S. Chantel(1), P. Martin(1), P.Y. Petit(2), D. Massignon(3), F. Saulnier(2), M. Benoist(2), S. Granger(2), P. Maire(4,5), R.W. Jelliffe(5), G. Aulagner(1)
(1) Department of Pharmacy, Hôpital Cardiologique, Bron, France; (2) Department of Anaesthesia, Centre Hospitalier Lyon-Sud, Pierre-Bénite, France; (3) Department of Haematology, Centre Hospitalier Lyon-Sud, Pierre-Bénite, France; (4) ADCAPT, Department of Pharmacy, Hôpital A. Charial, Francheville, France; (5) Laboratory of Applied Pharmacokinetics, University of South California, Los Angeles, USA
Introduction: The use of enoxaparin to prevent venous thromboembolism after orthopaedic surgery at a fixed dose (40 mg subcutaneously once daily) leads to an important pharmacokinetic variability, which may have consequences on the treatment’s efficacy. This variability is partly explained by covariates, however it still remains a residual variability.
Aim: To evaluate MAP (Maximum A posteriori Probability) Bayesian fitting, using one anti-Xa activity measured 3-4 hours after injection, to predict future anti-Xa levels.
Methods: Population pharmacokinetic parameters were estimated from clinical and biological data of 32 patients (group 1) receiving enoxaparin for venous thromboembolism prophylaxis after total hip replacement (NPEM2 program version 11.7). Future anti-Xa activities were simulated for 16 other patients (group 2) using USC*Pack following 3 methods: 1) MAP Bayesian procedure after one anti-Xa measurement (at day 1), 2) A priori method including body weight and creatinine clearance (AP1) and 3) A priori method without covariates (AP2). Simulated and measured anti-Xa levels at day 5 were compared (bias, precision).
Results: Age, gender, body weight and creatinine clearance were not statistically different between the 2 groups of patients. Simulations of anti-Xa activity by MAP Bayesian procedure, AP1 and AP2 showed respective bias ± standard deviation : 0,03 ± 0,12, 0,08 ± 0,12 and -0,10 ± 0,11 IU/mL, and precisions : 0,0148, 0,0206 and 0,0226. No correlation was found between individual error by MAP Bayesian procedure and body weight (r=0,028; NS), age (r=0,431; NS) nor creatinine clearance (r=0,369; NS).
Conclusion: MAP Bayesian procedure seems to be the less biased and the most precise method to predict future anti-Xa activities for patients receiving enoxaparin at prophylactic regimen. Inclusion of covariates in the A priori model does not improve predictions. Further investigations may be required to evaluate the improvement of pharmacokinetic variability control by MAP Bayesian procedure.
Reference: PAGE 14 () Abstr 802 [www.page-meeting.org/?abstract=802]
Poster: poster