Criteria to assess the clinical value of drug monitoring
The criteria for TDM to be clinically useful for a particular drug may therefore be developed and summarised as follows:
• Poor correlation between the dose given and the plasma concentration obtained in different patients (wide inter-individual pharmacokinetic variability). Clearly, if the dose given is an effective predictor of plasma concentration in all patients, then measure ment of an individual’s plasma concentration is superfluous. Compliance testing (deter mining whether patients actually take the drugs – often now referred to as concordance testing) may be a useful adjunct to TDM programmes (see below), but is unlikely to justify such a programme in the absence of other indications for monitoring.
• Good correlation between plasma concentration and pharmacological effect in different patients (low inter-individual pharmaco dynamic variability). If plasma concentration measurements do not give accurate information about response, they are at best useless and at worst misleading. So the two premises stated above must be satisfied (plasma drug concentration predicts receptor concentration and receptor drug concentration predicts response). Active metabolites are generally undesirable as they contribute to the effect but make a variable (or zero) contribution to the concentrations, depending on the assay system; for example, a metabolite may have 10% of the biological activity of the parent drug, but show 100% cross-reactivity in an immunoassay. This criterion also normally requires that the action of the drug at the receptor site be essentially reversible. If this is not the case, and the drug binds irreversibly to the receptor, the pharmacodynamic half-life (or duration of effect) may be markedly longer than the pharmacokinetic half-life (or length of time the drug can be detected in the circulation). In this situation, the drug may still be exerting an effect when no drug can be detected in the plasma, and it is difficult to argue that TDM has a useful contribution to make. The exception to this general rule may be some anti-cancer agents, for which an index of the body’s total exposure to the drug may predict subsequent response.
• TDM is only clinically relevant for drugs that show significant toxic or undesirable effects at plasma concentrations only slightly above those required for useful effects. If there is a wide margin between effective concentrations and undesirable effects, as is the case for pencillin, effective therapy may be achieved by giving the drug to all patients in large excess, with no need for individualisation of therapy. In contrast, the aminoglycoside antibiotics have relatively narrow margins between effective concentrations and those that produce unacceptable toxicity, so concentration monitoring has an essential role in ensuring maximal effect with minimal toxicity.
• Similarly, TDM is redundant for drugs for which there is a better clinical marker of effect, for example blood pressure, plasma glucose concentration or prothrombin time. Plasma drug concentrations have little to offer in this situation, except in the elucidation of rare cases where a high dose of drug fails to produce the desired effect, when TDM may help differentiate between noncompliance, poor drug absorption, rapid drug metabolism(cytochrome-P450(CYP)status), receptor dysfunction or the use of the wrong drug for the situation.
Application of these criteria results in a relatively short list of drugs for which concentration monitoring has a proven clinical role. Measurement of other drugs may contribute in isolated cases or particular clinical circumstances, but it cannot be said to be generally useful in dosage.
