Acute flaccid myelitis associated with enterovirus D68: An emerging infectious disease

Imagine your 5-year-old child having an unremarkable “cold,” then waking up 3 days later unable to walk and to move his left arm. There is no treatment, and the prognosis is uncertain. In Canada, where polio has been eliminated, the most likely diagnosis is acute flaccid myelitis (AFM).

To understand this relatively novel condition, it is worthwhile to reflect on poliomyelitis, a disease that most of us have never seen. The earliest records of poliovirus (PV) infection date back to the Eighteenth Dynasty of Egypt (1403–1365 BC), with carvings illustrating a man with a shrunken limb. PV was the first enterovirus (EV) to be discovered, in 1908. There are three serotypes of PV (PV1, PV2, and PV3) which are spread by the fecal–oral route with prolonged shedding in the stool, making stool testing the optimal diagnostic test (1–3). PV caused a wide clinical spectrum of disease, including acute flaccid paralysis (AFP) in approximately 0.1% of the cases. The World Health Organization (WHO) defines AFP as the sudden onset of paralysis/weakness in any part of the body in a child younger than 15 years of age (4,5). Since the introduction of polio vaccines in 1955, the worldwide incidence of polio has been reduced by 99.9% with a few endemic cases of wild-type PV1 disease still reported in Afghanistan, Pakistan, and Nigeria (6).

Human non-PV EV are ubiquitous and are subdivided into four species (A, B, C, and D). As many as 80 different EV serotypes can infect humans (4,7). Most EV cause only mild clinical syndromes, such as hand–foot–mouth disease, but some can also lead to devastating non-polio AFP, as documented in a systematic review recently published by Suresh et al (8).

EV-D68 had only been reported sporadically worldwide between 1962 and 2012, but rose to prominence in 2014 when the largest EV-D68 outbreak in history occurred, with up to 2,287 cases worldwide (1,4). Among cases with respiratory illness, there was a new worrisome identification of myelitis resulting in paralysis. During this period, 120 AFP cases were declared in the United States, and a Canadian group described 25 cases in four provinces (Ontario, British Columbia, Alberta, and Manitoba) (9). Based on the clinical and radiological features of these cases, the Centers for Disease Control and Prevention (CDC) adopted a standardized case definition of AFM. A confirmed AFM case is defined as acute focal limb weakness with magnetic resonance imaging (MRI) showing mainly grey matter lesions involving ≥1 spinal cord segments, whereas a probable AFM case is defined as acute focal limb weakness with cerebrospinal fluid (CSF) pleocytosis (>5 cells/mm³) (10).

Changing epidemiology

The EV-D68 virus was first described in 1962 in California, where it was isolated from the oropharynx of children hospitalized with acute wheezing and pneumonia. Distinctive features of EV-D68, as compared with other EV, include: lower optimal growth temperature, leading to better replication in the nasal cavity; and acid sensitivity, hence the inability to survive the passage in the stomach and thereby making it difficult to recover from stool (1,2,4). A study published in 2010 showed that the EV-D68 genome evolved tremendously since its initial discovery in 1962 which may have contributed to increased virulence and fitness culminating in the large outbreak that occurred in 2014 (4).

Since then, multiple reports have been published worldwide linking EV-D68 and AFM cases (9,11–18). Moreover, CDC surveillance data has revealed a clear seasonal and biennial pattern (1). A Morbidity and Mortality Weekly Report published in November 2018 noted a three-fold increase in the number of confirmed AFM cases (n = 80) when compared with the same period the previous year (19). Similarly, in a report published in February 2019, the United Kingdom AFP Task Force also highlighted the recent rapidly rising number of AFM cases associated with EV-D68 (20).

Evidence of causality

Identification of EV-D68 in clinical specimens from confirmed AFM cases worldwide supports an association between the virus and the clinical syndrome. However, the usual absence of EV-D68 in neurologic specimens (i.e., CSF) and, until recently, the absence of an animal model had left the causal relationship between EV-D68 and AFM unproven.

In 2017, a study published on an experimental mouse model showed that EV-D68 strains from the 2014 outbreak induced a paralytic disease resembling human AFM. The mice inoculated with EV-D68 strains developed paralytic disease. Spinal cord homogenates from paralyzed mice were then made into a lysate, which showed a cytopathic effect in cell culture. The supernatant was then collected and inoculated back into a new mouse that subsequently developed paralytic disease. This demonstrated the fulfillment of Koch’s postulates, thus establishing a causal role between EV-D68 and paralytic disease resembling AFM in an animal model. Subsequently, the presence of EV-D68 viral proteins was seen in motor neurons of paralyzed mice. The authors favoured the hypothesis of a direct viral injury as the pathophysiological process rather than an immune-mediated post-infectious phenomenon to explain the development of neurological damage secondary to EV-D68 infection (21).

To further strengthen the purported causality link, the authors of a study published in The Lancet Infectious Diseases applied the Bradford Hill criteria to EV-D68 and AFM (22). They concluded that seven out of nine criteria are met, including strength of association, consistency, temporality, plausibility, coherence, experiment, and analogy. The only unmet criteria are: specificity, given that multiple pathogens can induce AFM; and biological gradient, given that no dose-response relationship has been demonstrated between EV-D68 exposure and severity of clinical presentation (23).

To our knowledge, only one pathological report from a fatal human central nervous system (CNS) EV-D68 infection has been published. In this case, identification and sequencing of EV-D68 in the CSF, as well as the characteristic histopathologic pattern seen in the CNS, were able to establish the virus as the cause of death (24).

Clinical and radiological features of AFM

AFM predominantly affects pre-school age children, and in the majority of cases, neurologic symptoms are preceded by a non-specific viral prodrome. Paralysis is generally of rapid-onset (hours to days) and is almost always asymmetrical. The pattern of weakness is consistent with a lower motor neuron neuropathy, including hypo or areflexia as well as hypotonia. Cranial nerve involvement is not uncommon, and bowel/bladder dysfunction is occasionally reported. The absence of sensory deficits is also a hallmark of AFM. Imaging by MRI usually shows a predominant T2 hyperintense signal of the spinal cord grey matter. The anterior horn is classically affected, but this finding is not universal. CSF pleocytosis is virtually always present if lumbar puncture is done early in the course of the disease (25–31).

Outcome and therapeutics

So far, studies have shown that patients with AFM experience limited motor recovery associated with significant persistent disability requiring intensive rehabilitation, as illustrated by the experience of the Johns Hopkins Transverse Myelitis Center published in December 2018 (27).

Therapeutic options are currently limited since no targeted therapies or interventions (such as corticosteroids, intravenous immunoglobulins, plasma exchange, or antiviral medication) have been proven effective against AFM (29–33).

Microbiologic diagnosis of AFM

Identification of an etiologic agent in AFM cases is time-dependent. Establishing a microbiologic diagnosis requires sampling from all appropriate sites early in the clinical course of the disease. Furthermore, although molecular testing is the gold standard for diagnosis of EV infections, most commercial multiplex polymerase chain reaction (PCR) assays are not able to differentiate between EV and rhinovirus isolates. Therefore, specific EV in-house developed PCR assays have better analytical sensitivity for detection of EV compared with multiplex PCR assays (1,4). Respiratory samples (both from the nasopharynx and throat) should always be collected early in the clinical course when trying to document EV-D68 infection. Genotyping or serotyping is necessary (and mandatory) in cases of AFP, and obtaining stool samples for PV viral testing is essential for this purpose. Other infectious causes of AFP should also be considered (i.e., Powassan or West Nile virus, herpes viruses) (31,34–36).

AFM surveillance in Canada

In Canada, AFP cases in children <15 years old are monitored based on WHO recommendations, to ensure the country remains polio-free. This surveillance is the joint project of the Canadian Paediatric Society and the Public Health Agency of Canada (PHAC) through the Canadian Paediatric Surveillance Program. AFM is included in the Canadian AFP surveillance program, which also captures cases of other neurologic syndromes such as Guillain–Barré and transverse myelitis (5,37).

According to data on the PHAC website (last accessed on March 19, 2019), there are between 27 and 51 cases of AFP reported annually. Between January 1, 2018 and March 6, 2019, 77 cases of AFP were declared, with 49 cases confirmed. In comparison, only 25 cases were confirmed in 2017, supporting the biennial epidemiology (38,39).

Conclusions

AFM associated with EV-D68 is a worrisome emerging infectious disease among children worldwide. A high index of suspicion is essential in order to establish the appropriate diagnosis. Prompt collection of relevant microbiologic samples is critical to identify the etiologic agent. Ruling out poliomyelitis is also crucial in suspected AFM cases to maintain confidence in polio eradication worldwide. Whether vaccines can be developed for non-PV EV–causing neurologic disease remains to be seen.

Competing Interests:

The authors have nothing to disclose.

Ethics Approval:

N/A

Informed Consent:

N/A

Registry and the Registration No. of the Study/Trial:

N/A

Animal Studies:

N/A

Funding:

No funding was received for this work.

Peer Review:

This article has been peer reviewed.