Room: ePoster Forums
Purpose: To assess changes in VMAT treatment plan delivery as a result of large dose changes without plan re-optimization.
Methods: Three highly modulated head and neck VMAT plans with a total of nine beams were calculated in a phantom using an average of 64 control points ranging from 11 to 90 points per every four degrees of gantry rotation. The beams were then recalculated with doubled MU without re-optimization. Both the original beams and the ones with doubled MUs were delivered on a Varian TrueBeam machine. The log files for each beam were analyzed to study the details of MLC movement. The delivered MUs and MUs multiplied by the corresponding segment areas were compared between the original beam and double-MU beam. Dual plots of the segment area vs. the accumulating MUs were used to observe the difference in MLC movement between the original beam and double-MU beam.
Results: When doubling the MU from each beam, the difference between the expected and delivered MU ranged from -0.195% to 0.174%, with an absolute average of 0.043%. The difference between the expected and delivered MU*segment area ranged from -0.195% to 0.173%, with an absolute average 0.047%. This confirms that without re-optimization, simply changing MUs does not change the plan segments by much. However, as shown in the dual plots, when the MLC areas reach a peak or trough, discrepancies can be observed.
Conclusion: Without re-optimization, VMAT treatment plan delivery does not change significantly as a result of large dose changes. However, slower movement of MLC is observed for higher MU beams, which may be contributed by the finite sampling time of the MLC controller. These results are still considered preliminary and require an increase in sample size to assess the overall impact large dose changes have on VMAT treatment plans.