NIMG-27. EVALUATING THE POTENTIAL OF 18F-FLUCICLOVINE POSITRON EMISSION TOMOGRAPHY TO DETECT WHOLE BRAIN TUMOR BURDEN: A PRECLINICAL STUDY

Abstract

Abstract BACKGROUND Accurately identifying the boundary of invasive brain tumors is critical for planning effective treatment. Conventional imaging includes T1-weighted contrast-enhanced MRI to identify regions where tumor angiogenesis has broken down the blood-brain barrier. However, conventional MRI has limited sensitivity for identifying invading tumor cells. Amino acid PET with 18F-fluciclovine may offer improved sensitivity. We investigate whether 18F-fluciclovine PET provides a superior measure of tumor burden than MRI. METHODS Rats were implanted with patient-derived glioblastoma xenografts that had been transduced to enable fluorescent imaging of the TdTomato protein. When the tumors grew to a sufficient size, the rats were scanned with 18F-fluciclovine PET followed immediately by contrast-enhanced MRI. Rats were sacrificed and whole brains were removed. Western blots were performed to measure amino acid transporter levels. Fluorescent imaging of optically cleared brain slices enabled visualization of the labelled tumor and determination of whole tumor burden. A methodology for registering in vivo PET/MRI slices to ex vivo fluorescent images of optically cleared brain slices (1-mm thickness) has been developed. Tumor volumes are manually segmented on all images. RESULTS The ratio of tumor to normal brain uptake (TNB) of 18F-fluciclovine peaked ten minutes post-injection (mean value 7.1) and was followed by a continual decrease in the TNB ratio (mean value 4.6 at fifty minutes post-injection). The mean percentage of MRI segmented tumor volume that was inside the PET segmented tumor volume was 95%. The mean percentage of PET segmented tumor volume that was inside the MRI segmented tumor volume was 40%. Tumors with the highest level of amino acid transporters had the highest 18F-fluciclovine uptake. CONCLUSIONS The majority of 18F-fluciclovine PET identified tumor regions were not identified using conventional MRI, indicating the potential of this approach to detect invasive tumor cells and provide a more robust assessment of whole tumor burden.