543 - Retinopathy of Prematurity and Development of Brain Abnormalities on Structural MRI at Term in Infants Born Very Preterm

Poster Number: 543
Publication Number: 543.1346

Background: Retinopathy of prematurity (ROP) is an abnormal vascular proliferation in the retina and the leading cause of blindness in very preterm infants. Some studies show an association between ROP and neurodevelopmental impairment, however the mechanism by which ROP affects neurodevelopment is unclear. The association between ROP and neurodevelopmental impairment in one large cohort study (Schmidt B, JAMA 2003) was independent of brain injury on ultrasound. Brain MRI is more sensitive than cranial ultrasound and may identify associations between ROP and brain abnormalities.

Objective: Determine if ROP is associated with brain injury and/or dysmaturity on structural MRI at term equivalent age in infants born very preterm.

Design/Methods: We prospectively recruited a cohort of 395 very preterm infants (≤ 32 weeks gestational age [GA]) from five regional Cincinnati neonatal intensive care units. Eligible infant underwent routine ROP screening and diagnosis per international screening guidelines at 31 weeks corrected GA or 5 to 6 weeks of age (whichever came later). Patients were classified as having any ROP or severe ROP (stage 3 ROP or requiring treatment). We obtained structural MRI at 39 to 44 weeks postmenstrual age for all infants. Maturation as well as structural abnormalities were identified by a single neuroradiologist and abnormality burden was quantified using the Kidokoro global brain abnormality score. We performed multiple linear regression to determine the association between ROP and brain abnormality score while controlling for confounders listed in Table 1.

Results: Baseline characteristics for 395 very preterm infants are listed in Table 1. 261 (66%) infants had no ROP, 112 (28%) infants developed mild ROP, and 22 (5.6%) infants developed severe ROP. Any ROP was associated with a higher mean global brain abnormality score than the no ROP cohort (8.75 vs 4.72, p < 0.001). Severe ROP was associated with an even higher mean global brain abnormality score (11.86, p < 0.001). In multiple linear regression analyses, any ROP (β= 1.52 (95% CI: 0.27, 2.78) and severe ROP (β=2.53 (0.45, 4.61) remained significantly associated with global brain abnormality score, independent of known cofounders (Tables 2 and 3).

Conclusion(s): In our large cohort of very preterm infants, any ROP and severe ROP were significantly associated with brain injury and/or maturational abnormalities on structural MRI at term. This suggests that ROP may increase the risk of neurodevelopmental impairment through its adverse effects on brain injury/delayed maturation.