Incidence and outcomes after infolding or collapse of thoracic stent grafts

Objective: Device-related complications in the thoracic aorta are partly due to the unavoidable proximal angulation and increased flow-related forces. The present study evaluated the incidence, predictors, and outcome of the complication of infolding with the GORE TAG thoracic endoprosthesis (TAG device) to better understand the factors that might help predict these events. Methods: We reviewed all complaints reported to W. L. Gore and Associates (Flagstaff, Ariz) related to device infolding after the use of the GORE TAG device on or before December 2008. Events related to device infolding were evaluated. Reporting physicians and local company representatives were contacted, when necessary, to assemble all available imaging, data, and outcomes related to these case reports. When available, computed tomography images were reviewed to confirm aortic landing zone diameters, which were subsequently compared with the implanted device size. Results: From 1998 through December 2008, device infolding was reported in 139 patients (mean age, 40 +/- 17 years; 73.4% men) from 33,289 device implants (reported incidence, 0.4%). Events were noted in implants for trauma (60%), dissection (19%), aneurysm (10%), and other (9%) and unknown (2%) etiologies. In 77 patients with available imaging, the average minimum aortic diameter was 21.4 +/- 4.4 mm. The mean device diameter was 28.5 +/- 3.5 mm, with an average oversizing of nearly 33%. Of reported patients, 51% were asymptomatic, with the diagnosis being made on routine chest imaging. Time to diagnosis was 76 +/- 222 days (median, 9.5 days). Only 16 patients received no intervention after the diagnosis of device infolding, all of whom were asymptomatic. The other 123 patients underwent 135 interventions. Of these, 30 patients (24%) underwent open surgical conversion and complete or partial endograft removal. The other interventions included a variety of endovascular techniques, such as large balloon-expandable stent(s) in 40%, relining with additional endograft(s) in 31%, and repeat ballooning in seven patients. Ten patients died after device infolding, all after one or more attempts to repair the infolded device: five died of symptoms related to the infolding and five secondary to the intervention undertaken to correct the device infolding. Conclusions: TAG device infolding appears to be an infrequent event, primarily occurring in young trauma patients secondary to excessive oversizing and severe proximal aortic angulation. However, there clearly exists a need for devices that treat such patients. As a result, future device designs should consider the compression failure mode when being designed in order to help prevent such events. (J Vasc Surg 2012; 55: 652-8.)