• A Population Based-Analysis of Amprenavir Trough Concentrations
  Authored by Kurowski M., Sternfeld T., and Sawyer A.W.
  Poster P187

The subject of therapeutic drug monitoring (TDM) has assumed increased importance in HIV patient management, particularly as we have gained a fuller understanding of the importance of maintaining drug plasma concentrations above Cmin levels in order to both assure continuing antiviral effect and minimize the likelihood of selection of drug resistant forms. This study from two academic centers in London and Berlin set out to assess levels of amprenavir (APV, Agenerase) in 611 samples collected from 281 patients treated with amprenavir during 2001 and who were followed as part of a TDM protocol. These authors developed a predictive model of plasma trough levels of amprenavir in individuals receiving defined drug regimens.

Patients who also received lopinavir (LPV, Kaletra) were excluded from this analysis because of well-described amprenavir/lopinavir interactions in regard to plasma drug levels. This factor led to a final analysis of only 89 samples from 87 individuals. The data were fit to curves using a formula termed the "Maxim Likelihood Method" that maximizes the chances of observing a sample that displays the same characteristics as do actual data. This analysis was also enhanced through use of a software package termed "Mathematica" and a "Likelihood Equation" to perform differential calculus.

Data were analyzed in regard to 89 patients who received amprenavir/ritonavir (RTV, Norvir) regimens of 600/100 or 600/200 mg twice daily, in combination with two other antiviral drugs. Two distinct populations were revealed with trough levels more than and less than 500 mg amprenavir/ml. The geometric mean was 1,618 mg amprenavir/ml. They then fit a truncated normal distribution to amprenavir trough concentrations through use of a normal probability plot (NPP). Through these measurements, the authors concluded that mean plasma trough levels of amprenavir are in accordance with the results of clinical studies in which amprenavir/ritonavir was employed at 600/100 mg twice daily. In this regard, amprenavir seems to show lower inter-patient variability in regard to trough levels than other protease inhibitors (PIs). Patients with trough levels less than 500 mg/ml appear to be outliers, and these cases may be due to non-adherence to antiretroviral therapy regimens, pharmacogenetic factors (e.g., drug transporter genes), and/or other considerations that require further study.

Unfortunately, most clinicians are unlikely to find this type of analysis useful because the complicated mathematical calculations render the interpretation of data extremely difficult. In addition, this paper contains some degree of inherent confusion at the cut-off in regard to outliers and distinct populations in the Abstract is presented as 700 ng/ml in regard to APV trough levels and this figure was revised to 500 ng/ml in the poster. When confronted on this issue, the authors stated that they had recalculated their results from the time that the poster was written. This answer did not leave the reviewer with a sense of increased confidence about any facile interpretation of results.