The DWI 'reversal sign' of white matter hypoxic ischaemic injury in older children: an unusual MRI pattern for age

Prognostic value of magnetic resonance imaging performed during the subacute phase in adult patients with hypoxic-ischemic encephalopathy for long-term neurological outcomes

OBJECTIVE

To investigate the value of a model based on brain magnetic resonance imaging (MRI) performed in the subacute phase (between the 1st and 30th day) in predicting long-term neurological outcomes of adult hypoxic-ischemic encephalopathy (HIE) patients.

METHODS

Ninety-six adult HIE patients who underwent conventional MRI and diffusion-weighted imaging (DWI) during the subacute phase were retrospectively analyzed. Favorable (Cerebral Performance Categories (CPC) 1-2) and unfavorable outcome (CPC 3-5) groups were created based on patient neurological status approximately three months after the onset of hypoxic-ischemic events. A multivariate stepwise regression model was applied after univariate analysis of MRI findings, and then the overall MRI score, Alberta Stroke Program Early Computed Tomography Score (ASPECTS), Bilateral ASPECTS (Bi-ASPECTS), modified ASPECTS (mASPECTS) and Bi-ASPECTS combined with posterior circulation ASPECTS (PC-ASPECTS) were calculated based on MRI findings. Receiver operating characteristic (ROC) curves were used to assess prognostic accuracy.

RESULTS

Both univariate and multivariate analyses showed the cerebral cortex and cerebellum, neostriatum, hippocampus, brainstem and postanoxic leukoencephalopathy were independent prognostic factors for unfavorable outcomes. The multivariate regression analysis resulted in an overall classification accuracy of 84.4%, a sensitivity of 84.2% (95% CI, 71.6-92.1%), and a specificity of 92.3% (95% CI, 78.0-98.0%) for unfavorable outcomes. The model had an areas under the ROC curves (AUC) of 0.944 (95% CI, 0.901-0.987); the MRI overall scores were 0.918 (95% CI, 0.866, 0.971), ASPECTS 0.839 (95% CI, 0.755, 0.923), Bi-ASPECTS 0.837 (95% CI, 0.753, 0.922), mASPECTS 0.851(95% CI, 0.771, 0.931) and Bi-ASPECTS+PC-ASPECTS 0.876 (95% CI, 0.806, 0.946).

CONCLUSIONS

The multivariate model based on conventional MRI combined with DWI performed in the subacute phase could help predict the prognosis of adult HIE with high performance.

Can Apparent Diffusion Coefficient Predict the Clinical Outcome in Drowned Children?

Introduction

Pediatric cerebral hypoxic-ischemic injury frequently results in severe neurological outcome. Imaging with diffusion-weighted magnetic resonance imaging (DWi) demonstrates that the acute cerebral injury and apparent diffusion coefficient (ADC) allow the assessment of the severity of brain damage. The main objective was to examine if spatial distribution of reductions in ADC values is associated with clinical outcome in drowned children.

Methods

This is a retrospective study of 7 children (7 examinations) suffering from a hypoxic-ischemic event who underwent DWi. Seven subjects with normal DWi served as controls. The mean patient age was 4.88 ± 2.93 years and the male-to-female ratio was 5:2. The neurological outcome was divided into 2 categories: 4 children with Apallic syndrome and 3 deaths. We analysed the differences between the drowned children and the control group regarding clinical data, DWi abnormalities, and ADC values.

Results

The ADC values in the occipital and parietal grey matter were significantly different between the drowned children (765.14 ± 65.47 vs 920.95 ± 69.62; P = .003) and the control group (670.82 ± 233.99 vs 900.66 ± 92.72; P = .005). The ADC showed low values in the precentral area also ( P = .044).

Conclusion

The ADC reduction may be useful to predict the poor outcome in drowned children and can be a valuable tool for clinical assessment.

Anatomical patterns and correlated MRI findings of non-perinatal hypoxic-ischaemic encephalopathy

Non-perinatal hypoxic-ischaemic encephalopathy (HIE) has varying anatomical patterns dependent on the type of insult, the degree and duration of cerebral hypoxia, or presence and degree of hypoperfusion. Profound insults can affect the entire cerebral cortex or just the perirolandic cortex, the cerebellum and the deep grey matter structures. Less severe insults may affect only the watershed regions. The objective of this article is to review the anatomical patterns of non-perinatal HIEs by MRI.

CT and MR in non-neonatal hypoxic-ischemic encephalopathy: radiological findings with pathophysiological correlations

Non-neonatal hypoxic-ischemic encephalopathy is a clinical condition often related to cardiopulmonary arrest that demands critical management and treatment decisions. Management depends mainly on the degree of neurological impairment and prognostic considerations. Computed tomography (CT) is often used to exclude associated or mimicking pathology. If any, only nonspecific signs such as cerebral edema, sulci effacement, and decreased gray matter (GM)/white matter (WM) differentiation are evident. Pseudosubarachnoid hemorrhage, a GM/WM attenuation ratio <1.18, and inverted GM attenuation are associated with a poor prognosis. Magnetic resonance (MR) imaging is more sensitive than CT in assessing brain damage in hypoxic-ischemic encephalopathy. Some MR findings have similarities to those seen pathologically, based on spatial distribution and time scale, such as lesions distributed in watershed regions and selective injury to GM structures. In the acute phase, lesions are better depicted using diffusion-weighted imaging (DWI) because of the presence of cytotoxic edema, which, on T2-weighted images, only become apparent later in the early subacute phase. In the late subacute phase, postanoxic leukoencephalopathy and contrast enhancement could be observed. In the chronic phase, atrophic changes predominate over tissue signal changes. MR can be useful for estimating prognosis when other tests are inconclusive. Some findings, such as the extent of lesions on DWI and presence of a lactate peak and depleted N-acetyl aspartate peak on MR spectroscopy, seem to have prognostic value.