Assessment of pharmacokinetic variability of fondaparinux in 809 patients treated after major orthopedic surgery: the POP-A-RIX study
Delavenne X (1), Laporte S (1,2), Deygas B (1,2), Baylot D (3), Zufferey P (4), Barré J (5), Nguyen P (6), Borg JY (7), Charret F (8) and Mismetti P (1), for the GETHCAM investigators
(1) Dept. of Clinical Pharmacology ; EA 3065 ; University Hospital . (2) Inserm, CIE3, University Hospital of Saint-Etienne, France.(3) Clinique Mutualiste de Saint-Etienne, France.(4) Dept. of Anesthesiology and Intensive Care ; EA3065 ; University Hospital of Saint-Etienne, France.(5) Dept. of Anesthesiology, Robert Debré hospital, Reims, France.(6) Central Lab. of Hematology, Robert Debré hospital, Reims, France.(7) Hemostasis Unit, University Hospital of Rouen, France.(8) Clinique des Cévennes, Annonay, Franceof Saint-Etienne, France.
Objectives: Fondaparinux is a synthetic antithrombotic agent with specific anti–factor Xa activity. Despite a favorable benefit risk ratio in average, fondaparinux as other antithrombotics increase the risk of hemorrhage. One of the objectives of this study is to characterize pharmacokinetic of this drug according to patient characteristics using a population model, which may help to predict the risk of hemorrhage in specific populations.
Methods: Prophylactic dosage (2.5 mg once a day) of fondaparinux was administrated during at least 5 days. One to four samples were collected throughout the duration treatment. Plasma concentrations were assayed by enzymatic anti-Xa activity method. Population pharmacokinetic parameters and inter-individual variability were estimated on a random splitting of the dataset in the 2/3 of patients using NONMEM VI software. A covariate analysis was performed to explain a part of inter-individual variability of parameters. Model validation was based on visual predictive check and remaining 1/3 of patients as the test set.
Results: A total of 809 patients were included in the study, 566 in the training set and 243 in the validation set. A two-compartment model with first order absorption best fitted the plasma concentrations. Covariate analysis showed that creatinine clearance (CrCl) and sex were associated with an increased value of clearance (CL) (p<0.001). In addition, significant correlation was identified with body weight and central volume of distribution (V2) (p<0.001). For a typical woman (ClCr = 67 ml/min, body weight = 70 kg), the population (CL) was estimated to 0.24 L/h and V2 to 8.03 L, with inter-patient variabilities equal to 65% and 56% respectively. The Cl decreases from 0.24 L/h to 0.19 L/h in the typical woman would present with a CrCl equal to 50 ml/min. Coefficient of variation and standard deviation of the residual error were 12.7% and 0.04 mg/L, respectively. The visual predictive check evaluation confirmed that the full model was a good description of data. Finally, the external validation step showed an improvement in the predictive performance of the full model compared to the model without covariates.
Conclusions: It is the first pharmacokinetic developed of fondaparinux in a large population of patients after major orthopedic surgery. CrCl, body weight and sex were identified as explaining a part of inter-individual variabilities but the need for a therapeutic drug monitoring is questionable without more clinical investigation. To answer to this question, PK parameters will then be included in a multivariable analysis to assess their correlation with the risk of hemorrhage.
