Effects of static dosimetric leaf gap on MLC-based small-beam dose distribution for intensity-modulated radiosurgery

The aim of the present study was to evaluate the effect of various specific dosimetric leaf gaps on the multileaf collimator (MLC)-based small-beam dose distribution. The dosimetric static leaf gap was determined by comparing the profiles of small MLC-based beams with those of small collimated fields (square fields of 1, 2, 3, and 4cm). The results showed that an approximately 2-mm gap was optimal with the Millennium 120-leaf MLC (Varian Medical Systems, Palo Alto, CA) and a Varian 21EX 6-MV photon beam. We also investigated how much the leaf gap affects the planning results and the actual dose distribution. A doughnut-shaped planning target volume (PTV, 6.1 cm(3)) and inner organ at risk (OAR, 0.3 cm(3)) were delineated for delicate intensity-modulated radiosurgery test planning. The applied leaf gaps were 0, 1, and 2 mm. The measured dose distributions were compared with the dose distribution in the treatment planning system. The maximum dose differences at inside PTV, outside PTV, and inner OAR were, respectively, 22.3%, 20.2%, and 35.2% for the 0-mm leaf gap; 17.8%, 22.8%, and 30.8% for the 1-mm leaf gap; and 5.5%, 8.5%, and 6.3% for the 2-mm leaf gap. In a human head phantom (model 605: CIRS, Norfolk, VA) study, large dose differences of 1.3%-12.7% were noted for the measurements made using the MLC files generated by the three different leaf gaps. The planned results were similar, and measurements showed a large dose difference associated with the various leaf gaps. These results strongly suggest that plans generated by a commercial inverse planning system commissioned using general collimated field data will probably demonstrate discrepancies between the planned treatments and the measured results.