Highlights
- Large multicenter TORPIDO 30/60 trial found no difference in death or brain injury between initial FiO2 of 0.6 and 0.3 in extremely preterm infants.
- Individual participant data meta-analysis suggests high initial FiO2 (≥0.90) may reduce mortality, though certainty is low.
- Deferred cord clamping with 100% oxygen reduced early hypoxemia without increasing morbidity.
Background
Hypoxia and hyperoxia can both harm preterm infants due to their underdeveloped lungs and immature antioxidant systems. Delivery room management of oxygen—specifically the fraction of inspired oxygen (FiO2) used for initial resuscitation—remains controversial. While guidelines have historically recommended starting low and titrating up, recent evidence challenges whether this approach optimizes survival and neurological outcomes for infants born before 32 weeks’ gestation.
Study Design
Three major studies form the evidence base:
TORPIDO 30/60 Randomized Clinical Trial
Conducted in 31 hospitals across 6 countries, infants at 23–28 weeks’ gestation were randomized to initial FiO2 of 0.6 or 0.3. FiO2 was titrated to standard saturation targets in the first 10 minutes. Primary endpoint: death or brain injury at 36 weeks’ corrected gestational age.
IPD Network Meta-Analysis
Systematic review and Bayesian network meta-analysis of randomized trials enrolling infants <32 weeks’ gestation comparing low (≤0.3), intermediate (0.5–0.65), or high (≥0.90) FiO2. Primary outcome: all-cause mortality at hospital discharge.
Deferred Cord Clamping (DCC) with High Oxygen Trial
Double-blind RCT of 140 infants (22–28 weeks) receiving CPAP or positive pressure ventilation during DCC, randomized to 30% or 100% FiO2 via face mask before cord clamping. Primary outcome: achieving SpO2 ≥80% at 5 minutes.
Key Findings
TORPIDO 30/60 Trial
1423 infants analyzed. Death or brain injury rates were similar: 46.9% in the 0.6 FiO2 group vs 47.8% in the 0.3 FiO2 group (RR: 0.98; 95% CI: 0.89–1.09). Escalation to FiO2 of 1.0 was comparable (~40% in both arms). No harm or benefit was observed, suggesting moderate FiO2 levels may not change major outcomes.
IPD Network Meta-Analysis
Data from 1055 infants in 12 RCTs. High FiO2 (≥0.90) was associated with reduced mortality compared to low FiO2 (OR 0.45, 95% CrI 0.23–0.86, low certainty) and intermediate FiO2 (OR 0.34, 95% CrI 0.11–0.99, very low certainty). Probability of ranking first for mortality reduction: 97%. Effects on severe morbidities were inconclusive.
DCC High Oxygen Trial
SpO2 ≥80% at 5 min achieved in 69% of high oxygen group vs 39% of low oxygen group (adjusted OR 3.74; 95% CI 1.80–7.79; P<0.001). Absolute risk difference: 30% lower hypoxemia risk in high oxygen group. No difference observed in severe intraventricular hemorrhage or death before 40 weeks postmenstrual age.
Expert Commentary
These studies highlight critical nuances in neonatal resuscitation: TORPIDO suggests no outcome difference between moderate FiO2 levels, while the IPD-MA points to potential survival benefits from high FiO2—albeit with low certainty. The DCC trial introduces a physiologically plausible approach: during intact placental circulation, 100% oxygen may improve early oxygenation without increasing oxidative stress risk due to oxygen mixing with umbilical venous blood. Clinicians should interpret these findings with caution, considering patient-specific factors including gestational age, condition at birth, and available monitoring tools.
Conclusion
Optimal initial oxygen concentration for preterm resuscitation remains unresolved. Moderate FiO2 (0.3–0.6) may not markedly impact survival or brain injury, but high FiO2 (≥0.90) might offer survival advantages in certain contexts, particularly during deferred cord clamping. Future large, high-quality trials are needed to balance benefits of early oxygenation against potential oxidative injury risks, and to refine guideline recommendations.
Funding & Clinical Trials Registration
TORPIDO 30/60: ANZCTR ACTRN12618000879268. DCC High Oxygen: ClinicalTrials.gov Identifier NCT04413097.
References
Oei JL et al. JAMA. 2025 Dec 10:e2523327.
Sotiropoulos JX et al. JAMA Pediatr. 2024;178(8):774-783.
Katheria AC et al. JAMA Pediatr. 2025;179(9):971-978.