Neuroimaging Findings in CHANTER Syndrome: A Case Series

Radiographic Features

The 3 patients in our series had consistent NCCT findings. The initial CT findings showed cerebellar edema with eventual mass effect on the fourth ventricle and intraventricular obstructive hydrocephalus requiring shunting and decompression. A varying degree of involvement of the basal ganglia and cortex was noted on the initial CT, with subsequent development of these findings on CT the following day. All patients had vessel imaging showing no large-vessel occlusion to suggest arterially mediated infarctions. No evidence of hemorrhage was seen on NCCT in any of these cases. CTP and regional CBF images in 1 patient showed no evidence of a stroke or decreased blood flow to the basal ganglia and cerebellum. This finding suggests that the injury in CHANTER is not related to vascular flow restriction but rather likely internal cellular injury.

While patients who present in an unconscious state are likely to have NCCT as their first imaging study, MR imaging findings, especially diffusion restriction, are the distinct radiologic hallmark of CHANTER and necessary for diagnosis. The initial MR imaging findings among our 3 patients were also consistent. On initial MR imaging, all patients showed diffusion restriction in the cerebellum, hippocampi, basal ganglia, and cortex. Only 1 of the patients had subcortical WM involvement of the centrum semiovale. Initial MR imaging was performed within several hours for 2 patients but was delayed in patient 3. A follow-up MR imaging generally obtained 2–3 weeks later in all cases showed expected injury evolution with normalization of signal on DWI, FLAIR hyperintensity, and postcontrast enhancement in these areas, consistent with evolving injury. All patients had TOF-MRA demonstrating the absence of occlusion in large vessels.

The radiologic findings among the 3 patients in our series are largely consistent with those of the 6 patients in the Jasne et al¹ series, including the initial CT findings, development of obstructive hydrocephalus following cerebellar edema, and diffusion restriction in the cerebellar cortex, hippocampi, and basal ganglia, typical of CHANTER. The only substantial discrepancy was the presence or absence of cerebral cortical involvement. Jasne et al reported that all 6 of their patients had the typical diffusion restriction findings but sparing of the cerebral cortex. However, all 3 patients in our series had some extent of cerebral cortical involvement, though this was not a dominant feature. The involvement of subcortical WM was rare in both case series. Only 1 patient in the series of Jasne et al showed areas of restricted diffusion in the subcortical WM. In contrast, no patients in this series exhibited such findings.

Differential Diagnoses and Classification

Acute ischemic stroke is unlikely to be confused with CHANTER because patients with CHANTER do not have any vessel occlusion on imaging. Additionally, patients with CHANTER typically have injuries in multiple vascular territories. A related pathology that is closer with respect to radiologic features to CHANTER is HIE. Patients with HIE can present in an unresponsive state, similar to that in patients with CHANTER. However, they do not necessarily have acute exposure to drugs of abuse. The outcomes are typically extremely poor compared with patients with CHANTER syndrome.² This pattern of injury tends to affect areas of high metabolic demand. Notably, cerebellar and hippocampal edema can be seen in HIE. However, when this occurs, other areas, such as the cerebral cortex, are often diffusely involved.^8,9 In HIE, NCCT will typically show a diffuse loss of gray-white matter differentiation and effacement of the sulci with cytotoxic edema.^10,11 These findings were not seen in any patients classified as having CHANTER.

Additionally, in HIE, the cerebellum is more likely to be spared, giving rise to the white cerebellum sign, contrary to the cerebellar hypoattenuation seen on CT in all patients with CHANTER syndrome. Last, while obstructive hydrocephalus developed in all patients with CHANTER syndrome, it is not common in patients with HIE.¹² However, it is possible that some patients described here and previously with CHANTER syndrome might also have some degree of superimposed HIE, which might partially account for the variability in outcomes.

Heroin-associated spongiform leukoencephalopathy is a toxic leukoencephalopathy with distinct imaging features. Because it is exclusively associated with heroin vapor inhalation, it is also known as “chasing the dragon” leukoencephalopathy. Although it can present with cerebellar restricted diffusion, it almost always presents with changes in the WM, including symmetric confluent deep WM hyperintensity on T2 and FLAIR sequences, with sparing of the gray matter, which is not typical of CHANTER.^11,13 Contrary to CHANTER, it also presents during a more extended period.

Another recent case remarkably similar to CHANTER consisted of cerebellar edema in a young child following accidental opioid ingestion and fentanyl administration.^14,15 It appears that this case, as well as several others in the pediatric population, may fit under the recently described constellation of POUNCE.^{6,16⇓⇓⇓⇓-21} Children with POUNCE are generally accidentally exposed to opioids and present with a decreased level of consciousness. Similar to CHANTER, the prominent feature of POUNCE is malignant cerebellar edema, with diffusion restriction occasionally reported.^14,18,22 This cerebellar edema can lead to obstructive hydrocephalus.¹⁴ However, contrary to CHANTER, these cases rarely have hippocampal and basal ganglia involvement, instead involving WM areas with lesions that are hyperintense on T2 and hypoattenuated on CT.^{14,15,18,20,22} Some of these differences in injury patterns have been speculated to be due to the differences in opioid receptor expression between childhood and adulthood.^23,24

Another syndrome that has some overlapping features with CHANTER is OAA. Patients with OAA present with new-onset amnesia after exposure to opioids, most notably fentanyl.^{7,25⇓⇓⇓-29} Radiologically, the prominent finding is symmetric diffusion restriction and FLAIR hyperintensity of the hippocampi. Edema in the cerebellum and basal ganglia occurs only rarely, and patients do not present with obstructive hydrocephalus. Presentation and outcomes are milder than in CHANTER, with amnesia as the defining deficit.

The similarities between patients with CHANTER, POUNCE, and OAA perhaps indicate that these syndromes exist along a spectrum with several overlapping clinical and imaging features and a common underlying etiology (Table). It has been thought that CHANTER represents the more severe end of this spectrum, while OAA represents a milder form.¹

Febrile infections, including acute cerebellitis, would also be on the differential. However, apart from a different clinical history, cerebellitis is more likely to have leptomeningeal and cortical enhancement in the acute setting and less likely to have extensive supratentorial involvement, for example, in the basal nuclei, cortex, and hippocampi as in CHANTER.^29,30

Etiology and Mechanism

There is a clear association of the presentation of CHANTER syndrome and acute intoxication with drugs of abuse, particularly opioids. Four of the 6 patients identified by Jasne et al¹ had acute exposure to opiates. In our series, all 3 patients had a fentanyl overdose. Thus, although the pathogenesis of CHANTER is still uncertain, it appears that direct opioid toxicity plays a key role. The injury pattern is thought to be due to a combination of neurotoxicity and hypoxic effects, resulting in mitochondrial failure with anoxic injury.¹ Other proposed mechanisms for the effect of opioids on the cerebellum have been proposed in POUNCE syndrome, including direct oligodendrocyte toxicity, demyelination, and apoptotic upregulation.^6,31,32 An underlying susceptibility to CHANTER syndrome due to genetic or environmental factors might also be present. The expression of the OPRM1 gene, which encodes the μ-opioid receptor, which is the primary target of morphine-derivative opioids, is higher in key regions affected in CHANTER, particularly the cerebellum.^33⇓-35

In the adult literature, heroin-associated spongiform leukoencephalopathy has preferential cerebellar involvement. It has been postulated that this is due to the high concentration of opioid receptors in the cerebellum,^22,36 which are also highly expressed in affected regions of the WM.^33⇓-35 The injury in OAA is thought to be due to fentanyl-induced hippocampal excitotoxicity in the context of hypoventilatory hypoxemia.³⁷ Accordingly, we propose that the regional topography of CHANTER might be due to an overlap of brain regions most vulnerable to both opioid toxicity and hypoxic-ischemic injury, with similar conditions falling on either side of this spectrum. However, although the patients in our case series all had acute exposure to fentanyl specifically, the exposures reported by Jasne et al¹ appear more heterogeneous, including 2 not having any acute opioid exposures.