Prolonged postoperative respiratory support after proximal thoracic aortic surgery: Is deep hypothermic circulatory arrest a risk factor?

Purpose: In addition to the pulmonary risks associated with cardiopulmonary bypass, thoracic aortic surgery using deep hypothermic circulatory arrest (DHCA) may subject the lungs to further injury. However, this topic has received little investigation to date.
Materials and Methods: A prospective cohort review was performed on all patients undergoing proximal thoracic aortic surgery with (n = 478) and without (n = 135) DHCA between July 2005 and February 2013 at a single institution. The primary outcome was prolonged postoperative respiratory support (PPRS), defined as any of the following: >1 day of mechanical ventilation at either fraction of inspired oxygen >0.4 and/or positive end-expiratory pressure >5 mm Hg, >2 days of supplemental O2 requirement of at least 2.5 L/min, or discharge with new O2 requirement. Independent risk factors for PPRS were identified using multivariable logistic regression.
Results: Postoperative respiratory support was required in 100 patients (20.9%) with and 30 patients (22.2%) without DHCA (P = .74). Independent predictors of PPRS after proximal aortic surgery included the following: age, diabetes, history of stroke, preoperative creatinine, American Society of Anesthesiologists class 4, redo-sternotomy, total arch replacement, and transfusion requirement. Use of DHCA was not an independent risk factor for PPRS in the entire cohort. Subanalysis of only DHCA patients revealed that longer DHCA times were independently associated with PPRS.
Conclusions: Prolonged postoperative respiratory support is common after proximal aortic surgery. The use of DHCA was not associated with this complication in the overall cohort, although longer DHCA times were predictive when only the subset of patients undergoing DHCA was analyzed. Knowledge of the risk factors for PPRS after proximal aortic surgery should improve preoperative risk stratification and postoperative management of these patients.
Competing Interests: ☆☆ Financial/nonfinancial disclosures: The authors have no relevant conflicts of interest to disclose. ★ Conflict of interest: No conflicts of interest exist for any of the authors of this manuscript.
(Copyright © 2015 Elsevier Inc. All rights reserved.)