We have recently described severe pulmonary hypertension as a cause of reverse aortic arch flow (RAAF) in the neonate (J Amer Soc Echo 1996; 9:915-7). This study was designed to evaluate the incidence of RAAF in neonates with pulmonary hypertension and to compare these patients with those with similar diagnosis and normal aortic flow. Medical records from 102 newborn infants with pulmonary hypertension referred to a tertiary neonatal intensive care unit from August 1994 to April 1996 were reviewed. Twenty seven patients were excluded. The 75 infants entering the study were divided in 2 groups: Group 1 with 68 patients with normal aortic arch flow, and, Group 2 with 7 patients with RAAF. There were no differences in prenatal complications(maternal diabetes, infection, maternal hypertension, maternal drug abuse), method of delivery, sex, gestational age, Apgar score and diagnosis(diaphragmatic hernia, meconium aspiration, sepsis). All infants underwent echocardiographic examination on admission and all had structurally normal hearts; cerebral arteriovenous malformation was ruled out by cranial ultrasound. Echocardiograms were reviewed for anatomy, aortic arch flow, left ventricular output, and flow at ductus arteriosus and foramen ovale. Infants in Group 2 were significantly younger at the time of admission, required significantly higher ventilatory parameters on conventional or high frequency ventilation, had significantly lower PaO2 regardless of the type of ventilation, and had significantly lower echo derived left ventricular output. There was no significant difference on the need for ECMO therapy: 76% in Group 1 vs. 86% in Group 2. Both, the overall survival (Group 1 81%, Group 2 43%) and the survival after ECMO (Group 1 81%, Group 2 30%), were significantly lower for infants in Group 2.

CONCLUSIONS: RAAF in newborn infants with pulmonary hypertension, in absence of other causes of flow reversal, indicates reduced left ventricular output. This group of patients tend to be more difficult to oxygenate and to have higher mortality rate regardless of ECMO.