CASE REPORT

A 32 year old man with a one month history of mild inflammation of the right elbow presented to our hospital on 19 May 2000. He had an unremarkable past medical history, which did not include any toxic habits or recent trauma. Bursal aspiration showed that the synovial fluid had inflammatory characteristics (leucocyte count 4.9×10⁹ cells/l (54% neutrophils), and a glucose level of 3.8 mmol/l), but there were no crystals and a fluid culture was negative. A diagnosis of olecranon bursitis was established, and conservative management (fluid aspiration) was decided on. Bursal effusion was repeated over the next four days, so a further aspiration was carried out and local injection with triamcinolone acetate (20 mg) was given. However 24 days later the pain worsened and swelling of the elbow recurred; laboratory synovial findings showed a leucocyte count of 15.7×10⁹ cells/l (60% neutrophils) and a low glucose level (0.8 mmol/l). Culture yielded a few colonies of Candida spp, but antifungal treatment was not started because it was considered that this might be caused by contamination. One month later (28 July), the patient presented to the emergency room owing to development of a new extremely painful episode of bursitis. After joint aspiration, a steroid injection was again given, but this time a fluid culture was not carried out.

On 1 August clinical symptoms persisted. Physical examination showed an increase in the size of the olecranon bursa. The patient had never presented with fever, arthralgias, or any general complaints. Laboratory studies, including a test for antibodies to HIV, were normal or negative. Magnetic resonance imaging was performed showing multiseptate bursitis; the adjacent structures were normal. A removal of 10 ml bursa fluid again yielded a positive culture for Candida that was later identified as C parapsilosis (Majadahonda (Madrid), National Centre for Microbiology). Antifungal sensitivity testing showed a minimal inhibitory concentration for amphotericin B of 1 mg/l, 5-fluorocytosine 0.25 mg/l, fluconazole 0.25 mg/l, itraconazole 0.03 mg/l, and ketoconazole 0.015 mg/l. By the end of August, oral fluconazole was started at a dose of 400 mg/day for seven days, and then 200 mg daily. Recovery was slow and the patient needed repeated drainage. As follow up cultures were still positive, on 27 September it was decided to carry out surgical debridement with complete excision of the olecranon bursa. This material was not cultured, but histopathological analysis was performed demonstrating pseudohyphae structures, without granulomatous reaction or foreign bodies. After bursectomy, the patient continued fluconazole treatment (same maintenance dose) for six weeks more. Six months later he is completely asymptomatic.

Infection of superficial bursae (olecranon, prepatellar, and infrapatellar) is generally associated with different occupations or physical activities. Local trauma may predispose micro-organisms to penetrate by the transcutaneous route.¹ Similarly, intrabursal steroid injection, a habitual therapeutic procedure, may lead to infection. Weinstein et al noted that development of infection after this procedure occurred in 12% of a series of cases.⁴ Most frequently bacteria cause infections, but unusual pathogens like fungi have also been described.¹Candida septic bursitis is extremely rare. After a thorough review of the Medline database (from 1966 to January 2001) using medical subject headings, and keyword searches that included “septic bursitis” and “Candida”, we found only five reports.^3,5–,⁸ Two caused by C albicans, two by C tropicalis, and another one by C lusitaniae (table 1⇓). Characteristically, in all the cases, and in the present report, different risk factors or underlying diseases were found. Four cases were caused by haematogenous spread and two induced by direct penetration, including our case. The olecranon bursa was affected in three cases, including the present report.

C parapsilosis is a well known cause of arthritis that has been described secondary to systemic dissemination in intravenous drug users,⁹ and also by direct inoculation secondary to catheterisations¹⁰ or intra-articular injections.⁷ It is not strongly associated with immunocompromised hosts, but rather with invasive procedures or prosthetic devices.¹¹ More recently C parapsilosis has emerged as an important nosocomial pathogen. This is the Candida species that is most commonly isolated from the hands of healthcare workers.¹² In contrast with other Candida species, colonisation with C parapsilosis rarely occurs before the onset of invasive infection, suggesting an exogenous source of infection.

Appropriate antifungal drugs to treat Candida infections are available, but appropriate drug levels in osteoarticular structures are difficult to achieve. So for successful treatment of this infection, surgery is sometimes required. Half of the patients with Candida SB reviewed needed surgery for complete resolution (table 1⇑).

We would like to summarise several aspects of the present report: Firstly, steroid injection must be carefully prescribed in order to avoid probable side effects like infection.⁴ Secondly, most cases of Candida SB are produced by haematogenic spread, secondary to disseminated infection, whereas the present case was almost certainly through direct inoculation. Thirdly, isolation of C parapsilosis was neglected at the start so that antifungal treatment was delayed, leading to the need for surgery. We consider that the diagnostic delay together with a rather low maintenance dose of fluconazole were critical for the very slow resolution of the infection; probably 400 mg/day would have been more suitable for an infection in a deep compartment.

Because unusual micro-organisms are difficult to recognise and anti-inflammatory drugs may mask the symptoms, a higher degree of awareness is necessary to achieve prompt diagnosis and successful treatment. Nevertheless, special care must be taken to avoid complicating side effects in iatrogenic manipulations, so preventive measures to reduce the incidence of infection must never be omitted.