Session Item

Purpose: In modern radiotherapy, different technologies coexist within the same department. This is not only in relation to the delivery system but also the treatment planning system (TPS) or dedicated software employed in the clinical workflow for specific tasks. Often, it is necessary to transfer a patient’s DICOM dataset from one system to the other for several reasons, such as re-treatment, plan summation, image registration or DVH dose prediction. The aim of the study is to verify and quantify the effect of DICOM RT dataset transfer between two TPSs.

Materials and methods: Twenty-five patients treated in a 0.35T MR-Linac (MRidian, Viewray Inc.) (VR) for locally advanced pancreatic cancer (LAPC), were enrolled. Patients received 40Gy or 50Gy in five fractions, prescribed to the 80% isodose line or to target mean. For each patient, a nominal dose distribution was optimized on the planning MRI, fused with a planning CT acquired within the same day. Each fraction was daily adapted if needed: clinical and planning target volume were daily re-contoured, as well as OARs. Plan re-optimization was performed to shape dose distribution on the daily anatomy, considering the clinical OARs constraints. Each plan was then exported from the VR to another available TPS in the department dedicated to external beam radiotherapy (Eclipse, v15.6 Varian). A comparison between two planning systems was performed considering the PTV and OARs sizes (cc), as well as coverages and clinical constraints respectively. The OARs considered were duodenum, stomach and bowel.

Results: From the twenty-five enrolled patients, 139 plans were included in the data comparison: 11 fractions were not adapted since the anatomical variation was not relevant in terms of dosimetric variation and therefore clinical constraints were within tolerance. Regarding the percentage PTV variation, the median values obtained by boxplot (figure 1) are about 10,8% for each fraction, while PTV coverages at the prescribed dose are similar. Boxplots of the percentage OARs variation was also performed and shown in figure 2A exclusively for duodenum. The median values are approximately 15,9% for duodenum, 6,9% for stomach and 10,3% for bowel for each fraction. The percentage variations of the dose constraints between the two TPS is depicted in figure 2B: the median values are about 41,2% for duodenum, 60,5% for stomach and 54% for bowel for each fraction.

Conclusions: This study has preliminary demonstrated a non-negligible variation, both in (cc) and dosimetric parameters, when DICOM datasets are transferred from two specific TPSs. Therefore, such variations should be clinically considered, especially for OARs close to the target. To this aim, current investigations are now focalized on the DICOM structure algorithm employed by the TPSs during the transfer, to improve the understanding of the main possible cause of such variations.