Staphylococci, streptococci and Gram negative bacilli have been reported to be the most common causative organisms in infections after arthroplasty. Prosthetic infections caused by Brucella spp. are rarely described in the literature⁵. In addition, cultures are often unsuccessful because of the slowly growing and fastidious nature of this bacterium. In recent years, blood culture systems have been introduced into clinical practice, resulting in reduced detection times. Experience accumulated with the BACTEC 9240 instrument has demonstrated that the system enables detection of B. melitensis from blood cultures within the routine 7-day incubation period⁴.

Here, we report a case of brucellar monoarthritis of the knee, which underwent a total joint replacement 4 years ago, detected with blood culture system while no growth was detected on the conventional blood agar and chocolate agar media.

Figure 1: AP radiography of the knee

Figure 2: Lateral radiography of the knee

Cefazolin (1g, 3x1, i.m.) was started and she underwent arthroscopic lavage and debridement with the initial diagnosis of prosthetic knee infection. Sero-purulent synovial fluid was obtained during arthroscopy and was sent to our laboratory in a closed sterile syringe immediately, demanding a conventional culture. The loosening of the tibial and femoral component was checked with probe and no revision was made. Villous synovial hypertrophy was observed during the arthroscopy and arthroscopic synovectomy was performed.

In laboratory, synovial fluid was divided into two aliquots. One-half of the fluid (1.5 ml) was inoculated into an aerobic “Peds Plus blood culture bottle” and monitored by the BACTEC 9120 instrument (Becton Dickinson Diagnostic Instrument Systems, Sparks, MD, U.S.A.), and the second aliquot was plated onto sheep blood agar, chocolate agar, and MacConkey agar plates. After 3 days of incubation blood culture bottles were indicated as positive by the instrument. Detection of positive bottles was followed by performance of a Gram stain and subculture of the broth in sheep blood agar, chocolate agar, and MacConkey agar plates, and incubated at 350C in 5% CO₂. After 48 h, minute translucent colonies were seen on blood agar plates. Gram stain from blood culture bottles and blood agar culture showed gram-negative coccobacillus that was catalase, oxidase, and urease positive within 3 minutes.

These initial findings suggested the micro-organism to be Brucella spp. Further evaluation with Brucella antisera revealed that the bacterium was B. melitensis and the organism was confirmed as B. melitensis with VITEK II (Biomerieux, France) in a reference laboratory. After the isolation of Brucella in culture, the Brucella standard tube agglutination test was performed from the sera of the patient and it was positive, with a titre of 1/320. The patient was treated with doxycycline (2×100 mg/day orally) and rifampicin (1x 600mg /day orally) for 6 weeks. The symptoms slightly relieved, and she was discharged with the suggestion of weekly follow-ups comprising tube agglutination + liver enzyme level controls. A year later, the patient remained free of knee pain and able to perform her daily activity.

No epidemiological information could be obtained regarding the source of the infection. No previous contact with livestock or dairy products was reported.

Our patient presented initially with symptoms suggestive of prosthetic knee infection, but Brucella was not suspected on preoperative evaluation. To our knowledge according to pub med scanning, there are 3 previous reports of prosthetic knee infection^5,8,9. Our patient is the fourth case reported in the literature of a Brucella infection involving prosthetic knee, and the diagnosis was recognized postoperatively with the use of blood culture system for the culture of synovial fluid taken during arthroscopic surgery. Since the diagnosis was not suspected on preoperative evaluation, laboratory was not warned that Brucellae are suspected. This is not surprising since brucellosis has a variable and a wide spectrum of clinical manifestations and only searching for it when suspicious exposure exists, may lead to identification. Moreover, the laboratory data may usually be normal, making the diagnosis of brucellosis difficult for the physician¹. So, we would not be able to isolate the causative pathogen in this case, if we did not use the blood culture system besides conventional methods.

In conclusion, we emphasize the fact that Brucella infection of prosthetic joints should be considered in the differential diagnosis of prosthetic joint infection especially in endemic areas, so determination of Brucella titers preoperatively is warranted in these patients and the diagnosis of Brucella arthritis may be underestimated when inappropriate culture media are used.