Introduction

Capnocytophaga canimorsus, known formerly as “dysgonic fermenter-2”, is part of the normal flora in the mouths of dogs and, less commonly, cats. Data suggest that nearly a quarter of dogs carry this organism in their mouths with a lower proportion among cats [1]. Transmission to humans may occur via bites, scratches, or licking; occasionally cases have been reported after non-contact-related exposures. Clinical infections caused by C. canimorsus include bacteremia, septic shock, arthritis, endocarditis, and rarely meningitis [2, 3]. While Capnocytophaga is thought to be pathogenic primarily among immunocompromised persons, especially those with a prior splenectomy, cases among immunocompetent persons have been described [3].

Capnocytophaga is a fastidious, gram-negative bacillus. Diagnosis can be made by culture; however, since the organism grows slowly (2–7 days) and requires blood or chocolate agar incubated with 10% carbon dioxide, the diagnosis can be missed. Polymerase chain reaction (PCR) can be a valuable tool in establishing the diagnosis in cases of suspected infection, especially in the setting of negative cultures and/or prior antibiotic exposure. While treatment is generally with beta-lactam antibiotics, including penicillins, there is a lack of clinical data on the optimal treatment type and duration especially for severe infections such as meningitis.

We report a case of C. canimorsus meningitis and provide a comprehensive review of the English literature reviewing 36 additional published cases. Given the rarity of this condition, we summarize the risk factors, exposure history, symptoms, cerebrospinal fluid (CSF) parameters, diagnosis, treatment, and outcomes of this condition.

Methods

We report a case of C. canimorsus meningitis that was diagnosed using PCR technology. CSF was sent for cell counts and cultures and to the University of Washington for testing using a multiplex PCR. Informed consent was obtained from the individual participant reported on in this article.

In addition to our case, a comprehensive search of peer-reviewed English literature (1961–present) was performed to identify published cases of Capnocytophaga meningitis and compile relevant clinical data from these cases. The MEDLINE database was interrogated with the following MeSH terms: “Capnocytophaga” [MeSH], “Meningitis” [MeSH] and “Gram-Negative Bacterial Infections” [MeSH] with additional keyword searches with the terms “Capnocytophaga,” “dysgonic fermenter-2,” and “df-2.” Additionally, a search of Embase was performed with the following Emtree Subject Headings: “Capnocytophaga,” “Meningitis,” and “Gram-Negative Infection.”

Cases involving species other than C. canimorsus were excluded (e.g., C. gingivalis [5], C. cynodegmi [6]), as were cases not published in English literature or that did not include individual case data [7]. Additionally, C. canimorsus causing other types of central nervous infections (e.g., brain abscess) were excluded. Additional cases included in this review were identified through a review of articles on Capnocytophaga canimorsus meningitis as well as reviews of severe Capnocytophaga canimorsus infections.

Data collected included demographic and clinical information including age, sex, preexisting medical conditions, history of splenectomy, animal contact, symptoms, serum white cell count, time to positive blood cultures, results of head imaging, CSF counts and culture results, antibiotic selection and duration, and patient outcome. A total of 37 cases of Capnocytophaga canimorsus meningitis were identified including the current case.

Results

Case Report

A 71-year-old Caucasian female was found altered at home and transferred to an outside hospital. On presentation, she complained of fevers and severe headaches; however, she was uncertain of the length of her illness. Vitals signs included a temperature of 38.8 °C, pulse of 100 beats per minute, respiration 20 breaths per minute, and blood pressure of 182/86 mmHg. She had an altered mental status and meningismus including neck stiffness. There were no focal neurologic deficits or other examination findings.

Past medical history was remarkable for recently diagnosed lung cancer status-post lobectomy; she did not require adjunctive chemoradiation therapy. She also had a history of hypertension and chronic subdural hematoma. She denied diabetes, alcohol abuse, or prior splenectomy. She lived in Southern California and reported no recent travel history. She owned a dog and frequented a dog park with contact with several canines on a regular basis; she reported no dog bites.

Laboratory data on presentation were notable for a white blood count of 19,900 cells/mm3 (92% neutrophils), hemoglobin 13 g/dl, and platelets of 196 × 103/mm3. Lactate level was 2.4 mmol/l, creatinine was 1.0 mg/dl, and glucose was 109 mg/dl. Liver function tests and urinalysis were within normal limits.

A magnetic resonance image (MRI) with and without gadolinium showed left occipital and parietal acute infarcts without mass effect and stable small bilateral frontal subdural hematomas. A lumbar puncture was performed during the first 24 h of admission and revealed neutrophilic pleocytosis (590 white cells/mm3; 95% polymorphonuclear cells), red cell count of 2600 cells/mm3, low glucose of 12 mg/dl (reference range, 40–70 mg/dl), and elevated protein level of 413 mg/dl (reference range, 15–59 mg/dl). Gram stain did not show any organisms.

The patient was started on empiric antibiotic therapy with intravenous vancomycin and piperacillin-tazobactam prior to the lumbar puncture, but antibiotics were changed after the lumbar puncture revealed meningitis to intravenous vancomycin 500 mg every 8 h, ampicillin 2 g every 4 h, and ceftriaxone 2 g every 12 h; no steroids were administered.

Blood cultures (two sets) were drawn on admission using BD Plus Aerobic/F and BD Lytic/10 Anaerobic/F media. The first set was positive at approximately 4 days (98 h) and the second set at 4.6 days (110 h), in both the BD Lytic/10 Anaerobic/F media. The gram stain was reported as gram-negative rods with bacterial growth on plates by day 6. Matrix Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry (MALDI-TOF) was performed, but no identification was obtained. A Remel RapID ANA II test system was used for a biochemical identification, and Capnocytophaga was identified. The organism could not be successfully grown on culture for susceptibility testing.

Cerebrospinal fluid (CSF) cultures obtained on admission were sterile; however, antibiotics had been given prior to the lumbar puncture. Due to high suspicion of Capnocytophaga canimorsus as the causative organism given the patient’s exposure to dogs, a CSF specimen was sent to the University of Washington for multiplex broad-range bacterial polymerase reaction (PCR) testing, which was positive for C. canimorsus.

Antibiotic therapy was modified to meropenem 2 g IV q8 h. A transthoracic echocardiogram did not show vegetations and a chest, abdominal, and pelvic CT scan was unremarkable except for post-surgical finding consistent with prior lung lobectomy. The spleen appeared within normal limits. Repeat blood cultures after the initiation of antibiotics showed no growth.

The patient’s meningismus resolved during her hospital stay, and at the time of discharge her headache was significantly improved. She completed a 21-day total course of antibiotics and subsequently made a full recovery. She was educated on the potential infectious risks associated with dog ownership/exposure.

Review of the Literature

A total of 37 cases of Capnocytophaga canimorsus meningitis were identified including the current case (Table 1) [2, 4, 8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28]. Median age at presentation was 63 years (range, 12 days to 83 years) with a male predominance (28/37, 76%). While C. canimorsus meningitis has classically been characterized as a disease of immunocompromised patients, particularly patients who are asplenic, only 16% (6/37) of published cases occurred in patients with splenectomies. Five percent (2/37) had active steroid use. When summating all possible immunosuppressive states (e.g., medications or the presence of splenectomy or hematologic/autoimmune condition that impairs the immune system), 24% (9/37) of cases had one of these conditions: 5 splenectomy, 1 splenectomy and lymphoma, 2 steroid use, and 1 rheumatoid arthritis. Additionally, alcoholism was noted in 19% (7/37) and was the most common single medical condition identified.

Table 1 Summary of Capnocytophaga canimorsus meningitis cases in the published literature, n = 37
Full size table

Regarding animal exposure history, the majority of patients (59%; 22/37) reported a recent dog bite. The timing between the dog bite and presentation was a median of 6 days (range, 3 to 14 days). A smaller proportion reported non-bite dog exposures (22%; 8/37) and single cases of non-bite exposures to both dog and cat (3%, 1/37) and a cat bite (3%; 1/37). In addition, there was one reported case of indirect contact through two health providers who owned dogs. Overall, 11% (4/37) indicated no known animal contact prior to development of meningitis.

Presenting symptoms often included fever, headache, neck stiffness, altered mental status, and photophobia (Table 1). Other manifestations included rash in six cases, which was described as macular/papular in most cases; two cases had a rash that had a petechial/purpuric appearance. Other symptoms included seizures, myalgias, vomiting, fatigue, and hearing loss. Serum white blood cell (WBC) count was reported elevated in 11 (69%) of 16 cases that reported these data with a median WBC count of 13,500 cells/mm3 (range, 8000–25,000 cells/mm3).

CSF studies demonstrated a pattern consistent with bacterial meningitis with elevated CSF white counts of a median of 1024 cells/mm3 (range, 0–15,630 cells/mm3) with only one case with a normal white count. Most cases had neutrophilic predominance (81%, 26/32 cases with data). In addition, an elevated CSF protein was noted in 92% (22/24) of cases with a median value of (190 mg/dl) and low CSF/serum glucose ratio in all cases (median, 0.23). In cases that reported brain imaging, it was typically normal but in three cases (19%; 3/16 with imaging data), including the present case, acute infarcts were noted, and one additional case had cerebritis.

Blood cultures were found to be positive in 20 (54%) cases with a median growth time of 4 days (range, 2–9 days); 9 (24%) had negative blood cultures and 8 (22%) did not report results. CSF cultures were positive in 26 (70%) of cases with a median growth time of 5 days (range, 1–9 days). Due to the relatively long time for cultures to become positive, PCR is emerging as an important diagnostic technique. PCR was utilized to establish the diagnosis in the current case and overall in eight (22%) cases in this review.

Treatment involved extended courses of antibiotics with median treatment duration of 15 days (7–42 days). The most commonly utilized antibiotics were penicillin, ampicillin, and cephalosporins. Antibiotic susceptibilities were presented in eight cases in the literature with many cases lacking data, often because of insufficient growth of the bacteria on culture media. Overall, penicillins (including ampicillin), second- and third-generation cephalosporins, carbapenems (e.g., imipenem), and fluoroquinolones were susceptible in all cases reporting data for these antibiotics (Table 2).

Table 2 Antibiotic susceptibilities for the Capnocytophaga canimorsus isolates among meningitis cases in the published literature
Full size table

Outcomes were generally favorable with only one (3%) death observed in the published cases; this death may have been unrelated to the infection as the patient died of a cardiac arrest 10 days after discharge and had known coronary artery disease. Overall, 19% of survivors (7/36) had chronic sequelae of the disease including four with hearing loss, one with chronic headaches/disorientation, one with extremity amputations and chronic neurologic abnormalities, and one with extremity amputations. Both patients who underwent amputations also had brain imaging showing acute infarcts. Of the seven patients with chronic sequelae, three had a history of alcohol abuse and one was asplenic. No clear relationship between antibiotic duration and adverse outcomes was noted.

Discussion

Since the first reported case of Capnocytophaga canimorsus meningitis in 1976 [8], there have been 36 additional cases including the present case. We provide a comprehensive summary and provide a novel case of Capnocytophaga canimorsus meningitis to add to the existing literature. While this pathogen was previously reported to primarily cause severe disease in immunocompromised patients (e.g., history of splenectomy), our literature review revealed that most cases occurred among healthy, immunocompetent persons. As most cases had a recent dog exposure, C. canimorsus should be suspected in cases of bacterial meningitis with this exposure history.

Regarding underlying medical conditions, only 24% had an underlying immunosuppressive medical condition. While classically asplenia was deemed to be the characteristic condition associated with this pathogen, in our review only 16% (6 patients) with C. canimorsus meningitis were asplenic and 5% (2 patients) were receiving chronic steroids. The most common risk factor found in our review was alcoholism present in 19% (7 patients). These data suggest that Capnocytophaga is pathogenic and can cause severe disease in all hosts of all ages regardless of underlying risk factors.

A history of exposure to dogs was common in our review and emphasizes the importance of this historical feature in the early suspicion and diagnosis of this pathogen. Dog exposure, mostly common via a dog bite, was present in 86% of cases. Cat exposure was less commonly (3%) reported; only 11% of cases reported no known animal exposure. The incubation period was relatively short between animal exposure and presentation (median 6 days between bite and presentation). Whether antibiotic exposure after dog or cat bites would have prevented the reported cases of C. canimorsus meningitis is unclear, but prophylactic antibiotics did not prevent its occurrence in one previously published case although erythromycin was used rather than a β-lactam [18]. Among cases reviewed, there was a notable male predominance (3:1) perhaps related to an increased risk of canine bites or exposures among men.

Patients with C. canimorsus meningitis commonly presented with classic meningismus symptoms. In several of the cases, the course was fulminant and mimicked that of meningococcal disease. Of note, some C. canimorsus cases had purpuric or petechial rashes at presentation and required subsequent extremity amputations similar to the clinical course of Neisseria meningitidis [14]. CSF analysis was consistent with bacterial meningitis with an elevated leukocyte count with neutrophil predominance, elevated protein, and low glucose values occurring in most cases. Given the lack of differentiating clinical or laboratory findings for C. canimorsus meningitis compared with other bacterial organisms, clinicians must have a high index of suspicion suspecting this pathogen among patients presenting with findings of bacterial meningitis along with recent animal exposure.

The identification of Capnocytophaga sp. is supported by its growth using specific culture conditions as well as its colony morphology and microscopic characteristics [29]. The organism grows best at 35–37 °C on 5% sheep’s blood or chocolate agar (but not MacConkey’s) using anaerobic or aerobic conditions with 5–10% CO2. Even with ideal growing conditions, growth can take up to 7 days, and identification can be missed if culture plates are discarded at the traditional day 5 of incubation. The organism displays fingerlike projections on agar and gliding motility on microscopy, which may provide clues to its identification. A variety of biochemical identification systems and strips can be utilized for species determination. MALDI-TOF mass spectrometry using an enriched database may also aide in species level identification [30].

Capnocytophaga canimorsus is a relatively difficult organism to identify given its long incubation and specific culture requirements as noted above. In this review, only 54% of blood cultures and 70% of CSF cultures were positive with median growth times of 4 and 5 days, respectively. Two cases in the literature had gram stains showing gram-negative bacilli, but negative cultures. In cases where gram stains and/or cultures are negative, PCR can be helpful in definitively identifying the causative organism and improving the rapidity of the diagnosis [31], as was done in eight (22%) of the cases in our literature review. The value of PCR technology is particularly important in cases involving slow-growing or fastidious organisms (such as Capnocytophaga sp.) and in the setting of prior antibiotic exposure before cultures are obtained. Commercial PCR tests are becoming more widely available; however, C. canimorsus may not be included in all testing platforms (e.g., BioFire FilmArray Meningitis/Encephalitis Panel) [32] and hence may require broad-range PCR technology.

Regarding treatment, penicillin has been reported as the drug of choice in the past, although clinical comparative trials are lacking [12]. Since beta-lactamase activity has been reported among some Capnocytophaga species [33], clinical isolates should be tested for β-lactamase production to inform the optimal treatment regimen. Because the organism can be difficult to grow, susceptibilities may not be feasible. To provide further guidance on treatment options, we reviewed the published susceptibility data from C. canimorsus meningitis cases. These data suggest that C. canimorsus is typically susceptible to penicillins (including ampicillin), third-generation cephalosporins, carbapenems such as imipenem, and fluoroquinolones. Despite being a gram-negative organism, aminoglycosides and aztreonam are typically inactive. Isolates were also typically resistant to trimethoprim-sulfamethoxazole. In our review, the median treatment time was 15 days although the optimal duration is unknown. The current Infectious Disease Society of America (IDSA) Meningitis Guidelines suggest 21 days for meningitis involving aerobic gram-negative organisms (other than Haemophilus influenzae) [34].

The outcome of C. canimorsus meningitis was overall favorable and had a lower overall mortality rate compared with both other bacterial causes of meningitis (~ 14–25%) [35] and C. canimorsus sepsis (~ 30%) [3, 18]. Only one patient in our review died, and this event was unlikely related to the infection. Adverse sequalae, however, were notable with 19% of patients suffering from a residual deficit, which was most commonly hearing loss. Additionally, 19% of those with C. canimorsus meningitis with reported brain imaging had evidence of acute infarcts. Cerebral infarctions are potential complications of bacterial meningitis and in a prior study were reported among 36% of patients with pneumococcal meningitis, 9% with meningococcal meningitis, and 28% due to other bacteria; patients with infarctions are often older and have an immunocompromised state [36].

Limitations of this review include the overall small number of C. canimorsus meningitis cases in the literature and the lack of complete individualized data for some cases. Additionally, our review only included cases in the English literature. Strengths include this being the largest and most comprehensive review to date. Given that animal exposures are increasingly common [29], understanding this disease entity is important. Unlike other bacteria pathogens associated with animal bites (e.g., Pasteurella, streptococci), this organism is not typically associated with local wound infections, but rather presents as severe, disseminated infections including bacteremia, sepsis, and meningitis.

Conclusion

Capnocytophaga canimorsus is a rare but important cause of bacterial meningitis and should be considered in cases following a recent dog bite or exposure. Given the increasing incidence of infections associated with animal exposures, knowledge regarding the clinical manifestations, diagnosis, and treatment of this clinical entity is important. This case and review provides the most up-to-date information about the clinical management of this important infection. Furthermore, it highlights the value of PCR technology to enhance the timely diagnosis of C. canimorsus meningitis.