The domestic dog is the most significant reservoir for rabies virus and dog bites are responsible for the great majority of human deaths from rabies [1]. Although rabies in dogs has been controlled and eliminated in many regions of the world, it still persists throughout Africa and Asia causing thousands of human deaths every year. Despite the availability of effective vaccination, it is only through control of dog rabies that human deaths due to rabies will be reduced [2]. The United Kingdom (UK) has been rabies free since 1922 [3]. The gradual reduction of rabies throughout western Europe, where sylvatic rabies had been endemic since the late 1960s, has further reduced the risk of rabies entering the country from mainland Europe. A similar situation now exists in much of Europe where the introduction of oral-vaccination campaigns has been highly effective at eliminating fox rabies [4]. However, the danger of importation is always present and there have been a number of well documented cases of rabies introduction, particularly through movement of dogs from Morocco into Spain and France [5]. For the UK, two methods of preventing rabies entry have been in use over the past decade. The first has been quarantine of companion animals and exotic species, particularly bats and carnivores. Figures for the numbers of dogs and cats entering quarantine between 1999 and 2005 are shown in Table 1. A decline in the number of animals entering quarantine was observed as a result of the introduction in 2000 of the second method of controlling entry of companion animals, the UK Pet Travel Scheme (PETS). This enabled immediate entry for dogs, cats and ferrets provided the animal complies with the requirements of the scheme, principally demonstration of vaccination against rabies. This scheme has proven highly popular with companion animal owners with approximately 750,000 animals entering the UK by this route between 2000 and 2010 (Table 2). Both these mechanisms of companion animal entry into the UK will be harmonized with those of other European Union member states at the start of 2012 [6].

Meldrum (1988) [7] reported 29 cases of rabies in quarantine animals up to 1988. On two occasions, cases of rabies have resulted in dogs after the six-month quarantine period [8] and as a consequence compulsory vaccination was introduced in 1974 for all dogs and cats entering quarantine. In both cases, the incursions were dealt with swiftly and no further incidence of the disease was reported. To date, the UK has never resorted to the use of vaccination to control the disease in wildlife although this contingency would be considered by the Department for Environment, Food and Rural Affairs (Defra), the government department responsible for animal disease control. Prior to 2008, the most recent rabies associated death in quarantine occurred in a dog imported from Zambia in 1990. There is a legal obligation to submit the head of any animal that dies in quarantine for rabies diagnosis. In 2008, there were 29 deaths in UK quarantine.

Two preliminary reports have previously described the clinical history of the case, rabies diagnosis [9] and the public health investigation that occurred [10]. Briefly, the dog was imported from Sri Lanka to the UK by an animal welfare charity. The dog arrived at Heathrow airport on the 17 April 2008, and spent one night at the Animal Reception Centre before being transferred to quarantine kennels in north-east London. The dog had a history of severe diarrhoea and convulsed violently before it died on the morning of the 25 April. The initial diagnosis was for infection with Hepatozoon canis for which it had been treated with doxycycline. Confirmation of rabies infection was made within 24 hours of the death of the dog. A further four dogs that had contact with the dog underwent euthanasia as a precautionary measure. Eleven individuals in the UK received post-exposure prophylaxis (PEP) following direct physical contact with the dog.

This study confirms the epidemiological link between this case of rabies and the epidemic of dog rabies in Sri Lanka. We also report the histological findings in brain sections of this animal and note the absence of Negri bodies in this particular case.

Rabies virus infection was confirmed by the fluorescent antibody test and was corroborated by amplification of a genotype 1 lyssavirus by a differential reverse transcriptase-Polymerase Chain Reaction (RT-PCR) TaqMan assay from RNA extracted from the brain. Rabies virus was isolated from brain homogenates prepared directly from the animal and 100% of mice inoculated with this homogenate developed rabies encephalitis on day nine post-inoculation. A nested RT-PCR amplified a 606 bp fragment of the rabies nucleoprotein gene and this was used to derive 400 bp of genomic sequence from the infecting virus. Sequence alignment with a panel of isolates from Asia, and particularly from published sequences from Sri Lanka (Table 3), revealed that the RABV isolated from the dog had >99% sequence identity with RABV isolates from Sri Lanka. The phylogenetic tree (Figure 1 and inset) demonstrated that the isolated virus sequence (JN968375) clustered closely with a lineage of viruses isolated from India and the south-western provinces of Sri Lanka. This supports the evidence provided in the case history.

The presentation of brain sections revealed a mild non-suppurative encephalitis and leptomeningitis consistent with viral encephalitis. The characteristic changes were observed in all regions of the brain examined, including hippocampus, brainstem and cerebellum. They consisted of neuronal degeneration and necrosis, with occasional neuronal vacuolation, shrunken hyperchromatic neurons and images of central chromatolysis and nuclear margination, satellitosis, focal gliosis and lymphohistiocytic perivascular cuffing of no more than one cell of thickness [Figure 2(B)]. Inflammatory changes were more prominent in the brainstem. Negri bodies were not observed in any of the samples examined.

Immunolabelling against rabies nucleocapsid was detected in the perikaryon and neuropil of neurons in all examined regions [Figures 2(C) and 3]. No staining was observed in the corresponding regions of a normal dog brain used as a control [Figures 2(A) and 3(A)]. The region with a higher number of immunolabelled neurons and with a more prominent immunolabelling in their perikaryon was the brainstem (Figure 2). Immunolabelling in the hippocampus was observed in a small number of pyramidal neurons [Figure 3(D)]. Detection of viral antigen in the cerebellum was occasional in the Purkinje cell layer [Figure 3(B)] and more frequent in the cells of the granular and molecular layer.

Quarantine of dogs entering the UK has been an effective measure for preventing the entry of rabies since the beginning of the 20th century. Despite the introduction of the UK and EU PETS, which caused a noticeable drop in the number of animals entering quarantine, this has been an option for those wishing to import dogs into the country with over 3,000 entries in 2005 (Table 2). Despite the introduction of vaccination either in the country of origin or upon entry into quarantine, the risk still exists that an animal may have become infected during the period prior to entering the UK and that infection has advanced to a point where vaccination becomes ineffective. A further factor that may reduce the effectiveness of vaccination is the age of the dog [24]. Two recent cases of rabies in juvenile dogs in North America have highlighted this issue [25,26]. Vaccine manufacturers do not recommend rabies vaccination of dogs before twelve weeks of age due to possible presence of maternal antibodies that may inhibit the immune response [27]. This provides a window of susceptibility for those young dogs that do not have maternal antibody protection, as is likely in the case described here. Rabies is endemic in Sri Lanka with 68 cases of human rabies and 705 cases of dog rabies reported in 2006 [28]. However, these figures may be higher due to under reporting [29]. This has been in the face of large scale dog vaccination campaigns with approximately 1 million dogs vaccinated in 2006. A phylogenetic comparison of the sequence derived from the UK case with a panel of RABV sequences previously published by studies on the phylogeny of RABV from Sri Lanka [17,18] demonstrated a clear link between the case in quarantine and RABV circulating in Sri Lanka. These earlier studies have shown that the RABVs endemic in Sri Lanka are closely related to the viruses present in southern India. The Sri Lankan RABV isolated in the UK showed almost 100% sequence homology with some members of the panel and suggests that the dog was infected in the south-western provinces of Sri Lanka.

Although the disease symptoms observed in this animal were not typical of a rabies infected dog, investigation of samples by the FAT test demonstrated that the brain was heavily infected with RABV. This was confirmed by immunohistochemistry, which revealed widespread labelling of rabies nucleoprotein in the brainstem/medulla, hippocampus and, to a lesser extent, the cerebellum. The histological changes were consistent with viral encephalitis displaying characteristic degenerative and inflammatory changes. The nature and distribution of the lesions in this case were similar to those reported in a case of rabies in a 10-week-old puppy [25]. However, in the UK case and distinct from that observed in the Canadian dog, there was a complete absence of Negri bodies in all samples examined despite this being considered a common histological sign of rabies. This reflects the observation that up to 30% of street rabies cases do not produce Negri bodies [30,31] and emphasizes the requirement for virus-specific primary diagnostic tests such as the FAT and RT-PCR in cases of viral encephalitis of undetermined origin. The relatively high detection of viral antigen in the medulla/brainstem may be a consequence of the peripheral infection, as the medulla acts as crossing of motor tracts between the spinal cord and the brain. This in turn suggests that the brainstem should also be considered as a diagnostic sample when investigating animal cases suspected of rabies.

In March 2003 Defra launched DACTARI (Dog and Cat Travel and Risk Information). This is a national voluntary scheme for the investigation of the possible occurrence of exotic diseases in dogs and cats in Great Britain such as leishmaniasis, babesiosis, ehrlichiosis and dirofilariasis (of which leishmaniasis and ehrliciosis are zoonotic). Since 2003, there have been 100 reports of exotic diseases. Sixty-seven percent of which were detected in animals imported under PETS and 25% were detected in animals within quarantine.

Rabies virus is endemic in domestic dog populations in many regions of the world and as a result of movement of companion animals between countries the risk persists of transfer of infected animals between different areas, resulting in new outbreaks. Within rabies-free areas of the world, the consequences of such introductions are varied and range from a need to control the initial outbreak to investment in surveillance to demonstrate freedom from disease. For zoonotic viruses, there is a risk to public and veterinary health, which for rabies would result in human and animal deaths. Compulsory vaccination of dogs and restrictions on dog movement would also be required. In the case described in this study, no onward transmission occurred and a potential outbreak was controlled. However, the case demonstrates the need for constant vigilance by public health bodies in detecting incursions of potentially rabies infected animals and providing a rapid response to protect human contacts from this fatal disease.