利奈唑啉
药代动力学
医学
中枢神经系统
重症监护医学
药理学
生物
金黄色葡萄球菌
内科学
万古霉素
细菌
遗传学
作者
Lvchang Zhu,Xin-Xin I. Zeng,Yi Shi,Yuhang Wu,Xiaoshan Zhang,Shanshan Xu,Xuben Yu,Lisu Huang
出处
期刊:PubMed
日期:2025-04-03
标识
DOI:10.1093/infdis/jiaf169
摘要
Linezolid shows therapeutic potential for pediatric gram-positive bacterial central nervous system infections (CNSIs). However, its efficacy, safety profile, and cerebrospinal fluid (CSF) pharmacokinetics require detailed evaluation. This prospective two-center observational study enrolled children with confirmed or suspected gram-positive CNSIs. Clinical outcomes and adverse events were compared between linezolid-treated patients and a matched vancomycin cohort. Population pharmacokinetic (PopPK) modeling with nonlinear mixed-effects analysis quantified linezolid exposure in plasma and CSF. Among 45 matched pediatric CNSIs patients per group, linezolid demonstrated a 91.1% clinical response rate and 68.9% cure rate (vancomycin cure rate: 68.9%). However, non-inferiority to vancomycin was not established for the primary endpoint, possibly influenced by intergroup baseline variability and extended treatment duration. Adverse events occurred more frequently with linezolid, including gastrointestinal (48.9% vs. 24.4%, p = 0.02) and hematologic effects (73.3% vs. 53.3%, p = 0.05). Plasma trough concentrations > 7 µg/mL were correlated with elevated risk of leukopenia and neutropenia (odds ratio [OR] 9.38, 95% confidence interval [CI] 1.21-72.6; and OR 40.2, 95% CI 2.15-748.50). However, no treatment discontinuations occurred due to adverse events. The PopPK model analyzed 135 linezolid concentrations (90 plasma/45 CSF), identifying body weight as the primary covariate influencing distribution. Plasma and CSF trough concentrations showed a strong correlation (r = 0.87, 95% CI 0.75-0.98). Linezolid demonstrated favorable clinical efficacy and tolerability in pediatric CNSIs, with CSF concentrations that correlated with plasma levels and exhibited predictable pharmacokinetics.
科研通智能强力驱动
Strongly Powered by AbleSci AI