2014 - Alicante - Spain

PAGE 2014: Drug/Disease modeling - Oncology
Emilie Schindler

PKPD-Modeling of individual lesion maximal standardized uptake value (SUV) in Gastro-Intestinal Stromal Tumors (GIST) patients treated with sunitinib

Emilie Schindler (1), Michael Amantea (2), Mats O. Karlsson (1), Lena E. Friberg (1)

(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden; (2) Pfizer Global Research and Development

Objectives: Tumor glucose metabolism - determined by the maximal standardized uptake value (SUV) assessed by [18F]-fluorodeoxyglucose positron emission tomography - has been suggested as an alternative to tumor size to assess early tumor response to therapy in cases where clinical benefit is observed but a change in tumor size is limited or delayed, e.g. for cytostatic drugs [1]. This analysis aims to investigate potential relationships between sunitinib exposure and individual lesion SUV time-course and to characterize both inter-individual (IIV) and inter-lesion variability (ILV) in SUV response in gastro-intestinal stromal tumors (GIST) treated with sunitinib, a multi-targeted tyrosine kinase inhibitor.

Methods: Baseline and post-baseline SUV data (n=607) were available for 172 lesions from 66 patients followed for a median time of 10 weeks of treatment with three different oral doses of sunitinib under three different treatment schedules [2]. Indirect response (IDR) models with inhibition of the production or stimulation of the loss of response with linear, power and Emax drug-effect relationships were investigated to describe SUV time-course. IIV and ILV were allowed in all model parameters. ILV was implemented in NONMEM in a manner similar to inter-occasion variability. The individual lesions’ SUV that had been assessed from the same FDG-PET scan were allowed to have different residual error values, but were assumed to be sampled from the same variability distribution and correlated.

Results: Log-transformed SUV data were well characterized by an IDR model with stimulation of the loss of response through a linear drug effect model driven by daily AUC. The model predicted a typical decrease in SUV of 60% after 14 days of sunitinib treatment (50 mg qd). A linear disease progression was included and predicted a typical increase in SUV of 14% per year. The estimated IIV was larger than the estimated ILV for both SUV baseline (33% CV for IIV vs 23% CV for ILV) and the drug effect parameter (59% CV for IIV vs 49%CV for ILV). The typical doubling time of SUV for return to baseline during off-treatment periods was 2 weeks.

Conclusions: Significant IIV and ILV in SUV response could be identified in the developed lesion model. VPCs illustrated the capability of the model to predict the drug effect on individual lesion SUV, as well as the sum of SUV at each time point. The predictive ability of individual lesion SUV on overall survival is under investigation.

Acknowledgements: This work was supported by the DDMoRe (www.ddmore.eu) project.

[1] Wahl, R.L., et al., From RECIST to PERCIST: Evolving Considerations for PET response criteria in solid tumors. J Nucl Med, 2009. 50 Suppl 1: p. 122S-50S
[2] Demetri, G.D., et al., Molecular target modulation, imaging, and clinical evaluation of gastrointestinal stromal tumor patients treated with sunitinib malate after imatinib failure. Clin Cancer Res, 2009. 15(18): p. 5902-9.

Reference: PAGE 23 (2014) Abstr 3210 [www.page-meeting.org/?abstract=3210]
Poster: Drug/Disease modeling - Oncology
Click to open PDF poster/presentation (click to open)