Evaluation of Potential Biomarker for Subtle Cerebral Injury in Neonates and Infants Undergoing Surgery for Congenital Heart Disease: Association with Perioperative Cerebral Tissue Oxygenation

Objectives: Mild to moderate neurodevelopmental impairment is common in children with congenital heart disease (CHD) requiring cardiac surgery as neonates or infants. Perioperative hypoxic-ischemic brain injury might contribute to adverse neurodevelopment. Evaluation of monitoring techniques and markers for perioperative neurologic injury are of special interest. We analyzed and compared the value of two potential biomarkers for neuronal cell damage in neonates and infants (< 1 year) undergoing cardiac surgery. The relation between biomarker levels and perioperative cerebral tissue oxygenation was evaluated.

Methods: Serum levels of S100 and glial fibrillary acidic protein (GFAP) were measured before and 0, 12, 24 and 48 hours after surgery (S100: electrochemoluminometric immunoassay, ELECSYS, Roche; GFAP: Human GFAP ELISA, Abbexa). Cerebral tissue oxygenation (ScO₂) was derived by near infrared spectroscopy (INVOS, Covidien) for 12 hours before, during and for 48 hours after surgery. Cerebral oxygen extraction was estimated by calculating the difference between arterial saturation and cerebral tissue oxygen saturation (ΔSaO₂ScO₂).

Results: Biomarker levels and ScO₂ were obtained in 20 neonates and 35 infants. Pre- and postoperative S100 levels were significantly higher in neonates compared to infants. Highest S100 levels were measured directly after surgery in both groups (1.30 ±1.42 µg/L vs. 0.54 ±0.21 µg/L, p = 0.028). After 48 hours, S100 was elevated in 30 and 23% of cases, respectively (p = 0.749). In infants with elevated S100, mean postoperative ScO₂ (64 ± 7% vs. 70 ± 7%, p = 0.016) was lower and ΔSaO₂ScO₂ (31 ± 7% vs. 24 ± 6%, p = 0.005) higher than in infants with normalized values. Postoperative GFAP levels at 12, 24, and 48 hours were higher in neonates. Elevated GFAP levels after 48 hours were found in 57% of the neonates and 21% of the infants (p = 0.035). In neonates with elevated GFAP, mean intraoperative ScO₂ was lower compared to cases with normal GFAP values (70 ± 9% vs. 85 ±4%, p = 0.002). Early postoperative ΔSaO₂ScO₂ tended to be higher in cases with elevated GFAP (40 ± 9% vs. 28 ± 13%, p = 0.073).

Conclusion: A relationship between cerebral oxygenation and serum levels were found for both biomarkers. GFAP might be a useful biomarker for subclinical neuronal cell damage in neonates undergoing surgery for CHD, while S100 might be suitable for infants. Neurodevelopmental outcomes and their association to S100 and GFAP levels still need to be determined.