Article Text


Isolation of avian strains of Salmonella enterica serovar Typhimurium from cats with enteric disease in the United Kingdom
  1. A. W. Philbey, BVSc, PhD, MACVSc(Path), MRCVS1,
  2. D. J Taylor, MA, VetMB, PhD, DipECVPH, DipECVPHM, MRCVS2,
  3. H. A. Mather, MSc, CSci, FIBMS3 and
  4. J. E. Coia, BSc, MD, FRCPath, DipHIC3
  1. 1 Division of Pathological Sciences
  2. 2 Division of Animal Production and Public Health, University of Glasgow Veterinary School, Bearsden, Glasgow G61 1QH
  3. 3 Scottish Salmonella Reference Laboratory, Stobhill Hospital, North Glasgow University Hospitals Division, 133 Balornock Road, Glasgow G21 3UW

Statistics from

AVIAN-ADAPTED strains of Salmonella enterica serovar Typhimurium have been detected in wild birds, particularly finches (family Fringillidae), in Europe, Scandinavia and North America (Daoust and others 2000, Tauni and Österlund 2000, Pennycott and others 2006). In the uk, these strains of S Typhimurium include phage types (definitive types) dt2, dt40, dt41, dt56 variant (dt56v), dt99 and dt195 (Pennycott and others 2006). Wild bird strains have been isolated at relatively low frequency from domestic livestock in the uk and appear to be uncommon in human beings, cats and dogs (Pennycott and others 2006).

Salmonellosis in cats has been associated with a wide variety of serovars, most frequently S Typhimurium (Shimi and Barin 1977, Dow and others 1989). An established source of infection with S Typhimurium in cats is from the consumption of raw meat (Stiver and others 2003). Feline salmonellosis has also been called ‘songbird fever’, reflecting a long-recognised association with the predation of birds (Scott 1988). An outbreak of gastroenteritis due to S Typhimurium dt40 in cats and human beings in central Sweden in 1999 was associated with infection and mortality in wild birds (Tauni and Österlund 2000). This short communication reports the isolation of wild bird strains of S Typhimurium from nine cats with enteric disease in the uk, and also records the isolation of these strains from human beings in Scotland.

S Typhimurium phage types dt40 or dt56v were isolated from faecal samples or intestinal contents of nine cats, aged one to 14 years, with enteric disease in the uk from 2003 to 2007 (Table 1). All cats had a history of hunting small birds. Affected cats were usually dull, pyrexic, inappetent or anorexic and often developed diarrhoea, sometimes with dysentery, for periods of 36 hours to three weeks. In addition to S Typhimurium, β-haemolytic Escherichia coli was recovered from the faeces of cats 2 and 3 and Clostridium perfringens was recovered from the faeces of cat 3. However, S Typhimurium was considered to be the primary pathogen in these cases. No blood or tissue samples were available for bacterial culture from any of the affected cats.

Six of the seven S Typhimurium dt56v isolates from the cats were sensitive to ampicillin (50 μg/ml), cefotaxime (1 μg/ml), chloramphenicol (20 μg/ml), ciprofloxacin (0·5 μg/ml), furazolidone (20 μg/ml), gentamicin (20 μg/ml), kanamycin (20 μg/ml), naladixic acid (40 μg/ml), netilmicin (20 μg/ml), spectinomycin (100 μg/ml), streptomycin (20 μg/ml), sulphonamide (100 μg/ml), tetracycline (10 μg/ml), trimethoprim (2 μg/ml) and ciprofloxacin (0·125 μg/ml) by the breakpoint (minimum inhibitory concentration) method (Frost 1994). The dt56v isolate from cat 8 was resistant to kanamycin and streptomycin, the dt40 isolate from cat 3 was resistant to sulphonamide and streptomycin, and the dt40 isolate from cat 4 was resistant to sulphonamide and trimethoprim.

Eight of the cats recovered following treatment with antibiotics and supportive care, but one (cat 4) died before treatment could be instigated (Table 1). Clinical improvements were noted in five cases following treatment with clavulanate-potentiated amoxicillin (Synulox; Pfizer) alone. Cat 1 was initially treated with clavulanate-potentiated amoxicillin by injection and then treatment was continued with oral marbofloxacin (Marbocyl; Vétoquinol). Cat 3 showed clinical improvement following initial treatment with amoxicillin (Duphamox; Fort Dodge), followed by treatment with clavulanate-potentiated amoxicillin. Cat 5 initially failed to respond to clavulanate-potentiated amoxicillin and treatment was then changed to enrofloxacin (Baytril; Bayer) one week later, which could explain the more prolonged clinical course in this cat. Repeat faecal samples from cat 8 were positive for S Typhimurium at one week after the onset of clinical signs, but negative at three weeks. Faecal samples were negative by culture for Salmonella species when collected from cat 2 four weeks after the initial clinical presentation, and from Cat 5 two weeks after presentation. No follow-up testing was performed on the other five surviving cases.

Cat 4 was a two-year, five-month-old female domestic shorthair from Lochinver in north-west Scotland. In early January 2006, it became dull and anorexic for 24 hours. The cat exhibited signs of distress and abdominal discomfort and then vomited unidentified yellow fluid. It developed seizures and died 36 hours after the onset of clinical signs. At postmortem examination, there was dilatation and thickening of several 1 cm segments of the jejunum and a 3 cm segment of the ileum. Portions of intestine and lymph node were submitted for histological examination. The mucosa of the ileum was eroded and necrotic, with destruction of villi and crypts and exudation of neutrophils and necrotic cellular debris into the intestinal lumen (Fig 1a). The intestinal lamina propria was extensively infiltrated with neutrophils and sheets of macrophages. Bacteria were present on the luminal surface of the mucosa and also infiltrated the lamina propria. There was depletion of lymphoid cells from Peyer's patches in the ileum, with replacement by infiltrating macrophages. Less severe changes were present in the caecum. The subcapsular, intertrabecular and medullary sinuses of the ileocaecal lymph node were infiltrated with macrophages and lesser numbers of neutrophils (Fig 1b) and there was focal necrosis in some areas. The intestinal contents were negative for feline parvo-virus by pcr. S Typhimurium dt40 was recovered in profuse growth from the small intestine. The pyogranulomatous necrotising enteritis and pyogranulomatous lymphadenitis in this cat were consistent with fatal enteric salmonellosis.


(a) Pyogranulomatous necrotising ileitis and (b) pyogranulomatous lymphadenetis in a cat (cat 4) with enteric salmonellosis due to Salmonella enterica serovar Typhimurium phage type dt40. Haematoxylin and eosin. Bars=50μm

Since 2001, the Scottish Salmonella Reference Laboratory has identified 47 isolates of S Typhimurium dt40 and 29 isolates of S Typhimurium dt56v from human beings. All of the dt40 isolates and 27 (93 per cent) of 29 isolates of dt56v were fully sensitive to all antibiotics tested routinely; 52 per cent of the S Typhimurium dt56v isolates and 38 per cent of the S Typhimurium dt40 isolates were obtained from children under five years of age.

dt40 and dt56v are considered to be avian-adapted strains of S Typhimurium, with a relatively narrow host range and a wide geographical distribution (Rabsch and others 2002). S Typhimurium dt40 and dt56v have been isolated from European greenfinches (Carduelis chloris), house sparrows (Passer domesticus) and chaffinches (Fringilla coelebs), dt40 has been isolated from European goldfinches (Carduelis carduelis), dt56v has been isolated from tree sparrows (Passer montanus) and both strains have been isolated occasionally from other species of birds in the uk (Pennycott and others 1998, 2006). The outbreak of gastroenteritis due to S Typhimurium dt40 in cats and human beings in central Sweden in 1999 was associated with infection and mortality in wild birds, notably common redpolls (Carduelis flammea) and Eurasian siskins (Carduelis spinus) (Tauni and Österlund 2000).

S Typhimurium dt40 has also been implicated in epizootics of mortality in wild birds from winter 1997 to summer 1998 in eastern North America, including the Canadian Atlantic maritime provinces of New Brunswick, Novia Scotia, Prince Edward Island, Newfoundland and Labrador (Daoust and others 2000). The species most commonly affected in these areas were common redpolls, pine siskins (Carduelis pinus), purple finches (Carpodacus purpureus), evening gros-beaks (Coccothraustes vespertinus) and American goldfinches (Carduelis tristis) (Daoust and others 2000). Recurrent winter outbreaks of salmonellosis due to S Typhimurium dt160 have been identified in house sparrows in New Zealand (Alley and others 2002, Connolly and others 2006). Human cases of salmonellosis due to dt160 were associated with direct or indirect contact with birds (Alley and others 2002, Thornley and others 2003).

In the present investigation, the seven feline isolates of S Typhimurium dt56v were recovered from September to February, which closely overlaps the most frequent months of isolation of this phage type from wild birds in the uk (November to March) (Pennycott and others 2006). The two dt40 isolates were recovered from cats in December and January, which also closely overlaps the most frequent months of isolation of this phage type from wild birds in the uk (January to March). Outbreaks of salmonellosis in wild birds appears to be associated with congregation around artificial feeding stations such as bird tables in the winter (Pennycott and others 1998, 2005). Cats are likely to be exposed to S Typhimurium through predation of infected birds at these sites during the cooler months. The cats in this investigation were located in rural or suburban areas and had histories of hunting wild birds, which represent the likely source of infection. The two isolates of dt40 were obtained from cats in northern Scotland and Northern Ireland, whereas dt56v isolates were recovered from cats in southern Scotland and northern England. Although the number of isolates from cats is relatively small, these geographical trends are similar to those observed in birds in the uk (Pennycott and others 2006).

Healthy cats, as well as sick cats without enteric disease, may carry S Typhimurium, although the reported prevalence of infection varies widely (Hill and others 2000, Spain and others 2001, Van Immerseel and others 2004). This investigation demonstrates that avian strains of S Typhimurium cause diarrhoea and mortality in cats in the uk, probably associated with predation of birds at feeding stations during the cooler months. Cats have been recognised as a source of antimicrobial-resistant S Typhimurium (Wall and others 1995, Van Immerseel and others 2004), but the isolates of S Typhimurium dt40 and dt56v from the present cases were not highly multidrug-resistant strains. Human infection with S Typhimurium dt40 has been demonstrated in Sweden (Tauni and Österlund 2000) and, as reported here, human infections with S Typhimurium dt40 and dt56v have been demonstrated in Scotland. Isolates from human beings in this study were generally sensitive to commonly used antibiotics. Avian strains of S Typhimurium should therefore be regarded as potentially zoonotic. Bird feeding stations may play a central role in the epidemiology of avian strains of S Typhimurium in birds, cats and human beings.


The authors thank Audrey Adams, Patrick Argyle, Gregory Dunlop, Sarah Harker and Tim Pearson for submitting samples and providing clinical details for the affected cats. Bacterial culture and identification were performed by Kathleen Reynolds and Manuel Fuentes, and histological samples were processed by Lynn Stevenson at the University of Glasgow Veterinary School.


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.