SARS-CoV-2 in Domestic UK Cats from Alpha to Omicron: Swab Surveillance and Case Reports

Although domestic cats are susceptible to infection with SARS-CoV-2, the role of the virus in causing feline disease is less well defined. We conducted a large-scale study to identify SARS-CoV-2 infections in UK pet cats, using active and passive surveillance. Remnant feline respiratory swab samples, submitted for other pathogen testing between May 2021 and February 2023, were screened using RT-qPCR. In addition, we appealed to veterinarians for swab samples from cats suspected of having clinical SARS-CoV-2 infections. Bespoke testing for SARS-CoV-2 neutralising antibodies was also performed, on request, in suspected cases. One RT-qPCR-positive cat was identified by active surveillance (1/549, 0.18%), during the Delta wave (1/175, 0.57%). Passive surveillance detected one cat infected with the Alpha variant, and two of ten cats tested RT-qPCR-positive during the Delta wave. No cats tested RT-qPCR-positive after the emergence of Omicron BA.1 and its descendants although 374 were tested by active and eleven by passive surveillance. We describe four cases of SARS-CoV-2 infection in pet cats, identified by RT-qPCR and/or serology, that presented with a range of clinical signs, as well as their SARS-CoV-2 genome sequences. These cases demonstrate that, although uncommon in cats, a variety of clinical signs can occur.


Introduction
Many mammalian species, including domestic cats and captive big cats, are susceptible to SARS-CoV-2 infection by reverse zoonosis [1]. Infected domestic cats can transmit infection to other cats in experimental settings [2][3][4][5], whereas experimentally infected dogs are less likely to shed infectious virus [4,5]. Domestic cat-to-human transmission, with supportive viral sequencing data, has been reported [6], and in a zoo setting, lion-to-human transmission was concluded to have caused a cluster of human cases [7]. Consequently, the World Organisation for Animal Health (WOAH) has recommended surveillance of SARS-CoV-2 infections in animals [8].
The prevalence of SARS-CoV-2 in pet cats, detected by either RT-qPCR or by serology, was higher in households containing positive-testing humans, compared to households where pets were not known to have been exposed to humans with COVID-19 [9]; indeed, most studies reporting results of RT-qPCR testing of domestic cats have targeted COVID-19 affected households, which increases the probability of detecting cases. Early in the pandemic, between April and May 2020, 19 cats from COVID-19 households in the USA tested negative by RT-qPCR [10]. Conversely, 6/50 (12.0%) cats from COVID-19 households in Hong Kong tested positive between February and August 2020 [11], and 5/65 (7.7%) of Figure 1. Line graph of the estimated fortnightly average number of people testing positive for SARS-CoV-2 in England, data from the Office for National Statistics (ONS) [43]. The timing of feline cases presented in this report is shown above the graph, and the sampling periods for the active surveillance component of this study are shown below it. The infected cat identified by active rather than passive surveillance (Cat X) is denoted by an asterisk (*) and the two cases which were positive on serology only (cases 3 and 4) are denoted by blood tube symbols.
A single swab from the Delta-dominant period tested positive by RT-qPCR (1/175, 0.57%), with Ct values of 22.2 for the N1 assay and 23.0 for the N2 assay (260,000-320,000 viral copies per µL RT-qPCR template). No positive samples were detected from the 374 swabs tested after the emergence of Omicron BA.1 and subsequent variants. The cat that tested positive during the Delta dominant period was a 14-year-old, female neutered (FN), domestic shorthair (DSH) (denoted Cat X to distinguish her from the cats from the four Figure 1. Line graph of the estimated fortnightly average number of people testing positive for SARS-CoV-2 in England, data from the Office for National Statistics (ONS) [26,43]. The timing of feline cases presented in this report is shown above the graph, and the sampling periods for the active surveillance component of this study are shown below it. The infected cat identified by active rather than passive surveillance (Cat X) is denoted by an asterisk (*) and the two cases which were positive on serology only (cases 3 and 4) are denoted by blood tube symbols.
A single swab from the Delta-dominant period tested positive by RT-qPCR (1/175, 0.57%), with Ct values of 22.2 for the N1 assay and 23.0 for the N2 assay (260,000-320,000 viral copies per µL RT-qPCR template). No positive samples were detected from the 374 swabs tested after the emergence of Omicron BA.1 and subsequent variants. The cat that tested positive during the Delta dominant period was a 14-year-old, female neutered (FN), domestic shorthair (DSH) (denoted Cat X to distinguish her from the cats from the four highlighted case reports, Table 1). The animal displayed mild respiratory signs of sneezing and nasal discharge and had shown only minimal improvement with antibiotic therapy. After amplicon sequencing, the complete SARS-CoV-2 genome was obtained and classified as Delta AY.4.2.1 lineage by PANGOLIN [44] and deposited in GISAID [45] (accession number EPI_ISL_17971927). There were no spike mutations not seen in closely related human-derived sequences (Table A3, Figure A3).

Passive Surveillance
We received one swab from a cat with suspected SARS-CoV-2 infection during the Alpha wave. This animal (Cat 1A) tested positive by RT-qPCR and further details are given below (case one). We received swab samples from ten animals where SARS-CoV-2 infection was a differential diagnosis during the Delta wave. One animal that tested RT-qPCR positive (Cat 2A) is detailed below (case two) and there was an additional weakly positive result (Cat Y, no associated detailed case report) (2/10, 20%). In the latter case only a small proportion of the SARS-CoV-2 genome was reconstructed after sequencing, which was insufficient to confirm the variant present (Appendix A.2.1, Appendix B). This nine-year-old, FN, Siamese cat (Cat Y) had nasal discharge and a history of recent exposure to SARS-CoV-2. Since BA.1 became dominant, samples from eleven cats suspected of being infected with SARS-CoV-2 were received (rectal/faecal only from one animal); none tested RT-qPCR positive (0/11, 0%).

Seropositive Cases Identified Previously
Seventy seropositive cats were detected in our active serosurveillance study, which was reported previously [19], and included animals sampled from April 2020 to February 2022. The information supplied by the submitting veterinarians, together with available diagnostic test results for the seventy seropositive cats identified in that study, are reviewed here.
No clinical histories were available for 22/70 seropositive animals. Nine animals had been sampled for routine disease screening, three for treatment monitoring and five following positive in-house tests for other pathogens. Thirty cats were reported to be unwell with a range of clinical signs; these included eight animals with gingivitis, five with pyrexia, three with ascites, three with ocular signs, three with neurological signs, one with tachypnoea and one with inappetence (data in Supplementary Materials). Tests requested by the submitting veterinarians included haematology, biochemistry, plasma proteins (albumin, globulin), A1-AGP (acute phase protein), T4, effusion analysis and tests for feline leukaemia virus, feline immunodeficiency virus, feline coronavirus, feline calicivirus, feline herpesvirus, Toxoplasma gondii, Chlamydia felis and Mycoplasma felis.
Of the 30 cats reported to be unwell at the time of sampling, other pathogens/diseases detected likely accounted for the reported clinical signs in eight cases, but in 22 cases they did not. Of these 22 cases, nine displayed clinical signs similar to those reported in people infected with SARS-CoV-2 (pyrexia, inappetence, conjunctivitis and tachypnoea), and eight had signs not, or only rarely, reported in humans with COVID-19, which were likely unrelated (polyuria, gingivitis, jaundice and increased appetite). In addition, two animals displayed neurological signs, one was anaemic, one had uveitis and one had collapsed.

Case One: Acute Pyrexia
In February 2021, Cat 1A, a 9-month-old, male neutered (MN), DSH, developed anorexia and pyrexia for 24 h that resolved following treatment with a non-steroidal antiinflammatory drug (NSAID). This illness occurred during a household COVID-19 outbreak, and the cat was suspected to have been infected with SARS-CoV-2. Three of four humans in the household tested positive for SARS-CoV-2 RNA by RT-qPCR (samples not sequenced). A nasal/oral swab from this cat tested positive in our SARS-CoV-2 N1 and N2 RT-qPCR assays (Ct 31; 200 viral copies per µL RT-qPCR template) and the complete viral genome was reconstructed from the sequence data (GISAID accession number EPI_ISL_17971924), with PANGOLIN classifying the sequence as an Alpha lineage. The sequence contained one synonymous spike mutation, L821L, not previously detected in any closely related human-derived viral sequences (Table A3, Figure A1

Case Two: Sudden Death
Cat 2A was an eight-year-old, MN, DSH with no known health problems, owned by the same household since kittenhood. The household also contained another MN DSH (Cat 2B) who was eleven years old.
All three people in the household developed COVID-19, testing positive by RT-qPCR of nasopharyngeal swab samples in October 2021 (samples not sequenced). During their ten-day isolation, both cats developed mild respiratory signs including sneezing ( Figure 2). The sneezing abated, leaving both cats with mild epiphora and Cat 2B frequently licking its lips. Five days after the sneezing was most frequent, Cat 2A made an unusual sound, yawned and died suddenly. Dry swab samples were collected approximately 15 h post mortem from the nose, oral cavity, trachea and rectum of Cat 2A. Both the SARS-CoV-2 N1 and N2 RT-qPCR assays tested positive for one or more swabs. Ct values of between 17 and 19 (9.5-16.5 million copies per µL of template) for both assays were obtained from the nasal and tracheal swabs. The Ct was 37.5 (16 copies per µL, which was close to the limit of detection [approximately five copies not including RT step]) for N2 for the rectal swab, with no tested positive for one or more swabs. Ct values of between 17 and 19 (9.5-16.5 million copies per µL of template) for both assays were obtained from the nasal and tracheal swabs. The Ct was 37.5 (16 copies per µL, which was close to the limit of detection [approximately five copies not including RT step]) for N2 for the rectal swab, with no target detected using N1 primers, and no viral RNA was detected from the oral swab (Table A2). A near complete viral genome was reconstructed from the sequence data, except for one 485-nucleotide section due to a failed amplicon. The variant was classified as belonging to the Delta AY.4 lineage by PANGOLIN (GISAID accession EPI_ISL_17971926) and contained no novel spike mutations when the sequence was compared to the most closely related human-derived sequences (Table A3, Figure A2).
Cat 2B continued to show excessive lip licking for several weeks. At a veterinary examination 2.5 weeks after Cat 2A's death, no abnormalities were detected. A blood sample was collected and neutralising antibody titres of 62, 57 and 1019 were measured for the B.1/D614G, Alpha and Delta variants, respectively, indicating that this cat had been infected with the Delta variant.

Case Three: Anorexia
Cat 3A was an eight-year-old, MN, Ragdoll cat. The household had owned Cat 3A and his full brother, Cat 3B (MN), since kittenhood; the cats had no medical issues. The patient's owners both tested positive for SARS-CoV-2 by RT-qPCR of nasopharyngeal swabs in November 2021 (samples not sequenced). Towards the end of their isolation period, both cats became lethargic, listless, had watery ocular discharge and were sneezing ( Figure 3). Cat 3B recovered uneventfully.  After the ocular/respiratory signs resolved, Cat 3A was presented to his usual veterinary practice with a primary complaint of three days of inappetence. Lethargy, mild vomiting of froth and chewing of plants had been observed by the owners. Clinical examination was unremarkable and an anti-emetic (maropitant, Vetemex, 1 mg/kg SC) and an H2 receptor antagonist (famotidine, 5 mg bid PO) were prescribed.
Four days later the patient re-presented at the practice having still not eaten. He had been noted to drink small amounts and had mild occasional vomiting and diarrhoea. Clinical examination revealed a tense abdomen and weight loss of 340g since the first consultation. The patient was admitted, and he remained in hospital for a week. Throughout this time the cat showed no desire to eat.
Haematology was unremarkable and serum biochemistry/electrolytes were normal, other than a mildly raised amylase and hypochloraemia. Pancreatic lipase immunoreactivity was normal and unsupportive of a diagnosis of pancreatitis. Intravenous fluid ther- After the ocular/respiratory signs resolved, Cat 3A was presented to his usual veterinary practice with a primary complaint of three days of inappetence. Lethargy, mild vomiting of froth and chewing of plants had been observed by the owners. Clinical examination was unremarkable and an anti-emetic (maropitant, Vetemex, 1 mg/kg SC) and an H 2 receptor antagonist (famotidine, 5 mg bid PO) were prescribed.
Four days later the patient re-presented at the practice having still not eaten. He had been noted to drink small amounts and had mild occasional vomiting and diarrhoea. Clinical examination revealed a tense abdomen and weight loss of 340g since the first consultation. The patient was admitted, and he remained in hospital for a week. Throughout this time the cat showed no desire to eat. Haematology was unremarkable and serum biochemistry/electrolytes were normal, other than a mildly raised amylase and hypochloraemia. Pancreatic lipase immunoreactivity was normal and unsupportive of a diagnosis of pancreatitis. Intravenous fluid therapy was administered and was stopped after urine specific gravity indicated good hydration. Abdominal radiography and ultrasonography were largely unremarkable; areas of gastric wall close to the pylorus appeared thicker on the first ultrasonographic examination but normal on a subsequent scan four days later.
Whilst hospitalised, the patient was administered the anti-emetic, anti-nausea drugs maropitant and metoclopramide, and H 2 receptor antagonist omeprazole. Mirtazapine was used as an appetite stimulant throughout the hospitalisation period and a single dose of vitamin B was given. Mild discomfort on abdominal palpation was noted on day two, but not at other times. Analgesia was provided in the form of buprenorphine and meloxicam ( Table 2). A nasogastric feeding tube was placed on day two of hospitalisation and was well tolerated by the patient. It was kept in place throughout the patient's hospital stay; he was discharged with it in place and the owners were trained to administer tube feeds. The patient was discharged on metoclopramide and amoxycillin-clavulanate, the latter administered due to increased, harsh respiratory sounds.
Once back at home, Cat 3A was noted to behave differently towards his owners and sibling, appearing not to recognise them; he was wary of his sibling, with whom he had previously slept and engaged in mutual grooming. From a week post-discharge onwards, Cat 3A began to resume more typical social behaviour with Cat 3B. The feeding tube was removed two weeks post-discharge, by which point the cat was eating approximately 25% of his normal food intake. A week later his food intake had increased to 65-75% of normal, at which time all medications were stopped, and by a further month later his intake was considered normal.
No swab samples were available from Cat 3A. Blood was collected close to the date of hospital discharge and neutralising antibody titres of 286, 198, 1424 and 55 were detected to the B1/D614G, Alpha, Delta and Omicron BA.1 variants, respectively, consistent with infection with the Delta variant. to her usual veterinary practice. Haematology and serum biochemistry showed mild changes mostly consistent with the cat's underlying medical issues and treatment. The cat was treated for presumed 'cat flu' with bromhexine hydrochloride (0.1mg/kg PO sid, Bisolvon oral power, Boehringer Ingelheim, Bracknell, UK).
Twenty-four hours later, the cat was taken to an emergency veterinary service as her condition had worsened, with increased lethargy, inappetence, difficulty swallowing and persistent reverse sneezing. Dexamethasone (0.08 mg/kg IV) was administered. The next day, she was examined at a referral hospital under sedation and increased upper respiratory sounds and lack of airflow through the left nostril were noted. The patient received fluid therapy and oxygen.
Computed tomography (CT) was performed under a general anaesthetic. This revealed a well delineated, partially obstructive nasopharyngeal mass measuring 0.9 × 0.5 × 1 cm. There was also mild, generalised bronchial wall thickening consistent with age, asthma, bronchitis or bronchopneumonia. Sampling of the nasopharyngeal mass led to the diagnosis of an abscess with dysplastic cells. Pasteurella multocida infection was diagnosed and treated with amoxycillin and clavulanic acid (Clavaseptin 50 mg, Vetoquinol, Towcester, UK) for eight weeks; the patient recovered from the respiratory disease during this time but experienced an episode of polyuria/polydipsia, haematuria and stranguria. A month later the polyuria/polydipsia was still present, and the patient had lost weight and was azotaemic. She was diagnosed with Stage 2 Kidney Disease by IRIS staging [46]. Additionally, corticosteroid therapy was discontinued, and a hydrolysed diet supplied. There was no recurrence of intestinal signs following this change, making inflammatory bowel disease the most likely cause of the cat's previous intestinal signs, rather than lymphoma.
SARS-CoV-2 RNA was not detected in a swab sample taken on day seven of the clinical signs. The patient showed neutralising antibody titres of 79, ≤50, ≤50 and 155 to SARS-CoV-2 variants B.1/D614G, Alpha, Delta and Omicron BA.1, respectively.

Discussion
Less than 1% of all swabs tested in the active surveillance part of this project were found to be positive. It was higher during the Delta wave (0.57%) compared to the first year of the pandemic (0.26% [26]). Additionally, 20% of swabs submitted specifically for SARS-CoV-2 testing were positive during the Delta wave. Although the number of positives was low, making it difficult to draw conclusions, an increase in test positivity had been expected, given the greater proportion of the UK human population infected during the Delta wave compared to earlier waves (Figure 1), as a result of the greater transmissibility of Delta compared to previous variants. This finding is consistent with the higher seroprevalence in animals reported in 2021 compared to 2020 in the UK [18] as well as the results of a study that identified several feline cases in SARS-CoV-2-positive households during the Delta wave in Switzerland [24].
A survey conducted by the Office for National Statistics (ONS) [43] estimated that the following percentages of the UK human population were infected when each variant was most common: pre-Alpha 7.0% (26 April-7 December 2020), Alpha 8.1% Given the large number of people infected since the emergence of Omicron BA.1, greater numbers of infected cats were predicted during this period compared to earlier periods. However, no RT-qPCR positive feline samples were detected despite screening 385 swabs across both sampling methods after the emergence of Omicron BA.1. This concurs with our serosurveillance data; BA.1 dominant antibody profiles were not observed when this variant displaced Delta [19].
Following an appeal to veterinary surgeons for samples from potential cases of SARS-CoV-2 infection, in line with the APHA case definition and guidelines [42], we accepted swabs for diagnostic testing. Early in the pandemic there were national protective measures and widespread human testing in place to reduce human-to-human transmission. This meant that people with current or recent household infections should not visit a veterinary surgeon unless there was an emergency. Consequently, the samples tested during this period were likely biased towards non-COVID-19 households and this could have accounted for the low number of positive swabs detected during the early waves of the pandemic. More recently, with the lifting of restrictions and the downturn in human testing, the likelihood of cats from infected households being taken to visit a veterinary surgery has increased. Despite these changed circumstances, no increase in feline SARS-CoV-2 infection prevalence was observed.
Cats experimentally challenged with Omicron BA.1.1 (B.1.1.529) were reported to remain sub-clinically infected, whereas cats challenged with B.1/D614G or Delta B.1.617.2 developed lethargy and pyrexia [47]. Likewise, seven naturally Omicron infected, RT-qPCR positive cats were reported to show no clinical signs [48]. While a proportion of the swabs submitted to VDS were collected from healthy animals, for example for prebreeding screening, most were from cats with clinical signs of ocular, respiratory or oral disease. It may be hypothesised that if Omicron-infected cats do not, in general, show clinical signs, they will rarely be sampled for diagnostic testing. Cat 4A, which displayed an Omicron-dominant antibody profile, had an atypical, severe presentation of respiratory distress. Before presenting with respiratory disease, this cat had been administered daily corticosteroids for bowel disease, which could have made her more susceptible to SARS-CoV-2 infection. It has also been reported that Omicron-inoculated cats shed less virus and have lower levels of neutralising antibodies than B.1/D614G-or Delta-inoculated cats [47]. It is therefore possible that when cats are, or have been, infected with Omicron, there is less evidence of infection than was the case with earlier variants. Lastly, cats may be less susceptible to contracting Omicron than earlier variants.
Where humans in a household had been tested by RT-qPCR and sequence data were obtained from their cat, we attempted to obtain sequence data from the human samples; only a proportion of positive swab samples from humans in the UK were sequenced by public health bodies during most of our study period. Unfortunately, there were no sequence data for humans in our case households, which precluded comparison of viral sequences from linked human and feline cases. When our feline-derived viral sequences were compared to published human-derived viral sequences, only one cat had a single, synonymous spike mutation not seen in human samples (Table A3). This suggests that the virus is, at least initially, well conserved following human-to-cat reverse zoonosis, presumably maintaining the likelihood of onwards transmission from cats to humans, a phenomenon which has been reported [6]. In case 2, the chronology of SARS-CoV-2 diagnostic detection in the household was human, human, cat, cat and finally human. It is therefore possible that the virus could have been transmitted from cat-to-human, but this can neither be confirmed nor refuted with the data available. It remains unclear whether sustained cat-to-cat transmission, or sustained transmission between cats and other nonhuman species, could lead to novel viral mutations or variants, as reported in white-tailed deer [49] and mink [50].
Many of the seropositive animals identified by our serosurveillance [19] showed no clinical signs, mild clinical signs or had signs likely unrelated to SARS-CoV-2 infection at the time of sampling. This finding was consistent with numerous reports of mild or subclinical feline infections [2][3][4][5]9,11,[20][21][22][23]. Additionally, seropositive animals might have experienced historical infections. However, in this study we also detected SARS-CoV-2 infections in animals with more severe disease, including individuals who suffered prolonged anorexia (Cat 3A), severe respiratory disease (Cat 4A) and sudden death (Cat 2A).
Of the four household cases described in detail, two of the cases (Cats 1A and 2A) tested positive by RT-qPCR, and viral sequencing confirmed the presence of SARS-CoV-2 variants that were circulating widely at the time of their illnesses. Cats in both households were seropositive, with the highest SARS-CoV-2 neutralising antibody titres evident against the variants detected in swabs. No swabs were tested from Cat 3A as it was assumed the timeframe for obtaining a positive SARS-CoV-2 RNA result had passed. However, this cat displayed a high neutralising antibody titre against the Delta variant, which was consistent with his illness and his owner's positive tests occurring during the Delta wave. Cat 4A tested negative by RT-qPCR but showed the highest neutralising antibody titre against Omicron BA.1, consistent with infection with the Omicron variant which dominated at the time of the owner's illness and the cat's respiratory signs. This cat may have tested positive for SARS-CoV-2 RNA if sampled earlier or, alternatively, might not have shed detectable levels of virus. Experimental studies have shown that cats shed the ancestral variant for approximately five days [5], and field cases with longer shedding periods were identified during the pre-Alpha and Alpha waves [9,22]. However, a more recent study [48] reported that cats naturally infected with Omicron were PCR-positive on only the first day of four to five days of consecutive sampling, shedding low levels of virus.
A combination of anorexia, hyporexia and inappetence is a common and non-specific presentation of illness in cats. Underlying causes and mechanisms include pyrexia, pain, inflammation, neoplasia and diseases of multiple organs or systems [51]. Many potential causes of Cat 3A's inappetence were excluded by diagnostic testing and imaging and no definitive diagnosis was reached. The timeline of Cat 3A's illness is consistent with SARS-CoV-2 infection being the underlying cause. Potential mechanisms include gastrointestinal infection, nausea and a lost or altered sense of smell and taste. The anorexia continued long after the diarrhoea had resolved, suggesting nausea and a lost or altered sense of smell were more likely at that stage; loss of smell and/or taste is a recognised symptom in SARS-CoV-2-infected humans. Nausea might also explain the lip licking of Cat 2B which took weeks to resolve. Anosmia has been documented to persist months after initial infection and to reduce appetite and cause weight loss in humans [52], and therefore might explain the prolonged anorexia of Cat 3A. It was reported that a working dog with neutralising antibodies to SARS-CoV-2 had suspected anosmia, suggesting anosmia might affect nonhuman species [53]. Anosmia could also explain Cat 3A's altered behaviour towards his sibling and owners. Indeed, SARS-CoV-2-infected people have reported a loss of intimacy with their partner due to not being able to detect their usual, natural scent [52].
As well as reporting lack of smell and/or taste, affected people have also reported that food smells disgusting and that they smell things that are not present, a phenomenon known as phantosmia. Once these clinical signs occur there is a possibility that learned food aversions will develop because the patient associates eating or food with bad smells or nausea [51]. Cat 3A was tube fed, which allows nourishment without coaxing or syringe feeding, both of which could induce secondary food aversion.
Unfortunately, a broad surveillance approach, such as the one taken in this study, cannot definitively demonstrate a causal link between the clinical signs reported and SARS-CoV-2 infection. No post mortem examination or tissue sampling of the cat (2A) that died suddenly were undertaken, so it is unknown whether cardiac lesions indicative of SARS-CoV-2 infection, or another disease process, were present. It was reported previously that a SARS-CoV-2-infected cat developed severe respiratory distress and thrombocytopaenia during a household infection, but it was concluded that an unrelated, pre-existing cardiomyopathy caused the observed pathology [54]. Here, RT-qPCR testing demonstrated that Cat 2A, a previously healthy middle-aged cat, was infected with SARS-CoV-2, which could cause, or exacerbate, cardiac pathology in cats [30][31][32]. SARS-CoV-2 related myocarditis and pneumonia was described in a cat with hypertrophic cardiomyopathy (HCM) [32], and sudden cardiac death in humans who appeared to be recovering from COVID-19 has been reported [55].
In addition, it was reported that a cat with a cardiac murmur developed severe and progressive respiratory disease after being exposed to SARS-CoV-2 [31]. Following a post mortem investigation, high loads of viral RNA and infectious virus were detected in the upper and lower respiratory tract and heart, and viral RNA was also detected in other organs. There was acute myocardial degeneration and necrosis, and viral particles were visualised in heart tissue, supporting a diagnosis of viral myocarditis. Pre-existing HCM and a pleural effusion were noted; the authors suggested that HCM, which can be subclinical, could be a risk factor for severe clinical signs in cats with SARS-CoV-2 infection due to the overexpression of ACE2 in HCM hearts.
Cat 4A developed respiratory distress following infection with SARS-CoV-2. This is consistent with other reports of SARS-CoV-2-induced respiratory pathology in cats and with imaging of Cat 4A revealing lower airway pathology. Cats with SARS-CoV-2-associated respiratory disease are mostly reported to have mild clinical signs, but more severe presentations have also been reported, including laboured breathing [25], pneumonia [28] and acute dyspnoea with air hunger and rales [29]. Cat 4A was later diagnosed with a bacterial upper airway infection and abscessation. Bacterial infections are a common sequel to viral respiratory infections; this might have been the case here, or the infections could have been co-incidental. Renal disease is common in cats, and Cat 4A's Stage 2 disease could also have been a co-incidental rather than related event. However, research has shown that humans are at an increased risk of acute kidney injury and chronic disease following SARS-CoV-2 infection, even if they were not hospitalised [56], and this cat's renal disease might therefore also be related to SARS-CoV-2 infection.
It can be difficult to demonstrate causal links between viral infection and observed clinical signs, particularly when the aetiopathogenesis of a viral disease is not yet fully understood, clinical presentations are diverse and potentially informative diagnostic materials such as tissue samples are not available from field cases. While associative links might be established using epidemiological data, this approach requires large numbers of cases and extensive testing for infection. Our findings suggest that more widespread testing of cats for SARS-CoV-2 infection would likely provide useful insights to the range of clinical signs associated with infection and determine where causal links exist. Continuing circulation of SARS-CoV-2 and the emergence of new variants in humans provide ongoing opportunities for reverse zoonosis. With the evidence published to date showing that different variants have different levels of infectivity and virulence in cats as well as people, ongoing monitoring of the feline population should be considered an essential aspect of a comprehensive SARS-CoV-2 surveillance programme.

Conclusions
This study demonstrated a low prevalence of RT-qPCR positivity in samples submitted for diagnostic testing from UK cats, and an unexpected dearth of cases after the emergence of the Omicron variant, despite its high infectivity in humans. The SARS-CoV-2 infected cats that were identified displayed a range of clinical signs that included mild respiratory disease, pyrexia, sudden death, chronic lip licking, chronic anorexia, respiratory distress and chronic renal disease.  Libraries were sequenced in a R9.4.1 flow cell (Oxford Nanopore Technologies, Oxford, UK) using MinKNOW version 21.05.25, raw FAST5 files were base called using Guppy. For Cat 1A Guppy version 4.0.11 was used with high-accuracy mode and a minimum quality score of 7, whilst for Cats 2A and Y version 5.0.16 was used with super-accurate mode and a minimum quality score of 10; in both cases, barcodes were required at both ends for demultiplexing. Nanopore reads were processed using the ARTIC network artic-ncov2019 bioinformatics protocol and conda environment (https://github.com/artic-network/artic-ncov2019, accessed on 14 May 2021). Briefly, reads are size filtered, mapped against the reference strain Wuhan-Hu-1 (MN908947), amplicon primer trimmed, variant called with Nanopolish, followed by consensus sequence generation. Phylogenetic subtrees for Cats 1A, 2A and X are presented in Figures A1-A3. Cat Y only yielded sequence reads from one amplicon in ORF1ab, with no mutations present with respect to the reference sequence, and therefore a consensus sequence could not be generated, nor the lineage determined. Table A3. Spike mutations present in each feline SARS-CoV-2 genome sequence with respect to the ancestral Wuhan-Hu-1 reference sequence. The synonymous mutation L821L (highlighted in red) was the only mutation not seen in phylogenetically close human SARS-CoV-2 sequences.

. Phylogenetic and Mutation Analysis of Cat Sequences
Each of the three feline SARS-CoV-2 genome sequences were uploaded to the UShER (Ultrafast Sample placement of Existing tRee) [59] web server (https://genome.ucsc.edu/cgi-bin/hgPhyloPlace, accessed on 7 July 2023) to analyse the sequence neighbourhood surrounding each cat's sample. For each feline SARS-CoV-2 genome sequence, mutations with respect to the Wuhan-Hu-1 reference sequence were compared to those from the surrounding human population. Phylogenetic subtrees for Cats 1A, 2A and X are presented in Figures A1-A3. Cat Y only yielded sequence reads from one amplicon in ORF1ab, with no mutations present with respect to the reference sequence, and therefore a consensus sequence could not be generated, nor the lineage determined. Figure A1. Cat 1A SARS-CoV-2 UShER phylogenetic local subtree using sequences from the public databases, with the cat sequence highlighted in red. Figure A1. Cat 1A SARS-CoV-2 UShER phylogenetic local subtree using sequences from the public databases, with the cat sequence highlighted in red.  Figure A2. Cat 2A SARS-CoV-2 UShER phylogenetic local subtree using sequences from the public databases, with the cat sequence highlighted in red. Figure A2. Cat 2A SARS-CoV-2 UShER phylogenetic local subtree using sequences from the public databases, with the cat sequence highlighted in red.

Appendix B: Supplementary Information
Viruses 2023, 15, 1769 17 of 21 Figure A3. Cat X SARS-CoV-2 UShER phylogenetic local subtree using sequences from the public databases, with the cat sequence highlighted in red.  Figure A3. Cat X SARS-CoV-2 UShER phylogenetic local subtree using sequences from the public databases, with the cat sequence highlighted in red.