Contemporary HIV therapeutics is dominated by increasing frustration at the lack of durable virological suppression and the unaccaptable toxicity profile of the currently available antiretroviral drugs. There is also an increasing emphasis on tailoring treatment regimens for patients both in terms of their lifestyles and the sensitivity of their viral isolates. Hence it is not surprising that the wisdom of our current ‘hope based strategy of one dose fits all’ for antiretroviral therapy is now being questioned, and the potential role of therapeutic drug monitoring in optimizing the clinical benefit of currently available antiretroviral agents is being investigated. In particular, the protease inhibitors as substrates for the 3A4 isoform of cytochrome P450 demonstrate marked inter- and intrapatient variability and are potential candidates for therapeutic drug monitoring.
The paper by Patricia Baede-van Dijk et al.  examines the sources of variability in plasma concentrations of the parent compound nelfinavir and its active M8 metabolite in 355 HIV-infected patients attending outpatient clinics in a Dutch setting.
The authors noted marked variability in plasma nelfinavir and M8 levels in the 618 plasma samples analysed in this study. The median ratio between M8 and nelfinavir was 0.29, which was independent of the time following ingestion of the medication. However, lower M8 concentrations were identified in patients who were receiving inducers of CYP3A4 or inhibitors of CYP2C19 and in Black and Asian patients. Despite variation in the M8 metabolite in these subpopulations, the total concentration of nelfinavir plus M8 was essentially unchanged.
Consequently, assuming that M8 is as equipotent as the parent compound and assuming additive virological efficacy, the authors reasonably conclude that it is unnecessary to monitor plasma M8 concentrations in addition to nelfinavir levels for the purposes of therapeutic drug monitoring. However, it must be emphasized that there are now a number of approaches to the therapeutic drug monitoring of antiretroviral drugs.
In this particular study, the observed plasma drug concentration was compared with the population median on a pharmacokinetic population curve, while other centres favour estimation of trough plasma drug concentrations or random plasma samples from which trough plasma concentrations can be extrapolated. While dose adjustment of protease inhibitor therapy based on random or trough plasma levels is likely to be beneficial in the aggregate, these approaches may be suboptimal for a number of reasons.
First, in contrast to antibiotics, our understanding of the pharmacokinetics–pharmacodynamics of antiretroviral therapy is incomplete. While a number of studies have demonstrated that there is a concentration–effect relationship for protease inhibitor therapy, few studies have evaluated which pharmacokinetic–pharmacodynamic surrogate marker is the best predictor of antiretroviral activity. Measurement of trough plasma drug concentrations are only valid if the protease inhibitors exhibit time-dependent killing in a manner similar to β-lactam and glycopeptide antibiotics. Therefore, the utility of estimating single plasma drug concentrations is unclear at this time
Second, the possibility of synergy, antagonism or indifference between all of the components of the combination antiretroviral regimen is not considered. Third, in the case of nelfinavir, trough plasma concentrations may not represent the minimum plasma concentration over the dosing interval, as demonstrated in this paper. Finally, the use of drug concentration ratios or comparing trough plasma levels to fixed 95% inhibitory concentrations (IC95) merely explores the host–drug aspect of the host–drug–pathogen interaction. The IC95 values are derived from i n vitro data and are adjusted for plasma protein binding and, consequently, may not reflect the true in vivo situation for a particular patient. With the increasing availability of phenotype testing, it is likely that future endeavours in therapeutic drug monitoring will combine information from phenotypic sensitivity testing with the measured plasma drug concentration in an effort to truly individualize patient therapy. If the M8 metabolite is demonstrated to be equipotent in vivo with the parent compound, it is certainly possible that determination of M8 will be important, particularly for patients with highly resistant viral isolates.
1. Schentag J, Nix D, Forrest A, Adelman M. AUIC – the universal parameter within the constraint of a reasonable dosing interval
. Ann Pharmacother 1996, 30: 1029 –31.
2. Baede-van Dijk P, Hugen PWH, Verwij-van Wissen CPWGM, Koopmans P, Burger DM, Hekster YA. Analysis of variation in plasma concentrations of nelfinavir and its active metabolite M8 in HIV- positive patients
. AIDS 2001, 15: 993 –1000.