Ultrasound detection of white matter injury in very preterm neonates: practical implications

Abstract

Diffuse white matter injury is not well detected by cranial ultrasonography (CUS). The aim of this study was twofold: (1) to assess in very preterm neonates the predictive values of individual CUS abnormalities for white matter injury on MRI and neurological outcome; (2) to develop a strategy optimizing CUS detection of white matter injury. Very preterm neonates (n=67; 44 males, 23 females) underwent serial CUS and single MRI. Predictive values of CUS findings for a white matter classification on MRI, individual MRI findings, and neurological outcome at 2 years corrected age were calculated. The effects of timing and frequency of CUS were evaluated. Periventricular echodensities (PVEs) predicted abnormal white matter on MRI, but absence of PVEs did not predict absence of white matter changes. Peri- and intraventricular haemorrhage (P/IVH) was highly predictive of abnormal white matter on MRI. Frequency and timing of CUS did not influence predictive values. P/IVH and abnormal ventricular size/shape were reasonably predictive of unfavourable outcome, whereas absence of CUS abnormalities predicted a favorable outcome. (1) If PVEs are present, there is a significant chance of abnormal white matter on MRI. (2) Increasing frequency of CUS does not increase its diagnostic performance for white matter injury. (3) P/IVH is highly predictive of abnormal white matter on MRI and reasonably predictive of unfavourable outcome. (4) Absence of PVEs and P/IVH on CUS does not guarantee normal white matter, but predicts a favourable outcome.