The first p27 ELISA tests were based on polyclonal antibodies; such tests had the advantage of allowing quantitation of p27 but had a tendency to produce false-positive results as the antibodies did not detect only viral proteins but occasionally also non-viral components [Lutz et al., 1980b; Lutz et al., 1980c]. Improved ELISA tests based on monoclonal antibodies to p27 were introduced later to detect p27 capsid protein of exogenous FeLV present in blood or serum [Lutz et al., 1983a; Lutz et al., 1983b]. This assay utilizes a single monoclonal antibody specific for an epitope (A) of p27 fixed to a solid phase. The serum sample to be tested is mixed with one or two additional monoclonal antibodies specific for epitopes B and C of p27, and the mixture is then added to the solid phase. Hence the presence of p27 leads to insolubilisation of the enzyme-conjugated antibodies and the resulting colour change is indicative for the presence of p27, a marker of infection (but not always of viraemia as soluble p27 may be detected in the absence of infectious virus). ELISA procedures have the advantage of high diagnostic sensitivity and specificity – which, however, depends on the gold standard used for comparison.
In a field study in which the gold standard was proviral PCR, the diagnostic sensitivity was found to be 90 %, i.e. about 10 % of all 597 cats tested and found to be PCR positive were not recognized by p27 ELISA due to the fact that they are not antigenaemic; the specificity was very close to 100 % in that none of the p27 positive samples turned out to be PCR-negative (Hofmann et al. 2001). If the gold standard is virus isolation, the diagnostic sensitivity is in the range of 90 % and the diagnostic specificity of >98 % [Hartmann et al. 2001].
These tests are based on the same principle as the ELISA but small beads less than one micron in size are coated to the revealing antibodies rather than enzymes. The diagnostic sensitivity and specificity of immune chromatography tests was shown to be comparable to those of the ELISA [Hartmann et al., 2007; Hartmann et al., 2001; Pinches et al., 2007; Robinson et al., 1998].
The first method that allowed FeLV detection in viraemic cats under field conditions was the indirect immunofluorescent assay (IFA), introduced in 1973 [Hardy et al., 1973]. It was based on the observation that granulocytes, lymphocytes, and platelets in viraemic cats contain gag components, which may be detected by IFA in blood smears. The diagnostic sensitivity of IFA compared to virus isolation as the gold standard is significantly lower than 100 %; but positive cats are usually persistently viraemic [Hawks et al., 1991]. If a viraemic cat has leukopenia or if only a small percentage of peripheral leukocytes are infected, the presence of FeLV infection may be overlooked using IFA tests. Furthermore, all eosinophils have a tendency to bind the FITC conjugates used for IFA resulting in false positive tests if slides are not read carefully [Floyd et al., 1983].
Virus isolation in cell culture has been considered to be the ultimate criterion for FeLV infection. [Jarrett 1980; Jarrett et al., 1982]. Indeed, in the early phase of infection, detection of infectious FeLV is often the most sensitive parameter (Lehmann et al. 1991). In view of difficult logistics, this test is no longer considered for routine testing.
Since every cat cell carries between 12 and 15 copies of endogenous FeLV, it proved to be somewhat difficult to determine sequences that allowed only detection of exogenous provirus [Jackson et al., 1996]. The value of PCR techniques was greatly enhanced by the development of real-time PCR that not only allows detection but also quantitation of FeLV provirus [Hofmann-Lehmann et al., 2001]. PCR procedures have the highest analytical and diagnostic sensitivity and – provided the laboratory tests are run with all precautions of clean work and separated labs and with all necessary controls under conditions defined by good laboratory practice – also a very high specificity.
PCR for the detection of provirus may be useful for the clarification of inconclusive p27 antigen test (see chapter section 4.5.3).
The detection of viral RNA added a new aspect to the diagnosis of FeLV infection [Tandon et al., 2005]. Using this test, viral RNA present in whole blood, serum, plasma, saliva or faeces, is extracted, reverse transcribed into a cDNA, which is then amplified by real-time PCR. This technique permits the detection and quantitation of virus in the absence of cells. RNA PCR does not provide the same information as DNA provirus PCR. Many cats that have overcome FeLV viraemia remain provirus positive but do not produce detectable viral RNA in plasma, saliva or faeces [Gomes-Keller et al., 2006a]. However, detection of viral RNA is a reliable parameter of viraemia.
In most situations, cats are tested for FeLV individually. However, in circumstances where the cost of testing is a limitation, it is possible to use the RNA PCR test to screen pooled saliva samples, as the test is sufficiently sensitive to detect a single infected cat in up to 30 pooled samples. This approach may be advantageous when screening multicat households [Gomes-Keller et al., 2006b].
Although it is possible to measure antibodies against FeLV, the results are difficult to interpret because many cats develop antibodies to their own endogenous FeLV. Therefore such tests are currently of little clinical value. In some research laboratories, the so-called FOCMA (feline oncornavirus associated cell membrane antigen) test was used to detect antibodies to what was believed to be a tumour-associated antigen. It was later found that FOCMA was indeed a combination of several viral components; as this test is difficult to establish and to standardise, is not considered to be of clinical value. There are tests for virus neutralising antibodies but these are not widely available (mainly restricted to the UK) and are used only infrequently.
The first test that becomes positive after FeLV infection is usually virus isolation, followed within a few days by DNA and RNA PCR, ELISA, and later by IFA [Hofmann-Lehmann et al 2006]. Persistently viraemic cats are usually positive by all tests.
The most widely used tests for the diagnosis of FeLV infection in practice are antigen ELISA and immunochromatography. As the prevalence of FeLV infection seems to be decreasing in many European countries, there is a tendency for increasing false positive results. Therefore a doubtful positive result in a healthy cat should always be confirmed, preferably using provirus PCR (DNA PCR) offered by a reliable laboratory. A positive test in a cat with clinical signs consistent with FeLV infection is more reliable as in such cats the prevalence of FeLV is likely to be considerably higher.
Cats testing positive may overcome viraemia after two to sixteen weeks or in rare cases even longer. Therefore every positive cat without clinical signs should be separated and retested after several weeks or months; depending on discussions with the owner, retesting can occur at later time points up to one year when there is only a very small possibility that the cat will clear viraemia.
Cats that clear infectious virus from the plasma will be negative by VI, ELISA, immunochromatography, IFA, and RNA PCR, but remain positive by DNA PCR [Gomes-Keller et al., 2006a]. These cats should be considered latently infected, although the clinical significance is low in most cats. However, in rare instances, chronic stress, immune suppression or co-infection with other viruses may lead to reactivation in these cats. The mean proviral load in cats that overcome viraemia is several hundred times lower than in cats with persistent viraemia. A small proportion (2-3 %) of cats remain positive by ELISA and immunochromatography although no infectious virus can be isolated from the plasma. These cats have foci of infection outside the bone marrow from which soluble p27 is released into the circulation and such cats should be considered as potential sources of infection [Lutz et al., 1980c].
In summary, cats can be initially tested for p27. If the result is inconclusive for any reason, the test should be repeated by a qualified laboratory, using an alternate format, preferably PCR for provirus.