The epidemiological fitness cost of drug resistance in Mycobacterium tuberculosis

• DOI: 10.1073/pnas.0902437106
• Publication Date: 2009-08-14T02:14:08Z
• Authors: Fabio Luciani; Scott A. Sisson; Honglin Jiang; Andrew R. Francis; Mark M. Tanaka
• Keywords: Approximate Bayesian Computation

The emergence of antibiotic resistance in Mycobacterium tuberculosis has raised the concern that pathogen strains are virtually untreatable may become widespread. acquisition to antibiotics results a longer duration infection host, but this come at cost through decreased transmission rate. This raises question whether overall fitness drug-resistant is higher than sensitive strains—essential information for predicting spread disease. Here, we directly estimate drug resistance, rate which evolves, and relative resistant strains. These estimates made by using explicit models evolution M. , approximate Bayesian computation, molecular epidemiology data from Cuba, Estonia, Venezuela. We find sensitivity can be as low 10%, evolves rates ≈0.0025–0.02 per case year, comparable with Furthermore, contribution very high compared acquired due treatment failure (up 99%). Estimating such parameters vivo will critical understanding responding resistance. For instance, projections our suggest prevalence decline successful treatment, proportion cases associated likely increase.