A Prospective Study of an Offline Matching Protocol for Head and Neck Cancer on Helical Tomotherapy: The F5 Study

Abstract

Daily online matching is the default standard for image guided intensity modulated radiation therapy on the helical tomotherapy platform, but requires more time and involves higher imaging dose. Following a hypothesis that carefully performed offline protocols like the No-action-level (NAL) protocol may be appropriate for head and neck cancers where appropriate immobilization is possible, we had first performed a retrospective audit on 100 head and neck cancer patients spanning 2,858 daily online images revealing that offline imaging may be performed safely if the PTV margin was 5 mm (Clin Oncol (R Coll Radiol). 2016;28(3):178-84). In this study, we aimed to prospectively verify these findings and determine the savings in time and imaging dose associated with offline imaging. A cohort of 62 head and neck cancer patients planned for radiation therapy with curative intent on helical tomotherapy were identified prospectively. Patients underwent daily online image verification for the first 5 fractions, and the mean shift in each translational axis was then implemented for all subsequent fractions prior to treatment. The residual errors (RE) with this method was verified and tabulated by daily imaging as part of study. These were used for a recursive verification PTV margin (PTVverif) calculation. For every patient, the time taken to do daily setup, verification imaging and correction and treatment delivery were measured. Imaging dose at central axis of a cheese phantom for a 15 cm scan length at different scan resolutions were measured. Anatomical primary sites comprised nasopharynx 10 (16.1%); oropharynx 7 (11.3%); oral cavity 20 (32.3%); larynx/hypopharynx 21 (33.9%) and others 4 (6.4%). The treatment intent was radical in 38 (61.3%) and adjuvant in 24 (38.7%). Data from a total of 1853 treatment fractions were analyzed. RE were small, with >5 mm deviations seen in 51 (3.3%), 25 (1.6%) and 83(5.4%) fractions in the X (lateral), Y (craniocaudal) and Z (anteroposterior) axes respectively. The systematic errors were 1.1, 0.8, and 1.3 mm, while the random errors were 2.1, 1.8, and 2.2 mm in these axes. Based on the RE, the PTVverif margins in these axes were 4.3, 3.3, and 4.7 mm, confirming that 5 mm PTV margins are safe with this offline protocol. The mean time for daily imaging verification imaging and correction was 4.3 mins corresponding to 32.3% of in-room time. The daily imaging dose was 3.2 cGy, 1.6 cGy and 1.1 cGy for fine, normal, and coarse imaging modes, corresponding to reduction of imaging dose by 64 cGy, 32 cGy, 22 cGy per treatment course (assuming imaging avoidance in 20 of 30 fractions). To our knowledge, this is the only prospective study verifying the application of offline imaging during head and neck cancer IMRT on helical tomotherapy. Our results suggest that with PTV margins of 5 mm, an offline F5 NAL protocol could be performed safely and could lead to substantial savings in time and imaging dose.