Anatomic exclusion from endovascular repair of thoracic aortic aneurysm

Study design 

We identified all patients evaluated for endovascular repair of isolated TAA at the Hospital of the University of Pennsylvania between 2000 and 2004. Patients with aneurysmal segments proximal to the left common carotid artery (LCCA) or distal to the celiac axis were not included in this study. All patients in whom TEVAR was indicated underwent cross-sectional imaging by computed tomography angiogram and subsequent three-dimensional modeling by Medical Metrix Solutions (MMS, West Lebanon, NH). All patients were evaluated for either the Gore TAG endoprosthesis (W. L. Gore & Associates, Flagstaff, Ariz) or the Medtronic Talent graft (Medtronic, Minneapolis, Minn), or both. Patients were excluded from TEVAR repair on the clinical judgment of the surgeons participating in the trials, who were familiar with the anatomic exclusion criteria defined by the controlled clinical trials sponsored for the two devices (Table I). After each patient’s anatomy was analyzed, the excluding features were grouped according to broad categories: hostile proximal neck, hostile distal neck, or inadequate access vessels.

Table I. Exclusion criteria for sponsored, controlled clinical trials

	Gore TAG	Medtronic Talent
	⁎Nonaneurysmal segment at either neck <2 cm	⁎Nonaneurysmal segment at either neck <2 cm
	Diameter at either neck <23 mm or >37 mm	Diameter at either neck <18 mm or >42 mm
	Occluded LCCA or CA	Clearance at LCCA or CA <20 mm
	“Significant thrombus” at either neck	“Significant mural thrombus” at either neck
	Proximal neck angulated <60°	Access vessel “precludes safe insertion of delivery system
	Taper at either neck > 4 mm

LCCA, Left common carotid artery; CA, celiac artery.

⁎ Including, where necessary, the segment involving the takeoff of the left subclavian artery.

Outcomes and follow-up 

All patients initially rejected for TEVAR repair of a TAA for anatomic reasons were followed up to assess clinical outcome and eventual open or endovascular repair.

Statistical analysis 

All statistical calculations were performed by using SPSS software (SPSS Inc, Chicago, Ill). The Fisher exact test with a two-tailed assumption was used to assess for gender-based differences in rejection for TEVAR and for gender-based differences in the anatomic features leading to exclusion. Data are presented as mean ± standard deviation, unless otherwise indicated. Kaplan-Meier statistics were used to analyze survival of the untreated patients.

Results 

Between 2000 and 2004, 126 patients (73 men and 53 women) with isolated TAA were evaluated (P = .05), and 33 (26%; 16 men, 17 women, P = .22) were rejected for TEVAR on anatomic grounds. These patients were generally >70 years old (25, 75%), their aneurysms were of significant size, 6.4 ± 1.4 cm, and most (85%) were excluded for meeting more than one criterion (mean, 2.5 ± 1.2 criteria).

Clinical outcomes for those patients excluded from TEVAR are summarized in Table II. Of the 33 patients excluded, 25 did not undergo repair of their aneurysms at our institution owing to some combination of relatively small aneurysms, medical unsuitability for open operation, and patient unwillingness to undergo open repair. The exclusion criteria were grouped according to category: 28 patients had hostile proximal neck anatomy, 19 had hostile distal neck anatomy, and 10 had access anatomy making TEVAR difficult or impossible. Both the specific excluding anatomic features and their broad categorization are presented in Table III.

Table II. Clinical outcomes of the 33 patients excluded from endovascular stent grafting

	Outcome	n
	TEVAR after LCCA-LSCA bypass	2
	TEVAR with partial occlusion of CA	1
	Open repair with left atrial-femoral artery bypass	3
	Open repair with hypothermic circulatory arrest	2
	Not repaired at this institution	25

TEVAR, thoracic endovascular aneurysm repair; LCCA, left common carotid artery; LSCA, left subclavian artery; CA, celiac artery.

Table III. Rejection criteria for all 33 patients, in compendium

	Criteria	n	Category	n
	Short proximal neck	16	Proximal neck anatomy	28
	Wide proximal neck	10
	Excessive proximal taper	6
	Angulated proximal neck	2
	Proximal neck thrombus	9
	Proximal neck calcification	2
	Short distal neck	7	Distal neck anatomy	19
	Wide distal neck	5
	Excessive distal taper	1
	Distal neck thrombus	13
	Narrow or tortuous iliac arteries	9	Access anatomy	10
	Iliac occlusive disease	1
	Narrow aorta	1

There was no gender-based difference in proportion of aneurysms with hostile proximal necks (13 men and 15 women, P = .66), hostile distal necks (11 men and 8 women, P = .29), or both (9 men and 7 women, P = .49). Women had a tendency toward being more likely to have difficult access anatomy (3 men and 7 women, P = .25), but this difference did not reach statistical significance.

Excluded patients were then categorized by excluding anatomic feature(s). This analysis is best depicted in a Venn diagram (Fig 1). These different patient groups were analyzed according to clinical outcome. Three groups were of particular interest, and their anatomic characteristics and clinical outcomes are discussed here.

View Large Image Download to PowerPoint

Fig 1. A Venn diagram demonstrates the clustering of patients according to exclusion criteria category. Note that fully 28 of 33 patients were excluded at least partly because of an anatomically hostile proximal neck.

First, the group excluded solely because of proximal neck anatomic features comprised nine patients with proximal descending TAAs. One of these patients underwent open repair with hypothermic circulatory arrest for an enlarging symptomatic aneurysm.

Second, 12 patients comprising a cohort of patients with type I thoracoabdominal aortic aneurysms were excluded for a combination of proximal neck anatomy and distal neck anatomy,22 and their clinical outcomes are summarized in Table IV. One patient underwent TEVAR and required subsequent celiac axis stent placement because the thoracic aortic graft had partially covered the celiac ostium.

Table IV. Clinical outcomes of 12 patients excluded from thoracic endovascular aneurysm repair for both hostile proximal and hostile distal necks

	Outcome	n
	TEVAR repair after LCCA-LSCA bypass	1
	TEVAR repair with partial occlusion of CA	1
	Open repair with left atrial-femoral artery bypass	1
	Refused open repair	1
	Not repaired	8

TEVAR, thoracic endovascular aneurysm repair; LCCA, left common carotid artery; LSCA, left subclavian artery; CA, celiac artery.

Third, those patients excluded solely because of distal neck anatomic features represented patients with type V thoracoabdominal aortic aneurysms, arising at or below the 6th intercostal space and ending at or above the renal arteries.22 Two patients were in this category. One underwent open repair with left atrial-femoral arterial bypass for a large expanding aneurysm at our institution. The other underwent open repair at another institution for a symptomatic aneurysm.

Two patients who were initially excluded from TEVAR eventually underwent carotid–subclavian bypass and endografting (Table II). One was initially excluded because of a hostile proximal and distal neck anatomy, and the other was initially excluded because of narrow and calcified access vessel anatomy as well as hostile proximal neck anatomy. In the latter patient, access at operation was obtained through the left common iliac artery; at the conclusion of the procedure, this artery was narrowed at the site of arteriotomy repair and required stenting. Also as indicated in Table II, one patient underwent TEVAR after initially being excluded because of a hostile distal apposition site. At operation, the celiac artery ostium was unintentionally partially covered with the aortic stent, requiring celiac artery stenting at the conclusion of the procedure.

The mean size of the TAAs in the 25 untreated patients was 6.3 cm (range, 4.0 to 9.6 cm). Survival of these untreated patients was analyzed using Kaplan-Meier statistics (Fig 2). As discussed, one patient underwent repair for symptoms at another institution and was censored from the analysis on the date of the surgery. Twelve patients died, and at least 3 of these deaths were attributed to aneurysm rupture. The mean survival was 3.8 years (95% confidence interval [CI], 2.7 to 4.9 years). In contrast, mean survival of the first 186 TAAs repaired endovascularly in an elective fashion at the same institution was 4.7 years (95% CI, 4.2 to 5.3 years) (Dr Wilson Y. Szeto, unpublished data, 2006).

View Large Image Download to PowerPoint

Fig 2. Kaplan-Meier survival for those 25 patients who did not undergo operative repair at our institution.