Pediatrics
Pediatric Neurometabolic Disease: Hypomyelination vs Destruction
Fellowship-level neuroradiology teaching by E. Brooke Schrickel, MD. Open this topic in the interactive reading room →
Core clinical idea
Don't chase the enzyme — answer two sequential questions the genetics workup turns on: (1) is this too little myelin ever made (hypomyelination) or myelin made then destroyed (demyelinating leukodystrophy), decided on T1 and tissue integrity, not T2; and (2) if destructive, map symmetry, topography, gradient, deep gray, cord, and head size to narrow the entity. Set the age-corrected myelination baseline first.
Bottom line
Myelination clock first, then hypo- vs destructive on T1 — then map symmetry, gradient, and head size to the entity.
Core workstation questions
- Is myelination appropriate for age, corrected for prematurity in the first year?
- Is T1 intermediate with preserved volume and no cavitation (hypomyelination), or destructive with cavitation and volume loss (leukodystrophy)?
- Is the pattern symmetric and confluent (genetic) or asymmetric with gray-matter involvement (acquired ADEM/MOG)?
- Which way is the gradient marching — posterior splenium-to-biparietal (ALD) or frontal with external/extreme capsule (Alexander)?
- Is the child macrocephalic, and have I checked deep gray, cerebellum, cord, and cranial-nerve/root enhancement?
- Would spectroscopy confirm (e.g., towering NAA in Canavan) or detect early disease before conventional sequences?
What changes reporting / management
- State whether you think this is hypomyelination vs a destructive leukodystrophy — it points the workup in opposite directions (hypomyelinating gene panels vs metabolic/peroxisomal/lysosomal).
- Name the gradient and the enhancing leading edge (posterior centrifugal = X-linked ALD; frontal periventricular garland = Alexander) rather than just calling 'leukoencephalopathy.'
- Use head circumference as a first-pass filter: macrocephaly + leukoencephalopathy is a short differential (Alexander, Canavan, MLC, vanishing white matter).
- Protocol contrast and add spine/IAC coverage when the process looks destructive; cranial-nerve and nerve-root enhancement can clinch entities like metachromatic leukodystrophy.
- Finish the whole scan in a seizure patient — look for mesial temporal sclerosis perpetuating refractory seizures.
Practical traps
- T2 is bright in essentially all of these, so it does not separate hypomyelination from demyelination — read T1 and whether tissue is preserved or cavitated.
- Overcalling delayed myelination on peripheral/subcortical white matter that is simply not yet myelinated for age (and conversely, calling a diffusely unmyelinated 2-year-old 'normal').
- Satisfaction of search — stopping once a leukodystrophy is identified and missing a second actionable finding such as mesial temporal sclerosis.
Teaching pearls
- T2 is bright in all of them — the answer is on T1 and in whether tissue is preserved or cavitated.
- Correct for prematurity before calling delayed myelination.
- Symmetric and confluent = genetic leukodystrophy; asymmetric with gray-matter hits = acquired (ADEM/MOG).
- ALD marches posterior; Alexander marches frontal.
- Macrocephaly + leukoencephalopathy is a short list — use head size as a filter.
- Towering NAA = Canavan.
- Finish the scan — don't let the leukodystrophy hide a mesial temporal sclerosis.
Source lectures
- Pattern Approach to Pediatric Neurometabolic Disorders
Educational material for radiology residents and neuroradiology fellows. Nothing here drives individual patient care, and it contains no patient data.