Statistics from Altmetric.com
Septic arthritis should be excluded in patients with new onset inflammatory polyarthritis and a history of organ transplant on immunosuppressive drugs or diabetes. Unusual organisms or polymicrobial joint infection may be present, even in the absence of bacteraemia or abnormal radiographs, and a history of prior infections may provide valuable clues.
A 35 year old white man was seen for pain and swelling in both wrists and the right knee for about three weeks. He was diagnosed with Wilson’s disease at age 23 and underwent orthotopic liver transplant for cirrhosis one year later. Past history included personality disorder, choreiform movements, hypertension, diabetes mellitus, and renal insufficiency (serum creatinine 220–330 μmol/l). There were no symptoms of carpal tunnel syndrome. The drugs he was taking were prednisone 10 mg/day, mycophenolate mofetil 1000 mg twice daily, atenolol, and insulin. He had taken cyclosporin until six months previously, but it was discontinued because of worsening renal insufficiency. He had an intrahepatic abscess due to Enterococcus faecalis six months previously, which was treated with drainage and several weeks of intravenous ampicillin and gentamicin through a Hickman catheter in the left subclavian vein. The catheter was removed about two months before the initial clinic visit; the tip was not sent for culture.
On examination, there was proliferative synovitis in both wrists without limitation of range of motion. There was a large effusion and limited flexion of the right knee. Radiographs of the hands, wrists, knees, and shoulders were unremarkable; no chondrocalcinosis was noted. Seventy five millilitres of serosanguinous fluid were aspirated from the right knee (white blood cell count 9.75×109/l; red blood cell count 8.5×109/l). No crystals were observed and the Gram stain was negative. Synovial fluid culture grew Enterococcus faecalis, Enterobacter sakazakii, and Candida albicans. Arthrocentesis of the left wrist yielded no fluid. Blood cultures were negative.
Abdominal magnetic resonance imaging (MRI) disclosed no hepatic abscess. Left wrist MRI showed proliferative synovitis but no abscess. Because of renal insufficiency, negative blood cultures, and absence of abscesses, he was treated with intravenous ampicillin and fluconazole rather than amphotericin. After one week in hospital, all three joints improved dramatically, and the arthritis completely in another week. He was given a six week course of oral ciprofloxacin and fluconazole after discharge. One month after completing a course of antibiotics, he presented with recurrent pain and swelling in both wrists and the right knee. Synovial fluid from the right knee again grew Candida albicans; fluconazole was re-instituted. His joint symptoms resolved completely and have not recurred during suppressive treatment, which he has taken for about 12 months.
Given this patient’s history of diabetes, prior hepatic abscess, and Hickman catheter, and immunosuppressive drugs (mycophenolate and corticosteroids), septic arthritis was a strong possibility. Candidal arthritis is rare in immunocompetent subjects, but occurs in intravenous drug users and immunocompromised subjects. In this patient, transient candidaemia, possibly from a contaminated Hickman catheter, might have led to septic arthritis. Candidal arthritis most commonly affects the knee or hip, followed by the shoulder and ankle,1 and axial joints may be affected in intravenous drug users. Blood cultures are usually negative. In addition to adequate drainage of the joint, amphotericin B and oral or intravenous fluconazole have been used successfully for the treatment of Candida arthritis.2 Because of the patient’s renal insufficiency and lack of systemic illness, fluconazole was given for a prolonged course. Typically, in the patient with normal renal function, 2–3 weeks of amphotericin B followed by prolonged treatment (6–12 months) with fluconazole is required for successful treatment.2,3
Enterococcus faecalis, previously isolated from a hepatic abscess, might have seeded the joints through haematogenous spread. Enterobacter sakazakii (formerly called yellow-pigmented E cloacae) is an opportunistic Gram negative bacillus. Enterobacter strains rarely cause primary human disease, but frequently colonise patients in hospital, and can cause wound, respiratory, and urinary tract infections.4Enterobacter sakazakii meningitis or sepsis has been reported in neonates fed contaminated powdered milk. Only four cases of infection have been reported in adults,5 including two bronchopneumonia (both coinfected with Staphylococcus aureus), and one patient each with biliary sepsis and sepsis with acute respiratory distress syndrome. The role of Enterobacter sakazakii in this patient is difficult to ascertain; it may represent a non-pathogenic coinfection.
Wilson’s disease (hepatolenticular degeneration) is a rare autosomal recessive condition caused by mutations in the ATP7B gene, which encodes a copper transporting P type ATPase expressed primarily in the liver and kidney.6 As liver function declines, excess copper accumulates in the brain, cornea, and kidney. Arthropathies associated with Wilson’s disease are fairly common7–9 and usually affect large joints, but a distribution similar to rheumatoid arthritis has been reported. Radiographic changes include chondrocalcinosis and premature osteoarthritic changes such as osteophytes, subchondral cysts, and cartilage loss.10 The joint disease is thought to be in part due to aberrant copper deposition in the cartilage.9 Liver transplant generally ameliorates neurological and psychiatric manifestations,6 but it is unclear whether it helps arthropathy.
In summary, this case highlights the differential diagnosis of new onset arthritis in diabetic transplant patients taking immunosuppressive drugs. Arthritic problems of patients with Wilson’s disease are discussed. Successful suppressive treatment of Candida arthritis with fluconazole was accomplished in this patient.
The authors thank Dr Gene Ball for critical review of the manuscript.