Impact of Collimator Angles During VMAT Planning and Delivery
C Yang*, N Mistry, Y Chen, K Doxsee, Monmouth Medical Center, Long Branch, NJSU-E-T-390 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To evaluate the dosimetric impact from collimator angles during the VMAT planning and delivery.
Methods: An early stage lung cancer treatment with Pinnacle SmartArc™ planning system was evaluated, for both planning and delivery with dosimetric differences on an Elekta Synergy™ platform. Conventional planning process was utilizing 450 collimator angle to maximize the necessary PTV margins. A single 3590 arc with 10 MV photons were adopted. We compared three various collimator angles (00, 450 and 900) with generated planning and delivery variations; SunNuclear 3DVH™ module was also implemented to review the dose delivery discrepancy associated with the VMAT techniques corresponding to various collimator angle settings.
Results: VMAT delivery with various collimator setting introduces dosimetric differences. PTV D95% dose coverage for all angles is comparable (00, 450, 900 = 93.3%, 92.9%, 92.7%) while 450 collimator introduces more leakage doses through the superior/inferior directions. OARs of 00 and 900 collimators have different distributions such as 1% of cord volume doses of 47.9 cGy and 40.7 cGy per fraction, but all within clinical tolerances. Total lung doses are within limits with V20 values of 17.2%, 17.8%, 16.7% for 00, 450, 900 collimator, respectively. Treatment MUs for all collimators are within 2% variation due to the similar control point distributions.
Conclusion: 450 collimator may not the best choice for VMAT delivery with tumor size less than the MLC limitation. The convention was defined to suit the LINAC limitation with compromise at initial implementation. To avoid extra MLC leakage, 00 or 900 collimator may be a proper choice. Conformity with target shape or avoidance with the OAR locations with BEV optimization in initial planning stage is also important to validate the collimator angles. Currently there is no optimum collimator choice built in the planning system, user experience is also an influential factor.