BACKGROUND:

The pediatric medical community has long been concerned about the increased incidence of cerebral dysfunction among newborns with severe cyanotic congenital heart disease. There has been concern about a great number of potential predisposing factors including the effects of hypoxemia and acidosis, trauma during cardiac catheterization, and the duration of prolonged decreased arterial oxygen saturation. Many infants undoubtedly undergo significant brain damage prior to surgery, but may be subject to further insult in the peri-operative period. The proposed trial studied circulatory arrest versus low-flow bypass methods of surgical support in a homogeneous group of patients who were managed in identical fashion in all other aspects of their treatment. Major factors, such as the influence of cardiac catheterization and presurgical status were specifically incorporated into the primary statistical analysis. This was important in order to identify significant predisposing causes of brain injury beyond the surgical techniques being compared, and to increase the power of the comparison.

DESIGN NARRATIVE:

Randomized, partially blinded. Infants were assigned to deep hypothermia and either circulatory arrest or low-flow bypass support during arterial switch surgical repair of transposition of the great arteries. Randomization was stratified by diagnosis (with or without septal defect) and by specific surgeon of three, thus yielding six strata within each treatment group. Randomization was also blocked to minimize imbalances in treatment assignment between strata. The study was partially blinded, in that outcome data were obtained by investigators who were unaware of management, surgical care, and intensive care were standardized. Hemodynamic assessment was made early in the post-operative period. The outcomes of primary interest (magnetic resonance imaging, neurologic examination, and psychometric testing) were reviewed at the one-year evaluation, to maximize relevance to the long-term prognosis. Pre-operative and seven-to-ten day post-operative neurologic examination was also carried out. Developmental status at one year was measured by the Bayley developmental scales and the Fagan test of infant intelligence. Brain function was assessed by neurologic examination at one year. The recruitment period extended from April 1988 to February 1992.

The trial was extended through December 1997 to allow four years of follow-up. The follow-up allowed the assessment of a primary outcome variable in the neurologic domain and a primary outcome variable in the developmental domain. The primary outcome variable of the neurologic examination resulted in an ordered classification of three categories of normal, minor, and major dysfunction, such as apraxia or cerebral palsy. Development was assessed using standardized measures of cognition, behavior, language, attention, and motor skills. The primary outome variable for the developmental evaluation was the intelligence quotient on the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-R). Secondary outcome variables included measurements of behavior and performance in specific domains of speech and language, visual motor integration, attention, motor function, and temperament.

Beginning in January 1998, the study was extended for an additional four years through December 2001 to compare the effects of the two intraoperative methods with respect to developmental and neurologic status at eight years of age. Developmental status was assessed by the Wechsler Individual Achievement Test (primary outcome) and by the presence of learning disabilities and performance in specific neuropsychologic domains (secondary outcomes). Neurologic status was determined by neurologic exam (primary outcome) and by specific types of dysfunction such as motor problems. The study also compared the academic and intellectual performance of the cohort at age eight years to population norms and used the data to identify correlates of the children's eight year outcomes, as well as their development between ages one and eight.