Purpose or Case Report: Timely intracranial imaging enables diagnosis, directs treatment, and provides prognostic information of infant malformation or injury. MRI is the gold standard for neuroimaging, but can be inhibited by transportation risk and incompatible support devices. Ultrasound (US) is portable, rapid, and inexpensive, but limited by lower sensitivity and specificity. Contrast-enhanced ultrasound (CEUS) improves diagnostic accuracy of US in other clinical applications. Our aim was to prospectively compare CEUS of the neonatal brain to MRI.
Methods & Materials: Infants in our NICU with vascular access undergoing brain MRI were recruited. A GE Logiq E9 or E10 US system (GE Healthcare Milwaukee, WI) was used, performed by a radiologist blinded to MRI results. The contrast agent (Lumason, Bracco Diagnostics, Monroe Township, NJ) was used off-label, with dose of 0.03mL/kg IV, per departmental CEUS protocol. Infants were monitored for adverse events. CEUS followed routine US. Coronal plane cinematic clips were saved during wash-in of contrast, followed by coronal and sagittal sweeps of the entire brain parenchyma, with trans-mastoid sweeps last.
Results: Twenty-three neonates (mean age 17 days, mean gestational age 37+4 weeks) were imaged on average 17.5 hours following MRI. MRI indications included: 6 infection, 6 seizure, 4 suspected hypoxic ischemic injury, 4 follow-up malformations, 2 abnormal prior screening US, 1 prematurity. CEUS examinations were diagnostic quality and improved diagnosis compared to US. MRI and CEUS were concordant in 13 patients (57%) and discordant in 10 (43%). Discordances, with diagnoses seen only on MRI, included tiny acute infarct, small ischemic hemorrhagic transformation (ischemia was concordant), hemosiderin staining from old intraventricular hemorrhage, small posterior fossa subdural hemorrhages, and brainstem calcifications, several of which were clinically insignificant. Two MRI abnormalities not seen on CEUS had normal follow up MRI, including focal T2 signal hyperintensity and focal hyperperfusion on arterial spin labeling. One case of MCA infarct luxury reperfusion was seen at CEUS but not MRI. There were no false positive CEUS cases and no adverse events with CEUS were encountered.
Conclusions: Neonatal brain CEUS is feasible, safe, and increases diagnostic ability of US. Additionally, CEUS directly visualizes dynamic vascular perfusion that may prove to be an added benefit of CEUS over MRI. Although further study is needed, CEUS may be a valuable additional diagnostic tool.