Initial cannulation strategy impacts perioperative outcomes of acute type A dissection in high-volume centers.

OBJECTIVE

We performed an intention-to-treat analysis of initial cannulation strategy to assess the impact on perioperative outcomes in acute type A dissection using multicenter data.

METHODS

All patients undergoing surgical repair of acute type A dissection from a multicenter national registry of 9 high-volume aortic centers were analyzed. Cannulation strategies included in the analysis were axillary, femoral, direct aortic, and innominate. Among 950 patients, we excluded those with chronic syndromes, type B dissections, and unknown initial cannulation strategy. Patients with multiple cannulation strategies were included if the sequence in which strategies were initiated was known. The final cohort consisted of 936 patients. Primary outcomes were stroke and death. Multivariable logistic regression was performed to adjust for baseline differences. P values represent Tukey's post hoc comparisons.

RESULTS

Among 936 patients, cannulation strategies in descending order included axillary (n = 502, 53%), femoral (n = 268, 29%), aortic (n = 104, 11%), and innominate (n = 59, 6%). Of these patients, 46 (5%) had a change in the initial cannulation strategy before initiating circulatory arrest, mainly for poor axillary flow or initial femoral cannulation for hemodynamic instability followed by axillary. Patients in the femoral group were younger (61.3 ± 13.8 years) than patients in the aortic group (66.4 ± 12.52 years, P = .01) and more likely to present with malperfusion (n = 123, 45.9%) compared with patients in the aortic, axillary, and innominate groups (P < .01). Patients in the femoral group also had the longest duration of cerebral ischemia (femoral: 16.9 ± 16 minutes, aortic: 11.5 ± 11.8 minutes; axillary: 4.41 ± 10.3 minutes; innominate: 2.53 ± 6 minutes, P < .01 for all vs femoral). Unadjusted risk of death, stroke, and prolonged ventilation was lowest in the axillary and innominate groups. Length of stay was also reduced in the innominate group. Multivariable regression demonstrated axillary (odds ratio [OR], 0.52; 0.36-0.75; P = .004) and innominate (OR, 0.19; 0.07-0.54; P = .009) cannulation to be associated with a significantly reduced risk of stroke. A nonsignificant indication of reduced death in patients receiving axillary cannulation remained (OR, 0.66; 0.45-0.96; P = .07).

CONCLUSIONS

In high-volume aortic centers, an initial cannulation strategy using axillary access is associated with reduced risk of stroke compared with femoral access. Axillary cannulation should be the preferred strategy in experienced centers if anatomy and stability allow.