Jonathan Chauvin (1), Geraldine Ayral (1), Pauline Traynard (1)
(1) Lixoft, Antony, France
Objectives: Therapeutic drug monitoring (TMD) is increasingly advocated in clinical practice. Most of the time, target concentration windows are established using retrospective studies [1]. To better assess the value of TMD, clinical trials are required. Simulation of clinical trials including TMD and dose adaptation can greatly help to better design those trials and maximize their probability of success.
However, the implementation of such simulation with standard simulation tools can be cumbersome, as it requires the user to run a first simulation, analyze the results and decide for a new dose, run a simulation again, etc. The procedure can be further complicated if the simulation and analysis are done with different applications.
We propose a doseAdaptation() function as an extension of the simulx function from the mlxR R package (http://simulx.webpopix.org/) for simulation (which is part of the MonolixSuite) to automatize the simulation of therapeutic drug monitoring and dose adaptation.
Methods: The doseAdaptation() function works as an extension of the simulx function, which permits to easily simulate clinical trials.
In addition to the usual simulx arguments needed to define a simulation (model, population parameters, number of individuals, covariates, outputs), the doseAdaptation() function requires the definition of an initial treatment and one or several adjustment rules:
- Initial treatment:
- Dose
- Inter-dose interval
- Rule(s):
- The model variable on which the rule applies (for instance the drug concentration)
- The type of the model variable (continuous or event)
- The condition that should be checked (for instance the comparison of the measured concentration to a threshold)
- The time point(s) at which the measurements are made and the condition is checked
- How to adapt the dose if the condition is not met: by increasing/reducing the dose by a certain factor (multiplicative) or a certain value (additive)
- The factor or value by which the dose should be increased or reduced
Results: The use of the doseAdaptation() function is shown on two examples.
In the first example, we simulate a clinical trial with dose-adaptation of everolimus. Everolimus is a promising candidate for TMD [2] as there is evidence for the relationship between exposure, safety and efficacy [3,4,5]. We reuse published models for the pharmacokinetics and exposure-toxicity relationship to develop a joint PK-TTE model taking into account the pharmacokinetics and the adverse events whose hazard is related to the exposure. The model is used to simulate a clinical trial with two arms: one where dose adaptation (reduction) is only done if toxicity events appear, and one where dose adaptation is performed following therapeutic drug monitoring and adverse events. Using simulation replicates, the power of the study can be estimated for different numbers of patients per arm.
In the second example, we compare the sample size needed to achieve a given power for dose-controlled trials (arms are defined by a given dose) versus concentration-controlled trials (arms are defined by a given plasma concentration). As effect is usually related to the concentration or exposure, concentration-controlled trials are expected to have a higher power [6]. We investigate this hypothesis using a typical PK/PD model with Emax response.
Conclusion: The doseAdaptation() function allows to easily simulate dose adaptation following therapeutic drug monitoring by defining adaptation rules for continuous variables or events in a flexible and efficient way. This allows to explore “what if” questions or plan the design of clinical trials.
References:
[1] Lankheet, N. A. G., Desar, I. M. E., Mulder, S. F., Burger, D. M., Kweekel, D. M., van Herpen, C. M. L., … van Erp, N. P. (2017). Optimizing the dose in cancer patients treated with imatinib, sunitinib and pazopanib. British Journal of Clinical Pharmacology, 83(10), 2195–2204.
[2] Verheijen, R. B., Yu, H., Schellens, J. H. M., Beijnen, J. H., Steeghs, N., & Huitema, A. D. R. (2017). Practical Recommendations for Therapeutic Drug Monitoring of Kinase Inhibitors in Oncology. Clinical Pharmacology & Therapeutics, 102(5).
[3] Ravaud, A., Urva, S. R., Grosch, K., Cheung, W. K., Anak, O., & Sellami, D. B. (2014). Relationship between everolimus exposure and safety and efficacy : Meta-analysis of clinical trials in oncology. European Journal of Cancer, 50(3), 486–495.
[4] de Wit, D., Schneider, T. C., Moes, D. J. A. R., Roozen, C. F. M., den Hartigh, J., Gelderblom, H., … van Erp, N. P. (2016). Everolimus pharmacokinetics and its exposure – toxicity relationship in patients with thyroid cancer. Cancer Chemotherapy and Pharmacology, 78(1), 63–71.
[5] Deppenweiler, A. M., Falkowski, S., Monchaud, C., Picard, N., Laroche, M., Tubiana-mathieu, N., … Picard, N. (2017). Towards therapeutic drug monitoring of everolimus in cancer ? Results of an exploratory study of exposure-effect relationship . Pharmacological Research, 121, 138–144.
[6] Kraiczi, H., Jang, T., Ludden, T., & Peck, C. (2003). Randomized concentration-controlled trials: Motivations, use, and limitations. Clinical Pharmacology & Therapeutics, 74(3), 203–214.
Reference: PAGE 28 (2019) Abstr 9058 [www.page-meeting.org/?abstract=9058]
Poster: Clinical Applications