Peptostreptococcus canis and Bacteroides pyogenes prosthetic joint infection

We present a case of acute prosthetic joint infection caused by Peptostreptococcus canis and Bacteroides pyogenes in a 53-year-old patient with systemic lupus erythematosus, operated on successfully with a primary knee replacement. At discharge, she had an eschar on the surgical wound, which is why it started to drain after her stitches were removed. She was readmitted 20 days later with a fever, suppuration and wound dehiscence. The patient was operated on urgently; the wound was cleaned, surgical debridement performed and the polyethylene replaced. Empirical treatment with intravenous vancomycin and ceftazidime was initiated. Four intraoperative samples (osteoarticular biopsy, wound exudate, joint fluid and synovial tissue) and the replaced polyethylene were sent to the microbiological laboratory for culturing.

When the samples were Gram-stained, with the exception of the polyethylene, abundant polymorphonuclear leukocytes and gram-positive cocci were observed. Culturing was done in blood agar, MacConkey agar, Schaedler agar and Schaedler kanamycin–vancomycin agar (Becton Dickinson). The polyethylene was processed using sonication and was cultured in the above media as well as in thioglycollate broth (Becton Dickinson). Turbidity was observed in the broth after 24 h and growth in the Schaedler agar and Schaedler kanamycin–vancomycin agar after 48 h of incubation in anaerobic conditions. The isolates were identified using mass spectrometry (MALDI-TOF) as P. canis and B. pyogenes.

Susceptibility testing was performed with gradient-diffusion strips (E-test, bioMérieux) in Schaedler agar with a reading after 48 h of incubation in an anaerobic atmosphere.¹ Both isolates were susceptible to penicillin, amoxicillin/clavulanic acid, piperacillin/tazobactam, imipenem, clindamycin and metronidazole. Antibiotic therapy was replaced with parenteral penicillin and, subsequently, rifampicin was introduced. Ten days after the debridement, due to persistent signs of infection, the prosthesis was removed and a gentamicin- and clindamycin-loaded spacer was inserted. Two intraoperative samples (osteoarticular biopsy and joint fluid) and the prosthesis were sent to the laboratory for culturing. Polymorphonuclear leukocytes were observed in the staining of the joint fluid sample. The culture of the samples was negative, except in the sonication fluid of the prosthesis, in which B. pyogenes was isolated. Progress was favourable, with the patient treated with intravenous ampicillin and, at discharge, with oral clindamycin until six weeks of treatment were completed. Six months after the prosthesis was removed, a new prosthesis was re-implanted. Cultures of intraoperative samples were negative.

In our literature review we did not find any reported case of prosthetic joint infection caused jointly by P. canis and B. pyogenes. P. canis is a gram-positive anaerobic coccus found in the subgingival plaque of dogs, while B. pyogenes is a gram-negative anaerobic bacillus found in the oral microbiota of dogs and cats.^2,3 Wound infections caused by contact with animal saliva tend to be polymicrobial and reflect their oral microbiota. Prosthetic joint infection due to these microorganisms is rare. When it does occur, it tends to arise in knee prostheses and be associated with bites, scratches or licks from animals.⁴ Our patient lived with a dog, although she denied having been bitten or scratched. Contact with a dog in the presence of an open surgical wound supports the aetiology of the infection. Close contact with pets may be a risk that these patients should be warned about.

Prosthetic joint infection caused by anaerobic microorganisms is rare (3–6% of cases), but is associated with significant morbidity and a poor prognosis.⁵ Most are caused by Propionibacterium spp., which is mainly responsible for chronic infections. Peptostreptococcus spp. and Bacteroides spp. are reported less frequently, generally in polymicrobial cultures, and associated with acute, post-operative or haematogenous infections.⁵ Infections caused by Propionibacterium acnes are more indolent, while those caused by Peptostreptococcus spp. and Bacteroides spp. are associated with greater inflammation and suppuration.⁵ The aetiological diagnosis is based on proper sample taking and preserving them in anaerobic conditions until they are processed.⁶ The advance in microbiological diagnostic techniques, such as the use of MALDI-TOF, has allowed these rarely reported microorganisms to be identified with greater precision.

Antibiotic therapy is primarily administered with beta-lactams, in accordance with the antibiogram results. In acute, streptococcal prosthetic joint infections treated with implant retention, the combination of a beta-lactam (or levofloxacin) with rifampicin has been suggested due to its alleged activity against stationary-phase bacteria in biofilms according to the findings of two observational studies, where a better prognosis is described.^6,7 However, its in vitro synergy and clinical efficacy in implant-related infections due to anaerobes and streptococci is yet to be demonstrated.