Background. Therapeutic drug monitoring (TDM) for levofloxacin, a key multi-drug-resistant tuberculosis (MDR-TB) drug, is highly recommended but widely impractical. Saliva sampling with limited sampling strategy offers a non-invasive and more feasible alternative to blood for TDM, yet no validated saliva model exists to guide dosing.
Aim. To develop a saliva population pharmacokinetic (PK) model for levofloxacin in Vietnamese MDR-TB patients.
Methods. Adults receiving levofloxacin for ≥7 days had blood and saliva samples collected at 0, 2, and 5 hours post-dose at steady state (ACTRN12620000681954). Levofloxacin concentrations were quantified using HPLC-MS. One and two-compartment models with additive, exponential, and combined error models, and lag or transit absorption models were tested through baseline model development, and potential covariates were tested on the parameters using stepwise covariate modelling (scm) and diagnostics checks (NONMEM 7.4.4 and PsN 5.3.1). After validating plasma model, saliva component was integrated either as a scaling factor or as a distinct compartment. Sampling importance resampling technique (SIR) was used to assess the robustness of the final parameter estimates.
Results. Fifty-seven patients were included with 342 paired saliva-plasma samples after removing concentrations below limit of quantification. One-compartment population PK model with 1st order absorption and elimination, in which a saliva compartment was integrated, best described the PK in both plasma and saliva. Distinct compartment method outperformed scaling method in terms of objective function (dOFV = ̶ 4.949, p < 0.01, at α = 0.05 with 1 degree of freedom), covariance/correlation steps, model stability and precision. No significant covariates were identified. The final model was optimised using the FOCE-I with PRIOR approach, yielding population parameters: Ka = 4.18 h-1, Tlag = 0.95 h, CL/F = 10.3 L/h, V/F = 278.8 L, K23 (transfer rate from plasma to saliva) = 4.9 h-1, K30 (elimination rate from saliva) = 5.1 h-1, IIVCL = 41.2%, IIVV = 69.7%, additive error = 0.1 mg/L and respective proportional residual errors of 22.4% (plasma) and 34.5% (saliva). Visual diagnostics confirmed a good predictive performance (Fig. 1).
Conclusion. By leveraging Bayesian-guided estimation with limited samples, the validated saliva levofloxacin model offers a promising framework for enabling saliva-based TDM such as informing TDM implementation strategies or future clinical trials.
Figure 1. a) Diagnostic goodness-of-fit plot of the final levofloxacin saliva model: observed versus individual predicted concentrations; and b) Visual predictive check of the model for plasma (CMT:2) and saliva (CMT:3).