Microbiologic Outcome of Interventions Against Mycobacterium avium Complex Pulmonary Disease: A Systematic Review

Abstract

Pulmonary disease (PD) caused by Mycobacterium avium complex (MAC) is increasing worldwide. We conducted a systematic review of studies that include microbiologic outcomes to evaluate current macrolide-based treatment regimens. We searched literature published before April 2017 by using the MEDLINE, Cochrane, and Embase databases. Risk of bias in randomized trials was assessed using the Cochrane tool. We retrieved 333 citations and evaluated 42 studies including 2,748 patients: 18 studies were retrospective chart reviews, 18 were prospective, and six were randomized. The weighted average proportion of sputum culture conversions in macrolide-containing regimens after subtracting posttreatment microbiologic recurrences was 52.3%(95%CI, 44.7%-59.9%). Using the triple-drug regimens recommended by the American Thoracic Society (ATS) achieved treatment success in 61.4%(95%CI, 49.7%-72.5%), which further increased to 65.7%(95%CI, 53.3%-77.4%) when drugs were taken for at least 1 year by patients who were macrolide susceptible and had previously untreated MAC. The overall risk of bias was low in five of the six randomized trials. However, selective outcome reporting because of a posteriori exclusion of initially included patients (14.0%), uncompleted treatment (17.6%), and inconsistent use of outcome parameters (17 definitions of treatment success) hampered the comparison of nonrandomized trials. To date, randomized studies on treatment outcome in patients with MAC PD are scarce. Long-term treatments with ATS-recommended regimens for patients who are macrolide susceptible are superior to other macrolide-based therapies. A standardized definition of treatment success and genotypic distinction between reinfection and relapse by means of pretreatment and posttreatment identification of MAC species in cases of microbiologic recurrences may help to optimize evaluation of treatment regimens in the future.