Therapeutic drug monitoring

Therapeutic drug monitoring

Prescribers choose treatment regimens on the basis that the benefits that accrue will outweigh the adverse effects. However, that outcome is never certain for individual patients even if the best evidence suggests that it is likely. The uncertainty arises for many of the reasons outlined in previous sections (e.g. pharmacodynamics, pharmacokinetics, pharmacogenetics). For this reason, whenever possible, prescribers should try to measure the effects of the drug, beneficial and harmful. This will help to inform decisions about dose titration (up or down) or discontinuation of treatment. Monitoring can be achieved in several different ways. It might be achieved subjectively be asking the patient about their symptoms. More objective approaches might be to measure the actual clinical effect. If this is not possible then alternatives are to measure a biomarker that acts as a surrogate for the clinical endpoint. Alternatively, if the pharmacodynamic effects of the drug are difficult to assess then it may be possible to measure the plasma drug concentration on the basis that it will be closely related to the effect of the drug.

Monitoring drug therapy - using drug concentration

Therapeutic drug monitoring is the term usually applied to the monitoring of drug therapy using plasma drug concentration. Three criteria must be met if plasma concentration is to be a useful marker for clinical endpoints:

  • Clinical endpoints and other pharmacodynamic (surrogate) effects are difficult to monitor. If either were available they would be preferable to a more ‘remote’ marker like plasma drug concentration.
  • The relationship between plasma concentration and clinical effects is predictable. If this relationship is unpredictable then the drug concentration will be a poor predictor of clinical benefit or toxicity. Amongst anti-epileptic drugs phenytoin is exceptional in having a good correlation between its plasma concentration and clinical effects.
  • The therapeutic index is low. For drugs with a high therapeutic index any variability in plasma concentrations is likely to be irrelevant. Knowledge of the dosage alone may be sufficient to predict whether the plasma concentration is within the therapeutic range.

Monitoring drug therapy - using drug effect

The optimal approach to monitoring is to assess the clinical endpoints directly (e.g. control of ventricular rate in a patient with atrial fibrillation). The drug dosage can be titrated to achieve the desired heart rate. Sometimes this is impractical because the clinical endpoint is a future event that cannot be detected until it is inevitable or irreversible (e.g. prevention of cardiovascular events by statins) or because it cannot be assessed at a time that allows it to be used a guide for therapy (e.g. treating a chest infection with antibiotics). In these circumstances it may be possible to select a surrogate endpoint that will serve to predict success or failure in achieving the desired outcome. The surrogate end point may be either:

  • Directly related to the clinical endpoint by being an obligatory intermediate step in the pathophysiological process (e.g. serum cholesterol as a surrogate for myocardial infarction)
  • Indirectly related to the clinical endpoint because, although it indicates the progress of the pathophysiology is not a key factor in progression to the clinical endpoint (e.g. serum C-reactive protein as a surrogate marker for resolution of inflammation).

When the surrogate marker is a biological characteristic it is termed a ‘biomarker’.