Early Prediction of Radiation Response of Brain Metastases with [ 18F]-ML-10: A Novel Molecular PET Imaging Agent for Apoptosis

Abstract

A. M. Allen, A. Shirvan, E. Mishani, A. Steinmetz, A. Reshef, I. Ziv, E. Fenig Davidoff Center, Petach Tikva, Israel, Aposense LTD, Petach Tikva, Israel, Hadassah Medical School, Jerusalem, Israel, Aposense LTD., Petach Tikva, Israel, Aposense, Petach Tikva, Israel Purpose/Objective(s): Radiographic response to radiation therapy often lags 4-6 weeks behind the biological effect. The [F]ML-10 is a novel PET tracer that has been shown in preclinical models to effectively image an early apoptotic response to radiation therapy. We sought to test this imaging technique in a prospective study of patients undergoing WBRT for brain metastasis. Materials/Methods: Patients with brain metastasis were enrolled into a prospective Phase II trial of [F]-ML-10-PET scanning before and after WBRT. The study was approved by the Helsinki commission of the Rabin Medical Center. Enrollment was limited to patients with lesions .1.5 cm in diameter on pretreatment MRI scans. All patients received 30 Gy in 10 daily fractions following CT simulation and treatment planning. Prior to RT, each patient underwent [F]-ML-10-PET scan on GE Discovery STE scanner, with a 15.7 cm axial and 70 cm transaxial field of view (FOV). The PET voxel intensities (Bq/cm) were normalized to the superior sagittal sinus and were recorded. Following 9-10 fractions of radiation patients underwent a second [F]-ML-10 PET scan. Six to 8 weeks following the completion of therapy patients underwent a follow-up MRI scan to measure the change in tumor size. All PET and MRI scans were coregistered using a Normalized Mutual Information algorithm (version 2.9 build 2 by PMOD Technologies Ltd). A voxel based analysis was carried out to determine the change in PET voxel intensity in the target lesions from initial to follow-up PET scans (DPET) using a homegrown program written with MATLAB (version 7.4.0 by MathWorks). Response was assessed by MRI, based on the percentage of change in tumor size. The percent of ‘‘changed’’ voxels for each lesion was calculated and tested for correlation with the change in the anatomical size according to MRI.