A mechanism-based pharmacokinetic/pharmacodynamic analysis of polymyxin B-based combination therapy against carbapenem-resistant <i>Klebsiella pneumoniae</i> isolates with diverse phenotypic and genotypic resistance mechanisms

Increased resistance to β-lactams/β-lactamase inhibitors by mutations in β-lactamase genes, porins, and efflux pumps complicates the management of carbapenem-resistant <i>Klebsiella pneumoniae</i> (CRKP). Polymyxin B (PMB)-based combination therapy is the best alternative treatment for middle and low-income countries that cannot access the latest medicines. It is crucial to know both phenotypic and genotypic characteristics of a pathogen to understand the killing effect of each drug and its combinations. Hence, our objective was to incorporate mechanistic insights gained from resistance mechanisms of each isolate to develop a mechanism-based pharmacokinetic/pharmacodynamic model. Six clinical CRKP isolates with diverse genotypic resistance expressing <i>bla</i>_KPC, <i>bla</i>_NDM, porin, and mgrB mutations were used for static concentration time kill (SCTK) assays to evaluate the rate and extent of killing by monotherapy, double and triple combinations using PMB (0.5-64 mg/L), meropenem (10-120 mg/L), and fosfomycin (75-500 mg/L). Isolate BRKP28 expressed non-functional MgrB (a regulatory protein) and high-level phenotypic resistance (PMB MIC: >128 mg/L). In line with the observed resistance, the model estimated that BRKP28 had a reduced maximum killing rate constant for PMB (3.61 h⁻¹) relative to other isolates. The mechanistic synergy of PMB, due to outer membrane disruption, was incorporated into three isolates with porin mutations. PMB demonstrated 83%-88% mechanistic synergy with meropenem and 81%-98% with fosfomycin. The model further estimated that a very low concentration of PMB (0.49-0.64 mg/L) was sufficient to achieve 50% of the maximum synergy. Simulations using population pharmacokinetic models showed that combination therapy of PMB (1 mg/kg q12h) and fosfomycin (8 g q8h) achieved >73% reduction in area under the bacterial load-versus-time curve across four isolates. The triple combination therapy achieved a 67.7% reduction in non-carbapenamase producing isolate. These findings demonstrates that a low PMB dosing regimen (1 mg/kg q12h) can produce synergistic effects in combination therapy and may be effective in managing infections caused by CRKP, including PMB resistant isolates.