Stent-graft migration: a reappraisal of analysis methods and proposed revised definition

To demonstrate the need for a radiographic definition of migration that accurately describes a specific failure mode of an aortic stent-graft.

The diagnosis of endograft migration, as defined by the Society for Vascular Surgery/American Association for Vascular Surgery (SVS/AAVS) standards, requires a synthesis of clinical and/or radiographic observations. Radiographic studies and clinical reports of 704 consecutive patients treated over a 6-year period with abdominal aortic endografts were retrospectively reviewed. According to the current SVS/AAVS standards, 25 patients were identified as having endograft migration. Follow-up computed tomographic scans and radiographs available for 24 of these patients were scrutinized from discharge to the time of any observed proximal or distal fixation system movement based upon the discharge or 30-day CT scan. Proximal migration was defined with respect to the origin of the superior mesenteric artery (SMA) as movement in a caudal or cranial direction of >/=10 mm or >2 times the reconstructed resolution of the imaging study (whichever measurement was less). Distal migration was similarly defined using the aortic bifurcation and respective hypogastric artery origin as reference points. In an effort to assess the scan-to-scan variation, the distance between the SMA and lowest renal artery, which was expected to remain consistent, was measured.

Film analysis with application of the revised migration definition confirmed fixation system failure with respect to the native arterial system in 12 (50%) of the 24 patients. Subjects judged to have endograft migration according to the reporting standards but not to have radiographic evidence of migration based on the modified criteria included 2 proximal endoleaks without evidence of device movement treated with proximal extensions, 1 procedure-related migration, 2 type III endoleaks treated with a second prosthesis implanted within the first, 1 distal endoleak treated with a limb extension, 1 rupture with presumed distal limb migration, and 2 cases of component separation. Three limb extensions were placed in the absence of leak or migration. These 12 patients all had radiographic evidence that the proximal and distal aspects of the originally implanted device did not move with respect to the native arterial vasculature, thus confirming stability of the respective fixation system.

Adherence to this comprehensive definition of device migration, which differs from the published reporting standards, allows differentiation of the specific mechanisms of device failure. Data viewed in this context will aid in the understanding of device strengths and weaknesses, potentially improve patient assessment, and encourage design modifications to address specific aspects relating to fixation failure.