Antibiotic Tolerance and Combination Therapy

The recent paper by Haaber et al. (1) highlights another mechanism by which bacterial pathogens may evade the effects of antibiotics. The combination of colistin with vancomycin against Staphylococcus aureus may be antagonistic and may lessen the effectiveness of glycopeptide antibiotic. Colistin, it seems, induces gene expression in S. aureus which mimics the vancomycinintermediate (VISA) phenotype. This induction is reversible when the colistin is removed. This article is very timely. Therapeutic options for many infections, particularly those caused by Gram-negative bacteria, are diminishing. Many researchers and clinicians are looking for alternative treatment options and combination antibiotic therapy for recalcitrant multidrug-resistant (MDR) infections. Synergies between agents that might otherwise seem counterintuitive have been described. For example, combination therapy with colistin and vancomycin has been shown to be effective in vitro and in in vivo models against MDR Acinetobacter baumannii (2, 3). Driven by desperation for new therapeutic options for these highly resistant organisms, a number of researchers have shown interest in the combination of colistin with glycopeptide agents (see Claeys et al. [4] for a review). While colistin and vancomycin are an unlikely combination for treating S. aureus infection, the findings of Haaber et al. suggest that caution should be exercised if combination therapy is to be employed clinically. Simultaneous colonization of patients with S. aureus and A. baumannii has been documented (5). Moreover, it is not uncommon for infections to have a polymicrobial etiology. For example, Sancho et al. determined that 20.2% of bacteremia cases in critically ill patients are polymicrobial in nature, with A. baumannii and S. aureus being one of the more common combinations (6). Treating such a polymicrobial infection with the colistin-vancomycin combination might be effective against the Gram-negative pathogen, but the colistininduced vancomycin-intermediate phenotype might exacerbate the Gram-positive infection. Ultimately, the challenge to maximize both the effectiveness and the longevity of our current antibiotic inventory might require innovative approaches to how we use them. This serves as a useful reminder that such applications have to be considered holistically—new treatment regimens may have unforeseen implications. Clinicians should exercise caution testing colistin-vancomycin combination therapy when there is a high likelihood of polymicrobial infection.