Comparative assessment of seller’s staining test (SST) and direct fluorescent antibody test for rapid and accurate laboratory diagnosis of rabies. assessment of test (SST) and direct antibody test for rapid and accurate laboratory

Background: Rabies causes 55, 000 annual human deaths globally and about 10,000 people are exposed annually in Nigeria. Diagnosis of animal rabies in most African countries has been by direct microscopic examination. In Nigeria, the Seller’s stain test (SST) was employed until 2009. Before then, both SST and dFAT were used concurrently until the dFAT became the only standard method. Objective: This study was designed to assess the sensitivity and specificity of the SST in relation to the ‘gold standard’ dFAT in diagnosis of rabies in Nigeria. Methods: A total of 88 animal specimens submitted to the Rabies National Reference Laboratory, Nigeria were routinely tested for rabies by SST and dFAT. Results: Overall, 65.9% of the specimens were positive for rabies by SST, while 81.8% were positive by dFAT. The sensitivity of SST in relation to the gold standard dFAT was 81.0% (95% CIs; 69.7% - 88.6%), while the specificity was 100% (95% CIs; 76% - 100%). Conclusion: The relatively low sensitivity of the SST observed in this study calls for its replacement with the dFAT for accurate diagnosis of rabies and timely decisions on administration of PEP to prevent untimely deaths of exposed humans. Comparative staining fluorescent


Introduction
Rabies is a fatal zoonotic disease of mammals caused by a highly neurotropic virus belonging to the Lyssavirus genus, in the family Rhabdoviridae 1 . The viral agent is transmitted to humans mainly through bites and scratches from infected animals 2 . Rabies is present on most continents, causing at least 55, 000 human deaths per year and is endemic in most African and Asian countries 3 where most human deaths occur 4 .
Rabies is endemic in Nigeria with the domestic dog as the major vector of the causative agent 5,6 . Since its first occurrence in the country in humans in 1912 and first laboratory confirmed diagnosis in a dog in 1925 7 , hu-man and animal rabies cases have been recorded in all the regions and ecological zones of Nigeria annually, 8 . For instance, in 1942, rabies was reported from Kano Zaria Borno and Ilorin 6,8,9 and from Plateau, Makurdi and Enugu in 1946 10,11 . Serological evidence of Lagos bat virus circulation among Nigerian fruit bats 13 , and rabies virus antigens in apparently healthy dogs in Yola, Nigeria 14 were also reported.
According to Nawathe 15 , no fewer than 10,000 Nigerians are exposed to rabies annually, and about 1, 000 annual cattle mortality had been reported. In Nigeria, human and animal rabies cases are on the increase annually despite the availability of vaccines for its control and prevention 12 .
Although majority of the Veterinary Laboratories in Africa have adequate personnel capacity to diagnose rabies in animals, routine diagnosis is often constrained by lack of laboratory equipment and reliable reagents 16,17 . For instance, the direct microscopic examination by the SST, which due to its low sensitivity and reliability 18,19 , had long been substituted with the more sen-sitive, specific and reliable direct fluorescent antibody test (dFAT) in most rabies laboratories globally, had not been employed in the diagnosis of rabies in Nigeria, until 2009. Although the dFAT was introduced in the laboratory since 2005 20 , its full implementation only began in September 2009 when the two year (2010 to 2011) "south-south twinning" agreement between the OIE reference laboratory for rabies, Onderstepoort Veterinary Institute, Pretoria, South Africa and NVRI, Vom, Nigeria, was signed. Before the full implementation of the dFAT however, there had been 8 months (March to October, 2009) of transition, during which both techniques were concurrently utilized in routine diagnosis of rabies. This study was therefore designed to compare the sensitivity and specificity of the SST with the 'gold standard' dFAT in routine diagnosis of rabies in animals in the rabies National Reference laboratory, Nigeria.

Materials and methods Data collection
A total of 88 animal brain specimens were received for routine diagnosis of rabies in Central Diagnostic Laboratory, NVRI, Vom, Nigeria, from March to October 2009. Specimen submissions were from rabies-suspect animals and from animal carcases received for routine post mortem examination in the laboratory. Touch impression smears of the hippocampus on slides were fixed in chilled acetone and tested for the presence of rabies virus antigens by the dFAT as described by Dean et al. 21 ; using rabies fluorescein isothiocyanate conjugated (FITC) Anti-Rabies Monoclonal Globulin (FDI FU-JIREBIO Diagnostic, Inc.), according to manufacturer's instructions. Slides were examined under royal blue LED (455nm) excitation filter at X10 eye piece and X20 objective lenses using Fluorescence Microscope (Axioskop G/115-230).
Smears of the hippocampus were also made on another set of slides and stained with Sellers stain as described by Tierkel and Atanasiu 19 . The stained smears were examined under oil immersion (X100) magnification using a light microscope (Olympus CH) for presence of eosinophilic stained rabies virus particles inclusions (Negribodies) in the cytoplasm the brain cells.

Data analysis
Data collected was compiled, processed and analysed. The sensitivity, specificity and predictive values of the SST were determined using the VassarStats (http:// www.vassarstats.net/clin1.html) and by Chi-square methods as described by Akobeng 22 .

Discussion
In 72 of the 88 specimens tested by dFAT, 81.8% where positive for rabies while oval or round magenta red coloured Negri bodies were seen in only 58 (65.9%) of specimens stained with seller's stain. This indicated that majority of the animals were infected with rabies virus. However, the results showed that only 58/72 (80.5%) of the rabies positive specimens by dFAT showed presence of Negribodies in SST (Table 1, Figure 1). The lower positivity of the SST (65.9%) compared to the dFAT (81.8 %) (Figure 1) and the ability of the dFAT to detect rabies antigen in 14 of the 30 specimens that tested negative by the SST (Table 1) is in harmony with the lower sensitivity and reliability of the SST technique 23 compared to the 'gold standard' test. Negri bodies seen in 65.9% of the specimens by SST in this study were comparable with the 50-60% reported earlier 24 .
The sensitivity of SST in relation to the "gold standard" dFAT was 81.0% (95% CIs; 69.7% -88.6%), while the specificity was 100% (95% CIs; 76% -100%). This showed that SST missed rabies virus antigen in some positive samples in the study resulting in false negative results. However, the high specificity of the SST means there were no false positive results from samples that tested positive by the SST. Therefore, a positive result guarantees that the sample is truly positive while negative result does not. This further implies that, the Sell-er's method is only most useful to the clinician when the test result of rabies suspect animal is positive, but doubtful when result is negative.
Given that rabies is a deadly disease with 100% case fatality rate once the virus has entered the central nervous system (CNS), a highly sensitive, rapid, reliable and standardized technique is desirable for its accurate diagnosis. The dFAT has been reported to be 98-100% sensitive 23 and 99.6% specific while the SST has 90.5% sensitive and 98.8% specific 25 . This showed that although Negri bodies are specific for rabies virus infection, SST is less sensitive and specific as it only detects affinity of Negri bodies for acidophilic stains 26 . It has also been shown that Negri bodies are not developed in all infected individuals 26 and cannot be demonstrated by the SST in all animals from which the virus can be isolated 27 . Consequently, false negative results may occur in some infected materials and false positive results may occur if nonspecific inclusion bodies are present in tissue samples tested 26 . On the contrary, the dFAT has a high degree of sensitivity and specificity 28 and provides a reliable diagnosis in 98-100% cases 21 . With sufficient care and experience, fluorescent antibody techniques can be used with considerable speed and accuracy for the determination of the presence of rabies virus in the salivary glands of animals 27 .
Therefore, it is recommended to routinely use dFAT for rapid and accurate diagnosis of rabies in rabies-suspect animals in order to take right and timely decision on administration of PEP to exposed humans. An accurate laboratory diagnostic result obtained by dFAT within few hours [1-2 hours] could save the victim from avoidable death if the animal is rabid and the full course of PEP is administered to the victim within the recommended time. An accurate result would also save persons exposed to rabies from trauma and financial burden associated if the animal is not rabid 23 .

Conclusion
Although the SST has the advantages of yielding results under one hour, inexpensive laboratory equipment needed to perform the test and the need to keep specimens cold after fixation is avoided, it can no longer be recommended for rabies routine diagnosis because it has comparatively low sensitivity and should be replaced with the dFAT 24 , wherever it is still being employed in rabies routine diagnosis.