Abstract #4207

Hierarchical Versus Voxel-Wise Models for DCE-MRI in a Head and Neck Study with Lapatinib

Brandon Whitcher¹, Volker J. Schmid², David Collins³, Matthew Orton³, Dow-Mu Koh^3,4, Josep M. del Campo⁵, Kevin Harrington⁶, Iman A. El-Hariry⁷

¹Clinical Imaging Centre, GlaxoSmithKline, London, UK; ²Institute of Biomedical Engineering, Imperial College, London, UK; ³CRUK Clinical Magnetic Resonance Research Group, Institute of Cancer Research, Sutton, Surrey, UK; ⁴Department of Radiology, Royal Marsden Hospital, Sutton, Surrey, UK; ⁵Department of Medical Oncology, Vall d'Hebron Hospital, Barcelona, Spain; ⁶Targeted Therapy Team, Institute of Cancer Research, London, UK; ⁷Oncology Medicine Development Centre, GlaxoSmithKline, London, UK

We compare the results from a quantitative analysis of DCE-MRI data using summary statistics from a non-linear regression analysis, using both optimization and Bayesian methods, and the output from a Bayesian hierarchical model in a phase II study of lapatinib in patients with locally advanced squamous cell carcinoma of the head and neck.