PubMed2006
OBJECTIVE: Endovascular repair of abdominal aortic aneurysm (AAA) is complicated by the potential for stent graft migration over time. Factors including the type of fixation, initial proximal fixation length, and dilation and elongation of the infrarenal aortic neck may contribute to device migration. We sought to determine when device migration is a real phenomenon with actual device movement that compromises aneurysm exclusion. METHODS: Computed tomographic (CT) scans and computer reconstructions of all patients undergoing endovascular AAA repair with a passive fixation device at our institution from June 1996 to October 2004 were retrospectively reviewed. The distance from the distal renal artery to the proximal end of the stent graft at the time of initial deployment was determined for each patient. Migration was defined as a distance increase greater than 5 mm in the follow-up period; proximal fixation length, aortic neck enlargement and elongation, and neck angle were then measured. Data were further analyzed with respect to AAA growth, development of endoleak, AAA rupture, and the need for reintervention. RESULTS: A total of 308 patients with endovascular AAA repairs using a passive fixation device had complete postoperative imaging data sets; 48 patients (15.6%) with stent graft migration of 5 mm or more were identified, and 25 (8.1%) of these had a migration of 10 mm or more. Seventeen (35.4%) of 48 migration patients had a total loss of the proximal seal zone (loss patients); their average migration distance was 17.7 +/- 12.0 mm, with a mean neck shortening of 13.6 +/- 14.2 mm, and the average proximal fixation length loss was 14.0 +/- 7.6 mm. Those 31 patients with an intact proximal seal zone (nonloss patients) showed an average migration of 9.4 +/- 3.7 mm, with a mean neck lengthening of 9.6 +/- 8.4 mm and an average proximal fixation length change of 0.7 +/- 8.0 mm. Univariate analysis demonstrated significant differences between the loss and nonloss patients in follow-up duration (65.9 +/- 20.4 months vs 45.9 +/- 26.4 months; P = .01), neck dilatation at the distal renal artery (4.6 +/- 4.5 mm vs 1.8 +/- 1.9 mm; P = .026), stent graft migration distance (17.7 +/- 12.0 mm vs 9.4 +/- 3.7 mm; P = .001), change in aortic neck length (-13.6 +/- 14.2 mm vs 9.6 +/- 8.4 mm; P < .0001), change in proximal fixation length (-14.0 +/- 7.6 mm vs 0.7 +/- 8.0 mm; P < .0001), change in AAA size (1.8 +/- 7.1 mm vs -3.6 +/- 9.7 mm; P = .033), and use of a stiff body stent graft (47.1% vs 19.4%; P = .043). However, only change in aortic neck length was statistically significant on multivariate analysis (odds ratio, 0.75; 95% confidence interval, 0.591-0.961; P = .022). There were no differences between the loss and nonloss patients in time to migration discovery, initial AAA size, initial aortic neck diameter or length, initial device oversizing, initial neck angle, neck angle increase, type II endoleak, or AAA rupture. Eight of the 17 loss patients have been treated with proximal aortic cuffs; the remainder have refused reintervention, died of unrelated causes, or elected to have open repair. CONCLUSIONS: Postoperative elongation of the infrarenal aortic neck may create the radiographic perception of migration without necessarily causing a loss of proximal stent graft fixation. Patients with a total loss of the proximal seal zone actually have infrarenal aortic neck shortening, with a degree of neck dilatation beyond initial device oversizing that may compromise proximal fixation length. Conversely, those with an intact proximal seal zone demonstrate aortic neck elongation equivalent to migration, with no loss of proximal fixation length; these patients have a benign natural history without intervention. Thus, aortic neck dilatation beyond oversizing, aortic neck shortening, and loss of proximal fixation length are more clinically relevant predictors of proximal stent graft failure than simple migration distance.