Adaptive PD-optimal design of a pilot study for a clotting time test for enoxaparin

Background: Dosing of enoxaparin, like other anticoagulants, may result in bleeding following excessive doses and clot formation if the dose is too low. There is no standard measure of enoxaparin clinical effectiveness. We recently showed that a Xa clotting time test could potentially assess the effect of enoxaparin on the clotting system [1]. Preliminary in vitro experiments using plasma from a range of healthy volunteers showed wide variability in clotting times. Effectiveness of an additional substrate “Z” in reducing the variability in the clotting time was assessed in a single healthy volunteer. There is a need to optimally choose the concentrations of Xa and “Z” for the pilot study that will be followed by the proof-of-concept study. Optimal adaptive design can be used to address this issue as the parameter estimates obtained from experiments with volunteer 1 would be used to design the experiments for the next volunteer. In addition, D- and P-optimality criteria can be combined so that they provide a means to estimate parameters and also maximize the probability of getting a successful outcome.

Aim: To develop an adaptive PD-optimal pilot study for a clotting time test to assess the anticoagulant effect of enoxaparin.

Methods: The preliminary data from one single healthy volunteer was modelled and the estimated parameter values were used to locate PD-optimal concentrations of Xa and “Z”. This set of design variables (PD1) will be used for the next healthy volunteer of the pilot study. The second volunteer will be enrolled and data collected according to PD1. The data collected will then be modelled and the PD-optimal design (PD2) will be located. This iterative process will be repeated until all healthy volunteers (n=6) are enrolled. The data collected from the healthy volunteers will be in the form of time taken for their plasma sample to clot after activation by the PD-optimal Xa concentration in the presence of CaCl2 and PD-optimal “Z” concentration. This will be in the absence or presence of enoxaparin (0.25 IU/ml and 1.0 IU/ml). A successful clotting time test would be a test that gives a clotting time between 10 and 30 sec in the absence of enoxaparin. And is prolonged at least 1.5-fold in the presence of 0.25 IU/ml enoxaparin but less than 5-fold when 1.0 IU/ml enoxaparin is present.

Results and Discussion: Assuming there are no random effects in the parameter KDXa (affinity of Xa), precise parameter estimates were obtained in the adaptive design. Using estimated parameter values based on the data from one single healthy volunteer, PD-optimal concentrations of 15 nM of Xa and 0.25% of “Z” were located. These concentrations will be used to carry out the in vitro experiments with the second volunteer and have been predicted to give >80% chance of a successful clotting time test. Optimal adaptive design provided us with the most informative data with the most efficient and parsimonious use of resources. The resulting PD-optimal design after the 6th volunteer will be used as an initial design for the proof-of-concept study.

[1] Gulati A, Faed, J. M., Isbister, G. K., Duffull, S. B. (2012). Development and evaluation of a prototype of a novel clotting time test to monitor enoxaparin. Pharmaceutical Research 29:225-235.