Background and aims: To compare findings of cranial ultrasound (US) and magnetic resonance imaging (MRI) of the neonatal brain. Methods: A retrospective population based cohort study of premature infants with a very low birth weight (VLBW) of less then 1500 grams who were admitted to our neonatal unit on day 1 (period: 2007 to 2009). All infants underwent cranial US on day 1, 3, 7, 14, 28, 42 monthly thereafter and at term corrected age. Infants who were eligible for this study obtained a MRI at term (37 to 42 weeks gestational age) corrected age. Retrospectively, we compared the cranial US findings as a predictor of a wide spectrum of pathology on MRI. Results: Paired MRI and US studies were performed in (n=140) VLBW infants who were born at a median gestational age of 28 (range: 22+1 to 34+5) weeks and a median birth weight of 1020 (range: 335 to 1495) grams. US predicted some MRI findings accurately: germinal layer haemorrhage (GLH), cystic lesions, intraventricular haemorrhage (IVH) and severe white matter (WM) echogenicity on US for the presence of WM haemorrhagic parenchymal infarction on MRI. Other MRI changes were less well-predicted: delay in maturation and myelination, reduced cortical folding, congenital malformations, punctate lesions and mild or no WM echogenicity on US for the presence of normal (n=112) WM signal intensity on MRI. Conclusions: MRI of the neonatal brain might shed light on the origin of brain lesions causing long-term neurodevelopmental sequelae and might change our perinatal and neonatal management.

Placental syncytiotrophoblasts are known to express the efflux transporter proteins P-glycoprotein (ABCB1) and multidrug resistance-associated protein 2 (ABCC2), which are supposed to be a functional part of the human placental barrier. With advancing gestational age, expression of ABCB1 decreases progressively, whereas ABCC2 is more expressed. To evaluate to which extent they contribute to placental barrier function at term, permeability of talinolol, a substrate of both carriers, was measured using a validated human placenta perfusion model. We identified in randomized, crossover experiments a unidirectional transfer of talinolol in the fetomaternal direction because the maternofetal transfer was significantly lower (0.663 ± 0.188 versus 0.394 ± 0.067 relative to creatinine permeability, p = 0.012). Maternofetal permeability was increased by the ABCC2 inhibitor probenecid (0.59 ± 0.15 versus 0.68 ± 0.13, p = 0.028) and the nonspecific inhibitor verapamil (0.53 ± 0.09 versus 0.66 ± 0.16, p = 0.028) but was not influenced by the ABCB1 inhibitor valspodar (PSC833) (0.48 ± 0.11 versus 0.46 ± 0.09, p = 0.345). Genetic polymorphisms of ABCB1 and ABCC2 lacked significant influence on expression of the carriers and permeability of talinolol, respectively. In conclusion, maternofetal transfer of talinolol is restricted by a unidirectional process that is influenced by inhibitors of ABCC2.