An unusual case of Erysipelothrix rhusiopathiae prosthetic joint infection from the Canadian Arctic: whole genome sequencing unable to identify a zoonotic source

Erysipelothrix rhusiopathiae is the causative organism of erysipeloid, a localized cutaneous infection [1‐3]. Less commonly, it is known to cause systemic infection with or without endocarditis [1, 2, 4]. Infection is usually associated with occupational exposure to infected host animals and is most frequently reported following contact with domestic swine [1‐3, 5]. We report a case of E. rhusiopathiae prosthetic joint infection (PJI) in a woman with exposure to multiple wild host species in the Canadian Arctic. E. rhusiopathiae has recently been linked to multiple muskox die-offs [6, 7] as well as the emergence of a new disease syndrome in Arctic fox [8] in these Arctic areas. This has generated concerns that an increase in human infections might be observed, especially in regions where exposure to potentially infected animals is expected to be greater [6].

E. rhusiopathiae is a ubiquitous, non-spore forming, facultative intracellular, gram-positive bacillus with a global distribution [2]. While E. rhusiopathiae is a commensal organism and a pathogen of a variety of domestic and wild animals, it is also known to exist as an environmental saprophyte [1]. Human infection with this organism is considered a zoonosis [1, 2]. The clinical spectrum of E. rhusiopathiae infection in humans consists of three major forms of disease: localized cutaneous infection, diffuse cutaneous infection, and systemic infection. Localized cutaneous infection, also known as erysipeloid, is the most common presentation and is usually described as a subacute cellulitis at the site of inoculation, which often resolves spontaneously within 3 weeks [1, 3]. Diffuse cutaneous disease and systemic disease that involves bacteremia with or without endocarditis is uncommon [4]. Other uncommon manifestations reported include pneumonia, abscesses, meningitis, endophthalmitis, osteomyelitis, and septic arthritis [1, 2].

Joint infection is a common occurrence in animals infected with E. rhusiopathiae [5]. However, septic arthritis of both native and prosthetic joints is a rarely reported manifestation of the disease in humans. Only 6 native joint infections and 4 PJIs with this organism have been reported previously, which are summarized in Table 1. The majority of clinical presentations were of a large joint chronic monoarthritis that developed over several months. Two cases had an acute presentation within days of presumed exposure, one after a penetrating injury to the shoulder [10] and the other occurring early after arthroscopic knee surgery [11]. Systemic symptoms and signs were absent in most cases, with only 2 patients reporting fever. CRP was measured in most and was moderately elevated.

A previously surmised finding is that many of the reported cases had immunosuppressive conditions including chronic lymphocytic leukemia, systemic lupus erythematosus, diabetes mellitus, chronic renal failure, hemodialysis, and corticosteroid use that likely predisposed them to invasive E. rhusiopathiae infection [12]. In this case, no immunosuppressive conditions were known to exist. However, previous and recent surgery as well as the presence of prosthetic material also likely increase patient propensity to develop intra-articular infection with this organism, if conditions for exposure exist.

A definitive diagnosis of E. rhusiopathiae infection largely still relies on culture techniques, which have variable sensitivity depending on the organism burden present in the specimen collected [1]. While organisms were only seen on direct gram stain of surgical tissues from the current joint infection case and specimens of one of the previously reported cases, E. rhusiopathiae was eventually recovered from all cases by culture methods. Identification using basic and automated biochemical systems as well as mass spectrometry technology, such as matrix assisted laser desorption/ionization time of flight, reliably identifies this organism [13]. Molecular diagnostic techniques, including sequencing of the conserved regions of the gene encoding 16S rRNA, are also a specific way to confirm identification [14]. Use of species-specific molecular diagnostic tests (i.e. polymerase chain reaction) directly on clinical specimens such as joint fluid may also be a sensitive and specific way to detect the presence of E. rhusiopathiae genetic material, particularly if there is high clinical suspicion and culture methods fail to recover the organism.

E. rhusiopathiae is intrinsically resistant to aminoglycosides and vancomycin but susceptible to penicillins, broad spectrum cephalosporins and fluoroquinolones, with no documented resistance to these agents [1, 2, 15]. The treatment of choice for localized and systemic infections is penicillin or ampicillin [1‐4]. The ideal antibiotic regimen and duration for joint infections with this organism is unknown. It is recommended that native joint infections caused by other more common organisms be managed with a combination of drainage or surgery and antibiotic therapy [16, 17], and that PJIs be managed with surgery and antibiotics [18]. Management of previously reported E. rhusiopathiae joint infection cases has largely involved a combination of surgical intervention and antibiotic therapy (Table 1). Antibiotic decisions should be based on in vitro susceptibility results and individualized for each patient.

Transmission of E. rhusiopathiae to humans occurs after exposure to animals carrying or infected with the organism, or to their products [1‐4]. Transmission following exposure to environmental sources harbouring the organism has also been reported [1]. The organism usually enters through non-intact skin, although infection following ingestion has been reported as well [1, 3]. In 4 of the 10 previously reported joint infection cases, E. rhusiopathiae was thought to gain access to the affected joint by contiguous spread from an overlying skin infection. One joint was presumed to be have been seeded hematogenously from an infected aortic valve [19]. The pathogenesis of the remaining 4 cases was unknown. The pathogenesis of the present case is unclear, however she did have a healing operative wound from the recent surgery when she reported handling wild muskox and other animal meat. E. rhusiopathiae may have been inoculated into this wound while she was handling meat from these potential source animals and, like previous cases, spread contiguously into her knee joint.

Most infections with E. rhusiopathiae are described in the setting of an occupational exposure to domestic animals with veterinarians, butchers, farmers, and fishermen being among those at highest risk [1, 2, 20]. Swine are described as the major reservoir for human infection that occurs in occupational settings, however multiple other animals are known to carry the organism, including Arctic marine mammals, fish, foxes and muskoxen [1, 2, 5, 6, 8]. Therefore, hunting and food practices of local populations in the Canadian Arctic are also known to increase risk for infection with E. rhusiopathiae [1, 6]. Multiple large-scale die-offs of muskoxen due to E. rhusiopathiae that have been reported in the Canadian Arctic in recent years, many on the same remote island as the case reported here [6], signal a changing epidemiology of this infection in mammals that may have an impact on human health in the region.

A comparison of whole genome sequences (WGS) generated from E. rhusiopathiae isolates found in these muskox carcasses between 2010 and 2013 previously revealed that a single strain was associated with die-off events across two large islands [21]. As part of an ongoing surveillance program in the Canadian arctic, further isolates collected in 2017 from muskoxen and seals on the same islands, and from muskoxen on a neighbouring but distant island identified that this same strain is still broadly circulating (Mavrot et al., unpublished). However, based on WGS comparisons, the isolate from this human case of PJI is not genetically related to the E. rhusiopathiae strain found in these arctic animal species and is more closely related to isolates found in poultry species as well as swine isolates from Canada and Belgium (Fig. 1), although the patient had no reported exposure to poultry or swine. This suggests that the patient acquired her infection from another animal or environmental source which WGS was not able to elucidate in this case given lacking epidemiologic linkages. Further investigations into the epidemiology of this pathogen in the Canadian Arctic, including the use of WGS and expanding testing to migratory birds and fish, could help to assess the likelihood of other animal and/or environmental sources as the cause of human zoonotic infections [21].

The prevalence of zoonotic infections in humans, like those caused by E. rhusiopathiae, is influenced by a number of factors including the burden of disease within the animal reservoir(s) [22, 23]. Animals may become more susceptible to a particular organism as changes to host-environment-pathogen interactions occur [6, 23]. One such driving factor cited for the changing epidemiology of many zoonotic infections is climate change [22‐24]. It has been postulated that the prevalence of E. rhusiopathiae in certain animal host populations in the Canadian Arctic, such as muskoxen, is increasing partly as a result of climate change [6]. Kutz et al. who have been examining the epidemiology of E. rhusiopathiae in the Canadian Arctic expressed concern about the potential for zoonotic infection and emphasized the need for public health education and messaging to ensure that these important food sources continue to be harvested and handled in a way that prevents human exposure [6]. Guidelines for hunters in the Canadian North recommend that animals found dead not be touched or eaten, advise against cutting into animal parts that look abnormal, and whenever in doubt, to cook the meat or fish well [25, 26].

Infections with E. rhusiopathiae are not reportable to local or national health authorities in Canada. Without an epidemiological database for tracking infections, it is unknown if incidence of this infection in humans is increasing. Nonetheless, changes to the burden of disease in host animal populations are likely to result in an increased risk to those who rely on and are exposed to these animals and their environment on a regular basis. Our case experienced consecutive zoonotic infections of a prosthetic joint with B. suis followed by E. rhusiopathiae. Given the spread of E. rhusiopathiae among local wildlife in the Canadian Arctic, and the increased potential for human infection, consideration should be given to public health surveillance of this infection in this geographic region.