FeLV in Portuguese stray cats: molecular detection and subgroups
BMC veterinary research · 2025
In plain language
The study screened 129 stray cats in Portugal and performed molecular tests only in the 11 that tested positive for p27, a viral protein. FeLV DNA was detected in the blood of all 11: five samples had FeLV-A, and six had both FeLV-A and FeLV-B. One cat had no detectable viral RNA in its blood, but a single sampling cannot confirm how the infection develops over time. Selecting cats for molecular testing in this way may have missed infections with very small amounts of the virus.
Scientific detail
This observational in vivo study, published in BMC Veterinary Research in 2025 (doi: 10.1186/s12917-025-04691-2), screened 129 stray cats in Portugal. Molecular characterization was restricted to the 11 cats positive for p27 antigen. The reported claims carry evidence level C and describe natural infections, not a controlled infection experiment. The 129 cats were not a population screened by PCR independently of antigen results; this design cannot establish the frequency of antigen-negative, provirus-positive infections.
All p27-antigen-positive cats tested positive for FeLV proviral DNA by blood qPCR. The cat classified as likely having a regressive infection had no detectable blood viral RNA by RT-qPCR despite testing positive for p27. This discordance was observed at a single sampling point: it does not demonstrate complete absence of viral replication in all tissues or establish regressive infection longitudinally. Proviral DNA detection also provides no data on host-virus integration junctions or integration-site distributions.
Nested PCR detected FeLV-A alone in five positive samples and FeLV-A together with FeLV-B in six, with no FeLV-C amplification. These env-based subgroup assignments apply to the antigen-positive subset. Co-detection of A and B does not directly demonstrate a recombination event in an individual cat, and failure to amplify C does not establish its absence from the wider population. The study found no statistically significant difference in blood viral RNA loads between the A and AB groups. The numerical difference between group means does not establish greater replication associated with AB detection; a nonsignificant comparison in this small observational sample does not establish equivalence either.
The authors identify sample size and restriction of molecular testing to antigen-positive samples as principal limitations, potentially missing regressive infections with very low viral loads. This does not demonstrate that infections were missed or estimate their number; the observed predominance of high-load infections cannot characterize every infection state in the sampled population. Antibodies were not quantified, and one-off sampling prevented longitudinal confirmation of infection course. The study documents FeLV DNA, RNA and subgroup detection and provides data on the relationship between proviral burden, detectable RNA and subgroup composition. It does not directly investigate integration sites, replication kinetics, the origin of individual FeLV-B recombinants or causal effects of subgroup identity on replication or disease. Direct observations of recombination and functional subgroup interference assays are not reported.
What we know
FeLV proviral DNA was detected in the blood of all 11 cats selected through a positive p27 antigen result. Five samples had FeLV-A alone and six had both FeLV-A and FeLV-B; there was no FeLV-C amplification. One cat classified as likely having a regressive infection had no detectable blood viral RNA at that sampling point. No statistically significant difference in viral RNA loads was found between the A and AB groups. These are observational findings with evidence level C, limited to the antigen-positive subset.
What we don't know
The study does not establish the frequency of antigen-negative, provirus-positive infections, or whether any regressive infections were missed and how many. It does not confirm infection course over time or the absence of replication in all tissues of the cat with no detectable blood RNA. It also does not establish the absence of FeLV-C from the wider population, equivalence between the A and AB groups, integration sites, replication kinetics, the origin of individual FeLV-B recombinants or causal effects of subgroups on replication or disease.
Evidence
EvidenceLevel C — Preliminary evidenceConfidence 100/100
Molecular characterization was restricted to cats with positive p27 antigen screening results.
Excerpt verified in the source — verbatim, in the source's own language“All samples with a positive serologic test were subjected to the following molecular analyses: FeLV RNA and DNA qPCR, FeLV genetic characterization including subgroups identification and detection of Mycoplasmas Haemo Group in FeLV (+) samples.”
The molecular findings describe an antigen-selected subset, not an independently PCR-screened population of 129 cats. This design cannot establish the frequency of antigen-negative, provirus-positive infections.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
All p27-antigen-positive cats were positive for FeLV proviral DNA by blood qPCR.
Excerpt verified in the source — verbatim, in the source's own language“All cats that serologically tested positive for Ag p27 were also FeLV(+) by qPCR blood testing.”
The study reports concordant antigen and proviral DNA detection within the antigen-positive subset. It does not report host-virus integration junctions or establish integration-site distributions.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
The cat classified as likely regressive had no detectable blood viral RNA by RT-qPCR despite a positive p27 antigen test.
Excerpt verified in the source — verbatim, in the source's own language“No viremia was detected by RT-qPCR in the “likely regressive” case, although the ELISA p27 antigen test was positive in this cat.”
This is a single-time-point discordance between antigen detection and detectable blood viral RNA. It does not demonstrate complete absence of viral replication in all tissues or longitudinally establish regressive infection.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
Nested PCR detected FeLV-A alone in five positive samples and FeLV-A plus FeLV-B in six, with no FeLV-C amplification.
Excerpt verified in the source — verbatim, in the source's own language“Subgroup results by nested PCR showed that of the eleven FeLV-positive samples, 5 were amplified for subgroup A and 6 were amplified for subgroups A and B with a combination of AB and no amplification for subgroup C and/or ABC. Nested PCR also allowed the identification of FeLV subgroups, with 45.5% (95% CI 27.3%-81.8%) for subgroup A and 54.5% (95% CI 18.2%-72.7%) for subgroup AB. Confidence intervals were calculated using bootstrap resampling.”
These are molecular subgroup assignments in the antigen-positive subset. Co-detection of A and B does not directly demonstrate a recombination event within a particular cat, and failure to amplify C does not establish its absence from the wider population. The confidence intervals are reproduced as reported.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
The study did not find a statistically significant difference in blood viral RNA loads between the A and AB groups.
Excerpt verified in the source — verbatim, in the source's own language“These descriptive findings suggest different distributions of viral load results across FeLV subgroups; however, inferential analysis using Mann-Whitney did not suggest a statistically significant difference between subgroups ( p = 0.429).”
The numerical difference in group means does not establish greater replication associated with AB detection. A nonsignificant comparison in this small observational sample also does not establish equivalence.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
The authors acknowledge that antigen-gated molecular testing may have missed regressive infections with very low viral loads.
Excerpt verified in the source — verbatim, in the source's own language“The main limitation of this study, besides the sample size, is that the molecular characterization was only performed on samples that tested positive in the serologic screening tests. This may have caused us to overlook regressive states with very low viral loads.”
This is an acknowledged ascertainment limitation, not evidence that missed infections actually occurred or an estimate of their number. The observed predominance of high-load infections cannot characterize every infection state in the sampled population.
Provenance
Source: Prevalence and genetic characterization of feline leukemia virus in portuguese stray cats.
AVA