Fuchs Aline (1), Rotzinger Aurélie (2)(3), Cavassini Matthias (4), Schneider Marie Paule (2)(3), Csajka Chantal (1)(2)
(1) Division of Clinical Pharmacology, Service of Biomedicine, Department of Laboratory, Centre Hospitalier Universitaire Vaudois and University of Lausanne, Lausanne, Switzerland, (2) School of Pharmaceutical Sciences, University of Geneva, University of Lausanne, Geneva, Switzerland, (3) Community pharmacy, Department of ambulatory care & community medicine, University of Lausanne, Lausanne, Switzerland, (4) Infectious Disease Service, Centre Hospitalier Universitaire Vaudois and University of Lausanne, Lausanne, Switzerland
Objectives: To evaluate the impact of adherence measurement on efavirenz (EFV) and lopinavir (LPV) pharmacokinetic parameters estimation and to investigate whether real dosing history as measured by electronic monitoring (EM) compared to patient self-reported last dose intake modify the interpretation of concentrations measured within the frame of therapeutic drug monitoring (TDM) program.
Methods: All adherence and pharmacokinetic data from HIV-positive outpatients followed at the medication adherence enhancing program combining EM, pill count and motivational interviews [1], with at least one LPV or EFV concentration were analysed. Population pharmacokinetic modeling was performed based on previously published studies with NONMEM® [2, 3] using both dosing time recording methods (EM and self-report). Differences in predicted individual clearance (CL) were evaluated using a paired t-test and Bland-Altman plot. Bias and precision characterizing predictive performance between the two approaches were estimated by MPE and RMSE [4].
Results: 108 EFV levels from 55 patients and 131 LPV levels from 65 patients were collected. For both drugs, population estimates of CL were similar between EM and self-report recording methods, with a good precision on parameter, whereas the volume of distribution (Vd) differed, with a higher imprecision on parameter estimate. CL interindividual variability was similar between the two approaches. No significant difference in the estimated individual CL between the two dosing time recording methods was observed for both drugs (P=0.08 and P=0.4 respectively). The Bland-Altman plot suggested that 3 patients for EFV and 6 patients for LPV had a different estimation of individual CL according to the dosing history used. Relative predictive performance were similar for self-reported and EM program for both drugs (MPE: EFV=18 ng/mL; LPV=-52 ng/mL) but with a high imprecision for LPV (RMSE: EFV=173 ng/mL; LPV=1098 ng/mL).
Conclusions: CL was statistically insensitive to reported dosing method but Vd was more affected. These differences seem however of limited clinical importance since concentration predictions were similar between the two dosing time recording methods. Despite such results, combination of both methods may help to capture complementary aspects of patients’ adherence in routine care: longitudinal monitoring of patients’ actual behaviour through EM and actual quantitative dosing through TDM.
References:
[1] Krummenacher I, Cavassini M, Bugnon O, Spirig R, Schneider MP. Antiretroviral adherence program in HIV patients: a feasibility study in the Swiss HIV Cohort Study. Pharm World Sci. 2010 Dec;32(6):776-86.
[2] Arab-Alameddine M, Di Iulio J, Buclin T, Rotger M, Lubomirov R, Cavassini M, Fayet A, Decosterd LA, Eap CB, Biollaz J, Telenti A, Csajka C, Swiss HIVCS. Pharmacogenetics-based population pharmacokinetic analysis of efavirenz in HIV-1-infected individuals. Clinical pharmacology and therapeutics. 2009 May;85(5):485-94.
[3] Lubomirov R, di Iulio J, Fayet A, Colombo S, Martinez R, Marzolini C, Furrer H, Vernazza P, Calmy A, Cavassini M, Ledergerber B, Rentsch K, Descombes P, Buclin T, Decosterd LA, Csajka C, Telenti A, Swiss HIVCS. ADME pharmacogenetics: investigation of the pharmacokinetics of the antiretroviral agent lopinavir coformulated with ritonavir. Pharmacogenetics and genomics. 2010 Apr;20(4):217-30.
[4] Sheiner LB, Beal SL. Some suggestions for measuring predictive performance. Journal of pharmacokinetics and biopharmaceutics. 1981 Aug;9(4):503-12.
Reference: PAGE 23 (2014) Abstr 3152 [www.page-meeting.org/?abstract=3152]
Poster: Drug/Disease modeling - Infection