Optimizing polymyxin B exposure in carbapenem-resistant organism Hospital-Acquired Pneumonia: a retrospective observational study

Background The area under the concentration-time curve over 24 hours (AUC _0-24h ) is a critical pharmacokinetic parameter influencing the clinical efficacy of polymyxin B (PMB). However, due to substantial population heterogeneity among critically ill patients, the correlation between the AUC _0-24h of PMB and clinical treatment success, as well as the optimal therapeutic threshold remains inadequately elucidated. Objectives This study aimed to investigate the relationship between PMB AUC _0-24h and clinical efficacy in patients with carbapenem-resistant hospital-acquired pneumonia (HAP), and to identify the optimal therapeutic AUC _0-24h range for PMB. Methods We conducted a retrospective observational study of carbapenem-resistant HAP patients receiving intravenous PMB with therapeutic drug monitoring (TDM). The plasma concentrations of PMB were determined using validated ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC–MS/MS). The primary outcome was 14-day clinical treatment success rates, the secondary outcomes included the incidence of AKI. Logistic regression analyses and restricted cubic spline analyses were employed to investigate the optimal therapeutic threshold of PMB in carbapenem-resistant HAP patients. Result A total of 138 patients were ultimately included in the final analysis, with the median age of 68 years, and 110(79.7%) patients were male. Clinical success was achieved in 63(45.7%) patients. The pathogen of infection in 77% of patients was Carbapenem-Resistant Acinetobacter baumannii (CRAB), and all pathogens isolated from the included patients were sensitive to PMB, with a minimum inhibitory concentration (MIC) ≤ 2mg/L. The median AUC _{0 − 24h} of PMB was 77.5 (55.6, 105.6) mg·h/L. In enrolled patients, an AUC _{0 − 24h} >77 mg·h/L enabled to predict the clinical treatment success, while AUC _{0 − 24h} >110 mg·h/L enabled to predict the AKI incidence. In the multivariate logistic regression model, AUC _{0 − 24h} < 77 mg·h/L was an independent risk factor for reduced clinical treatment success (OR: 0.33, 95% CI: 0.13–0.78; p = 0.013). AUC > 110 mg·h/L significantly increased AKI risk (OR: 0.33, 95% CI: 1.06–10.47; p = 0.043). Conclusion This study confirms that the AUC _{0 − 24h} of PMB is closely associated with both 14-day clinical treatment success rates and the incidence of AKI in patients with carbapenem-resistant HAP, with an optimal range from 77 to 110 mg·h/L.