Quantification of dosimetric impact of implementation of on-board imaging (OBI) for IMRT treatment of head-and-neck malignancies.

Implementation of daily kilovoltage imaging for setup verification improves the reproducibility of treatment by eliminating small random setup errors. We evaluate the dosimetric consequences of such shifts, not yet evaluated, in a group of head-and-neck cancer patients (ENT) treated with intensity modulated radiation therapy (IMRT) at Emory University. Twelve patients with ENT malignancies were analyzed. On-Board Imaging (OBI) was used in at least 70% of each patient's treatment sessions. An isodose distribution was generated for each fraction, with the isocenter shifted to its calculated location prior to OBI repositioning. These plans were summed and then compared to the simulation plan for coverage of target structures. For these 12 patients, there were a total of 18 planning target volumes (PTV). The mean (range) percent reduction in minimum dose was 12.1% (-1.0 to 43.3). For 10 right necks and 9 left necks treated, the mean percent reduction in minimum dose was 11.8% (-0.6 to 39.7) and 13.3% (-3.6 to 31.2), respectively. The mean reduction in mean dose to the PTV was 1.3% (0 to 5.1). The mean reduction in mean dose to the right and left necks was 1.0% (0.2 to 3.9) and 1.13% (0.4 to 3.4), respectively. From this analysis, we conclude that the shifts made were small and random, with essentially no change in mean dose delivered to target structures. There is, however, significant improvement in the minimum dose delivered. Underdosing even a small portion of the tumor potentially sacrifices the probability of local control; correcting these setup errors seems desirable.