The objective of this secondary analysis was to examine the impact of 2 perfusion modalities on S100 calcium-binding protein B (S100B) levels and neural injury outcomes in patients undergoing congenital heart surgery with cardiopulmonary bypass (CPB).

Pediatric patients (aged <18 years) were randomly assigned to nonpulsatile or pulsatile perfusion. S100B levels were measured in patients at 5 time points (baseline before incision, 3 to 5 minutes on pump, right before weaning from CPB, 1 hour after CPB, and 24 hours after CPB). All patients were assessed for short-term clinical outcomes and with chart review for postsurgical neural injury: stroke, seizure, white matter injury, developmental delays, behavioral issues, memory loss, unexplained loss of consciousness, hearing impairment, and attention-deficit/hyperactivity disorder.

A total of 144 consecutive patients were enrolled (70 nonpulsatile, 74 pulsatile). Pulsatile perfusion produced CPB characteristics similar to those of nonpulsatile perfusion and created a greater pulsatility index in the middle cerebral artery and arterial line. S100B levels at all measured time points were similar between both perfusion modalities. A total of 36 patients had neural injuries (15 nonpulsatile, 21 pulsatile). Additionally, S100B levels at all measured time points were similar in patients with and without neural injury. In all groups, S100B levels rose during CPB and were greatest before weaning from CPB and 1 hour after CPB. The 24-hour S100B levels had a weak association with clinical outcomes (0.25 < r_S <0.5).

The choice of perfusion modality did not appear to affect S100B levels. Additionally, the measurements of S100B may not be sufficiently sensitive to predict neural injury reliably in patients with congenital heart defects who are undergoing CPB.