Increased Flexibility of AneuRx Stent-Graft Reduces Need for Secondary Intervention Following Endovascular Aneurysm Repair

583 Increased Flexibility of AneuRx Stent-Graft Reduces Need for Secondary Intervention Following Endovascular Aneurysm Repair Frank R. Arko, MD; W. Anthony Lee, MD; Bradley B. Hill, MD; Paul Cipriano, MD; Thomas J. Fogarty, MD; and Christopher K. Zarins, MD Division of Vascular Surgery, Stanford University Medical Center, Stanford, California, USA Purpose: To evaluate the impact of a change in the manufacturing of the AneuRx stentgraft on the long-term results of endovascular abdominal aortic aneurysm (AAA) repair. Methods: The first 70 AAA patients treated with the AneuRx stent-graft between October 1996 and December 1998 were reviewed. The early stiff bifurcated design (STIFF) was used in 23 patients (mean age 71.7 9.3 years, range 45 87) and the current flexible bifurcated design (FLEX) in 47 mean age 75.0 7.3 years, range 61 96). Data on patient demographics, aneurysm morphology, technical success, complications, secondary procedures, and outcomes were compared using Kaplan-Meier estimates to evaluate patient survival and freedom from surgical conversion, rupture, and secondary interventions at 6, 12, and 24 months. Results: The 2 groups were equally matched with regard to age, preoperative comorbidities, proximal neck dimensions, and aneurysm diameter. Mean follow-up times were 22.42 11.72 months (range 1 46) for the STIFF cohort and 18.08 6.14 months (range 1 30) for the FLEX (p 0.057). Eleven (48%) of 23 STIFF patients required secondary interventions versus 6 (13%) of 47 FLEX patients (p 0.05). There were no ruptures. At the 24- month interval, survival estimates were 86% for STIFF and 76% for FLEX (p ); freedom from surgical conversion was 100% for STIFF and 97% for FLEX (p ) and freedom from secondary interventions was 18% for STIFF and 90% for FLEX (p 0.05) at 24 months. Conclusions: The AneuRx stent-graft was effective in achieving the primary objective of preventing aneurysm rupture in all patients. However, increasing the flexibility of the bifurcated module significantly improved the primary success rate by reducing the need for subsequent secondary interventions. J Endovasc Ther Key words: endovascular repair, abdominal aortic aneurysm, endoleak, endograft migration, fixation Endovascular abdominal aortic aneurysm (AAA) repair has become an increasingly at- Address for correspondence and reprints: Christopher K. Zarins, MD, Division of Vascular Surgery, Stanford University Medical Center, 300 Pasteur, Rm H3600, Stanford, CA 94305 USA. Fax: 1-650-498-6044; E-mail: zarins@ stanford.edu tractive therapy compared to open surgery, offering patients the benefits of less blood loss, shorter hospital stay, reduced morbidity, and earlier return to function. 1 5 As our experience with endografting grows, the introduction of new technology should result in improved outcomes. However, in evaluating 2001 by the INTERNATIONAL SOCIETY OF ENDOVASCULAR SPECIALISTS

584 ANEURX FLEXIBILITY J ENDOVASC THER stent body, which now consisted of 7 individual 1-cm-long nitinol rings joined together end to end. The segmented body construction enhanced the flexibility of the bifurcated graft throughout its length and eliminated the unbending 5-cm proximal segment. The junction gate markers on the new device were increased by 1.5 cm to achieve a longer landing zone in the iliac limb module (Fig. 1). There was no change to the iliac limb design; the length of the outer nitinol rings measured 9.6 mm for 16-mm-diameter stent rings, 9.8 mm for 15-mm stent rings, 9.9-mm for 14-mm stent rings, and 7.4 mm for 13- and 12-mm stent rings. Figure 1The original stiff configuration (left) with a single 5-cm nitinol ring as compared to the current flexible configuration (right) with individual 1- cm stent-graft rings in the main body. Note the 1.5- cm longer gate of the flexible graft on the right. long-term results, many studies have included patients treated with older prostheses that are no longer available commercially, 6 so that the overall outcome may not reflect the results that can be obtained with newer devices. Therefore, it is important to know if advances in stent-graft design and delivery have translated into improved results. The purpose of this study was to evaluate the impact of a 1998 manufacturing change in the AneuRx stent-graft on the long-term results of endovascular AAA repair. METHODS In April 1998, manufacturers of the bifurcated AneuRx stent-graft (Medtronic AVE, Santa Rosa, CA, USA) changed the design to make the main component more flexible compared to the original stiff bifurcated device. The main module of the initial stent-graft was constructed using a single-unit, 5-cm-long nitinol bifurcated stent coupled proximally and distally to individual 1-cm-long nitinol stent rings joined together end to end (Fig. 1). The change in construction of the endograft involved the entire length of the bifurcated Study Design A database review was conducted to identify all patients treated with the AneuRx bifurcated stent-graft from October 1996 through December 1998 at Stanford University Hospital. Seventy consecutive patients were identified as participants in the ongoing phase-i and II multicenter clinical trial of this device during the observation period. Data on patient demographics, procedural success, complications, conversion, secondary procedures, and outcomes were retrieved. Surgeons had been blinded to the device type selected at the procedure, so the graft lot number was used retrospectively to determine graft type, which was confirmed by postoperative image analysis. Data from computed tomographic angiography (CTA) and 3-dimensional reconstructions were used to ascertain preoperative proximal neck diameter, proximal neck length, aortoiliac tortuosity, and maximum aneurysm diameter and volume. Postoperative measures recorded included stent-graft position relative to the renal arteries, neck diameter, maximum aneurysm diameter and volume, and presence/absence and type of endoleak. Plain abdominal radiographs were used to visualize the position and configuration of the stent-graft. Patients were divided into 2 groups based on the type of stent-graft that was implanted: there were 23 patients (mean age 71.7 9.3 years, range 45 87) who received the early stiff device (STIFF) and 47 patients (mean age

ANEURX FLEXIBILITY 585 TABLE 1 Preoperative Comorbidities in the Patient Groups STIFF FLEX (n 23) (n 47) p Chronic obstructive pulmonary disease 23% 31% Hypertension Congestive heart failure Diabetes Coronary artery disease 51% 23% 17% 66% 57% 14% 14% 54% Cerebral vascular disease Peripheral vascular disease Chronic renal failure Dysrhythmia not significant. 17% 17% 3% 34% 23% 17% 8% 26% 75.0 7.3 years, range 61 96) who received the current flexible device (FLEX). Definitions Primary success was defined as survival with freedom from aneurysm rupture, surgical conversion, and secondary interventions. Secondary success was freedom from aneurysm rupture and surgical conversion. Secondary interventions referred to the placement of proximal and/or distal extender cuffs for any reason after the primary procedure, including repair of type-i endoleaks, graft migration, or inadequate proximal/distal fixation. Procedures to restore graft limb patency were included, but postprocedural angiography and percutaneous coil embolization of collateral branches for type-ii endoleaks were excluded as unrelated to the device change. Statistical Analysis Results are expressed as mean standard deviation. Statistical differences between groups were evaluated with the 2-tailed student t test, log rank, and Fisher exact test as appropriate. Differences were considered significant at p 0.05. Kaplan-Meier estimates were used to evaluate survival, freedom from surgical conversion, freedom from rupture, and freedom from secondary interventions. TABLE 2 Preoperative Aneurysm Morphology in Patients Receiving the STIFF Versus the Current Flexible AneuRx Stent-Graft Neck diameter, mm Neck length, mm Aneurysm diameter, mm Aneurysm volume, STIFF (n 23) 21.8 1.9 22.8 9.2 59.5 8.8 FLEX (n 47) p 22.4 2.2 22.6 9.2 56.6 9.0 0.08 ml 185.6 24.3 197.4 31.3 not significant. RESULTS The 2 groups were equally matched with regard to preoperative comorbidities (Table 1); there were no significant differences in the proximal neck dimensions or aneurysm size (Table 2). The degree of aortoiliac tortuosity in the STIFF group was minimal in the majority (70%, 16/23); 13% (3/23) had moderate and 17% (4/23) had severe tortuosity. In the FLEX group, somewhat more than half of the aortoiliac segments were moderately (30%, 14/ 47) or severely tortuous (23%, 11/47), while 46% (22/47) had minimal tortuosity. At the primary procedure, fewer proximal extender cuffs (9%, 2/23) were required with the STIFF body design compared to the FLEX (17%, 8/47), but more distal extenders were used with the STIFF model (48% [11/23] versus 23% [11/47] with the FLEX). All extender cuffs placed at the primary procedure were to secure proximal and/or distal fixation in both groups. There were no significant differences between the groups with regard to the primary procedure (fluoroscopy time, amount of contrast utilized, or blood loss). There were no deaths in either group at 30 days, and morbidity was low (9%, 6/70) overall. There was no difference in preoperative and the most current mean aneurysm diameters between the designs (Fig. 2) over mean follow-ups of 22.4 11.7 months (range 1 46) for the STIFF group and 18.1 6.2 months (range 1 30) for FLEX (p ). In patients with the STIFF endograft, more than half (57%, 13/23) showed a decrease in aneurysm diameter, and 17% (4/23) had no change. Six

586 ANEURX FLEXIBILITY J ENDOVASC THER Figure 2Scatter plot comparing the preoperative and current aneurysm sizes for the STIFF device (A) and FLEX model (B) at latest follow-up: mean 22.4 months for STIFF and 18.1 months for FLEX. (26%) aneurysms increased; over an average 3.3 years (range 2 4), the mean diameter increase in these 6 patients was 0.8 cm (range 0.4 1.5), for an annual enlargement rate of 0.26 cm/y. In the FLEX group, 42% (20/47) had a decrease in aneurysm diameter and 38% (18/47) showed no change. Nine (19%) aneurysms expanded a mean 0.52 cm (range 0.3 1.4) over a 2.1-year average follow-up (range 2 3); the annual enlargement rate was 0.24 cm/y in this group. Similarly, the preoperative and current aneurysm volumes did not differ appreciably between the 2 groups (Fig. 3). In the STIFF cohort, 30% (7/23) had a decrease and 70% (16/23) had no change; no aneurysm increased in volume. In those patients with the FLEX design, 17% (8/47) had a decrease, 74%

ANEURX FLEXIBILITY 587 Figure 3Preoperative and current aneurysm volumes in patients with the STIFF device (A) versus those with the FLEX device (B).

588 ANEURX FLEXIBILITY J ENDOVASC THER TABLE 3 Endoleaks in Patients Receiving the STIFF Versus the Current Flexible AneuRx Stent-Graft Type I Type II (6 mo) Type III Type IV FLEX (n 47) 4% (2/47) 17% (8/47) 0% 0% STIFF (n 23) p 8% (2/23) 17% (4/23) 8% (2/23) 0% (35/47) had no change, 8% (4/47) increased in volume. There were no differences in the presence of endoleaks between the STIFF and FLEX device (Table 3). In terms of secondary interventions, 11 (48%) STIFF patients required secondary interventions versus 6 (13%) FLEX patients (p 0.05). STIFF patients required a total of 6 proximal extender cuffs to seal endoleaks (n 2), secure proximal fixation (n 2), and for graft migration (n 2). They also needed 6 distal extender cuffs to repair distal type-i endoleaks (n 2), secure distal fixation (n 2), and repair junctional endoleaks (n 2). FLEX patients received 2 proximal and 3 distal extender cuffs to secure fixation; 1 distal extender cuff was implanted to repair a distal type-i endoleak. There was no morbidity or mortality related to any secondary intervention. Kaplan-Meier estimates (Fig. 4) at 6-, 12-, and 24-months for freedom from secondary intervention for STIFF was 82%, 61%, and 18% compared to 100%, 98%, and 90% for FLEX (p 0.05). One FLEX patient underwent surgical conversion during the initial insertion of the device; freedom from surgical conversion was 100% for STIFF and 97% for FLEX patients at all 3 time intervals (p ). There was no aneurysm rupture recorded in either group. Survival estimates (Fig. 5) for STIFF patients were 100%, 86%, and 86% compared to 96%, 94%, and 76% for FLEX patients (p ). DISCUSSION After a decade of experience with aortic endografting, longer-term outcomes are being Figure 4Kaplan-Meier estimates for freedom from secondary interventions following successful endovascular aneurysm repair for STIFF and FLEX grafts.

ANEURX FLEXIBILITY 589 Figure 5Kaplan-Meier estimates for survival in STIFF and FLEX patients. published, and it is not uncommon for them to include first and second-generation designs of the same device. 4,7 In the case of the AneuRx stent-graft, the inclusion of the prototype stiff bifurcated graft with the new flexible grafts in studies of endograft performance 4 may not accurately reflect the true results using the commercially available device. While it is assumed that the introduction of new design technology will lead to improved outcome, this is not necessarily the case. When comparing the outcome of different stent-grafts, it is important to determine that the patient groups are equally matched with regard to comorbidities and aneurysm morphology, which was the case in our study. Any adverse outcomes will thus likely be related to the graft itself. There was only one difference between the groups, the degree of aortoiliac tortuosity. There were a significantly greater number of patients in the FLEX group who were considered to have moderate to severe aortoiliac tortuosity as compared to the STIFF patients. Thus, the FLEX patients had less favorable anatomy and may be expected to have worse outcomes. However, despite more favorable anatomy in the STIFF patients, there were a greater number of secondary interventions compared to the FLEX cases over an equivalent period of time. There was no difference between the grafts in overall incidence of endoleak. However, there was a general trend for the more flexible graft to have fewer type-i and III endoleaks, which require prompt endovascular or open surgical repair to achieve aneurysm exclusion. The incidence of type-ii endoleaks was the same in both groups, which is not surprising since retrograde flow is not related to the stent-graft but to patent side branches. The flexibility of the graft did not have an effect on long-term aneurysm morphology. In both groups, aneurysm diameter and volume remained unchanged compared to the preoperative measures. However, in patients with the STIFF device, those with the greatest aneurysm shrinkage and change in morphology were most likely to require proximal and distal extender cuffs. This was not seen with the flexible device. The stiff design was not

590 ANEURX FLEXIBILITY J ENDOVASC THER able to conform to the dynamic changes of the aneurysm over time, which increased the likelihood of graft migration with loss of proximal or distal fixation in these patients. The greatest difference between the stentgraft groups was the need for secondary interventions. Nearly half of the patients with the STIFF design needed a secondary procedure to secure either proximal and/or distal fixation compared to only 13% of patients with the flexible stent-graft design. However, it is notable that this study was comprised of the first 70 procedures at our institution, which included the learning curve with the AneuRx device. This may have affected the incidence of secondary interventions as compared to a more recent subset of patients. However, both grafts were placed during the initial learning curve so that it is unlikely that this alone accounts for the difference in secondary interventions. May et al. 8 reported that the adverse event rates were similar in patients treated early in their experience as compared to a later period. Thus, the need for a secondary intervention appears to be an inherent risk of endovascular repair rather than a complication that occurs during the initial learning phase of a new technique. Weaknesses of this study include the fact that the first 23 patients received the stiff bifurcated design, which in itself may have been influential, particularly as regards patient selection. Also, patients who received the stiff endograft had longer follow-up, which could well have an impact on the incidence of secondary interventions. However, the 4-month difference in mean follow-up was small and not significant. Furthermore, the relatively small sample size in both groups raises the possibility of a type-2 error in the statistical analysis. According to our data on perioperative mortality and survival, endovascular aneurysm repair is a safe procedure in our experience, regardless of the structural modifications of the prostheses. Freedom from surgical conversion between patients receiving the stiff and flexible graft was similar. Only one patient in the flexible group required surgical conversion at the original procedure, but this was not related to the graft itself, as the delivery system and deployment of the two grafts were identical. The conversion occurred because a severely angled proximal neck blocked access. Better preoperative recognition of the excessive tortuosity would have indicated open repair as the more appropriate therapy from the beginning. The primary objective of endoluminal therapy is to prevent aneurysm rupture and death. Both stent-graft designs were successful in this regard. Zarins et al. 9,10 have demonstrated that the risk of aneurysm rupture following endovascular repair in the US AneuRx multicenter clinical trial was higher in patients receiving the stiff body design (3.4%) than in patients receiving the flexible stentgraft (0.4%). It was felt that the inability of the stiff module to flex in tortuous or shrinking aneurysms exacerbated the angulation forces at the proximal and distal ends of the graft, resulting in fixation failure. Review of the causes of aneurysm rupture using the AneuRx stent-graft revealed that each instance of aneurysm rupture was related to insecure fixation; moreover, evidence of insecure fixation was apparent on postprocedural CT and abdominal radiographs months and years prior to rupture. 11 These previous studies did not look at the need for secondary procedures required after the initial procedure, which is significantly higher with the stiff device according to our data. Therefore, physicians following patients who have received the early stiff design should be cognizant of this potential for secondary intervention and maintain vigilant surveillance. In conclusion, the AneuRx stent-graft was effective in achieving the primary objective of preventing aneurysm rupture in all patients in this study. However, increasing the flexibility of the main module significantly improved the primary success rate by reducing the need for subsequent secondary interventions. References 1. Moore WS, Rutherford RB. Transfemoral endovascular repair of abdominal aortic aneurysm: results of the North American EVT phase 1 trial. J Vasc Surg. 1996;23:543 553. 2. White GH, Yu W, May J, et al. Three-year experience with the White-Yu Endovascular-GAD Graft for transluminal repair of aortic and iliac aneurysms. J Endovasc Surg. 1997;4:124 136.

ANEURX FLEXIBILITY 591 3. Blum U, Voshage G, Lammer J, et al. Endoluminal stent-grafts for infrarenal abdominal aortic aneurysms. N Engl J Med. 1997;336:13 20. 4. Zarins CK, White RA, Schwarten D, et al. AneuRx stent graft versus open surgical repair of abdominal aortic aneurysms: multicenter prospective clinical trial. J Vasc Surg. 1999;29: 292 308. 5. White GH, May J, McGahan T, et al. Historic control comparison of outcome for matched groups of patients undergoing endoluminal versus open repair of abdominal aortic aneurysms. J Vasc Surg. 1996;23:201 212. 6. Harris PL, Vallabhaneni SR, Desgranges P, et al. Incidence and risk factors of late rupture, conversion, and death after endovascular repair of infrarenal aortic aneurysms: the EUROSTAR experience. J Vasc Surg. 2000;32:739 749. 7. Stelter WJ, Umscheid T, Ziegler P. Three-year experience with modular stent-graft devices for endovascular AAA treatment. J Endovasc Surg. 1997;4:362 369. 8. May J, White GH, Waugh R, et al. Comparison of first- and second-generation prostheses for endoluminal repair of abdominal aortic aneurysms: a 6-year study with life table analysis. J Vasc Surg. 2000;32:124 129. 9. Zarins CK, White RA, Moll FL, et al. The AneuRx stent graft: four-year results and worldwide experience 2000. J Vasc Surg. 2001; 33:S135 145. 10. Zarins CK, White RA, Fogarty TJ. Aneurysm rupture after endovascular repair using the AneuRx stent graft. J Vasc Surg. 2000;31:960 970. 11. Zarins CK, White RA, Hodgson KJ, et al. Endoleak as a predictor of outcome after endovascular aneurysm repair: AneuRx multicenter clinical trial. J Vasc Surg. 2000;32:90 107.