- Email: [email protected]
A Systematic Review of Endovenous Stenting in Chronic Venous Disease Secondary to Iliac Vein Obstruction
Selected Abstracts from the January Issue of the European Journal of Vascular and Endovascular Surgery A. Ross Naylor, MBChB, MD, FRCS, Editor-in-Chief, and Philippe H. Kolh, MD, PhD, Senior Editor
Stroke/Death Rates Following Carotid Artery Stenting and Carotid Endarterectomy in Contemporary Administrative Dataset Registries: A Systematic Review Paraskevas K.I., Kalmykov E.L., Naylor A.R. Eur J Vasc Endovasc Surg 2016;51:3-12. Background: Randomised trials have reported higher stroke/death rates after carotid artery stenting (CAS) versus carotid endarterectomy (CEA). Despite this, the 2011 American Heart Association (AHA) guidelines expanded CAS indications, partly because of the Carotid Revascularization Endarterectomy versus Stenting Trial, but also because of improving outcomes in industry sponsored CAS Registries. The aim of this systematic review was: (i) to compare stroke/death rates after CAS/ CEA in contemporary dataset registries, (ii) to examine whether published stroke/death rates after CAS fall within AHA thresholds, and, (iii) to see if there had been a decline (over time) in procedural risk after CAS/CEA. Methods: PubMed/Medline, Embase, and Cochrane databases were systematically searched according to the recommendations of the PRISMA statement from January 1, 2008 until February 23, 2015 for administrative dataset registries reporting outcomes after both CEA and CAS. Results: Twenty-one registries reported outcomes involving more than 1,500,000 procedures. Stroke/death after CAS was significantly higher than after CEA in 11/21 registries (52%) involving “average risk for CEA” asymptomatic patients and in 11/18 registries (61%) involving “average risk for CEA” symptomatic patients. In another five registries, CAS was associated with higher stroke/death rates than CEA for both symptomatic and asymptomatic patients, but formal statistical comparison was not reported. CAS was associated with stroke/death rates that exceeded risk thresholds recommended by the AHA in 9/21 registries (43%) involving “average risk for CEA” asymptomatic patients and in 13/18 registries (72%) involving “average risk for CEA” symptomatic patients. In 5/18 registries (28%), the procedural risk after CAS in “average risk” symptomatic patients exceeded 10%. Conclusions: Data from contemporary administrative dataset registries suggest that stroke/death rates following CAS remain significantly higher than after CEA and often exceed accepted AHA thresholds. There was no evidence of a sustained decline in procedural risk after CAS. Endovascular Versus Open Repair as Primary Strategy for Ruptured Abdominal Aortic Aneurysm: A National Population-based Study Gunnarsson K., Wanhainen A., Djavani Gidlund K., Björck M., Mani K. Eur J Vasc Endovasc Surg 2016;51:22-8. Objective/Background: In randomized trials, no peri-operative survival benefit has been shown for endovascular (EVAR) repair of ruptured abdominal aortic aneurysm (rAAA) when compared with open repair. The aim of this study was to investigate the effect of primary repair strategy on early and midterm survival in a non-selected population based study. Methods: The Swedish Vascular Registry was consulted to identify all rAAA repairs performed in Sweden in the period 2008-12. Centers with a primary EVAR strategy (treating >50% of rAAA with EVAR) were compared with centers with a primary open repair strategy. Peri-operative outcome, midterm survival, and incidence of rAAA repair/100,000 inhabitants aged >50 years were assessed. Results: In total, 1304 patients were identified. Three primary EVAR centers (pEVARc) operated on 236 patients (74.6% EVAR). Twenty-six primary open repair centers (pORc) operated 1068 patients (15.6% EVAR). Patients treated at pEVARc were more often referrals (28.0% vs 5.3%; P < .01), had a higher rate of respiratory comorbidity (36.5% vs 21.9%; P < .01), and higher pre-operative systolic blood pressure (84.3 vs 72.3 mmHg; P < .01). There was no difference in mortality based on primary treatment strategy at 30 days (pEVARc 28.0%, n ¼ 66; pORc 27.4%, n ¼ 296 [P ¼ .87]), 1 year (pEVARc 39.9%, n ¼ 93; pORc 34.7%, n ¼ 366 [P ¼ .19]), or 2 years (42.1%, n ¼ 94; 38.3%, n ¼ 394 [P ¼ .28]), either overall or in subgroups based on age or referral status. Overall, patients treated with EVAR were older (mean age 76.4 vs 74.0 years; P < .01),
and had a lower 30 day mortality (EVAR 21.6%, n ¼ 74; odds ratio 29.6%, n ¼ 288 [P # .01]). Incidence of rAAA repair was lower in pEVARc regions (6.07, 95% confidence interval [CI] 5.01-7.13) when compared with pORc regions (8.15, 95% CI 7.64-8.66). Conclusion: There was no difference in mortality after rAAA repair among centers with a primary EVAR approach when compared with a primary open repair strategy, either peri-operatively or in the midterm. The study supports the early findings of the randomized controlled trials in a national population based setting. Association Between Fibulin-1 and Aortic Augmentation Index in Male Patients with Peripheral Arterial Disease Paapstel K., Zilmer M., Eha J., Tootsi K., Piir A., Kals J. Eur J Vasc Endovasc Surg 2016;51:76-82. Background: Fibulin-1 (FBLN-1), a newly identified biomarker for vascular stiffness in type 2 diabetes, may participate in the pathophysiological processes leading to progression of arterial stiffness in atherosclerosis. In the present study, the relationship between FBLN-1 and arterial stiffness was examined in patients with atherosclerosis and in healthy subjects. Methods: Thirty-eight patients with peripheral arterial disease (PAD) (age 62.4 6 9.0 years), 38 patients with coronary artery disease (CAD) (age 64.0 6 9.5 years), and 30 apparently healthy controls (age 61.1 6 6.4 years) were studied. Serum FBLN-1, oxidized low density lipoprotein (oxLDL), resistin and plasminogen activator inhibitor-1 (PAI-1) levels were measured using the enzyme linked immunosorbent assay method. The technique of applanation tonometry was used for non-invasive pulse wave analysis and pulse wave velocity assessments. Results: The levels of FBLN-1 (PAD ¼ 9.4 [4.9-17.8] vs CAD ¼ 7.1 [4.8-11.8] vs controls ¼ 5.6 [4.1-8.4] mg/mL; P ¼ .005), carotid-femoral pulse wave velocity (cf-PWV) (9.8 6 2.2 vs 9.5 6 2.2 vs 8.3 6 2.2 m/s; P ¼ .023) and the heart rate corrected augmentation index (AIx@75) (29.4 6 7.2 vs 19.2 6 7.2 vs 15.4 6 7.1%; P < .001), differed among the three groups. A correlation between FBLN-1 and AIx@75 was observed only in patients with PAD (rho ¼ 0.37, P ¼ .021). The relationship retained statistical significance in a multiple regression model after adjustment for potential confounders. Conclusions: An independent association was demonstrated between serum FBLN-1 and AIx@75 in the PAD group. Thus, the findings suggest that FBLN-1 may play a role in arterial stiffening in patients with atherosclerosis. A Systematic Review of Endovenous Stenting in Chronic Venous Disease Secondary to Iliac Vein Obstruction Seager M.J., Busuttil A., Dharmarajah B., Davies A.H. Eur J Vasc Endovasc Surg 2016;51:100-20. Objectives: Deep endovenous stenting to relieve chronic venous disease (CVD) secondary to post-thrombotic or non-thrombotic iliac vein obstruction is becoming increasingly well described. However, current and adequately reported systematic reviews on the topic are lacking. This report aimed to produce a systematic review and meta-analysis of the available data, reported to the Preferred Reporting Items for Systematic reviews and Meta-Analyses guideline. Methods: MEDLINE, EMBASE, and the Cochrane Central Register for Controlled Trials databases and key references were searched. Results: Sixteen studies were included (14 before-and-after studies, 1 controlled before-and-after study, and 1 case series) encompassing successful deep venous stenting in 2373 and 2586 post-thrombotic or non-thrombotic limbs and patients respectively. The data were too heterogeneous to perform a meta-analysis. There were significant improvements in validated measures of the severity of CVD and venous disease-specific quality of life. Persistent ulcer healing rates ranged from 56% to 100% in limbs that had often already failed conservative management. Primary and secondary
555
556 Abstracts
stent patency ranged from 32% to 98.7% and 66%-96% respectively. The major complication rate ranged from 0 to 8.7% per stented limb. A GRADE assessment demonstrated the quality of the evidence for five outcomes to be “Very Low” and one to be “Low” (ulcer healing). Conclusions: The quality of evidence to support the use of deep venous stenting to treat obstructive CVD is currently weak. The treatment does however appear promising and is safe and should therefore be considered as a treatment option while the evidence base is improved. Post-operative Venous Thromboembolism in Patients Operated on for Aorto-iliac Obstruction and Abdominal Aortic Aneurysm, and the Application of Pharmacological Thromboprophylaxis Pawlaczyk K., Gabriel M., Dzieciuchowicz L., Stanisic M., Begier-Krasinska B., Gabriel Z., Olejniczak-Nowakowska M., Urbanek T. Eur J Vasc Endovasc Surg 2016;51:121-6. Objective/background: In light of the methods generally used to assess the risk of venous thromboembolism (VTE), major vascular operations should be regarded as high risk procedures. Nevertheless, no principles for implementing and maintaining thromboprophylaxis have so far been developed. The aim of this study was to determine the frequency and nature of VTE occurrence in patients routinely applying pharmacological thromboprophylaxis following implantation of an aorto-bifemoral prosthesis. Methods: The prospective non-randomized study included 105 patients with aortoiliac obstruction and 119 patients with abdominal aortic aneurysm (AAA) treated surgically. During hospitalization pharmacological thromboprophylactic procedures were observed. A duplex test was performed on the day before surgery, on the day of discharge, and 30 days after the patients had left the hospital. Results: VTE was detected in 18.1% of the patients with aortoiliac obstruction (9.5% of patients during hospitalization and 8.6% of patients after discharge). VTE was diagnosed in 21.0% of patients with AAA (15.1% of patients during hospitalization and 5.9% of patients after discharge). The incidence of VTE was comparable in both groups, both during hospitalization (P ¼ .51) and in the 30 day period following the end of hospitalization (P ¼ .48). It is advisable that before hospital discharge routine duplex ultrasonography tests should be conducted on the venous systems of all patients who have undergone major vascular operations.
JOURNAL OF VASCULAR SURGERY February 2016
Conclusions: It is likewise advisable to consider whether thromboprophylaxis for vascular patients should be extended beyond their discharge from hospital. Natural History of Common Autologous Arteriovenous Fistulae: Consequences for Planning of Dialysis Access Wilmink T., Hollingworth L., Powers S., Allen C., Dasgupta I. Eur J Vasc Endovasc Surg 2016;51:134-40. Objectives: The aim was to study primary failure, maturation times, and survival of common arteriovenous fistulae (AVF) to aid planning for vascular access, and to assess which strategy results in most dialysis days. Methods: This was a longitudinal cohort study. Two databases of access operations and dialysis sessions over 9 years with 12-year follow-up were reviewed. Functional dialysis use is defined as achieving six consecutive dialysis sessions with two needles on AVF. Primary failure (PF) is failure to achieve functional dialysis use. Maturation time, calculated only for patients on dialysis with a central line at AVF operation, is defined from the operation date to the functional dialysis date. Cumulative patency, including PF, is calculated from the operation to date of AVF abandonment and is compared using KaplaneMeier curves and adjusted hazard ratios (HRs). Results: A total of 1206 AVF, 689 (57%) radiocephalic AVF (RCAVF), 383 (32%) brachiocephalic AVF (BCAVF), and 134 (11%) brachiobasilic AVF (BBAVF), were analysed. PF was 23%. PF was lower for BCAVF (17%) than RCAVF (26%) and BBAVF (26%) (P ¼ .006). PF was higher for women (OR 1.59, 95% CI: 1.21-2.09) and patients with vascular kidney disease (OR 1.69, 95% CI: 1.19-2.59). Median maturation time was 10.3 weeks. Cumulative patency was worse for BCAVF (HR 1.36, 95% CI: 1.03-1.81) and BBAVF (HR 1.63 95% CI: 1.12-2.38), for patients on dialysis at AVF creation (HR 1.55, 95% CI: 1.13-2.12), and diabetics (HR 1.55, 95% CI: 1.12-1.85). RCAVFs resulted in 3% more dialysisperson-years (py) per 100 operations for all patients and in 15% more dialysis-py in the over 80s. Conclusion: RCAVFs have higher PF, but better survival than other AVF, and result in more dialysis time. AVF created pre-dialysis have better survival. An average maturation time of 10 weeks should be considered if planning to start dialysis on an AVF.
Stroke/Death Rates Following Carotid Artery Stenting and Carotid Endarterectomy in Contemporary Administrative Dataset Registries: A Systematic Review Paraskevas K.I., Kalmykov E.L., Naylor A.R. Eur J Vasc Endovasc Surg 2016;51:3-12. Background: Randomised trials have reported higher stroke/death rates after carotid artery stenting (CAS) versus carotid endarterectomy (CEA). Despite this, the 2011 American Heart Association (AHA) guidelines expanded CAS indications, partly because of the Carotid Revascularization Endarterectomy versus Stenting Trial, but also because of improving outcomes in industry sponsored CAS Registries. The aim of this systematic review was: (i) to compare stroke/death rates after CAS/ CEA in contemporary dataset registries, (ii) to examine whether published stroke/death rates after CAS fall within AHA thresholds, and, (iii) to see if there had been a decline (over time) in procedural risk after CAS/CEA. Methods: PubMed/Medline, Embase, and Cochrane databases were systematically searched according to the recommendations of the PRISMA statement from January 1, 2008 until February 23, 2015 for administrative dataset registries reporting outcomes after both CEA and CAS. Results: Twenty-one registries reported outcomes involving more than 1,500,000 procedures. Stroke/death after CAS was significantly higher than after CEA in 11/21 registries (52%) involving “average risk for CEA” asymptomatic patients and in 11/18 registries (61%) involving “average risk for CEA” symptomatic patients. In another five registries, CAS was associated with higher stroke/death rates than CEA for both symptomatic and asymptomatic patients, but formal statistical comparison was not reported. CAS was associated with stroke/death rates that exceeded risk thresholds recommended by the AHA in 9/21 registries (43%) involving “average risk for CEA” asymptomatic patients and in 13/18 registries (72%) involving “average risk for CEA” symptomatic patients. In 5/18 registries (28%), the procedural risk after CAS in “average risk” symptomatic patients exceeded 10%. Conclusions: Data from contemporary administrative dataset registries suggest that stroke/death rates following CAS remain significantly higher than after CEA and often exceed accepted AHA thresholds. There was no evidence of a sustained decline in procedural risk after CAS. Endovascular Versus Open Repair as Primary Strategy for Ruptured Abdominal Aortic Aneurysm: A National Population-based Study Gunnarsson K., Wanhainen A., Djavani Gidlund K., Björck M., Mani K. Eur J Vasc Endovasc Surg 2016;51:22-8. Objective/Background: In randomized trials, no peri-operative survival benefit has been shown for endovascular (EVAR) repair of ruptured abdominal aortic aneurysm (rAAA) when compared with open repair. The aim of this study was to investigate the effect of primary repair strategy on early and midterm survival in a non-selected population based study. Methods: The Swedish Vascular Registry was consulted to identify all rAAA repairs performed in Sweden in the period 2008-12. Centers with a primary EVAR strategy (treating >50% of rAAA with EVAR) were compared with centers with a primary open repair strategy. Peri-operative outcome, midterm survival, and incidence of rAAA repair/100,000 inhabitants aged >50 years were assessed. Results: In total, 1304 patients were identified. Three primary EVAR centers (pEVARc) operated on 236 patients (74.6% EVAR). Twenty-six primary open repair centers (pORc) operated 1068 patients (15.6% EVAR). Patients treated at pEVARc were more often referrals (28.0% vs 5.3%; P < .01), had a higher rate of respiratory comorbidity (36.5% vs 21.9%; P < .01), and higher pre-operative systolic blood pressure (84.3 vs 72.3 mmHg; P < .01). There was no difference in mortality based on primary treatment strategy at 30 days (pEVARc 28.0%, n ¼ 66; pORc 27.4%, n ¼ 296 [P ¼ .87]), 1 year (pEVARc 39.9%, n ¼ 93; pORc 34.7%, n ¼ 366 [P ¼ .19]), or 2 years (42.1%, n ¼ 94; 38.3%, n ¼ 394 [P ¼ .28]), either overall or in subgroups based on age or referral status. Overall, patients treated with EVAR were older (mean age 76.4 vs 74.0 years; P < .01),
and had a lower 30 day mortality (EVAR 21.6%, n ¼ 74; odds ratio 29.6%, n ¼ 288 [P # .01]). Incidence of rAAA repair was lower in pEVARc regions (6.07, 95% confidence interval [CI] 5.01-7.13) when compared with pORc regions (8.15, 95% CI 7.64-8.66). Conclusion: There was no difference in mortality after rAAA repair among centers with a primary EVAR approach when compared with a primary open repair strategy, either peri-operatively or in the midterm. The study supports the early findings of the randomized controlled trials in a national population based setting. Association Between Fibulin-1 and Aortic Augmentation Index in Male Patients with Peripheral Arterial Disease Paapstel K., Zilmer M., Eha J., Tootsi K., Piir A., Kals J. Eur J Vasc Endovasc Surg 2016;51:76-82. Background: Fibulin-1 (FBLN-1), a newly identified biomarker for vascular stiffness in type 2 diabetes, may participate in the pathophysiological processes leading to progression of arterial stiffness in atherosclerosis. In the present study, the relationship between FBLN-1 and arterial stiffness was examined in patients with atherosclerosis and in healthy subjects. Methods: Thirty-eight patients with peripheral arterial disease (PAD) (age 62.4 6 9.0 years), 38 patients with coronary artery disease (CAD) (age 64.0 6 9.5 years), and 30 apparently healthy controls (age 61.1 6 6.4 years) were studied. Serum FBLN-1, oxidized low density lipoprotein (oxLDL), resistin and plasminogen activator inhibitor-1 (PAI-1) levels were measured using the enzyme linked immunosorbent assay method. The technique of applanation tonometry was used for non-invasive pulse wave analysis and pulse wave velocity assessments. Results: The levels of FBLN-1 (PAD ¼ 9.4 [4.9-17.8] vs CAD ¼ 7.1 [4.8-11.8] vs controls ¼ 5.6 [4.1-8.4] mg/mL; P ¼ .005), carotid-femoral pulse wave velocity (cf-PWV) (9.8 6 2.2 vs 9.5 6 2.2 vs 8.3 6 2.2 m/s; P ¼ .023) and the heart rate corrected augmentation index (AIx@75) (29.4 6 7.2 vs 19.2 6 7.2 vs 15.4 6 7.1%; P < .001), differed among the three groups. A correlation between FBLN-1 and AIx@75 was observed only in patients with PAD (rho ¼ 0.37, P ¼ .021). The relationship retained statistical significance in a multiple regression model after adjustment for potential confounders. Conclusions: An independent association was demonstrated between serum FBLN-1 and AIx@75 in the PAD group. Thus, the findings suggest that FBLN-1 may play a role in arterial stiffening in patients with atherosclerosis. A Systematic Review of Endovenous Stenting in Chronic Venous Disease Secondary to Iliac Vein Obstruction Seager M.J., Busuttil A., Dharmarajah B., Davies A.H. Eur J Vasc Endovasc Surg 2016;51:100-20. Objectives: Deep endovenous stenting to relieve chronic venous disease (CVD) secondary to post-thrombotic or non-thrombotic iliac vein obstruction is becoming increasingly well described. However, current and adequately reported systematic reviews on the topic are lacking. This report aimed to produce a systematic review and meta-analysis of the available data, reported to the Preferred Reporting Items for Systematic reviews and Meta-Analyses guideline. Methods: MEDLINE, EMBASE, and the Cochrane Central Register for Controlled Trials databases and key references were searched. Results: Sixteen studies were included (14 before-and-after studies, 1 controlled before-and-after study, and 1 case series) encompassing successful deep venous stenting in 2373 and 2586 post-thrombotic or non-thrombotic limbs and patients respectively. The data were too heterogeneous to perform a meta-analysis. There were significant improvements in validated measures of the severity of CVD and venous disease-specific quality of life. Persistent ulcer healing rates ranged from 56% to 100% in limbs that had often already failed conservative management. Primary and secondary
555
556 Abstracts
stent patency ranged from 32% to 98.7% and 66%-96% respectively. The major complication rate ranged from 0 to 8.7% per stented limb. A GRADE assessment demonstrated the quality of the evidence for five outcomes to be “Very Low” and one to be “Low” (ulcer healing). Conclusions: The quality of evidence to support the use of deep venous stenting to treat obstructive CVD is currently weak. The treatment does however appear promising and is safe and should therefore be considered as a treatment option while the evidence base is improved. Post-operative Venous Thromboembolism in Patients Operated on for Aorto-iliac Obstruction and Abdominal Aortic Aneurysm, and the Application of Pharmacological Thromboprophylaxis Pawlaczyk K., Gabriel M., Dzieciuchowicz L., Stanisic M., Begier-Krasinska B., Gabriel Z., Olejniczak-Nowakowska M., Urbanek T. Eur J Vasc Endovasc Surg 2016;51:121-6. Objective/background: In light of the methods generally used to assess the risk of venous thromboembolism (VTE), major vascular operations should be regarded as high risk procedures. Nevertheless, no principles for implementing and maintaining thromboprophylaxis have so far been developed. The aim of this study was to determine the frequency and nature of VTE occurrence in patients routinely applying pharmacological thromboprophylaxis following implantation of an aorto-bifemoral prosthesis. Methods: The prospective non-randomized study included 105 patients with aortoiliac obstruction and 119 patients with abdominal aortic aneurysm (AAA) treated surgically. During hospitalization pharmacological thromboprophylactic procedures were observed. A duplex test was performed on the day before surgery, on the day of discharge, and 30 days after the patients had left the hospital. Results: VTE was detected in 18.1% of the patients with aortoiliac obstruction (9.5% of patients during hospitalization and 8.6% of patients after discharge). VTE was diagnosed in 21.0% of patients with AAA (15.1% of patients during hospitalization and 5.9% of patients after discharge). The incidence of VTE was comparable in both groups, both during hospitalization (P ¼ .51) and in the 30 day period following the end of hospitalization (P ¼ .48). It is advisable that before hospital discharge routine duplex ultrasonography tests should be conducted on the venous systems of all patients who have undergone major vascular operations.
JOURNAL OF VASCULAR SURGERY February 2016
Conclusions: It is likewise advisable to consider whether thromboprophylaxis for vascular patients should be extended beyond their discharge from hospital. Natural History of Common Autologous Arteriovenous Fistulae: Consequences for Planning of Dialysis Access Wilmink T., Hollingworth L., Powers S., Allen C., Dasgupta I. Eur J Vasc Endovasc Surg 2016;51:134-40. Objectives: The aim was to study primary failure, maturation times, and survival of common arteriovenous fistulae (AVF) to aid planning for vascular access, and to assess which strategy results in most dialysis days. Methods: This was a longitudinal cohort study. Two databases of access operations and dialysis sessions over 9 years with 12-year follow-up were reviewed. Functional dialysis use is defined as achieving six consecutive dialysis sessions with two needles on AVF. Primary failure (PF) is failure to achieve functional dialysis use. Maturation time, calculated only for patients on dialysis with a central line at AVF operation, is defined from the operation date to the functional dialysis date. Cumulative patency, including PF, is calculated from the operation to date of AVF abandonment and is compared using KaplaneMeier curves and adjusted hazard ratios (HRs). Results: A total of 1206 AVF, 689 (57%) radiocephalic AVF (RCAVF), 383 (32%) brachiocephalic AVF (BCAVF), and 134 (11%) brachiobasilic AVF (BBAVF), were analysed. PF was 23%. PF was lower for BCAVF (17%) than RCAVF (26%) and BBAVF (26%) (P ¼ .006). PF was higher for women (OR 1.59, 95% CI: 1.21-2.09) and patients with vascular kidney disease (OR 1.69, 95% CI: 1.19-2.59). Median maturation time was 10.3 weeks. Cumulative patency was worse for BCAVF (HR 1.36, 95% CI: 1.03-1.81) and BBAVF (HR 1.63 95% CI: 1.12-2.38), for patients on dialysis at AVF creation (HR 1.55, 95% CI: 1.13-2.12), and diabetics (HR 1.55, 95% CI: 1.12-1.85). RCAVFs resulted in 3% more dialysisperson-years (py) per 100 operations for all patients and in 15% more dialysis-py in the over 80s. Conclusion: RCAVFs have higher PF, but better survival than other AVF, and result in more dialysis time. AVF created pre-dialysis have better survival. An average maturation time of 10 weeks should be considered if planning to start dialysis on an AVF.