目的·从蛋白结合率的角度，评估接受间歇性血液透析（intermittent hemodialysis，IHD）的重症患者中抗生素的清除率以及药代动力学/药效动力学（pharmacokinetics/pharmacodynamics，PK/PD），通过探讨药物蛋白结合率与透析清除的一般规律，为临床制定抗生素在血液透析中的给药方案提供理论依据。方法·纳入于上海交通大学医学院附属第九人民医院肾脏内科就诊的19名接受低通量血液透析并接受抗生素治疗的患者，按抗生素种类分为美罗培南组（n=7）、万古霉素组（n=5）、头孢曲松组（n=7）。采用液相色谱-串联质谱（liquid chromatography with tandem mass spectrometry，LC-MS/MS）建立人血浆/血清和透析液中美罗培南、万古霉素、头孢曲松抗生素的检测方法。使用MATLAB软件建立二室药代动力学模型。计算抗生素瞬时清除率和透析总清除率，并分析PK/PD参数。结果·受试者的临床特征在组间差异无统计学意义。各组透析清除率：美罗培南组（5.14~5.97 L/h）>万古霉素组（2.87~3.77 L/h）>头孢曲松组（1.21~1.90 L/h），P<0.001。3种抗生素在含透析液废液室的二室药动学模型中拟合良好（fval%<2），PK/PD参数与既往文献相近，美罗培南最低抑菌浓度（minimal inhibitory concentration，MIC）为2、8、16 μg/mL的血药浓度高于MIC占给药间隔的比率（fraction of time that the free drug concentration is above the MIC，%fT_>MIC）值分别为95.2%、60.8%和32.4%，头孢曲松MIC为0.25、4、16 μg/mL（游离浓度）的%fT_>MIC值均>45.0%，万古霉素仅14.0%达到15~20 mg/L谷浓度范围。结论·3种抗生素均符合二室模型。血浆蛋白结合率对低通量IHD中抗生素的透析清除率有显著影响，血浆蛋白结合水平较高的抗生素在血液透析中的清除率较低。0.5 g/d美罗培南和2.0 g/d头孢曲松的用药方案在接受低通量IHD的患者中大多能达到疗效，而万古霉素方案（负荷剂量1.0 g、0.5 g/2 d维持）有治疗失败的风险。

Objective ·To evaluate the clearance and pharmacokinetics/pharmacodynamics (PK/PD) of antibiotics from the perspective of protein binding rates in critically ill patients undergoing intermittent hemodialysis (IHD), in order to explore the association between protein binding rate and dialysis clearance of antibiotics, and to provide theoretical basis for developing antibiotic dosing regimens during hemodialysis. Methods ·Nineteen patients undergone low-flux hemodialysis and received antibiotic therapy at the Department of Nephrology, Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine, were enrolled and divided into the meropenem group (n=7), the vancomycin group (n=5) and the ceftriaxone group (n=7) according to the type of antibiotics. A liquid chromatography with tandem mass spectrometry (LC-MS/MS) method was established to detect meropenem, vancomycin, and ceftriaxone in human plasma/serum and dialysate. A two-compartment pharmacokinetic model was established using MATLAB. Instantaneous and total dialysis clearance rates were calculated, and PK/PD parameters were analyzed. Results ·No significant differences were found in the clinical characteristics of subjects among the three groups. The dialysis clearance rates were as follows: meropenem group (5.14‒5.97 L/h) > vancomycin group (2.87‒3.77 L/h)> ceftriaxone group (1.21‒1.90 L/h), with statistically significant differences (P<0.001). All three antibiotics showed good fit in the two-compartment pharmacokinetic model with a dialysate chamber (fval%<2), and the calculated PK/PD parameters were consistent with previous literature. For meropenem, the fraction of time that the free drug concentration remained above the minimal inhibitory concentration (%fT_>MIC) values were 95.2%, 60.8% and 32.4% at minimal inhibitory concentration (MIC) values of 2, 8 and 16 μg/mL, respectively. For ceftriaxone (free concentration), the %fT_>MIC values were all above 45.0% at MICs of 0.25, 4 and 16 μg/mL. For vancomycin, only 14.0% of the trough concentrations reached the target range of 15‒20 mg/L. Conclusion ·The three antibiotics are well described by the two-compartment model. The plasma protein binding rate has a significant effect on the dialysis clearance of antibiotics in low-flux IHD, with higher protein binding associated with lower clearance. The regimens of meropenem (0.5 g/d) and ceftriaxone (2.0 g/d) are generally effective among patients undergoing low-flux IHD, while the vancomycin regimen with a loading dose of 1.0 g and a maintenance dose of 0.5 g/2 d carries a risk of treatment failure.