Facebook Twitter YouTube


Perinatal Clinical Predictors of Cerebral Microstructural Development On Diffusion Tensor Imaging (DTI) In Extremely Preterm Infants (EPI)

Friday, October 19, 2012
Room R02-R05 (Morial Convention Center)
Ulana Pogribna, MD/MPH1, Robert E. Lasky1, Katrina Burson1 and Nehal A. Parikh, DO, MS2, (1)Pediatrics, Division of Neonatology, University of Texas Health Science Center Houston, Houston, TX, (2)Pediatrics, Nationwide Children's Hospital, Columbus, OH


A majority of EPI survivors (≤28 weeks GA) exhibit neurodevelopmental impairments (NDI). DTI can sensitively detect microstructural brain injury/aberrant development in very preterm infants, but clinical correlates have not been well characterized. The purpose of this study was to comprehensively investigate and validate perinatal clinical predictors of DTI abnormalities at term-equivalent age in two prospective cohorts of EPI.


Following consent, DTI (3 Tesla) was performed on 50 extremely low birth weight (ELBW) infants and 16 healthy term newborns at 38 weeks post-menstrual age (PMA). Tract-based spatial statistics were performed to identify vulnerable white matter regions in ELBWs. Region of interest based fractional anisotropy (FA) and mean diffusivity (MD) were measured for 9 such regions (ICC>0.93). Over 50 clinical factors were correlated with FA and MD. For each region of interest, all variables with P<0.25 in bivariate analyses (controlling for PMA at MRI) were entered in multiple linear regression analyses with elimination of factors with the highest P values one at a time manually. Bootstrapping confirmed internal validity. We then sought external validation by testing all significant predictors and models in a second independent cohort of 36 EPI, following the same data collection and imaging methods.


ELBW infants had a mean (SD) GA of 25.2 (1.7) wks and BW of 751 (143) g. The second cohort of EPI exhibited significantly higher mean (SD) GA of 26.5 (1.6) wks, BW of 873 (208) g, and higher mean FA and lower MD in multiple regions as compared to ELBWs, suggesting more maturation. They also exhibited lower morbidity rates and less brain injury than ELBWs. Chorioamnionitis and indomethacin use for patent ductus arteriosus were significant (P<0.05) risk factors for aberrant development/injury in both cohorts. White matter injury on cranial ultrasound was a significant risk factor in ELBWs, with similar trends in the EPI cohort (P<0.1). Longer duration of human milk use was a protective factor in both cohorts (P<0.05).


Several perinatal factors known to be associated with NDI also correlate with microstructural cerebral development/injury. Robust correlation with NDI (ongoing) will be essential to validate DTI measures as surrogate end points for clinical trials in infants.