Scholarlink

Introduction <p>Multidrug-resistant Gram-negative bacteria (MDR-GNB), especially carbapenem-resistant strains, pose a major therapeutic challenge in intensive care units and are associated with high morbidity and mortality. Polymyxin B (PMB) and colistin sulfate (CS) are the last-line agents for MDR-GNB infections; however, their clinical use is limited by nephrotoxicity. Although the steady-state 24-h area under the curve (AUC<sub>ss,24h</sub>) has been suggested as a predictor of nephrotoxicity, prior studies have mainly applied semiparametric approaches that cannot fully describe the risk across exposure levels. Parametric time-to-event (TTE) analysis offers a more robust framework but has not been applied to polymyxin-induced nephrotoxicity.</p> Aim <p>This study aimed to identify clinical and pharmacological factors influencing PMB- and CS-associated nephrotoxicity in critically ill patients with MDR-GNB infections and to establish AUC<sub>ss,24h</sub> thresholds predictive of acute kidney injury (AKI) using parametric TTE modeling.</p> Method <p>We retrospectively analyzed real-world data from 562 patients with MDR-GNB infections treated with PMB (n = 354) or CS (n = 208) at Xiangya Third Hospital, Central South University, between July 2018 and July 2023. Pharmacokinetic profiles were simulated using published models, and drug exposure parameters (AUC<sub>ss,24h</sub>, C<sub>ss,max</sub>, and C<sub>ss,min</sub>) were estimated. Propensity score matching was used to balance the baseline covariates. Kaplan–Meier curves and log-rank tests were used to compare the AKI incidence between the groups. Parametric TTE models were developed using NONMEM (version 7.5), incorporating exposure parameters and covariates. The model performance was validated using bootstrap and visual predictive checks. Classification and regression tree (CART) analyses were used to determine the exposure thresholds.</p> Results <p>Overall, 39.4% of patients developed AKI, with a significantly higher incidence in the PMB group than in the CS group (51.7% vs. 18.4%). The final PMB model identified AUC<sub>ss,24h</sub>, sepsis, transplant history, and vancomycin co-administration as independent risk factors, with an EC50 of 80.4&#xa0;μg·h/mL for PMB. For CS, AUC<sub>ss,24h</sub> and multisite infections predicted AKI with an EC50 of 57.5&#xa0;μg·h/mL. CART analysis revealed nephrotoxicity thresholds of 101&#xa0;μg·h/mL for PMB and 44&#xa0;μg·h/mL for CS administration. Simulation showed that increasing PMB AUC<sub>ss,24h</sub> from 50 to 125&#xa0;μg·h/mL raised 14-day AKI risk from 25 to 75%, while for CS, increasing AUC<sub>ss,24h</sub> from 25 to 50&#xa0;μg·h/mL elevated risk from 20 to 60%.</p> Conclusion <p>In critically ill patients with MDR-GNB infections, higher plasma exposure to PMB and CS was strongly associated with increased nephrotoxicity. Exposure thresholds of AUC<sub>ss,24h</sub> ≥ 101&#xa0;μg·h/mL for PMB and ≥ 44&#xa0;μg·h/mL for CS significantly elevated the AKI risk. Therapeutic drug monitoring should be integrated into clinical practice to optimize polymyxin dosing, reduce toxicity, and improve patient outcomes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Nephrotoxicity of polymyxin B and colistin sulfate in patients with multidrug-resistant gram-negative bacteria infections: a parametric time-to-event analysis

  • 期刊论文

  • 发表日期：

    2025年11月7日

查看全文
  • Yuanfang Qin,
  • Qin Ding,
  • Shuqi Huang,
  • Ruwei Yang,
  • Shengnan Zhang,
  • Tingting Wu,
  • Jingjing Liu,
  • Qi Pei

摘要

Introduction

Multidrug-resistant Gram-negative bacteria (MDR-GNB), especially carbapenem-resistant strains, pose a major therapeutic challenge in intensive care units and are associated with high morbidity and mortality. Polymyxin B (PMB) and colistin sulfate (CS) are the last-line agents for MDR-GNB infections; however, their clinical use is limited by nephrotoxicity. Although the steady-state 24-h area under the curve (AUCss,24h) has been suggested as a predictor of nephrotoxicity, prior studies have mainly applied semiparametric approaches that cannot fully describe the risk across exposure levels. Parametric time-to-event (TTE) analysis offers a more robust framework but has not been applied to polymyxin-induced nephrotoxicity.

Aim

This study aimed to identify clinical and pharmacological factors influencing PMB- and CS-associated nephrotoxicity in critically ill patients with MDR-GNB infections and to establish AUCss,24h thresholds predictive of acute kidney injury (AKI) using parametric TTE modeling.

Method

We retrospectively analyzed real-world data from 562 patients with MDR-GNB infections treated with PMB (n = 354) or CS (n = 208) at Xiangya Third Hospital, Central South University, between July 2018 and July 2023. Pharmacokinetic profiles were simulated using published models, and drug exposure parameters (AUCss,24h, Css,max, and Css,min) were estimated. Propensity score matching was used to balance the baseline covariates. Kaplan–Meier curves and log-rank tests were used to compare the AKI incidence between the groups. Parametric TTE models were developed using NONMEM (version 7.5), incorporating exposure parameters and covariates. The model performance was validated using bootstrap and visual predictive checks. Classification and regression tree (CART) analyses were used to determine the exposure thresholds.

Results

Overall, 39.4% of patients developed AKI, with a significantly higher incidence in the PMB group than in the CS group (51.7% vs. 18.4%). The final PMB model identified AUCss,24h, sepsis, transplant history, and vancomycin co-administration as independent risk factors, with an EC50 of 80.4 μg·h/mL for PMB. For CS, AUCss,24h and multisite infections predicted AKI with an EC50 of 57.5 μg·h/mL. CART analysis revealed nephrotoxicity thresholds of 101 μg·h/mL for PMB and 44 μg·h/mL for CS administration. Simulation showed that increasing PMB AUCss,24h from 50 to 125 μg·h/mL raised 14-day AKI risk from 25 to 75%, while for CS, increasing AUCss,24h from 25 to 50 μg·h/mL elevated risk from 20 to 60%.

Conclusion

In critically ill patients with MDR-GNB infections, higher plasma exposure to PMB and CS was strongly associated with increased nephrotoxicity. Exposure thresholds of AUCss,24h ≥ 101 μg·h/mL for PMB and ≥ 44 μg·h/mL for CS significantly elevated the AKI risk. Therapeutic drug monitoring should be integrated into clinical practice to optimize polymyxin dosing, reduce toxicity, and improve patient outcomes.