Branch pulmonary artery stenosis after arterial switch operation: The effect of preoperative anatomic factors on reintervention

Background: We hypothesized that preoperative patient characteristics and branch pulmonary artery (PA) size might influence the rate of postoperative branch PA reintervention in patients with transposition of the great arteries who undergo the arterial switch operation (ASO). Methods: The retrospective single-center study included 262 consecutive (2008-2017) newborns who underwent the ASO. Demographic characteristics, echocardiography, and clinical outcomes were reviewed. Competing risk analysis modeled incidence of branch PA reintervention and cause-specific hazard regression for predictors analyses. Results: Median age and weight were 7 (range, 5-11) days and 3.4 (range, 3.1-3.8) kg, respectively. Various types of early branch PA reinterventions (concomitant revision or reintervention during the intensive care unit stay) were required in 28 (10.7%) patients. These patients had prolonged ventilation (P < .001), intensive care unit duration (P < .001), worse right ventricular function (P = .043), and high in-hospital mortality (P = .010). Branch PA dimensions significantly decreased immediately after ASO compared with baseline measurements. The median follow-up duration was 20.8 (range, 0.9-44.7) months. Branch PA reintervention was common among survivors without early reinterventions (9.4%), and even more frequent among those with early reinterventions (25%). Subsequent reintervention (all catheter-based) was necessary for more than one-third of patients after initial branch PA reintervention. The multivariable analysis showed preoperative dimension of the left PA (hazard ratio, 0.527 [95% CI, 0.337-0.823]; P = .005), and right PA (hazard ratio, 0.503 [95% CI, 0.318-0.796]; P = .003) were independently associated with late branch PA reinterventions. Conclusions: Branch PA reintervention was common and often required surgical or catheter-based reinterventions after ASO. PA branch diameters became significantly smaller after ASO. Smaller preoperative branch PA predicted late branch PA reintervention, indicating a smaller margin of geometrical tolerance to this effect.