Selection of rifampicin-resistant Staphylococcus aureus during tuberculosis therapy: concurrent bacterial eradication and acquisition of resistance.

BACKGROUND AND OBJECTIVES Acquired antimicrobial resistance is commonly attributed to regimens that expose bacteria to subinhibitory concentrations; consequently, eradication of susceptible cells is advocated. The mutant selection window hypothesis predicts that resistance can be acquired even when inhibitory concentrations are exceeded and susceptible bacteria are eradicated. The objective was to test that prediction clinically. METHODS Tuberculosis patients (n = 372) were sampled for nasal colonization by Staphylococcus aureus at the beginning of anti-tuberculosis therapy with rifampicin-containing regimens and again after 2 and 4 weeks. Rifampicin susceptibility of S. aureus was determined, and S. aureus isolates from patients developing acquired resistance were examined by molecular strain typing. Diabetes patients (n = 200) served as untreated controls. RESULTS Nasal colonization was 17% and 20% for the tuberculosis and diabetes patients, respectively. Four patients were initially colonized with rifampicin-resistant S. aureus and were excluded from further sampling. Initiation of anti-tuberculosis therapy eradicated S. aureus nasal colonization in 53/58 tuberculosis patients while allowing acquisition of rifampicin resistance in 5/58. Pulsed-field gel electrophoresis (PFGE) band patterns and protein A repeat sequence determination differed in S. aureus isolated from different patients but was identical in isolates obtained from the same patient before and after acquisition of resistance. No resistance was acquired in untreated control patients, which differed statistically from treated patients (P = 0.025). CONCLUSIONS Acquired resistance and eradication of susceptible bacteria can occur concurrently; restricting acquired resistance may require direct suppression of mutant growth and viability in addition to elimination of susceptible bacteria.