Corticosteroid Use for Paradoxical Reactions during Antibiotic Treatment for Mycobacterium ulcerans

Buruli or Bairnsdale Ulcer (BU) is a neglected infectious disease caused by Mycobacterium ulcerans and is characterized by necrotic cutaneous lesions. Infection is challenging to treat, and the ideal combination of surgery and antimicrobial therapy continues to evolve. M. ulcerans has been endemic to the Bellarine peninsula in Victoria, Australia, since 1998, with more than 250 cases of infection. Studies have illustrated the safety and efficacy of antimicrobial therapy [1–5], and our standard treatment practice has evolved over the last 15 years to comprise limited surgical debridement with combination antimicrobial therapy. Immune reconstitution inflammatory syndrome (IRIS) is a paradoxical reaction occurring during treatment of an infection, recognized clinically by deterioration after initial improvement. These reactions are well described in tuberculosis and leprosy, where effective antimicrobial killing may be accompanied by (transient) clinical deterioration during treatment [6,7], predominantly in HIV-infected patients after the introduction of antiretroviral therapy [8]. IRIS reactions may also occur in patients with a competent immune system [9]. Our group and others have described paradoxical reactions occurring during the treatment of M. ulcerans infection with antimicrobials [10,11]. In cases of Mycobacterium tuberculosis (TB) infection complicated by IRIS, steroid therapy is recommended. A randomized placebo-controlled trial of IRIS in TB found that prednisone reduced the need for hospitalization and therapeutic procedures and hastened improvements in symptoms, performance, and quality of life [12]. In our group’s description of paradoxical reactions during therapy for BU, we proposed that adjunctive corticosteroid therapy may improve healing and prevent the need for further surgical intervention [10]. In the last 2 years our group has therefore acquired an early experience with the use of steroid therapy to treat severe IRIS in patients with M. ulcerans infection. From 1998 through the end of 2011, our group has treated 163 patients with M. ulcerans infection with antimicrobials. We have assessed both retrospectively (until 2009) and prospectively (since 2009) that 31 patients (19%) developed paradoxical reactions. To date, five patients have been treated with steroid therapy for severe paradoxical reactions. The patients treated with corticosteroids were aged between 9 and 84 years. All patients were ambulatory and managed as outpatients. These patients developed IRIS 2–13 weeks after commencing combination antimicrobial therapy after experiencing an initial clinical improvement in the erythema and induration surrounding the lesion. The clinical findings that were indicative of a severe paradoxical reaction included markedly increased inflammation and induration surrounding the M. ulcerans lesion, copious wound discharge, the appearance of new secondary lesions, and necrotic eschar formation (see Table 1). Clinical deterioration during antibiotic treatment may be interpreted as treatment failure, leading to further potentially disfiguring and unnecessary surgery and either a change or prolongation of antimicrobial therapy [10]. Therefore, we believe that confirmation of a paradoxical reaction via histopathological assessment of a biopsy specimen and mycobacterial cultures are important in the therapeutic decision-making process for patients with clinical deterioration during antimicrobial therapy for M. ulcerans. Histopathological findings of a paradoxical reaction include ulceration and necrosis, a florid mixed inflammatory infiltrate with multinucleated giant cells, usually sparse or absent acid fast bacilli (AFB), and negative mycobacterial cultures (see Figures 1 and 2) [10,13], although excised tissue may remain positive for M. ulcerans via polymerase chain reaction. In all of our cases with a paradoxical reaction, mycobacterial cultures were negative regardless of whether AFBs were seen on microscopy. We recognize that biopsy and histopathology may not be readily available in resourcelimited settings managing patients with BU. In these cases, a diagnosis of paradoxical reaction based on clinical parameters will be most appropriate. In our patients a provisional diagnosis of severe IRIS was followed by commencement of prednisone at a dose of 0.5–1 mg/ kg daily with the aim of reducing further tissue destruction, preserving skin grafts, and limiting the extent of further surgery. In all patients with M. ulcerans treated with steroid therapy, there was a marked clinical improvement in the appearance of the lesion within days to weeks, and the duration of prednisone therapy used was 4–6 weeks, with gradual tapering of the dose after the first 2–3 weeks of therapy. All patients tolerated prednisone well with no side-effects, and antimicrobial therapy was not changed nor prolonged beyond 12 weeks despite the occurrence of a paradoxical reaction. All patients underwent debridement of their M. ulcerans lesions before or after the commencement of the predni-