2012 - Venice - Italy

PAGE 2012: Study Design
Alexandre Sostelly

Dose and dose schedule optimization of anticancer drugs

Alexandre Sostelly (1,2), Joakim Nyberg (2), Mats O. Karlsson (2), Andrew C. Hooker (2)

(1) EMR3738, Ciblage Thérapeutique en Oncologie, Faculté de Médecine et de Maïeutique Lyon-Sud Charles Mérieux, Université Claude Bernard Lyon 1, Lyon France; (2) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala Sweden

Objectives: Anticancer drug dose and dose schedule are known to affect the tumor response. Tumor growth inhibition (TGI) models establish the relationship between tumor growth dynamics and drug effects and can help to identify optimal dose schedules. Moreover, numerous factors, such as tumor resistance and dose-effect relationship, can affect drug effect and have to be accounted for to maximize the tumor response. We aim at optimizing anticancer dose and dose schedule using population optimal design methodology by taking into account both tumor resistance and the dose-effect relationship.

Methods: We used the TGI model from Claret et al.[1] applied to a theoretical phase II capecitabine study as a basis for this work. Optimization of dose and dose schedule was performed in PopED v.2.11 [2]. Optimizations maximized the change in tumor size from baseline after 2 treatment cycles (6 weeks) within clinical constraints using a penalty function. The criterion is affected by variability and the mean criterion was computed using 500 LHS samples taken from the parameter distributions. To obtain a clinically relevant dose schedule, we constrained the dose interval to 1 day and doses to be the same for each week of the cycle. The total dose was fixed to that used in clinical practice and daily doses were not allowed to exceed 5g.m-2.
Impact of tumor resistance has been evaluated by changing the resistance development rate value in the TGI model and impact of the dose-effect relationship (linear and Emax type) has been investigated by modifying the drug exposure parameter. To compare the optimized dose schedules, the mean dose time (MDT) was computed to reflect the dose density within the cycle.

Results and Discussion: In case of linear exposure-response, the optimal schedule frontloads doses in the 1st week of the cycle (MDT=4g.m-2.d-1) for any degree of tumor resistance development rate. In the case of Emax type exposure, the optimal dose schedule depends on the magnitude of resistance development rate. At maximal effect, the optimal schedule equally distributes doses in the cycle for low and medium degree of resistance (MDT=10.3g.m-2.d-1) whereas it frontloads doses for high degree of resistance (MDT=4.7 g.m-2.d-1). At minimal effect, the optimal schedule frontloads doses in the 1st week for any resistance degree (MDT=4g.m-2.d-1). Between maximal and minimal effect, MDT decreases when resistance degree increases. Our approach allows optimizing anticancer dose and dose schedule based on clinically relevant criterion and within clinical constraints.

References:
[1] Claret L, Girard P, Hoff PM, et al. Model-based prediction of phase III overall survival in colorectal cancer on the basis of phase II tumor dynamics. J Clin Oncol. 2009
[2] Nyberg J, Ueckert S, Karlsson MO, Hooker AC. PopED v2.11 2011 (http://poped.sourceforge.net)




Reference: PAGE 21 (2012) Abstr 2373 [www.page-meeting.org/?abstract=2373]
Poster: Study Design
Top