Therapeutic Drug Monitoring
When and how to measure drug levels to optimize therapy, including sampling strategies, target ranges, and clinical decision-making.
## What Is Therapeutic Drug Monitoring?
Therapeutic drug monitoring (TDM) uses measured drug concentrations in biological fluids (usually plasma or serum) to individualize dosing and optimize therapeutic outcomes. TDM bridges pharmacokinetics and clinical practice by replacing empiric dosing with data-driven adjustments.
## When Is TDM Indicated?
TDM is valuable when all of the following conditions are met:
1. **Narrow therapeutic index**: small difference between effective and toxic concentrations
2. **Large interindividual variability**: patient factors make plasma levels unpredictable from dose alone
3. **Defined concentration-response relationship**: target ranges are established and validated
4. **Available assay**: accurate, precise, timely analytical method exists
5. **Actionable results**: measured levels lead to dose changes that improve outcomes
## Drugs Commonly Monitored
| Drug Class | Examples | Target Range | Key Concern |
|-----------|---------|-------------|-------------|
| Aminoglycosides | Gentamicin, tobramycin | Peak 5-10, trough < 2 mcg/mL | Nephrotoxicity, ototoxicity |
| Vancomycin | — | AUC/MIC 400-600 | Nephrotoxicity |
| Antiepileptics | Phenytoin, carbamazepine, valproate | Drug-specific | Seizure control vs toxicity |
| Lithium | — | 0.6-1.2 mEq/L | Narrow index, renal changes |
| Digoxin | — | 0.8-2.0 ng/mL | Cardiac arrhythmias |
| Immunosuppressants | Cyclosporine, tacrolimus, sirolimus | Drug-specific | Rejection vs toxicity |
| Theophylline | — | 10-20 mcg/mL | Seizures, arrhythmias |
| Methotrexate | — (high-dose) | < 0.05 mcM at 48h | Leucovorin rescue timing |
## Sampling Strategy
### Trough Levels
Collected immediately before the next dose (Cmin). The most common TDM sample because timing is straightforward and trough levels correlate with both efficacy and toxicity for many drugs. Used for vancomycin, antiepileptics, immunosuppressants, digoxin, and lithium.
### Peak Levels
Collected 30-60 minutes after IV infusion completion or 1-2 hours after oral dose. Important for aminoglycosides (peak correlates with bactericidal efficacy).
### AUC-Based Monitoring
Vancomycin guidelines now recommend AUC/MIC-guided dosing rather than trough-only monitoring. AUC can be estimated from two-point sampling (peak + trough) using Bayesian software or from a single trough using population PK models.
## Interpreting Results
Before adjusting doses based on drug levels, confirm:
- **Timing**: Was the sample drawn at the correct time relative to the dose?
- **Steady state**: Has the patient been on a stable dose for at least 5 half-lives?
- **Adherence**: Has the patient been taking doses consistently?
- **Assay specificity**: Some immunoassays cross-react with metabolites (e.g., digoxin-like immunoreactive substances in renal failure)
- **Free vs total levels**: For highly protein-bound drugs (phenytoin, valproate), free levels are more accurate in hypoalbuminemia
## Bayesian Dose Optimization
Modern TDM incorporates Bayesian forecasting, which combines population PK models with individual measured concentrations to estimate personalized PK parameters. Software tools (DoseMeRx, InsightRX, MwPharm) can then simulate various dosing regimens and recommend optimal adjustments.
## Key Takeaways
- TDM is indicated for drugs with narrow therapeutic indices and high PK variability
- Correct sample timing and confirmation of steady state are prerequisites for accurate interpretation
- Trough levels are the most commonly used TDM samples
- AUC-based monitoring is replacing trough-only approaches for vancomycin
- Bayesian dose optimization combines population models with individual data for precision dosing