INTRODUCTION
Infection following total hip arthroplasty (THA) is a devastating complication affecting 1% to 2% of all THA patients1,2 and is the cause of about 10% of all THA revisions.3 Brucellosis is a zoonosis that generally causes systemic infection and affects osteoarticular tissue in 10% to 85% of patients.4,5 Brucellar infection usually affects the large joints or sacroiliac joints causing spondylitis, tenosynovitis, or osteomyelitis. Nonetheless, adequate antibiotic treatment can control the disease. Brucellar infection of implants following total joint arthroplasty is very rare.6
CASE REPORTS
Case 1
In December 1998, a 66-year-old woman (weight, 62 kg; height, 1.61 m) presented to the Hospital Ramon y Cajal in Madrid, Spain with pain in the thigh and groin areas. The pain was initially mild and mechanical but had increased and begun appearing during the night in the previous weeks requiring the use of analgesics. The patient was a housewife with a history of chronic low back pain and prolonged contact with cattle. She had undergone an uncemented THA to treat a hip fracture 3 years previously. She had no history of immunosuppressive conditions, diabetes, or corticosteroid use. There was no fever or local swelling. C-reactive protein level and globular sedimentation rate were increased. Radiological examination demonstrated a well-fixed femoral stem (Fig. 1a), but radiolucent lines appeared in more than 50% of the surface of the acetabular cup. Articular aspiration was performed, and cultures of the articular liquid were found positive for Brucella abortus.
A 2-stage revision THA was planned to correct septic loosening of the implant. Abundant purulent liquid was present in the hip joint and both components were found loosened intra-operatively. Frozen tissue pathology demonstrated acute and chronic inflammation. A spacer made with a Thompson hemiarthroplasty stem was covered with a mix of methyl methacrylate 40 g, cefotaxime 3 g, and streptomycin 1 g (Fig. 1b). Operating time was 130 minutes and blood loss was 1200 cm^sup 3^.
Oral rifampicin 900 mg and doxycycline 200 mg were administered daily for 6 weeks postoperatively. Four months later, a hybrid THA was implanted and the defect in the metaphyseal femur was filled with morselised allograft prior to cementing the femoral stem. No antibiotic therapy was used after reimplantation.
At postoperative 5.5 years, the patient was asymptomatic and able to walk without crutches. Radiography revealed a normal THA with well-fixed components (Fig. 1c).
Case 2
In June 1994, a 71-year-old man (weight, 82 kg; height, 1.74 m) was admitted to the Hospital Ramon y Cajal in Madrid, Spain with persistent mechanical hip pain that was not controlled by non-steroidal anti-inflammatory drugs. Medical history did not indicate immunosuppressive conditions, diabetes, or corticosteroid use. The patient had worked in an agricultural field during his active life. He had severe osteoarthritis and received a porous-coated uncemented THA. The patient developed mechanical hip pain again 3 years later, and radiography revealed a loosened acetabular component. A revision THA was then performed to remove the cup and re-implant a new uncemented cup. The new cup was complicated by postoperative dislocation and required open reduction and modification of the orientation.
In October 1999, the patient was admitted to our hospital again with painless suppuration of the hip joint through the scar from previous surgeries. Radiography demonstrated a well-integrated THA (Fig. 2a). Urgent drainage and debridement were performed, and both components were left in situ as they were solidly integrated. Operating time was 70 minutes and blood loss was 400 cm^sup 3^. Antibiotic treatment with teicoplanin 400 mg daily was started.
Postoperative cultures were found positive for Brucella melitensis. Indirect Brucella Coombs test showed a titer of 1/2560 (positive, >1/40) and a Brucella standard tube agglutination test was positive with a titer of 1/640. Antibiotic treatment was changed to rifampicin 900 mg, streptomycin 1 g, and doxycycline 200 mg daily. One month later, streptomycin was stopped and the rest was continued for 6 months.
At postoperative 5 years, the patient was asymptomatic and able to walk without crutches. Radiography revealed a normal THA (although radiolucent lines were shown in less than 30% of the acetabular surface, we considered the implant was well fixed) [Fig. 2b], and the results of the Brucella agglutination test had dropped to 1/8.
DISCUSSION
The incidence of infection following primary total hip or knee arthroplasty varies from 1% to 5.6%,1,2,7 and is mainly caused by aerobic gram-positive microorganisms2 including coagulase-negative staphylococci and Staphylococcus aureus. Gram-negative aerobic microorganisms and anaerobes are rare and account for less than 10% of infections.2 Fungal or mycobacterial infections are extremely unusual. The treatment of choice is a 2-stage management comprising extraction of the implants and reimplantation after a period of antibiotic treatment with or without a spacer.7 Conservation of the implants and the use of suppressive antibiotic therapy are rarely indicated and are reserved for patients with well-fixed implants or those who are unfit for surgery.7
Brucellosis is a common zoonosis worldwide but predominantly affects the Persian Gulf, South America, and Mediterranean countries.8 It is caused by a variety of microorganisms from the Brucella family and can be transmitted by inoculation through conjunctivae, cuts and abrasions in the skin, or by ingestion of unpasteurised milk or infected dairy products. Brucellosis is a systemic infection with non-specific symptoms such as sweats, anorexia, fatigue, weight loss, fever, lymphadenopathy, and hepatosplenomegaly.8 The most frequent osteoarticular infections are sacroiliitis and arthritis of peripheral large joints (hips and knees).4 The mechanism of infection is presumed to be haematogenous dissemination of the microorganism.
Only a few cases of brucellar infection following total joint arthroplasty have been reported, and thus the correct management is still unclear. Diagnosis is often difficult as patients usually present with no systemic symptoms. Only local symptoms (night pain, swelling, and suppuration) or mechanical loosening are typically present.6 Radiological studies can be used to diagnose component loosening. Technetium or gadolinium bone scans can help to determine if there is loosening, infection, or both. Aspiration of the joint is a controversial diagnostic tool for infection after arthroplasty2,3 due to a false-positive rate of 13%.9 Nonetheless, the procedure can be used to determine the causal agent if the diagnosis is unclear.
Adequate treatment depends on how and when the diagnosis is achieved. When one or both of the components are loosened and the preoperative aspiration, bone scan, or intra-operative findings are suggestive of septic loosening, the treatment of choice is a 2-stage procedure with extraction of the components, antibiotic therapy, and subsequent reimplantation. Good results have been reported in 3 patients (2 hips and one knee) treated with 2-stage procedure and antibiotic therapy (doxycycline 100 mg and rifampicin 600 mg daily) for 6 weeks before and after re-implantation.6 Good results were also obtained in THA patients who underwent 2-stage reimplantation using a longer pro-implantation antibiotic scheme of streptomycin 1 g daily for 3 weeks and doxycycline 100 mg and rifampicin 600 mg daily for 3 months, without post-implantation antibiotics.10 These results were nonetheless considered preliminary because of the very short follow-up period of less than one year. The antibiotic regimen for our first patient was a combination of doxycycline and rifampicin for 6 weeks between extraction and re-implantation. The results were satisfactory after more than 5.5 years' follow-up. Jones et al.11 reported a THA implant loosening caused by B abortus treated successfully with a one-stage re-implantation and antibiotic therapy for one year; however, with a follow-up period of only 2 years, there is not yet enough evidence to recommend this course of treatment. Kasim et al.12 reported one case of brucellar infection diagnosed through intra-operative cultures during a revision THA for aseptic implant loosening. The patient was treated with doxycycline 100 mg and rifampicin 600 mg daily for 5 months postoperatively. The infection seemed to be controlled after 4 years of follow-up.
Correct management for brucellar infection diagnosed through aspiration or over a well-fixed THA or total knee arthroplasty (TKA) is still unclear. Agarwal et al.13 reported a case of bilateral TKA infected with B melitensis treated with a prolonged course of antibiotic therapy (rifampicin and trimethoprim-sulphamethoxazole) of 19 months. Shorter antibiotic courses have also given good results: an early postoperative TKA infection was successfully treated with streptomycin 1 g for 3 weeks and doxycycline 100 mg daily for 5 months, which was then changed to trimethoprim-sulphamethoxazole for 2 more months.14 A TKA infection that developed 14 months after surgery was treated successfully with streptomycin 1 g for 3 weeks and doxycycline 200 mg and rifampicin 900 mg daily for 6 weeks.15 However, the follow-up periods of these 3 cases were only 19 months, 24 months, and 8 months, respectively. This is insufficient to evaluate the usefulness of this suppressive therapy. In our second patient, antibiotic treatment for 6 months seemed to control the infection completely even after 5 years, indicating that the described antibiotic cycle may control the disease and perhaps even cure the infection.
Both B melitensis and B abortus have been isolated in THA and TKA infections, with B melitensis being the most frequently isolated species. The low prevalence of these infections prevents a detailed analysis of the specific pathogenicity of each bacterial strain.
We believe that a standard 2-stage revision arthroplasty is the treatment of choice for loosened THA or TKA infected with Brucella. When the implants are well fixed, debridement and antibiotic treatment without re-implantation surgery is suggested, because the Brucella pathogen is highly susceptible to antibiotics, less aggressive, and intracellularly ubicated. These characteristics can facilitate effective suppressive antibiotic treatment-an option generally discounted in other infections.
© 2006 Western Pacific Orthopaedic Association Provided by ProQuest LLC. All Rights Reserved.
Source: Journal of Orthopaedic Surgery
