Mycobacterium avium Brain Abscess at the Initiation of Highly Active Antiretroviral Therapy

Reported here is a case of brain abscess due to Mycobacterium avium-complex (MAC), diagnosed on the basis of pathology and microbiological investigations, which occurred in a HIV-1-infected patient with a previous history of disseminated MAC infection. This is only the third case of MAC brain abscess to be reported to date; in the two previous cases, one patient was not immunocompromised [1] and the other patient had been receiving highly active antiretroviral therapy (HAART) [2]. In our patient, HAART had just been initiated and immune function parameters had not been restored at the time the MAC brain abscess was diagnosed. This contrasts with previous reports of patients receiving HAART who developed other atypical MAC manifestations. Clinicians should thus be aware that brain abscesses due to MAC can occur in HIV-infected patients who do not have sustained elevation of their CD4+ cell count. A 40-year-old man was hospitalized in an intensive care unit for treatment of status epilepticus. HIV-1 infection had been diagnosed in 1989. In 1996 he had cytomegalovirus retinitis, and in March 1997 disseminated Mycobacterium avium infection was diagnosed. Clarithromycin (2,000 mg/d), ethambutol (20 mg/kg/d) and rifabutin (300 mg/d) were discontinued in April 1999. In August 2001 HAART was initiated with lamivudine (300 mg/d), stavudine (80 mg/d), tenofovir (300 mg/d), efavirenz (600 mg/d) and pentafuside (180 mg/d). Within 2 months plasma HIV-RNA levels decreased from 152,000 to 3,570 copies/ml, but the CD4+ cell count remained at 29/mm3. In October 2001, relapsing CMV retinitis was treated with foscavir (12 g/d); 1 month later, the patient presented in epileptic crisis associated with a Glasgow score of 7. Blood investigation revealed a leukocyte count of 7.77 109/l, with 31 CD4+ cells/mm3 and a plasma RNA level of 40,000 copies/ml. Cerebrospinal fluid examination showed 7 leukocytes, 190 erythrocytes/mm3, and no Cryptococcus sp. (India ink staining). The cerebrospinal fluid remained sterile, and a CMV pp65-antigenemia test (CINA Kit, HCMV PP65 Rapid Antigenemia Complete Kit; Argene, Varilhes, France) was negative, but CMV DNA was detected (Herpes Consensus Generic; Argene, Varilhes, France). Computerized tomography of the brain showed a solitary, 2-cm lesion in the left occipital lobe with perilesional edema. Treatment with clindamycin (4 g/d), cotrimoxazole (5.8 g/d) and 5FU (100 mg/d) failed and the cerebral lesion was excised. A brain biopsy was tested for Toxoplasma gondii using PCR, and the result was negative. Hematoxylin-eosin-saffron-stained sections of formalin-fixed, paraffin-embedded tissue showed tumorlike features largely composed of spindle-shaped cells arranged in a fascicle-like pattern (Fig. 1A). ZiehlNeelsen staining revealed widespread intra-cytoplasmic acid-fast bacilli (Fig. 1B). The tissue was strongly and diffusely immunoreactive for CD68 (KL-1; Dako, Ely, UK). Mycobacterium avium was isolated in pure culture from the biopsy and identified using DNA Accuprobe (GenProbe, San Diego, USA). Treatment was changed to clarithromycin (500 mg/d), ethambutol (1,200 mg/d) and ciprofloxacin (1,000 mg/d) for 6 months followed by azithromycin (600 mg/d) and ethambutol (1,200 mg/d). One year after the introduction of HAART, the patient was asymptomatic and computerized tomography of the brain showed complete resolution of the lesion. P. Berger · M. Drancourt ()) Unit des Rickettsies CNRS UMR 6020, IFR 48, Facult de M decine, 27, Boulevard Jean Moulin, 13385 Marseille C dex 05, France e-mail: Michel.Drancourt@medecine.univ-mrs.fr Tel.: +33-4-91385517 Fax: +33-4-91830390