Exposure to supra-physiologic levels of oxygen has been associated with end-organ damage and worse outcomes across multiple populations related to oxidative stress and inflammation. There is limited evidence to guide practice in children undergoing surgery with cardiopulmonary bypass, including neonates with cyanotic congenital heart disease, who are particularly vulnerable to oxidative stress due to immature and depleted antioxidant defenses.

We conducted the first randomized trial of perioperative normoxia (goal PaO₂ 60-100 mmHg during cardiopulmonary bypass and for the first 24 hours) versus hyperoxia (goal PaO₂ 200-300 mmHg during cardiopulmonary bypass) exclusively in cyanotic neonates to examine feasibility, systemic oxidative stress response, and safety. The primary outcome was a serum measure of oxidative stress: thiobarbituric acid reactive substances (TBARS). Additional biomarkers and clinical outcomes were collected.

Twenty-nine neonates (median age 5 days) were enrolled. Mean PaO₂ in the normoxia arm (n=15) during cardiopulmonary bypass was 106±27 mmHg (vs hyperoxia 256±29 mmHg, p<0.0001). Mean post-operative TBARS expressed as fold-change from baseline was lower in the normoxia arm at 2 hours (1.21±0.26 vs 1.93±0.63), 6 hours (1.09±0.25 vs 1.77±0.7), and 24 hours (0.96±0.16 vs 1.41±0.29); all p<0.01. Other biomarkers of oxidative stress were significantly lower in the normoxia arm (all time points p<0.01). Safety outcomes including in-hospital mortality and morbidities were similar between arms.

Controlled normoxia during cardiopulmonary bypass in cyanotic neonates is feasible, safe, and associated with significant improvement in measures of oxidative stress. These data are critical in informing larger-scale studies of end-organ injury and clinical outcomes.