Saliva point-of-care testing for suboptimal levofloxacin and linezolid exposure in multidrug-resistant tuberculosis

Background Therapeutic drug monitoring may optimise treatment outcomes in patients with multidrug-resistant tuberculosis, yet current methods are invasive and resource intensive. This study evaluated the feasibility of using saliva point-of-care testing to screen for suboptimal exposure to levofloxacin and linezolid. Methods We conducted a community-based cross-sectional study in Vietnam (2021–2023). Drug concentrations were measured in paired plasma and saliva samples at 0, 2 and 5 h post-dose, using a point-of-care test (NanoPhotometer NP80) in comparison with liquid chromatography-mass spectrometry. Drug exposure (area under the drug concentration–time curve from 0 to 24 h (AUC 0–24 )) was calculated using a Bayesian population pharmacokinetic approach (Edsim++). Therapeutic targets were defined as a free AUC 0–24 to minimum inhibitory concentration ratio of 160 for levofloxacin and 125 for linezolid. Receiver operating characteristic curve analyses were conducted to determine optimal saliva thresholds for predicting subtherapeutic or supratherapeutic plasma exposure. Results A saliva exposure threshold of 70.3 mg·h·L −1 accurately identified 16 out of 19 patients with subtherapeutic levofloxacin plasma exposure (sensitivity 84%, specificity 78%; n=60). For linezolid (n=17), saliva testing correctly identified the only patient with subtherapeutic exposure and the only patient with supratherapeutic exposure. However, no saliva exposure thresholds could be determined. Conclusions Saliva-based point-of-care testing may predict the plasma drug exposure of levofloxacin and linezolid, thus enabling the detection of patients at risk of suboptimal drug exposure. This non-invasive tool warrants further investigation to establish a saliva-based threshold for linezolid and to evaluate the clinical impact and cost-effectiveness of saliva-guided drug dosing.