Free Drug Monitoring by Liquid Chromatography and Implications for Therapeutic Drug Monitoring

Abstract

Free (non-protein bound) drug monitoring is very complex but new advances in separation of protein bound from free drug by ultrafiltration can facilitate free concentration therapeutic drug monitoring. Free drug can be separated from protein-bound drug by means of ultrafiltration in which the free drug passes through the membrane filter while the protein bound drug is retained. The ultrafiltrate is then subjected to chromatographic analysis (HPLC). Limitations include a sufficient sample for these highly protein bound drugs and detector sensitivity at the low free drug concentrations. Free drug fractions that have been successfully analyzed by HPLC include disopyramide, phenytoin, carbamazepine, and propranolol. HPLC offers important advantages over immunoassays, since chromatographic methods can also measure the amount of metabolite present. The detection of metabolites can be important, e.g. N-desisopropyl disopyramide which displaces disopyramide from its binding sites on alpha-one acid glycoprotein. The main controversies concerning free drug concentration monitoring have revolved around the time expenditure as well as clinical significance of changes in free fraction. Obtaining free and total drug concentrations of monitored drugs permits an assessment of the free fraction of drug and provides pharmacokinetic parameters for more accurate dose prediction and dose adjustment, in which HPLC can offer an attractive method of analysis.

Free drug concentration monitoring is controversial and is not indicated for all highly protein bound drugs (e.g., warfarin and the sulfonylureas). Candidates for free drug monitoring must have a relationship between pharmacologic effect and drug concentrations, or considerable variance in the free fraction of drug in serum. Potentially toxic drugs demonstrating poor correlations by linear and orthogonal regression of total drug concentrations versus free drug concentrations are ideal for such monitoring. For such drugs, nomograms developed to predict free drug concentrations are often unhelpful, particularly with seriously ill patients. Likewise highly protein bound drugs with metabolites capable of displacing parent drug may also unpredictably affect the free drug concentrations.