Monte Carlo simulation for dosage optimization of the best available therapy for bloodstream infections secondary to carbapenemase-producing Klebsiella pneumoniae in critically ill patients.

Objectives: We aimed to use Monte Carlo simulation, based on pharmacokinetic/ pharmacodynamic (PK/PD) targets, to investigate and determine the optimal dosage of the available combination therapies for carbapenem-resistant Klebsiella pneumoniae (CRKP) in critically ill patients.

Methods: We collected CRKP clinical isolates from Phramongkutklao Hospital between October 2020 and June 2022. A molecular study of resistant genes was performed using polymerase chain reaction (PCR). Broth microdilution checkerboards were used to evaluate the mono- and synergistic antibiotic activities. Monte Carlo simulation was used to determine the optimal antibiotic regimens, based on the probability of target achievement (PTA) and cumulative fraction of response (CFR).

Results: The 54 CRKP isolates were resistant to tigecycline (100%), colistin (75.9%), amikacin (70.4%), and gentamicin (63.0%). The most common carbapenemase genotype was bla_{oxacillinases (OXA)-48-like} (42.6%), followed by bla_{New Delhi metallo beta lactamase (NDM)} (29.6%) and coexistence of bla_OXA-48-like and bla_NDM (22.2%). Based on the PTA, synergistic and additive activities against CRKP isolates were observed with appropriate dosages of tigecycline-colistin (67.9%), tigecycline-gentamicin (62.2%), and tigecycline-amikacin (51.4%).

Conclusions: Tigecycline-colistin was the best available combination therapy for critically ill patients with CRKP, especially NDM. When used in combination with tigecycline, a colistin creatinine clearance (CrCl) of <90 mL/min can raise the CFR target and less nephrotoxicity.