The relationship between oral factor Xa (FXa) inhibitor du-176b pharmacokinetics (PK) and the probability of bleeding events (BE) in patients with atrial fibrillation (AF)
Abstract number: OC-WE-003
Giugliano1 R., Rohatagi2 S., Kastrissios3 H., Green3 M., Carrothers3 T.J., Mendell4 J., Antman1 E., Salazar5 D.
11Brigham and Women's Hospital, Boston 22Operations and Management, Daiichi Sankyo Pharma Development, Edison 33Pharsight Corporation, Mountain View 44Clinical Pharmacology 55Translational Medicine and Clinical Pharmacology, Daiichi Sankyo Pharma Development, Edison, USA
How-to-cite Giugliano R, Rohatagi S, Kastrissios H, Green M, Carrothers TJ, Mendell J, Antman E, Salazar D. The relationship between oral factor Xa (FXa) inhibitor du-176b pharmacokinetics (PK) and the probability of bleeding events (BE) in patients with atrial fibrillation (AF). Journal of Thrombosis and Haemostasis 2009; Volume 7, Supplement 2: Abstract OC-WE-003
Background: DU-176b is being developed to reduce stroke and systemic embolic events in AF. A Phase IIb dose-ranging study in AF patients assessed the safety and PK of DU-176b.
Methods: DU-176b PK and BE data from a randomized, parallel group, multinational safety study in AF patients (CHADS2 ≥ 2) were analyzed. Patients (n = 1145) were randomized to either double-blind DU-176b or open-label warfarin (INR 2.0–3.0) for 3 months. BE were blindly adjudicated by an independent CEC using ISTH definitions. PK samples were drawn at pre-dose and 1–3 h post-dose on Day 28. Population PK methods were used to estimate individual exposure metrics at steady-state concentrations: maximum (Cmaxss), minimum (Cminss) and area-under-the-curve (AUCss). Relationships between exposure metrics and probability of BE (all bleeds) were explored using logistic regression.
Results: The probability of BE significantly correlated with all three exposure metrics, but most strongly correlated with Cminss (P = 0.01 for major bleeds and major plus clinically relevant bleeds, P < 0.0001 for minor bleeds and P < 0.00001 for all bleeds).
Table 1 Incidence (%) of BE: Log Regression of DU-176 Cminss vs. BE
| Dose (mg) | Observed | Predicted (95% PI) |
|---|
| 30 QD | 5.5 | 7.1 (5.4, 8.8) |
| 60 QD | 8.1 | 8.8 (7.0, 10.7) |
| 30 BID | 13.1 | 11.7 (9.6, 13.8) |
| 60 BID | 18.9 | 17.1 (15.4, 20.4) |
Conclusions: DU-176 levels strongly correlated with the probability of BE in AF patients. Cminss, the best predictor of BE, provided insight into the observed dose-BE response and allowed optimization of DU-176b dosing that is now underway in a phase III trial.
Disclosure of interest: R Giugliano/TIMI Study Group, Daiichi Sankyo, Grant/Research-Support.
S Rohatagi, Daiichi Sankyo, Employee.
H Kastrissios, Daiichi Sankyo, Consultant.
M Green, Daiichi Sankyo, Consultant.
T Carrothers, Daiichi Sankyo, Consultant.
J Mendell, Daiichi Sankyo, Employee.
E Antman/TIMI Study Group, Daiichi Sankyo, Grant/Research-Support.
D Salazar, Daiichi Sankyo, Employee.
