Capnocytophaga canimorsus meningitis and bacteraemia without a dog bite in an immunocompetent individual

Abstract

We describe the case of an immunocompetent 75-year-old man with Capnocytophaga canimorsus bacteraemia and meningitis. C. canimorsus is commonly found in the oral flora of dogs with human infection typically occurring following a bite. Unusually, while our patient was a dog owner, there was no history of bite nor scratch mark. Admission blood cultures flagged positive for Gram-negative bacilli, but prolonged molecular analysis was required before C. canimorsus was isolated in blood and cerebrospinal fluid. There is a high mortality rate in invasive infection, and in our patient’s case, antibiotic therapy was commenced prior to laboratory confirmation with our patient making a complete recovery. This case highlights the importance of including C. canimorsus in the differential diagnosis of unwell patients who keep dogs, even without a bite. This case occurred amid heightened awareness of COVID-19, which may represent predisposition for zoonoses during social isolation and increased human–pet contact.

Background

Capnocytophaga canimorsus is a commensal bacterium of the oral flora of dogs and cats. It can be transmitted to humans by bite, scratch or skin abrasion or less commonly as in our patient’s case by exposure to animal saliva.¹ The incidence of C. canimorsus infection is low; reported as 0.67–4.1 per million population in two studies.^{2 3} Infections in humans are usually wound infections, but it can cause bacteraemia and rarely meningitis.⁴ At the time of writing, there was only one other case of C. canimorsus meningitis reported in Ireland.⁵ While infections with this pathogen were initially reported as occuring in those with immunodeficiency such as asplenism, alcohol abuse, smoking, corticosteroid therapy and haemato-oncological diseases; more recently, cases among immunocompetent persons have been described.⁶ Prompt treatment with antibiotic therapy is crucial in invasive infection due to the exceptionally high mortality rate of up to 30% with C. canimorsus bacteraemia.⁷ Diagnosis of C. canimorsus may be delayed due to slow growth using traditional culture techniques that can result in the diagnosis being missed.⁸ We report the case of a 75-year-old immunocompetent man with C. canimorsus meningitis and bacteraemia. We present this case to highlight the importance of considering C. canimorsus infection in patients with exposure to dogs even where there is no evidence of inoculation by bite or scratch to ensure early treatment of this infection despite potential delays in diagnosis.

Case presentation

A 75-year-old man with no significant previous medical history presented in June 2020 to the emergency department (ED) of a city-centre hospital having been found by his son with confusion, abdominal pain, faecal incontinence and fever. Prior to admission, he was fit and well, independent and living alone. He was last seen well by his neighbours 2 days previously out walking his dog who he walked everyday for seven miles. He was admitted under the infectious diseases team. On examination, he was feverish with a temperature of 39.2°C, respiratory rate of 28 bpm, blood pressure of 155/84 mm Hg and heart rate of 98 bpm. His Glasgow Coma Scale was 11/15 and there was no focal neurological deficit on examination. Abdominal palpation revealed mild generalised abdominal tenderness with no features of an acute abdomen.

Investigations

Admission blood tests showed raised inflammatory markers, white cell count 13.3 x 10⁹/L (reference range 4–11 x 10⁹/L), C reactive protein >320 mg/L (reference range 0–5 mg/L). Arterial blood gases on room air showed a respiratory alkalosis, PH 7.56 (range 7.35–7.45), PaCO2 3.0 (range 4.7–6 kPa), PaO2 11.7 (range 11–14 kPa), bicarbonate 20.6 (range 22–26 mmol/L), lactate 1.5 (range <1.6 mmol/L). The aerobic blood culture collected on admission flagged positive after 72 hours and small Gram-negative bacilli were identified, the anaerobic bottle remained sterile. Agar plates were set up from the aerobic bottle and incubated in air, CO2 and anaerobic conditions. Growth was only observed on the blood agar plate, which was incubated in anaerobic conditions after 72 hours. No pathogens were detected using in-house molecular methods;Biofire FilmArray Blood Culture ID panel. Initial attempt at identification using MALDI-TOF after 48 hours of incubation failed. A further 5 days of incubation in anaerobic conditions was required before the organism was successfully identified as a C. canimorsus using MALDI-TOF. The isolate failed to grow for antibiotic susceptibility testing. Stool cultures were negative for bacterial pathogens including Salmonella, Campylobacter and Shigella. Mid-stream urine cultures were negative. Chest X-ray revealed no abnormality. CT brain showed no intracranial mass, haemorrhage or infarct with evidence of inflammatory mucosal thickening in ethmoid, sphenoid and frontal sinuses. CT abdomen and pelvis showed no intra-abdominal source of infection. Due to the patient’s ongoing confusion and drowsiness, a lumbar puncture was performed. Cerebrospinal fluid (CSF) analysis demonstrated an elevatedwhite cell count of 914 x 10⁶/L (reference range 0-5 x 10⁶/L), but no organisms were identified on Gram stain or using the in-house Biofire FilmArray Meningitis Encephalitis panel. CSF was culture negative after prolonged incubation. CSF was referred to an external laboratory for 16S ribosomal RNA (rRNA) PCR gene sequencing and C. canimorsus was identified.

Admitting ECG revealed previously undiagnosed atrial fibrillation. A transthoracic echocardiogram showed a severely dilated left atrium and ventricle with global hypokinesia and a severely reduced ejection fraction of 20%–25% in the setting of his severe sepsis. He had no evidence of valvular vegetations.

Differential diagnosis

On the patient’s admission to ED, given his myriad symptoms and signs including fever, abdominal pain, confusion, tachycardia, tachypnea; the immediate presumed diagnosis was sepsis of unknown source. The main differential diagnoses at this point included sepsis secondary to intrabdominal pathology, respiratory infection, intracranial infection or urosepsis, and a work-up was undertaken to investigate his syndrome. SARS-CoV-2 swab was negative and his admission chest X-ray revealed no obvious lesions. By day 2 of admission, with stool, blood and urine cultures remaining negative, CT abdomen was performed which showed no pathology to account for the patient’s symptoms. We then proceeded to focus on meningitis as a possible differential diagnosis, given the patient’s ongoing confusion. Lumbar puncture demonstrated meningitis with raised CSF white cell count of 914 x 10⁶/L ((reference range 0-5 x 10⁶/L) (white cell count differential not available)) and a markedly raised CSF protein of 2 g/L (reference range 0.15-0.45 g/L), although a FilmArray panel for common bacteria and viruses was negative and CSF remained culture negative. We subsequently received notification that his blood cultures were positive for a Gram-negative bacillus although a formal identification was not available at that point. After prolonged incubation of the blood culture bacteria using anaerobic conditions as well as 16S rRNA PCR gene sequencing on CSF by an external laboratory, a final diagnosis was made of C. canimorsus bacteraemia and meningitis associated with domestic dog exposure.

Treatment

Due to the unclear diagnosis, initial antibiotic therapy started in the ED was broad spectrum and included intravenous ceftriaxone 2 g two times per day, metrondiazole 500 mg three times per day and aciclovir 800 mg three times per day. This was combined with intravenous crystalloid fluid resuscitation. On receipt of positive blood cultures for Gram-negative bacilli along with the subsequent elevated white cell count in CSF with Herpes simplex virus DNA not detected on PCR by film array, the antibiotic therapy was changed to intravenous meropenem 2 g three times per day. This led to a subsequent improvement of the patient’s overall status and to a decrease in serum inflammatory markers. On receipt of confirmed growth of C. canimorsus in blood cultures on day 3, antibiotic therapy was switched to intravenous ceftriaxone 2 g two times per day, which was continued for total of 3 weeks of intravenous antibiotic therapy.

Within 3 days of commencement of antibiotic therapy, the patient was completely asymptomatic of presenting symptoms and was apyrexial. His confusion on admission had resolved entirely.

He was discharged home from hospital 21 days after initial presentation.

Outcome and follow-up

At 1-month post discharge, he was reviewed in the infectious diseases outpatient clinic. At this point, he was very well and reported being back to his normal activities including walking his dog everyday. He was discharged from the infectious diseases clinic. He is due to be followed up with the cardiology team for a repeat echocardiogram in the next few months.

Discussion

C. canimorsus is a Gram-negative bacillus and facultative anaerobe that is part of the normal gingival flora of cats and dogs. It was first reported as causing infectious syndromes in humans in 1976 and can cause a wide range of clinical syndromes including cellulitis, ocular infections, septicemia, meningitis, endocarditis and disseminated intravascular coagulation.^{2 9} Although rare, case reports of C. canimorsus infections have been reported worldwide. Men are more commonly infected than women with an age range of 4 months to 77 years.⁶ Patients with a history of immunosuppression, asplenia, alcoholism, haematological malignancies and cirrhosis are at higher risk of infection but there have been case reports in immunocompetent patients.^{6 10}

Over half of infections with C. canimorsus are due to an animal bite, 8.5% caused by scratches with varying rates up to 27% by exposure to animals.⁶ Our patient denied any history of nor was there any clinical evidence of inoculation with a bite or scratch from the patient’s dog; however, he did describe the dog as being very affectionate and regularly licking him and we would have to presume that this exposure was the route of transmission. The vast majority of bites, scratches or exposure to animals do not lead to hospital admission and only few patients develop C. canimorsus septiciemia or meningitis. C. canimorsus septicemia is exceptional in its extremely high mortality rate of up to 30% whereas the mortality rate for reported cases of C. canimorsus meningitis is much lower at 5%.^{6 11}

The median time from exposure to sepsis is 3 days and 7 days for meningitis.^{5 12} Early identification of infection can be difficult. Barriers to identification in the microbiology laboratory that have been suggested include the lack of familiarity with the organism, its growth requirements and inability to identify it in some laboratories resulting in an overall correct identification rate suggested to be about 32%.¹³ Blood cultures and wound cultures do not commonly identify the organism and molecular techniques such as 16s rRNA gene sequencing are increasingly being used for identification.¹⁴ Of note, meningitis has been much less commonly described than septicemia, and it has been suggested that diagnosis may be delayed due to poor or slow growth using traditional culture techniques.⁸ In our patient’s case, it was through peripheral blood cultures and 16s rRNA gene sequencing of CSF that the pathogen was identified as causing the patient’s bacteraemia and meningitis. It should be noted that initial culture of the CSF fluid was negative, and it was only on further PCR testing that the pathogen was identified. Although diagnosing C. canimorsus can pose challenges, it is generally susceptible to a broad range of antimicrobial agents including beta lactams and third-generation cephalosporins.¹⁵ In our patient’s case, treatment with a third-generation cephalosporin was effective in eradicating the pathogen.

This case occurred shortly after the first wave of COVID-19 in Ireland, during a time of ongoing social distancing and lifestyle modifications, and may represent an increased predisposition for zoonoses during times of social isolation where human–pet contact may be increased.

Intellectual property rights assignment or licence statement

Learning points.

• Capnocytophaga canimorsus is a rare yet significant cause of sepsis. Human infection typically follows a bite or scratch and less commonly following contact with animal saliva such as from dog licking. In our case, there was no evidence of inoculation with neither bite nor scratch mark. This case highlights the importance of a history of animal exposure, which can raise clinical suspicion and result in early diagnosis and treatment.

• Our case also reminds us that while immunocompromised patients are at higher risk of infection, immunocompetent patients can also be susceptible.

• Increased awareness of C. canimorsus as a cause of septicemia and meningitis is of paramount importance for patient survival, given the high mortality rate associated with this infection.

• Clinicians should be aware of additional diagnostic assays such as 16s ribosomal RNA gene sequencing, which can aid significantly in the detection of pathogens from culture-negative sites, where a high index of clinical suspicion for infection remains. As in this case, detection of the C. canimorsus from cerebrospinal fluid ensured that the patient received the correct antibiotic for the correct duration and can aid antimicrobial stewardship efforts within the hospital setting.

Ethics statements

Patient consent for publication

Obtained.