Purpose: To assess inter-institution variability of treated volumes in preoperative radiotherapy for rectal cancer among Austrian radiotherapy institutions in the framework of a multi-centre phase-III clinical trial.
Materials and methods: All eleven Austrian radiotherapy departments were invited to participate in this pre-study dummy-run. They received a short history of two 'dummy patients' (case A and case B); three computer assisted tomography (CT) slices; simulation films; and the protocol describing the radiation technique to be used. Participants were asked to prepare a treatment plan for either case on the basis of the materials provided and to use their computerized planning systems. Additionally and independently of the CT-based treatment plans, they were asked to delineate the fields to be treated on the simulation films.
Results: Nine of eleven departments participated. All participants used a three or four field technique as requested. The variation of beam widths and planning target volumes (PTVs) in the central plane was 6-11% and 11-16%, respectively. The standard deviations (SD) were 21 and 24% for the two cases for mean treated volumes of 2.1 and 2.9 l, respectively. The variation of beam widths in the central plane was less pronounced in the simulation based treatment plans as compared with the CT-based treatment plans for the dorsal fields, the opposite was true for the laterals.
Conclusion: Considerable variation of treated volumes is inevitable in multi-institution trials despite detailed treatment guidelines. Simulator based treatment fields seem to result in less pronounced inter-institution variations compared with CT-based treatment planning, if bony landmarks can be used as is the case in rectal cancer. Continuous quality control is thus warranted in multi-centre trials to increase homogeneity of volumes treated.