Premium
SU‐FF‐T‐489: Simulating the Effects of Anti‐Angiogenic Therapy Using Voxel‐Based Biological Parameters Derived From Functional Imaging
Author(s) -
Titz B,
Adhikarla V,
Kozak K,
Jeraj R
Publication year - 2009
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3181987
Subject(s) - bevacizumab , tumor hypoxia , radiation therapy , hypoxia (environmental) , medicine , imaging biomarker , voxel , nuclear medicine , perfusion , angiogenesis , antiangiogenic therapy , cancer research , chemotherapy , radiology , chemistry , magnetic resonance imaging , organic chemistry , oxygen
Purpose: Although anti‐angiogenic agents are being used in combination with standard radiotherapy or radiochemotherapy, the biological response mechanisms are currently not well understood. This study investigates therapeutic response to bevacizumab, a monoclonal antibody to VEGF‐A, using a functional imaging‐based tumor simulation model. Method and Materials: A multiscale tumor growth and radiation therapy response model based upon patient‐specific parameters was extended by implementing response mechanisms to bevacizumab. PET/CT images of cellular proliferation ([18 F ]FLT) and hypoxia ([61 Cu ]Cu‐ATSM) in human HNSCC tumors served as simulation input. Treatment with bevacizumab according to a clinical treatment regimen was simulated as an example of anti‐angiogenic therapy. Heterogeneous drug delivery within the tumors was simulated using a four‐compartment pharmacokinetic model coupled with perfusion parameters assessed by DCE‐CT. Within each voxel, response was modeled based on changes in VEGF‐A levels, simulated changes in microvessel density (MVD), and resulting changes in oxygenation levels. Results were compared to follow‐up PET/CT scans of hypoxia and proliferation. Results: Simulating response to a single bevacizumab dose of 10 mg/kg yielded decreased [18 F ]FLT uptake values as observed in all follow‐up scans. The mean correlation coefficient between simulated and follow‐up [18 F ]FLT scans was 0.79. However, trends in [61 Cu ]Cu‐ATSM uptake values could not be reproduced in all simulation cases, indicating that response time, as well as quantitative changes in hypoxia might depend on additional patient‐specific biological parameters. Simulated MVDs were found to be highly time‐dependent. Conclusion: The implemented model successfully reproduced trends in proliferative response following bevacizumab monotherapy and can be readily integrated into a previously developed radiation response model. Results also suggest that MVD should be used carefully when characterizing treatment response to anti‐angiogenic therapy due to a strong time‐dependence. Upon successful validation, the implemented model could serve as a prospective tool to help identify personalized dosing and sequencing of combination therapies.