Achieving the relocation accuracy of stereotactic frame-based cranial radiotherapy in a three-point thermoplastic shell.

AIMS: To compare the accuracy of fractionated cranial radiotherapy in a standard three-point thermoplastic shell using daily online correction with accuracy in a Gill-Thomas-Cosman relocatable stereotactic frame. MATERIALS AND METHODS: All patients undergoing fractionated radiotherapy for benign intracranial tumours between March 2009 and August 2010 were included. Patients were immobilised in the frame with those unable to tolerate it immobilised in the shell. The ExacTrac imaging system was used for verification/correction. Daily online imaging before and after correction was carried out for shell patients and systematic and random population set-up errors calculated. These were compared with frame patients who underwent standard departmental imaging/correction with fractions 1-3 and weekly thereafter. Set-up margins were calculated from population errors. RESULTS: Systematic and random errors were 0.3-0.7 mm/° before correction and 0.1-0.2 mm/° after correction in all axes in the frame, and 0.6-1.5 mm/° before correction and 0.1-0.4 mm/° after correction in the shell. Isotropic margins required for patient set-up could be reduced from 2 mm to <1 mm in the frame and from 5 mm to <1 mm in the shell. CONCLUSION: Similar set-up accuracy can be achieved in the standard thermoplastic shell as in a relocatable frame despite less precise immobilisation. The use of daily online correction precludes the need for larger set-up margins.