The hidden threat of unidentified agents of disease in human and veterinary biologicals.

Theiler's disease, colloquially referred to as “serum hepatitis”, refers to a mysterious hepatitis of horses that occurs about 1–3 months after they receive a biological of equine origin [1]. First described in 1919, the cause of Theiler's disease has remained unknown despite the well-established epidemiological link between administration of the biological and equine hepatitis. Outbreaks of Theiler's disease, have been reported in North America and Europe following administration of immune serum for prevention of anthrax or encephalitis, or antitoxin preparations for prevention of tetanus or botulism. The consequences of this disease can be devastating, with 50–90% mortality in horses who become symptomatic [2], and the existence of this disease, although rare, has raised understandable concerns regarding the safety and acceptable risk:benefit ratio of equine blood-derived biological products.

In an article published in the March 2013 issue of the Proceedings of the National Academy of Sciences, Chandriani et al. report the identification of a novel virus that is the likely cause of Theiler's disease [3]. Named “Theiler's Disease-Associated Virus” (TDAV), the virus was identified after the research team responded to an outbreak of Theiler's disease at a farm in upstate New York. Eight horses at this farm were stricken with the disease after receiving the same antitoxin preparation for prevention and treatment of botulism. To identify TDAV, the researchers employed an elegant technique that uses unbiased next-generation sequencing, or “deep” sequencing, to identify novel and hitherto unknown pathogens. Use of deep sequencing for pathogen discovery is a “needle-in-a-haystack” approach, involving analysis and classification of millions of sequence reads of nucleic acids present in clinical specimens, and producing sequences or contigs (contiguous assemblies of sequences) corresponding to candidate pathogens [4]. In principle, this approach is able to simultaneously detect nearly all potential infectious agents, since all pathogens, with the exception of prions, are DNA/RNA-based. Previously, a similar strategy was used to identify a porcine circovirus as a contaminant in a human rotavirus vaccine preparation [5], and recently in 2012, to recover the whole-genome sequence of a novel virus, the Bas-Congo rhabdovirus, associated with a cluster of hemorrhagic fever cases in central Africa [6].

TDAV was found in all eight of the diseased horses, as well as in the antitoxin preparation. This virus is in the family Flaviviridae, which also includes human hepatitis C virus (HCV). However, the genome sequence of TDAV reveals that it is slightly more similar to “GB viruses”, which are in the recently proposed Pegivirus genus (with 35.3% amino acid identity to GBV-D, the closest viral match to TDAV in GenBank) [7], instead of the Hepacivirus genus that includes HCV. Members of the Hepacivirus genus are considered to be pathogenic, with HCV and GBV-B causing hepatitis in humans and monkeys, respectively [7]. In addition, canine hepacivirus is associated with pulmonary infections in dogs [8], although, strangely enough, also infects yet is not known to cause disease in horses [9], [10]. In contrast, none of the members of the Pegivirus genus to date (GBV-A in monkeys, GBV-C in humans, and GBV-D in bats) are linked to hepatitis or any other disease. Thus, the strong causal association between TDAV and hepatitis in horses reported in this study provides evidence of pathogenicity by at least one member of the Pegivirus genus. Clearly, more research on the diversity, host range, and pathogenicity of these viruses is needed.

The researchers correctly report that Koch's postulates have not been strictly fulfilled for TDAV. This is largely because of the lack of a suitable culture system for TDAV to enable purification of the virus in vitro. Although experimental inoculation of 4 horses with the subject antitoxin failed to demonstrate clinical signs of acute hepatitis, one horse did develop elevated liver enzyme levels in the blood and elevated TDAV titers, comparable to the levels seen in natural infections. Furthermore, the totality of the evidence in support of TDAV as the cause of Theiler's disease is convincing: (1) TDAV was the only detectable virus common to the antitoxin preparation and index cases, (2) TDAV was detected in all animals with hepatitis, and (3) there was a direct link between administration of toxin and development of Theiler's disease. It is still unclear, however, whether TDAV is the only cause of Theiler's disease of horses, or whether other as-yet unidentified agents might be responsible for some cases of the disease.

Whether or not TDAV poses a threat to humans is also unclear. The origin of TDAV remains unknown, and future research should seek to track down the host reservoir and modes of transmission for the virus in the wild. As several antitoxins certified for use in human patients are derived from horse serum [11], it also will be important to ascertain whether the virus is infectious to humans. The discovery of TDAV as the likely etiology of Theiler's disease makes it clear that routine purification procedures used during preparation of equine antitoxins, such as filtration and affinity chromatography, may not be sufficient to exclude infectious viral particles, thus potentially calling into question the safety of equine-derived biologics for human use. Efforts should also be made to characterize the full spectrum of disease caused by TDAV, if any, in horses and other mammals, including humans.

There has been tremendous interest in the use of broad-spectrum surveillance technologies such as deep sequencing to screen biologicals for adventitious agents. The discovery of TDAV is a direct validation of this approach, as the unbiased deep sequencing strategy was critical in identifying a highly divergent viral pathogen and neatly solving a nearly century-old mystery. PCR-based and serological tests derived from knowledge of the TDAV genome can be used to screen for the virus in equine blood products in the future, thereby likely preventing or minimizing the incidence of Theiler's disease and increasing our confidence in the safety of these biologicals. Importantly, the researchers found no evidence of horse-to-horse transmission of TDAV, and all documented cases were simply the result of a single exposure to the same antitoxin. Thus, although the current focus in diagnostic microbiology may be on rapidly characterizing the zoonotic causes of highly publicized outbreaks of communicable infectious diseases in animals and humans such as avian influenza and coronavirus EMC [12], [13], [14], we must not forget the ongoing hidden threat posed by both known and unidentified agents that may be circulating in biologics.