Moreno Ursino (1), Sarah Zohar (1), Frederike Lentz (2), Corinne Alberti (3), Tim Friede (4), Nigel Stallard (5) and Emmanuelle Comets (6,7)
(1) INSERM, UMRS 1138, team 22, CRC, University Paris 5, University Paris 6, Paris, France, (2) Federal Institute for Drugs and Medical Devices, Bonn, Germany, (3) INSERM, UMR 1123, Hôpital Robert-Debré, APHP, University Paris 7, Paris, France, (4) Department of Medical Statistics, University Medical Center Göttingen, Göttingen, Germany, (5) Statistics and Epidemiology, Division of Health Sciences, Warwick Medical School, The University of Warwick, UK, (6) INSERM, CIC 1414, University Rennes-1, Rennes, France, (7) INSERM, IAME, UMR 1137, University Paris Diderot, Paris, France.
Objectives: To highlight the benefits of conducting a sequential adaptive dose-finding clinical trial by adding information given by PK measurements. We compared several methods for dose selection, including or not PK data, by looking at the percentage of selection of the maximum tolerated dose (MTD). We evaluated whether taking into account PK information brings any benefit.
Methods: The objective phase I dose-finding studies in oncology of is to determine the MTD while limiting the number of patients exposed to high toxicity. Several Bayesian methods to estimate probability of toxicity were compared through simulations. Of the six methods we tested, some estimate the probability of toxicity directly versus dose, others instead estimate the probability of toxicity passing through the AUC. AUC is treated as covariate for the link function of probability of toxicity and/or as dependent variable in linear regression versus dose.
For simulation, we assumed toxicity to be related to AUC and considered 6 possible dose levels.
We simulated trials based on a model for the TGF-β inhibitor LY2157299 in patients with glioma [1]. The PK model was reduced to a one-compartment model with first-order absorption as in [2]. Toxicity occurred when AUC was above a given threshold. For each scenario, 7 in total, we simulated 1000 trials with 30 patients to 60. We evaluated the ability of each method to estimate the dose-toxicity relationship by considering the estimate of the probability of toxicity for each tested dose.
Results: Methods which incorporate PK measurements had comparable performance to those without PK data in terms of percentage of MTD selection. Regarding the ability to estimate the dose-toxicity relationship, looking at the credible intervals built by the first and the third quartile, all the methods are able to estimate properly the probability of toxicity at MTD and at adjacent doses in each scenario; however, only methods which includes PK as a dependent variable are able to estimate adequately the probability associated to all the doses.
Conclusions: Incorporating PK values did not alter the efficiency of estimation of MTD but it increased the ability to estimate the entire dose-toxicity curve. This aspect is very important in case of data extrapolation for further clinical trials.
References:
[1] Gueorguieva, I., Cleverly, A. L., Stauber, A., Pillay, N. S., Rodon, J. A., Miles, C. P., Yingling, J. M. and Lahn, M. M. (2014). Defining a therapeutic window for the novel TGF‐β inhibitor LY2157299 monohydrate based on a pharmacokinetic/pharmacodynamic model. British Journal of Clinical Pharmacology, 77: 796 – 807.
[2] Lestini G., Dumont C., Mentré F. (2015). Influence of the size of cohorts in adaptive design for nonlinear mixed effects models: An evaluation by simulation for a pharmacokinetic and pharmacodynamic model for a biomarker in oncology. Pharmaceutical Research, 32(10): 3159 – 3169.
Acknowledgments: This research has received funding from the European Union’s Framework Programme for research, technological development and demonstration under grant agreement no 602144. This work was part of the InSPiRe project but does not necessarily represent the view of all InSPiRe partners.
Reference: PAGE 25 () Abstr 5862 [www.page-meeting.org/?abstract=5862]
Poster: Methodology - Other topics