Dosing Guidance for Intravenous Colistin in Critically Ill Patients
Nation RL, Garonzik SM, Thamlikitkul V, et al. Clin Infect Dis. 2017; 64: 565-71.
Study Question: What are the population pharmacokinetics (PK) of colistimethate (CMS) and colistin and can they be used to form an improved dosing algorithm?
Study Description: An open-label, non-randomized study evaluated the population PK of CMS and colistin in 214 critically ill adults treated for pneumonia or bloodstream infection caused by gram-negative bacteria. Exclusion criteria were patients with cystic fibrosis and those receiving a concomitant inhaled polymyxin. All dosing regimens were at the discretion of the treating practitioner (median daily dose 200mg colistin base activity (CBA), range 75-600mg CBA). Steady state plasma concentrations of colistin were evaluated across a dosing interval. The dosing algorithm was constructed to achieve an average steady state concentration (Css,avg) of 2mg/L or more in > 80% of patients, 1.5mg/L or more in > 90% of patients, and a concentration of 4 mg/L or more in < 30% of patients.
Results: Median colistin Css,avg was 2.35 mg/L, but varied widely (range 0.24-9.92 mg/L). Significant covariates included creatinine clearance for clearance of CMS and colistin and body weight for the volume of peripheral and central CMS compartments. Using the gathered PK data, a derived algorithm (see article) was calculated to achieve Css,avg ≥ 2mg/L in at least 80% of all creatinine clearance groups except those with creatinine clearance ≥ 80mL/min (likely due to daily dose capping at 360mg CBA). All creatinine clearance ranges achieved Css,avg > 4mg/L in less than 30% of patients.
Conclusion(s): The derived dosing algorithm is predicted to reliably achieve a desired Css,avg to maximize effectiveness of colistin therapy while minimizing nephrotoxicity. Practitioners treating patients with creatinine clearance ≥ 80mL/min should strongly consider combination therapy.
Perspective: Historically, dosing of colistin has been challenging due to wide inter-patient variability. This study provides dosing recommendations based on the largest compilation of CMS/colistin PK data to date, and greatly enlarged the available PK data in patients requiring renal replacement therapy. In addition, the dosing “look-up table” simplifies use of the algorithm in comparison to the algorithm recommended following the interim analysis of this dataset. Of note, the algorithm may under-dose patients with a high creatinine clearance (≥ 80 mL/min) and organisms with elevated minimum inhibitory concentrations (MIC ≥ 1 mg/L). This study likely provides the best overall recommendations for colistin dosing published to date.