331 - Intermittent Hypoxemia and Increased Brain Injury Biomarker (S100B) in Preterm Infants

Poster Number: 331
Publication Number: 331.3057

Background: Intermittent Hypoxemia (IH) is a common and significant problem in preterm infants. We have previously shown IH frequency increases with postnatal age, peaking at 4 wks of life, and plateaus thereafter. Evidence from both animal models and clinical studies links IH to brain injury and neonatal impairments. However, early biomarkers for IH-related brain injury are currently lacking, making it challenging to effectively monitor disease, optimize care, and predict outcomes.

Objective: Assess for the first time in preterm infants, the relationship between IH and S100B, a brain injury biomarker shown to increase in adults with brain injury.

Design/Methods: We enrolled preterm infants ≤32 gestation age (GA). To quantify an IH profile, oxygen saturation (SpO2) was collected using high-resolution (1Hz sampling, 2s averaging) pulse oximeters. IH profile included measures (e.g. frequency, duration, time in hypoxemia) at multiple thresholds (SpO2 < 85, < 80) and categorized events of short ( < 1-min) and long (≥1-min) durations. S100B was measured in urine samples obtained at multiple time points during NICU stay. We excluded infants with severe intraventricular hemorrhage (contributor to brain injury), and sepsis/NEC. To better demonstrate the relationship between IH and S100B, we stratified IH burden into quartiles (Low, Mild, Moderate, High IH). Secondary analyses categorized infants based on GA: extremely (22-27wks) and very (28-32wks) preterm. We used two-tailed parametric/non-parametric statistical tests.

Results: A total of 29 infants were prospectively enrolled. After exclusions as above, we analyzed 52 samples from 21 infants with a median GA of 28.3wks (IQR 26.2–30.1) and birth weight of 1060g (IQR 810–1330g). There was a statistically significant association between S100B levels and IH, i.e. S100B increased with a higher IH frequency and events of longer duration (N=52, all p< 0.05). We then assessed S100B and IH at peak frequency (1 sample per patient closest to 4 wks); similarly, there was a strong correlation (Table 1). S100B levels increased in infants with higher IH quartile (Figure, p=0.01). Similarly, analyses by age group showed increased S100B with higher IH, especially of longer duration (%IH-SpO2 < 80, all p< 0.05).

Conclusion(s): This study shows increased release of brain injury biomarker S100B with increased IH burden in preterm infants. This novel finding is the first step towards developing an early IH-related brain injury biomarker that will help define the thresholds of IH injury, optimize IH care, and predict outcomes. A larger study is in progress to further assess these relationships.