Resolution of recurrent fusarium arthritis after prolonged antifungal therapy.

S everal recent reports have highlighted the emerging importance of Fusarium as a pathogen in immunosuppressed patients. Fusarium is a genus of filamentous molds occurring abundantly in soil and plants. Fusarium does not usually cause serious disease in humans with intact immunity. When infection does occur, nails and corneas are the commonest sites. In immunocompromised patients, however, invasive or disseminated infection can occur, and mortality rates as high as 75% have been described. Hematologic malignancy and severe neutropenia appear to represent special risk factors for dissemination, most commonly to blood, skin, sinuses, and lungs. Bone and joint infections are less common but have been described; in one 10-year review, Fusarium accounted for some 20% of fungal osteoarticular infections in patients with hematologic malignancy. In this report, we describe a case of joint infection with Fusarium, a mold of emerging importance in immunocompromised individuals. The patient was a 62-year-old man with acute myeloid leukemia (AML), presenting 5 weeks after induction chemotherapy with Ara-C and daunomycin. He complained of 1 week of left ankle pain and swelling, fevers, and drenching night sweats. On examination, his left ankle was warm, and an effusion was present. His left second metacarpophalangeal and proximal interphalangeal joints were swollen and tender. M4/M5 AML had been diagnosed 2 months previously. Induction chemotherapy was complicated by febrile neutropenia and erythematous skin nodules. Skin biopsy showed fungal hyphae (Fig. A), and culture of skin and blood grew Fusarium species. He was treated with intravenous voriconazole with resolution of fever and rash and discharged on oral voriconazole. At the onset of current joint symptoms, he continued on oral voriconazole and was not neutropenic. Analysis of synovial fluid from the left ankle showed a white cell count of 22,000 (75% neutrophils, 12% lymphocytes), with negative Gram stain and no crystals. Radiographs of the left ankle and hand showed soft-tissue swelling with preserved joint space. He was admitted for presumed polyarticular septic arthritis. Ceftriaxone and vancomycin were added, and his ankle effusion was aspirated daily. Initial blood and synovial fluid cultures were negative. On hospital day 8, the second of 5 synovial fluid cultures grew fungus, later confirmed to be Fusarium. Voriconazole and antibiotics were discontinued, and liposomal amphotericin (AmBisome) was started with azithromycin for synergy. The effusion resolved, but new skin lesions consistent with Fusarium subsequently appeared. Sensitivities requested during his initial fungemia returned, showing intermediate sensitivity (minimal inhibitory concentration, 2) to amphotericin B and voriconazole and sensitivity to caspofungin (minimal inhibitory concentration, 1.5). Caspofungin was added, his skin lesions resolved, and hewas discharged on azithromycin, AmBisome, and caspofungin. He was readmitted for consolidation chemotherapy with Ara-C and daunomycin 5 weeks after initiation of AmBisome. Despite ongoing AmBisome/azithromycin and caspofungin, his left ankle effusion recurred after onset of neutropenia, highlighting the critical role of neutrophils in Fusarium infection. A left ankle magnetic resonance imaging (MRI) showed a loculated joint effusion and a lesion consistent with osteomyelitis (Figs. AYC). Voriconazole and broad-spectrum antibiotics were added without improvement. When neutropenia resolved 8 days later, he defervesced, and ankle swelling improved. He received a total of 2.5 months of AmBisome/azithromycin, 6 months of caspofungin, and 8 months of oral voriconazole. Subsequent MRI showed resolution of the ankle effusion and osteomyelitis. Four years later, he remains in remission from his AML and has had an excellent functional recovery from his ankle arthritis. Early diagnosis and treatment of invasive fusariosis are critically important. In contrast to some fungal pathogens, blood cultures are frequently positive. Biopsy of affected tissue can permit rapid identification of hyphae, while culture is awaited. In our case, only 1 of 5 synovial cultures grew Fusarium, emphasizing that multiple arthrocenteses may be needed for diagnosis. Voriconazole, amphotericin, or posaconazole forms the mainstay of empiric therapy. Local resistance trends should assist selection of empiric agents in individual cases. Although susceptibility testing methods are not standardized, sensitivity testing remains crucial. In the present case, synovial fluid isolates were sensitive to caspofungin, despite the fact that Fusarium is generally considered resistant to echinocandins. Insufficient data exist, at present, to recommend firmly between the use of single agents or combination therapy, and duration of therapy is also unclear. Surgical resection of heavily infected tissue, removal of conduits such as potentially infected lines, and the use of colonystimulating factors to hasten neutrophil reconstitution should also be considered as adjunctive measures. In this case, surgical debridement was considered, but deferred pending reconstitution of his bone marrow. Recovery from neutropenia represents a particularly important prognostic factor in invasive fusariosis and often coincides with clinical improvement, as demonstrated dramatically in this case. However, as our case also illustrates, antifungal therapy should be continued while any clinical or radiographic features of active disease persist, and at least until all cultures are negative. It is hoped that the present report will serve to inform clinicians of this invasive organism’s potential to cause joint and bone infection in immunocompromised patients, CONCISE REPORT