Anti-SU Antibody Responses in Client-Owned Cats Following Vaccination against Feline Leukaemia Virus with Two Inactivated Whole-Virus Vaccines (Fel-O-Vax® Lv-K and Fel-O-Vax® 5)

A field study undertaken in Australia compared the antibody responses induced in client-owned cats that had been vaccinated using two inactivated whole feline leukaemia virus (FeLV) vaccines, the monovalent vaccine Fel-O-Vax® Lv-K and the polyvalent vaccine Fel-O-Vax® 5. Serum samples from 428 FeLV-uninfected cats (118 FeLV-vaccinated and 310 FeLV-unvaccinated) were tested for anti-FeLV neutralising antibodies (NAb) using a live virus neutralisation assay to identify 378 FeLV-unexposed (NAb-negative) and 50 FeLV-exposed (NAb-positive; abortive infections) cats, following by anti-surface unit (SU) FeLV-A and FeLV-B antibody ELISA testing. An additional 42 FeLV-infected cats (28 presumptively regressively infected, 14 presumptively progressively infected) were also tested for anti-SU antibodies. NAb-positive cats displayed significantly higher anti-SU antibody ELISA responses compared to NAb-negative cats (p < 0.001). FeLV-unexposed cats (NAb-negative) that had been vaccinated less than 18 months after a previous FeLV vaccination using the monovalent vaccine (Fel-O-Vax® Lv-K) displayed higher anti-SU antibody ELISA responses than a comparable group vaccinated with the polyvalent vaccine (Fel-O-Vax® 5) (p < 0.001 for both anti-FeLV-A and FeLV-B SU antibody responses). This difference in anti-SU antibody responses between cats vaccinated with the monovalent or polyvalent vaccine, however, was not observed in cats that had been naturally exposed to FeLV (NAb-positive) (p = 0.33). It was postulated that vaccination with Fel-O-Vax® 5 primed the humoral response prior to FeLV exposure, such that antibody production increased when the animal was challenged, while vaccination with Fel-O-Vax® Lv-K induced an immediate preparatory antibody response that did not quantitatively increase after FeLV exposure. These results raise questions about the comparable vaccine efficacy of the different FeLV vaccine formulations and correlates of protection.


Introduction
Feline leukaemia virus (FeLV), a member of the Retroviridae family, was first reported in 1964 following its discovery by Bill Jarrett and colleagues during the investigation Viruses 2021, 13 of a time-space cluster of cats with T-cell lymphoma [1]. Subsequent research reported persistently viraemic cats with progressive FeLV infections to be 62 times more likely to develop lymphoma or leukaemia than FeLV-uninfected cats [2,3]. A potential link between transiently viraemic cats with regressive FeLV infections and lymphoma has also been suggested [4][5][6].
In addition to providing an early example of viral oncogenesis, abortive FeLV infections provide one of the few examples of a retroviral infection from which some animals can completely recover. In contrast, other retroviruses such as bovine leukaemia virus, feline immunodeficiency virus (FIV), feline foamy virus, equine infectious anaemia virus, caprine arthritis encephalitis virus and human immunodeficiency virus-1 (HIV-1), cause life-long infections [7]. Additionally, FeLV is one of the few retroviruses for which commercial vaccines exist (in addition to FIV and Jembrana disease in cattle) [8][9][10][11][12][13][14][15][16]. Consequently, the analysis of immune responses facilitating complete recovery in FeLV-exposed cats with abortive infections, and those protecting cats following FeLV vaccination, are important to assist with current efforts to develop effective vaccines against other retroviruses, such as HIV-1 [17,18].
Antibody-mediated (humoral) and cell-mediated immunity are both important in protecting against FeLV infection [8,19]. Following exposure to FeLV, virus-neutralising antibodies (NAbs) directed predominantly against the surface unit (SU) envelope glycoprotein gp70 have been observed in cats with abortive infections, in which FeLV replication is restricted to oropharyngeal tissues before being cleared [20][21][22]. Anti-FeLV NAbs have also been detected in cats with transient viraemia (regressive infections) [19,[23][24][25]. Kittens fed colostrum from queens that had recovered from natural FeLV infection, and then subsequently challenged with FeLV, were protected from persistent viraemia (progressive infection) due to the passive transfer of anti-FeLV NAbs [26]. Cats that become persistently viraemic following FeLV challenge typically develop neither NAb nor high levels of FeLV-specific cytotoxic lymphocytes, indicative of inadequate humoral and cell-mediated immune responses, respectively [23][24][25]27]. FeLV vaccination likely primes the humoral and cellular immune responses; thus, anti-FeLV NAbs are usually not detectable in FeLVvaccinated cats pre-challenge but develop in response to experimental or natural exposure to FeLV [8,9,11,19,21,[27][28][29].
The aim of this study was to investigate the anti-SU antibody responses of clientowned cats in Australia that had been vaccinated against FeLV using two commercially available, inactivated whole-virus (IWV) FeLV vaccines (Fel-O-Vax ® Lv-K and Fel-O-Vax ® 5).

Study Population
Residual samples from a study investigating FeLV infection in healthy client-owned cats, which included client-owned cats from 13 veterinary hospitals around Australia [14], and cats from two rescue facilities in Sydney, New South Wales (NSW) that were sampled in response to FeLV-associated disease outbreaks [21], were utilised.
Animal ethics approval for the sampling of the client-owned cats was granted by the University of Sydney Animal Ethics Committee (Approval number N00/1-2013/3/5920), while the rescue cats were sampled and tested at the request of the facility managers following the diagnosis of progressive FeLV infections.

Vaccination History
Vaccination histories were extracted from the medical records of client-owned cats. When multiple vaccines were administered concurrently, they were given as separate injections, in different syringes, at different sites. All vaccines were injected subcutaneously into the dorsal interscapular area, except for Fel-O-Vax ® Lv-K, which was injected into the right flank fold. In this study, "on-time" vaccination was defined as a primary course of two vaccines given one month apart followed by annual re-vaccination as per manufacturer's guidelines, with the last FeLV vaccine being administered within 18 months of sampling. Although Fel-O-Vax ® Lv-K and Fel-O-Vax ® 5 are registered as annual vaccines in Australia, the duration of immunity for FeLV vaccines has been shown to exceed 12 months [10,[30][31][32][33][34]. The term "overdue" vaccination was used to define cats for which the last FeLV vaccine had been given more than 18 months prior to sampling.

Determination of FeLV Exposure/Infection Status
A combination of FeLV p27 capsid antigen testing and FeLV proviral real-time polymerase chain reaction (qPCR) testing was used to classify cats as FeLV-uninfected or FeLV-infected, using accepted definitions [35]. FeLV p27 testing of whole blood was performed using three commercially available FeLV point-of-care (PoC) antigen test kits (SNAP Combo ® , IDEXX Laboratories, Westbrook, ME, USA; Witness ® , Zoetis Animal Health, Lyon, France; and Anigen Rapid ® , BioNote, Gyeonggi-do, Korea). FeLV proviral qPCR testing was performed in-house according to a published protocol, using primers designed to amplify a section of the unique region (U3) of the long terminal repeat (LTR) of the three main subtypes of FeLV [35,36].
Cats testing p27-negative/qPCR-negative were considered FeLV-uninfected, and NAb results were used to further categorise these cats as FeLV-unexposed (NAb-negative) or FeLV-exposed (NAb-positive; abortive infections). FeLV-vaccinated cats testing NAbpositive were classified as abortive infections, since FeLV-vaccinated cats do not produce NAbs prior to exposure to FeLV [8,9,11,19,21,[27][28][29]. Cats testing p27-negative/qPCRpositive were considered presumptively regressively infected, while cats testing p27positive/qPCR-positive were considered presumptively progressively infected ( Figure 1). FeLV infection status was considered "presumptive", since testing had been performed at a single time point and therefore it was not possible to confirm whether viraemia (as determined by p27 positivity) was transient or persistent.

Figure 1.
Algorithm used for classifying the FeLV exposure and infection status of cats recruited for the current study (grey cat with black outline). Cats unexposed to FeLV are represented as a white cat with a black outline; FeLV-uninfected cats that had been exposed to FeLV (NAb-positive; abortive infections) are represented as a white cat with a black outline and surrounded by FeLV, to demonstrate their robust immune response to clear early FeLV infection and protect them from further FeLV challenge; presumptively regressively infected cats are represented as a grey cat with a white outline and a question mark to represent a possible predisposition to developing lymphoma; and presumptively progressively infected cats are represented as a black cat with a white outline and a tombstone to represent their poor prognosis. FeLV-infected cats were classified as "presumptively" infected since testing was only performed at a single time point. NAb = neutralising antibody, qPCR = real-time polymerase chain reaction.

FeLV NAb Testing
NAb were detected using a focus reduction test [29]. Briefly, QN10 cells were incubated overnight in 12-well plates (4 × 10 4 cells/mL) before the addition of 4 × 10 2 FFU/mL of FeLV-A/Glasgow-1 that had been incubated for 2 h with an equal volume of two-fold, serial dilutions of plasma samples (1/4, 1/8, 1/16 and 1/32). Residual infectivity was measured in quadruplicate wells. NAb titres were recorded as the reciprocals of the plasma dilutions that visibly reduced the number of focus forming units by 75% compared to the virus control wells incubated without plasma. A NAb titre of 4 was considered weakly NAb-positive, while 32 or greater was considered a strong NAb-positive value.

FeLV Anti-SU Antibody ELISA Testing
Anti-SU antibody ELISA testing was performed to assess humoral response to FeLV vaccination, since FeLV-vaccinated cats have been shown to produce pre-challenge antibodies against whole FeLV and p45 (the non-glycosylated form of gp70) quantifiable by ELISA [37], whereas FeLV-vaccinated cats do not produce NAb without FeLV exposure [8,9,11,19,21,[27][28][29]. Plasma samples were tested for anti-SU antibodies, using both Fc-tagged FeLV-A SU and Fc-tagged FeLV-B SU as capture antigens, to ensure a comprehensive analysis of antibody responses against FeLV SU and to maximise test sensitivity [38]. The FeLV-B SU antibody response was interpreted as an antibody response to an epitope shared between FeLV-A and FeLV-B SU.
Positive and negative controls were included on each test plate. The positive control was a pooled sample of plasma from specific pathogen-free (SPF) cats that had recovered from experimental FeLV infection and tested strongly positive for anti-FeLV NAbs. The negative control was a pooled sample of plasma from SPF cats that had negligible reactivity to FeLV SU by immunoblot. Normalised optical density (NOD) values were determined using the formula NOD = [(Sample OD-Negative control OD)/(Positive control OD-Negative control OD)]. Samples were tested in triplicate and tests were for the current study (grey cat with black outline). Cats unexposed to FeLV are represented as a white cat with a black outline; FeLV-uninfected cats that had been exposed to FeLV (NAb-positive; abortive infections) are represented as a white cat with a black outline and surrounded by FeLV, to demonstrate their robust immune response to clear early FeLV infection and protect them from further FeLV challenge; presumptively regressively infected cats are represented as a grey cat with a white outline and a question mark to represent a possible predisposition to developing lymphoma; and presumptively progressively infected cats are represented as a black cat with a white outline and a tombstone to represent their poor prognosis. FeLV-infected cats were classified as "presumptively" infected since testing was only performed at a single time point. NAb = neutralising antibody, qPCR = real-time polymerase chain reaction.

FeLV NAb Testing
NAb were detected using a focus reduction test [29]. Briefly, QN10 cells were incubated overnight in 12-well plates (4 × 10 4 cells/mL) before the addition of 4 × 10 2 FFU/mL of FeLV-A/Glasgow-1 that had been incubated for 2 h with an equal volume of two-fold, serial dilutions of plasma samples (1/4, 1/8, 1/16 and 1/32). Residual infectivity was measured in quadruplicate wells. NAb titres were recorded as the reciprocals of the plasma dilutions that visibly reduced the number of focus forming units by 75% compared to the virus control wells incubated without plasma. A NAb titre of 4 was considered weakly NAb-positive, while 32 or greater was considered a strong NAb-positive value.

FeLV Anti-SU Antibody ELISA Testing
Anti-SU antibody ELISA testing was performed to assess humoral response to FeLV vaccination, since FeLV-vaccinated cats have been shown to produce pre-challenge antibodies against whole FeLV and p45 (the non-glycosylated form of gp70) quantifiable by ELISA [37], whereas FeLV-vaccinated cats do not produce NAb without FeLV exposure [8,9,11,19,21,[27][28][29]. Plasma samples were tested for anti-SU antibodies, using both Fc-tagged FeLV-A SU and Fc-tagged FeLV-B SU as capture antigens, to ensure a comprehensive analysis of antibody responses against FeLV SU and to maximise test sensitivity [38]. The FeLV-B SU antibody response was interpreted as an antibody response to an epitope shared between FeLV-A and FeLV-B SU.
Positive and negative controls were included on each test plate. The positive control was a pooled sample of plasma from specific pathogen-free (SPF) cats that had recovered from experimental FeLV infection and tested strongly positive for anti-FeLV NAbs. The negative control was a pooled sample of plasma from SPF cats that had negligible reactivity to FeLV SU by immunoblot. Normalised optical density (NOD) values were determined using the formula NOD = [(Sample OD-Negative control OD)/(Positive control OD-Negative control OD)]. Samples were tested in triplicate and tests were repeated if the standard deviation was >0.1. Anti-SU ELISA results were not categorised as "positive" or Viruses 2021, 13, 240 5 of 18 "negative"; rather, the range of antibody responses against SU were compared among the cats tested.

Determination of FIV Status
Cats were considered FIV-infected if they tested positive for FIV antibodies using two commercially available FIV PoC antibody test kits (Witness ® , Zoetis Animal Health, Lyon, France; and Anigen Rapid ® , BioNote, Gyeonggi-do, Korea), which can be used to differentiate FIV-infected and FIV-vaccinated cats [38]. In addition, client-owned cats were tested for FIV proviral DNA using a commercially available qPCR assay (FIV RealPCR TM , IDEXX Laboratories, East Brisbane, QLD, Australia), and virus isolation (VI) was performed in rare, discrepant cases (Yamamoto Laboratory, The University of Florida, Gainesville, FL, USA; and Veterinary Diagnostic Services, The University of Glasgow, Scotland, UK) [39,40].

Statistical Analysis
Numerical analyses were performed using the statistical software Genstat 18th Edition (VSN International, Hemel Hempstead, UK) and R Version 3.6.2 (The R Foundation, Vienna, Austria). Ordinal logistic regression (OLR) testing was used to compare sex and breed compositions between groups. A Shapiro-Wilk test was used to assess data for normality. Normally distributed data (cat ages) were analysed using means and two-sample t-testing, while non-normal data (anti-SU NOD values) were analysed using medians, Mann-Whitney U-testing (paired samples) and restricted maximum likelihood (REML) modelling (multiple groups). Fisher's exact testing was used to compare binomial outcomes. Statistical significance was considered when p < 0.05.
The categorisation of 428 FeLV-uninfected cats and 42 FeLV-infected cats according to FeLV vaccination history and FeLV infection status is shown in Table 1. Four FeLV-infected cats (three presumptively regressive, one presumptively progressive) had been vaccinated against FeLV, although their FeLV infection status prior to primary vaccination had not been determined.
In total, 33 cats were FIV-infected, including four cats co-infected with FIV and FeLV (2 presumptively progressive FeLV infections with FeLV cycle threshold (C T ) values of 16.8 and 21.3, respectively; 2 presumptively regressive FeLV infections with C T values of 31.2 and 37.2, respectively). Table 1. Summary of FeLV infection status and FeLV vaccination history of the cats in the current study (n = 470). On-time vaccination referred to the previous vaccination being given within the past 18 months, while overdue vaccination referred to the last vaccination being given more than 18 months previously. FeLV-uninfected/FeLV-unexposed cats that were overdue for vaccination were excluded from all analyses, while FeLV-uninfected/FeLV-exposed cats (abortive infections) overdue for vaccination were retained. None of the four FeLV-vaccinated/FeLV-infected cats had been tested for FeLV infection prior to vaccination.
The positive control sample included on each plate had a mean absorbance of 2.6 and the negative control had a mean OD of 0.5.

FeLV Anti-SU Antibody ELISA Testing in Unvaccinated Cats and Abortive Infections
The anti-SU antibody ELISA levels against FeLV-A of unvaccinated/unexposed cats (n = 303; considered to be the field control group) were significantly lower than those of presumptively progressively infected cats (n = 14) (0.9 vs. 1.1, p = 0.012; Mann-Whitney U-test), but unvaccinated/unexposed cats had significantly higher anti-SU antibody ELISA levels against FeLV-B than presumptively progressively infected cats (1.1 vs. 0.7, p = 0.009; Mann-Whitney U-test). A wide range of anti-SU ELISA values was observed in this field control group (0.07-2.9 against FeLV-A, 0.1-2.5 against FeLV-B). Presumptively regressive infections (n = 28) had significantly higher anti-SU antibody ELISA levels (2.5/1.9 against FeLV-A/FeLV-B) than both unvaccinated/unexposed cats and presumptively progressively infected cats (p < 0.001 for both; Mann-Whitney U-tests). Anti-SU antibody ELISA responses were similar between presumptively regressive infections and pooled vaccinated/unvaccinated abortive infections (n = 50; 2.2/2.3 against FeLV-A/FeLV-B) (p = 0.10 for FeLV-A, p = 0.62 for FeLV-B; Mann-Whitney U-tests; Figure 3). Excluding anti-SU antibody results from four infected cats (three regressive, one progressive) that were vaccinated, to exclude possible confounding effects of vaccination and infection on antibody production, had no effect on the findings. ELISA levels against FeLV-B were not significantly different between NAb-positive samples grouped according to reciprocal titre (NOD values were 2.3, 2.1, 2.3 and 2.2 for 4, 8, 16 and ≥32, respectively) (p = 0.32; REML testing). All NAb-positive groups, when considered individually (i.e., 4, 8, 16 and 32), were significantly higher against both FeLV-A and FeLV-B SU than NAb-negative samples (p < 0.001; least significant difference testing) (Figure 2).  The positive control sample included on each plate had a mean absorbance of 2.6 and the negative control had a mean OD of 0.5.

Comparing FeLV Anti-SU Antibody Responses between Fel-O-Vax
Unfortunately, group sizes were too small to compare vaccinated/unexposed and vaccinated/exposed cats based on the number of annual FeLV re-vaccinations administered.

Discussion
In this study, anti-FeLV SU antibody responses were measured by ELISA, comparing the results between FeLV-vaccinated and FeLV-unvaccinated cats given core vaccines concurrently under natural conditions. Surprisingly, despite the two IWV FeLV vaccines being derived from the same cell line and strain of FeLV (FeLV-A/61E), using the same adjuvant formulation and being produced by a single manufacturer in the same facility [41], a quantitatively different humoral response was observed.
Vaccination using the monovalent FeLV vaccine (Fel-O-Vax ® Lv-K, given concurrently with Fel-O-Vax ® 3) induced significantly higher anti-SU antibody levels in FeLV-unexposed cats against both FeLV-A and FeLV-B compared to vaccination using a polyvalent vaccine that included a FeLV component (Fel-O-Vax ® 5). Following natural exposure to FeLV, however, antibody responses to the different vaccines were similar. While vaccination with the polyvalent Fel-O-Vax ® 5 reproduced the response reported from experimental studies, in which FeLV vaccination led to minimal antibody production but primed the cat's humoral immune system in case of FeLV exposure [19], vaccination with the monovalent Fel-O-Vax ® Lv-K resulted in a robust humoral response irrespective of FeLV exposure. Similarly, a strong humoral immune response following vaccination (and prior to experimental FeLV challenge) with a recombinant p45 FeLV vaccine (Leucogen ® ) has been observed, using an ELISA to measure anti-p45 antibodies [42]. This is the first time that a difference in antibody response between cats vaccinated with Fel-O-Vax ® Lv-K and Fel-O-Vax ® 5 has been reported. Previously, most Fel-O-Vax ® vaccine studies have only considered one of the two vaccine formulations at a time and have been primarily concerned with vaccine efficacy as determined by the presence or absence of antigenaemia or viraemia after challenge (Table 3). Only two groups to date have concurrently reported the efficacies of the monovalent and polyvalent Fel-O-Vax ® FeLV vaccines, but neither measured anti-FeLV antibody levels following vaccination or FeLV challenge [30,41,43]. The effect of FIV infection on the anti-SU antibody response following FeLV vaccination in the current study could not be examined because of insufficient sample sizes, although it was reported that healthy experimentally FIV-infected cats were able to mount a sufficient humoral immune response to vaccination with a recombinant FeLV vaccine in the early phase of FIV infection [42].
Any anti-FeLV SU antibody result should be considered alongside results from p27 antigen and proviral PCR testing, particularly when testing has only been performed at a single time point [44]. As predicted, presumptively progressively infected cats in the current study had low anti-SU antibody levels, indicative of poor humoral (and presumably cellmediated) immune responses despite ongoing viraemia (antigenaemia). There was a wide range of anti-SU responses detected by ELISA for the field control group (unvaccinated and unexposed); we suspect that this spread of results could have reflected anti-SU antibodies that were detected by ELISA but did not neutralise the virus used in the NAb assay, since it was demonstrated previously that some FeLV-infected cats produce antibodies that do not contribute to virus neutralisation [45]. It is likely that the field control group contained some cats that had recovered from FeLV exposure after developing protective cellular immunity mediated by FeLV-specific cytotoxic T lymphocytes, in the absence of NAb, while developing antibodies that were detectable by ELISA [19,23,24]. Presumptively regressively infected cats displayed higher anti-SU antibody levels than presumptively progressively infected cats and the field control group, and similar levels to the antibody responses in uninfected/exposed cats (abortive infections), providing evidence of strong and effective humoral immune responses in naturally infected cats with regressive FeLV infections [21,22,28,44,46]. This conclusion was also supported by the higher proportion of presumptively regressively infected cats that tested positive for anti-FeLV NAbs, compared to only a small number of presumptively progressively infected cats.
The reason for the difference in antibody response observed between cats vaccinated with the monovalent FeLV and polyvalent FeLV vaccine is not known. A saturation phenomenon with the anti-SU antibody ELISA, that might have limited higher NOD values (e.g., for cats vaccinated with Fel-O-Vax ® Lv-K and then naturally exposed to FeLV), was considered unlikely on the basis of internal testing that demonstrated a concentration dependent decrease in absorbance levels, even with a range of starting sample dilutions. This finding contrasts with a study that compared the immunogenicity of a recombinant FeLV p45 vaccine given as a monovalent or polyvalent formulation (Nobivac ® FeLV vs. lyophilised Nobivac ® Forcat reconstituted in one dose of Nobivac ® FeLV), reporting that both vaccinations induced comparable antibody levels against FeLV as measured by an anti-p45 antibody ELISA [47]. One possible explanation for the unexpected results from the current study is that the route of vaccine administration impacted the antibody response, since Fel-O-Vax ® Lv-K was injected into the right flank fold and Fel-O-Vax ® 5 was injected into the dorsal interscapular region. In humans, a reduced seroconversion rate (up to 17-fold) was reported when an intramuscular hepatitis B vaccine was administered in the buttock rather than the arm, with delayed vaccine antigen release, or a lower number of macrophages, T and B lymphocytes in the injected area, hypothesised to be responsible for the difference in antibody response [48]. In rats, the highest antibody levels following vaccination with a commercial core canine vaccine were found in animals vaccinated subcutaneously in the houhai acupuncture site (the dorsal midline between the anus and tail base), and the lowest antibody levels were in animals vaccinated subcutaneously in the back at the level of the last thoracic vertebra on the dorsal midline [49]. Additional imaging investigations by the same researchers suggested that the enhanced humoral response observed in rats was as a result of increased lymphocyte activation and drainage in the houhai acupoint compared to the dorsal interscapular area [50]. To the best of the authors' knowledge, no studies have investigated differences in anti-FeLV antibody responses in cats associated with the site of vaccination, in part because vaccine licensing studies usually involve administering vaccinations subcutaneously into the interscapular region [33]. The European Advisory Board on Cat Diseases (ABCD) guidelines on feline injectionsite sarcoma recommend that veterinarians should avoid administering any vaccinations subcutaneously into the interscapular region. Vaccination in the distal limb or tail is recommended as surgery is more likely to be curative if a sarcoma develops at one of those sites compared to the interscapular area [51,52].
A second possible explanation for this unexpected finding is that an adjuvant effect was observed, with approximately two times the volume A third possible explanation for the difference in antibody responses between the monovalent and polyvalent FeLV vaccine may have been a "batch effect". Different batches of FeLV vaccines are known to vary within a range of approved potency values that have been demonstrated to be safe and efficacious. It is possible, especially given the low rate of FeLV vaccination in Australia, that the two veterinary hospitals using Fel-O-Vax ® Lv-K (located approximately 5 km apart, owned by the same veterinarian, but run separately including stock ordering) used FeLV vaccines from the same highly immunogenic batch during the three-year study period. Unfortunately, vaccine batch numbers were not recorded in the medical records. The phenomenon of "antigenic competition" was also considered as a potential explanation for the weaker antibody response that was observed in FeLV-unexposed cats vaccinated with the polyvalent FeLV vaccine compared to the monovalent FeLV vaccine. Antigenic competition has been reported in sheep vaccinated against the foot rot pathogen Dichelobacter nodosus. With this example, however, interference was associated with immunologically related pilus antigens, and a high level of protection with the polyvalent vaccine against all nine antigens was still achieved [55]. We hypothesise that a competitive effect would not be anticipated in a polyvalent vaccine such as Fel-O-Vax ® 5 that contains five antigenically distinct organisms [56]. Furthermore, Fel-O-Vax ® Lv-K vaccinated cats also received a killed-adjuvanted trivalent core vaccine (Fel-O-Vax ® 3) concurrently, but at a different site. Although antigenic competition is theoretically more likely to occur with Fel-O-Vax ® 5 than concurrent Fel-O-Vax ® 3/Fel-O-Vax ® Lv-K administration, due to the local component of antigenic competition, there is no published evidence that suggests that administering multiple antigens mixed in the same syringe (e.g., Fel-O-Vax ® 5) decreases the antibody response in vaccinated cats. It seems conceivable, however, that the pronounced difference in humoral immunogenicity observed in the present study with the administration of two different vaccine formulations, both containing the same FeLV IWV antigen potency, could have been the result of the monovalent vaccine (Fel-O-Vax ® Lv-K) being administered to a different area of the cat (right flank fold) and a different draining lymph node to the other vaccine antigens.
Further research is required, testing larger cohorts of cats, to determine if the observed difference in antibody response between cats vaccinated using Fel-O-Vax ® Lv-K and Fel-O-Vax ® 5 is correlated with a difference in protection from natural FeLV challenge. Since FeLV exposure/infection status had not been ascertained prior to FeLV vaccination in any cats, including for the four FeLV-vaccinated/FeLV-infected cats (three regressive, one progressive), due to the retrospective nature of the current study, inferences could not be made about FeLV vaccine effectiveness in the field. To date, only laboratory-based studies involving SPF cats, rather than field-based vaccine efficacy studies, have been performed to report the preventative fractions of Fel-O-Vax ® Lv-K and Fel-O-Vax ® 5, involving different challenge viruses and routes of inoculum administration (Table 3) [11,27,28,30,41,43,53,54]. In one of only two studies to directly compare the efficacy of Fel-O-Vax ® Lv-K and Fel-O-Vax ® Lv-K IV (identical to Fel-O-Vax ® 5), Legendre et al. (1991) (with results reported in Sebring et al. 1991) demonstrated that no kittens in either group, vaccinated subcutaneously in the flank region, became progressively infected, and both FeLV vaccines demonstrated preventative fractions of 100% [41,43]. The second study to examine both vaccine formulations reported efficacies of 86% and 100% for Fel-O-Vax ® Lv-K and Fel-O-Vax ® Lv-K IV, respectively, but did not perform any statistical analysis to assess the possible significance of this difference [30].
The efficacies of FeLV vaccines other than the Fel-O-Vax ® range, administered as either monovalent or polyvalent formulations, have been shown to be comparable. For example, in a study to determine the efficacy of Versifel ® FeLV (Zoetis Animal Health; an IWV vaccine containing FeLV-A, FeLV-B and FeLV-C), administered at the same time as a modified live-virus (MLV) trivalent core vaccine, no difference in FeLV vaccine efficacy was observed when the two vaccines were given concurrently (the FeLV component was injected subcutaneously at the base of the neck and the MLV component was injected subcutaneously in the left thoracic wall) compared to simultaneous administration (the MLV component was reconstituted using the FeLV vaccine and the entire contents were administered subcutaneously at the base of the neck) [57]. Similarly, in a study investigating the efficacy of a canarypox virus-vectored FeLV vaccine (Purevax ® FeLV, Merial), no difference in vaccine efficacy was observed, whether the FeLV component was administered as a monovalent or polyvalent vaccine [58]. Field-based vaccine efficacy studies for all FeLV vaccines are required to determine vaccine performance under natural challenge conditions and in different jurisdictions where different FeLV strains are circulating, similar to studies testing the efficacy of Fel-O-Vax ® FIV in the field [14,59].

Conclusions
In many countries around the world, including Australia, the prevalence of FeLV infection has decreased in part due to the use of efficacious vaccines. In other countries, including many in Asia, South America and some parts of Europe, the prevalence of FeLV remains high and vaccination programs are urgently required. It will be important to determine whether the increased humoral antibody response against the FeLV SU observed in FeLV-unexposed cats vaccinated with the monovalent FeLV vaccine (Fel-O-Vax ® Lv-K) compared to the polyvalent vaccine (Fel-O-Vax ® 5) is correlated with increased efficacy, and to address whether the site of vaccine administration affects the anti-SU antibody response and hence the efficacy of vaccination. Encouragingly, after natural exposure to FeLV, FeLV-vaccinated cats had comparable anti-SU antibody levels, suggesting that vaccination with either the monovalent or polyvalent Fel-O-Vax ® FeLV vaccine initiates a strong and protective antibody response on challenge. Veterinarians should continue to vaccinate any cat likely to be exposed to FeLV-infected cats, particularly kittens and young adult cats in multi-cat situations, to reduce the risk of progressive FeLV infection and the development of disease. FeLV testing prior to vaccination is advisable and will assist studies of the efficacy of FeLV vaccination in the field.