35 Rabies and rabies-related lyssaviruses

Rabies virus is epidemic in most parts of the world. It can replicate in all warm-blooded animals in which it causes a devastating neurological illness, which almost invariably results in death. Rabies is a disease of animals and human infection is a ‘spillover’ event occurring most commonly following a bite from an infected dog. Infection is seen in different patterns; rabies with little or no wildlife involvement, sometimes known as urban or street rabies, or in the wildlife population with spillover into domesticated animals (sylvatic).

Eleven distinct species of lyssavirus are now recognized: species 1 is the most common strain found predominately in terrestrial animals. Species 2–11 are detected in bat species with the exception of Mokola virus (species 3). Despite the availability of effective vaccines substantial mortality still occurs, mostly in the tropics. The majority of rabies free countries are islands which are able to remain rabies free by import controls. Effective animal vaccines are available and dog rabies is well controlled in most parts of the developed world with dog vaccination. However, rabies remains an intractable problem in many countries in Asia and Africa due to a lack of infrastructure, cost of vaccines and difficulty in controlling dog populations. In recent years progress in controlling wildlife rabies has been achieved in western Europe using vaccine in bait, which offers promise for other regions with complex epidemiology.

Despite the availability of effective vaccines, rabies virus remains endemic across much of the globe with two thirds of the world’s population living in a rabies endemic region. The ability of the virus to infect and replicate in a wide variety of hosts mean that as well as circulating in specific species, cross species transmission (CST) can also occur. From a zoonotic perspective, human infection constitutes a CST event from the infected animal reservoir. Whilst transmission of virus from infected animals into the human population is historically through the bite of an infected dog, the elimination of rabies from domestic canine populations across the developed world has highlighted other species as important transmission vectors. Several terrestrial wildlife species harbour the virus, often within distinct populations. Occasionally, transmission between different species drives the generation of different rabies virus biotypes. Modern genetic analyses have been used to highlight mutual adaptation of virus variants and host populations. Epidemiologically, within an endemic area, disease is usually sustained within a single host species, although interactions between different species can cause a CST event to occur. This event occurs most often when clinical disease is identified.

It has long been established that whilst dogs are considered as the principal terrestrial reservoir for rabies virus, bats (Order Chiroptera) constitute an increasingly important reservoir of rabies and rabies-related lyssaviruses (Fooks 2004; Vos et al. 2007; Banyard et al. 2009). Indeed, ten of the defined lyssavirus species have been isolated from bats, the exception being Mokola virus (species 3). Few isolates of this virus are available for study and its epidemiology is poorly understood. Rabies virus was identified in vampire bats with records associating transmission of rabies from haematophagous bats to humans dating back as far as the sixteenth century (Blancou 1994). In the first half of the twentieth century, insectivorous and frugivorous bats were also proven to be harbouring rabies virus variants and over the past 90 years more data associating different bats species across the world with the transmission of lyssaviruses has been generated. Genetic characterization of bat derived lyssaviruses has shown that, although isolates can be grouped as the same species as terrestrial isolates, in reality the viruses remain quite distinct with several species of rabies virus variants being derived solely from bats. Certainly, isolates from different continents, whilst sharing many common characteristics at the molecular level, remain genetically diverse.

A number of reporting systems have been implemented to record individual disease outbreaks. For rabies and related lyssaviruses several organizations monitor cases of rabies and make data available to the general public, these include: RabNet (www.who.int/rabies/rabnet), the Rabies Bulletin in Europe (www.who-rabies-bulletin.org/), the Bulletin of Epidemiological Surveillance of Rabies in the Americas (www.paho.org/English/AD/DPC/VP/rabia.htm) and the Office International d’Epizooties (OIE) (www.oie.int/eng/en_index.htm) all act as reporting systems. These surveillance reports are essential in monitoring the status of countries for presence of the virus. However, rabies remains endemic in almost all continents although vaccination of dogs and wildlife, quarantine and surveillance have greatly reduced disease incidence. Currently Antarctica, as well as a number of island nations such as the United Kingdom, Japan, Barbados, Fiji, the Maldives, the Seychelles, New Zealand and Hawaii are free of classical rabies isolates. Such island nations are able, with relative ease, to remain rabies-free by the application of importation controls such as quarantine. In addition, much of Western Europe and parts of northern and southern continental Europe such as Greece, Portugal, Sweden, Norway as well as some countries within Latin America (e.g. Uruguay and Chile) are also free of terrestrial rabies. Bat lyssaviruses continue to pose a threat, albeit low, to the human population in many of these areas.

Despite the successes of island nations in controlling rabies and the elimination of canine rabies in parts of the developed world, the control of terrestrial rabies remains a challenge and the virus is still epizootic across much of Eastern Europe, Africa, Asia, and Latin America. The vast majority of human deaths from the disease occur within the developing world. Whilst figures estimated for annual human deaths within these areas are high (up to 50,000), it is widely believed that these figures are a gross underestimate of the actual number of human deaths (Fooks 2005; Fooks 2007; George Baer, personal communication). This affects our understanding of the global epidemiology of the disease and in many regions the incidence of disease is largely unknown. In some regions surveillance has identified where disease occurrence is either linked with endemic infection within the canine populations, often referred to as ‘street’ or ‘urban’ rabies, or with CST from wildlife reservoirs into domestic animals and man.

Rabies virus is a member of the Order Mononegavirales, Family Rhabdoviridae, genus lyssavirus and contains a non-segmented negative strand genome of approximately 12 kilobases (kb) in length. As well as rabies virus, the order Mononegavirales includes a range of other important diseases of man including members of the Filoviridae (Ebola virus) and Paramyxoviridae (measles virus). The Rhabdoviridae are classified into several serologically and genetically distinct genera with the capability to infect a diverse spectrum of hosts including vertebrates, invertebrates and plants. Within this diverse family two genera are classified as being able to infect mammalian species: the vesiculoviruses and lyssaviruses. The vesiculovirus genus includes the viruses causing Vesicular Stomatitis virus (VSV) and bovine ephemeral fever virus (BEFV), two antigenically related viruses of economic importance in cattle. The lyssavirus genus includes rabies virus, the rabies-related viruses, and several other recently classified viruses that share only a distant relationship to rabies virus.

The lyssavirus genus was originally subdivided into four distinct serotypes and further, through genetic typing, into seven major genotypes. More recently, classification of members of the lyssavirus genus has altered from being genotypes into distinct species with the inclusion of previously unclassified rabies-related viruses. Initial characterization of the different virus isolates into the four serotypes was made by cross-immunization experiments in animals and by antigenic typing of isolates using monoclonal antibodies. Serotype 1 contained the classical rabies viruses and Australian bat lyssavirus (ABLV); serotype 2 covered isolates of Lagos bat virus (LBV); serotype 3 included Mokola virus (MOKV); and serotype 4 included Duvenhage virus (DUVV) and both variants of the European bat lyssaviruses, -1 and -2 (EBLVs). Molecular typing then enabled these viruses to be classified into seven distinct genotypes on the basis of genetic sequence data. Genotype 1 included all isolates of classical rabies virus; genotype 2 contained LBV; genotype 3 included the MOKVs; genotype 4 included DUVVs; genotype 5 included isolates of EBLV-1; genotype 6 included EBLV-2 isolates and genotype 7 included ABLV (Tordo et al. 2004). With the exception of MOKV isolates all genotypes have been isolated from bats (Kuzmin et al. 2005). Further genetic characterization of individual isolates has then identified MOKV and LBV isolates as being distinct from the other genotypes, ultimately leading to the classification of genotypes into phylogroups where MOKV and LBV make up phylogroup 1 and the remainder of the genotypes fall into phylogroup 2 (Nadin-Davis et al. 2007). In addition, it is likely that MOKV isolates and LBV can be divided further using genetic characterization that reflect their geographical origins.

Isolation of several rabies-related viruses from bats across Eurasia has further altered the way that these viruses are classified. Molecular data have been used to classify several other rabies-related lyssaviruses that were originally been proposed as new members of the genus. These include; Aravan virus (ARAV), isolated in southern Kyrgystan in 1991 from a lesser mouse-eared bat (Arai et al. 2003); Khujand virus (KHUV) isolated from the brain material of a whiskered bat in 2003 in northern Tajikstan (Kuzmin et al. 2003); Irkut virus (IRKV) isolated in 2002 from a greater tube-nosed bat and West Caucasian Bat virus (WCBV) isolated in 2002 from a bent-winged bat in the Caucasus mountains (Kuzmin et al. 2005). Initially, these isolates were unclassified but establishment of the relationship between these new viruses and the currently defined members of the lyssavirus genus have been studied both at the molecular level as well as using more modern techniques such as antigenic cartography which seek to define the relationship between amino acid composition and antigenicity (Nel and Markotter 2007). These studies with the Eurasian lyssavirus-like isolates has further enabled the classification of the seven lyssavirus genotypes into eleven distinct species. The current classification of these viruses is detailed in Table 35.1.

Individual rabies virus particles were first visualized using electron microscopy in the early 1960s (Matsumoto et al. 1962; Davies et al. 1963). The virions are bullet-shaped with a ‘spiky’ appearance. The spikes or peplomers are 9 nanometer (nm) protrusions, spaced approximately 5nm apart within the lipid bilayer that consist of the viral glycoprotein (G). Virions have an average length of 180 (130–300) nm and diameter of 75 (60–110) nm. A number of researchers have reported differences in virion length and have attributed them to the presence of defective interfering particles (DI). These truncated forms of the virion are thought to be aberrations of virus replication whereby truncated nascent genomes lacking segments of the virus genome are generated and packaged into virions. DI particles can vary greatly in length and, as they are incapable of autonomous replication, can interfere with the replication of full-length virus by usurping proteins that they are unable to generate themselves, thus reducing the replicative ability of the full-length, non-defective virus. DI particles are often seen after multiple passages in tissue culture (Holland and Villarreal 1975; Grabau and Holland 1982; Finke and Conzlemann 1999).

The minimal replicative unit for these viruses is the ribonucleoprotein complex (RNP) which consists of the RNA together with the nucleocapsid protein (N), the phosphoprotein (P) and the large polymerase protein (L). This RNP is then surrounded by the lipid containing envelope, 7.5–10nm thick, that the virus acquires from the plasma membrane of the infected cell upon exit. The matrix protein (M) plays a role in virion structure either sitting within the coiled coil, as suggested for vesicular stomatitis virus (VSV), or lining the inner surface of the virus envelope being in contact with both the RNP and the cytoplasmic tails of the glycoprotein (G) (Barge et al. 1993; Nakahara et al. 1999).

The lyssavirus genome consists of a single strand of negative sense RNA that can vary between 10 and 12kb in length. The negative strand genome contains 5 or more distinct open reading frames that each encode one of the viral proteins. The 3’ end of the genome contains an untranslated region of 58 nucleotides that precedes the first gene, called the leader. This region shows a high degree of complementarity with the terminal 5’ end of the genome where a similar region exists often referred to as the trailer region. Studies with other negative strand RNA viruses such as the paramyxoviruses, the filoviruses and the bornaviruses have identified distinct domains within these regions that play significant roles within the viral transcription and replication strategies (Banyard et al. 2005). Often referred to now as the genome and antigenome promoters (leader and trailer regions, respectively), these short untranslated regions contain all the necessary information to drive transcription to produce messenger RNA for the production of viral proteins; replication, from the genome promoter, to form a positive strand replicative intermediate; and generation of nascent negative strand genome RNA from the antigenome promoter at the 3’ end of the positive sense replicative intermediate. The exact nucleotides and viral protein complexes that interact at these regions to form viral transcriptase and replicase complexes are largely unknown although cis-acting signals for a transcriptase complex have been suggested (Conzelmann and Schnell 1994; Whelan and Wertz 1999).

Viral transcription is known to proceed from the 3’ end of the genomic RNA to generate a short leader RNA followed by messenger RNAs encoding for each of the five viral proteins: the nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G), and polymerase (L) (Tordo and Kouknetzoff 1993). Each of the newly generated mRNAs are capped and polyadenylated by the multifunctional polymerase (L) protein. The transcriptase complex generates a monocistronic mRNA from all but one of the five coding regions. The exception is the phosphoprotein (P) whereby poor Kozak consensus sequence around the primary P gene methionine leads to leaky transcription initiation and the complex initiates transcription at any of the three or four (depending on the virus isolate) in-frame initiation codons downstream of the primary AUG. These amino-truncated products are of unknown function although whilst the full length P is found exclusively in the cytoplasm, truncated forms can be found in the nucleus (Takamatsu et al. 1998).

Each of the genes are separated by non-translated intergenic regions, which in turn are flanked by gene start and stop sequences recognized by the transcriptase. This sequence of untranslated gene signals at each of the gene boundaries leads to a transcriptional gradient of mRNAs being synthesized. This gradient is generated as the transcriptase complex may become detached from the template at each intergenic region and, as it can only reinitiate transcription at the 3’ promoter, the 3’ proximal genes are generated in greater quantities than each of the respective downstream genes. At some point after initiation of transcription the viral proteins N, P and L, acting as a transcriptase complex, switches to a replicase form whereby it is now able to ignore the gene start and stop signals and instead generate a full-length positive strand antigenome RNA. This antigenome RNA then serves as the template for the generation of nascent genome sense negative strand RNA, the nascent negative strand being generated as the replicase complex binds to the 3’ end of the antigenome and synthesizes across the length of the genome. It is now recognized for several nonsegmented negative strand viruses that the protein complexes that make up the transcriptase and replicase forms contain different components. It is possible that individual proteins form different structures through the interactions with different domains to play these different roles within the viral life cycle although the mechanisms behind each of the different RNP activities remains largely unknown (Gupta et al. 2003). Studies using different lyssavirus isolates have attempted, to define molecular determinants of pathogenicity and N, P, M, and G have all been implicated (Mita et al. 2008; Pulmanausahakul et al. 2008; Shimizu et al. 2007; Takayama-Ito et al. 2006; Wirblich et al. 2008).

The mechanisms by which viral transcriptase and replicase complexes act differently at the intergenic regions as well as the stop/start gene sequences that flank them remain unknown. Whilst for a number of the negative strand viruses these regions are conserved in length if not in sequence, for rabies virus, these intergenic regions differ in length at each gene boundary. Between N and P a dinucleotide region is present whilst at both the P-M and the M-G gene boundaries a pentanucleotide is present. However, between the G and L genes some members of the lyssavirus genus contain a very long intergenic region of 423 nucleotides (Marston et al. 2007). Whilst some have described this long 3’ non coding region as a pseudogene (Ψ) others have suggested that it is of no evolutionary significance and is unlikely to represent a remnant gene (Tordo et al. 1986; Ravkov et al. 1995). Currently, the role of this long untranslated region remains unknown although it has been postulated that it may play a role in regulating transcription of the polymerase gene (L) as the level of L gene transcription correlates with the extraordinary length of Ψ. More recent studies have indicated that this region may play a role in the neuroinvasiveness of rabies virus strains (Faber et al. 2004).

The principal mechanism of human infection with rabies viruses is through the bite of an infected animal that contains virus in its saliva. The lyssaviruses are unable to infect humans through the dermal barrier unless the skin is broken. However, infection may occur at exposed mucous membranes including the conjunctivae, the nasal lining, the oral cavity, the anus and external genital organs. Infection through inhalation of aerosolized virus in bat caves has been reported (Irons et al. 1957; Brass 1994) and experimental studies have confirmed the possibility of infection via this route for both rabies and other lyssaviruses (Constantine 1962; Winkler 1968; Winkler et al. 1973; Johnson et al. 2006a). These studies suggest that this method is a potentially important mechanism by which virus is sustained within a colony where vast numbers of animals live in close proximity. Human infection via this route, however, remains extremely rare (Johnson et al. 2006). Accidental human exposure to aerosolized virus within a laboratory has also been reported. In the incident, a veterinarian succumbed to disease whilst another laboratory worker recovered although the individual sustained substantial neurological impairment (Winkler et al. 1973; Tillotson et al. 1977).

Human-to-human transmission has also been reported albeit very rarely (Fekadu et al. 1996) and evidence of transplacental transmission has been suggested by a single case (Sipahioglu and Alapaut 1985). In 2004, recipients of kidneys, a liver and an arterial segment from a common donor developed fatal rabies and live virus was recovered from the recipients of the transplanted organs (Srinivasan et al. 2005; Burton et al. 2005). Several human-to-human cases of rabies transmission have also occurred through corneal transplantation (Houff et al. 1979; Gode and Bhide 1988; Sureau et al. 1981). Many cases of rabies reported annually have no origin cited as the source of infection and such cases are assumed to be as a result of transmission through minor scratches from infected animals or bites from small bat species that have gone unnoticed (Messenger et al. 2003). These cases underline the need for education of local populations and tourists visiting endemic areas (Fooks et al. 2003). Deliberate release of rabies virus remains an important and yet remote likelihood (Fooks et al. 2009).

The ability of rabies virus to enter cells is dependent on the ability of its glycoprotein, present on the virion surface as a trimer, to bind to cellular receptors (Gaudin et al. 1992). Entry of virus may occur either directly into the central nervous system (CNS) or after initial replication within muscle tissue. It has been suggested, though not proven, that rabies virus may replicate in non-neuronal tissue around the bite site. Such low level replication at the bite site in a tissue type that is not highly permissive for virus may explain long incubation periods that are occasionally observed following infection (Lafon 2005). Evidence from in vitro studies suggests that the virus enters motor nerves primarily through neuromuscular junctions and antigen has been detected within sensory spindles, proprioceptors and stretch receptors (Lewis et al. 2000; Murphy 1977).

Currently, three rabies virus receptors are proposed: the nicotinic acetylcholine receptor (nAchR), responsible for interneuronal communication within the CNS and the peripheral nerve network; neural cell adhesion molecule (NCAM), present at the nerve termini and deep within the neuromuscular junctions at the postsynaptic membranes; and to a lesser extent the neurotrophin receptor (p75NTR), which plays a role in cellular death, synaptic transmission and axonal elongation (Dechant and Barde 2002; Tuffereau et al. 2007). It is currently unknown to what extent each of these proposed receptors are utilized by rabies viruses and their use by other lyssaviruses also remains to be elucidated.

Once the virus has bound to a permissive receptor on the surface of a neuron or cell, entry is believed to occur either by direct fusion between the virion membrane and the plasma membrane or by receptor mediated endocytosis. In the latter case membrane fusion of the virus particle and the endosomal membrane is as a result of a lower pH environment within the endosomal compartment. Either way, the outcome of attachment and penetration into the cell is the release of the viral genetic material in the form of a ribonucleoprotein complex into the cell cytoplasm. Once released the molecular events that drive transcription and replication are initiated and the virus replicative cycle begins.

Transport of rabies virus from the axons of the peripheral nervous system to the CNS occurs by retrograde axoplasmic flow. Studies with human dorsal route ganglia neurons showed that the virus was able to move relatively quickly in a retrograde manner traveling at 50–100mm/day (Tsiang et al. 1991) and such fast axonal transport may be a result of the interaction of the virus phosphoprotein with actin- and microtubule-based motility networks such as dyenin LC8 (Jacob et al. 2000; Raux et al. 2000). However, more recent studies in vivo have suggested that mutation of the phosphoprotein LC8 binding domain has only minor effects on viral spread suggesting that other, as yet unidentified, interactions may be key to virus mobility (Mebatsion 2001; Rasalingham et al. 2005). Recent studies using the CVS strain of rabies virus have identified key stages of infection using the virus as a transneuronal tracer of neuronal connections. Intracellular transport of viruses from dendrites of infected neurons to the presynaptic terminals of connected neurons has been shown. The ubiquitous distribution of suitable receptor molecules within the CNS have been shown to drive the rapid movement of virus whilst peripheral infection is restricted to motor endplates and axons limiting initial propagation of virus at the periphery (Ugolini 2008).

The neurotropic nature of rabies virus and other lyssaviruses generally dictates the rate of progression of disease with virus entry at a highly innervated site often leading to swift progression of virus into the CNS and the onset of clinical disease. However, as the virus is often transmitted through the bite of an infected animal, the level of innervation at the bite site may vary greatly and as a result, there is a noticeable variation in disease progression. This difference is often reflected in the length of the incubation period observed between infection and onset of clinical disease. The nature of the infecting virus particles, the dose transmitted and presence or absence of permissive receptors on peripheral nerve networks may also limit virus entry and motility (Ugolini 2008).

Historically, rabies virus infection was considered as having an incubation period of between 7 and 100 days in the sixth century (Sun Si Miao of China); between 9 days and 7 years in the thirteenth century (Bernard de Gordon); between 20 days and several years in the sixteenth century (Fracastoro) (Blancou 1994). Variation in the length of incubation period or prodromal period is often a reflection of the difficulty of establishing the day on which infection occurred, especially where animals are concerned. Whilst infection with rabies virus generally has an incubation period of 20–90 days, in extreme cases this period may be substantially greater, and is more variable than that noted for any other acute infection. Extremes of preclinical incubation period following rabies virus infection has been reported as being as long as 14 to 19 years although with such cases secondary exposure cannot be discounted (Fishbein 1991).

Incubation periods for the other lyssaviruses are also difficult to determine, largely due to individuals being unable to account for a possible exposure time. Cases of infection with Duvenhage virus are known to have had incubation periods of at least four weeks (Meredith et al. 1971; Paweska et al. 2006) whilst a case of European Bat lyssavirus type-1 (EBLV-1) in an adult human had a 45 day incubation period (Botvinkin et al. 2005). The first case of death attributed to rabies in the UK for over 100 years, that of an infection of a 55-year old bat conservationist with European Bat Lyssavirus type-2 (EBLV-2), suggested that infection may have been due to a bite 19 weeks prior to the first signs of disease although a history of previous bat bites was recorded (Fooks et al. 2003). For the Australian Bat Lyssaviruses a much longer incubation period was observed. A human case of ABLV in 1998 reported that infection of a 37-year old woman led to clinical disease some twenty months after the presumed exposure to virus (Hanna et al. 2000; Johnson et al. 2008a).

Factors influencing the incubation period must include the site and severity of the bite, with bites nearer the head having a shorter incubation period; degree of innervation of the bite site, highly innervated regions of the body such as the face, neck, and hands are thought to be more dangerous, presumably because of their rich nerve supply; the quantity of virus ‘inoculated’ with small amounts of virus possibly being transmitted through bites from bats compared to the larger volume of virus excreted in the saliva of infected dogs; the nature of the infecting virus isolate with experimental differences observed between highly neuroinvasive strains and less neuroinvasive strains and the age and immune status of the host. The incubation period in children is reported to be shorter than that in adults and is probably associated with their infant stature (Warrell 2008; World Health Organization (WHO) 2008).

In regions where rabies virus is known to be circulating, when neurological symptoms of unexplained origin are observed, the possibility of a rabies virus infection should always be considered. Lack of history of a biting incident should not dismiss rabies for consideration, since such bites may have appeared trivial or have occurred many months before and been forgotten by the patient. Travellers to endemic areas should be made aware of the dangers of rabies infection as they are for other diseases such as malaria (Fooks et al. 2003).

Whilst infection from rabies virus was previously only associated with transmission of virus through the saliva of an infected animal, other means of infection are now recognized and care must be taken if at any stage an individual may have become exposed through ‘non-bite’ mechanisms such as infection of an open wound, scratches, and rarely inhalation of aerosolized virus (Vos et al. 2007; Johnson et al. 2006; Winkler et al. 1973). Fears surrounding potential exposure through the urine and faecal matter of an infected animal remain unresolved but are considered to be low risk mechanisms of virus transmission.

Observation of what may constitute clinical presentation characteristic of infection with a rabies virus or a related lyssavirus can be difficult. As discussed previously, incubation times can vary greatly depending on the nature of the infecting agent as well as the site of infection and the species affected. Early symptoms in rabies patients may resemble those of tetanus, typhoid, and malaria, or of viral encephalitides such as those caused by measles virus, mumps virus, herpesvirus, or enteroviruses. Furthermore, differential diagnosis can be problematic due to similarities in early clinical observations with other pathogen based and medical conditions such as transmissible spongiform encephalopathies, tetanus, listeriosis, poisoning, Aujeszky’s disease and most commonly with Guillain-Barré Syndrome (Hemachudha et al. 2002; Solomon et al. 2005). Secondary infections can also mask the presence of rabies infection, occasionally leading to an incorrect diagnosis of disease (Mallewa et al. 2007). Even during late stage infection, highly informative technologies such as Magnetic Resonance Imaging (MRI) cannot conclusively differentiate between paralytic and furious rabies cases although such scanning may enable other sequelea to be ruled out (Laothamatas et al. 2008). In the absence of clinical disease that may be specifically attributed to rabies infection a 2–20 day prodromal period may occur where the victim may suffer from a number of non-specific symptoms including general malaise, weakness, and loss of appetite, fever and headaches. Paraesthesia or itching at the site of the bite is often reported during this initial prodromal phase. The disease may then progress to the clinical stage whereby acute dysfunction of the nervous system ensues leading to the symptoms that, in the public eye are responsible for the general fear attributed to infection with rabies; hydrophobia, madness and death. Approximately 80% of human cases go on to develop ‘furious’ rabies whilst the remaining 20% develop the paralytic, or ‘dumb’ form of the disease. Once clinical disease is apparent in the vast majority of cases will ultimately result in death. The few reported survivors have suffered severe neurological impairment.

All mammals are susceptible to infection, however, there are no species specific symptoms apart from altered behaviour as a direct result of neuronal involvement. As with human infection, a prodromal phase occurs where the infected animal may become noticeably lethargic and display signs of weakness. Domesticated animals will often become less interested in interacting with owners whilst wild animals that would normally avoid contact with man may seek out human interaction.

Similarly to human infection, rabies may manifest in a variety of different ways, although clinical progression is usually described as either a ‘furious’ or a ‘dumb’ form. However, these two distinct clinical presentations of the disease are extremes of infection and an individual animal may exhibit clinical manifestations specific for either form of disease or elements of both. It is thought that the clinical progression of disease may be linked to the primary site of viral replication within the CNS, which must be linked to the site of virus entry and the amount of virus transmitted. A number of ongoing studies are attempting to relate disease pathogenesis to involvement of different regions of the brain although currently the exact mechanism of pathogenesis within the brain is unclear (Hicks et al. 2009).

With furious rabies, after a brief prodrome where animals may appear more alert than normal, an acute neurological phase often occurs with signs similar to acute disease in man. Infected animals have been observed exhibiting parasthesia, with animals being seen to excessively groom or scratch areas presumed to be the site of infection. Parasthesia may progress to extremes of self-mutilation and even self-consumption of the affected region leading to severe trauma. Often animals will snap at anything within reach and will hold onto anything they are able to bite hold of with great tenacity. Neurological signs then develop progressively becoming almost continuous until death which is generally due to respiratory arrest and organ failure.

Cases of dumb rabies in dogs often initiate the cycle of clinical disease with the animal becoming lethargic and often showing a lack of coordination. This situation may rapidly worsen or remain stable for a short while before further clinical signs develop. Motor paralysis often develops initially as a ‘weakness’ of the hind limbs. As infection progresses the paralysis spreads affecting all limbs leaving the infected animal quadriplegic. Mandibular paralysis may also occur causing a loss of the swallowing reflex. Such ascending paralysis is commonly observed in wild as well as domestic animals. Once dumb rabies has progressed to this stage the infected animal may succumb to infection without any additional clinical signs.

Early signs of infection in dogs generally lasts for two to five days before either the development of paralytic rabies, in approximately 75% of infected animals or the aggressive form of disease, furious rabies in the remainder. Paralysis and death will often occur four to eight days after the onset of clinical disease. Although typically thought of as a disease of canines, felines can be infected and more often develop the furious form of the disease and can pose a greater threat to man as their role as a companion lap animal means that statistically they are more likely to scratch at the face and neck of their owners with claws harboring infected saliva.

The earliest reported experimental diagnosis of rabies virus was by the German scientist, Georg Gottfried Zinke (1804), and specific diagnostic tests have been the mainstay of case confirmation and surveillance. The majority of the traditional tests are, however, relatively slow and insensitive compared to today’s techniques, and in some cases require the use of live animals. Here, we describe the techniques currently used in the diagnosis of rabies and lyssaviruses and highlight the importance of more recent molecular based technologies. Unfortunately, in countries where there is the most frequent need for robust, timely and accurate diagnostic tools, mainly across the developing world where the virus remains endemic, clinical diagnosis with or without an examination for Negri bodies is often the only diagnostic method that is attempted. However, whenever feasible clinical suspicions should be confirmed by laboratory tests. The accurate diagnosis of infection with rabies and lyssaviruses is generally confirmed by internationally standardized laboratory tests undertaken on CNS tissue removed from the brain at post mortem. Recommended regions of the brain include sites where high levels of virus antigen can be readily detected such as the hippocampus, cerebellum and the medulla oblongata. In suspect cases, it is usual for the entire head of the animal to be submitted to an approved laboratory for confirmation of infection. However, if a large number of animals are involved, brain material from individual animals may be submitted using the ‘straw technique’ whereby a plastic straw-like rod is passed through the occipital foramen to collect suitable material for testing (OIE 2004).

There are no gross lesions characteristic for rabies virus infection visible at autopsy. In most animals, microscopic non-specific lesions suggestive of viral encephalomyelitis with ganglioneuritis may be observed in the nerve centers, as well as histolymphocytic cuffs and gliosis. The most notable lesions are usually in the cervical spinal cord, hypothalamus and pons. The only specific lesions consist of intracytoplasmic eosinophilic inclusions termed ‘Negri bodies’ that are discussed below.

Some of the earliest diagnostic methods for the detection of rabies virus infection in post mortem material were based on histopathological examination. Specific round to oval inclusions described by Negri (1903) usually identified in the cytoplasm of undamaged nerve cells and particularly in the hippocampus were routinely used for diagnosis using both histological and immunological techniques. These ‘Negri bodies’, which consist of a reticulogranular matrix containing tubular structures measuring from 4 to 5µm are contiguous with maturing virus particles. However, absence of these disease markers does not exclude the disease. As will be described later, these markers of infection may be detected in almost any region of the brain, but are primarily seen in the Ammon’s horn of the hippocampus, the ganglioneurones and neurones of the cerebellum, the motor area of the cerebral cortex, and in the medulla. Their frequency of occurrence, size, and shape may be influenced by the host species, the infecting virus strain, and the clinical phase.

Rabies diagnosis was previously based upon the detection of Negri bodies and the use of histological tests such as Sellers’s staining techniques or Mann’s fixation protocols (WHO 1996). These assays are now rarely undertaken as they have been superseded by more accurate and reliable tests and are now considered unreliable, particularly when decomposed material is submitted (McElhinney et al. 2004). However, despite histopathological techniques being discontinued for use in a diagnostic capacity, a molecular based method has been developed for the detection of species 5 and 6 lyssaviruses (Finnegan et al. 2004). This technique is a robust, highly sensitive and specific in-situ hybridization technique, and employs digoxigenin labelled riboprobes for the detection of lyssavirus RNA in mouse-infected brain tissue. Using this method, both genomic and messenger RNAs have been detected. The ability to detect messenger RNA is indicative of the presence of replicating virus. Whilst this assay is of diagnostic value, interest in such new protocols is driven by the need to investigate lyssaviruses other than the classical rabies species (Hicks et al. 2009).

Over the last ten years a number of novel techniques have been developed for diagnostic confirmation of rabies viruses in clinical specimens, yet the most commonly used test is the fluorescent antibody test (FAT) developed over 35 years ago (Dean and Abelseth 1973). This test, still recommended by the OIE and WHO as the ‘Gold Standard’ for diagnosis of rabies virus in clinical specimens, depends on the reaction between virus antigen in the brain and fluorescently labelled anti-rabies antibodies directed against the nucleocapsid protein. Not only is it routinely used for confirmation of the presence of antigen in brain smears but can also be used for virus isolation on passaged material as well as virus identification intra vitam from skin biopsy material. Whilst this test can give accurate (95–99%) results within hours of receipt of a fresh clinical specimen, its sensitivity can be affected by the quality of the tissue examined, autolysis, presence of different lyssavirus antigen and the region of the brain submitted for analysis.

The rapid rabies enzyme immunodiagnosis (RREID) technique, based on the detection of virus nucleocapsid antigen in brain tissue, was the first technique to take rapid accurate post-mortem diagnosis out of the laboratory as it did not require microscopy and, with the aid of a special kit, could be used under field conditions. The RREID test can be used to examine partially decomposed tissue specimens for evidence of rabies virus but it cannot be used with specimens that have been fixed in formalin, and comparative studies against fluorescent antibody based tests suggested that its sensitivity may not be as effective as that of the fluorescent antibody based methods (Perrin et al. 1986; Franka et al. 2004). Despite this drawback, such methodologies have been shown to be of benefit for large scale surveillance having the potential to be automated (Bourhy et al. 1989; Perrin and Sureau 1987; Bouhry and Perrin 1996). More recently, an avidin-biotin amplified dot-blot enzyme immunoassay has shown good sensitivity and specifity in small scale trials using infected brain tissues but has not yet been tested on clinical samples such as saliva for ante-mortem diagnosis of human cases (Madhusudana et al. 2004). Use of ELISA in serological assays will be discussed later.

The OIE guidelines for diagnosis of the rabies virus infection primarily recommends the use of FAT on post-mortem brain material. Confirmatory virus isolation must, however, be used to support the initial diagnosis, particularly when FAT results are equivocal and where human exposure is involved. To this end the mouse inoculation test (MIT) was developed. This test relies on the detection of virus antigen in the positive FAT sample and requires processing of post-mortem tissue to generate a homogenate of brain material. As for the FAT test, regions are carefully chosen and pooled to form a homogenate, which is then inoculated intracerebrally into weanling mice. Clinical signs (positive result) may be seen as early as six to eight days for RABV although it can often take considerably longer for mice to succumb and considerable variation in the onset of clinical signs seen when using this method has been reported (Johnson et al. 2003). Mice should be observed for a period of 28 days. This in vivo test is time-consuming, expensive and involves the use of animals. The WHO recommends that it should be avoided for routine diagnosis if validated in vitro methods, such as isolation of virus on tissue culture cells, are established within the laboratory (WHO 2005).

Tissue culture isolation of virus was once problematic due to the fact that lyssaviruses rarely cause cytopathic effect that can be observed using the light microscope. Some viral strains can cause the formation of large fusions of cells, termed ‘syncytia’, but not all strains of virus have fusogenic ability and thus infection of a cell monolayer is generally difficult to interpret.

Cell culture techniques such as the Rabies Tissue Culture Inoculation Test (RTCIT) used the fluorescently conjugated antibody developed for the FAT test to detect virus nucleocapsid in tissue culture. The test involves the inoculation of the sample into a neuroblastoma cell line that has been shown to be highly permissive for the majority of lyssaviruses. Positive results are commonly obtained within 2–4 days. The FAT is then used to confirm the presence of rabies virus antigen in either infected mice or cell monolayers. From a research perspective, these virus isolation techniques provide an opportunity to generate individual virus stocks which can be further characterized using molecular methods as described below.

Advances in molecular techniques over the last decade have seen PCR assays replace many of the more traditional screening methods for lyssavirus diagnosis. These are now becoming more widely accepted and accessible for the diagnosis of rabies and, whilst not recommended by the WHO for routine post-mortem diagnosis of lyssaviruses, the use of the reverse transcriptase polymerase chain reaction (RT-PCR), nested or hemi-nested RT-PCR and other PCR based formats are increasingly being used. Many laboratories have now enforced strict quality control procedures and with trained staff with demonstrable experience and expertise have applied these molecular techniques for confirmatory diagnosis and epidemiological surveys. As well as post-mortem diagnoses, RT-PCR has been used to confirm rabies infection intra-vitam in suspect human cases, where conventional diagnostic methods have failed and post-mortem material is not available (Smith et al. 2003). RT-PCR techniques can be used not only to detect rabies virus RNA in a wide variety of biological fluids and samples (e.g. saliva, cerebrospinal fluid, tears, skin biopsies and urine) but is also now used to differentiate between different species using both conventional and real time RT-PCR assays (Wakeley et al. 2005). Real time assays have recently been used for the rapid diagnosis and species confirmation of EBLV-2 infected Daubenton’s bats in the UK (Fooks et al. 2004) as well as in the intra-vitam detection and genotyping of two human rabies cases in the UK in 2002 and 2005 (Fooks et al. 2003b; Solomon et al. 2005) and for the detection of a rabies in a quarantined puppy in UK (Fooks et al. 2008).

Molecular tools such as PCR and sequencing continue to enable the characterization of new virus isolates and to date a number of lyssavirus isolates have been sequenced in their entirety with at least one virus from each of the 11 species having been sequenced (Table 35.2).

Detection of virus-specific antibodies is important when assessing past-exposure to infection, especially when virus can persist in a population. However, infection with rabies virus is invariably fatal and post-infective rabies antibodies are rarely detectable before death so serological techniques are superfluous in assessing the presence of virus within a population. However, the presence of virus-specific antibodies within a vaccinated population can be used to assess vaccination status and is an important tool in both monitoring individuals that receive vaccine as a requirement of their work as well as the movement of vaccinated animals between countries. Serological assays such as the rapid fluorescent focus inhibition test (RFFIT) or the fluorescent antibody virus neutralization (FAVN) test can be used to evaluate seroconversion within a vaccinated population and estimate an individual’s potential likelihood of being protected from infection with a wildtype virus. Whilst this situation remains valid for rabies viruses, in recent years a number of other lyssaviruses have been isolated in bats and it seems that, within this host species, the virus may circulate in the absence of clinical disease (Serra-Cobo et al. 2002) and that animals may be serologically positive for having encountered the virus (Brookes et al. 2005; Hayman et al. 2008; Kuzmin et al. 2006).

Both the FAVN and RFFIT are OIE prescribed tests for international travel of companion animals. The principle examined with both tests is the ability of antibodies within suspect serum samples to neutralize virus. Non-neutralized virus is essentially what is detected by fluorescent antibody staining and the dilution at which the test sera was able to completely neutralize virus is quantitated. Standardized reference sera from the OIE and WHO are used as comparators that allow quantitation of virus neutralizing antibody present in samples from animals and humans. These tests are internationally accepted as the ‘Gold Standard’ for rabies virus serological assessment and are universally accessible. From a logistical point of view, however, use of these tests in areas where the virus remains endemic is problematic as facilities for handling live virus are frequently unavailable. This fact has been the driving force behind attempts to replace these tests with novel methodologies that do not require high security containment facilities and vaccinated staff. Recent progress has been made using novel ELISA techniques that have been shown to be suitable for the detection of rabies virus-specific antibodies in serum samples from companion animals and humans (Cliquet et al. 2004; Servat and Cliquet 2006; Servat et al. 2007; Feyssaguet et al. 2007). Recently, the companion animal ELISAs have been accepted by the OIE as a screening tool or alternative to the FAVN but await approval for use in the EU. Once fully validated these new ELISAs will overcome the problems associated with the current tests and should allow development of local and national diagnostic facilities across the developing world.

Advances in molecular biology have identified lentiviral vectors as suitable expression vectors for antigen generation for a number of viruses. Recently, this approach has been extended to members of the lyssavirus genus whereby glycoprotein genes of different lyssavirus species have been cloned into lentiviral vectors to be co-expressed with HIV/MLV gag-pol and GFP/luciferase in human epithelial cells (Wright et al. 2008). When used in neutralization assays these lyssavirus pseudotype constructs replace the use of live lyssaviruses in these tests and therefore overcome a major obstacle in transferring rabies diagnostic capabilities to the developing world, and the need for high security facilities to undertake serological work becomes redundant. The suitability of lentiviral pseudotyped lyssaviruses is currently being assessed by Rabies Reference Laboratories.

Rabies vaccines that are virtually 100% effective in preventing disease have been available for over 100 years. The real constraint to vaccination is the cost of vaccine production and delivery, awareness of the need for pre-exposure vaccination and the lack of both social and economical infrastructures in regions where the threat of human infection is highest and the virus is endemic. Although a number of vaccines developed have been associated with clinical disease and mortalities in vaccinees, primarily using nerve tissue based vaccines (NTVs), millions of lives have been saved through vaccination. Therefore the value of vaccination cannot be underestimated although it is economically prohibitive or even impractical to vaccinate local populations, especially when the regions at risk are some of the poorest populations in the world. Currently, additional cost-effective strategies for use in developing countries are required to mitigate the omnipresent risk to human health.

The first rabies vaccines developed were essentially a crude suspension of desiccated brain material and spinal cord material from an infected animal. Often live virus vaccines were prepared in sheep, rabbit or goat brains. Use of these preparations as a live virus vaccine led to many cases of vaccine associated rabies and understandably people became reluctant to be vaccinated. A further drawback to vaccines generated from infected nervous tissue was the risk of neuroallergenic responses to inoculation. Later, vaccines were inactivated using the techniques of Fermi and Semple, at first using phenol and later beta-propriolactone, vastly increasing the safety of the preparations (Bugyaki et al. 1959). As far back as 1973 the WHO recommended that the use of crude nerve tissue based vaccines be stopped, however, recent publications suggest that these phenol inactivated preparations may still being used in some of the poorest parts of the world where the disease remains endemic (Ayele et al. 2001; Parviz et al. 2004).

In some parts of the world vaccines of crude nervous tissue origin have now been replaced by vaccines made through growth of virus in the brains of suckling mice. These vaccines are prepared using mice aged less than 1 day to reduce the level of potential encephalitogenic substances and were shown to be considerably safer, notably reducing the number of cases of neuroallergenic side effects seen post-vaccination (Fuenzalida and Palacios 1955). However, crude nerve tissue based vaccines are thought to still be used extensively across Latin America, Africa and Asia (Briggs and Hanlon 2007). Vaccine preparations were again improved with the generation of vaccines in embryonated eggs. Vaccines like the duck embryo vaccine were used for almost 20 years across Europe and North America. Specific responses generated to the duck embryo vaccine were, however, relatively poor and low immunogenicity deemed that it required 14 to 20 doses before suitable protection was conferred to the vaccinee and thus its production was abandoned (Shaul et al. 1969; Plotkin 1980).

Whilst it is believed that a number of the early NTVs and embryo based vaccines are still being used in some of the poorest parts of the developing world, modern tissue culture derived vaccines have largely replaced the use of these potentially dangerous early vaccines. The breakthrough came in the early 1960s with the development of the Human Diploid Cell Vaccine (HDCV) (Wiktor et al. 1964). This vaccine has no doubt saved millions of lives and was pushed into the global spotlight through being used in a WHO vaccine ‘field-trial’ to treat humans that had suffered severe bite wounds from rabid wolves (Bahmanyar et al. 1976). The ability of the HDCV to generate highly protective virus neutralizing antibody responses made it the ‘vaccine of choice’ for a number of years but very high production costs and low virus yield made the vaccine unaffordable in the parts of the world that needed it most.

With production costs being a pre-eminent consideration of use, further tissue culture based vaccines have been generated in an attempt to make pre-exposure vaccination affordable to the developing world, especially in regions where people live on less than US$1 per day. More recent vaccines include purified chick embryo cell vaccine (PCECV) and purified Vero rabies vaccine (PVRV). These vaccines have successfully replaced the use of controversial NTVs in several areas where the cost of administering HDCV is unaffordable, such as across Thailand. Despite their safety and antigenicity, there still remains a potential problem with successive booster injections where, as with HDCV vaccination, occasionally allergic responses, presumably to the beta-propriolactone used to inactivate the virus, have been reported. It has been suggested that individuals who have experienced such reactions should receive no further vaccination with these vaccines unless they are actually exposed to rabies virus (Nicholson 1990). PVRV vaccine is currently licensed across much of the globe including much of Europe and Latin America where it is replacing use of the suckling mouse brain vaccine that has been in use since the 1950s. However, at the global level it is recognized that several countries are still vaccinating individuals with out-dated vaccine types. Bangladesh, the Union of Myanmar, Pakistan, Peru, and Argentina have all still to convert to cell-culture based vaccines with Pakistan still reportedly using NTVs (Burki 2008). In 2008, India announced their explicit intention to phase-out and halt the use of NTVs and to replace their use with tissue-culture based vaccines for human use.

Whilst modern technological advances seek to improve vaccines, those currently available and recommended by the WHO are considered to be safe, highly immunogenic and are being used successfully across the globe (Quiambo et al. 2000; Jones et al. 2001).

The value of pre-exposure immunization is very high. However, in parts of the world where either modern, safe vaccines are not available or where medical professionals lack awareness of the current tools available to protect local populations, significant numbers of individuals continue to die annually from the disease, a large percentage of which, tragically are children. In the developed world, where resources are available to enable vaccination, pre-exposure immunization has saved many lives. It is clear that all individuals that either work with the virus or who may come into contact with infected animals (especially where the virus remains endemic) must be vaccinated.

Current recommended immunization schedules vary depending on the vaccine being used and the severity of risk. For scientists working with the viruses, it is essential that antibody titre is measured post-immunization and that an antibody titre of at least 0.5IU/ml is achieved (OIE 2004; WHO 1996). If the antibody titre of someone working routinely with the virus falls below this level then a booster injection is required although generally, boosters or antibody tests are recommended every three to five years for other risk groups such as bat workers. The discovery of a number of novel lyssaviruses, often from bat species, has prompted studies into the protection afforded by current rabies vaccines to these lyssaviruses. Whilst protection studies have been undertaken, exact protective titres are unknown (discussed further below). Recent studies using recombinant vaccinia viruses encoding the glycoprotein genes of rabies virus, Mokola virus and West Caucasian Bat Virus, either expressed individually or in pairs have shown promise as potential multivalent vaccines (Weyer et al. 2007). Development of human vaccines that protect against lyssaviruses as well as classical rabies strains are currently being developed as alternatives that could confer a cross protective response in the vaccinee (Nel 2005; Hanlon et al. 2005).

The key to post exposure prophylaxis (PEP) following a potential transmission of lyssaviruses to an individual is timing. The risk assessment of a potential exposure to rabies and lyssaviruses must be conducted urgently. PEP should be administrated rapidly following exposure. Supplies of human and equine rabies immunoglobulin are expensive, and relatively speaking in short supply, and are therefore generally given in high risk exposures. Immunoglobulins used consist of either human or equine serum that contains neutralizing antibodies directed to discrete epitopes on the rabies virus. These serum, when administered rapidly, provides passive rabies virus neutralizing antibody during the two to three day post infection ‘window’ where the immune system has not yet responded to the inactivated virus vaccine. As well as neutralizing infectious virus particles it is thought that RIG may also participate in the elimination of infected cells by antibody-dependent cellular cytotoxicity. In the first instance, as much RIG as is possible should be applied to the wound and any exposed mucus membranes surrounding the wound. Guideline doses are 20 IU RIG per kg bodyweight of human RIG or 40 IU RIG per kg if equine RIG is used. Any remaining RIG should be injected intramuscularly into the area around the wound site (WHO 2005). Several different regimens currently exist for the administration of RIG.

Alongside PEP, and always prior to it being administered, wound care is the principal factor that will greatly influence the possibility of survival from infection. The wound should be thoroughly washed immediately with soap and water or detergent and then, if available, it is recommended that either 70% ethanol or an aqueous solution of iodine should be used to further sterilize the affected area (WHO 2005). This immediate wound treatment serves to remove or inactivate infectious virus particles (and/or other potential pathogens) from the site before they have an opportunity to replicate or to reach the nerve endings. Importantly, the wound must never be sutured as blood flow from the wound allows for additional virus removal from the bite site.

Despite the effectiveness of PEP in practice it is rarely administered, and it is estimated that only 5% of cases actually receive PEP as recommended by the WHO. This failure is generally due to lack of availability, which is intrinsically linked to the cost of prophylactic treatment and which often far outweighs the resources available.

At the same time a suitable vaccine, usually whatever vaccine is available, should be inoculated but at a site distant to the region treated with RIG to avoid the vaccine being neutralized by complexing with the RIG. Several approaches to post exposure vaccination are used across the globe (reviewed in Briggs and Mahendra 2007). There are four basic vaccination regimens recommended for use as PEP with cell culture based rabies vaccines and their administration differs across the globe. Two intramuscular and two intradermal regimens currently exist. However, the intramuscular Essen ‘four-dose’ regimen is most often applied across North America and Europe (WHO 2005). Some European countries use the ‘Zagreb’ or ‘2-1-1’ regimen which reduces the number of doses required compared to the Essen regimen. Intradermal vaccination regimens have been shown to be effective, although drawbacks include requirement of storage at 2–8°C post reconstitution meaning that it is often not cost effective to treat one or two individuals with these preparations as the reconstituted batch may expire. Furthermore, intradermal vaccination requires highly skilled staff. Two intradermal regimes currently exist, the ‘Thai red cross’ and the ‘Eight site’ regimen (Warrell 2008; WHO 2005). Vaccination regimens as PEP, when applied are very effective but require concomitant application of RIG. Human treatment failures where PEP has been used can generally be attributed to the incorrect PEP as the short incubation period linked to bite of the head and neck. (Arya 1999; Sriaroon et al. 2003; Parviz et al. 2004). However, the effectiveness of PEP, when administered properly, cannot be overstated.

With the biggest threat of infection coming from rabid dogs, the first vaccines were tested extensively in canines by Pasteur and those that followed up his initial research (Bunn 1991). Nerve tissue vaccines and vaccines developed subsequently derived through passage in embryonic chicken eggs were all developed but often either caused post-vaccinal allergic reactions that could kill the vaccinee or, with egg derived vaccines, those that were effective in adult dogs, often caused rabies in younger dogs, cats and cattle (Bunn et al. 1991).

Live virus vaccines were also developed and are still used in parts of the world where safer but more costly vaccines are unavailable. Passage of Pasteur’s ‘fixed’ virus strain was continued for many years by passage in rabbit spinal cord material and mice and its common descendant is what we now term the ‘Challenge virus strain’ (CVS). CVS, after repeated passage in baby hamster kidney cells (BHKs) was denoted CVS-11 and was used extensively as a modified live virus vaccine in dogs although its use was confined to dogs due to concerns about safety in other animals. A number of other cell culture derived modified live vaccines have been generated using ERA, SAD and CEO strains that are used in different parts of the world (reviewed in Reculard 1996).

Genetically engineered vaccines have also been developed and used widely, in attempts to control rabies in wildlife species. Oral vaccines based on vaccinia virus expressing rabies virus glycoprotein were first produced in the early 1980s (Kieny et al. 1984; Pastoret et al. 1988). This new type of vaccine overcame fears surrounding the use of live attenuated virus vaccines and in various forms has been applied with considerable success. In 1995, after trials both in the USA and France, an oral vaccine was licensed for use against rabies in wildlife and has been successfully used to contain and eliminate rabies across much of Canada and the USA (Krebs et al. 2005; Slate et al. 2005). Oral vaccine preparations are constantly being optimized in an attempt to eliminate the virus from areas where it continues to pose a threat to the human population (Slate et al. 2005; Cross et al. 2007).

New vaccines based on using other viruses as vectors for glycoprotein sequence include those based on canarypox virus, such as the combination rabies vaccine currently licensed in the USA (Anon 2006) and a recombinant adenovirus-vectored vaccines that is showing promise as a potential canine rabies virus vaccine (Tims et al. 2000; Hu et al. 2006, 2007).

Unfortunately, despite the number of deaths annually attributed to lyssavirus infection and the recognition that this number is likely to be vastly underestimated, rabies is not considered a priority by many governments and pharmaceutical companies whose input would be needed to try and combat the disease on a global level. The discovery of a number of novel lyssaviruses that are genetically related to classical rabies virus and are important human pathogens has increased the profile of these viruses. However, it is clear that even this emerging threat will do little to drive investment into new generation vaccines that may be needed to prevent countless further deaths. Improved oral vaccines, which are cheaper to produce and with improved bait specificity may help the current situation. Such vaccines need to be applied across much of Africa and Asia where the virus remains endemic in both domestic and wild populations (Cross et al. 2007). Of the few studies currently underway, those incorporating cheaper alternatives to the tools currently available such as generating vaccines in plants look promising (Girard et al. 2006; Lodmell et al. 2006; Yusibov et al. 2002).

Alternatives to RIG as post exposure prophylaxis are also currently being explored to overcome the expense that precludes the uses of RIG across much of the developing world. Currently, no effective antiviral strategies to combat infection with rabies or rabies-related lyssaviruses exist. Recent studies with a number of different negative strand RNA viruses have suggested that silencing virus-specific gene expression using RNA interference (RNAi) molecules to disrupt viral replication may be an effective measure in instances where antiviral agents are not available.

RNAi is the mechanism by which mRNA degradation is induced by intracellular double-stranded RNA in a targeted sequence-specific manner. This pathway is thought to be an evolutionary mechanism for protecting the host and its genome against viruses and rogue genetic elements that use double-stranded RNA (dsRNA) in their life cycles. Until recently, this intracellular process was unknown within the scientific community as the response is masked by the cellular non-sequence specific responses to generation of dsRNA moieties greater than 30 nucleotides in length that triggers the interferon pathway (Elbashir et al. 2001). The observation that RNAi is able to silence gene expression in a sequence specific manner has driven the application of this technology to studies with a number of viral pathogens. RNAi is a powerful tool amenable to the development of anti-viral drugs. Recent advances with this technology have shown that siRNA molecules tagged to a peptide generated from a region of the rabies glycoprotein that is known to bind acetylcholine are able to cross the blood brain barrier and inhibit viral replication in the brain (Kumar et al. 2007). Whilst such studies offer exciting possibilities for antiviral treatment through transvascular delivery of molecules to the CNS, a post exposure prophylaxis that can be administered in and around the ‘bite site’ is also of huge importance.

The EU has made huge steps forward into the elimination of terrestrial rabies from a number of its member states over the last 20 years (OIE 2005). However, the circulation of the virus in both fox and raccoon-dog populations continues to be a major threat to domestic dog populations and hence to human health.

Historically, the disappearance of terrestrial rabies from central Europe during the first half of the twentieth century was attributed to a reduction in the number of stray animals across the continent as well as the disappearance of wolf populations through hunting and human incursion into wolf territory. This left great swathes of Central Europe largely free of rabies throughout the 1900s although wild carnivore species had become infected and so the disease remained. The disease re-emerged across much of Europe through the establishment of infection within the red fox population which through the 1940s spread rabies across the Russian-Polish border then over the following four decades through much of the rest of Europe. Rabies infection within domestic dog populations has now been controlled, largely through oral vaccination and quarantine rules set in place throughout the 1970s although the virus still exists in wildlife species, principally the red fox (Vulpes vulpes) and the raccoon dog (Nyctereutes procynonoides) (Steck and Wandeler 1980; Johnson et al. 2003b). Whilst the fox population has long been defined as a reservoir, it is still not clear whether the raccoon-dog population is a distinct reservoir or a spillover host. Rabies infection of other species include stoats, weasels, deer, water voles, hedgehogs, badgers, voles, stone martens and free-roaming cats, the latter of which is believed to play an important part in the epidemiology of the virus. The implications of infection within such populations are unclear as molecular and serological data for cases within these species are few.

The UK has a long history of terrestrial rabies although through the latter half of the nineteenth century and the start of the twentieth century the disease was rarely reported in wildlife species and terrestrial rabies was, through animal destruction and quarantine, finally eliminated in 1922 (Fooks et al. 2004). As an island nation, the UK was then able to, through very strict import regulations including the quarantine of imported dogs and cats, maintain its rabies free status. However, when cases have occurred in quarantine, such as recently with the importation of infected puppies from Sri Lanka, the disease has been identified and infected animals eliminated with no further risk of disease (Fooks et al. 2008). The introduction in 1971 of vaccination of dogs and cats on arrival in quarantine has resulted in the successful maintenance of a rabies free status. The last indigenously acquired human rabies (rabies species) case occurred in Wales in 1902 with the last case of indigenous terrestrial rabies in 1922. However, since 1946 a further 25 human deaths have been recorded in the UK, all of which were the result of infection transmitted to individuals during travel abroad (Johnson et al. 2005; Banyard et al. 2010) most recently that of a young woman returning from South Africa having been previously bitten by a rabid dog (Hunter et al. 2010).

A further threat of rabies to the European Union comes from the identification of rabies-like viruses in the endogenous bat populations across Europe. The presence of bat rabies poses a low but serious threat to human health although infections have been reported. Four human fatalities have been attributed to EBLV infection: in Russia and the Ukraine (EBLV-1); in Finland and the UK (EBLV-2) (Lumio et al. 1988; King 1991; Selimov et al. 1989; Fooks et al. 2003). Furthermore, CST events into other species including Danish sheep (Ovis aries) and a stone marten (Martes foina) in Germany have also been reported although these events have only been reported for EBLV-1 (Muller et al. 2004; Stougaard and Ammendrup 1998; Tjornehoj et al. 2006) (Table 35.3). Experimentally, sheep have also been shown to be susceptible to EBLV-2 infection (Brookes et al. 2007). The viruses identified as being the causative agents of bats rabies within Europe were originally characterized through monoclonal antibody typing as being Duvenhage-like viruses and were grouped into serotype 4. However, more modern genetic characterization has enabled a distinction between viruses isolated and they are now defined as European Bat Lyssaviruses-1 and -2 (EBLV 1 and 2), however, both species are genetically distinct. The reservoirs for these viruses have been identified as principally being Eptesicus serotinus for EBLV-1 and both Myotis dasycneme and Myotis daubentonii for EBLV-2 although for both viruses infection of other bat species has been reported (Schneider and Cox 1994). Recently, in 2002 a bat conservationist died in Scotland following CST from a Daubenton’s bat. The virus was characterized and typed as EBLV-2 (Fooks et al. 2003). Furthermore, in August 2007 a Daubenton’s bat in Germany was reported to be harbouring EBLV-2, the first isolation of the virus in this country suggesting that these viruses may have a broader geographical range than originally thought (Fig. 35.1a) (Muller et al. 2007; Freuling et al. 2008).

Oral vaccination campaigns in wildlife species have done much to reduce the incidence of rabies across Western Europe, with the decline of rabies cases in Germany giving a strong indication of the reduction in cases across the remainder of countries following vaccination programmes (Fig. 35.1b) (Brouchier et al. 1991). However, rabies virus still exists in regions of Eastern Europe including the Russian Federation and Belarus. Currently rabies infection in Europe is seen, from a geographical perspective, as being divided into two areas. If a line is drawn between the Baltic Sea down to the Black Sea it is clear that any future cases of rabies virus infection are more likely to occur to the east of this defined boundary, although incursions may occur where rabies still exists within areas along the boundary line such as in Turkey (Johnson et al. 2006b). To the East of this boundary are some of the poorest regions of Europe including the Baltic States where high numbers of rabies cases have been reported. The prevalence of the raccoon dog across parts of north-eastern Europe perpetuates the threat of the virus in these areas.

Clearly, within these areas rabies infection is not the only threat to human health and what financial resources exist may well be allocated to issues deemed to be of higher priority. New strategies are required to counter this continuing threat and the EU must focus on generating coordinated efforts between the affected countries to try and maximize the cost effectiveness of a strategy and that attempts to finally free Europe from this preventable disease. Rabies cases reported across Europe over the last decade are detailed in Fig. 35.2.

Rabies infection in Asia continues to be vastly underreported for a wide variety of reasons; many countries do not have the infrastructure or the necessary veterinary resources to actively report cases of the disease, other regions suffer the huge problem of unknown stray dog populations that harbour the disease whilst tragically, many infected individuals are kept at home, often left to suffer, with no medical attention or PEP being available. Much of the information reported on the incidence of rabies across Central Asia comes from a small number of countries that are either willing and/or able to report such data to the WHO and the OIE with only 40% of the territory of central Asia reporting data. Again, dogs and cats are the primary domestic species affected with foxes being the most affected wildlife species.

Canine rabies of both domestic and wild populations exists across much of Asia with Afghanistan, Bangladesh, Cambodia, China, India, Indonesia, Laos, the Middle East, Nepal, Pakistan, the Philippines, Sri Lanka, most of the former Soviet Republics, Thailand, and Vietnam being endemic for canine rabies. Canine rabies therefore is the biggest threat to human health.

Across India alone, it is estimated that 20,000 people die each year as a result of canine rabies, just under two thirds of the estimated number of human lives lost to the disease annually across the whole of the Asian continent (Pradhan et al. 2008). In India, dogs remain as both the main reservoir for the virus and the primary transmitter to the human population, with a large proportion of cases involving children under 14 years of age. Current efforts at eliminating rabies from India are centered around vaccine campaigns for domesticated dogs as well as those deemed to belong to a community (those that are not individually owned but are fed waste food and therefore continually return to a local area to be fed) and the use of oral vaccines present in bait for vaccination of wild canine populations.

In Eastern Asia, studies have shown that, the dog is the most reported reservoir. In Thailand, transmission from dogs to the human population accounts for 70–95% of rabies deaths (Kasempimolporn et al. 2008). Dog owners who fail to get their pets vaccinated cause additional problems in controlling rabies in domestic animals across much of Asia. Even if facilities exist for vaccination, the cost and general ignorance of the importance of such actions often leads to animals not being vaccinated. Recent studies in southern Asia have suggested through accumulated history records that as many as 78% of owned domestic dogs were not

vaccinated against rabies (Singh and Sandhu 2008).

The steady increase in cases of rabies across China has also provoked public fears of the disease (Hu et al. 2008a). A recent study has shown that since the start of the new millennium, the incidence of rabies in southern and south western territories of China has risen dramatically (Fig. 35.3). Four defined regions have been found to be endemic for the virus and the necessary resources and infrastructure, as well as ample supply of human vaccine, are lacking in these areas. Again, as seen globally, the victims of the disease are mainly children or teenagers that have been bitten by a rabid dog and again, improper wound treatment, lack of PEP and short supplies of rabies immunoglobulin have meant that fatalities have been greater than expected. To counter this increase in cases, both local and national authorities need to act to raise awareness and provide the necessary tools to protect those at risk as well as those exposed using conventional PEP regimens (Si et al. 2008, Tang et al. 2005; Hu et al. 2008a).

Whilst dogs represent the greatest risk of infection with rabies across much of Asia, in the Middle East foxes, jackals and wolves also constitute important wildlife reservoirs. The red fox and the golden jackal are considered the most problematic carnivores spreading rabies virus in the Middle East although minor epidemiological roles are also thought to be played by other carnivorous mammals such as the grey wolf, the mongoose and the badger (Seimenis 2008). Several countries across the Middle East, including Iran, Israel, Oman, Saudi Arabia, Turkey, and Yemen, are experiencing an increase in cases of rabies in wildlife species with the red fox and the golden jackal carrying the greatest burden of disease (Vos 2004).

Despite the presence of bat rabies across much of Europe and the Americas, bat rabies is rarely reported from Asia, although it should be noted that bats are not generally examined unless captured following human exposure. However, a number of unclassified viruses (Aravan, West Caucasian Bat Virus, Khujand, and Irkut virus) have been isolated in and around the geopolitical areas that separates Europe, the Middle East and Asia (Arai et al. 2003; Kuzmin et al. 2003, 2005).

Rabies has been present across much of Africa for centuries, although its history prior to 1900 is fragmentary and largely anecdotal. Characterization of viruses from Africa remains problematic due to the lack of economic infrastructure and, as a result, well equipped laboratories. Indeed, in sub-Saharan Africa, disease recognition only followed the establishment of diagnostic laboratories at the turn of the twentieth century. The origins of rabies in Africa are still questioned. However, modern phylogenetic analyses have defined a number of rabies species variants as being members of a ‘cosmopolitan lineage’. This lineage includes isolates that, from molecular data are genetically related, but have been isolated from all corners of the globe. It has been suggested that that progenitors of the strains classified within this lineage were distributed widely across the globe during colonial times, giving the possibility that Europe may have served as a source of rabies for the rest of the world. Current phylogenetic analysis shows that since this period, isolates now cluster into a number of distinct clades based on geographic origin (Nadin-Davis and Bingham 2004). However, other studies suggest that the virus may have originally evolved in West Africa (Swanepoel et al. 1993). This hypothesis is supported by the presence of three of the four serotypes and the historical occurrence of rabies viruses in dogs which are less virulent than conventional street viruses across this region (Blancou 1988). However, it may be that rabies was introduced to West Africa some time after ad 1500 by Europeans, since nucleotide sequencing has shown similarities between European, New World, and some West African isolates (Smith and Seidel 1993). Although it is unlikely to have been the first incidence of rabies in southern Africa, the first irrefutable diagnosis was made in 1893 during an epizootic involving dogs, cats, and domestic ruminants in the vicinity of Port Elizabeth on the south east coast of South Africa (Van Sittert 2003).

The introduction of modern molecular methods and surveillance mechanisms has led to a better understanding of the problems faced across Africa (Bouhry et al. 2008). Today, molecular tools have divided African isolates into different phylogenetic clusters based on similarities and differences at the nucleotide level. Three main African clades are now defined that include viruses that cluster closely with one another, either in line with geographical location or associated with the nature of the host from which the viral genomic material was derived. One cluster, referred to as ‘Africa 1’ (Kissi et al. 1995) or ‘Africa CD1’ (Nadin-Davis et al. 2002) includes viruses that seem adapted to both domestic and wild members of the Canidae that are found widely distributed throughout Africa. This virus lineage, however, contains two distinct clusters. Viruses from clade 1 or ‘Africa 1a’ have been found in Algeria, Ethiopia, Gabon, Madagascar, Morocco, Tunisia, and the Sudan (Johnson et al. 2004a); Clade 1b or Africa 1b contains isolates from the Central African Republic, Kenya, Mozambique, Namibia, Tanzania, Zaire, Zambia, Zimbabwe, and South Africa (Kissi et al. 1995; Sabeta et al. 2003; Mansfield et al. 2006a; Ngoepe et al. 2009).

Isolates from dog populations from West Africa form another clade, Africa 2, whilst another distinct clade, ‘Africa 3’ represents viruses that are circulating within mongoose populations in southern African countries including South Africa, Botswana and Zimbabwe (Johnson et al. 2004b). Apart from domestic dogs and cats, representatives of at least 30 different species belonging to all five families of carnivore native to southern Africa have been diagnosed with rabies (Thomson and Meredith 1993). Some species, notably the side striped jackal (Canis adustus), the yellow mongoose (Cynictis penicillata) (Nel and Rupprecht 2007), and the bat-eared fox (Otocyon megalotis) (Sabeta et al. 2007) seem to be maintaining different biotypes of the virus. Further investigation of these virus groups is needed to determine if the strong geographical segregation of these viruses truly reflects host restriction or whether this viral lineage may extend into other, as yet undefined, parts of Africa (Nel 1993; Nel and Rupprecht 2007).

Despite technological advances, the continued lack of adequately equipped regional laboratories and surveillance systems in Africa make the reporting of rabies cases, be it of human or animal occurrence, problematic. The proportion of rabies mortalities on a global scale that were reported to WHO in 1999 suggested that African cases of rabies made up only 3% of the worldwide incidence (Knobel et al. 2005). Continued political and economic instability across much of Africa has meant that this situation is showing few signs of improving and whilst some countries have improved capabilities, others still have no reporting or diagnostic capacity. With the domestic dog still considered to be the principal reservoir of rabies (WHO 2004), dog-associated rabies has continued to increase throughout sub-Saharan Africa. In this tropical zone, humans and animals are more widely distributed than in North Africa, and there has been a greater tendency for epidemics of dog rabies to spread over large areas and for the disease to be observed in both domestic and wild vertebrates (Blancou 1988; Nel and Rupprecht 2007). Estimates of the incidence of human rabies across Africa further suggest that the problem is much greater than previously thought. In urban and rural areas of Africa, current overall estimates of rabies incidence are 2.0 and 3.6 persons per 10⁵ persons, respectively. These values highlight a shortfall in the WHO estimates of 35,000 to 50,000 rabies human mortalities globally and underpin the need for better diagnostic capabilities, reporting and surveillance within some of the poorest regions of the world.

From a conservation standpoint, rabies virus currently threatens several endangered species across the African continent. As human habitation has expanded, canine rabies has spread to wildlife species and some already endangered canids, such as the African wild dogs (Lycaon Pictus), the Ethiopian wolf (Canis simensis) and spotted hyenas (Crocuta crocuta) (Gascoyne et al. 1993) act as vectors for the disease. Certainly, the existence of Ethiopian wolves has been put under threat from the introduction of the virus, probably through the fox population (Sillero-Zubiri et al. 1996; Randall et al. 2004; Knobel et al. 2008).

More recently, a number of lyssaviruses of African origin have been described. Isolates of Lagos Bat virus, classified within the lyssavirus species 2, have been recovered from various countries across the African continent (Hayman et al. 2008; Kuzmin et al. 2008). These viruses, although few in number, show a high degree of genetic diversity with isolates from different regions varying (Bourhy et al. 1993; Johnson et al. 2002b; Nadin-Davis et al. 2002; Markotter et al. 2008a). Members of species 3, the Mokola viruses (MOKVs) have also been detected in a number of African countries although the majority of isolates characterized so far have been from South Africa and Zimbabwe (Nel et al. 2000; Markotter et al. 2008b). One MOKV isolate from Nigeria, West Africa, was found to be, on the limited genetic data available, distinct from the remaining isolates (Nadin-Davis et al. 2002; Markotter et al. 2008b). As with the MOKVs, members of the lyssavirus species 4, Duvenhage virus (DUVV) isolates are generally reported from South Africa and Zimbabwe. Isolates have been made from both humans and bats (Amenugal et al. 1997; Badrane et al. 2001; Johnson et al. 2002b; Nadin-Davis et al. 2002).

The Americas are generally divided into three major geographical areas encompassing a total of 48 defined countries, island nations and territories. North America includes the US of America and Canada; Central America includes all the mainland countries as well as the island nations of the Caribbean; South America, or Latin America includes some of the most densely populated countries of the Americas, south of Central America.

The development of oral vaccine strategies to prevent sylvatic rabies in the mid 1960s and their application across much of North America greatly reduced the incidence of terrestrial rabies (Krebs et al. 2003). Prior to this rabies cases were rarely reported in Canada throughout the start of twentieth century until the 1940s when rabies in foxes spread into Canada causing an increase in incidence. Skunks have also played a role in the maintenance of rabies in Canada especially through the 1950s and 1960s. However, the result of extensive oral vaccination in the 1960s lead to a notable decrease in cases and over the last 15 years continued vaccination policies have meant that very few cases of rabies have been reported from Canada (Belotto et al. 2005).

A similar picture has been seen across the USA. Historically, the USA has a long association with rabies cases and between 1940 and 1950 over 100,000 cases were reported. Again, in the 1960s, mass vaccination campaigns and the use of oral vaccines hugely reduced the incidence of rabies in the canine population. Whilst the USA is now free of canine rabies, the incidence of rabies in wildlife species seems to be increasing across much of the country. Whilst the infection of wildlife species across the US remains a threat to the human population, surprisingly few human cases are reported annually with only four cases across the USA being reported between 2006 and the end of 2007 (Blanton et al. 2008).

The rabies situation in wildlife in the USA is complicated due to the presence of different terrestrial vector species as well as the presence of several bat species that harbour species 1 virus. Furthermore cases of transmission from bats to humans are often cryptic, with the interaction between infected animal and exposed individual often going unnoticed (Messenger et al. 2002). Most frequently reported terrestrial wildlife cases across the USA come from three different vector species; foxes, raccoons and skunks. Despite the complexity of the situation, these terrestrial species do seem to be limited in their distribution and so dispersal of virus variants may be described according to their vector species within geographically defined locations. To this end, raccoons (Procyon lotor) act as a major reservoir of rabies, predominantly across much of eastern US (Childs et al. 2001). Fox rabies exists within distinct fox populations; the red fox (Vulpes vulpes) and the grey fox (Urocyon cineroargenteus) and the Arctic fox (Alopex lagopus). Grey foxes maintain rabies within geographically limited populations in Arizona and Texas although oral vaccination campaigns have greatly reduced the number of fox cases reported from Texas (Sidwa et al. 2005). Red and Arctic fox populations maintain the virus in Alaska and can also cause incursions of disease in Canada although, again, oral vaccination programs have greatly reduced case numbers (MacInnes et al. 2001; Mansfield et al. 2006b). Rabies associated with skunks (Mephitis mephitis, Spilogale putorius) covers much of north central and south central USA with almost 1,500 reported cases in 2007 (Blanton et al. 2008).

Whilst rabies virus attributed to the infection of specific species is clearly defined, spillover into other wildlife species occurs but rarely initiates a new cycle of maintenance of virus within the new host. Transmission between animals of the same species maintains the levels of circulating virus and at the regional level may lead to continued circulation of a virus variant within a local population for decades (Blanton et al. 2008; Childs et al. 2001). The transmission of rabies virus from bats to both the human population and wildlife populations has also been reported. Historically, transmission has occurred from insectivorous bats such as the Yellow Bat (Lasiurus intermedius) and the Hoary bat (Lasiurus cinereus) (Brass 1994). The current distribution of bats species across much of the US as well as the ability for unchecked movement of bat populations makes tracking the virus difficult. Occasionally, spillover of bat variants strains of rabies occurs into wildlife species, most recently documented through the transmission from big brown bats (Eptesicus fuscus) to skunks in Arizona in 2001. This CST event resulted in rabies outbreaks within the skunk population in an area of Arizona that had been previously free of rabies virus infection (Leslie et al. 2006). Such transmission events rarely result in cycling of the bat-variant virus within the infected terrestrial population although one report suggests the probable circulation of a bat variant of rabies virus in foxes in Canada (Daoust et al. 1996). In this instance, it was speculated that some degree of intraspecific transmission passed the virus from bats into foxes and that adaptation of bat rabies to enable circulation within foxes may occur (Daoust et al. 1996).

Rabies remains a serious economic problem as well as being a threat to human health across much of Central and South America. In recent years there has been a sharp decrease in the number of canine cases, as a result of several mass vaccination campaigns implemented across several of the worse affected countries. Furthermore, perhaps directly linked to the reduction seen in cases of canine rabies, the number of human cases has also decreased notably although an increase in availability and application of PEP must also have assisted in reducing human fatalities. Much of the successes seen in the reduction of cases has come since the formation of the ‘Plan of Action for the Elimination of Urban Rabies from the Principal Cities of Latin America’ organized by the Pan American Health Organization (PAHO). As a result, over the past 20 years an almost 90% decrease in the number of dog and human cases has been observed across the regions most greatly affected, mainly in the Spanish speaking Caribbean countries and across Latin America. In Mexico, mass parenteral vaccination of dogs over a fifteen year period resulted in more than 150 million vaccine doses being administered between 1990 and 2005. The result of this targeted campaign using modern cell culture based vaccines has been the elimination of human rabies and a 43 fold reduction in canine rabies cases across the country (Lucas et al. 2008). Rabies from wildlife reservoirs continues, however, to be a problem across much of Central and South America with a number of species being implicated in transmission but with the greatest number of reports involving infections of wildlife such as raccoon rabies, skunk rabies and cases involving cattle and bats (Fig. 35.4). From these data, however, it is worth emphasizing that within certain geographical regions cases linked to different wildlife species predominate with high numbers of cases of rabies in cattle and horses being reported in Brazil; cases of rabies involving skunks, raccoons and bat species are mainly reported from the USA and the problem of rabies in vampire bats being restricted to Central and South America (Belotto et al. 2005).

Historically, attempts to control rabies where outbreaks occur have been through slaughter, quarantine and mass vaccination of infected animals. Although such actions are still practiced, more modern approaches involving rapid identification of the source of infection, establishment of containment zones where possible and the application of oral vaccination strategies to combat further possible cases are implemented instead. In all cases a rapid, strategic plan must be implemented to minimize the potential of further spread of disease within wildlife and the likelihood of transmission to the human population. In island nations, rabies-free status is relatively easy to maintain through strict importation and quarantine programs but these must be successfully enforced to prevent reintroduction of the virus. Quarantine allows the animals to be assessed over a period of time within which they may develop obvious symptoms of disease. A number of island nations have successfully used quarantine to maintain a rabies-free status including Australia, New Zealand, Hawaii, Sweden, and the UK. However, recent events in a number of these countries have tested the effectiveness of the quarantine procedures in place but have successfully prevented reintroduction of the virus. Recently, in the UK, a puppy imported from Sri Lanka was placed into routine rabies quarantine and developed clinical signs of rabies, again highlighting the value of the animal control systems in place (Fooks et al. 2008). As well as preventing the disease from entering its own animal population, quarantine has an important role in the prevention of spread of rabies to other countries.

Controlling dog rabies across areas where the virus remains endemic is a major challenge in the global fight against the disease. The control of rabies within the domestic dog population is something that, within the developed world, has led to a dramatic reduction in the number of human cases reported. However, across much of the developing world, domestic dogs continue to act as the main reservoir for infection (Rupprecht et al. 2002) with transmission from wildlife into domestic dog populations often linking the virus circulating within wildlife populations to the human population. Regions where the human population lives in close proximity to populations of potentially rabid dogs would benefit from targeted vaccination campaigns. Unfortunately, the situation regarding dog rabies in such areas seems to be continuing to spiral out of control with a dramatic increase in the number of dog rabies cases being reported over the past few decades (Cleaveland 1998). A number of factors compound this situation including: the unchecked increase in dog populations in these areas; the high fluidity of human and dog movement; the lack of economic infrastructure that in turn effects the availability of both medical and veterinary services equipped to deal with rabies cases; poor public awareness through lack of educational facilities that leave much of the population unaware of simple PEP and a shortage of trained staff that can administer vaccine and PEP even where it is available.

As demonstrated across much of the developed world, the problem of dog rabies can be overcome through mass parenteral vaccination of owned dogs. However, across the developing world an inability to put mass vaccination strategies into practice often means that even where local vaccination campaigns are enforced, a temporary reduction in cases seen but the virus is soon reintroduced and the cycle of infection re-initiates. Several different studies have suggested that the level of vaccine coverage within a domestic dog population required to substantially reduce the presence of the disease range from 60 to 87% (Bogel and Joshi 1990; Perry 1995; Kayali et al. 2003; Kamoltham et al. 2003). Such high levels of vaccination coverage are immensely difficult to achieve as many dogs in such areas are considered ‘community dogs’ that wander between populated areas, occasionally fed waste food by different groups or individuals or are stray dogs that are not owned within a community and are often totally feral. However, where such target vaccination levels are achieved cases within populations have been shown to decrease dramatically (Cleaveland et al. 2003) with a sharp reduction in the number of human cases seen having a concomitant effect on the demand for PEP. In one province of Thailand, application of awareness campaigns in concert with vaccination and administration of intradermal PEP has completely eliminated human rabies deaths over a five year period (Kamoltham et al. 2003).

Application of oral vaccination campaigns to combat wildlife rabies across much of North America and Western Europe has suggested a role in such elimination methods where ‘community’ dog populations are largely inaccessible and/or parenteral vaccination campaigns have been unsuccessful. Several types of vaccines are licensed for delivery in this manner and bait formulations have been designed to appeal to different species. Strategies to apply such oral vaccines vary with different bait formulations being attempted in different canine populations (Bishop et al. 2001; Harischandra et al. 2001; Hu et al. 2008b).

Unfortunately, whilst estimates suggest that if dog rabies is controlled in these areas the economic burden associated with the disease, especially through transmission to the human population, would decrease dramatically, the financial resources and economic infrastructure are generally lacking preventing the administration the type of vaccination campaigns needed to remove the threat of disease (Fishbein 1991; Meslin 1994; Kaare et al. 2007). Whilst the necessary tools to combat rabies within domestic dog populations exist, the application of these tools to the developing world continues to be a major challenge.

In much the same way that canine rabies constitutes a huge economic burden across the developing world, rabies virus in wildlife populations is still problematic in much of the developed world. In North America, several different vectors are able to maintain the disease and this means that whilst much of North America is ‘free’ of canine rabies, the threat of spillover into the dog and human populations remain high. Attempts to control this situation using novel vaccination techniques have been partially successful in reducing the incidence of disease. However, in one species, limited success may be achieved while in another it has been totally unsuccessful. For example, oral vaccination using specific bait delivery systems to combat rabies within the raccoon population were applied to the skunk population with little success (Olson et al. 2000; Russell et al. 2005; Hanlon et al. 2002). Attempts to vaccinate raccoons with recombinant poxviruses virus expressing the rabies glycoprotein has also been problematic. In this instance, pre-existing antibodies generated through natural infection with raccoonpox affected the immunostimulatory potential of the vaccine preparation (Root et al. 2008). Often programs such as ‘trap-vaccinate-release’ have been applied successfully to contain rabies in the skunk population as well as in outbreak situations (Rosatte et al. 2001; Engeman et al. 2003). In addition to the successes seen through the use of inactivated oral vaccines, several live attenuated vaccines have been developed for oral application successfully used in small scale trials with wildlife populations (Vos et al. 2002; Follman et al. 2004).

Whilst across all of the Americas terrestrial rabies is the main problem, a further threat exists in the transmission of virus from the bat population. In North America, rabies in insectivorous bats has been responsible for many human cases (Gibbons et al. 2002), whilst across Central and South America the threat from bat rabies is most notably associated with vampire bats and the resulting economic losses through infection of cattle (Blancou and Fooks 2008). Many different strategies have been used in an attempt to control rabies in the vampire bat populations, including early methods to destroy bat populations with smoke, fire and explosives; capture and treatment with anticoagulants such as warfarin (Linhart et al. 1972; Thompson et al. 1972). As well as applying warfarin jelly to the backs of captured bats and to previous vampire bat wounds on cattle, warfarin injected into cattle will remain in the animals blood for several days, effectively killing any bats that feed on the treated animal. This systemic treatment of cattle with anticoagulant preparations is claimed to be very effective, reducing biting rates on cattle by 85–90% (Flores-Crespo and Arellano-Sota 1991). More recently, oral vaccine preparations have been trialed in bat populations with limited success. The cost effectiveness of trying to vaccinate bat populations has, however, been questioned and as vampire bat populations continue to increase, a more destructive mechanism to reduce population sizes may be the best way to control the situation (Almeida et al. 2005).

The lack of human and financial resources available to many African countries leads to an inability to put vaccination strategies into practice. Furthermore, the continued problem of canine rabies require that co-ordinated efforts to control rabies within domestic populations be made but this rarely occurs. Exceptions to this, however, include conservation efforts to protect endangered species such as the African wild dogs (Lycaon pictus) from potential extinction due to the introduction of the disease into the population in the early 1990s (Gascoyne et al. 1993) and more recently an oral vaccine preparation has been developed to attempt to vaccinate free ranging wild dogs (Knobel et al. 2002; Knobel and Toit 2003; Knobel et al. 2008). Since then, the introduction of the virus into the Ethiopian wolf (Canis simensis) population in 2003 led to a internationally co-ordinated effort to save these canids from the brink of extinction with live capture and vaccination programs being implemented (Knobel et al. 2008).

It is universally accepted that wildlife rabies plays a comparatively minor role in disease maintenance across much of Asia with canine rabies remaining the major threat to human health (Wilde et al. 2005). Several oral vaccine preparations are available and are in use to combat rabies where it remains a problem and where wildlife vectors such as fox species and in the Middle East, jackal species, exist. Oral vaccination campaigns have recently been applied with some success across parts of the Middle East including Egypt, Israel, Jordan and Palestine (Yakobson et al. 2006; Rosatte et al. 2007). Incidence of wildlife rabies in the Middle East is often only reduced by culling stray dog populations as well as those shown to be rabid (Seimenis 2008; Vos 2004). Central Asia has a much greater incidence of wildlife rabies where several species including arctic foxes, raccoon dogs and red foxes are thought to act as vectors for the disease. Reporting cases of disease across the majority of the continent, however, remains at a very low level despite the implementation of rabies awareness campaigns. Spread of the disease seems to be confined to areas through the presence of wild vector species and limited by geological boundaries rather than by human attempts to curb the incidence of disease. Exact population densities of vector species across many Asian countries are unknown although the implementation of domestic dog vaccination in some areas has helped the situation. Russia is the only country of Central Asia that is attempting to vaccinate wild carnivore populations and, using lessons learned from global elimination campaigns, have started to use oral vaccines to reduce incidence of the disease (Gruzdev 2008).

Countries of Eastern Asia also have rabies present within wildlife species with raccoon dog populations maintaining the disease in South Korea and serological evidence existing for lyssaviruses within bat populations across several Far East Asian countries such as Cambodia (Reynes et al. 2004), China (Tang et al. 2005), Thailand (Lumlertdacha et al. 2005) and the Phillippines (Arguin et al. 2002).

The current status of rabies throughout the EU has vastly improved over the past decade. Successful applications of oral vaccination campaigns have now removed the threat of terrestrial rabies from much of Western Europe. The principal reservoirs of disease, red foxes and raccoon dogs have been targeted in a number of campaigns that have eliminated the virus from a number of European countries (Steck and Wandeler 1980). However, continued expansion of the EU to the East has generated new challenges and the virus remains hugely problematic across Eastern Europe and within the Baltic states. Fox and raccoon dog population increases have enhanced the problem (Zienius et al. 2003). Many issues remain to be solved with current oral vaccination campaigns. Bait formulations and approach to bait drops need modernizing to optimize efficiency of delivery to target species and timing of drops can profoundly affect uptake by different target species depending on breeding life cycles (Vos 2003). Innovative ideas on how to find pragmatic solutions to these problems are needed if successful elimination of the virus from the EU territory is to be witnessed in the near future.

One of the biggest obstacles to successfully reducing the number of deaths attributed to rabies virus is in creating awareness of this vaccine-preventable disease. The death toll in children and young adults in areas where the virus is endemic remains high. In such areas, educating children of school age as well as informing local populations about the need to exercise extreme caution when in the presence of an animal behaving in an unusual or aggressive manner and careful wound care would greatly reduce the number deaths attributed to rabies infection. While many governments have neither the resources nor the infrastructure to provide PEP to local populations, better education and an awareness of the dangers from rabies within societies will help reduce incidence. Effective disease reporting in areas where little is known about the presence of the disease, especially in wildlife species would also be of huge value to elimination campaigns (Rupprecht et al. 2008).

With rabies now often falling under the banner of a ‘neglected tropical disease’ partnerships are being formed across the globe to try and shift the world’s attention back to this devastating disease. International collaborations between expert groups have generated new focus groups such as the Partnership for Rabies Prevention (PRP) and the Alliance for Rabies Control (ARC) to try and combat the virus in areas where it remains endemic and some of the world’s largest funding bodies have now recognized the need for a push towards elimination of the virus wherever possible.