Rabies is an infectious viral disease that is almost always fatal following the onset of clinical symptoms. The disease causes about 59,000 human deaths annually worldwide, most of them being in Asia and Africa, particularly in resource-constrained countries [1
]. Dog bites account for almost the entire incidence of human rabies, whereas rabies in animals could be attributed to a sylvatic cycle between wild as well as feral canines and other carnivores. Rabies is preventable, and domestic canine rabies has been eradicated or controlled in several countries by vaccination; however, sylvatic rabies still presents a danger in these countries [2
]. Given the importance of the spread of rabies to humans by canids and other carnivores, the protracted incubation period, the often delayed appearance of typical symptoms, rapid and accurate diagnosis of rabies in animals is critical for prognostication, and for initiating and implementing post-exposure prophylaxis, infection-control strategies and public health measures.
Since the recognition in the early 20th century of Negri bodies as being the pathognomonic histopathological lesion in brain or spinal-cord sections, an array of immunoassays and molecular techniques have been developed for the laboratory diagnosis of rabies [5
]. Whereas Seller’s technique to visualize intracytoplasmic virus-inclusion body aggregates (Negri bodies) is simple and rapid, it is suitable only for fresh specimens. Paraffin-embedded tissue samples can be used for staining, but the method is time-consuming, and like Seller’s staining, lacks sensitivity. Comparatively, the detection of virus antigen or nucleic acid is both more sensitive and rapid. Antigen can be detected by various immunoassays such as the fluorescent antibody technique (FAT), enzyme-linked immunosorbent assay, immunochemistry (e.g., direct rapid immunochemical test or dRIT, indirect rapid immunochemistry test or IRIT), or immunoblot (immunochromatography, dot-blot). Among them, the direct fluorescent antibody (DFA) test is the gold standard [9
]. This test detects the presence of rabies virus antigen in infected tissues, particularly in the brain. However, the higher cost involved in fluorescent microscopy, the requirement for specialized training, and its unsuitability for highly decomposed samples limit the wide usage of DFA in resource-limited countries. On the other hand, enzyme immunoassays, such as rapid rabies enzyme immunodiagnosis (RREID), are not only as sensitive and specific as FAT, but also can be applied to partially decomposed samples; however, brain tissues need to be homogenized, resulting in a potential exposure hazard to laboratory personnel. Of late, dRIT has been increasingly employed for the laboratory diagnosis of rabies, owing to its simplicity [11
]. Here, tissue smears are fixed, reducing exposure hazard, and the result can be read by using a simple microscope within an hour. Although dRIT can perform as well as DFA, it cannot also be applied to samples in advanced stages of decomposition, a common occurrence in developing countries where maintaining the samples under a cold chain during transportation is a challenge. The latter obstacle can be overcome by the application of nucleic acid-detection techniques such as reverse transcription, followed by polymerase chain reaction (RT-PCR), nucleic acid sequence-based amplification (NASBA), or loop-mediated isothermal amplification (LAMP), which not only can be applied to decomposed samples, but are also more sensitive and specific than DFA and/or dRIT [14
]; although the lack of standardized protocols and higher percentage of false positives are their disadvantages.
In this study, we compared the application of DFA, dRIT and RT-PCR for confirmatory diagnosis of rabies in suspected brain samples of animals resourced from different parts of India. Two hundred and fifty seven freshly collected and transported, or frozen (−80 °C) archived samples from cattle, buffalo, horse, goat, pig, dog, cat, jackal, leopard and wolf subjects were subjected to all the tests. We observed 100% concordance between DFA and dRIT, and also showed the utility of RT-PCR in detecting viral nucleic acid in a further 35 samples that were unsatisfactory for testing by both DFA and dRIT. The results support the utility of dRIT as a simple test that can be adopted to field conditions, and contribute to the epidemiology of rabies in India.
3. Results and Discussion
Despite an estimated 35% of all the human rabies deaths worldwide occurring in the country [21
], the disease is not notifiable in India. The lack of reporting is compounded by fear of touching cadavers, constraints in transporting the samples, and the availability of a limited number of laboratories capable of carrying out definitive diagnostic tests. As elsewhere, the major route of transmission of rabies virus to humans are dog bites, which constitute 91.5% of all animal bites in India [21
]. More than 96% of rabies cases in India are the result of contact with infected dogs [23
]. It has been estimated that India has one dog for every 36 persons, and the majority of these dogs are free=ranging or feral [23
]. In addition, rabies has also been reported to have been contracted through contact with infected jackals, cats, monkeys, mongooses and foxes [24
]. Thus, rabies in animals is not only a major concern for India, but also presents an opportunity for rapid action on post-exposure prophylaxis for humans, livestock and pets if it can be diagnosed quickly and easily. However, capacity-building in implementing validated or well established diagnostic tests and in instituting referral laboratories has been slow. In this context, an OIE twinning programme has recently been initiated at our rabies laboratory in Bengulugu. The work described here is part of a rabies diagnosis programme supported by Crucell and led to the development of the twinning programme [25
DFA is the test of choice for the laboratory confirmation of rabies [26
]. In the current study, positive samples showed bright green fluorescent foci of varying size scattered within the smear, sometimes being clearly visible within neurons (Figure 1
, left panels), mirroring similar descriptions by others [11
]. However, DFA has several drawbacks such as the need for an expensive fluorescent microscope, well-trained personnel, and quality controlled reagents (antibodies, conjugates), and varied parameters used during microscopy, and incubation times and temperatures, not to mention the subjectivity in interpretation of the test results [27
]. In addition, acetone used as fixative in DFA does not completely inactivate the virus, as demonstrated by the infectivity of acetone-fixed tissue for neuroblastoma cells [31
], posing a potential biohazard to laboratory personnel. Indeed, complete inactivation of cell culture-derived rabies virus appears to require >30% acetone [32
Some of the limitations of DFA can be overcome by dRIT. Whilst excellent concordance between dRIT and DFA has been observed with freshly prepared samples, dRIT could perform better than DFA with frozen or fixed samples [11
]. Positive results with dRIT can be declared by the presence of dark red- to brown-colored deposits scattered throughout the impression (Figure 1
, right panels), as has been shown earlier [34
]. Immunohistochemistry tests have been found to be as reliable as DFA for confirming rabies using tissues obtained from various animal species as well as those inoculated experimentally, even when the tissues had been stored frozen for various lengths of time and/or fixed [36
]. Another advantage of dRIT is the use of formalin for fixing the tissue smears. Titres of cell culture-derived virus have been shown to be reduced by three orders of magnitude with 3–4% formaldehyde treated for 30 min [32
], and complete inactivation can be achieved with 10% formaldehyde treatment for as little as 11 min [38
], although it is arguable that cell culture-derived and tissue-embedded virus could be differentially affected by the same treatment.
The dRIT typically uses monoclonal antibodies (MAbs) to detect rabies virus antigen. However, it is possible that there could be slight variation in the amino acid sequence of the N protein targeted by these MAbs, resulting in varied sensitivity and specificity of the assay. In addition, variability in the quality of conjugates could also influence the assay sensitivity, potentially leading to inconclusive results [29
]. Polyclonal antibodies have been recently proposed as an alternative, and shown to have slightly higher sensitivity and specificity in detecting the antigen [12
]. It might, therefore, be necessary for OIE reference laboratories to produce and distribute standard reagents for use by any laboratory worldwide. Alternatively, an assay using conjugated secondary antibodies [40
] may be explored.
A simple conventional RT-PCR has been found to be highly congruent to DFA in declaring positivity for rabies [35
]. An example of the RT-PCR profile is depicted in Figure 2
. The positive control yielded the expected amplicon of 605 bp, whereas no amplification was seen in negative or no template controls (NTC).
Analysis of data (see Table 2
) from our studies (see Table 2
) revealed that 1/1, 3/5, 92/144, 18/51, 8/9, 1/1, 1/1, 17/19, 3/3, 18/18 and 5/5 samples from Andhra Pradesh, Gujarat, Karnataka, Kerala, Maharashtra, Manipur, Pondicherry, Punjab, Rajasthan, Tamil Nadu and Uttar Pradesh, respectively, were positive by both DFA and dRIT, amounting to 167/257 (64.98%) of all the samples tested being positive by both the tests. It is to be noted that there was 100% agreement between the two tests in interpreting the results as positive, negative or inconclusive. Thirty six of the 257 samples (Table 2
) had decomposed to various levels, and consequently produced considerable levels of non-specific auto-fluorescence in DFA (data not shown), making the interpretation ambiguous. This is in line with the earlier findings that there is a considerable loss of sensitivity of DFA when decomposed tissues are subjected to testing [14
All of the 167 brain samples positive by DFA and dRIT were positive by RT-PCR (Table 2
). In addition, 35 of the 36 samples, which were deemed to be unfit for either DFA or dRIT due to decomposition of the tissue, also yielded the expected amplicons by RT-PCR. Indeed, RT-PCR can detect the presence of nucleic acid in samples in decomposing conditions and collected several days earlier, transported at ambient temperatures, archived in frozen or fixed condition for several years, exhumed bodies, or in some cases, as an intravitam diagnostic assay [42
]. This is not surprising since nucleic acids, especially fragments, are likely to be more resistant to tissue decomposition than proteins. This is of practical significance as delayed or improper transportation is often the cause of deterioration of clinical samples, leading to the inability to interpret the results or provide an indeterminate result. It should be noted, however, that a product of 605 bp, as amplified in our study, could miss some samples as the target region is more susceptible to fragmentation and/or degradation as compared to a smaller amplicon.
The ASSURED (affordable, sensitive, specific, user-friendly, rapid, equipment-free, door-step) criterion is especially applicable for the diagnosis of rabies since the disease is invariably fatal and occurs in resource-limited countries. The dRIT is a simple test for the diagnosis of rabies, and could be considered in place of the current standard test, the DFA. However, requirement for a microscope and trained personnel, as well as refrigeration of reagents is still an impediment to the widespread use of dRIT. Various laboratories have, therefore, developed immuno-chromatographic dip-stick tests using lateral flow assays for the diagnosis of rabies as well as other related lyssaviruses [53
]. However, these assays need a lot of standardization since a wide range of sensitivities, specificities and batch-to-batch variation have been observed when compared side by side [57
], and incomplete inactivation of the virus could be a biohazard [61
]. In addition to lateral flow tests, nucleic acid-based assays which are suitable for field use, such as the loop-mediated or other isothermal amplification techniques [63
], hold promise for the diagnosis of rabies, despite challenges in the standardization and validation of these molecular tests [66
]. Table 3
presents a comparison of the various tests developed or employed for the diagnosis of rabies.