Influence of setup errors on spinal cord dose and treatment plan quality for cervical spine tumours: a phantom study for photon IMRT and heavy charged particle radiotherapy

Publisher Website
Google Scholar
Abstract
Tumours partly surrounding the cervical spine may be treated by conformal radiotherapy (RT) using intensity-modulated RT (IMRT) with photons or heavy charged particle RT. For both, a high setup accuracy is required to spare the radiosensitive spinal cord, if a high dose is to be delivered. A phantom study was performed to determine the variation of the dose to the spinal cord surface by predefined setup errors. The measured doses were compared to those calculated by the treatment planning programme. The influence of systematic setup errors on characteristic parameters of the treatment plan quality was quantified. The largest variation of the mean and maximum doses to the spinal cord due to setup errors was significantly larger for carbon ions than for IMRT (mean: 11.9% versus 3.9%, max: 29.2% versus 10.8% of the prescribed dose). For the comparison of measured and calculated doses, mean deviations of 3% (IMRT) and 6% (carbon ions) of the prescribed dose were obtained. These deviations have to be considered, when the spinal cord dose is assessed from the treatment plan and they may also influence the dose prescription. Carbon ions yield better values for coverage (99.9% versus 93.1%) and conformality (110% versus 126%) of the PTV as compared to IMRT, while the spinal cord is better spared. Dose distributions produced with carbon ions, however, are more sensitive to setup errors, which have to be considered during treatment.