Analysis and Comparison of Treatment Planning Software for Normal Tissue Dose for Large and Small brain Tumors Treated with a Non-Invasive Stereotactic Radiosurgery Instrument

Abstract

<h3>Purpose/Objective(s)</h3> Patients with metastatic brain tumors may be treated with a stereotactic radiosurgery (SRS) platform or a linear accelerator (LINAC)-based platform. Recently, Leksell GammaPlan (LGP) received an update that made an optimization algorithm. Our objective was to review and compare older, completed plans with the newer "optimized" plans to establish if there were improvements on plan parameters, namely, Gradient Index (GI), Selectivity Index (SI), and Paddick Conformity Index (PCI). <h3>Materials/Methods</h3> We reviewed our institutional database in order to identify and gather patients with brain metastases treated by SRS. We chose patients with varying lesion sizes to see if there were any significant differences in how the optimization algorithm would treat large lesions (>5 cc), what we defined as very small (<2 cc), small (2-3 cc) and medium-sized lesions (3-4 cc). All lesions were originally treated with a 900 cGy x3 fractionation scheme. Optimization within the LGP software was performed on these plans while maintaining at least 99.51% coverage of the lesion volume as well as normalizing each plan to prescribe to the 50% isodose line in order to preserve the integrity of the original SRS plans. Because PCI is SI * coverage, and coverage was considered to be unity, only values of SI were directly observed. <h3>Results</h3> The cohorts had improvements overall in both GI and SI. 30 patients with metastases were identified that were treated using SRS. 15 of which had large lesions, 5 of which had medium-sized lesions, 5 of which had small lesions, and 5 of which had very small lesions. The GI improved for the very small on average by 0.492 (14.88%), for the small on average by 0.482 (16.27%), for the medium on average by 0.37 (12.66%), and for the large by 0.176 (6.63%). The SI improved for the very small on average by 0.144 (25.71%), for the small on average by 0.096 (12.83%), for the medium on average by 0.132 (22.92%). For the large tumors, SI on average lowered by 0.017 (2.05%). It is important to note that SI for only large tumors has appeared to marginally decrease. <h3>Conclusion</h3> From the advent of the new LGP optimization algorithm, utilizing this new update may be associated with improvement in dose falloff and local control of the tumor volume. Our data suggests that for lesions of all sizes, the optimization significantly improves all parameters in question for Gamma Knife treatment planning.