Dose verification of volumetric modulated arc therapy (VMAT) by use of in-treatment linac parameters.

Linac parameters such as the multi-leaf collimator (MLC) position and jaw position, cumulative monitor units (MUs), and the corresponding gantry angle were recorded during the clinical delivery of volumetric modulated arc therapy for prostate, lung, and head/neck cancer patients. Then, linac parameters were converted into the beam-data format used in the treatment planning system, and the dose distribution was reconstructed. The dose-volume histogram and the dose difference (DD) were compared with the corresponding values in the treatment plan. A reproducible error of in-treatment linac parameters was observed when a sudden change of beam intensity or MLC/jaw speed occurred. The maximum cumulative MU error was more than 4 MU for lung cancer cases, and the maximum MLC position exceeded 5 mm for prostate and head/neck cancer patients. However, these errors were quickly compensated for at the next control point. All treatments analyzed in the present study were delivered within 0.4% accuracy at the planning target volume. The cumulative dose agreed with that of the plan within 3% of the prescribed dose. The 1% DD was 93.9, 99.9, and 93.4% of the prescription dose for prostate, lung, and head/neck cancer patients, respectively.