One hundred ninety-six Mycobacterium avium isolates from blood samples recovered from 93 AIDS patients for several months were typed by serotyping, by IS1245 restriction fragment length polymorphism (RFLP) analysis and in some cases RFLP analysis with plasmids pVT2 and pLR7 as probes, and by pulsed-field gel electrophoresis (PFGE). PCR typing of single colonies was also used to detect polyclonal infections. Strains belonged mainly to serotypes 1, 4, and 8. pVT2-and pLR7-related plasmids were detected in strains from 49% of the patients. The IS1245 RFLP and PFGE analyses showed a 96.8% diversity of the M. avium strains from the 93 patients. The vast majority (95.2%) of infections were monoclonal, indicating that recent infection is unlikely, even at an advanced stage of AIDS. For one patient, sequential isolates gave divergent patterns of sensitivity and resistance to clarithromycin, but all were identified as the initial clone. RFLP analysis and PCR typing of single colonies allowed for the detection of three polyclonal infections during the bacteriological follow-up. Among strains from patients whose samples were positive by culture after treatment for 2 to 15 months, 97.4% were the same as the initial strain. In conclusion, relapses and failures were mostly due to the initial strain. These relapses and failures resulted either from the selection of resistant mutants or the reappearance of sensitive strains, suggesting the persistence of nonsterilized tissue reservoirs.

We conducted a randomized, open-label trial in 42 French hospitals to compare the clinical and bacteriologic efficacy of combination therapy with clarithromycin/clofazimine (Clm/Clof) with that of combination therapy with clarithromycin/rifabutin/ethambutol (Clm/Rib/Eth) as treatment for Mycobacterium avium bacteremia. One hundred forty-four human immunodeficiency virus-seropositive patients older than 18 years of age who had CD4 lymphocyte counts of <100/mm3 and a blood culture positive for M. avium were enrolled in the study. The main measures of outcome were blood cultures, abatement of clinical symptoms (fever), and survival. Treatment success (defined as patient living, either no fever or a reduction of > or = 1 degrees C in initial body temperature, and a blood culture negative for M. avium) was similar in both treatment groups at months 2 and 6. However, following initial resolution of infection, relapse of M. avium bacteremia occurred in more patients in the Clm/Clof group than in the Clm/Rib/Eth group (22 vs. six, respectively; P < .001); these relapses were accompanied by emergence of strains resistant to clarithromycin in 21 and two patients, respectively. In conclusion, combination therapy with Clm/Rib/Eth prevented relapse of mycobacterial disease and, compared with combination therapy with Clm/Clof, was associated with a significant decrease in the emergence of resistant M. avium strains in HIV-infected patients treated for at least 28 weeks.

No abstract

No abstract