12. Intervention for Arch Obstruction in the Interstage Period Following Norwood: Prevalence, Risk Factors, and Practice Variability
Paul Joseph Devlin1, Brian W. McCrindle1, Pirooz Eghtesady2, Bahaaladin Alsoufi3, Eugene H. Blackstone4, James M. Meza5, *William M. DeCampli6, *James K. Kirklin7, *Jeffrey P. Jacobs8, *Ali Dodge-Khatami9, *Kristine J. Guleserian10, James E. O'Brien, Jr.11, *Erle H Austin, III12, *Peter J. Gruber13, Tara Karamlou14
1The Hospital for Sick Children, Toronto, ON, Canada; 2St. Louis Children's Hospital, St. Louis, MO; 3Children's Healthcare of Atlanta, Atlanta, GA; 4Cleveland Clinic, Cleveland, OH; 5Duke University Medical Center, Durham, NC; 6Arnold Palmer Hospital for Children, Orlando, FL; 7University of Alabama at Birmingham, Birmingham, AL; 8Johns Hopkins All Children's Heart Institute, St. Petersburg, FL; 9University of Mississippi Medical Center, Jackson, MS; 10Nicklaus Children's Hospital, Miami, FL; 11Children's Mercy Hospital, Kansas City, MO; 12Norton Children's Hospital, Louisville, KY;1 13Children's Hospital of Philadelphia, Philadelphia, PA; 14Phoenix Children's Hospital, Phoenix, AZ

Objective: Arch obstruction after the Norwood procedure is common and it contributes to morbidity and mortality. We analyzed the prevalence, risk factors, and practice variability of intervention for arch obstruction after Norwood in a multicenter cohort of neonates with critical left heart obstruction.

Methods: 
From 2005 - 2017, 593 neonates in the CHSS (Congenital Heart Surgeons’ Society) Critical Left Heart Obstruction cohort underwent a Norwood procedure. Competing risks methodology determined simultaneous risk and associated incremental risk factors for arch intervention and heart transplant or death. Variables analyzed included demographics, baseline echocardiography, and Norwood operative details.

Results: 
Of the 593 infants, 119 (20%) underwent 151 interventions for arch obstruction after Norwood: catheter (n=115) or surgical (n=36) prior to, or during, a stage II procedure at a median age of 3.9 months (IQR: 2.6 - 5.3). Of the catheter procedures, 21 (18%) occurred at pre-stage II catheterization, while 22 (61%) of the surgical aortic repairs occurred during stage II procedure, heart transplant, or biventricular repair. From Norwood, competing risks analysis to first intervention demonstrated the distribution of patients among mutually-exclusive end-points (Figure 1). Intervention for arch obstruction was represented by a single early phase hazard. Interdigitation of the distal aortic anastomosis was protective against arch intervention (p=0.02, reliability: 78%), while coarctectomy in general was not significant. Risk factors for arch intervention in the interstage period included the presence of a native tissue pulmonary artery to aorta anastomosis (p=0.02, reliability: 78%), coronary sinusoids on preoperative echocardiography (p=0.05, reliability=66%), and longer cardiopulmonary bypass time (p=0.01, reliability 67%). Variables that were not significant in the final multivariable model included shunt type, patch type, pre-operative aortic sizes, and aortic atresia. Among institutions, there was variation in the proportion of patients undergoing interstage arch intervention (range: 0 - 46% among the 18 institutions that contributed 5 or more patients to the cohort) and threshold (median pre-intervention gradient: 20.0mmHg, IQR: 9.5 - 30.5, range 2.0 - 62.0, n=68/115) for catheter arch intervention.

Conclusions: 
Interdigitation of the distal aortic anastomosis and inclusion of patch material in the aortic anastomosis during the Norwood procedure decreased the high risk of subsequent aortic obstruction. Serial surveillance for arch obstruction, including assessing changes in systemic RV function and tricuspid insufficiency, is needed to further define the impact of arch obstruction. A standardized definition of arch obstruction is needed to improve analysis of its role in morbidity and mortality during the interstage period.