3123 - Impact of Lung Tissue-Air Interface on PET Signal Statistics and BgRT Dose Delivery under Respiratory Motion
Presenter(s)
S. Sharma, T. Yeung, A. Groll, and A. Da Silva; RefleXion Medical, Inc., Hayward, CA
Purpose/Objective(s): Biology-Guided radiotherapy (BgRT) uses real-time FDG signal for target localization and delivery, which may be affected by lung tissue-air interfaces due to heterogeneous attenuation. Two controlled phantoms were developed to simulate (1) an intraparenchymal tumor fully surrounded by aerated lung and (2) a lung-based tumor with partial air/tissue adjacency. These configurations generated distinct PET signal profiles: higher activity concentration (AC) and normalized target signal (NTS) in the fully aerated condition and reduced NTS in the interface condition, due to increased background variability. The objective was to assess whether BgRT with a new system under development† maintains robust planning, tracking, and dose delivery under these interface-driven PET distributions with matched respiratory motion.
Materials/Methods: An anthropomorphic motion phantom consisting of a 28.1 L water tank mixed with an 18F-FDG background that mimicked lung uptake (8 kBq/mL) was used.
Two 26 mm FDG fillable targets were designed to mimic: (1) a fully intraparenchymal tumor surrounded by aerated lung-equivalent material and (2) a tissue-interface tumor with ~50% air and 50% tissue adjacency. Each target comprised two hemispheres sandwiching a radiochromatic X-ray film in the coronal plane. Perforated styrofoam inserts in water simulated lung tissue–air heterogeneity.
Targets were filled to a tumor-to-background ratio (TBR) of 8:1. The intraparenchymal configuration yielded AC of 21.25 kBq/mL and NTS of 14.47. The tissue-air interface (2nd) configuration yielded AC 18.33 kBq/mL and a reduced NTS of 4.52. A C-shaped cardiac organ-at-risk (OAR) with coronal film was evaluated for dose agreement with respect to the planned dose. Identical 3D respiratory motion was applied during experiments.
Results: Plan conformality was comparable between configurations, with a conformity index (CI) of 1.03 for both cases. Target coverage was 95% for the case mimicking a fully intraparenchymal (air-surrounded) target and 95.1% for the case mimicking the tissue-interface configuration. Treatment time was approximately 17 minutes for both deliveries.
Film-based analysis demonstrated a PTV radial margin loss of 1.2 mm for the intraparenchymal case, with 3%/3 mm gamma agreement of 93.4% relative to the treatment plan. For the tissue-interface case, margin loss was 1.7 mm, and the 3%/3 mm gamma agreement was 97.6%. The 2D dose profiles showed good agreement with the planned dose on both the target and OAR films for both cases.
Conclusion: Despite reduced NTS in the tissue-interface condition, BgRT on the new system under development achieved comparable conformality, coverage, and delivery accuracy in both cases. Film verification confirmed consistent dose agreement and minimal margin loss. These results demonstrate BgRT’s robustness to interface-driven FDG signal variability in heterogeneous lung environments.
† New BgRT system requires regulatory clearance and is not available for sale.