OBJECTIVES: Neonates with hypoplastic left heart syndrome have significant hemodynamic threats to cerebral perfusion and are at risk of reduced neurodevelopmental performance. We hypothesized that cerebral hypoxia, detectable by near-infrared spectroscopy in the early postoperative period, would be related to later neurodevelopmental performance. METHODS: The study population was a sequential cohort of patients who had undergone stage 1 palliation of hypoplastic left heart syndrome under standard conditions, including neonatal perioperative monitoring with cerebral near-infrared spectroscopy, and who had undergone a neurodevelopmental assessment at age 4 to 5 years. The neonatal demographic and 48-hour perioperative hemodynamic parameters, including cerebral oxygen saturation, were tested for their relationship to 4 domains of neurodevelopmental performance, including visual-motor integration in childhood in univariate and multivariate models. The neurodevelopmental scores were classified as low if less than 85 (-1 standard deviation) and abnormal if less than 70 (-2 standard deviations). RESULTS: For the 51 patients in the surgical cohort, the early survival was 94%, the cumulative survival was 86%, and the neurodevelopmental assessment was completed by 21 (48%) of the survivors, without evidence of an ascertainment bias. At the test age of 56.3 ± 5.5 months, the composite neurodevelopmental index, constructed from equally weighted measures in 4 domains, was 97.6 ± 9.6, not different from the age-based norms, with 3 of 21 in the low range and none abnormal. The mean visual-motor integration was 93.4 ± 14, slightly less than the population norm (P < .05), with 2 of 21 having low scores and 1 abnormal scores. In patients with low to abnormal visual-motor integration, the perioperative stage 1 palliation cerebral oxygenation saturation was significantly lower (63.6 ± 8.1 vs 67.8 ± 8.1, P < .05). Two patients had discrete embolic strokes after their initial hospitalization; the occurrence of late stroke reduced the visual-motor integration performance but was not related to the early cerebral oxygen saturation. Nonlinear relationships of cerebral oxygen saturation to the neurodevelopmental measures found cerebral oxygen saturation thresholds of 49% to 62%. The hours at a cerebral oxygen saturation less than 45% and 55% were related to low visual-motor integration and neurodevelopmental index scores in the univariate and multivariate models. A multivariate model of age and weight at stage 1 palliation, cerebral oxygen saturation, arterial oxygen saturation, cardiopulmonary bypass and deep hypothermic circulatory arrest times, and later stroke predicted visual-motor integration to an important degree (R(2) = 0.53, P < .001). The actual and predicted visual-motor integration and neurodevelopmental index were normal when a cerebral oxygen saturation less than 45% and other risk conditions were avoided. CONCLUSIONS: Neurodevelopmental performance was related to demographic, neonatal perioperative physiologic, and later factors. Perioperative cerebral oxygenation assessed by near-infrared spectroscopy can detect hypoxic-ischemic conditions associated with injury and reduced neurodevelopmental performance and was the most significant physiologic factor identified. These data suggest that efforts to avoid cerebral hypoxia are likely to improve the outcomes in this high-risk population.
OBJECTIVES: Neonates with hypoplastic left heart syndrome have significant hemodynamic threats to cerebral perfusion and are at risk of reduced neurodevelopmental performance. We hypothesized that cerebral hypoxia, detectable by near-infrared spectroscopy in the early postoperative period, would be related to later neurodevelopmental performance. METHODS: The study population was a sequential cohort of patients who had undergone stage 1 palliation of hypoplastic left heart syndrome under standard conditions, including neonatal perioperative monitoring with cerebral near-infrared spectroscopy, and who had undergone a neurodevelopmental assessment at age 4 to 5 years. The neonatal demographic and 48-hour perioperative hemodynamic parameters, including cerebral oxygen saturation, were tested for their relationship to 4 domains of neurodevelopmental performance, including visual-motor integration in childhood in univariate and multivariate models. The neurodevelopmental scores were classified as low if less than 85 (-1 standard deviation) and abnormal if less than 70 (-2 standard deviations). RESULTS: For the 51 patients in the surgical cohort, the early survival was 94%, the cumulative survival was 86%, and the neurodevelopmental assessment was completed by 21 (48%) of the survivors, without evidence of an ascertainment bias. At the test age of 56.3 ± 5.5 months, the composite neurodevelopmental index, constructed from equally weighted measures in 4 domains, was 97.6 ± 9.6, not different from the age-based norms, with 3 of 21 in the low range and none abnormal. The mean visual-motor integration was 93.4 ± 14, slightly less than the population norm (P < .05), with 2 of 21 having low scores and 1 abnormal scores. In patients with low to abnormal visual-motor integration, the perioperative stage 1 palliation cerebral oxygenation saturation was significantly lower (63.6 ± 8.1 vs 67.8 ± 8.1, P < .05). Two patients had discrete embolic strokes after their initial hospitalization; the occurrence of late stroke reduced the visual-motor integration performance but was not related to the early cerebral oxygen saturation. Nonlinear relationships of cerebral oxygen saturation to the neurodevelopmental measures found cerebral oxygen saturation thresholds of 49% to 62%. The hours at a cerebral oxygen saturation less than 45% and 55% were related to low visual-motor integration and neurodevelopmental index scores in the univariate and multivariate models. A multivariate model of age and weight at stage 1 palliation, cerebral oxygen saturation, arterial oxygen saturation, cardiopulmonary bypass and deep hypothermic circulatory arrest times, and later stroke predicted visual-motor integration to an important degree (R(2) = 0.53, P < .001). The actual and predicted visual-motor integration and neurodevelopmental index were normal when a cerebral oxygen saturation less than 45% and other risk conditions were avoided. CONCLUSIONS: Neurodevelopmental performance was related to demographic, neonatal perioperative physiologic, and later factors. Perioperative cerebral oxygenation assessed by near-infrared spectroscopy can detect hypoxic-ischemic conditions associated with injury and reduced neurodevelopmental performance and was the most significant physiologic factor identified. These data suggest that efforts to avoid cerebral hypoxia are likely to improve the outcomes in this high-risk population.