An easy-to-use predictive score for identifying patients at risk of suboptimal linezolid exposure and candidates for therapeutic drug monitoring

Abstract

Objectives

Due to the high interindividual pharmacokinetic variability, only between 34.5–57.5% patients treated with linezolid fall within its therapeutic range with standard dosing. This study aimed to develop and validate a simple-to-use predictive score to identify patients at risk of sub- or supra-therapeutic linezolid exposure who may benefit from therapeutic drug monitoring (TDM).

Methods

A retrospective observational study was conducted using data from 819 patients treated with linezolid and undergoing TDM (2011–2022). Linezolid trough concentrations were classified as sub-therapeutic, therapeutic and supra-therapeutic (< 2 mg/L, between 2 and 7 mg/L and > 7 mg/L, respectively). A multinomial logistic regression model was used to develop a predictive score, which was externally validated in a separate cohort of 73 patients. Discrimination performance was assessed using ROC curve.

Results

Only 31.5% of patients achieved therapeutic concentrations. Median linezolid trough concentrations were 3.9 mg/L (range: 0.5–63.7 mg/L). Independent predictors of sub-therapeutic concentrations were a younger age, a low linezolid dose/body weight, augmented renal clearance and intensive care unit admission. Predictors of supra-therapeutic concentrations were a higher dose/body weight, renal impairment and liver cirrhosis. The predictive model showed high accuracy for identifying patients at risk of supra-therapeutic exposure (AUC 82% derivation; 90% validation), while performance was moderate for underexposure (AUC 63% derivation; 49% validation). An online calculator was implemented to facilitate score use in clinical practice.

Conclusions

The predictive score provides clinicians with an effective tool to support individualised linezolid therapy and guide TDM implementation, reducing the risk of toxicity or treatment failure.