P1.212: Low cord blood progesterone predicts worse neurocognitive outcomes in preterm infants.

Background: Progesterone (P4) modulates placental function, maternal immune response, and fetal development, rising 5-10 fold in pregnant mothers in the 3rd trimester. P4 supplementation for short cervical length reduces risk of preterm birth. In animal models, P4 receptor-mediated pathways enhance recovery after brain injury. Our prior study investigated umbilical cord blood P4 levels in term infants. No prior studies have investigated the association between cord P4 and postnatal outcomes in preterm infants.

Our objective was to determine perinatal P4 level in preterm infants and investigate its relation to neurodevelopmental outcomes.

Methods: This is an IRB-approved prospective cohort study. We enrolled 216 inborn infants with GA < 31 weeks (mean 26.7). We excluded infants with congenital anomalies or prenatal plan to withhold neonatal care. We collected samples from cord blood as well as newborn blood at NICU admission and 24 hours of life. We analyzed samples using a CLIA certified P4 assay. We correlated samples with individual infant scores on follow up neurodevelopmental assessments, including Bayley Scales of Infant & Toddler Development III at 12, 24, and 36 months. Mean term cord P4 (631 ng/mL) was determined from our prior study. One standard deviation below this mean (285 ng/mL) was used as the cutoff for this cohort, and infants were stratified into low cord P4 (n=75) and normal cord P4 groups (n=69) for analysis. Data analysis utilized Pearson’s chi-square or Fisher’s exact tests and independent sample t-tests or correlation analysis for categorical and continuous variables, respectively. We used an alpha of 0.05.

Results: Cord blood samples were collected on 66.7% (144) of total subjects enrolled. Of those, 6.9% (10) were under the P4 assay limit of detection. Cord levels significantly increased with GA (p=0.0034), but serum levels decreased 10-fold by 24 hours of life (estimated blood half-life of < 6h in neonates). Cord P4 levels were >1 SD lower than the term infant mean in 52% (75) of subjects ( < 285 ng/mL). Lower GA was associated with significantly worsened measures of neurodevelopment including head circumference at 28 days of life (p < 0.0001), lower 3 month TIMP z-score (p=0.0269), higher 6 month EMPP score (p < 0.0001), and lower Bayley III scores at 12 and 24 months in domains of cognition, language, and motor development. In preterm infants with low cord P4 levels, 24 month Bayley III social-emotional, cognitive, and general adaptive composite scores were significantly lower than in preterm infants with higher levels (>285 ng/mL).

Conclusion: Birth GA was strongly associated with both umbilical cord blood P4 levels and neurodevelopmental outcomes. Additionally, low cord P4 levels were predictive of worse neurocognitive outcomes in preterm infants in this study. Further investigation of the impact of P4 on early life neurodevelopment, and potential replacement therapeutic strategies, are warranted.