OUP user menu

Helicobacter canis bacteraemia in a 7-month-old child

Jørgen Prag, Jens Blom, Karen A. Krogfelt
DOI: http://dx.doi.org/10.1111/j.1574-695X.2007.00271.x 264-267 First published online: 1 July 2007


On the basis of biochemical, phenotypic and 16S rRNA analyses, Helicobacter canis was isolated and identified from an otherwise healthy 7-month-old girl with intermittent fever. Blood cultures signalled bacterial growth after 5 days that was characterized as small gram-negative spiral rods. Subculturing on Colombia plates with 5% sheep blood, chocolate agar and brucella agar, aerobically and anaerobically as well as in a microaerophilic atmosphere, showed scanty growth after an additional 4 days. Secondarily seeded with fluid from the original bottle, the paediatric blood bottles repeatedly signalled growth after one night's incubation, whereas the conventially treated bottles did not support growth after 7 days' incubation. From the secondary seeded paediatric bottles a pure culture was isolated on chocolate agar plates, and identified as H. canis. This case indicates that blood culture systems should be compared and improved for their capacity to detect Helicobacter and related pathogenic bacteria species. Further studies are also needed to determine the importance of H. canis as a primary pathogen, and the role of cats in the possible zoonotic spread of H. canis to humans.

  • Helicobacter canis
  • paediatric blood culture
  • infant


The genus Helicobacter includes at least 26 formally named species (Whary & Fox, 2004). The tropism of the genus Helicobacter ranges from the stomach, cecum and colon to the liver or genital tract of mammals and birds. Helicobacter spp. comprise a diverse group of potentially emerging pathogens in multiple species, ranging from humans and nonhuman primates to pets, household animals and a variety of wildlife (Whary & Fox, 2004). Therefore great attention should be paid to the development of suitable media in order to grow these bacteria.

The original descriptions of Helicobacter canis were from the faeces of healthy and diarrheic dogs (Stanley et al., 1993), a child with enteritis (Burnens et al., 1993), Bengal cats with diarrhoea (Foley et al., 1999) and a dog with severe hepatitis (Fox et al., 1996).

Given the biochemical characteristics and the diversity of this microaerophilic bacterial genus, the requirements for diagnostic media is high. Conventionally, automated blood culture systems are used in hospital settings. The chances to vary culturing parameters are therefore very limited and this restricts the development of new culture media.

We demonstrate that the media in the standard blood culture bottles can be improved with meat or yeast extract in order to detect Helicobacter septicaemia in addition to the already covered spectrum of pathogenic bacteria.

Furthermore, additional studies are needed to determine the importance of H. canis as a human pathogen, and the role of cats in the possible zoonotic spread of H. canis to humans.

Materials and methods

Case history

An otherwise healthy 7-month-old girl was admitted to a paediatric department after 3 days of intermittent fever and four acrocyanotic attacks. She was initially treated with intravenous ampicillin and gentamicin, followed by oral mecillinam for 10 days, against a urinary tract infection with susceptible Escherichia coli. Neutrophils 15 × 109/L and CRP 45 mg L−1 normalized 5 days after admission. The patient recovered but still experienced fever spikes. The family has a healthy cat, but no contact with dogs.

Bacterial culture

Automated blood-culture system bottles Bactec Peds Plus/F (Becton Dickinson, USA) are used routinely for children weighing <10 kg. The bottles are then incubated for 7 days. Colombia plates with 5% sheep blood (BD), chocolate agar (SSI) and brucella agar (Oxoid), as well as 5% horse blood agar plates with yeast extract (SSI), under both aerobic and microaerophilc conditions were used for subculturing. Reagents for biochemical tests were all obtained from SSI Diagnostica (e.g. catalase, oxidase, urease, indoxyl acetate and nitrate).

Electron microscopy

Electron microscopy was carried out using a FEI Morgani D268 electron microscope at 80 kW. Bacteria grown for two days were harvested from blood agar plates and gently resuspended in distilled water. Bacterial ultra-structure was examined by the negative staining technique using 1.5% phosphotungstic acid.

16S rRNA analysis

Sequencing was performed on lysates of the pure culture using the MicroSeq® ID kit (Applied Biosystems, USA) as previously described (Fontana et al., 2005). A 527-bp fragment of the 16S rRNA gene of the bacterial strains was amplified. Amplified products were purified prior to sequencing. Forward- and reverse-sequencing reactions were performed for each amplified product. All sequencing analyses were performed in an ABI Prism 310 genetic analyzer.

Sequence data analysis

Sequencing data were analysed using microseq software. The analysis steps include assembly at the forward and reverse sequences into a consensus sequence, editing of the consensus sequence to resolve discrepancies between the two strands, and comparison of the consensus sequence in the microseq database. The database contains full-length 16S rRNA gene sequences for more than 1400 different bacteria. In most cases, each species was represented by the type strains, thus ensuring that the database was created using the ‘most typical’ strain of the species. The database comparison, using the full alignment tool of the microseq 500 software, generates a list of the closest matches with a distance score. This score indicates the percentage difference between unknown sequences and the database sequence. The consensus sequences were also compared with universal databases in the NCBI data bank. No incongruence with microseq ID was found.


Bacterial culture and identification

After 5 days' incubation, the blood-culture Bactec Peds Plus/F taken at admission showed growth of small, immobile, gram-negative spiral rods, 2–3 µm long (Fig. 1, Table 1). Secondary inoculation from the original bottle on Colombia plates with 5% sheep blood (BD), chocolate agar and brucella agar, aerobically and anaerobically as well as in a microaerophilic atmosphere, showed scanty growth with very small colonies after 4 days.

Figure 1

Spiral rods, as seen by electron microscopy, from the Bactec Ped Plus Bottle.

View this table:
Table 1

Ingredients in the blood culture media used, and growth of Helicobacter strains in the three blood-culture flasks. Major differences are indicated in bold type.

Bactec blood-culture
IngredientsPEDS (% w/v)Aerobic (% w/v)Anaerobic (% w/v)
Processed Water (mL)402525
Soybean-Casein Digest Broth2.752.752.75
Yeast Extract0.250.250.4
Animal Tissue Digest0.100.05
Sodium Pyruvate0.10
Sodium Polyanethol Sulfonate (SPS)
Pyridoxal HCI (Vitamin B6)0.0010.001
Sodium Citrate0.02
Potassium Phosphate0.24
Nonionic Adsorbing Resin10.016.016.0
Cationic Exchange Resin0.61.01.0
Growth of
    H. pylori+
    H. cinaedi++
    H. canis+

Secondary seeded Bactec Peds Plus (BD) from the original bottle including patient blood repeatedly signalled growth after one night's incubation at 35°C, whereas the Bactec Aerobic and Anaerobic Plus signalled no growth after 7 days' incubation, confirmed by microscopy. Bacterial growth was improved when using 5% horse blood agar plates with yeast extract (SSI). Secondary seeded bottles from agar plates showed growth in Peds Plus after 12 days but no growth was seen in the aerobic or anaerobic bottles after 14 days.

Direct examination of the colonies revealed the presence of curved gram-negative rods. The isolates grew at 37 and 42°C. The organisms were oxidase-positive, and catalase, urease and indoxyl acetate negative, neither reduced nitrate to nitrite nor hydrolyzed hippurate, but they did grow in the presence of 1.5% bile. The isolates were resistant to cephalothin (30 µg) but sensitive to nalidixic acid (30 µg). The organisms isolated from blood were provisionally identified as H. canis.

Electron microscopy of the pure culture revealed a helicobacter-like organism with bipolar flagella and a characteristic sheath (Fig. 2a and b).

Figure 2

Pure culture of Helicobacter canis isolated from blood. (a) Bacterium with bipolar flagella. (b) Flagella with characteristic sheath.

Molecular identification

16S rRNA sequencing was performed on the pure culture to confirm the biochemical findings. The sequence of the human isolate had a 99% similarity with the type strain of H. canis (L13464).


To our knowledge, this is one of the few reports of H. canis infection in humans. Helicobacter canis has been previously reported in a child with gastroenteritis (Burnens et al., 1993), dogs with and without diarrhoea (Stanley et al., 1993), and a puppy with necrotizing hepatitis (Fox et al., 1996).

Culturing of Helicobacter species has been a difficult task and it was first achieved in the beginning of the 1980s (Marshall & Warren, 1983). During the last three decades the development and use of molecular techniques in research and diagnosis has been favoured. Unfortunately, this has caused a stagnation in the development of commercially available culture media.

It has been shown that H. pylori cannot grow in ordinary Bactec blood culture systems but can grow in brucella broth with the detergent SPS (Sodium Polyanethol Sulfonate) (Kehler et al., 1994) (Table 1). However, H. cinaedi has been isolated after 5 days' incubation from aerobic Bactec bottles (Orlicek et al., 1993; Burman et al., 1995).

Our strain of H. canis, probably the first strain isolated from a human blood-culture, did not grow in the standard aerobic or anaerobic bottle but in the paediatric bottle supplemented with extra yeast extract and animal tissue digest with less SPS. Our findings indicate that the media in the standard blood culture bottles should be improved with meat or yeast extract as in the paediatric bottle in order to stimulate growth of enteric Helicobacter spp. in routine blood cultures. This case demonstrates that the media in the standard blood culture bottles could probably be improved, at a low cost, with meat and/or more yeast extract as in the paediatric bottle in order to detect Campylobacter and Helicobacter septicaemia in addition to the already covered spectrum of pathogenic bacteria. Addition of H2 is another possibility. Further studies for improving culture media for diagnostics are urgently needed.


The authors would like to thank Berit Jensen for excellent technical assistance and Eva Litrup for performing the 16S rRNA sequencing analysis.


View Abstract