A systematic review of the clinical significance of pulmonary Mycobacterium kansasii isolates in HIV infection.

Guidelines recommend treating HIV-infected patients with pulmonary Mycobacterium kansasii infection only in the presence of multiple positive cultures and clinical and radiographic abnormalities. Some authors suggest a single positive culture warrants treatment. A systematic literature review was done to determine whether HIV-infected patients who had M. kansasii isolated from respiratory specimens may have an indolent infection (often referred to as colonization) not requiring treatment and to determine the utility of diagnostic criteria in distinguishing disease from indolent infection. Sixteen studies were included, with at least 573 patients: mean age 44 years; 91% male; 64% men who had sex with men; 35% injection drug users; and median CD4 lymphocyte count of 2-381 cells/microL. The median rate of indolent infection was 8%. Of the few patients who did not satisfy diagnostic criteria and were left untreated, outcomes were generally favorable. Overall, survival was longer in treated patients (mean 12 vs. 4 months). Indolent pulmonary infection with M. kansasii may exist in the setting of HIV, but published data do not provide adequate information to identify such patients. It is unclear whether unfulfilled diagnostic criteria necessarily imply the absence of disease in this context.