Ruminococcus gnavus infection of a metal-on-metal hip arthroplasty resembling a pseudo-tumour in a 72 year-old woman with no intestinal symptoms

A 72-year-old woman was referred to the emergency department of a tertiary hospital under suspicion of left total hip infection and/or urinary infection. The patient complained of dysuria and left groin pain of several days duration, as well as temperature up to 39.5 °C, with no trauma. She had been implanted with a metal-on-metal big-head total hip arthroplasty five years ago. Physical exam revealed pain with groin palpation and with passive motion, and she had difficulty to walk. Blood analysis showed 18,980 g/l leukocytes, 87% neutrophils, C-reactive protein 180 mg/l and procalcitonin 4.12 ng/ml. Intravenous empiric antibiotic treatment was begun with cephazollin 1 g/8 h, gentamycin 240 mg/24 h and clyndamicin 600 mg/8 h. After 3 days urine culture yielded >100,000 CFU of Klebsiella pneumoniae and treatment was changed to oral amoxicillin-clavulanate 875 mg for 10 days. The patient improved clinically and returned to walk without hip or groin pain. Along the following two months she suffered three episodes of left hip dislocation treated conservatively with closed reduction under general anaesthesia. The radiographic control after the last episode showed incomplete reduction (lack of full seating of head inside cup). The total hip was surgically revised to a conventional metal-on polyethylene because of suspicion of pseudotumor interposition inside the joint.

One-stage uncemented exchange was performed, but only the acetabular cup and the femoral head were replaced. At the time of surgery, all biochemical parameters resulted normal. The removed periarticular tissue was examined histopathologically by microscopy to confirm the presence of a chronic inflammatory process with coagulation necrosis, hemosiderin and replacement of normal tissue with scar tissue. Five intraoperative specimens were obtained for microbiogical purposes: three tissue samples of the periarticular and intraarticular granuloma/pseudotumour, the retired prosthetic components (big head and old acetabulum), and bone tissue from the reaming of acetabular bone.

Specimens were aseptically collected and sent for microbiological analysis and culture within 2 h. All samples were processed independently. Microscopic examination of the samples showed an absence of polymorphonuclear cells and/or microorganisms. Saline solution was added to the biopsies prior to homogenization. Ultrasonication of the removed prosthetic implants was performed. The resulting suspensions obtained were plated onto 5% sheep blood Columbia agar medium incubated at 36 °C under aerobic and anaerobic conditions; onto Chocolate agar incubated at 37 °C in a 5% CO₂ atmosphere; and onto Schaedler agar + 5% sheep blood, Pheniletanol blood agar and Schaedler Neomycin Vancomycin agar + 5% sheep blood incubated at 36 °C under anaerobic conditions (all media were from bioMérieux, Marcy l’Etoile, France). Additionally, 1 ml of the suspensions were injected into a pair of aerobic and anaerobic culture bottles BACTEC Plus Aerobic/F and Plus Anaerobic/F incubated for 7 days in a BACTEC 9240 Blood Culture System (Becton Dickinson Microbiology Systems, Sparks MD, USA).

Four samples yielded positive cultures after 3 days of direct incubation on anaerobic conditions: the three granuloma/pseudotumor samples and the sonicate from the prosthetic components. Microorganisms appeared as translucent small colonies and Gram staining showed short Gram-positive diplococci. Few colony forming units grew from the direct solid medium and positive cultures were also obtained from anaerobic bottles from the granuloma/pseudotumor samples and from the prosthesis sonicate; the same colonies appeared in the subcultured anaerobic plates after 48 h.

The microorganism was identified as Ruminococcus gnavus (score: 2.2) by matrix-assisted laser desorption/ionization time of flight mass spectrometry, Maldi-TOF MS (Maldi Byotyper 3.0 System, Bruker Daltonics GmbH, Leipzig, Germany). The test was confirmed twice with the same result. Susceptibility of R. gnavus to antimicrobial agents was performed using agar diffusion epsilometer testing (Oxoid, Basingstoke, UK) on 5% sheep blood Columbia agar medium at 37 °C in anaerobic atmosphere. Minimum inhibitory concentrations (MIC) were determined for penicillin (MIC = 0.064 μg/ml), tigecycline (MIC = 0.064 μg/ml), vancomycin (MIC = 0.19 μg/ml) and linezolid (MIC = 1.5 μg/ml). It was also tested against amoxicillin–clavulanic acid, meropenem, clindamycin and metronidazole but results could not be interpreted due to a contamination and the strain could not be recovered from the frozen vial. According to the European Committee on Antimicrobial Susceptibility Testing criteria, the isolate was found to be susceptible to penicillin and vancomycin.

Intravenous empiric antibiotic treatment was begun with cephazollin 1 g/8 h, gentamycin 240 mg/24 h and clyndamicin 600 mg/8 h during the immediate post operative period. Seven days later, de-escalation was done and oral clindamycin 600 mg/8 h along 6 months was prescribed by treating orthopaedic surgeries according to the protocols established in the Hospital.¹ At last revision in out-patient clinic (10 months post-surgery), erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) resulted normal, the patient was absolutely asymptomatic, did not need any walking aid, climbed stairs and used public transportation, was able to walk up to 1 km without stopping, and showed no signs of left hip prosthesis dysfunction.

R. gnavus is a strictly anaerobic Gram-positive non-spore-forming coccus that may be motile or non motile. It belongs to the Clostridia class of the Firmicutes division. R. gnavus is found to be part of the intestinal flora in humans and in the rumen of animals such as sheep, cattle and goats.²

The only previously described human infections were two cases of bacteremia associated with diverticulitis in men,³ one case of total hip arthroplasty infection in a man with ulcerative colitis,⁴ and one case of septic arthritis⁵; all these infections were described in patients with gastrointestinal diseases. The patient described in this case report had no intestinal history or symptoms at all, considering the case as a blood stream infection. The macroscopic perioperative findings were compatible with the presence of granuloma/pseudotumor tissue surrounding the prosthesis.

An increased rate of pseudotumor formation has been described with metal-on-metal hip prostheses. Pseudotumors can present variable granulomatous reactions. Pain is the strongest predictor for pseudotumor presence.⁶

Pseudotumors are often associated with subclinical infection, may destroy massively host bone and soft tissue, and usually require revision surgery for treatment.⁷

In the present case, the pseudotumor samples yielded positive for R. gnavus, the same microorganism as the prosthetic joint sample.

We conclude that the microorganism mentioned should be considered as a potential cause of joint prosthesis infection.