mja.com.au | The Medical Journal of Australia

Home | Issues | MJA shop | MJA Careers | Contact | Topics | Search | RSS  | Login | Buy full access

Diagnostic Dilemma

“My foot hurts”: a flare of rheumatoid arthritis?

Manish Dugar, Wayne A Rankin, Emily Rowe and Malcolm D Smith
MJA 2009; 190 (7): 392-393

A 56-year-old man with a history of rheumatoid arthritis presented with a 2-day history of worsening pain in his left foot. Treatment with high-dose steroids was of no benefit, hence a diagnosis of septic arthritis was considered. However, the patient’s condition deteriorated despite empirical antibiotic therapy. Following persistent investigation, the cause was identified as a fastidious Legionella longbeachae infection, and appropriate antibiotic therapy led to complete resolution of the sepsis. This emphasises the importance of considering infections with atypical organisms in patients on immunosuppressive therapy.

Clinical record

In July 2008, a 56-year-old retired man presented with a 2-day history of worsening pain in his left foot. His medical history included seropositive active rheumatoid arthritis (RA), type 2 diabetes mellitus with microvascular complications, and idiopathic dilated cardiomyopathy managed with an implantable cardioverter-defibrillator and permanent pacemaker (left ventricular ejection fraction: 12% in 2006, 42% in 2008). His RA had been difficult to control — intermittent flares, usually involving his feet, were managed with high-dose steroids. Before presentation, the patient was taking weekly oral methotrexate (20 mg) and daily prednisolone (7.5 mg). Leflunomide, hydroxychloroquine and sulfasalazine therapy had failed to control his RA in the past and were withdrawn because of adverse reactions. Treatment with rituximab, an anti-CD20 monoclonal antibody, in September 2007 (two 1000 mg infusions, administered 2 weeks apart) had also failed to control his RA, and tumour necrosis factor-α (TNF-α) inhibitors were contraindicated owing to his cardiac disease.1

On examination, the patient had diffuse swelling with tenderness over the small joints of his left foot and ankle, with minimal erythema. He was not constitutionally unwell; however, his C-reactive protein (CRP) level was 64 mg/L (reference range [RR], 0–10 mg/L), and he was anaemic (haemoglobin level, 86 g/L [RR, 130–175 g/L]) and leukopenic (white cell count, 2.0 × 109/L [RR, 4.0–11.0 × 109/L]; neutrophil count, 1.6 × 109/L [RR, 1.5–7.5 × 109/L]). A provisional diagnosis of RA flare was made, the patient was admitted to a private hospital, and high-dose oral prednisolone therapy (50 mg/day) was begun.

Over the next 2 days, there was no improvement in the patient’s left foot swelling and pain, and his haemoglobin and CRP levels worsened further (haemoglobin level, 77 g/L; CRP level, 102 mg/L). However, he remained afebrile. He was transfused with three units of packed red cells. A whole body bone scan was performed to test for underlying infection, and it revealed intense tracer uptake in the region of the fourth and fifth tarsometatarsal joints of the patient’s left foot (Box). He subsequently developed fever, chills, tachycardia and painful swelling in his left wrist. In view of the worsening clinical features and the increasing CRP level, an alternative diagnosis of septic arthritis or osteomyelitis was strongly suspected, rather than a flare of RA. Blood cultures were ordered, and intravenous flucloxacillin therapy was begun; methotrexate was withdrawn, and the dose of prednisolone was reduced to 25 mg daily. A computed tomography scan of the patient’s left foot showed only soft tissue swelling, with no cortical destruction, fracture or collection.

The patient was then transferred to our institution as his treating physician was to go on leave. On examination, he was febrile, had synovitis of his left wrist, and had marked swelling and erythema of his left foot, extending to the ankle. A Gram stain of an aspirate from his left wrist, collected on admission to our hospital, revealed occasional gram-negative bacilli. Further blood cultures were ordered and flucloxacillin was substituted with meropenem. The next day, meropenem was changed to cefepime for better coverage of possible Pseudomonas infection, pending blood culture results. The cellulitis of the patient’s left foot did not resolve during the next 2 days, and possible septic arthritis developed in the right fourth finger (distal interphalangeal joint).

Despite ongoing antibiotic therapy, the patient’s condition deteriorated. He became neutropenic (neutrophil count, 0.3 × 109/L), and the swelling of his left foot extended to the knee. His haemoglobin level dropped to 82 g/L, and he required transfusion of two additional units of red cells. He had daily temperature spikes, but blood and synovial fluid cultures were still showing no growth at 4 days after the initial aspirate and blood samples were taken. To identify the apparently fastidious bacterium noted in the initial wrist aspirate, synovial fluid from the left wrist was plated onto various selective media. Colonies grew on buffered charcoal yeast extract medium, and a Legionella species was suspected. Azithromycin and moxifloxacin were added to the treatment regimen 5 days after the patient was transferred to our institution, and cefepime was withdrawn. The neutropenia was treated with subcutaneous injections of granulocyte colony-stimulating factor.

The patient’s condition improved dramatically during the next few days; complete resolution of cellulitis and joint swellings occurred with ongoing appropriate antibiotic therapy. Polymerase chain reaction analysis confirmed that the isolated bacterium was a Legionella species, which was subsequently shown by 16S ribosomal RNA gene sequencing to be Legionella longbeachae. Extensive review of environmental risk factors did not reveal a recent source of exposure to this organism, and there had been no precedent pneumonia. He was discharged home and prescribed a 6-week course of oral azithromycin and moxifloxacin. On follow-up at 6 weeks, his CRP level and white cell count had normalised, with complete resolution of symptoms.

Discussion

This case shows how challenging it can be to differentiate between flare and infection in patients with RA who present with worsening local symptoms. Adding to the diagnostic challenge in this case was the involvement of a particularly fastidious organism. To our knowledge, this is the only reported case to date of septic arthritis with L. longbeachae without preceding pulmonary infection in a patient with RA, although two cases of septic arthritis with Legionella pneumophila in immunocompromised patients without lung involvement have been reported.2,3

Extrapulmonary infections by Legionella species are rare and are generally thought to arise from haematogenous spread after pulmonary infection. L. longbeachae is responsible for about 4% of cases of community-acquired legionellosis worldwide, and about 30% of cases in Australia and New Zealand.4 It is commonly found in soils and potting mixes, and occasionally in water.5 However, our patient was not a gardener, and we could not identify any other recent source of exposure to L. longbeachae.

Patients with RA have an increased risk of infection because of the disease itself and immunosuppressive therapy.6 Therefore, a differential diagnosis of underlying infection with typical and atypical organisms should be considered when patients with RA present with disease flares that do not respond to appropriate escalation of immunosuppressive therapy.

Several factors complicated the management of this patient’s condition. In particular, his RA was difficult to control; he also could not be administered TNF-α inhibitors (owing to his underlying moderate-to-severe heart failure) as per current expert consensus recommendations.1 Abatacept (a selective blocker of T-cell co-stimulation) is being considered as a future treatment option in this patient. He also had anaemia of chronic disease secondary to the active RA, for which erythropoietin therapy was begun after discharge; he did not require further blood transfusions over the 6 months after discharge. In addition, he was leukopenic; this was likely to be secondary to the sepsis because, although he had previously been treated with rituximab, his lymphocyte count was normal 1 month before presentation and at follow-up.

Our patient presented with an apparently simple flare of RA that developed into life-threatening sepsis with neutropenia. This case illustrates the need for a high index of suspicion of infection with unusual organisms when symptoms suggestive of active arthritis in such patients do not respond to standard treatment.

Delayed regional bone scans of a patient with persistent pain and swelling of the left foot

Intense tracer uptake is visible in the left foot — in the fourth and fifth tarsometatarsal joints, with flare extending along the shaft of the fourth metatarsal. A: Posterior view of both feet. B: Anterior view of both feet. C: Left lateral view of left foot. D: Right lateral view of left foot.

Competing interests

None identified.

Author detailsManish Dugar, MB BS, Registrar, Department of Immunology, Allergy and Arthritis1Wayne A Rankin, PhD, Medical Student2Emily Rowe, MB BS, Registrar, Department of Infectious Diseases1Malcolm D Smith, PhD, FRACP, Head of Rheumatology1

1 Flinders Medical Centre, Adelaide, SA.

2 School of Medicine, Flinders University, Adelaide, SA.

Correspondence: manish.dugarAThealth.sa.gov.au

References
  1. Danila MI, Patkar NM, Curtis JR, et al. Biologics and heart failure in rheumatoid arthritis: are we any wiser? Curr Opin Rheumatol 2008; 20: 327-333. <PubMed>
  2. Linscott AJ, Poulter MD, Ward K, Bruckner DA. Legionella pneumophila serogroup 4 isolated from joint tissue. J Clin Microbiol 2004; 42: 1365-1366. <PubMed>
  3. Bemer P, Leautez S, Ninin E, et al. Legionella pneumophila arthritis: use of medium specific for mycobacteria for isolation of L. pneumophila in culture of articular fluid specimens. Clin Infect Dis 2002; 35: E6-E7. <PubMed>
  4. Yu VL, Plouffe JF, Pastoris MC, et al. Distribution of Legionella species and serogroups isolated by culture in patients with sporadic community-acquired legionellosis: an international collaborative survey. J Infect Dis 2002; 186: 127-128. <PubMed>
  5. Montanaro-Punzengruber JC, Hicks L, Meyer W, Gilbert GL. Australian isolates of Legionella longbeachae are not a clonal population. J Clin Microbiol 1999; 37: 3249-3254. <PubMed>
  6. Doran MF, Crowson CS, Pond GR, et al. Frequency of infection in patients with rheumatoid arthritis compared with controls: a population-based study. Arthritis Rheum 2002; 46: 2287-2293. <PubMed>

(Received 18 Sep 2008, accepted 26 Jan 2009)


Home | Issues | MJA shop | Terms of use | MJA Careers | More... | Contact | Topics | Search | RSS 

mja.com.au | The Medical Journal of Australia  

©The Medical Journal of Australia 2009 www.mja.com.au PRINT ISSN: 0025-729X ONLINE ISSN: 1326-5377