Introduction
FeLV is an enveloped RNA retrovirus. The virus usually enters via the oral mucosa, and viral replication takes place in the local lymphatic tissue. Primary viremia occurs because of spread of infected lymphocytes and monocytes. When these cells reach the bone marrow, replication in blood precursors starts and a secondary viremia may develop.1 Some cats do not control the viremia and become p27 antigen-positive 3 to 6 weeks after FeLV exposure. In cats experimentally exposed to FeLV, p27 antigenemia increased after the second week post-infection and was associated with progressive disease with higher peripheral virus charge.1,2
The cornerstone of control of FeLV is the identification of infected cats. Point-of-care FeLV p27 serology screening tests are regularly performed in veterinary practice because of their affordable cost, wide availability, good sensitivity and specificity, and rapid results.3 If a point-of-care antigen test is positive, diagnosis references recommend confirmation of the results by running a second and different antigen test, testing RNA viral excretion using saliva, or submitting an EDTA blood sample for pro-virus PCR testing.1 This recommendation gives various options, and allows clinicians to confirm the diagnosis using different viral and molecular techniques. However, FeLV diagnosis confirmation is important as p27-positive antigenemia is associated with progressive disease and a hazard ratio of death of 3.4 compared with healthy cats.4 In addition, FeLV-infected cats with anemia are 3.5 times more likely to die than FeLV-infected cats without anemia4 and have a median survival time of 0.6 years.5
Feline retrovirus screening is routinely performed in cats with anemia. In a study involving 8,642 FeLV-infected cats presented to a North American teaching hospital, 11% presented with anemia.6 However, only about 10% of FeLV anemias are regenerative because of the common bone marrow suppression effect of the virus,7 supporting the weak evidence of FeLV causing immune-mediated hemolytic anemia (IMHA).8
Depending on host virus control, FeLV disease has different clinical and diagnostic presentations; the most common are abortive disease (p27 serology negative, pro-virus PCR negative), regressive disease (p27 serology usually negative, pro-virus PCR positive), and progressive disease (p27 serology positive, pro-virus PCR positive).1 P27 positive results with negative pro-virus is an uncommon presentation. These conflicting results between point-of-care p27 serology and PCR tests in cats with IMHA have been previously observed.9,10 Two cats described in 2016 in a large IMHA descriptive study had a positive SNAP ELISA FeLV test (SNAP Combo FeLV/FIV; Idexx Laboratories, Inc) with a negative pro-virus DNA PCR, but no further investigations were performed.9 Six cases of suspected false-positive results with SNAP ELISA tests and anemia of suspected immune-mediated origin (ASIMO) have been detected in Spain,10 and since then more patients have been identified in Europe.
The primary aim of this retrospective study was to describe the clinical findings, diagnosis, treatment, outcome, and survival in 18 cats with ASIMO with a discrepancy between FeLV diagnostic tools at admission, specifically a positive SNAP ELISA FeLV result with a negative PCR test for the detection of FeLV pro-virus. A second aim was to describe the most effective way to clarify the FeLV status of these patients. We hypothesized that these patients will have a clinical presentation and treatment response similar to cats with IMHA or precursor targeted immune-mediated anemia (PIMA), and the survival will differ from patients with anemia due to FeLV. We also suspect that point-of-care testing followed by quantitative PCR is the most accurate way to diagnose these patients.
Materials and Methods
Clinical and clinicopathological data of cats with ASIMO, positive results with SNAP ELISA FeLV tests (SNAP Combo FeLV/FIV; Idexx Laboratories, Inc), and negative FeLV PCR results were retrospectively reviewed at selected referral practices in Spain, Italy, and the United Kingdom between 2018 and 2022. Cases were recruited initially from the AniCura Ars Veterinària database by searching with the following keywords: “FeLV,” “FeLV PCR,” “IMHA,” and “PIMA,” and by contacting various centers by email and asking the clinicians to search in the medical record system for patients with positive point-of-care FeLV tests, HTC < 20%, and negative FeLV pro-virus PCR results. Data collected included signalment, clinical signs, physical examination findings, clinicopathological abnormalities, diagnostic tests, concurrent diseases, specific treatment, treatment response, outcome, and survival. Cats were considered to have regenerative anemia if the aggregate reticulocytes count on presentation was ≥ 50,000 cells/µL.11
Inclusion criteria were a diagnosis of anemia (HTC < 20%) with a suspected immune-mediated origin for ASIMO8,12,13 in the absence of an identified trigger other than p27 positivity, p27 positive result using a bidirectional-flow p27 antigen SNAP ELISA test and negative results on real-time pro-virus PCR (quantitative or qualitative) performed on peripheral blood. Different laboratory pro-virus PCR tests were performed depending on clinician preferences and availability. Veterinary Pathology Group laboratory used primers and cut-off values defined by Pinches et al in 2007.14 Specific primer sequences used by other laboratories were not able to be confirmed. Nevertheless, in the studies of validation of p27 ELISA15 and quantitative FeLV PCR16 performed by IDEXX, the primers used were those previously described by Tandon et al.17
A diagnosis supportive of IMHA was made if at least one of the criteria of immune-mediated erythrocyte destruction was found, following the ACVIM consensus.8 These criteria are as follows: positive saline agglutination test (SAT), positive direct antiglobulin test (DAT), positive flow cytometry (FC), and positive SAT that persists after washing. Positive results on macroscopic or microscopic evaluation, or both, were considered positive for SAT.
On the other hand, PIMA was diagnosed when destruction of the RBC precursors, rubriphagocytosis, was observed12 on bone marrow cytology, bone marrow biopsy, or spleen cytology,13 independently of peripheral immune-mediated RBC destruction.12 While the response to immunosuppressive treatment was recorded for all patients, it was not used as an inclusion criterion. Clinical follow-up was recorded for all cats from diagnosis to February 2022.
Patients were excluded if they did not have at least one of the criteria of immune-mediated RBC destruction,8 if they had positive results on pro-virus PCR, or if no investigation of triggering cause was performed.
Survival data were generated by the Kaplan-Meier method. For this analysis, any cat that died or was euthanized was classified as dead, and any cat that was lost to follow-up was censored. Statistical analyses were performed with SPSS Statistics for Macintosh version 25.0 (IBM), and descriptive statistics were used to report baseline data. Six cases included in this study had been previously presented as an oral presentation at the ECVIM Congress in 2020.10
Results
A total of 20 cats were reviewed: 8 cats were diagnosed at AniCura Ars Veterinària Hospital (Barcelona, Spain), 5 at the Veterinary Teaching Hospital of the University of Bologna, Italy, 4 at AniCura Glòries Hospital Veterinari (Barcelona, Spain), 2 at AUNA Veterinary Hospital (Valencia, Spain), and 1 at Dick White Referrals, Newmarket, United Kingdom.
Eighteen cats met the inclusion criteria; 2 were excluded because of lack of evidence of immune-mediated RBC destruction. Both excluded cats showed signs of hemolysis, but evidence of an immune-mediated cause was weak as only SAT was performed. One of the excluded cats died 3 days after the diagnosis, and the second one responded favorably to immunosuppressive treatment, being alive 2 years and 6 months after the diagnosis at the time of writing the manuscript.
Signalment and clinical signs
Eight cats were male (all neutered) and 10 were female (9 neutered and 1 intact). Sixteen cats were domestic shorthairs, 1 was a Maine Coon, and 1 was a Balinese crossbreed. Median age at the time of diagnosis was 1 year and 11 months (range, 9 months to 13 years). Only one cat had previously been vaccinated against FeLV; vaccination date was not recorded (Supplementary Table S1).
Sixteen cats were presented for lethargy or weakness (16/18 [89%]). Eleven had hyporexia or anorexia (11/18 [61%]). Three cats had pica, vomiting, or diarrhea (3/11 [17%]), and 2 had weight loss (2/18 [11%]). Other clinical signs detected were abdominal distention, pigmenturia, increased respiratory rate, sneezing, and over-licking (one cat each; 1/18 [6%]; Supplementary Table S1).
On physical examination, all cats had pale mucous membranes; however, jaundice was only observed in 6 patients (6/18 [33%]). Fourteen presented with tachycardia (14/18 [78%]) and 5 with tachypnoea (5/18 [28%]). Six cats had hyperthermia (6/18 [33%]) and 3 had hypothermia (3/18 [17%]). Five cats had a heart murmur (5/18 [28%]) and 2 had a gallop rhythm (2/18 [11%]) during auscultation. Hypotension, muscle wasting, and mental dullness were each detected in 1 cat (1/18, [6%]; Supplementary Table S1).
At presentation, 9 cats were receiving medication for a suspected haemolytic or immune-mediated process (9/18 [50%]; Supplementary Table S1). Six of these 9 cats were receiving doxycycline, and 4 were also on treatment with steroids. One cat was transfused with packed red blood cells (pRBCs) 2 days prior to being referred. Other medications previously administered to the patients were reported (Supplementary Table S1).
Diagnosis
Complete blood count and serum biochemistry results were available for all cats. Median HCT at presentation was 9.00% (range, 3.70% to 18.80%). Reticulocyte count was available in 17 out of 18 (94%) cats, with a median of 35,000 cells/µL (range, 200 to 356,000 cells/µL), and only 5 cats (5/18 [28%]) had regenerative anemia at diagnosis. The median automated platelet count was 81,500 cells/µL (range, 35,500 to 243,000 cells/µL), and genuine thrombocytopenia (platelets < 120,000 cells/µL without clumps in the blood smear) was present in 9 cats (9/18 [50%]). Median serum total protein concentration was 7.76 g/dL (range, 6.20 to 10.70 g/dL), and median serum globulin concentration was 4.60 g/dL (range, 2.80 to 7.80 g/dL). Hyperglobulinemia (≥ 5 g/dL) was present in 5 patients (5/18 [28%]). Hyperbilirubinemia (bilirubin ≥ 0.6 mg/dL) was present in 12 cats (12/18 [67%]). Six cats had a mild to moderate elevation of ALT (< 10 times the upper reference range; 6/18 [33%]) (Supplementary Table S2).
Blood smear evaluation was performed in all the patients. Ghost cells were present in 11 of the cats (11/18 [61%]; Supplementary Table S2). Thirteen of the cats had anisocytosis (13/18 [72%]). Polychromatophils or RBC precursors were present in 10 cats (10/18 [56%]); however, 7 of these patients did not show reticulocytosis on the CBC.
The SAT was performed in 17 patients. The dilution protocol performed for each laboratory was not routinely recorded for all the patients; however, 11 of 17 had a 4:1 saline to blood ratio. A positive SAT was present in 12 cats (12/17 [70%]); 7 had positive results on both evaluations, 2 showed macroscopic agglutination that was not microscopically evident, and 3 had microscopic agglutination only. The remaining 5 cats did not show agglutination (5/17 [30%]), with 2 of these receiving steroid therapy at the time of the evaluation. A Coomb test was only performed in 1 cat, with positive results (Supplementary Table S2).
Candidatus Mycoplasma hemominutum and Mycoplasma hemofelis PCR testing was performed in 15 cats (15/18 [83%]), and 7 of them were also tested for Candidatus Mycoplasma turicensis. All results were negative, but 3 cats were already on treatment with doxycycline when tested. Six cats had bone marrow sampling (6/18 [33%]). Five had erythroid hyperplasia and 1 had erythroid hypoplasia. Five cats had evidence of hemophagocytosis and rubriphagocytosis (Supplementary Tables 2).
Abdominal ultrasonography was performed in all patients. Splenomegaly was detected in 16 cats (16/18 [89%]), hepatomegaly in 6 (6/18 [33%]), and mesenteric lymphadenopathy in 7 (7/18 [39%]). One cat had minimal tricavitary effusion. Eleven patients underwent ultrasound-guided fine needle aspiration of the spleen and cytological examination (11/18 [61%]); the most common findings were extramedullary hematopoiesis (8/11 [73%]), reactive spleen (9/11 [82%]), hemophagocytosis (3/11 [30%]), and rubriphagocytosis (1/11 [10%]; Supplementary Table S2).
Patients were diagnosed as follows: 12 cats had a supportive diagnosis of IMHA (12/18 [66%]) and 6 cats were classified with PIMA (6/18 [33%]; Supplementary Table S2).
SNAP Combo FeLV/FIV testing was performed in all cats with positive results. In addition, 2 cats had positive results with another point-of-care FeLV test based on immunochromatography (IC; 2/18 [11%]; Uranotest FeLV-FIV; Urano). External laboratory serology was performed in 2 cats; 1 showed positive results using ELISA and the other had a negative result using IC (Supplementary Table S3). All cats were simultaneously tested for FIV antibodies and all were negative (SNAP Combo FeLV/FIV; Idexx Laboratories, Inc).
Pro-virus real-time FeLV PCR of whole blood was performed in all patients after initial SNAP ELISA testing. The median time from point-of-care testing to PCR was 1 day (range, 0 to 75 days; Supplementary Table S3). Seven PCR tests were quantitative and 11 were qualitative. One patient also underwent pro-virus real-time FeLV PCR testing of the bone marrow, with negative results. In 9 patients, a second SNAP Combo FeLV/FIV test was performed after remission of the clinical signs and normalization of the RBC count (9/18 [50%]; Supplementary Table S3). All follow-up SNAP ELISA tests had negative results. The median time to performance of the second test was 42 days after diagnosis (range, 17 to 159 days). One cat had serial SNAP ELISA tests performed with serum leftovers during the hospitalization period to evaluate when the patient became seronegative, despite SNAP ELISA being a qualitative test. Positive results were less clear after immunosuppressive treatment was started (Supplementary Figure S1) and 1 month after the test became negative. Finally, 2 patients that had negative follow-up SNAP ELISA test results showed an IMHA relapse 6 and 9 months after the diagnosis, and the FeLV SNAP ELISA tests were again positive. Pro-virus PCR testing was not repeated at that time.
Treatment
All patients were treated with prednisolone, initially administered IV and then PO upon diagnosis. Doses were adjusted at the clinician’s discretion (Supplementary Table S2). The median initial dose was 2 mg/kg (range, 1 to 3 mg/kg), once daily PO, and this initial dose was maintained a median of 1 month (range, 0.5 to 12 months) before initiating dose reduction. All patients were treated with doxycycline at 10 mg/kg once daily PO for a median of 3 weeks (range, 2 to 4 weeks). Six cats were treated with a second immunosuppressive drug (6/18 [33%]), including cyclosporine at 5 mg/kg twice daily PO (range, 5.4 mg/kg daily to 5 mg/kg twice per day) in 4 cases and mycophenolate mofetil at 10 mg/kg twice daily PO in 2 cases (Supplementary Table S2). One cat was switched from cyclosporine to MMF, and 1 month later to chlorambucil. Sixteen of the cats (16/18 [88%]) received a pRBC transfusion. Additional treatments administered at veterinary discretion were marbofloxacin (3/18 [17%] cats), clopidogrel (1/18 [6%]), and vitamin C (1/18 [6%]; Supplementary Table S2).
Survival
Follow-up was available for all the cats, and the median follow-up recorded was 168 days (range, 14 to 1,098 days; Supplementary Table S2). None of the cats showed signs of FeLV progressive disease. The mean survival time was estimated to be 769 days (Figure 1). The median survival time was not determined as the majority of cats were censored because they were still alive at the time of analysis and had good response to treatment, not because they had died or were lost to follow-up. Fourteen cats (14/18 [78%]) were alive at the time of writing this article, while 4 (4/18 [22%]) had died or were euthanized because of persisting or worsening anemia. Twelve cats had a follow-up longer than 2 months. In regard to the other 6 cats, 3 died before this time and 3 were enrolled with a follow-up shorter than 2 months. Of these 12 cats, one died because of refractory IMHA and the remainder were alive at the time of writing this article (11/12 [93%]).
Kaplan–Meier curve assessing survival time in a population of cats with suspected false positive results in point-of-care FeLV tests and concurrent immune-mediated anemia (n = 18, blue solid line). Mean survival time was estimated to be 769 days. The x axis shows time expressed in days, and the y axis shows cumulative survival (Cum Survival) expressed as a proportion of 1. Ticks on the line indicate censored items representing cats lost to follow-up.
Citation: Journal of the American Veterinary Medical Association 261, 10; 10.2460/javma.23.02.0059
Discussion
The feline population described in this study showed severe anemia, evidence of immune-mediated processes, and discordant results between SNAP ELISA p27 and real-time pro-virus PCR testing in peripheral blood. This situation has previously been reported in 2 cats with IMHA9 and in 6 additional patients.10 The latter 6 patients are included in this study.
In our population, diagnosis of PIMA or IMHA was achieved in 18 of 20 cats. Two patients showed signs of hemolysis, but an immune-mediated origin was not confirmed. The SAT was the unique test performed to investigate the immune-mediated origin of the anemia. It is possible that further investigations of the anemia of these cats could have supported an immune-mediated etiology of these 2 excluded cases.
FeLV has been sporadically described as a cause of IMHA; however, clear evidence of causation is lacking.8,9,18 In the study of Swann et al,9 2 cats with primary IMHA had positive results with p27 antigen testing using immunochromatography and negative results with pro-virus PCR. These patients were immunosuppressed, and response and outcome were similar to the other patients,9 in agreement with the findings described in this study.
In this study, median reticulocyte count at presentation (35,000 cells/µL; range, 200 to 356,000 cells/µL) was similar to that in previous studies of cats with primary IMHA (32,850 cells/µL; range 0.00 to 795,400 cells/µL).9 Regenerative anemia was present in 42% (7/18) of cats with a supportive diagnosis of IMHA. These results are in agreement with previous studies in which regeneration at diagnosis was present in 42% or 55% of cats with primary IMHA,9,18 but do not agree with characteristic anemia caused by FeLV infection because only around 10% of these are regenerative and most of them are associated with concurrent disease, such as mycoplasmosis.7
Bone marrow samples of FeLV-infected cats with progressive disease also differ from the results presented in this study. Non-neoplastic hematologic disorders described in association with anemia caused by FeLV include myelodysplastic syndrome, aplastic anemia (pancytopenia), and panleukopenia-like syndrome.19 Conversely, in this study, the most common bone marrow findings were erythroid hyperplasia and erythrophagocytosis.
At the time of writing this study, 78% (14/18) of the patients were alive and estimated mean survival time was 769 days (Figure 1). Previous studies in cats with immune-mediated anemia showed a median survival time of approximately 1.5 years, with 60% of the patients surviving more than 6 months.12,18 Conversely, in patients with progressive FeLV infection, anemia is correlated with a higher risk of death (hazard ratio of 3.5)4 and an average survival time of only 0.6 years.5
Treatment was administered at the clinician’s discretion, but all patients received prednisolone and doxycycline. Twelve patients did not receive additional immunosuppressive treatment; in 11, the disease was controlled, and 1 died without receiving a second immunosuppressive drug. Six cats received an additional immunosuppressive drug and 3 of them died because of nonresponsive disease despite the treatment.
In regard to the second aim of the study, recent studies showed that the test used in this study (SNAP Combo FeLV/FIV test) had a positive predictive value and negative predictive value of 100% in populations with an FeLV prevalence of 1%, 5%, and 10%.3 The 2020 American Association of Feline Practitioners retrovirus guidelines20 recommend confirming positive point-of-care serology results with a different brand antigen test, a microwell-plate ELISA for p27 antigen, or a PCR FeLV pro-virus test. However, in our study, and previously in another one,9 we found similar false-positive results in some microwell-plate external ELISA and different brand point-of-care tests (Supplementary Table S3). Therefore, we further emphasize that any anemic cat with positive screening results should always be confirmed by the presence of pro-virus FeLV DNA in whole blood.1,17 This is particularly relevant if FeLV prevalence or exposure risk is low as lower prevalence decreases the positive predictive value of the result, which increases the rate of false-positive test results.1 According to the most recent pan-European study, FeLV prevalence in the cat population is 0.7% in the United Kingdom, 2.6% in Spain, and 5.7% in Italy.21
Based on the results observed in this study, 2 main explanations are possible. The first group of explanations encompasses different possibilities of real false-positive results in the SNAP ELISA antigen test. The second one involves true FeLV infections without virus DNA detection.
False-positive results in p27 point-of-care FeLV tests
Cross-reaction between an unknown molecule and the point-of-care test—
We had previously hypothesized10 the existence of a cross-reaction between antibodies present in feline serum targeting unknown epitopes (against RBCs, their precursors, and/or platelets) and anti-p27 antibodies, reactants of the matrix, or the conjugate of the SNAP ELISA tests. Because of the presence of sero-negativization of all the patients tested after immune suppression and sero-positivization with relapse of ASIMO, our main theory is that a protein mimicking p27 is related to the immune-mediated destruction of RBCs in these patients.
The tests performed in the 2 cats described by Swann and colleagues were based on IC.9 This is meaningful because 2 of our patients also had positive results on tests based on IC carried out in their referral veterinary practice (Supplementary Table S3). This information highlights that both techniques (IC and ELISA) are able to detect p27 antigen or p27-like products, even when performed by different laboratories.
We cannot exclude that agglutination caused interference with the SNAP ELISA test. Agglutination is a complex process mostly generated by IgM, IgG, and complement deposit over the membrane of RBCs. It was detected in 70% (12/17) of our patients by the SAT, and its interaction with the test is not excluded by the manufacturing laboratory.22
Endogenous p27 expression—
Another theory that could explain the presence of p27 antigen in cats with immune-mediated anemia is related to the presence of retrovirus genetic material in the DNA of domestic cats. Retrovirus genetic material is an important part of the genome of the cat, as in most mammals. These integrated fractions are called endogenous retroviruses (ERVs). Usually, when ERVs are transcribed, non-infectious viral particles are produced.23 However, ERVs have displayed various functions in mammals: they have the capacity to produce infectious viruses, can undergo recombination with other retroviruses (such as FeLV B or C subtype),7 and can even function as host antiviral factors.23 This makes it possible that at some point, endogenous p27 protein was transcribed or recombined, starting or supporting the autoimmune disease.
From a diagnostic point of view, it is important that FeLV diagnostic tests can discriminate exogenous FeLV molecules from endogenous FeLV-like products.1 Consequently, p27 was chosen as a diagnostic tool because it has not been detected as an endogenous product, and cats are immunologically tolerant to p27 (they do not appear to immunologically respond to the protein and p27 remains in the blood). However, the p27 precursor gene (Gag gene) is part of the ERV,24 and p27 endogenous production has been alleged as one of the main explanations for the cat’s immunologic tolerance to this protein,24 making this replication conceivable.
To summarize, several ERVs have been identified in mammals, and it is unclear whether they remain inactive or could emerge as a future infection.1,25,26 The presence of FeLV Gag gene in the cat genome makes plausible the transcription of endogenous p27 protein, which could be another source of false-positive results in the SNAP ELISA p27 test; however, this has not been detected before.
Sample type interaction—
Finally, it is known that sample origin (whole blood, serum, plasma, saliva, etc) may interact with SNAP ELISA test results.27 Most of the tests should be performed with serum or plasma as whole blood in colloidal assays has led to higher rates of false-positive results.27 This is particularly relevant when the sample is hemolyzed and the test-line is red.27 In our case, because of the retrospective nature of the study, the type of sample used to perform the test was not consistently recorded in all centers; however, at least 12/18 (66%) of the diagnosis tests were performed using serum. Nevertheless, the technology used in all SNAP assays22 has a wash step in which the buffer removes unbound debris and conjugate from the matrix. This produces a clean background and the blue-colored positive results can be easily interpreted22; therefore, this interaction seems less probable. Bilirubinemia present in the hemolytic process has also been considered, although bilirubin does not seem to interact with the SNAP Combo FeLV/FIV test.22
Real FeLV antigenemia without the detection of virus DNA
False-negative results on pro-virus PCR—
Pro-virus PCR testing was performed in 5 different laboratories. Seven patients underwent quantitative PCR (qPCR) testing and 11 underwent qualitative testing as quantitative evaluation became more commercially available during the study period. Quantitative PCR tests are preferred because the amount of pro-virus DNA present in the blood of infected cats is correlated with the disease stage and outcome, as it is higher in patients with progressive disease.28 However, no preferences are defined for establishing diagnosis and cats with negative results are considered free of disease.1 It is interesting to note that in a study performed by a commercial laboratory to correlate the pro-virus load with the presence of p27 antigen in blood using quantitative ELISA testing, 24 of the 353 feline samples (7%) had FeLV pro-virus DNA below the limit of quantification but p27 antigen was detected.16 Clinicopathological data and evolution of these cats are not available.16
Antigenemia during primary viremia—
An alternative explanation of these results is the detection of p27 during the primary viremia stage. As p27 is a structural protein present in the virus membrane, theoretically it could be detected when the virus enters the blood. However, negligible amounts of p27 are detected in this primary phase, and in experimental studies, p27 blood concentration increases 2 weeks post-infection,2 which corresponds with the time that usually takes a cat with nonabortive disease to become FeLV pro-virus positive.1
Atypical FeLV presentation—
Atypical FeLV presentation was also debated. Diagnosis of atypical FeLV infection is confirmed by pro-virus positive results in the infected tissue. However, when whole blood tests are performed, inconsistent results among p27, virus, and pro-virus PCR tests are observed. FeLV antibodies are positive and FeLV RNA may be present. Focal or atypical FeLV infections are rare cases, uncommonly associated with systemic clinical signs, and do not show response to immunosuppressive therapy. Most of the cats with an “atypical” presentation were classified before 2000, when laboratory viral detection techniques were under development and the transport and storage of the samples had been shown to lead to false-negative results.29,30 This clinical outcome is currently uncommon and it was not considered a presentation option according to the latest American Association of Feline Practitioners retrovirus guidelines.21 In our cases, atypical FeLV infection is considered unlikely and false-positive SNAP ELISA antigen results are suspected. Nevertheless, negative results for FeLV antibodies, viral RNA PCR testing, and postmortem studies could had helped to clarify this hypothesis.
All the theories exposed above need further investigation to elucidate the underlying mechanism. However, until more clarity is achieved, positive results using SNAP ELISA FeLV diagnosis test should be interpreted with caution until molecular confirmation is completed.
The main limitations of this study are the small sample assessed and the inconsistency in the diagnosis and management of each case due to its retrospective nature. Better description, classification, and diagnosis could have been performed if an ordered approach was made. Moreover, it is already known that FeLV results can vary depending on the test used, the course of infection, and the time passed since the initial contact. In our case, sample type at initial testing was not always recorded and 5 laboratories performed the PCR tests with different primers, which were not disclosed by the laboratories on all occasions. Finally, the use of FeLV antibodies and virus RNA PCR tests could have helped to completely rule out atypical FeLV infection; however, they were not offered by contacted commercial laboratories. Follow-up for some patients is short (< 2 months) and survival analysis is compared with data extracted from other studies, making difficult to draw solid assumptions. A control group (either FeLV-infected patients with positive pro-virus PCR and anemia or patients with ASIMO with negative SNAP FeLV test) was not available because of the heterogeneity in the origin of the described population.
In conclusion, this study showed inconsistent results between FeLV SNAP ELISA and pro-virus PCR tests in cats with ASIMO. These results confirmed the necessity of confirming positive SNAP ELISA FeLV tests using further essays such as pro-virus PCR testing, especially if an immune-mediated process is suspected.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org
Acknowledgments
The authors declare that there were no conflicts of interest.
References
- 1.↑
Hofmann-Lehmann R, Hartmann K. Feline leukaemia virus infection: a practical approach to diagnosis. J Feline Med Surg. 2020;22(9):831-846. doi:10.1177/1098612X20941785
- 2.↑
Cattori V, Tandon R, Riond B, Pepin AC, Lutz H, Hofmann-Lehmann R. The kinetics of feline leukaemia virus shedding in experimentally infected cats are associated with infection outcome. Vet Microbiol. 2009;133(3):292-296. doi:10.1016/j.vetmic.2008.07.001
- 3.↑
Levy JK, Crawford PC, Tucker SJ. Performance of 4 point-of-care screening tests for feline leukemia virus and Feline Immunodeficiency Virus. J Vet Intern Med. 2017;31(2):521-526. doi:10.1111/jvim.14648
- 4.↑
Spada E, Perego R, Sgamma EA, Proverbio D. Survival time and effect of selected predictor variables on survival in owned pet cats seropositive for feline immunodeficiency and leukemia virus attending a referral clinic in northern Italy. Prev Vet Med. 2018;150:38-46. doi:10.1016/j.prevetmed.2017.12.001
- 5.↑
Helfer-Hungerbuehler AK, Widmer S, Kessler Y, et al. Long-term follow up of feline leukemia virus infection and characterization of viral RNA loads using molecular methods in tissues of cats with different infection outcomes. Virus Res. 2015;197:137-150. doi:10.1016/j.virusres.2014.12.025
- 6.↑
Cotter SM. Management of healthy feline leukemia virus-positive cats. J Am Vet Med Assoc. 1991;199(10):1470-1473.
- 7.↑
Hartmann K. Clinical aspects of feline retroviruses: a review. Viruses. 2012;4(11):2684-2710. doi:10.3390/v4112684
- 8.↑
Garden OA, Kidd L, Mexas AM, et al. ACVIM consensus statement on the diagnosis of immune-mediated hemolytic anemia in dogs and cats. J Vet Intern Med. 2019;33(2):313-334. doi:10.1111/jvim.15441
- 9.↑
Swann JW, Szladovits B, Glanemann B. Demographic Characteristics, survival and prognostic factors for mortality in cats with primary immune-mediated hemolytic anemia. J Vet Intern Med. 2016;30(1):147-156. doi:10.1111/jvim.13658
- 10.↑
Izquierdo Robert L, Feo Bernabé LJ, Seth M, Puig Prat J. Feline leukemia virus false positive results using an in-house test in cats with immune-mediated hemolytic anemia. Research communications of the 30th ECVIM-CA online congress. J Vet Intern Med. 2020;17(34):3058-3166.
- 11.↑
Winzelberg Olson S, Hohenhaus AE. Feline non-regenerative anemia: diagnostic and treatment recommendations. J Feline Med Surg. 2019;21(7):615-631. doi:10.1177/1098612X19856178
- 12.↑
Black V, Adamantos S, Barfield D, Tasker S. Feline non-regenerative immune-mediated anaemia: features and outcome in 15 cases. J Feline Med Surg. 2016;18(8):597-602. doi:10.1177/1098612X15588800
- 13.↑
Lucidi CA, de Rezende CLE, Jutkowitz LA, Scott MA. Histologic and cytologic bone marrow findings in dogs with suspected precursor-targeted immune-mediated anemia and associated phagocytosis of erythroid precursors. Vet Clin Pathol. 2017;46(3):401-415. doi:10.1111/vcp.12502
- 14.↑
Pinches MDG, Helps CR, Gruffydd-Jones TJ, Egan K, Jarrett O, Tasker S. Diagnosis of feline leukaemia virus infection by semi-quantitative real-time polymerase chain reaction. J Feline Med Surg. 2007;9(1):8-13. doi:10.1016/j.jfms.2006.05.008
- 15.↑
Buch JS, Clark GH, Cahill R, et al. Analytical validation of a reference laboratory ELISA for the detection of feline leukemia virus p27 antigen. J Vet Diagn Invest. 2017;29(5):654-659. doi:10.1177/1040638717710451
- 16.↑
Beall MJ, Buch J, Cahill RJ, et al. Evaluation of a quantitative enzyme-linked immunosorbent assay for feline leukemia virus p27 antigen and comparison to proviral DNA loads by real-time polymerase chain reaction. Comp Immunol Microbiol Infect Dis. 2019;67:101348. doi:10.1016/j.cimid.2019.101348
- 17.↑
Tandon R, Cattori V, Gomes-Keller MA, et al. Quantitation of feline leukaemia virus viral and proviral loads by TaqMan real-time polymerase chain reaction. J Virol Methods. 2005;130(1-2):124-132. doi:10.1016/j.jviromet.2005.06.017
- 18.↑
Kohn B, Weingart C, Eckmann V, Ottenjann M, Leibold W. Primary immune-mediated hemolytic anemia in 19 cats: diagnosis, therapy, and outcome (1998-2004). J Vet Intern Med. 2006;20(1):159-166.
- 19.↑
Stützer B, Müller F, Majzoub M, et al. Role of latent feline leukemia virus infection in nonregenerative cytopenias of cats. J Vet Intern Med. 2010;24(1):192-197. doi:10.1111/j.1939-1676.2009.0417.x
- 20.↑
Little S, Levy J, Hartmann K, et al. 2020 American Association of Feline Practitioners’ feline retrovirus testing and management. J Feline Med Surg. 2020;22:5-30. doi:10.1177/1098612X19895940
- 21.↑
Studer N, Lutz H, Saegerman C, et al. Pan-European study on the prevalence of the feline leukaemia virus infection - reported by the European Advisory Board on Cat Diseases (ABCD Europe). Viruses. 2019;11(11):1-27. doi:10.3390/v11110993
- 22.↑
O’Connor TP. SNAP Assay Technology. Top Companion Anim Med. 2015;30(4):132-138. doi:10.1053/j.tcam.2015.12.002
- 23.↑
Anai Y, Ochi H, Watanabe S, et al. Infectious endogenous retroviruses in cats and emergence of recombinant viruses. J Virol. 2012;86(16):8634-8644. doi:10.1128/JVI.00280-12
- 24.↑
Willett BJ, Hosie MJ. Feline leukaemia virus: half a century since its discovery. Vet J. 2013;195(1):16-23. doi:10.1016/j.tvjl.2012.07.004
- 25.↑
Kawasaki J, Nishigaki K. Tracking the continuous evolutionary processes of an endogenous retrovirus of the domestic cat: ERV-DC. Viruses. 2018;10(4):1-13. doi:10.3390/v10040179
- 26.↑
Polani S, Roca AL, Rosensteel BB, Kolokotronis SO, Bar-Gal GK. Evolutionary dynamics of endogenous feline leukemia virus proliferation among species of the domestic cat lineage. Virology. 2010;405(2):397-407. doi:10.1016/j.virol.2010.06.010
- 27.↑
Barr MC. FIV, FeLV, and FIPV: interpretation and misinterpretation of serological test results. Semin Vet Med Surg (Small Anim). 1996;11(3):144-153. doi:10.1016/S1096-2867(96)80026-0
- 28.↑
Hofmann-Lehmann R, Cattori V, Tandon R, et al. How molecular methods change our views of FeLV infection and vaccination. Vet Immunol Immunopathol. 2008;123(1-2):119-123. doi:10.1016/j.vetimm.2008.01.017
- 29.↑
Hartmann K, Werner RM, Egberink H, Jarrett O. Comparison of six in-house tests for the rapid diagnosis of feline immunodeficiency and feline leukaemia virus infections. Vet Rec. 2001;149(11):317-320. doi:10.1136/vr.149.11.317
- 30.↑
Levy J, Crawford C, Hartmann K, et al. 2008 American Association of Feline Practitioners’ feline retrovirus management guidelines. J Feline Med Surg. 2008;10(3):300-316. doi:10.1016/j.jfms.2008.03.002
