Emergence and re-emergence of viral diseases of the central nervous system

2.1. Expansion of viruses to new geographic regions

Important examples of the expansion of the range of encephalitic viruses have been seen with the arthropod-borne viruses. These viruses are maintained in natural cycles involving both vertebrate and invertebrate hosts and can increase their range and cause human disease if introduced into a new location that has appropriate hosts available to maintain the natural cycle. Modern transportation has introduced vectors that efficiently transmit arboviruses into new areas (e.g. the Asian tiger mosquito Aedes albopictus into North America and Europe) and pre-existing or new populations of competent vectors set the stage for successful establishment of viruses in new regions. The 1999 introduction of West Nile virus (WNV) into North America from the Middle East (Gubler, 2007), the spread of Japanese encephalitis virus (JEV) into India, Nepal and northern Australia (Solomon et al., 2003, van den Hurk et al., 2009) and the explosive 2005–2007 outbreaks of Chikungunya virus infection in the Indian Ocean islands and India (Powers and Logue, 2007, Renault et al., 2007) are examples of this mechanism.

Once introduced into New York, probably through air transport, WNV infected a variety of susceptible mosquito and bird species to establish an endemic cycle and spread rapidly across America. WNV has resulted in tens of thousands of cases of CNS disease in North America and is now endemic across the continent. WNV has the widest distribution of all flaviviruses as its geographic range now spans North and South America, Europe, the Middle East, Africa, Western Asia and Australia (Gubler, 2007). The lineage 1 strain of WNV introduced into the United States is a more neurovirulent strain that causes encephalitis than the lineage 2 African strains that cause febrile illness (Beasley et al., 2002).

Japanese encephalitis is the leading cause of viral encephalitis in Asia, with an annual incidence of 30,000–50,000 cases of neurologic disease. JEV is distributed in temperate and tropical areas of eastern and southern Asia and maintained in a zoonotic cycle between Culex mosquitoes and wading birds or pigs (van den Hurk et al., 2009). Recent geographic extension to Pakistan, Papua New Guinea and the Torres Strait of northern Australia (Fig. 1 ) may have involved viremic migratory birds or transport of infected mosquitoes by wind, land vehicles or planes. Emergence is also associated with the increases in endemic areas of population, pig production and irrigated rice farming (Erlanger et al., 2009). The case fatality rate is typically between 20% and 40% with children and the elderly at greatest risk. Effective vaccines are increasingly available and appropriate use should decrease the disease burden (Solomon, 2006).

Open in a separate window

Fig. 1

Current geographic distribution of Japanese encephalitis virus. Source: World Health Organization.

An additional recent dramatic example of arbovirus expansion has been the appearance of explosive outbreaks of rash and arthritis due to CHIKV infection on islands in the Indian Ocean with subsequent spread to India and Europe (Powers and Logue, 2007). The introduced CHIKV strain is a variant of the Central/East African genotype that is efficiently transmitted by Aedes albopictus (Chevillon et al., 2008). Although occasionally recognized previously, large numbers of neurologic complications have been reported only recently, although the actual frequency of neurologic disease is still unclear (Arpino et al., 2009). Neurologic complications are age-dependent with the greatest susceptibility in the very young and the elderly. Encephalopathy is the main manifestation in newborns, while meningitis and encephalitis occur in older children and adults (Arpino et al., 2009, Chandak et al., 2009).

2.2. Spread of viruses into humans from animal reservoirs

Exposure of humans to animals provides the opportunity for animal viruses to spread into human populations. This may result only in infection of the exposed human (e.g. rabies after a bite from a rabid animal) or in infection that has or gains the ability to spread within the human population (e.g. HIV, SARS coronavirus) (Wolfe et al., 2007). For instance, lentiviruses of primates have been transmitted on multiple occasions to humans, but only a few strains acquired the ability for human-to-human transmission to become the HIV strains that are currently prevalent in human populations around the world (Sharp et al., 2001). If, such as rabies, the virus remains a zoonosis with intermittent transmission to humans, then control of infection in animal populations is necessary to prevent human disease (Zinsstag et al., 2007). For rabies, regular vaccination of domestic dogs is the most important intervention, but much of the world remains at risk (Fig. 2 ) (Hampson et al., 2009, Rupprecht et al., 2008).

Open in a separate window

Fig. 2

Map of the geographic areas of the world where rabies is considered a high risk. Red areas are high risk from contact with rabid dogs and orange areas are high risk from contact with infected bats and other wildlife. Gray and green areas are low or no risk. Source: World Health Organization.

Often, there may be an intermediate host for the virus on the way to becoming a human disease. Bats are increasingly recognized as important hosts for a number of zoonoses that cause CNS infection (e.g. lyssaviruses, henipaviruses, coronaviruses and filoviruses) (Halpin et al., 2007). Disruption of the environment with changing agricultural practices has increased the likelihood that these viruses will be transmitted to humans. Recent outbreaks of neurologic disease in humans due to the pteropid bat viruses Hendra and Nipah illustrate this process.

Hendra is a paramyxovirus that was first identified when horses, and a few humans in close association with the horses in Australia, became ill with respiratory and neurologic disease that was often fatal. The related Nipah virus was recognized in Malaysia because of an outbreak of respiratory disease in swine and encephalitis in humans exposed to infected swine. For each, horses and pigs became infected through exposure to infected bats, and humans became infected through exposure to the infected horses and pigs. The viruses do not cause disease in the bat reservoir hosts. Recently, in Bangladesh, humans have developed fatal neurologic disease directly from exposure to infected bats with evidence of subsequent human-to-human spread (Epstein et al., 2006, Gurley et al., 2007).

2.3. Genetic mechanisms for virus adaptation

Adaptation to a new host requires the virus to change. Viruses, particularly RNA viruses, are adept at change and have several mechanisms for altering their genomes. The basic mechanisms are: shuffling gene segments by reassortment (applies to viruses with segmented genomes), mutation (applies to all viruses) and recombination (applies to most viruses).

Influenza virus has eight gene segments and is the most important virus that evolves by reassortment. When two different strains of influenza (e.g. an avian strain and a human strain) infect the same host cell, the new progeny viruses can reassort the two sets of eight genes with a potential for 256 different combinations of segments in individual virus particles. This process results in viruses with new properties. Most will be less “fit” for human infection than the original viruses, but others may be able to replicate well in humans that do not have immunity to the new virus. This is the process that resulted in the antigenic shift that occurred prior to the influenza pandemics of 1957, 1968 and, most recently, 2009 (Garten et al., 2009). Although most of the disease is confined to the respiratory tract, neurologic complications of influenza occur, particularly in children (CDC, 2009) and for H5N1 avian influenza virus, invasion of the CNS has been documented (De et al., 2005). After new strains are established in the human population, they evolve slowly by mutation (antigenic drift).

Mutation is an important mechanism of change for all RNA viruses because viral RNA-dependent RNA polymerases are error-prone and have no built in mechanism for correcting mistakes (Domingo et al., 2006). Polymerase errors, plus the lack of proofreading, results in one mistake approximately every 10,000 nucleotides, an average of one error in every genome. Furthermore, progeny viruses are produced in large numbers by infected cells. Most of these mutations will be deleterious to virus growth, but will occasionally be advantageous. It has recently been shown in animal models of polio and foot and mouth disease that this ability to vary is important for virulence (Vignuzzi et al., 2006, Sanz-Ramos et al., 2008).

An additional mechanism for virus evolution is recombination. In this case, two similar viruses that infect the same cell can recombine different portions of their genomes to produce new viruses. This is commonly observed for enteroviruses that can infect and be shed from the gastrointestinal tract for long periods of time. This is of current importance for poliovirus where the live virus vaccine includes three strains of polio that not only mutate, but also recombine with themselves and other enteroviruses to produce new vaccine-derived strains that may have increased virulence (Cherkasova et al., 2003).

3.1. Poliovirus

Poliovirus is an enterovirus that is spread by the oral route through contaminated food and water and causes asymptomatic infection in the majority of people. It is approximately two times more infectious than smallpox virus, but only one in 100–200 will develop virus spread to motor neurons in the CNS followed by acute flaccid paralysis due to motor neuron death or dysfunction. Thus, unlike smallpox, poliovirus can circulate widely in the population in the absence of symptoms. Furthermore, virus can continue to be shed from the gastrointestinal tract for weeks after infection. Therefore, laboratory tests are needed to identify poliovirus infection and surveillance relies on identifying cases of acute flaccid paralysis associated with poliovirus in the stool.

There are three poliovirus serotypes and immunity must be induced to all three for full protection from paralytic disease. This can be accomplished either through the use of an inactivated vaccine or a live attenuated virus vaccine. The inactivated vaccine induces serum antibody that prevents virus spread to the CNS, but does not necessarily prevent infection of the gut. The live attenuated vaccine is essential for eradication efforts because it is delivered orally, infects and induces immunity in the intestinal tract and protects against subsequent wild type virus infection and thus interrupts silent spread of wild type virus from person-to-person.

The original poliovirus eradication goal was 2000. At that time, two of six WHO regions, the Americas and Western Pacific, were certified free of wild type poliovirus and only a few countries in the other regions had continuing wild type virus circulation. Confirmed cases of poliovirus-induced paralysis worldwide had decreased from 350,000 to 2971.

However, in the same year a new problem was recognized. Viruses isolated from a cluster of children with paralysis on the Caribbean island of Hispaniola were identified as mutated vaccine viruses that had circulated in an under vaccinated population for about 2 years and had reacquired neurovirulence. Subsequently, several other polio outbreaks were identified as due to vaccine-derived polioviruses that can circulate widely (Wringe et al., 2008). The strains of live attenuated vaccine viruses have only a small number of genetic changes from the wild type viruses from which they were derived (Kew et al., 2005). Revertants that re-gain a neurovirulent phenotype are selected rapidly in the intestine and shed in stool (Minor et al., 1986, Chumakov et al., 2007). Despite this, disease in vaccinees is rare because infants are protected from spread to the CNS by maternal antibody. However, spread to nonprotected individuals can occur and result in paralytic disease. Recombination between the vaccine strains, as well as other related enteroviruses, such as Coxsackie A viruses, is also frequent, although importance of this for acquisition of virulence is less clear (Chumakov et al., 2007, Cherkasova et al., 2003, Minor et al., 1986).

In 2002, only India, Nigeria, Pakistan and Afghanistan had endemic circulation of wild type poliovirus, but since that time there has been spread from these countries, particularly Nigeria, to previously polio-free regions and the eradication effort has stalled (Fig. 3 ) (Pallansch and Sandhu, 2006, Chumakov et al., 2007). In addition, it appears that in some populations (e.g. the northern regions of India), protective immunity is not always achieved even after multiple doses of vaccine (Grassly et al., 2009). Because of prolonged shedding of vaccine virus, particularly by immunodeficient individuals, the current endgame strategy is unclear (Kew et al., 2005, Minor, 2009), but will probably require the development of a new type of polio vaccine (Chumakov and Ehrenfeld, 2008).

Open in a separate window

Fig. 3

September 2009 map of countries reporting indigenous wild poliovirus (red) or imported wild poliovirus (orange). Source: World Health Organization.

Work from the author's laboratory was supported by the National Institutes of Health (NS18596, NS038932 and AI023047) and the Bill and Melinda Gates Foundation (RG3522).