Aortic Arch Advancement and the Impact of Vocal Cord Dysfunction

Background: Vocal cord paresis (VCP) is a known complication of neonatal aortic arch repair and is associated with prolonged tube feeding and mechanical ventilation. The postoperative incidence of VCP after aortic arch advancement is unknown. We aimed to determine the incidence of VCP after aortic arch repair via arch advancement compared to other arch repair methods, and characterize the comorbidity and natural history of VCP.

Methods: A retrospective analysis was performed of neonates who underwent arch repair via median sternotomy at Lurie Children’s Hospital from 2010-2019. Patients were stratified by Single ventricle (SV) physiology or biventricular (BiV) physiology. Our primary outcome of VCP was identified by direct visualization of paretic vocal cord(s). Covariates included age and weight at time of surgery. Logistic regression models compared the incidence of VCP between patients who underwent aortic arch repair by arch advancement, patch or Norwood augmentation, and extended end-to-end anastomosis.

Results: A total of 154 patients (74 SV, 80 BiV) met inclusion criteria for the study. Median age at time of surgery was 6 days (IQR 5-9) and mean weight at time of surgery was 3.2 kg (SD 0.7). Within the BiV group, 36 subjects underwent aortic arch advancement (45%) whereas in the SV group, only 1 subject (1.4%) underwent arch advancement. VCP was found in 59 (23 BiV, 36 SV) patients (38.3%). Adjusted odds of VCP with aortic arch advancement is 5.41 times that of patients with other methods (p < 0.001) when controlled for both patient weight and age at time of surgery in the BiV group. Due to small sample size, no associations with repair type could be made in the SV group. Tube feeding at discharge was needed in 54/74 (72.9%) of the SV patients and 52/80 (65%) of the BiV patients. Median time to full oral feeds was 230 (IQR 49.5-465.5) for the SV group and 38.5 (IQR 12.8-109.5) days for BiV group. There was no significant difference in aortic re-intervention rates when comparing between arch repair methods (14% arch advancement, 27% other approaches, p=0.1).

Conclusion: Patients with biventricular physiology that undergo aortic arch advancement are more likely to have vocal cord paresis than patients who undergo repair via extended end-to-end anastomosis or patch aortoplasty. With further data on the effects of vocal cord paresis on feeding, breathing, and growth, as well as time to full oral feeds and resolution of VCP, providers will be able to use this data to more comprehensively counsel patient families. Further studies will determine if we can alter the natural history of VCP after arch repair with speech therapy or other interventions to improve quality of life and morbidity burden for these patients.