Late cerebral relapse of a Mycobacterium avium complex disseminated infection in an HIV-infected patient after cessation of antiretroviral therapy.

Abstract

Mycobacterium avium complex (MAC) is a common environmental organism that may cause disseminated disease in immunocompromised patients infected with HIV. Treatment is usually life long. We report here a case of late cerebral relapse that occurred in an HIV-infected patient who had been treated 5 years earlier for a disseminated MAC infection and who stopped highly active antiretroviral therapy (HAART) and secondary prophylaxis. A 34-year-old woman, who had been diagnosed with HIV infection in 1992, was admitted in March 1996 for fever, night sweats and anaemia. Her CD4 cell count was 9 cells/mm3. MAC infection was diagnosed on blood culture. The isolate was susceptible to clarithromycin and she was treated with rifabutin, ethambutol and clarithromycin. Control mycobacterial blood cultures became negative. Rifabutin was stopped after 2 months as a result of side-effects (arthralgia) but ethambutol and clarythromycin were continued for 3 years. Five months after beginning mycobacterial treatment, a combination of lamivudine, stavudine and indinavir was started. Her CD4 cell level increased to 551 cells/mm3 and the plasma HIV-RNA level became less than 50 copies/ml in 2 months. In January 2000 she decided to stop HAART (her CD4 cell count was 528 cells/mm3) and she was lost of follow-up. In August 2001 she was admitted for a 15-day history of severe headaches, dizziness, nausea and vomiting. Physical examination was normal except for weight loss. Laboratory investigations revealed a CD4 cell count of 73 cells/mm3, a plasma HIV-RNA level of 473 790 copies/ml. A computed tomography (CT) scan of the brain showed two ring-enhancing masses in the left frontal lobe with significant perilesional oedema and associated mass effect. She was first treated for a presumed abscess caused by Toxoplasma gondii with sulfadiazine and pyrimethamine. A CT scan performed 4 weeks later did not show any improvement of the lesion. A stereotactic CT-guided fine-needle aspiration of the left frontal lobe mass was performed. Histological examination of brain tissue revealed a non-granulomatous inflammatory reaction and acid-fast bacilli were seen by Ziehl–Neelsen staining. Mycobacterium avium was isolated from the brain biopsy specimen. The isolate was still susceptible to clarithromycin (MIC < 8 mg/l) and the patient was treated successfully with rifabutin, ethambutol and clarithromycin. HAART was resumed. To prove a cerebral relapse of MAC infection definitively, a polymerase chain reaction typing of M. avium strains isolated in 1996 and 2001 was performed as previously described [1] (see Fig. 1).Fig. 1. Similar genotyping profile of: Mycobacterium avium strains isolated in our patient in 1996 and 2001 (lanes 3 and 4, respectively). Lanes 1, 2 and 5 strains from unrelated patients. Lane 6, DNA molecular weight marker (100 bp ladders). Methods are described in Picardeau and Vincent [1]. Briefly, primers to the ends of insertion sequences 1245 and 1311 were used. DNA purified from several colonies of mycobacteria taken from Lo"westein-Jensen slants was amplified. Polymerase chain reaction products were analysed by electrophoresis on a 2% agarose gel and detected by ethidium bromide staining.As shown in Fig. 1, the two strains of M. avium isolated exhibited a similar molecular profile, demonstrating that the patient experienced a late relapse while discontinuing HAART, which resulted in severe immunodeficiency. Although MAC infections are usually disseminated in HIV-infected patients, a diagnosis of cerebral infection is very uncommon. To our knowledge, cerebral relapse of MAC infection has not been reported. Murray et al. [2] reported a similar case of a cerebral MAC infection that occurred in an HIV-infected patient 6 years after a disseminated infection. However, in that report, susceptibility to clarithromycin of the initial and the relapse strains showed different profiles, and genotyping analysis was not performed, suggesting in this case reinfection rather than a relapse. The safety of the cessation of primary or secondary prophylaxis against opportunistic infections in patients with a favourable immunological response to HAART has been demonstrated in several randomized trials and cohort studies [3,4]. The safety of this strategy in patients treated for disseminated MAC infections is less clear. Retrospective studies have shown that this strategy might be successful in patients with a favourable immunological and virological response to to HAART [5–9]. However, late reccurrences have been described in patients with low CD4 cell counts when being treated with HAART or after stopping HAART [8,9]. The patient described here experienced a late relapse of MAC disease despite 36 months of effective MAC therapy, when CD4 cell counts dropped (from 528 to 73/mm3) after the withdrawal of HAART. At a time when a general policy towards the discontinuation of secondary prophylaxis against opportunisitic infections is proposed, this observation, combined with previous studies, reinforces the recommandation that prophylaxis should be resumed in patients in whom CD4 cell counts decrease below 100/mm3.