Campylobacter fetus Invasive Infections and Risks for Death, France, 2000–2021

Campylobacter fetus accounts for 1% of Campylobacter spp. infections, but prevalence of bacteremia and risk for death are high. To determine clinical features of C. fetus infections and risks for death, we conducted a retrospective observational study of all adult inpatients with a confirmed C. fetus infection in Nord Franche-Comté Hospital, Trevenans, France, during January 2000–December 2021. Among 991 patients with isolated Campylobacter spp. strains, we identified 39 (4%) with culture-positive C. fetus infections, of which 33 had complete records and underwent further analysis; 21 had documented bacteremia and 12 did not. Secondary localizations were reported for 7 (33%) patients with C. fetus bacteremia, of which 5 exhibited a predilection for vascular infections (including 3 with mycotic aneurysm). Another 7 (33%) patients with C. fetus bacteremia died within 30 days. Significant risk factors associated with death within 30 days were dyspnea, quick sequential organ failure assessment score >2 at admission, and septic shock.

C ampylobacter is a genus of microaerophilic, fas- tidious, gram-negative, occasionally partially anaerobic, non-spore forming, motile bacteria with a characteristic spiral or corkscrew-like appearance (1).Such morphology enables the bacteria to colonize the mucosal surfaces of the gastrointestinal tract in humans and other animal species (2).In France, C. fetus is the most commonly isolated Campylobacter species, after C. jejuni and C. coli, found in fecal samples during diarrheal episodes in humans (3), and the leading species recovered from invasive infections, such as bacteremia and secondary localizations; both C. jejuni and C. coli have been identified in 43% of cases (4).
Using data for January 2000-December 2021, we conducted a retrospective observational and descriptive study in Nord Franche-Comté Hospital, located in eastern France.Our primary objective was to describe clinical and paraclinical features (including antimicrobial susceptibility) in patients with C. fetus infections by comparing patients with and without bacteremia.Our secondary objective was to evaluate the risk factors for 30-day mortality in patients with bacteremia caused by C. fetus.
Patient consent was obtained by sending patients a letter informing them of the use of their medical data for research purposes and receiving no objection by 30 days later.Because of the retrospective nature of the study, with no patient involvement and use of already available data, the local Ethics Committee of Nord-Franche-Comte Hospital determined that patient consent was sufficient.The confidentiality of participant data has been respected in accordance with the Declaration of Helsinki.

Study Population and Design
Nord Franche-Comté Hospital has a capacity of 1,216 beds across all sites.The Nord-Franche-Comté Hospital practice has ≈100,000 visits to its emergency rooms and ≈3,600 deliveries per year (12,13).Our study included all adults (>18 years of age) with a C. fetus infection, defined by identification of C. fetus in a microbiological sample (blood, fecal, or other site culture) of hospitalized patients over a 21-year period (January 1, 2000-December 31, 2021).

Data Collection
We collected clinical data regarding demographic and baseline characteristics and underlying conditions from patients' medical records.We also extracted laboratory and imaging findings, outcomes, and results of antimicrobial susceptibility to amoxicillin, amoxicillin/clavulanic acid, imipenem, gentamicin, azithromycin, doxycycline, and fluoroquinolones (ofloxacin and ciprofloxacin).

Definitions
We defined secondary localizations as a positive result on biopsy, graft, blood culture samples (or a combination of those) or evocative images on computed tomography or 18 F-fluorodeoxyglucose-positron emission tomography/computed tomography ( 18 F-FDG PET/CT).Endocarditis, also considered as a secondary localization, was defined by a positive valvular biopsy sample, blood culture, or both, associated with evocative images on echocardiography, 18 F-FDG PET/CT, according to the European Society of Cardiology 2015 modified criteria for diagnosing infective endocarditis (14).
According to the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) (15), adult patients with suspected infection can be rapidly identified as being more likely to have poor outcomes  (16).We considered antimicrobial treatment to be appropriate if the strain was susceptible to >1 of the drugs prescribed, according to the CA-SFM/EUCAST recommendations (17).C. fetus is naturally resistant to third-generation cephalosporins, ticarcillin, and piperacillin, so we considered those drugs to be inappropriate.Some strains were tested by automated broth microdilution system (Vitek-2; bioMérieux, https://www.biomerieux.com), and results could not be reinterpreted according to current recommendations.

Data Analysis
Unless otherwise indicated, we expressed discrete variables as numbers and percentages and continuous variables as mean/average, SD, and 95% CI.We performed comparisons among patients with and without C. fetus bacteremia by using a χ 2 or Fisher exact test for qualitative variables and a Student t or Wilcoxon test for quantitative data.Risk factors for death are expressed as odds ratios (ORs), and statistical analysis was performed by using univariate logistic regression.We used a significance level of p>0.05 and performed all analyses by using R version 4.2.1 (The R Project for Statistical Computing, https://www.r-project.org).We defined a significant trend as p<0.06.
During the study period, we identified 39 patients with culture-positive C. fetus infections; of those, 33 had complete records and underwent further analysis, 21 with documented bacteremia and 12 without (Figure 1).Among bacteremic patients, fecal cultures were negative for 20 (95.3%).For only 1 patient were simultaneous peripheral blood and stool cultures positive, isolating C. fetus, and that patient was included in the bacteremia group.With regard to patients without documented bacteremia, most (11 of 12) isolates were from fecal samples; gastroenteritis was reported for 10 (83%) patients, of which 9 had liquid diarrhea.Peripheral blood cultures were performed for 8 (66%) of 12 patients without documented C. fetus bacteremia and were negative.

Therapeutic Management and Outcomes
Among 29 patients receiving antimicrobial therapy, the most commonly used drug was amoxicillin/clavulanic acid ( 12    †Campylobacter fetus was isolated from 11 fecal cultures and from 1 bile fluid culture; only 1 patient had simultaneous positive peripheral blood and fecal cultures in which C. fetus was isolated with the same antimicrobial testing susceptibility pattern.‡All patients with connective tissue diseases (bacteremia; n = 2) had multiple sclerosis.§Defined by hematologic diseases, long-term steroid therapy, or immunomodulatory treatment.¶Defined by cardiac failure, arrhythmia, coronary heart disease, stroke, peripheral arterial obstructive disease and thromboembolic disease.#28-y-old patient sought care for bloody diarrhea without fever 8 d after a travel in Spain.**The 3 patients who underwent transthoracic echocardiography were different patients than those who underwent 18 F-FDG PET/CT.No patient had both endocarditis and mycotic aneurysm.
[40%]).Amoxicillin/clavulanic acid was prescribed for 7 (88%) of 8 patients with bacteremia treated with dual-regimen therapy.We found no significant difference in mean duration of treatment between the 2 groups (9 [SD 8] vs. 5 [SD 5] days; p = 0.2).Five patients underwent surgery, including 4 with bacteremia (2 for mycotic aneurysm, 1 for prosthetic valve endocarditis, and 1 for septic arthritis).Four patients were transferred to an intensive care unit for septic shock.Two patients experienced a relapse with fever as the main clinical sign after 26 and 50 days; 1 patient died of septic shock during the second episode.

Antimicrobial Susceptibility Testing
Among patients with C. fetus bacteremia, the rate of resistance was 10% (2/20) to both amoxicillin and †Mycotic aneurysm (n = 3) with infectious native aortic aneurysm (n = 2); prosthetic aortic valve and a positive culture of the aneurysm after surgery (n = 1); prosthetic valve endocarditis (n = 1) with typical oscillating vegetation (15 mm) confirmed by transthoracic echocardiography; abdominal aorta thrombophlebitis (n = 1); hematogenous medical device infection with a percutaneous implantable port-related infection (n = 1); osteoarticular (n = 1) with glenohumeral shoulder arthritis and a positive culture of the articular fluid after surgery, suggesting a contiguous infection.‡Four patients with bacteremia caused by C. fetus underwent surgery: mycotic aneurysm (n = 2), endocarditis (n = 1), and septic arthritis (n = 1).§Two patients exhibited a relapse with fever after 26 and 50 d; the second patient died of septic shock during the second episode.The first patient received ciprofloxacin orally for 5 d, and the second patient received IV vancomycin for 7 d.¶Among the 7 bacteremic patients who died, 2 died in the context of evolutive/expanding malignancy (independently of the bacteremia).

Discussion
The most commonly detected cause of Campylobacter bacteremia is C. fetus (5,6).However, cohorts or large case series exclusively involving patients with C. fetus bacteremia remain scarce.C. fetus is usually isolated from blood samples and is less frequently associated with enteritis (18,19).The Campylobacteremia Study (4), a retrospective multicentric study of Campylobacter spp.bacteremia in France, also showed that one of the regions with the highest rate of Campylobacter spp.infection is the Franche-Comté region.
Our study comprised 33 patients with C. fetus infection; the 21 patients with bacteremia were older than the patients without bacteremia, in keeping with data in the literature and previous reports.According to the medical literature of patients with Campylobacter bacteremia in one of the largest retrospective cohorts (n = 592), patients were elderly (median age 68 years) and most had underlying conditions, mainly immunosuppression (4).In our cohort, immunosuppression was more frequent among patients with bacteremia caused by C. fetus than among with patients with no bacteremia; the trend was significant (p = 0.06).The leading cause was malignancy or cancer (33%) (1,20).Two patients with documented C. fetus bacteremia had systemic sclerosis, which seems to be a predisposing condition among connective tissue diseases (1,21).Pacanowski et al. (5) showed that, compared with patients with bacteremia caused by other Campylobacter species, patients with C. fetus bacteremia were older and had underlying comorbidities (e.g., cardiovascular diseases, diabetes mellitus).That finding is consistent with our results and those of other reports (20,22).
Among patients with C. fetus bacteremia, one third exhibited secondary localizations with a predilection for vascular infections.A recent multicenter study in France (252 patients with C. fetus bacteremia) found that 11.5% patients had vascular localization and 4.4% had endocarditis (6).In our study population, we found more vascular localizations but less endocarditis.However, endovascular localizations were not systematically searched and might have been underdiagnosed.In our study, TTE and 18 F-FDG PET/CT were each performed for only 14% of patients with bacteremia, which is a major limitation.We suggest performing those radiologic examinations early for patients with C. fetus bacteremia (6,23).Late radiologic examination may partially explain the high mortality rate among patients with aneurysm rupture or endocarditis.
One of the major problems associated with C. fetus infection is empiric treatment.Infection with those fastidious bacteria is uncommon, and recommendations for treatment of bacteremia are lacking.The standard choice for empiric treatment of Campylobacter spp.enteritis remains fluoroquinolones and macrolides (18).However, in our cohort, 33% of bloodstream isolates were resistant to fluoroquinolones, and 10% were resistant to azithromycin.No strain was resistant to amoxicillin/clavulanic acid, aminoglycoside, or imipenem.The initial empiric treatment should be dual antimicrobial therapy (including amoxicillin/ clavulanic acid or imipenem with an aminoglycoside) (6,22,24).In our cohort, all dual-therapy regimens consisted of amoxicillin/clavulanic acid (7/8 [88%]) or imipenem (1/8 [12%]) with a second agent.We conclude that initial prescription of amoxicillin/ clavulanic acid and use of dual antimicrobial therapy were protective factors.Failure to administer appropriate antimicrobial therapy is strongly associated with fatal outcomes (4,5).
Other independent risk factors for death were immunosuppression, cancers, and surgery (5,6).In our cohort, risk factors for death within 30 days after C. fetus bacteremia were dyspnea, quick sequential organ failure assessment score at admission >2, and septic shock.We found no significant difference between survivors and nonsurvivors with regard to antimicrobial therapy duration (p = 0.8), which could be explained by the longstanding clinician behavior of avoiding short antimicrobial regimens, even for patients who have positive fecal cultures without bacteremia.
In our cohort, the mortality rate was high (33% of patients with C. fetus bacteremia).It should be noted that among those 7 patients, 2 died in the context of evolutive/expanding malignancy and 1 died in the context of recurrent bacteremia with septic shock.In addition, 3 of 7 bacteremic patients who died were receiving 3 antimicrobial drugs, which suggests that in some cases, the number of antimicrobial drugs may have been a marker of illness severity.
Among the limitations of our study were the retrospective method used and the limited number of patients.A prospective study might confirm and support our results.As we previously mentioned, secondary localizations are probably underdiagnosed because of lack of knowledge of this disease and therefore nonperformance of investigations.
In summary, we found that C. fetus bacteremia mainly affects patients who are elderly, are immunocompromised, or have underlying conditions.Infections are associated with high mortality rates, especially if no dual antimicrobial therapy including amoxicillin/clavulanic acid is prescribed.For patients with bacteremia caused by C. fetus, screening for secondary localizations may be warranted by performing TTE and 18 F-FDG PET/CT.

Figure 1 .
Figure 1.Flowchart of patient enrollment in study of Campylobacter fetus invasive infections and risks for death, Nord Franche-Comté Hospital, France, 2000-2021.
Campylobacter fetus accounts for 1% of Campylobacter spp.infections, but prevalence of bacteremia and risk for death are high.To determine clinical features of C. fetus infections and risks for death, we conducted a retrospective observational study of all adult inpatients with a con- 2190Emerging Infectious Diseases • www.cdc.gov/eid• Vol. 29, No. 11, November 2023w

Table 2 .
Secondary localizations, therapeutic management, and outcomes of patients with Campylobacter fetus infections among patients with and without bacteremia, Nord Franche-Comté Hospital, Trévenans, France, 2000-2021* *Values are no./total(%) except as indicated.Boldface indicates p<0.05 or a significant trend defined by p<0.06.Blank cells for p values indicate no p value was calculated.GBS, Guillain-Barré syndrome; NA, not applicable.

Table 3 .
Antimicrobial resistance of Campylobacter fetus strains isolated from patients with and without bacteremia, Nord Franche-Comté Hospital, Trévenans, France, 2000-2021* Values are no.resistant/no.tested (%) except as indicated.Blank cells for p values indicate no p value was calculated.†Susceptibility to amoxicillin was intermediate (susceptible with high doses) for 2 strains.‡Susceptibility to ofloxacin and ciprofloxacin was the same for all strains. *

Table 4 .
Risk factors for death within 30 d after Campylobacter fetus bacteremia, Nord Franche-Comté Hospital, Trévenans, France, Values are no./total(%) except as indicated.Blank cells for p values indicate no p value was calculated.qSOFA, quick sequential organ failure assessment; NA, not applicable; OR, odds ratio.†p<0.05indicates significance. *