Timing and Interpretation of MRI After Acute Profound Perinatal Asphyxia
A term male neonate (39 weeks 2 days, 3.4 kg) was delivered by emergency cesarean after a prolonged fetal bradycardia and placental abruption. Apgar scores were 1, 4, and 6; cord arterial pH was 6.82 with a base deficit of 19 mmol/L and an initial lactate of 14 mmol/L. He was intubated at 8 minutes and started on therapeutic hypothermia at 3.5 hours of life. At 30 hours he remains ventilated, has burst-suppression on amplitude-integrated EEG, and has received phenobarbital for two electrographic seizures. He is hemodynamically stable on low-dose dopamine. The NICU team asks whether to send him to the scanner now, at 30 hours, because the parents want to know how severe the injury is, and whether the cooling blanket can be removed for transport.
1. When should the first MRI be performed, and what would a scan obtained at 30 hours show or fail to show? Justify your answer with the pathophysiology of hypoxic-ischemic injury and the practical constraints of scanning this patient.
Read the worked answer
Recommendation: image at 72–96 hours of life, not at 30 hours, unless a specific clinical question requires earlier scanning.
- Why later. Cytotoxic edema evolves over 48–96 hours. At 30 hours, diffusion restriction in the ventrolateral thalami and posterior putamina may already be visible, but the full extent is not demarcated. T1 and T2 signal are frequently normal this early, so an early study risks a falsely reassuring result that drives prognostic counseling in the wrong direction.
- Pseudonormalization. ADC values normalize at roughly 7 days and can mask injury afterward; the 72–96 hour window captures maximal conspicuity on DWI/ADC before this occurs.
- Hypothermia effect. Cooling slows the evolution of injury, further arguing against a very early read.
- Practical risk. Off-unit transport, sedation, ventilator manipulation, and loss of thermoregulation all carry real risk. Cooling should be maintained with an MRI-compatible blanket and rewarming should follow protocol at 72 hours rather than being driven by the scanner schedule.
- What the team already knows. Burst-suppression on aEEG plus the neurologic exam is already strongly prognostic; MRI refines but will not change the next 48 hours of management.
If there is a compelling question (suspected stroke, unexplained seizures, surgical issue), scan earlier - but plan a repeat at 72–96 hours.
2. Profound acute asphyxia from abruption produces a different injury pattern than a prolonged partial insult. Which pattern do you predict here, which structures will you scrutinize, and which sequences and planes would you include?
Read the worked answer
Predicted pattern: acute profound (deep gray matter) injury. Cord pH 6.82 with a base deficit of 19 after abruption indicates a sudden, near-total interruption of placental blood flow rather than a slow partial one.
- Structures at risk. Ventrolateral thalami, posterior putamina, perirolandic (sensorimotor) cortex, hippocampi, brainstem tegmentum, and the posterior limb of the internal capsule (PLIC).
- Contrast pattern. A prolonged partial insult preferentially injures parasagittal watershed cortex and subcortical white matter. Real injuries are often mixed, so describe both distributions.
- Sequences.
- Axial DWI and ADC as the core sequence; small structures demand thin sections and no gap.
- Axial and coronal T1 inversion recovery to assess the normal T1 hyperintensity of the PLIC - its loss is the single most useful prognostic sign.
- T2, plus SWI or GRE for hemorrhage, thrombosis, and venous infarction.
- Single-voxel MR spectroscopy (TE 144) over the basal ganglia for a lactate peak and reduced NAA.
- Consider MRV if sinus thrombosis is a concern.
- Reporting. Describe each involved structure explicitly rather than writing 'hypoxic-ischemic encephalopathy'; the neonatal team uses the pattern and the PLIC sign to counsel parents and to plan feeding and follow-up.
3. The day-4 study shows restricted diffusion in both thalami and posterior putamina, loss of the normal T1 hyperintensity of the right PLIC, a lactate peak, and a small asymptomatic subdural collection from birth trauma. How do you frame the report and the follow-up imaging plan?
Read the worked answer
Framing the report.
- Bilateral deep gray injury with unilateral PLIC sign loss plus a lactate peak carries a high likelihood of abnormal motor outcome, particularly extrapyramidal/dyskinetic cerebral palsy. State this as a probability, not a certainty, and anchor it to the structures involved - the PLIC sign is the strongest single predictor and should be called out with laterality.
- Use graded language ('high risk,' 'consistent with') and avoid precise outcome percentages, which are not derivable from one scan.
- The subdural collection should be described, attributed to instrumentation/birth trauma, and explicitly labeled as not explaining the encephalopathy, so it is not over-interpreted by the family.
- Recommend early referral to neonatal follow-up, neurology, and early intervention rather than waiting for imaging to 'settle.'
Follow-up imaging.
- No routine repeat MRI in the newborn period. The injury pattern is established and repeat scanning adds sedation risk without changing management.
- Consider a single MRI at 3–6 months (or earlier if there is a shunt, progressive ventriculomegaly, or a new clinical question) to assess volume loss, cystic encephalomalacia, and myelination.
- Emphasize that serial neurologic examination and developmental assessment, not imaging, drive long-term prognostication after the initial study.