Polymyxin B plasma AUC/MIC predicts IV efficacy but not with adjunctive inhalation for MDR Gram-negative infections.

Treating multidrug-resistant (MDR) Gram-negative bacterial infections, particularly in critically ill patients, remains a significant challenge due to limited therapeutic options and high mortality rates. Polymyxin B, a last-line polypeptide antibiotic, is crucial for these infections, but its efficacy, especially in pulmonary infections, is often limited by poor lung penetration and potential nephrotoxicity. While intravenous (IV) administration is standard, adjunctive inhaled (IH) polymyxin B is increasingly used to achieve higher local drug concentrations in the lungs. However, the impact of this combined approach on systemic pharmacokinetics (PK) and the predictive value of plasma-based PK/pharmacodynamic (PD) targets, such as AUCss,24h/MIC, for clinical outcomes remains poorly understood, creating a critical gap in guiding optimal dosing strategies.

This prospective observational cohort study enrolled 43 critically ill adults diagnosed with multidrug-resistant Gram-negative bacterial infections who received Polymyxin B for ≥48h. Patients were assigned to either an intravenous (IV, n=22) or a combined intravenous plus inhaled (IV+IH, n=21) administration group. Plasma samples were collected and analyzed using validated HPLC-MS/MS to determine drug concentrations. Population pharmacokinetic (PPK) models were developed using NONMEM software to characterize Polymyxin B disposition. Clinical efficacy at the end of treatment was independently and blindly assessed as the primary endpoint, allowing for correlation with plasma AUCss,24h/MIC ratios.

The study enrolled 43 patients (IV, n=22; IV+IH, n=21), achieving an overall clinical success rate of 66.7%. A two-compartment PPK model best described Polymyxin B pharmacokinetics, with a typical clearance of 2.61 L/h. Creatinine clearance and total bile acids were identified as significant covariates influencing drug disposition. In the overall cohort, no significant differences in AUCss,24h or AUCss,24h/MIC were observed between patients with clinical success and those with failure (P=0.591 and P=0.143, respectively). However, a distinct pattern emerged when groups were analyzed separately:

In the IV monotherapy group, AUCss,24h/MIC was significantly higher in responders (P=0.005), and an ROC analysis suggested an exploratory efficacy threshold of 94.4. Conversely, no exposure-response relationship was observed in the IV+IH group for either AUCss,24h (P=0.398) or AUCss,24h/MIC (P=0.495), indicating that plasma-based PK/PD targets may not reliably predict efficacy when adjunctive inhalation is used, possibly due to high local pulmonary exposure.