Implementation of volumetric arc therapy for head and neck cancer

In volumetric modulated arc therapy (VMAT), treatment is delivered using a continuous arc motion of the gantry with simultaneous variation of mlc position, gantry speed and dose rate. Several manufacturers offer VMAT facilities based on conventional accelerator gantries while others offer tomographic therapy based on ring gantries. A number of centres have begun clinical VMAT treatments in recent years. The RapidArc solution for VMAT was introduced for clinical use by Varian Medical Systems, Inc in April 2008, with the first patient treated at Rigshospitalet, Copenhagen University Hospital, in May 2008. In May 2010, RapidArc was installed retrospectively on two Clinac 21iX linear accelerators at the Beatson West of Scotland Cancer Centre (BWoSCC), Glasgow. The first head and neck VMAT patient at the BWoSCC was treated in June 2010, just four weeks after the VMAT hardware upgrade on the linear accelerators. The following article describes our experience in commissioning and introducing head and neck VMAT to BWoSCC. Commissioning VMAT As has become common practice in new RapidArc installations, we engaged the help of consultancy company Medical Physics Services (MPS) for commissioning. MPS was on site for one week and assisted BWoSCC physicists in the tests required to commission VMAT on the linear accelerator and to commission the Eclipse treatment planning system for VMAT planning in line with previous published commissioning schedules. MPS also advised on establishing a patient specific pre-treatment quality assurance programme. Commissioning tests on the linear accelerator confirm the dose delivery accuracy and mlc positional accuracy with varying gantry speed and dose rate. The delivery of dose was evaluated by irradiating seven regions of equal dose using seven different combinations of dose rate, gantry position and gantry speed. The dose for each strip was within 2% of the average dose. The mlc leaf speed was confirmed by using a dynamic mlc (dmlc) test pattern that utilises four different combinations of leaf speed and dose rate to deliver four areas of equivalent dose. Again, the dose for each strip was within 2% of the average dose. The mlc positional accuracy was evaluated using a ‘picket fence’ dmlc test pattern. Positional errors can be detected visually on the film that is irradiated with the test pattern. The dose delivery and mlc position at different gantry angles was also evaluated using different mlc test patterns. VMAT planning calculations are performed with the Varian Eclipse system using a multi-stage inverse planning optimisation algorithm, known as the Progressive Resolution Optimisation Algorithm, or PRO. The accuracy of dose calculations depends on the accuracy of key measured data such as the dosimetric leaf separation and mlc leaf transmission. Measurements had been made previously for the department’s dmlc IMRT programme but were repeated using the latest measurement technique recommended by Varian. Moving from dmlc IMRT to VMAT for head and neck treatments Prior to the introduction of VMAT, suitable head and neck cancers were treated using dmlc IMRT. Primarily, patients with disease in the mouth and throat region were selected, although ethmoid sinus and nasopharynx patients also received treatment with this modality. A class solution of seven equally spaced fixed gantry angles was utilised, along with template dose objectives for PTV coverage and OAR sparing. Building on this experience, a multi-disciplinary team consisting of clinical oncologists, radiographers and physicists was established to investigate the feasibility of utilising VMAT for head and neck cancers. Treatment planning study A treatment planning study was performed to compare optimised dose distributions created by the PRO with those created by dmlc IMRT optimisation for head and neck cancers. Ten patients with floor of mouth disease that were previously treated using dmlc IMRT were randomly selected and replanned with a single 360° arc and double 360° arcs. Arcs were usually restricted to the transverse plain but couch rotations were allowed in exceptional cases where required to meet dose constraints. The treatment plans were evaluated by comparing various parameters, including PTV coverage, OAR dose constraints, dose bathing effect and comparison of monitor units. Use of different structures, margin creation and constraint sets was investigated.