Dextromethorphan-Associated Neurotoxicity with Cerebellar Edema Syndrome in Young Children: Neuroimaging Features

DISCUSSION

Dextromethorphan-associated neurotoxicity in young children is a little-known entity. We describe 3 pediatric patients who presented with significant neurologic deterioration after ingestion of over-the-counter antitussive medications containing dextromethorphan for their short febrile respiratory illness. The patients in our series had strikingly similar clinical features and MRI of the brain appearances like those of the recently described POUNCE syndrome.¹

POUNCE syndrome is a term coined by Kim et al¹ for a distinct clinico-radiologic syndrome caused by acute opioid toxicity in children. It is characterized by cerebellar edema in the form of T2-FLAIR hyperintensities with or without restricted diffusion. It is often accompanied by some degree of supratentorial white matter injury, predominantly in the deep and periventricular white matter, seen as punctate or confluent T2-FLAIR hyperintensities, which may show restricted diffusion in the early stages.¹ A few of the previously published reports on pediatric opioid toxicity have also described similar cerebellar and white matter injury, often with malignant cerebellar edema leading to complications, including cerebellar tonsillar herniation, compression of brainstem, and rapidly evolving obstructive hydrocephalus. These complications may lead to dreaded consequences, including death if not addressed in early stages by antiedema measures or posterior fossa decompression or placement of an external ventricular drainage tube if necessary.^{1⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓-12}

All the children in our report presented with profound encephalopathy and respiratory failure and had cerebellar edema on MRI. Two of them showed supratentorial involvement. None of the children showed hemorrhages or contrast enhancement, though hemorrhage has been reported in a few of the previously published cases of POUNCE.⁶ Our patients did not show significant compression of the brainstem or upstream hydrocephalus and did not require surgical measures, likely due to early recognition and administration of appropriate therapy.

Dextromethorphan, an over-the-counter antitussive medication, is a D isomer of levorphanol, which is a synthetic analog of codeine, an opioid receptor agonist. Its cough-suppressant actions in regular doses are mediated via its action on σ opioid receptors in the medulla. Due to its stereochemistry, it does not bind to μ and δ opioid receptors, which are commonly involved in manifestations of opioid toxicity.^13,14

Although definite histopathologic evidence is lacking to date, POUNCE is postulated to occur due to direct neurotoxic effects of the opioid agents on glial and neuronal cells, with contributions from factors like apoptotic upregulation due to mitochondrial injury and potentiation of these mitochondrial pathways from anoxic insults secondary to respiratory depression.^1,2,4 Predominant cerebellar involvement in POUNCE is proposed due to an abundance of μ receptors in the cerebellum, which are the primary target for opioid toxicity.¹

Although dextromethorphan primarily binds with σ opioid receptors, due to a striking resemblance of the clinical and radiologic syndrome in all the cases of our series with POUNCE, we postulate the occurrence of DANCE syndrome in small children with dextromethorphan toxicity and propose a similar mechanism with the possibility of cross-reactivity with μ receptors. Over-the-counter cough remedies containing dextromethorphan have been largely prohibited for children younger than 2 years of age and are not recommended for use in those younger than 4 years of age.^14⇓-16 The occurrence of DANCE syndrome in young children in our report further emphasizes the necessity of this precaution. A recently published case report has highlighted clinical features of opioid toxidrome secondary to dextromethorphan toxicity in children, but it is extremely rare and there is a paucity of knowledge about the associated neuroimaging features.¹⁷ Although never proved, the occurrence of dextromethorphan toxicity in only young children may be due to the enhanced sensitivity of the receptors in this age group, leading to a heightened effect even at doses less than the toxic range. Further research into this matter is warranted.

Because all the children in our series had a respiratory illness and showed good response to methylprednisolone, one of them had high interleukin-6 levels, and one had positive IgG for COVID-19, the possibility of an underlying parainfectious immunologic phenomenon and the role of viral infection in the increasing predisposition to DANCE syndrome also need to be investigated. In this context, the most important differential of DANCE syndrome remains viral/infectious cerebellitis. However, a specific MR pattern involving both the cerebellum (in all the children) and supratentorial white matter (in 2 of the 3), a remarkably similar clinical context with features including pinpoint pupils, a striking resemblance to POUNCE syndrome, a lack of contrast enhancement, and normal findings on blood and CSF investigations despite extensive infectious disease work-up make the diagnosis of an infectious/viral cerebellitis less likely than DANCE. Hypoxic-ischemic encephalopathy is another major clinical differential of DANCE syndrome in patients with sudden unresponsiveness and respiratory failure. Predominant involvement of the cerebellar white matter and centrum semiovale without cortical and basal ganglia involvement on MRI of the brain makes the distinction from hypoxic-ischemic encephalopathy possible, when the cerebellum is usually involved late.^1,18,19

A similar predilection for the cerebellum and white matter in cases of “chasing the dragon” leukoencephalopathy occurring due to chronic rather than acute inhalation of heroin and the recently described cerebellar, hippocampal, and basal nuclei transient edema with restricted diffusion syndrome and opioid-associated amnestic syndrome occurring secondary to acute opioid toxicity in adults suggests that all these disorders may represent a continuum of similar pathophysiology.^18⇓⇓-21 Nevertheless, there are differences (Online Supplemental Data) among these entities possibly due to varying binding affinity of opioid receptors in different parts of the brain in children and adults.²²

Our report has a few limitations. We could not determine the serum levels of dextromethorphan in the children in our series. The lack of supratentorial involvement in the third child in our study could not be explained. This may be related to the difference in the sensitivity of the brain receptors at different ages because the third child was the youngest of the 3 (younger than 2 years of age), but we refrain from making a definite assumption without any histopathologic or other conclusive evidence from the literature. Due to logistic reasons, MR follow-up was not available in 2 of the 3 patients. However, normal neurologic examination findings were reassuring and indicated a definite clinical improvement in these patients.

DANCE syndrome can have an extremely poor prognosis if untreated as could be seen in the sibling of the patient in case 1 who succumbed to death before reaching our hospital. However, if timely recognized and optimally managed, it can have a good clinical outcome despite a significant initial insult as seen in all the patients of our series, likely owing to the neuroplasticity of the pediatric brain.¹ This experience highlights the importance of early neuroimaging and the need to promptly recognize the specific DANCE pattern and urgently treat this rare entity. The case series reiterates the adverse effects of dextromethorphan in young children and the need to monitor its use, especially in children younger than 4 years of age.