2422 - Evaluation of BgRT Delivery Accuracy for Multiple Small Targets of Varying Sizes

M. Surucu¹, N. Kovalchuk¹, B. Han¹, G. Bal², R. Yang², A. Maniyedath², A. M. Ndlovu³, and R. T. Forbang³; ¹Department of Radiation Oncology, Stanford University, Stanford, CA, ²RefleXion Medical, Inc., Hayward, CA, ³Hackensack University Medical Center, Hackensack, NJ

Purpose/Objective(s): Biology guided radiotherapy (BgRT) on the medical technology platform promise to fulfill a unique need for treating multiple metastatic tumors guided by the PET signal emitted from each target using a single injection. We investigate the feasibility of sequentially treating five small targets of varying diameters, in the same session, using one FDG fill. Materials/

Methods: A custom patient-specific quality assurance (PSQA) insert containing five spherical targets of diameters 8, 9, 11, 13, 16 mm were each filled with FDG at a target-to-background (T:B) ratio of 20:1. The Modeling PET used to optimize the BgRT plan was acquired on medical technology within 25 minutes of filling the phantom. For treatment planning, the PTV was a 5 mm expansion of the PET-avid GTV, while the biology tracking zone (BTZ) was a 5 mm expansion of the PTV. Plan optimization and delivery to all five targets were completed on the same day. For each target, a 50 Gy in 5 fractions BgRT plan was optimized and were treated sequentially from small to large targets. During delivery, to test the effect of FDG decay on BgRT delivery, an injection to treatment time of 80 minutes is used for the 8 mm target while 230 minutes was used for the last target (16 mm). Activity concentration (AC, kBq/ml) and Normalized Target Signal (NTS) is measured for each target during modeling PET session and right before the BgRT delivery (PreScan PET). Margin loss and gain are defined as the maximum distance in mm between the planned and delivered BgRT 97% isodose lines. The absolute-dose gamma passing rates using 3%/2mm and 3%/3 mm criteria were used to evaluate the BgRT delivered doses captured using the patient-specific quality assurance (PSQA).
Results: All BgRT plans were successfully planned and delivered. For the five targets (in small to large order), the AC was 13.0, 20.0, 30.6, 42.4 and 56.8 during the PET modeling, and 10.4, 10.0, 12.4, 15.1 and 14.8 during PreScan PET preceding each BgRT plan delivery. The NTS for the five targets during PET modeling was 15.4, 25.9, 30.3, 26.2 and 31.4 while during PreScan PET was 10.6, 10.4, 14.0, 20.0 and 17.8. The patient-specific quality assurance (PSQA) pass rates at 3%/2mm were 88.3%, 98.0%, 95.5%, 99.2%, and 99.7% respectively, while 3%/3mm yielded 90.9%, 99.7%, 99.8%, 100% and 100%. The margin loss was 0, 1.0, 1.7, 1.6 and 1.5 mm, while the margin gain was 3.3, 0.5, 1.7, 0.5 and 1.3 mm.
Conclusion: As expected, despite the same T:B fill concentration of 20:1, the detected PET AC was inversely proportional to target diameter in the PET modeling, where all targets were imaged at once. The lower PET AC during treatment compared to modeling PET was due to the FDG radioactive decay (half-life: 109 minutes) but was greater than 5kBq/ml AC limit for each target. Despite the lower PET counts during delivery, dosimetric results of these static BgRT deliveries were in good agreement for targets larger than 9 mm in this experiment. More tests will be performed to establish BgRT delivery accuracy for small targets both in static and motion deliveries, using varying T:B ratios.