Use of Stents and Stent Grafts to Salvage Angioplasty Failures in Patients with Hemodialysis Grafts

Use of Stents and Stent Grafts to Salvage Angioplasty Failures in Patients with Hemodialysis Grafts Thomas M. Vesely, Mohammed Zaheer Amin, and Thomas Pilgram St. Louis, Missouri ABSTRACT To determine the efficacy of using stents and stent grafts for treatment of hemodialysis graft-related stenoses which have failed angioplasty. This retrospective study was composed of 51 patients with polytetrafluoroethylene hemodialysis grafts who underwent angioplasty of a graft-related stenosis but subsequently required insertion of a stent or stent graft. The study group included 40 patients with >30% residual stenosis and 11 patients with angioplasty-induced venous ruptures. The patient s medical records and radiology reports were reviewed. Patient demographics, indication for stent placement, and the type of stent were recorded. Stent patency was determined using Kaplan Meier analysis. A total of 70 stents were used; most (57.7%) were deployed across the venous anastomosis. A variety of different types of stents and stent grafts were used including 26 SMART, 18 Viabahn, 13 aspire, and 13 other stents. The technical success rate for insertion of the device was 100%. Following insertion of the device the primary patency of the vascular access was 81%, 70%, and 54% at 1, 3, and 6 months, respectively. The secondary patency of the vascular access was 89%, 82%, and 74% at 3, 6, and 12 months, respectively. The primary patency of the stent or stent graft was 96%, 93%, 87%, and 47% at 1, 3, 6, and 12 months, respectively. Stents and stent grafts are useful for salvaging failed angioplasty procedures and thereby maintaining patency of the hemodialysis graft. By inserting a stent, the resultant patency rates are similar to those following a successful angioplasty procedure. Importantly, the primary patency of the stent was substantially better than the primary patency of the entire vascular access. Angioplasty continues to be the primary percutaneous technique for treatment of venous stenoses associated with arteriovenous fistulas and prosthetic grafts used for hemodialysis (1). It remains the standard treatment to which other percutaneous techniques are compared. A variety of other percutaneous devices have been utilized to treat vascular access-related stenoses. These include the peripheral cutting balloon, cryoplasty, metal stents and stent grafts. Stents and stent grafts appear to be an ideal method to treat neointimal hyperplastic stenoses. These endovascular devices can oppose elastic recoil and optimize endoluminal dimensions thereby improving blood flow through the fistula or graft. However, the use of stents remains controversial. Early studies reported that the routine use of metallic stents failed to provide any additional benefit when compared with angioplasty alone (2,3). More recent publications have reported improved long-term patency rates when stents are used in selected situations (4 6). The usual mode of stent failure is ingrowth of neointimal Address correspondence to: Thomas Vesely, MD, 24 The Boulevard, Suite 420, St. Louis, MO 63117, or e-mail: tmvesely@aol.com. Seminars in Dialysis Vol 21, No 1 (January February) 2008 pp. 100 104 DOI: 10.1111/j.1525-139X.2007.00326.x hyperplasic tissue through the mesh of a metallic stent. The use of a covered stent or stent graft would seem to be a logical choice to prevent this problem. But once again, published studies have reported differing results and opinions regarding the value of stent grafts when compared with conventional angioplasty (7,8). The use of stents and stent grafts to treat vascular access-related stenoses continues to be a controversial subject. This retrospective review was performed to evaluate our outcomes when using stents and stent grafts for treatment of venous stenoses which failed to improve with angioplasty and for the repair of angioplastyinduced venous ruptures. Materials and Methods Patients A computer search of the interventional radiology database at our institution was performed to identify all patients with polytetrafluoroethylene (PTFE) hemodialysis grafts who received a stent or stent graft during the 4-year time period from April 2000 to April 2004. We did not include patients who had stents placed in the central veins or for treatment of a pseudoaneurysm. Fiftyone patients were included in this study. There were 34 female patients and the median age of all patients was 100

STENTS AND STENT GRAFTS IN HEMODIALYSIS 101 62.2 years. The indication for the patient s referral to interventional radiology included (i) decreased intragraft blood flow, (ii) elevated venous pressures, (iii) upper extremity swelling, or (iv) graft thrombosis. Thirty-two patients presented with patent, dysfunctional grafts and 19 patients presented with thrombosed grafts. Fortythree patients had loop configuration hemodialysis grafts and seven patients had straight grafts. Thirtythree grafts were located in the left upper extremity, 16 in the right upper extremity, and two in the right thigh. The angiographic images, procedural reports, and medical records of these 51 patients were retrospectively reviewed to document all vascular access-related events. Graft-related procedures that were performed following placementofthestentorstentgraftwerereviewedto determine the patency of the device, the location of venous stenoses, and the cause of vascular access failure. This information was used to determine whether subsequent graft failure resulted from restenosis within the stent or stent graft, or from the development of a new stenosis within the vascular access circuit Procedures An 18 g needle was used to enter the apex of the loopconfiguration PTFE graft. A 5 French angiographic catheter was advanced into the arterial limb of the graft andpositionedwiththedistaltipinthenativeartery adjacent to the arterial anastomosis. A thorough diagnostic fistulogram was performed to evaluate the hemodialysis graft and native outflow veins. Nineteen patients presented with a thrombosed hemodialysis graft. In these patients a thrombectomy procedure was performed after evaluation of the native veins and prior to angioplasty of venous stenoses. A 7 French or 8 French vascular introducer sheath (DialEase; Mallinckrodt Medical, St. Louis, MO) was inserted into the apex of the graft. Heparin (3000 U) was administered to patients undergoing a thrombectomy procedure but was not given to patients with patent grafts. In all 19 patients the thrombectomy procedure was performed using the percutaneous thrombolytic device (Arrow Int., Reading, PA). Angioplasty of venous stenoses was performed after completion of the thrombectomy procedure but before dislodgement of the arterial plug. Stenoses causing greater than 50% luminal reduction were treated with angioplasty. In the majority of patients the degree of stenosis was determined by visual assessment of the lesion on the digital subtraction images. When in doubt, the degree of stenosis was measured using the calibrated software within the digital imaging system. The venous stenosis was crossed with a 0.035 inch Bentson guidewire (Cook Inc., Bloomington, IN) allowing introduction of an appropriately sized angioplasty balloon. The diameter of the balloon was selected by the operating physician and was typically 10 20% larger than the diameter of a normal venous segment adjacent to the stenosis. The angioplasty balloon was inflated using an insufflator which contained dilute contrast material. Balloon inflation times were typically 60 seconds. Following completion of the angioplasty procedure the balloon catheter was removed and a completion fistulogram was performed by injecting contrast material through the sidearm of the vascular sheath. The postangioplasty fistulogram revealed that 40 patients had greater than 30% residual stenosis and 11 patients had a venous rupture at the site of angioplasty. Because of the failed angioplasty procedure these 51 patients underwent additional treatment with insertion of a stent or stent graft (Fig. 1). The existing introducer sheath was replaced with an 8 French or 9 French sheath. The 0.035 inch Bentson guidewire which was in position from the angioplasty procedure was utilized for the stent insertion procedure. The length of the venous lesion was measured using the calibrated software within the digital imaging system. The length of the stent or stentgraftwaschosensothatitextendedatleast10mm beyond each end of the lesion. Under fluoroscopic observation the stent delivery catheter was advanced into the graft, and the device was appropriately positioned and deployed across the stenotic segment. After deployment, an appropriate diameter angioplasty balloon was used to fully expand the stent or stent graft and to ensure complete apposition of the device against the vessel wall. A final fistulogram wasperformedtoevaluate the final position of the stent or stent graft (Fig. 1). At the completion of the procedure the introducer sheath was removed and a purse-string suture was used to close the access site at the graft apex. Anticoagulation or antibiotics were not administered during, or following, any of these procedures. Statistical Analysis The entire study group of 51 patients was included in the statistical analysis. Primary patency, assisted primary patency, and secondary patency rates were calculated using Kaplan Meier life table analysis. Primary patency of the vascular access circuit was defined as the time period from stent placement to the next graftrelated event. Assisted primary patency of the vascular access circuit was defined as thetimeperiodfromstent placement to graft thrombosis. Secondary patency was defined as the time period from stent placement to abandonment of the graft. Primary patency of the stent or stent graft was defined as the time period from device placement to restenosis or thrombosis of the device. Kaplan Meier life table analysis, log-rank, and Wilcoxon tests were used for analysis and comparison of subgroups. The three subgroups included (i) 18 patients with SMART stents, (ii) 10 patients with Viabahn stent grafts, and (iii) 10 patients with aspire stent grafts. Only those patients with one type of stent were included in the subgroup analysis. The other stent subgroups had less than five patients and were not included. Results The study group consisted of 51 patients with PTFE hemodialysis grafts who had an unsuccessful angioplasty procedure and underwent treatment with a stent or stent graft.

102 Vesely et al. (a) (b) (c) Stent graft Fig. 1. (a) A fistulogram reveals a tight stenosis at the venous anastomosis of a prosthetic hemodialysis graft. (b) Following angioplasty there is significant residual stenosis. (c) Following insertion of an aspire stent graft there is substantial improvement in luminal diameter. TABLE 1. Number of stents and stent grafts by type Stent stent graft type n TABLE 2. Patency rates of the vascular access circuit in 51 patients 1 month 3 months 6 months 12 months SMART 26 Viabahn 18 aspire 13 Wallstent 5 Flare 4 Luminexx 2 Genesis 1 Fluency 1 Total 70 A total of 70 devices were inserted; 34 metal stents, and 36 stent grafts. The number and types of stents and stent grafts is listed in Table 1. Thirty-seven patients received one stent, nine patients received two stents, and five patients received three stents. Four patients received two different types of stents or stent grafts. The technical success rate for insertion of the stent or stent graft is 100%. The majority (57.7%) of stents were deployed across the venous anastomosis. There were no complications associated with the stent or stent graft insertion procedures. The primary patency, assisted primary patency, and secondary patency rates of the vascular access circuit are listed in Table 2 and the Kaplan Meier survival data are showninfig.3. The patency of the stent or stent graft was determined separately from the overall patency of the vascular access circuit. The primary patency rate of the stent or stent graft is 96%, 93%, 87%, and 47% at 1, 3, 6, and 12 months, respectively. The primary patency rates of the three most commonly used stents; the SMART stent, the aspire stent graft, and the Viabahn stent graft, are Primary 81 70 54 13 Assisted primary 83 76 66 47 Secondary 89 89 82 74 TABLE 3. Primary patency by stent type. Includes only patients with one type of stent 1 month 3 months 6 months 12 months SMART (n = 18) 78 72 60 33 aspire (n = 10) 90 68 56 0 Viabahn (n = 10) 89 78 26 26 listed in Table 3 and the Kaplan Meier survival data are shown in Fig. 2. When comparing these three devices there was no difference in their primary patency rates (p =0.59). There was also no difference (p = 0.95) in the primary patency of the vascular access circuit when comparing the group of patients who presented with patent, dysfunctional grafts to those patients who presented with thrombosed grafts. Discussion The long-term follow-up of 51 hemodialysis patients who received a stent or stent graft revealed a 6-month primary patency rate of 54%. Our results are similar to those recently reported by Vogel and Parise who achieved a 6-month primary patency of 51% using the

STENTS AND STENT GRAFTS IN HEMODIALYSIS 103 (a) (b) Stent graft Intragraft stenosis Fig. 2. (a) A digital radiograph reveals a Viabahn stent graft crossing the elbow. (b) A fistulogram performed 4 months later demonstrates a patent stent graft and the development of a new stenosis within the hemodialysis graft. Survival 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 aspire SMART Viabahn 0 100 200 300 400 500 600 Days Fig. 3. Graph of primary patency rates for three different stents. nitinol SMART stent (9). In a more recent study, Vogel andpariseachievedevenbetterresultsandreporteda six-month primary patency of 67% when using the nitinol SMART stent for treating venous stenoses (10). These results are substantially better than previous studies which utilized stainless steel stents for the treatment of hemodialysis graft-related stenoses. Patel et al. reported a 6-month primary patency of only 27% when using Wallstents to treat venous anastomotic stenoses following percutaneous thrombectomy procedures (11). Hoffer et al. conducted a randomized, prospective clinical trial comparing angioplasty to the stainless steel Wallstent and reported a 6-month primary patency of only 12% for the 17 patients who received the Wallstent (12). As suggested by Clark, nitinol stents may have advantageous characteristics which provide superior long-term patency when compared with stainless steel stents (3). The current study revealed that the primary patency of the stent or stent graft was better than the primary patency of the entire vascular access. Our retrospective review of the diagnostic fistulograms revealed that in some patients the endpoint of primary patency was the development of a significant stenosis at a new location in the vascular access circuit (Fig. 2). The original lesion that was treated with a stent or stent graft often remained patent and required no additional intervention. Our study suggests that the subgroup of patients who received nitinol SMART stents had a superior primary patency rate at 6 months when compared with those patients who received the PTFE-encapsulated Viabahn stent graft. However, these results may be skewed by the small numbers of patients in each subgroup. The 3 and 12 month primary patency rates for those patients who received SMART stents were similar to the patency rates of those patients who received the Viabahn stent graft. Vascular stents can be beneficial for the treatment of angioplasty-induced venous rupture. In the event of a venous rupture the insertion of a stent or stent graft can be performed easily and rapidly. These devices can tamponade the damaged vascular wall and thereby minimize blood loss and perivascular hematoma formation. Previously published studies have demonstrated that metal stents can effectively repair a venous rupture and maintain patency of the graft with three month primary patency rates ranging from 46% to 63% (13,14). In summary, this study has demonstrated that when stents and stent grafts are used to treat angioplasty failures the long-term outcomes are similar to those achieved following a successful angioplasty procedure (1,3,15). Disclosure Dr. Vesely is a consultant for W. L. Gore and Bard Peripheral Vascular. References 1. Gray RJ: Angioplasty and stents for peripheral and central venous lesions. Tech Vasc Interv Radiol 2:189 198, 1999 2. Lorenz JM: Use of stents for maintenance of hemodialysis access. Semin Interv Radiol 21:135 140, 2004 3. Clark TW: Nitinol stents in hemodialysis access. J Vasc Interv Radiol 15:1037 1040, 2004 4. Maya I, Allon M: Outcomes of thombosed arteriovenous grafts: comparison of stents vs. angioplasty. Kidney Int 69:934 937, 2006 5. Sreenarasimhaiah VP, Margassery SK, Martin KJ, Bander SJ: Salvage of thrombosed dialysis access grafts with venous anastomosis stents. Kidney Int 67:678 684, 2005 6. Pan HB, Liang HL, Lin YH, Chung HM, Wu TW, Chen CY, Fang HC, Chen CK, Lai PH, Yang CF: Metallic stent placement for treating peripheral outflow lesions in native arteriovenous fistula hemodialysis patients after insufficient balloon dilatation. AJR 184:403 409, 2005 7. Vesely TM: Use of the aspire covered stent for treatment of hemodialysis graft-related stenoses. J Vasc Interv Radiol 13:S38, 2002

104 Vesely et al. 8. Haskal ZJ, Mietling S, Schuman E, Altman S, Bermann S, McLennan G, Trerotola S, Aruny J, Ross J, Vesely T, Patel N, Gray R, Durham J: Multicenter phase I results of the Bard PTFE stent graft trial for hemodialysis venous anastomotic graft stenoses. JVIR 14:S25, 2003 [Abstract] 9. Vogel PM, Parise C: SMART stent for salvage of hemodialysis access grafts. J Vasc Interv Radiol 15:1051 1060, 2004 10. Vogel PM, Parise C: Comparison of SMART stent placement for arteriovenous graft salvage versus successful graft PTA. J Vasc Interv Radiol 16:1619 1626, 2005 11. Patel RI, Peck SH, Cooper SG, Epstein DM, Sofocleous CT, Schur I, Falk A: Patency of Wallstents placed across the venous anastomosis of hemodialysis grafts after percutaneous recanalization. Radiology 209: 365 370, 1998 12. Hoffer EK, Sultan S, Herskowitz MM, Daniels ID, Sclafani SJ: Prospective, randomized trial of a metallic intravascular stent in hemodialysis graft maintenance. J Vasc Interv Radiol 8:965 973, 1997 13. Rajan DK, Clark TW: Patency of Wallstents placed at the venous anastomosis of dialysis grafts for salvage of angioplasty-induced rupture. Cardiovasc Interv Radiol 26:242 245, 2003 14. Welber A, Schur I, Sofocleous CT, Cooper SG, Patel RI, Peck SH: Endovascular stent placement for angioplasty-induced venous rupture related to the treatment of hemodialysis grafts. J Vasc Interv Radiol 10:547 551, 1999 15. Vesely TM: Endovascular intervention for the failing vascular access. Adv Renal Replace Therapy 9:99 108, 2002