Abstract #1238

MRI Predictors of Response to Pembrolizumab, Bevacizumab and Hypofractionated Stereotactic Irradiation in Patients with Recurrent High Grade Gliomas

Samuel Hawkins¹, Olya Stringfield², Nicolas Rognin¹, John Arrington³, Michael Yu⁴, Heiko Enderling⁵, Solmaz Sahebjam⁶, and Natarajan Raghunand¹

¹Cancer Imaging & Metabolism, Moffitt Cancer Center, Tampa, FL, United States, ²Cancer Imaging & Metabolism, Image Response Assessment Team Core, Moffitt Cancer Center, Tampa, FL, United States, ³Radiology, Moffitt Cancer Center, Tampa, FL, United States, ⁴Radiation Oncology, Moffitt Cancer Center, Tampa, FL, United States, ⁵Integrated Mathematical Oncology, Moffitt Cancer Center, Tampa, FL, United States, ⁶Neuro-Oncology, Moffitt Cancer Center, Tampa, FL, United States

Standard MRI scans of patients with recurrent high grade gliomas treated with pembrolizumab were analyzed to build a model to predict time-to-progression. Images from five standard MRI sequences were co-registered across multiple scan dates per patient and automatically segmented into normal and pathologic tissue types based on calibrated pixel intensities. 308 radiomic features describing size, shape, and texture were extracted per image type. The four most predictive features were used in a linear regression model that could predict time-to-progression to within an average of three months of actual progression in test patients.

This abstract and the presentation materials are available to members only; a login is required.

Join Here