2014 outbreak of enterovirus D68 in North America

BACKGROUND

From August to October of 2014, the largest and most widespread outbreak of enterovirus D68 (EV-D68) disease described to date occurred in North America. EV-D68 is a species D enterovirus in the Picornaviridae family of small nonenveloped viruses with single-stranded, positive sense RNA genomes [Pallansch and Roos, 2013]. It has an icosahedral capsid composed of four structural proteins: VP1-3, comprising the outer capsid, and VP4, the internal surface of the capsid shell [Liu et al., 2015]. In contrast to other enteroviruses, EV-D68 is biologically more similar to human rhinoviruses [Oberste et al., 2004]. In fact, human rhinovirus 87, discovered in 1963, was subsequently reclassified using molecular analysis as EV-D68 [Blomqvist et al., 2002; Ishiko et al., 2002]. Similar to rhinoviruses, EV-D68 optimally grows at 33°C compared to 37°C preferred by other enteroviruses, and is both heat and acid labile [Oberste et al., 2004]. These characteristics and affinity for α2-6-linked sialic acids typically found in the upper respiratory tract make the respiratory tract the preferred target for EV-D68 replication [Imamura et al., 2014a]. Likewise, transmission of EV-D68 is thought to occur primarily via respiratory secretions. This is unlike other enteroviruses which, with their heat and acid stability, survive and preferentially replicate in the gastrointestinal tract and are primarily spread via the fecal-oral route [Pallansch and Roos, 2013].

EV-D68 was first isolated by Schieble et al. [1962] in respiratory specimens from four children in California with lower respiratory tract disease [Schieble et al., 1967]. Between 1970 and 2005, 26 cases of EV-D68 infection were identified in the US by the passive Centers for Disease Control and Prevention (CDC) National Enterovirus Surveillance System (NESS), representing only 0.1% of reported enterovirus infections [Khetsuriani et al., 2006]. From 2008 to 2010, small clusters of EV-D68 respiratory illness were reported in North America (United States: GA, PA, AZ, NY), Asia (Philippines, Japan, Cambodia), and Europe (Netherlands, Italy, France) [Centers for Disease Control and Prevention, 2011; Hasegawa et al., 2011; Imamura et al., 2011; Kaida et al., 2011; Rahamat-Langendoen et al., 2011; Ikeda et al., 2012; Renois et al., 2013]. An expanding number of small clusters of EV-D68 respiratory disease were reported between 2010 and 2014 in Europe (England, Italy, France, Finland, Netherlands), Asia (Philippines, Thailand, China, Japan), Oceania (New Zealand, Australia), and Africa (Gambia, Senegal, S. Africa, Kenya) [Imamura et al., 2011; Meijer et al., 2012; Tokarz et al., 2012; Todd et al., 2013; Lu et al., 2014; Meijer et al., 2014; Opanda et al., 2014; Piralla et al., 2014; Furuse et al., 2015; Levy et al., 2015].

EPIDEMIOLOGY OF THE 2014 ENTEROVIRUS D68 OUTBREAK IN NORTH AMERICA

On August 15, 2014, clinicians at Children's Mercy Hospital in Kansas City, Missouri reported an unusual increase in children presenting with severe respiratory disease and increased detection of rhinoviruses/enteroviruses on multiplex PCR testing of respiratory specimens [Oermann et al., 2015; Schuster and Newland, 2015]. Sequencing of rhinovirus/enterovirus-positive samples conducted at the CDC identified EV-D68 as the predominant virus implicated in this outbreak. Similar increases in severe respiratory disease were reported soon thereafter from University of Chicago Comer Children's Hospital and Children's Hospital Colorado, with EV-D68 similarly identified in a majority of evaluated cases [Midgley et al., 2014, 2015a]. Subsequently, hospitals throughout the country reported surges in pediatric patient volumes and resource utilization during the outbreak period [ED Management Editors, 2014; Shaw et al., 2014].

In total, 1,153 EV-D68 infections were confirmed in 49 US states and the District of Columbia by the CDC [Centers for Disease Control and Prevention, 2015]. Because of the limited microbiologic confirmatory testing conducted, this number underestimates the likely millions of cases of milder disease that occurred throughout the US [Centers for Disease Control and Prevention, 2015]. For example, although only 101 samples had confirmatory EV-D68 testing at Children's Hospital Colorado, syndromic surveillance demonstrated 1185 excess emergency department visits, 387 excess hospitalizations, and 96 excess PICU admissions for respiratory disease above expected volumes from August to September 2014 during the EV-D68 outbreak [Messacar et al., in press]. Similarly, there were dramatic increases in emergency room visits, hospitalizations, and PICU admissions in Northern Illinois and Northwest Missouri during this same time frame in 2014 compared to the previous two years [Midgley et al., 2014]. During the outbreak, EV-D68-specific testing to confirm infection was only available through research laboratories and the CDC, and focused on those with severe illness. There were no active population-based surveillance efforts in place during the outbreak. Therefore published reports are convenience samples that may not accurately represent the extent or spectrum of EV-D68 epidemiology.

In September of 2014, clusters of severe respiratory disease with detection of EV-D68 were reported in Canada [Edwin et al., 2015]. A time-limited enhanced surveillance program of the Public Health Agency of Canada identified 268 cases of EV-D68-associated respiratory disease during September 2014 in Ontario, Alberta, and British Columbia [Edwin et al., 2015]. Syndromic surveillance in Alberta identified an increase in daily ED visits in August–September 2014 for influenza-like illness, shortness of breath, and cough/congestion compared with 2013, with a predominance of EV-D68 among respiratory viruses identified [Drews et al., 2015]. Subsequent to the 2014 North American outbreak, increased detection of EV-D68 was noted in Europe, including the Netherlands [Meijer et al., 2014; Poelman et al., 2015], Germany [Reiche et al., 2015], Denmark [Midgley et al., 2015b], Sweden [Dyrdak et al., 2015], Spain [Gimferrer et al., 2015], Italy [Esposito et al., 2015], and France [Lang et al., 2014; Bal et al., 2015]. Two additional cases were also noted in the southern hemisphere in Chile [Torres et al., 2015].

EV-D68-associated respiratory illness was predominantly reported in children during the 2014 North American outbreak. Preschool and school aged children were most frequently identified with EV-D68 infection, with a median age of 5 years amongst confirmed infections [Midgley et al., 2014]. However, the age range of those reported was wide (3 days—92 years) and case reporting derived primarily from children's hospitals likely skewed the data towards the pediatric age group [Midgley et al., 2014]. More than half of children identified with confirmed EV-D68 infection were noted to have a prior history of asthma and these children were more frequently admitted to the ICU [Midgley et al., 2014; Rao et al., 2015; Schuster et al., 2015]. Infections leading to severe respiratory disease in adults with hematologic malignancy and those undergoing hematopoietic cell and solid organ transplants were also reported [Waghmare et al., 2015].

MOLECULAR VIROLOGY

Molecular epidemiological analyses indicate that there has been rapid evolution of EV-D68 since the mid-1990s. This increased genetic diversification has resulted in the emergence of three clades (A–C), multiple sub-lineages, and novel variants [Meijer et al., 2012; Tokarz et al., 2012; Lauinger et al., 2012]. Antigenic variation among newly emerged EV-D68 strains, with loss of neutralization by pre-existing antibodies, has been proposed as a potential explanation for the recent worldwide emergence of EV-D68 disease [Imamura and Oshitani, 2015]. Sequencing of EV-D68 strains circulating during the 2014 outbreak in the U.S. demonstrate that one major lineage (within clade B) predominated (92%), with minor lineages accounting for a minority of cases [Midgley et al., 2015a].

Molecular clock analysis of the major lineage suggests that this strain likely emerged around 2010 and is similar to strains that were circulating in the US, Asia, and Europe in 2011–2012 [Midgley et al., 2015a; Greninger et al., 2015].

CLINICAL MANIFESTATIONS AND MANAGEMENT

Children with laboratory-confirmed EV-D68 during the 2014 North American outbreak most commonly presented with asthma-like symptoms of shortness of breath, cough, and wheezing; increased work of breathing, diminished breath sounds, and hypoxia were noted on exam [Midgley et al., 2015a; Rao et al., 2015; Schuster et al., 2015]. Fever was noted in only a quarter to half of children with confirmed infection [Midgley et al., 2015a; Schuster et al., 2015]. Chest radiographs were frequently abnormal, primarily consisting of nonfocal findings of airways disease, with focal infiltrates in a minority [Rao et al., 2015; Schuster et al., 2015]. Children generally had normal to mildly elevated white blood cell counts and C-reactive protein levels [Rao et al., 2015; Schuster et al., 2015]. Bacterial coinfections were uncommonly reported [Schuster et al., 2015] (Table I).

The vast majority of children with severe disease were managed with supplemental oxygen, bronchodilators, and steroids, typical of asthma management. Use of continuous albuterol and second line asthma medications including magnesium, terbutaline, aminophylline, and epinephrine is indicative of the high acuity observed among affected children [Rao et al., 2015; Schuster et al., 2015]. Of hospitalized children with confirmed EV-D68 infection described in the U.S., a majority (59%) were admitted to the ICU, although disproportionate testing among cases of severe disease skews these data [Midgley et al., 2015a]. In contrast, Canadian surveillance found that only 6.8% of hospitalized children with confirmed EV-D68 infection required ICU care [Drews et al., 2015]. Of U.S. patients in the ICU, many (44–70%) were managed with non-invasive positive pressure ventilation and a minority (7%) required ventilator support [Rao et al., 2015; Schuster et al., 2015]. Children in the ICU with EV-D68 infections had more accelerated presentations, more rapid recovery, and fewer complications than children in the ICU with pandemic H1N1 influenza A virus infections [Rao et al., 2015]. EV-D68 was detected in 14 U.S. patients who died; however, the role of the virus in these deaths is unknown [Centers for Disease Control and Prevention, 2015].

POTENTIAL ASSOCIATION WITH ACUTE FLACCID MYELITIS

In August–September 2014, a cluster of 12 children with acute flaccid paralysis and cranial nerve dysfunction was noted at Children's Hospital Colorado. This cluster geographically and temporally coincided with the EV-D68 respiratory outbreak and five of eleven children (45%) tested had EV-D68 identified in the nasophayrnx [Messacar et al., 2015; Pastula et al., 2014]. Magnetic resonance imaging demonstrated distinctive lesions in the brainstem cranial nerve motor nuclei and anterior horn of the gray matter of the spinal cord in a pattern suggestive of motor neuron tropism and similar to that observed with poliovirus and enterovirus 71 infections [Maloney et al., 2015]. The U.S. CDC conducted enhanced nationwide surveillance to identify cases meeting a case definition of acute flaccid myelitis: acute flaccid paralysis in children <21 years of age with imaging lesions predominantly affecting the gray matter of the spinal cord. To date, 118 children in 34 states with acute flaccid myelitis in 2014–2015 have been reported [Centers for Disease Control and Prevention, 2014]. Seven of 19 (37%) children with acute flaccid myelitis who had samples collected within 14 days of prodromal respiratory illness onset had EV-D68 identified in the nasopharynx; however, EV-D68 (or another virus) has not been identified in the cerebrospinal fluid (CSF) in any case [Greninger et al., 2015; Messacar et al., 2015; Centers for Disease Control and Prevention, 2014]. Two published cases of acute flaccid paralysis in which EV-D68 was identified in the CSF both occurred prior to the 2014 outbreak [Khetsuriani et al., 2006; Kreuter et al., 2011]. All EV-D68 strains identified from respiratory specimens of cases of AFM studied from Colorado and California were clade B1 strains, the predominant circulating strain during the outbreak [Greninger et al., 2015]. The association between EV-D68 and cases of acute flaccid myelitis observed in 2014-2015 remains under investigation [Centers for Disease Control and Prevention, 2014]. Subsequent to U.S. reports, cases of acute flaccid paralysis associated with EV-D68 infection were identified in Canada [Sherwood et al., 2014; Crone et al., 2015], France [Lang et al., 2014], Norway [Pfeiffer et al., 2015], Australia [Levy et al., 2015], and Great Britain [Varghese et al., 2015] (Table II).

DIAGNOSIS

EV-D68 is detectable in respiratory secretions from the upper and lower respiratory tract. In contrast, its acid and heat lability likely leads to lower yield in stool or rectal specimens. In one study of children with EV-D68 pneumonia, virus was detected in the blood of 12 (43%) of 28 patients, most commonly in the first 3 days of illness but up to 7 days post onset of symptoms [Imamura et al., 2014]. Detection of EV-D68 in CSF has been reported in only a small number of cases [Khetsuriani et al., 2006; Kreuter et al., 2011; Levy et al., 2015].

Most commercially available multiplex polymerase chain reaction (PCR) platforms for detection of respiratory viruses detect EV-D68, but these assays cannot differentiate rhinoviruses and enteroviruses due to their genetic similarity [McAllister et al., 2015]. Typing of enteroviruses and rhinoviruses can be conducted using seminested PCR amplification and sequencing of the VP1 or other gene segments; however, this is time and labor intensive [Nix et al., 2006]. Recently developed EV-D68 specific RT-PCR assays provide a more rapid method of detecting and confirming EV-D68, with excellent sensitivity and specificity [Wylie et al., 2015; Piralla et al., 2015; Centers for Disease Control and Prevention, 2015; Zhuge et al., 2015]. More widespread access to EV-D68 specific RT-PCR assays will lead to improved surveillance and understanding of the epidemiologic and clinical spectrum of this virus.

ANTIVIRAL TREATMENT AND PREVENTION

None of the three antivirals in clinical development for enterovirus or rhinovirus infections (pleconaril, vapendavir/BTA-798, pocapavir/V-073) was found to have consistent in vitro activity against 2014 circulating strains of EV-D68 [Rhoden et al., 2015]. Pleconaril was shown to inhibit the 1962 Fermon strain of EV-D68; however, its activity against the 2014 EV-D68 outbreak strain was found to be cell line-specific, with no activity when tested in human rhabdomyosarcoma cells but good antiviral activity when tested in HeLa H1 cells [Liu et al., 2015; Rhoden et al., 2015; Sun et al., 2015]. DAS181, a nebulized drug in clinical development for influenza and parainfluenza virus infections, cleaves α2,6-linked sialic acid receptors to which EV-D68 binds and strongly inhibits EV-D68 at nanomolar concentrations in vitro [Rhoden et al., 2015]. Fluoxetine is the only currently FDA-approved medication shown to have significant activity against 2014 strains of EV-D68 in vitro [Rhoden et al., 2015]. This activity is independent of selective serotonin reuptake inhibition and involves inhibition of RNA replication though viral protein 2C binding [Ulferts et al., 2013]. Commercially available intravenous immune globulin (IVIG) products have been shown to contain high levels of neutralizing antibodies against 2014 outbreak strains of EV-D68 [Zhang et al., 2015]. Thus, IVIG may have a role in the prophylaxis and treatment of EV-D68 infections, particularly in patients with humoral immunodeficiencies which predispose to severe and prolonged enterovirus infections. There are no available vaccines against EV-D68.

In the absence of proven antiviral therapies or vaccines, infection control and prevention strategies are essential to limiting transmission of EV-D68. Proper handwashing, covering the mouth and nose while coughing and sneezing, and cleaning of contaminated surfaces are general measures to prevent spread. In the hospital setting, isolation with standard, contact, and droplet precautions (gown, glove, and mask) should be enacted in the presence of respiratory symptoms. During periods of high EV-D68 activity, visitation restriction of symptomatic visitors and school-aged siblings has been utilized to limit hospital-acquired infections [Schuster and Newland, 2015; Messacar et al., in press]. Given the propensity for children with asthma to develop severe disease, asthma action plans should be in place and enacted early during a community-wide EV-D68 outbreak.

FUTURE DIRECTIONS

Prospective population-based surveillance is key to determining the epidemiologic patterns of EV-D68 and to guide hospital preparedness for surges in patient volumes and resource utilization. New EV-D68 specific RT-PCR tests will make real-time, widespread surveillance more feasible. As part of this surveillance, sequencing of circulating isolates will assist in tracking the molecular epidemiology of EV-D68. While bronchodilators and steroids certainly play an important role in the treatment of asthmatic children with exacerbation due to EV-D68 infection, the efficacy of these treatments for previously healthy children with EV-D68 lower respiratory tract infection merits systematic evaluation. The association between EV-D68 and acute flaccid myelitis should be further investigated through prospective syndromic surveillance, in vitro and animal models, host immune response and genetic studies. If EV-D68 continues to circulate and cause widespread severe respiratory disease and/or neurologic disease, development of effective antiviral therapies and vaccines will become scientific priorities.