Organs-at-Risk Sparing Using Midline Mlc Travel Limits for VMAT on Varian Ethos
Abstract
Purpose
Limiting jaw position across midline is a common inverse-planning strategy to minimize dose to normal tissue and improve conformality when treating irregular targets. However, jawless linacs lack an analogous, planner-controlled method. In this study, we evaluated the effect of restricting MLC leaf travel across midline on organ-at-risk (OAR) dose while maintaining target coverage in prostate+node VMAT plans.
Methods
Five prostate+node patients previously treated at our clinic were replanned using the open-source ORIBIT-RT Prostate+Nodes RapidPlan (Varian Ethos Photon Optimizer) model. For each patient, two VMAT plans were generated using the same prescription, field arrangement, planning objectives, and constraints. Both plans met the same target coverage criteria (V100%rx>99%). One plan allowed unrestricted leaf travel while the other applied symmetric MLC travel limits about the isocenter (±3.0 cm laterally, ±2.0 cm superior-inferiorly) using the Limit Leaves option in the Eclipse PO V18 optimizer. The resulting DVH metrics were compared for bladder, large bowel, small bowel, left and right femoral heads, penile bulb, and rectum. Paired, two-tailed t-tests were used for mean-dose and prescription-metric comparisons.
Results
Statistically significant mean-dose reductions were observed for bladder (p=0.004) and right femoral head (p<0.0001). Maximum dose was comparable or reduced for all OARs (mean change −0.38%). DVHs demonstrated the most consistent improvements across patients for bladder, large bowel, and femoral heads, with minimal change for remaining OARs. Additionally, clinical OAR metrics showed moderate improvement with V45Gy and V41Gy significantly lower for large bowel (p=0.02) and bladder (p =0.007), respectively.
Conclusion
Limiting leaf travel positions reduced OAR dose in prostate+node VMAT plans without sacrificing target coverage, suggesting it can function as a practical surrogate to limiting jaw motion on jawless linacs. This low-effort, planner-driven technique is readily deployable and with additional validation in larger patient cohorts could be extended to additional complex sites, such as head and neck.