Pre-treatment positron emission tomography image quality and inter-fractional tumor uptake variation for bone metastasis during biology-guided radiotherapy.

e15060 Background: In biology-guided radiotherapy (BgRT) with real-time position emission tomography (PET) guidance, PET images are acquired on the BgRT machine both before the start of each treatment fraction and during beam delivery. However, pre-treatment PET scans could have lower image quality as one couch pass was used compared to functional modeling PET scans on the BgRT machine with four couch passes (four-pass PET scans). This study is aimed at evaluating pre-treatment PET scan image quality and inter-fractional tumor uptake variations of F-18 fluorodeoxyglucose (FDG)-avid bone lesions in a BgRT treatment course. Methods: To evaluate pre-treatment PET scan image quality, a water-filled cylindrical PET imaging phantom was used. A spherical compartment of the phantom was injected with high FDG activity while the rest of the phantom was injected with low FDG activity. On two separate dates, a pre-treatment PET scan was performed to the phantom on the BgRT machine followed by a four-pass PET scan for reference PET Imaging data. Pre-treatment PET scan images were compared to four-pass PET scan images in terms of maximum and mean standard uptake value (SUV max and SUV mean ) for high and low FDG activity compartments. For inter-fractional tumor uptake variation analysis, daily pre-treatment PET images for all patients who received multi-fractionated BgRT to bone mets at our institution were retrieved. Five patients were selected who received BgRT for a prescribed dose ranging from 24 to 30 Gy in two to three treatment fractions. Prior to treatment planning, each patient received a CT simulation and a PET scan on the BgRT machine for baseline FDG activity. Before each treatment fraction, a pre-treatment PET scan was performed. Tumor uptake variations were evaluated based on normalized SUV parameters. The metabolic tumor volume (MTV) was calculated as the volume inside the GTV with at least 50% SUVmax in the first-fraction pre-scan PET image. Results: In the PET imaging phantom analysis, SUVmax and SUVmean values for both the high and low FDG activity compartments in pre-treatment PET scan images showed < 5% deviation from those in four-pass PET scan images. In tumor uptake variation analysis, compared to pre-treatment PET images in the first fraction, subsequent pre-treatment PET images showed an average SUVmax variation of 0.75±1.64 (range: -0.42 – 3.47), average SUVmean variation of 0.3±0.68 (range: -0.19 – 1.38), and average MTV variation of -0.16±0.75 cm 3 (range: -1.4 – 0.6 cm 3 ). Compared to the first-fraction pre-treatment PET images, subsequent pre-treatment pet images showed maximum deviation of 43.6% in SUV max , 23.6% in SUV mean , and 8.1% in MTV. Conclusions: Based on the phantom imaging study, pre-treatment PET images showed adequate agreement with four-pass PET scan images. Large variations ( > 20%) in tumor FDG uptake were observed in the BgRT course for bone lesions, indicating the need for adaptive radiotherapy based on daily PET imaging results.