Magnetic Resonance Imaging and Computerized Tomography in Central Hypoventilation

Publisher Website
Google Scholar
Abstract
Central hypoventilation syndrome (CHS) is a disorder of respiratory control. It may be an idiopathic primary disease or it may be the secondary consequence of an infectious process or Chiari II malformation. Clinical data suggest that the primary defect involves the brainstem respiratory centers. To date, pathologic evaluation has linked primary CHS with decreased density of neurons and myelinated nerve fibers in the medulla and brainstem gliosis, and absence of the external arcuate nucleus. Magnetic resonance imaging (MRI) is not considered the most reliable imaging technique for evaluating the brainstem. In the hope of finding gross structural abnormalities that might correlate with the clinical and pathologic features of CHS, we used MRI and computerized tomography (CT) to evaluate the brain, brainstem, and spinal cord of 17 infants and children with central hypoventilation (11 primary, 6 secondary). Each of the 11 children with primary CHS demonstrated a normal brainstem and spinal cord by MRI. However, 9 of 11 infants had mild ventricular dilatation and modest prominence of the sulci interpreted as either mild atrophy or mild extraventricular obstructive hydrocephalus. Results of studies in the one child with postinfectious disease were normal. The 5 infants with Chiari II malformation demonstrated characteristic brainstem changes. MRI results modified clinical care, resulting in symptomatic improvement after hindbrain decompression by cervical laminectomy in 1 infant and surgical drainage of syringobulbia in 1 child. In summary, in primary CHS, the MRI and CT provided evidence for a more diffuse CNS process rather than a specific brainstem lesion. Such a lesion may be too small or too subtle to resolve with MRI or CT. In secondary central hypoventilation, MRI modified surgical management in 2 cases, resulting in symptomatic improvement of hypoventilation. In the remaining Chiari II patients, anatomic findings did not necessarily correlate with clinical symptoms.