Saliva as an Alternative Matrix for Antibiotic Pharmacokinetics and Therapeutic Drug Monitoring: Current Evidence and Future Directions: A Narrative Review

Saliva is an attractive alternative matrix for pharmacokinetic assessment and therapeutic drug monitoring (TDM) because it is non-invasive, readily repeatable and potentially suitable for decentralized sampling. However, salivary drug concentrations are influenced by protein binding, ionization, lipophilicity, salivary pH and flow, collection procedures, oral contamination and analytical performance, limiting simple extrapolation from plasma. This narrative review synthesizes mechanistic, bioanalytical, pharmacokinetic and clinical evidence on salivary monitoring of antibacterial agents, with particular attention to antituberculosis drugs and the distinction between correlation and clinically meaningful agreement. Evidence varies substantially among drug classes. Linezolid has the most developed evidence base, with studies supporting saliva-based exposure estimation, although individual agreement, external validation and prospective evaluation of saliva-guided dosing remain insufficient for routine plasma substitution. Levofloxacin shows potential primarily as a screening tool for subtarget systemic exposure in multidrug- or rifampicin-resistant tuberculosis, whereas pyrazinamide may permit semiquantitative exposure assessment. Evidence for ofloxacin, isoniazid, rifampicin, metronidazole and selected macrolides suggests pharmacokinetic associations but remains inadequate for reliable individualized dosing. In contrast, current data do not support routine salivary TDM for aminoglycosides, vancomycin, most β-lactams, tetracyclines, ciprofloxacin or moxifloxacin. Across the literature, small cohorts, heterogeneous sampling methods, inadequate analytical sensitivity, and reliance on correlation rather than agreement or prediction-error analyses remain major limitations. Saliva should therefore be considered a drug- and application-specific investigational matrix rather than a general substitute for plasma. Future translation requires standardized collection and assay validation, prospective pharmacokinetic qualification, external validation of predictive models and demonstration that saliva-based dosing decisions improve clinical outcomes.