Clinical experience with a PET/CT scanner to account for intrafraction motion in radiation oncology

Abstract

1656

Objectives: To report on the feasibility of respiratory-gated lung radiotherapy with dependence on 4DPET and 4DCT for treatment planning rather than simulation with conventional CT.

Methods: Respiratory-correlated data from both a PET/CT system (Siemens Biograph 64) and a Philips CT system were acquired. Gated acquisition was performed with the Anzai and the Varian RPM systems, respectively. Correlation was measured between 4DCT and 4DPET, in phantom, based on respiratory phase by calculating the deviation from the mean tumor position for each phase relative to a fixed position (similar to bony anatomy). The accuracy of each external surrogate was also tested in phantom (Fig 1). Treatment planning and gated delivery were performed using two treatment planning systems in conjunction with a Varian linear accelerator. An end-to-end test for gated radiotherapy was performed and has shown reliable delivery of the intended plan developed from 4DPET/CT imaging. Five lung cancer patients have received respiratory-gated treatment using this methodology.

Results: When comparing between phases, less than 5% uncertainty in the phantom study provided good correlation between the 4DCT and 4DPET data. Insignificant differences were measured between the surrogate systems used for gated acquisition. Improvements of dose distributions on the moving targets improved with gated delivery. Localization uncertainty of the moving target was less than 3 mm.

Conclusions: We conclude that 4DPET/CT studies yield accurate respiratory-correlated information (spatial and temporal) of the tumor region. Also proven is the integration of a Siemens PET/CT system with BrainLab and Varian systems for gated radiotherapy.

Research Support: M.D. Anderson Cancer Center Orlando receives research funding from Siemens Medical Solutions, Inc.