Dosimetric Consequences of 3D Versus 4D PET/CT for Target Delineation of Lung Stereotactic Radiotherapy

Abstract

Lung tumor delineation is frequently performed using 3D positron emission tomography (PET)/computed tomography (CT), particularly in the radiotherapy treatment planning position, by generating an internal target volume (ITV) from the slow acquisition PET. We investigate the dosimetric consequences of stereotactic ablative body radiotherapy (SABR) planning on 3D PET/CT in comparison with gated (4D) PET/CT. In a prospective clinical trial, patients with lung metastases were prescribed 26 Gy single-fraction SABR to the covering isodose. Contemporaneous 3D PET/CT and 4D PET/CT was performed in the same patient position. An ITV was generated from each data set, with the planning target volume (PTV) being a 5-mm isotropic expansion. Dosimetric parameters from the SABR plan derived using the 3D volumes were evaluated against the same plan applied to 4D volumes. Ten lung targets were evaluated. All 3D plans were successfully optimized to cover 99% of the PTV by the 26 Gy prescription. In all cases, the calculated dose delivered to the 4D target was less than the expected dose to the PTV based on 3D planning. Coverage of the 4D-PTV by the prescription isodose ranged from 74.48% to 98.58% (mean of 90.05%). The minimum dose to the 4D-ITV derived by the 3D treatment plan (mean = 93.11%) was significantly lower than the expected dose to ITV based on 3D PET/CT calculation (mean = 111.28%), p < 0.01. In all but one case, the planned prescription dose did not cover the 4D-PET/CT derived ITV. Target delineation using 3D PET/CT without additional respiratory compensation techniques results in significant target underdosing in the context of SABR.