Drug mode of action and resource constraints modulate antimicrobial resistance evolution.

An increasingly important goal in the design of antimicrobial treatment regimens is to minimise the probability of resistance evolving, without harming individual patients' outcomes. A key characteristic to consider when choosing an antibiotic for treatment is its mode of action: bacteriostatic (growth-inhibiting) or bactericidal (mortality-inducing). We present a theoretical model comparing the efficacy of bacteriostatic, bactericidal, and intermediate drugs at preventing the evolutionary rescue of an initially susceptible bacterial population. We find that, all else equal, in resource-abundant environments, bacteriostatic drugs are best, as they constrain cell divisions and thus allow fewer resistance mutations to occur. This contrasts with the prevailing assumption that bactericidal drugs are best as they actively kill cells. When multiple drugs are employed, using one bacteriostatic and one bactericidal drug is usually optimal, because the cell division rate cannot fall below zero, so there are diminishing returns to bacteriostatic activity from two drugs. Severe resource constraints mean that growth rates are already low, and thus there is less benefit to bacteriostatic drugs further limiting growth, so bactericidal drugs are favoured. If these findings are empirically verified in the laboratory and in vivo, they could significantly guide clinical practice.