regional versus local anesthesia for arteriovenous fistula creation in end-stage renal disease: a systematic review and meta-analysis. J Vasc Access 2017; 18: 177–184.

22.

Rajan

Ebner

Desai

et al . Percutaneous creation of an arteriovenous fistula for hemodialysis access. J Vasc Interv Radiol 2015; 26: 484–490.

23.

Schinstock

Albright

Williams

et al . Outcomes of arteriovenous fistula creation after the fistula first initiative. Clin J Am Soc Nephrol 2011; 6: 1996–2002.

24.

Ghorbani

Aalamshah

Shahbazian

et al . Randomized controlled trial of clopidogrel to prevent primary arteriovenous fistula failure in hemodialysis patients. Indian J Nephrol 2009; 19: 57–61.

25.

Al-Balas

Lee

Young

et al . Predictors of initiation for predialysis arteriovenous fistula. Clin J Am Soc Nephrol 2016; 11: 1802–1808.

26.

Kim

et al . Effectiveness of beraprost sodium in maintaining vascular access patency in patients on hemodialysis. Int Urol Nephrol 2017; 49: 1287–1295.

27.

Shima

Miyamoto

Kubota

et al . Beraprost sodium protects against diabetic nephropathy in patients with arteriosclerosis obliterans: a prospective, randomized, open-label study. J Nippon Med Sch 2015; 82: 84–91.

28.

Kim

et al . Effect of beraprost sodium on arterial stiffness in patients with type 2 diabetic nephropathy. Trials 2013; 14: 275.

29.

Oudiz

Benza

Delcroix

et al.

Long-term follow-up in PAH patients dosed with beraprost sodium modified release (BPS-MR) tablets, an oral twice daily prostacyclin analogue. Am J Respir Crit Care Med 2011; 183: A5905.

30.

Sugawara

Kudo

Saito

et al . Novel effects of beraprost sodium on vasculatures. Int Angiol 2010; 29: 28–32.

31.

Hashiguchi

Ohno

Saito

Studies on the effectiveness and safety of cilostazol, beraprost sodium, prostaglandin E-1 for the treatment of intermittent claudication.

Yakugaku Zasshi 2004; 124: 321–332.

32.

Chue

Thant

Luo

et al . Comprehensive comparison of the performance of autogenous brachial-basilic transposition arteriovenous fistula and prosthetic forearm loop arteriovenous graft in a multiethnic Asian hemodialysis population. Biomed Res Int 2016; 2016: 8693278.