Asymptomatic or mild febrile cases of Madariaga – the base of the iceberg?

Abstract

Solicited opinion piece on accepted manuscript “Madariaga virus: Identification of a lineage III strain in a Venezuelan child with acute undifferentiated febrile illness, in the setting of a possible equine epizootic”

Madariaga virus, an alphavirus in the eastern equine encephalitis antigenic complex, has only rarely been shown to cause human disease even with the known capacity to cause outbreaks in equine populations in Latin America. When human cases have been detected, until recently, most have been associated with severe neurological disease. However, a follow up study to an outbreak of MADV in Panama in 2010 revealed both mild cases as well as evidence of asymptomatic infections [1]. Additionally, seroprevalence studies in the Amazonia region of Peru showed MADV antibodies in up to 3% of the residents [2] suggesting that infection with this virus may not be uncommon even in the absence of reported disease.

Blohm et al., in the work entitled “Madariaga virus: Identification of a lineage III strain in a Venezuelan child with acute undifferentiated febrile illness, in the setting of a possible equine epizootic” describe a case of Madariaga virus (MADV) causing only mild febrile illness in a child in Venezuela. With previous data (albeit limited) describing mild or asymptomatic infections with MADV, this case report provides one additional piece of evidence that these mild infections may indeed be much more common than neurological cases. This would not be unexpected for neurological arboviruses. West Nile virus (WNV) neurological disease has been clearly shown to result in only a small fraction of overall WNV infections [3]; this has resulted in the “iceberg” pattern of disease where the tip is the number of neuroinvasive disease cases while the mild and asymptomatic cases are represented by the underwater portion of the “iceberg”. A similar pattern may emerge for MADV as it becomes more epidemiologically characterized. What is perhaps more significant and potentially useful for clinicians and public health officials was how challenging it was to obtain this confirmation and characterization of an etiology of MADV.

The authors went through an extensive process of random primed PCR amplification, subcloning of the prominent amplicon, and Sanger sequencing using a genome walking methodology to confirm the nature of the infecting agent. While these efforts were certainly far beyond the capacity of a typical public health laboratory to identify an agent, they were useful for full genetic characterization of the virus. For simple diagnosis, sequencing of the “weak” amplicon generated from the universal alphavirus primer set could have provided a more straightforward diagnosis of the agent (which may have been performed but was not described in the manuscript). Interestingly, the amplicon generated from the universal alphavirus primer set was not an expected size which clearly shows the limits of using testing methods that are not fully vetted for all agents under consideration. One other interesting element to the testing design in this study was that the authors astutely tested a urine sample for viruses; the finding in this work was that the universal primers generated an alphavirus amplicon from this urine sample. Until the ZIKAV outbreaks of 2016–2017, urine was not typically a diagnostic sample type for most arboviruses. This sample type was explored extensively for use in detecting ZIKAV as it has been suggested that it may be more useful due to the simplicity of collection and the possibility that virus could be detectedfor a longer time in urine than in serum [4]. The finding by Blohm and team that urine was positive for MADV RNA suggests that it would be worthwhile to further evaluate this sample type for detection of a range of arboviruses, particularly those such as MADV or eastern equine encephalitis virus (EEEV) that have a long prodromal period and are therefore unlikely to be detected in serum samples.

In this short report, the authors also point out the need for laboratory diagnosis of mild febrile illnesses due to the overlapping clinical symptoms associated with a range of viruses. This is an important point particularly for regions where there are overlapping arboviruses that present with mild febrile symptoms. A clear example of this need for laboratory confirmation occurred during an outbreak of encephalitis in Panama in 2010 when both VEEV and MADV were found to have caused encephalitis and mild cases [5]. In this situation, there were only a small number of cases facilitating the laboratory distinction of these viruses. However, during large outbreaks, detecting multiple causative circulating agents is more problematic but perhaps even more important for patient management. An example of this was an outbreak of dengue-like illness reported in Jambi, Sumatra, in 2014–2015. While dengue was indeed a primary cause of the outbreak, it was only by testing dengue negative samples that cases of chikungunya (CHIK) and Zika were detected [6]. During large outbreaks, once a cause of the outbreak has been confirmed, laboratory testing of the majority of subsequent samples may not occur due to limited resources. However, this report is a reminder that laboratory testing throughout the course of large outbreaks is still essential.

Finally, while more cases of MADV, either mild or neurologic in nature, are likely to continue to be detected as awareness of this agent as a potential human pathogen increases, it is important to use caution in suggesting urban transmission, even speculatively, without evidence. This is tempting to do given the recent findings of significant levels of transmission of ZIKAV and CHIKV in human-mosquito-human cycles. However, with limited public health, research, and mosquito surveillance resources, efforts must be focused on the most likely transmission patterns. Whether MADV is truly an emerging pathogen or one that is simply now being recognized as a cause of mild and perhaps widespread febrile disease, Blohm et al. are correct in asserting that a better understanding of the epidemiology of this pathogen is needed.