The two major retroviruses of cats, feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV) are important causes of disease and mortality in cats worldwide. Testing for these two viruses is an integral part of everyday feline practice and serves to man aims
diagnosis of retroviral associated disease
control and management programmes particularly for breeding catteries and rescue shelters
There have been important developments in testing for the feline retroviruses in recent years, particularly related to the introduction of polymerase chain reaction (PCR) tests. However most cats are still tested using in practice test kits and the key issue to the practitioner is how much they can rely on these tests and when the use of more specialised confirmatory tests is indicated.
Prevalence of FeLV and FIV
A key factor in evaluating the reliability of the tests is the prevalence of infection and whether this is changing. A low and reducing prevalence will decrease the positive predictive value of a test and can increase concerns about false positives. This is particularly relevant to FeLV for which there is general agreement that the prevalence has decreased significantly in recent years, in countries where FeLV vaccination has been used extensively.
There is surprisingly little accurate data on the prevalence of FeLV and most information is based on analysis of the proportion of samples tested positive in diagnostic laboratories. This is clearly a biased population which may not reflect the general population. Analysis of results of diagnostic samples showed 2.3% of a large number of test (Levy et al, 2006) performed in North America and in the UK, 5.0% of samples tested as positive (Hosie et al, 1989). However many practitioners in the UK report that FeLV positive results are now quite a rarity and it is considered that the prevalence of infection is below 1% for the general cat population. In other countries a much higher prevalence is still reported and in Brazil proviral DNA was found in around 47% of a convenience sample of both healthy and sick cats (Coelho et al, 2008).
There is more information on the prevalence of FIV infection. This varies between countries but is typically around 12-18% in sick cats. In contrast to FeLV, a significant proportion of healthy cats test positive typically 3-6%. Surveys in Brazil have shown around 15% of cats to test positive - 14% in Sao Paul (Roche et al, 1999) and 17% in Rio de Janeiro. Clade B appears to be the only clade to have been identified in Brazil (Matins et al, 2008). Even in countries in which the prevalence of FeLV now appears to be low, screening for both FeLV and FIV is usually an early priority in laboratory investigation of sick cats and is the only practical way of identifying infection.
FeLV in practice test kits. The majority of cats are screened using in-practice test kits which are based on detection of P27, one of the major core proteins. The most commonly based test is the SNAP test (Idexx Ltd) which is based on ELISA methodology. Other in-practice test kits use immunochromatography methodology to detect p27 and are sometimes termed rapid immunomigration (RIM) TESTS. The most popular of these is the Witness test (Synbiotics Ltd) but a variety of other tests are available.
In-practice tests are simple to use but there are potential pitfalls which may lead to erroneous results. The most common pitfalls are:
using the test kit at an incorrect temperature
reading the test at an incorrect time
using whole blood
The sensitivity of the tests is very good approaching 100% (for detecting viremia) and the specificity is considered to be around 98% (Robinson et al, 1998). Negative results are therefore generally reliable but false positives can be a problem, particularly if testing cats with a low likelihood of infection. It is recommended that in these circumstances, particularly if testing healthy cats, positive results should be checked using a confirmatory test (see below)
Another anomaly that has been recognised for many years is that viremia cannot be detected by virus isolation in a significant proportion (up to 10-15%) of cats which test positive for p27 antigen. These cats are referred to as "discordant" cats. No satisfactory explanation has been established for discordance and clarification of whether or not this indicates genuine infection. One suggested explanation has been that these are latently infected cats but the application of new PCR tests (see below) has failed to confirm this. The experience is that discordant cats do not have significant infection in that they do not appear to be at significantly increased risk of FeLV related disease nor of transmitting infection.
This test is based on detecting viral components in infected white cells on blood smears using an indirect immunofluorescent technique. The test is widely used in the USA but less so elsewhere. Operator expertise is crucial for accurate interpretation of results but in experienced hands results correlate well with the results of virus isolation.
This test was used as the "gold standard" confirmatory test for many years mainly in the UK but has been largely replaced by PCR tests. It is surprisingly sensitive and will detect infection at an early stage. The main disadvantages of this technique is that it depends on specialist expertise/facilities and results take some time to generate - up to two weeks to confirm a negative result.
PCR tests are now generally more readily available. The most commonly used is based on detection of proviral DNA. The tests are usually based on real-time PCR and give a quantitative result of the amount of proviral DNA. High levels of provirus invariably correspond to viremia. However a significant number of cats have considerably lower (x 100 - x 1000) levels of proviral DNA and these cats are generally negative on other tests including p27 assays and virus isolation. It is believed that these cats have been previously exposed to FeLV and have controlled but failed to eliminate virus. This is now believed to be the normal outcome in cats which meet and control virus (Hofmann-Lehmann et al, 2001). A key question for these cats is, as for discordant cats, whether virus can subsequently be reactivated leading to FeLV related disease or transmission of infection. This does not appear to be the case and these cats usually have high levels of serum neutralizing antibody which would be protective against viremia.
PCR tests have also been developed for detecting viral mRNA but these are less widely available. They are extremely sensitive and will also detect virus in samples other than blood such as saliva. Presence of mRNA indicates viremia.
In practice test kits As with FeLV testing, most cats are screened using in-practice test kits and these often combine FeLV and FIV in the same test kit device. The tests differ from FeLV in practice tests in that they are based on detection of antibodies rather than antigen, usually antibodies against p24, a major core protein, and transmembrane antigen. The same issues about pitfalls in performing in-practice FeLV test kits also apply to the FIV tests. In one study in the USA, repeat testing of the same samples that had previously been tested as positive in practice, using the same test method performed in a specialist laboratory found 15% of samples interpreted as positive in practice to be negative (Barr et al, 1991).
Specificity of the in-practice test kits for FIV is generally good and positive results are reliable but sensitivity is not ideal. Studies have shown that a significant proportion (up to 20%) of cats which can be shown to be infected by virus isolation tests give negative results on antibody testing. The reason for this discrepancy is not known and early infection before seroconversion does not seem to be the explanation in most cases. If negative results are obtained for a case with a high index of suspicion, use of an alternative test method should be considered.
There are two situations in which the use of antibody detection is unsatisfactory. Firstly, it invalidates testing in young kittens since any antibody detected may be maternally derived. Only a relatively small proportion of kittens born to FIV infected queens develop maternally derived infection. Antibody testing cannot be used to detect infection in kittens until they have reached 56 months of age. This can be particularly relevant in rescue shelters when testing of kittens from a feral background is required and early homing is desirable to facilitate socialization.
The other situation in which the use of antibodies for testing is a disadvantage is in countries where vaccination against FIV is available and cats have been vaccinated. In-practice test kits will not discriminate between antibodies induced by vaccination and those resulting from natural infection although recently techniques have been described which will allow differentiation (Levy et al, 2008).
Western blotting and immunofluorescence
More specialised tests for detecting antibody are based on other methodologies, using western blotting or immunofluorescence, but these are less widely available. Whilst they may offer some advantages in accuracy in detecting antibodies they s till suffer from the same inherent disadvantages (outlined above) of using a test based on antibody detection.
This is regarded as the gold standard test but is very specialised and laborious making it unsuitable for routine diagnostic use.
PCR tests are becoming more widely available and have the significant advantage of avoiding some of the drawbacks of detecting antibody. The main difficulty inherent in PCR tests for FIV is the variability of isolates for a virus that exists in different clades and has a propensity to mutate. Early PCR tests consequently had a poor sensitivity, often no more than 70% (Crawford et al, 2005). We have attempted to circumvent this problem by combining two different sets of primers in our PCR test for FIV to improve the sensitivity (Pinches et al, 2007). The use of real-time PCRs enables quantitation of virus in the blood. This may be of advantage in assessing the significance of a positive result and in monitoring progress/disease progression.
1. Barr MC, et al (1991) Comparison and interpretation of diagnostic tests for feline immunodeficiency virus infection (FeLV). J. Am. Vet. Med. Assoc 199: 1377-1381
2. Caxito F, Coelho F, Oliveira M, & Rosende M (2006) Feline immunodeficiency virus subtype B in domestic cats in Minas Gerais, Brazil. Vet Res. Comm. 30: 953-956
3. Coelho FM, Bomfim MRQ, Caxito FA, Ribeiro NA, Luppi MM, Costa A, Oliveira ME, DaFonesca FG, Resende M (2008) Naturally occurring feline leukemia virus subgroup A and B infections in urban domestic cats. J. Gen. Virol. 89: 2799-2805
4. Crawford PC, Slater MR, Levy JK (2005) Accuracy of polymerase chain reaction assays for diagnosis of feline immunodeficiency virus infection in cats. J. Amer Vet Med Assoc 226: 1503-1507
5. Hofmann-Lehmann R, Huder JB, Gruber S, Boretti F, Sigrist B, Lutz H. (2001) Feline leukemia provirus load during the course of experimental infection and in naturally infected cats. J Gen Virol 82: 1589-96.
6. Hosie MJ, Robertson C, Jarrett O. (1989) Prevalence of feline leukemia virus and antibodies to feline immunodeficiency virus in cats in the United Kingdom. Vet Rec 125: 293-7.
7. Levy JK, Scott HM, Lachtara JL, Crawford PC. (2006) Seroprevalence of feline leukemia virus and feline immunodeficiency virus infection among cats in North America and risk factors for seropositivity. J Am Vet Med Assoc 228: 371-6.
8. Levy JK, Crawford PC, Jusuhara H, Motokawa K, Gemma T, Watanabe R, Arai S, Bienzle D & Hohdatsu T (2008). Differentiation of feline immunodeficiency virus vaccination, infection, or vaccination and infection in cats. J. Vet. Intern. Med. 22: 330-334.
9. Pinches MD, Diesel G, Helps CR, Tasker S, Egan K, Gruffydd-Jones TJ (2007). An update on FIV and FeLV test performance using a Bayesian statistical approach. Vet Clin Pathol 36: 141-7.
10. Robinson, A. et al (1998) Comparison of a rapid immunomigration test and ELISA for FIV antibody and FeLV antigen testing in cats. Vet. Rec. 18, 491-492.
11. Reche A Jr, Hagiwara MK & Lucas SRR (1999) Estudo clinico da síndrome de imunodeficiência adquirida em gatos domésticos de Sao Paulo. Braz. J. Vet. Res. Anima. Sci. 34: 152-155.
12. Souza HJM, Teixeira CH, & Grac R (2002) Estudo epidemiológico de infeceões pelo virus immunodeficiência feline, em gato domésticos do município do Rio de Jaeneiro. Clin Vet. 36: 14-21.