New Device Trials: The United States Multicenter Experience With the Powerlink

Introduction 

Recently approved by the FDA for sale in the United States, and having been employed in Europe since 1999, the Powerlink device (Endologix, Inc.) is a unibody bifurcated endograft for repair of abdominal aortic aneurysms (AAA). Its stented endoskeleton is self-expanding and is covered with a ePTFE fabric. The Powerlink is surgically implanted through one surgically exposed femoral artery and a contralateral 9 Fr puncture. This requirement for only one surgically exposed artery allows the Powerlink to be used in patients with disadvantaged access situations, and prompted investigators to employ local anesthesia in 22% of cases during the pivotal multicenter trial of the Powerlink device. The midterm results of this multicenter trial have been reported recently.¹

The Powerlink U.S. Multicenter Trial

A pivotal multicenter trial of the Powerlink device was performed in patients with infrarenal abdominal aortic aneurysms. The trial included both test patients receiving the Powerlink and concurrent surgical controls. A total of 258 patients (192 test and 66 control) were enrolled. The test and control groups were not statistically different with respect to demographic variables, with the exception that test patients were significantly older than control patients. Overall mean follow-up was 22 months and median follow-up 24 months. Technical success with respect to successful implantation of the Powerlink was achieved in 97.9% of cases. Technical failures included 3 failed insertions and 1 access failure. The primary conclusion of the trial was that the Powerlink device was safe and effective. Importantly, several advantages were noted for the Powerlink recipients when compared with those patients undergoing concurrent open surgical repair of their aneurysms during the perioperative period. There was a lower mortality and serious adverse event rate for Powerlink patients than in the open surgical control patients during the perioperative period. There was a significant reduction in blood loss, operative time, ICU utilization and hospital length of stay (LOS) for Powerlink patients when compared with open surgical controls. This offers the potential of cost savings and improved patient satisfaction for endovascular aneurysm repair (EVAR) with the Powerlink device. In addition, Powerlink recipients experienced significantly fewer myocardial infarctions, less renal failure and fewer gastrointestinal complications than the open surgical control cohort. Secondary interventions were more commonly required in the Powerlink cohort than in the open surgical control patients, with endoleaks being the most prevalent necessitating course. Overall, Kaplan-Meier analysis revealed a one-year endoleak-free survival of 78%. All but 2 of the endoleaks were of type II and only 6.2% of patients required a secondary intervention for treatment of endoleak, with most interventions occurring within the 30-day perioperative period. Importantly, there were no type III or IV endoleaks noted. Since the Powerlink is a unibody rather than a modular bifurcated graft, there is no risk of component separation producing later endoleak as the AAA shape changes with time. There is no type IV endoleak noted at the time of implantation, as the ePTFE material is non-porous. There were no incidences of fabric defects or erosions, which have been problematic for a number of other devices.^2,3 Over the course of the trial, three devices were explanted (2 at the time of autopsy after death from causes unrelated to AAA and one during conversion to open repair for treatment of a refractory type I endoleak). In each of these there was noted to be no evidence of material fatigue in either stents or the ePTFE fabric, in up to 24 months of patient use. Wire fractures have been observed with frequency in nitinol endovascular grafts,^{3, 4} but have never been observed with the Powerlink. At explant, only minimal wear was observed in the cobalt chromium alloy stents. Durability of the fabric and stents may in part be attributed to the single wire woven stent design and the absence of any suture holes in the ePTFE fabric, except only at its ends. Cuff usage was frequent in the pivotal trial, for a number of reasons. The Powerlink has a long body (80 or 100 mm), prompting a preferred strategy of implantation of the device at or near the aortic bifurcation, rather than at the level of the renal arteries. This provides a hedge against potential graft migration, and necessarily includes placement of a proximal aortic cuff to achieve the infrarenal seal. The cuffs themselves are long lengths (55 mm) providing generous overlap, avoiding the possibility of type III endoleak. This latter approach has been preferred in Europe. In addition, cuffs were placed to remedy proximal attachment endoleaks noted at the time of infrarenal Powerlink deployments. The addition of a 34 mm diameter proximal cuff should allow treatment of wider AAA necks (up to 32 mm diameter), as is currently performed with this device in Europe. The 34 mm cuff is expected to begin U.S. trials soon. Significant reduction in aneurysm diameter and volume was noted during follow-up at every time-interval examined. AAA volume reduction was noted in 64% of patients by 24 months. Sac expansion was noted in only 2% of patients at 24 months of follow-up. While endotension has been common with another EVAR device which employs PTFE fabric,⁵ it appears not to be prevalent with the Powerlink device. Remodeling of the excluded AAA sac appears to be strongly influenced by the presence of the Powerlink device. Recently, analysis of the Powerlink trial patients has demonstrated a tendency for the aorta treated with the device to straighten over time.⁷ This straightening phenomenon was noted in over 80% of patients. It is thought that the sac is shrinking around the Powerlink device. Owing to the columnar strength of the Powerlink, kinking is not introduced into the graft as the sac regresses. Graft migration (> 10 mm), frequently observed with other EVAR grafts^2,6 was noted in only 2.2% of patients. None of the observed migrations have resulted in an endoleak. Resistance to migration is aided by the single wire design and long main-body of the graft, adding columnar strength which resists migration. As mentioned above, the long main body allows placement of the graft on the aortic bifurcation, making migration unlikely. Thrombosis of endograft limbs was noted in only 1.0% of Powerlink limbs. The observed cases were felt to be due to extrinsic compression by underlying iliac disease. It is recommended that iliac lesions be treated by balloon angioplasty prior to Powerlink implantation to avoid difficulties with graft compression and delivery system retrieval. Special considerations apply to Powerlink patients when graft limb thrombosis does occur, owing to the presence of a stent graft endoskeleton. Thrombolytic therapy is preferable to catheter thrombectomy, as the latter may be complicated by the presence of the endoskeleton.

Conclusion

The Powerlink system safely and effectively protects patients from AAA rupture over the intermediate term. The endograft’s ePTFE fabric covering and cobalt chromium alloy stents have thus far been free from failure and fatigue. The requirement for only one surgically exposed femoral artery, with percutaneous contralateral limb deployment, facilitates graft placement in patients with disadvantaged access routes. Powerlink patients experienced significantly fewer major adverse events than open surgical control patients. Careful follow-up over the longer term is necessary to assure the durability of these results. jeffrey.carpenter@uphs.upenn.edu

Editor's Note This report on the clinical results of an FDA pivotal trial with the Powerlink AAA stent-graft is significant because it describes rather favorable patient outcomes with a device whose design is fundamentally different from other endografts currently available in the U.S. However, VDM readers must be warned these data relate to intermediate-term outcomes only, and that much remains unknown (or unstudied) vis-a-vis long-term device performance and integrity. We will look forward to reviewing such information when it becomes available in the future. Frank J. Criado, MD frank.criado@MedStar.net