Leirens Q, Valenzuela B, Lindauer A, Faelens R, Olsson P, Roeshammar D
SGS Exprimo, Mechelen, Belgium
Introduction: The main challenge during oncology drug development in the paediatric population is to improve the clinical development process making it faster to have access to safer and more effective treatments. Dose-finding studies in paediatrics have to be designed to avoid a large number of patients being treated with ineffective doses and, at the same time, to avoid overdosed patients. Regulatory agencies advocate the use of modelling and simulation of the available adult data to guide paediatric clinical trial designs [1,2]. One of the most intuitive and user-friendly software for performing clinical trial simulations is Simulo [3], a Java-based application running on an R backend with a graphical user interface. In Simulo it is easy to define the design of a paediatric clinical study and simulate using model parameters derived from adults and the body weight and age distribution for children.
Objectives: To show the capabilities of Simulo for simulating a paediatric phase I oncology clinical trial with the aim to predict a dose level that likely provides a similar exposure range as in adults. For this illustrative clinical trial simulation example, we selected to predict the potential outcome of an ongoing study [4] with trametinib, a MEK inhibitor approved in adult population for the treatment of V600E or V600K BRAF mutation-positive melanoma.
Methods: A previously developed population pharmacokinetic model for trametinib in adult patients [5] was implemented in Simulo. Simulations of the paediatric population were performed using the adult pharmacokinetic model, allometric scaling, and paediatric body weights sampled from the NHANES database [6], corresponding to children from 2 to 17 years. Three dose levels were simulated (0.0125 mg/kg, 0.025 mg/kg, and 0.040 mg/kg) in accordance with the ongoing clinical trial [6]. A total of 18000 different paediatric patients were simulated in order to generate 1000 virtual clinical trials, consisting of 6 patients per clinical trial and dose level. The percentage of “successful” trials under each dose level was defined as the fraction of 1000 virtual trials with exposure related parameters (AUC0-24,ss, Cmin,ss, and Cmax,ss) within the corresponding adult geometric mean value ± 20%. Moreover, the 90% prediction interval (PI) of adult exposures after 2 mg daily was generated by simulating 6000 profiles using the adult population pharmacokinetic model. The percentage of virtual paediatric clinicals trials where 6 out 6 patients were within the corresponding interval was then calculated.
Results: At the lowest dose of 0.0125 mg/kg, 100% of the simulated clinical trials were below the exposure targets for all three metrics. With the intermediate dose level, 23% of the simulated clinical trials had a Cmin,ss≥10 ng/mL (trough plasma concentration that has been found to be associated with progression free survival in an adult exposure-response analysis [5]), and none of the virtual trials were above the upper limit for the corresponding Cmax,ss in adults. For the highest dose level (0.040 mg/kg), an AUC0-24,ss above the upper limit was found in 42% of the virtual trials and more than 44% would have Cmin,ss≥14 ng/mL. For the intermediate dose only 29 %, 31% and 45% of virtual paediatric trials showed that 6 out of 6 patients were within the 90% PI of the adult AUC0-24,ss, Cmax,ss and Cmin,ss reference ranges, respectively.
Conclusions: Application of modelling and simulation using the previous knowledge from adult populations can help to explore different scenarios before enrolling paediatric patients into phase I clinical trials in order to avoid treatment with ineffective dose levels and to define the most convenient dose in a faster way. Other parameters like the number of children to be enrolled, different dosing strategies (flat dosing or tier-based dosing), can be also easily simulated using Simulo.
References:
[1] European Medicines Agency. Guideline on the role of pharmacokinetics in the development of medicinal products in the paediatric population. 2008. p. 1e8.
[2] U.S. Food and Drug Administration (FDA). Draft guidance for industry. General clinical pharmacology considerations for pediatric studies for drugs and biological products. 2014.
[3] Leirens Q, Faelens R, Gisleskog PO, et al. PAGE 26 2017. Abstr 7331 [www.page-meeting.org/?abstract=7331].
[4] Study to investigate safety, pharmacokinetic (PK), pharmacodynamic (PD) and clinical activity of trametinib in subjects with cancer or plexiform neurofibromas and trametinib in combination with dabrafenib in subjects with cancers harboring V600 mutations. https://clinicaltrials.gov/ct2/show/NCT02124772
[5] Ouellet D, Kassir N, Chiu J, et al. Population pharmacokinetics and exposure-response of trametinib, a MEK inhibitor, in patients with BRAF V600 mutation-positive melanoma. Cancer Chemother Pharmacol (2016) 77(4): 807-17.
[6] Patel D, Dombrowsky E, Barrett JS. A SAS-based solution for the extraction of population demographic priors from the CDC’s NHANES survey database: application in the creation of datasets for modeling and simulation projects. J Clin Pharmacol (2010) 50(9): 1085.
Reference: PAGE 27 (2018) Abstr 8708 [www.page-meeting.org/?abstract=8708]
Poster: Methodology - Study Design