Brucellosis in humans caused by Brucella canis: A scoping review

Abstract

Background

Brucella canis is a potential cause of brucellosis in humans, but this disease has been poorly characterized.

Procedure

A scoping review was completed.

Results

The review yielded 24 studies that described clinical B. canis infection in 68 individuals. The median age was 32 y (range: 17 mo to 71 y). Fever, fatigue, headache, chills, and malaise predominated.

Specific treatments were reported for 30 individuals. A tetracycline (tetracycline, doxycycline, or minocycline) was the sole treatment or part of a combination treatment for most (21/30, 70%) individuals, followed by rifampin (9/30, 30%), trimethoprim/sulfamethoxazole (7/30, 23%), and an aminoglycoside (gentamicin, streptomycin) (7/30, 23%). Clinical outcome was reported for 35 individuals. No mortalities were reported.

A known or suspected source of exposure was reported for 56 cases; 45 (80%) were linked to dogs and 11 (20%) to laboratory exposure. Contact with pet dogs that had aborted or with aborted fetuses accounted for 31% of the canine exposures.

Conclusion and clinical relevance

Brucellosis attributed to B. canis can cause a wide range of clinical presentations in humans, but was similar to brucellosis caused by other Brucella spp. This nonspecific nature highlights the importance of identifying risk factors for exposure, to determine whether B. canis infection should be considered.

INTRODUCTION

Brucellosis is a globally important zoonotic disease caused by bacteria of the Brucella genus. Most human infections are caused by Brucella species associated with food animals: B. melitensis, B. abortus, B. suis and B. ovis. Yet other Brucella species can also cause zoonotic infections.

Brucella canis is a dog-adapted Brucella species detected in dogs internationally (1–7). Similar to other Brucella spp., B. canis is a Gram-negative bacterium that can cause persistent infections in its reservoir species and spillover into humans. Reproductive disease is the most common clinical manifestation in dogs (7,8), but other diseases, such as diskospondylitis, can occur (9,10). However, most infected dogs have clinically inapparent infections (3,7,11).

In recent years, B. canis has received increasing attention because of concerns involving importation of infected dogs and the associated human risk (12–15). Yet this bacterium has been identified in a wide range of countries, and given the gaps in canine surveillance, endemic sources likely pose a risk in many, if not most, countries.

Since there has been limited study of B. canis in humans, the objective of this scoping review was to describe the clinical presentation, diagnosis, treatment, and exposure sources of clinical B. canis infections in humans internationally.

METHODS

Population

This scoping review included reports of brucellosis in humans that was attributed to B. canis.

Concepts

The key concepts were diagnostic approaches, clinical disease characteristics, exposure sources, treatment, and clinical outcomes.

Content

Academic literature published in English was included in this scoping review. This scoping review included publications worldwide, with no publication date restrictions.

Search strategy

The scoping study protocol was designed as per Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews (PRISMA-ScR) guidelines (16). The scoping review was conducted in accordance with JBI methodology for scoping reviews (17). A search of PubMed and Web of Science databases was completed on March 15, 2024, to identify publications reporting human infection with Brucella canis. Search terms were [Dog OR canine OR canid OR canis OR puppy] AND [Brucella OR brucellosis] AND [Human OR zoonotic OR infection]. Reference lists of included publications were reviewed for additional relevant studies.

Screening process

Only studies reporting clinical infections were eligible. “Infection” was defined as detection of B. canis by isolation, antigen detection (e.g., PCR), or serological testing in the presence of clinical signs and symptoms that were attributed to B. canis infection. No date range was applied to this search. Reviews were excluded from data extraction but reference lists of reviews were hand-searched to identify primary sources not identified during the initial literature search. Studies that described genomic characterization of isolates with no clinical or epidemiological data were excluded. Studies were excluded if full texts in English were not available.

Two independent reviewers screened the titles and abstracts against the inclusion/exclusion criteria (Level 1, detailed below). Abstracts that fit the Level 1 criteria or for which all aspects of the inclusion criteria could not be determined from the title and abstract were selected for full-text screening (Level 2, detailed below). In cases of nonagreement, the reviewers discussed the evaluation and reached consensus.

Level 1: Title and abstract screening

1. Is the title/abstract in English?

2. Does the title/abstract describe study of human infection with B. canis?

3. Does the title/abstract describe research that used primary or secondary data, including case reports, case series, or observational studies?

Level 2: Full-text screening

1. Is the full text available and in English?

2. Does the title/abstract describe study of human infection with B. canis, including clinical reports?

3. Were the reported infections associated with clinical disease attributable to B. canis infection (brucellosis)?

4. Does the title/abstract describe research that used primary or secondary data, including case reports, case series, or observational studies?

5. Are diagnostic, clinical, treatment, exposure, or outcome findings reported?

Data items

Full-text extraction recovered author, year of publication, title, study type (e.g., case report, case series, seroprevalence study), year(s) of sampling, country, number of infected individuals, age(s) of individuals, number of individuals < 18 y of age, clinical disease characteristics, diagnostic method, antimicrobial treatment, outcome, source of infection (if known), and mortality.

Synthesis of results

Data were summarized and described.

Ethical approval

Ethical approvals were not required because this study did not involve human or animal subjects.

RESULTS

The literature search yielded 1455 potentially relevant studies. After de-duplication and relevance screening, 24 studies describing clinical disease attributed to B. canis infection were determined to be relevant for this review and were included in data extraction (Figure 1). Details of the 24 studies included in this review are presented in Table 1. Sixteen (67%) were single case reports and 8 (33%) were case series.

FIGURE 1.

Flowchart describing the selection of articles, inclusion and exclusion of studies, and final number of studies for which data extraction was performed.

TABLE 1.

Studies included in final data extraction for a scoping review describing clinical Brucella canis infection in humans.

Study (first author, publication year)	Study type	Country/countries	Number of individuals
Ahmed-Bentley et al, 2021	Case report	Canada	1
Blankenship et al, 1975	Case report	US	1
Dentinger et al, 2015	Case report	US	1
Faigel et al, 1969	Case report	US	2
Fox et al, 1977	Case series	US	16
Ishihara et al, 2023	Case report	Japan	1
Javeri et al, 2014	Case report	US	1
Kawakami et al, 2019	Case report	Japan	1
Kolwijck et al, 2022	Case report	Netherlands	1
Lawaczeck et al, 2011	Case report	US	1
Lucero et al, 2005a	Case report	Argentina	1
Lucero et al, 2005b	Case series	Argentina	17
Lucero et al, 2010a	Case report	Argentina	4
Lucero et al, 2010b	Case report	Argentina	1
Marzetti et al, 2013	Case series	Argentina	3
Morisset et al, 1969	Case series	US	2
Munford et al, 1975	Case series	US	2
Nomura et al, 2010	Case series	Japan	2
Piampiano et al, 2000	Case series	US	2
Rifkin et al, 1978	Case report	US	1
Taylor et al, 1989	Case series	US	4
Tosi et al, 1982	Case report	US	1
Wallach et al, 2004	Case report	US	1
Ying et al, 1999	Case report	US	1

US — United States.

References: (18,25,29–43).

The included studies were published between 1969 and 2023; the number of studies per year is presented in Figure 2. Studies were from 5 countries: United States (n = 14, 58%), Argentina (n = 5, 21%), Japan (n = 3, 13%), and 1 study (n = 4.2%) each from Canada and the Netherlands. Details were reported for 68 affected individuals, with 1 to 17 reported per study (median: 1).

FIGURE 2.

Year of publication and country of origin for 24 studies describing brucellosis attributed to Brucella canis in humans. US — United States.

Age was described for 45 individuals, whereas 2 others were reported as adults. Affected individuals ranged from 17 mo to 71 y of age. The median age from studies that reported age was 32 y (25th quantile: 15 y, 75th quantile: 46 y) (Figure 3). Fourteen individuals (32%) were < 18 y of age.

FIGURE 3.

Scatterplot of ages of individuals (n = 45) diagnosed with brucellosis attributed to Brucella canis in humans.

Clinical details

Specific clinical details were provided for 31 individuals. Commonly reported clinical signs and symptoms are outlined in Table 2. Both chronic and acute presentations were described. One noteworthy case described fever, malaise, and other clinical signs, recurrent for 19 y, that started when the individual attended an “animal-care school” and handled whelping dogs.

TABLE 2.

Common clinical findings in 31 human individuals with Brucella canis infection.

Symptom	n (%)
Fever	21 (68)
Fatigue/lethargy	12 (39)
Headache	9 (29)
Chills	8 (26)
Malaise	8 (26)
Weight loss	4 (13)
Splenomegaly/hepatomegaly	4 (13)
Cough	4 (13)
Vomiting	2 (6.5)
Diarrhea	2 (6.5)
Decreased appetite	2 (6.5)
Nausea	2 (6.5)
Arthralgia	2 (6.5)
Myalgia	2 (6.5)
Lymphadenopathy	2 (6.5)

One study involving 17 individuals only reported “symptoms compatible with brucellosis” (18), whereas 2 studies (19,20) reported summary clinical information from 20 individuals with brucellosis caused by multiple Brucella spp. Since those reports focused on individuals with typical clinical signs, it is likely that the true prevalence of common signs and symptoms from Table 2, such as fever, chills, malaise, and headache, predominated in those reports.

In addition, some uncommon presentations were reported, including endocarditis (n = 2), and there were single reports of septic arthritis, transient left arm paresis, Guillain-Barré syndrome, peritonitis, and meningoencephalitis.

Diagnostic method

All studies described diagnostic methods. Twelve (50%) diagnosed B. canis infection via blood culture, 7 (29%) via a combination of blood culture and serology, and 5 (21%) via serology alone.

Treatment

Specific treatments were reported for 30 individuals. A wide range of antimicrobials or antimicrobial combinations were used (Table 3). A tetracycline (tetracycline, doxycycline, or minocycline) was the sole treatment or part of a combination treatment for most individuals (21/30, 70%), followed by rifampin (9/30, 30%), trimethoprim/sulfamethoxazole (7/30, 23%), and an aminoglycoside (gentamicin, streptomycin) (7/30, 23%).

TABLE 3.

Antimicrobial approaches for treatment of Brucella canis infection in 30 human individuals.

Drug(s)	n (%)
Tetracycline	5 (17)
Doxycycline + rifampin	4 (13)
Doxycycline + streptomycin	2 (6.7)
Trimethoprim sulfamethoxazole (TMS)	2 (6.7)
Doxycycline + rifampin + TMS	1 (3.3) each
TMS + rifampin
Ampicillin, then nothing
Ampicillin, then tetracycline
Gentamicin + doxycycline
“IV antibiotics”
TMS + gentamicin
Ceftriaxone, then TMS
Tetracycline + streptomycin
Doxycycline + ciprofloxacin
Ceftriaxone, then doxycycline + rifampin
Moxalactam, then chloramphenicol
TMS + minocycline
Ofloxacin + rifampin + doxycycline
Doxycycline + gentamicin
Gentamicin, then doxycycline + rifampin
No antimicrobial(s)

Some individuals were not treated for B. canis or had what was likely questionably effective treatment. One individual received no antimicrobials and infection was self-limiting. Another individual had clinical resolution without antimicrobials but was subsequently treated with tetracycline because of persistent but subclinical bacteremia. One received a short course of ampicillin (a drug not typically considered effective against Brucella), then was left untreated. Another received ampicillin, then 3 doses of tetracycline before treatment was discontinued.

Outcome

Clinical outcome was reported for 35 individuals. No mortalities were reported. All 35 individuals were reported to have responded clinically to treatment, including those that received no or limited antimicrobial therapy. Long-term consequences were reported for 1 individual who had Guillain-Barré syndrome attributed to B. canis infection. That individual was responding to treatment at the time of writing but was still undergoing rehabilitation therapy.

Exposure

A known or suspected source of exposure was reported for 56 cases; 45 (80%) were linked to dogs and 11 (20%) to laboratory exposure (Table 4). Detailed exposure information was often not reported but contact with pet dogs that had aborted or contact with aborted fetuses accounted for 14 cases, representing 31% of canine exposures. Although the risks are greatest from contact with breeding dogs, one infection was linked to an 8-week-old (and therefore sexually immature) puppy (21). In another report, the affected individual’s dog was a spayed female, but the source could not be determined because the individual had also interacted with infected dogs in the community (22).

TABLE 4.

Known or suspected sources of Brucella canis infection in humans (n = 56).

Known or suspected source	n (%)
Pet dog (no further details)	25 (45)
Pet dog that aborted or contact with aborted fetuses	12 (21)
Pet dogs: Breeding dogs	2 (3.6)
Imported breeding (n = 1) or rescue (n = 1) dog that had aborted	2 (3.6)
Community dog	1 (1.8)
Pet dog or community dogs	1 (1.8)
Pet dog: Newly purchased	1 (1.8)
Animal care school, handled whelping dogs	1 (1.8)
Laboratory	11 (20)

Brucella canis infection status of suspected sources was not always queried, but often (n = 33), contact with a dog known to be infected was identified. Typically, diagnosis of the dog’s infection occurred after recognition of disease in a human.

Laboratory exposure accounted for 11 infections. Nine were linked to exposure to laboratory cultures, whereas 2 were linked to contact with infected fetuses. Laboratory exposures were mainly documented in older reports, with 4 individuals exposed via that route in 1969, 6 in 1977, and 1 in 2004.

DISCUSSION

These data represent the reported and accessible subset of B. canis infections in humans but highlight some important clinical and management aspects of this endemic, zoonotic, canine-associated bacterium. Whereas B. canis is endemic in the dog population and has been detected worldwide, reports of disease in humans are limited. Underdiagnosis is an important concern with B. canis because commonly used human serological tests are designed to detect food-animal-associated Brucella species, not this rough-coated Brucella species (23). With an unknown degree of underdiagnosis and only a subset of diagnosed cases reaching the literature, the incidence of B. canis infection in humans cannot be ascertained. The majority of cases were reported from 2 countries, despite the presence of this bacterium in dogs worldwide, further suggesting underdiagnosis and underreporting.

Clinical signs were largely those common with brucellosis of any etiology, particularly nonspecific findings such as fever, chills, fatigue, malaise, and headache. This highlights potential clinical challenge for early recognition of B. canis infection if risk exposures are not identified and reported. The prevalence of individual signs and symptoms reported here was likely an underestimate, as publication bias tends towards reporting atypical disease. Further, specific clinical details were not provided for 37 individuals, yet 1 study involving 17 individuals indicated symptoms were compatible with brucellosis, presumably meaning presence of fever, chills, and other common abnormalities.

A wide range of treatments was used with apparent success. There are no guidelines for treatment of B. canis but use of tetracyclines, aminoglycosides, and rifampin, often in combination, is consistent with approaches to treatment of brucellosis caused by other Brucella spp. Use of tetracycline (versus doxycycline) and monotherapy was predominantly reported in older publications. Interestingly, multiple infections were apparently self-limiting clinically, with complete resolution in 1 patient, clinical (but not microbiological) resolution in another, plus 2 patients that received limited treatment.

Exposures were often not well investigated, but there were many situations in which dog contact was explored and in-contact dogs were tested. There was a high reported prevalence of known contact with aborting dogs or aborted fetuses, and it is possible that the prevalence was higher since limited details were reported in many studies. This underscores the high risk posed by these situations and the need for awareness and use of infection-control practices around breeding animals, particularly those with reproductive disease. Contact with nonbreeding dogs, particularly those that have been spayed or neutered, is thought to pose a lesser risk (24). Although that scenario could not be well investigated in this review due to limited dog-level data for many reports, a large percentage of infections were linked to breeding dogs and aborted fetuses. However, 1 case was linked to a sexually immature, 8-week-old puppy, highlighting the potential for transmission across a range of dog ages and reproductive statuses.

Laboratory exposures accounted for 20% of cases with a known or suspected origin. Most laboratory exposures were from the 1960s and 1970s (20,25,26), but 1 was from 2004 (27). Brucella canis, like other Brucella species, is a high-risk agent in the laboratory and is supposed to be handled under biosafety level (BSL)/containment level (CL) 3 conditions. Potential laboratory exposures continue to be a concern (28), particularly when B. canis is unintentionally isolated using lower containment levels through routine clinical testing.

Clinical outcomes were very good. All but 1 individual for whom information about treatment response was available responded well to treatment. The additional patient, who had Guillain-Barré syndrome, responded to treatment but still had abnormalities that were being addressed through rehabilitation. Outcome was not reported for almost 1/2 of affected individuals, but it is reasonable to hypothesize that fatal infections or severe manifestations would likely have been mentioned. Care must be taken not to overinterpret prognosis from a small dataset, but these data supported the effectiveness of treatment and the likely ability for some infections to resolve without intervention.

Various limitations should be considered. Underdiagnosis and underreporting limit the number of publications describing B. canis in humans and make it impossible to infer an incidence of disease from available studies. Exclusion of manuscripts when full text was not available also reduced the number of reported cases, particularly some of the earlier reports of human infections from the 1970s and 1980s. However, this should not have significantly affected the main findings, as there is no information to suggest that clinical aspects of B. canis infection in humans have changed. Whereas dogs were the clear source of natural (non-laboratory) infection, investigations and descriptions of exposures were variable and often superficial. This hampered efforts to more clearly define different contact risks, such as those from contact with nonbreeding pet dogs and, in particular, spayed or neutered household pets. This is important for risk communication and requires further study. This scoping review also targeted clinical disease caused by B. canis, not asymptomatic infections. Thus, it covers the most clinically relevant subset of infected humans but still only represents a subset. The long-term consequences of asymptomatic B. canis infection are not known, but lack of disease at one time point does not mean there are no potential clinical consequences.

These results should not be taken to represent the totality of human B. canis infections. In addition to cases missed due to underdiagnosis and lack of reporting, a small number of existing publications were not included because they were older manuscripts for which full text was inaccessible, or they were not published in English. Thus, this review provides an overview of results, with most reported infections, but is not a complete summary. CVJ