Encephalitis: A Global Problem Deserving of a Global Approach

(See the Major Article by Britton et al on pages 2517–26.)

It is estimated that more that 20 000 cases of encephalitis occur in the United States annually [1]. However, this number pales in comparison to the number of cases in the developing world where diseases such as rabies, Japanese encephalitis virus, and tick-borne encephalitis virus account for a many-fold greater number of cases per year. Historically, encephalitis has been an enigmatic syndrome, with a causative pathogen or disease identified in only a minority of cases. The development of molecular testing in the 1990s revolutionized the diagnosis of herpes simplex encephalitis, which until that time required brain biopsy for confirmation of infection [2]. This advance led to great optimism that cerebrospinal fluid (CSF) polymerase chain reaction (PCR) would provide not only rapid but also exquisitely sensitive detection of other causes of encephalitis. However, for nonherpes group viruses, the sensitivity and specificity of spinal fluid PCR was often disappointingly low, and the proportion of unknown cases remained substantial. Encephalitis syndromic surveillance studies funded through the Centers for Disease Control and Prevention Emerging Infections Program (EIP) in 3 states (New York state, the California Encephalitis Project, and the Tennessee Unexplained Encephalitis Study) in the early 2000s found that even if state-of-the-art PCR testing on CSF and extra-central nervous system specimens, in combination with traditional methods such as serology, is used, a pathogen was identified in <50% of cases [3, 4]. The impressive study by Britton et al, included in this edition of Clinical Infectious Diseases, shows just how far we have come in the last decade in terms of improving the diagnosis of encephalitis etiologies and just how far we must go in terms of improving outcomes of this often-devastating disease.

This prospective multicenter study utilized the Pediatric Active Enhanced Disease Surveillance network at 5 major children’s hospitals in Australia to identify almost 300 children who met a standardized case definition of encephalitis between May 2013 and December 2016. Infectious causes accounted for almost two-thirds of cases, with enteroviruses, parechovirus, and bacterial-meningoencephalitis the most frequently identified causes, followed by influenza, herpes simplex virus (HSV), and Mycoplasma pneumoniae. One-quarter of cases were attributed to immune-mediated causes, with acute disseminated encephalomyelitis and anti-N-methyl-d-aspartate receptor (anti-NMDAR) cases predominating. Notably, <20% of cases represented undiagnosed causes of encephalitis, a marked improvement compared to prior studies from the previous decade.

What is different about the Australian study compared to previous attempts to identify causes of encephalitis? For one, this study illustrates the major contribution of noninfectious etiologies. Anti-NMDAR encephalitis, first identified in 2007, has emerged as one of the most important (and treatable) causes of encephalitis and is often indistinguishable from infectious etiologies at presentation. Second, as investigators in the field for over 2 decades, we continue to question the true role of M. pneumoniae as a cause of encephalitis. The authors designate a positive Mycoplasma immunoglobulin-M as evidence of causality, but many studies have shown this to be an unreliable marker of acute infection. Third, the population was restricted to pediatric cases. Pathogens common in childhood encephalitis in Australia, like those identified in the United States, differ from organisms most commonly identified in adults, such as HSV-1, varicella zoster virus, and West Nile virus. Yet, Britton et al highlights that this is more nuanced, as even among children there is marked heterogeneity between neonates (<6 months) and older infants and children. The EIP studies excluded neonatal cases due to the challenge in differentiating sepsis and meningitis from encephalitis in this age group. This is underscored by the finding that bacterial meningoencephalitis with Streptococcus pneumoniae or Streptococcus agalactiae were only diagnosed in neonatal cases in the Australian study. Similarly, HSV-2 is almost never a cause of encephalitis after 6 months of age. Inclusion of children throughout the age continuum illustrates the unique pathogens seen in neonates compared to older children and adults, an important consideration for pediatric providers who manage patients with encephalitis. Finally, the comprehensive nature of the study is yet another reason for the small percent of “unknowns.”

The interpretation of laboratory results, both positive and negative, can be challenging. The authors used previously published criteria (the so-called Granerod criteria [5]) to standardize these classifications. At the time of publication, these criteria were a significant step toward developing a cohesive approach to consistent interpretation of laboratory testing. However, over time it has become apparent that the Granerod criteria are overly reliant on detection of organisms in the CSF, leading to the potential for misclassification bias. For instance, amplification of virus in CSF is not always consistent with acute infection. Detection of cytomegalovirus or Epstein-Barr virus in CSF often reflects reactivation of latent virus and not necessarily the cause of the encephalitis. Similarly, herpes human virus-6 detection may be due to the presence of chromosomally integrated virus unrelated to the etiology of encephalitis. Understanding the potential for false positive and false negative PCR results is especially relevant with the increasing use of commercial multiplex meningoencephalitis panels. Overreliance on molecular testing ignores the fact that for many pathogens, including most arboviruses, serology is significantly more sensitive than PCR.

We agree with Britton et al. that the Granerod criteria give insufficient weight to identification of pathogens from extra-central nervous system sites. For instance, enterovirus-71, a well-established cause of encephalitis, is inconsistently detected in spinal fluid by PCR and more likely to be identified by PCR of stool specimens [6]. Influenza, one of the leading pathogens found in the Australian study, is rarely identified in CSF, and a failure to test respiratory samples for this virus would likely lead to a missed diagnosis. Even with rabies, a quintessential cause of encephalitis, PCR of saliva is more sensitive than of CSF.

Perhaps the most clinically relevant finding from the Britton et al study is confirmation that most cases of childhood encephalitis are caused by well-substantiated and relatively common pathogens rather than rare outliers or novel organisms. Why is this significant? It suggests that the routine application of a tiered diagnostic algorithm has the potential to identify a cause in the majority of encephalitis cases. The Infectious Disease Society of America encephalitis guidelines, published in 2008, and now archived, were a landmark attempt to standardize diagnosis and management of this challenging syndrome. However, the guidelines were published before the recognition that up to one-quarter of cases of encephalitis are due to immune-mediated conditions such as anti-NMDAR encephalitis. Updating these guidelines to reflect our current understanding of encephalitis and the appropriate use of diagnostic testing needs to be a priority. In the interim, there exists an underappreciated and little-known resource developed by an ad hoc consortium of international stakeholders that includes diagnostic algorithms for both pediatric and adult encephalitis [7] These algorithms outline 2-tiered testing for both pediatric and adult patients with encephalitis, with both routine and conditional testing based on regional epidemiology (eg, Japanese encephalitis virus), specific laboratory and neuroimaging features, and clinical characteristics (see the Supplementary Tables). Almost all of the identified causes of encephalitis in the study by Britton et al would have been captured by adhering to this optimized testing algorithm, with the potential to maximize diagnoses and minimize expenditures through rational use of commercially available tests.

What about the almost 20% of cases in the study in which no cause was identified? It is conceivable that these represent infections caused by less common or even novel organisms. For this subset, next-generation sequencing (NGS), a new technique to allow unbiased pan-species molecular diagnostics, may play an important role. As illustrated in a recent publication of more than 200 adult and pediatric patients with meningoencephalitis, NGS of CSF identified an infectious cause in more than one-quarter of patients, many of which would not have been identified through traditional testing algorithms [8]. It is clear that NGS will, and should, have a role in testing of enigmatic cases of encephalitis. However, this new technology cannot replace a thoughtful and meticulous exposure history, careful review of clinical and laboratory features, and thorough laboratory evaluation of not just CSF but other potential sites of infection including nasopharynx, throat, serum, and stool.

In contrast to the significant progress made in identifying etiologic causes of encephalitis, improvement in outcomes has been minimal. In the Australian study, 5% of patients died during acute hospitalization and more than one-quarter of the patients had moderate to severe neurologic impairment at discharge. This highlights the continued morbidity of this syndrome and the pressing need for posthospital rehabilitation specific for pediatric patients with encephalitis. The lack of standardized encephalitis-specific rehabilitation protocols was part of the genesis for Encephalitis411 (see https://encephalitis411.org), a nonprofit organization founded by encephalitis survivors to advocate for patients in the United States.

The study on childhood encephalitis in this month’s issue of Clinical Infectious Diseases provides a framework to expand encephalitis research. Research priorities need to focus not just on the “what” but also on the “why” and the “how.” For instance, why is it that almost all children experience several enteroviral infections during their lifetime yet only an exceedingly small number develop encephalitis? Similar questions could be asked for influenza and herpesviruses. A research priority needs to be how to best manage patients with encephalitis. New agents for the treatment of encephalitis are urgently needed, yet antimicrobial therapies for viral encephalitis have not advanced since 1986 when acyclovir was shown to improve outcomes for herpes simplex encephalitis [9]. This likely reflects the lack of incentive for research and development by pharmaceutical companies for this uncommon syndrome. Yet, this does not explain the absence of clinical trials to explore the fundamental question as to the role of corticosteroids in patients with viral and unknown encephalitis. These and other key questions can only be answered if encephalitis research is considered a priority for federal funding agencies with the ability to support international multicenter collaborative research for this relatively infrequent but not rare syndrome. Britton et al confirm that the pathogens and challenges of encephalitis are not unique to Australia but common to those found throughout the world. The time has come to move beyond causality and toward the global goal of optimizing management for this devastating syndrome.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

References

Note

Potential conflicts of interest. The authors: No reported conflicts of interest. Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.