Purpose: Assess the feasibility of an MRI-guided linac system in treating brain metastasis by comparing the plan quality to a frameless automated linac-based SRS system for treatment of brain metastasis.
Methods: Ten patients with a single metastatic lesion were retrospectively planned using a frameless automated linac-based SRS planning system with HDMLC (2.5mm leaf-width) and an MRI-guided linac system with double-focused double-stacked multileaf collimator (4.15mm leaf-width). Lesion sizes ranged from 0.3cc to 27.1cc, with an average of 5.1cc. Energy for all plans was 6MV FFF and normalization set to cover 99.5% of the target volume with the prescription dose. The HDMLC plans utilized VMAT technique with 4-5 non-coplanar arcs at 1200 MU/min while the MRI-guided linac system used step-and-shoot technique with multiple coplanar beams at 600 MU/min. Plan quality was assessed by the following metrics: gradient index (GI), RTOG conformity index (CI(R)(T)(O)(G)), brain volume receiving 12Gy or more (V12Gy) and maximum dose (D(m)(a)(x)) for brainstem, optic chiasm, and optic nerves. Two-sided Wilcoxon sign-rank test (statistical significance at p<0.05) was used to test differences between metrics.
Results: All plans met OAR constraints. D(m)(a)(x) to the brainstem, optic chiasm and nerves were not significantly different between the plans. HDMLC-linac plans were statistically better in terms of GI (p=0.00), CI(R)(T)(O)(G) (p=0.00) and V12 (p=0.01).
Conclusion: Our study demonstrates that MRI-guided linac system can provide comparable organs-at-risk sparing to the HDMLC-linac for treatment of single metastatic lesions, although the HDMLC-linac plans had better GI, CI(R)(T)(O)(G), and V12. The GI for the MRI-guided plans improved with increasing target volumes, achieving acceptable values (<4.0) for targets greater than 3cc, suggesting it could be used for SRS treatment of single metastasis targets larger than this size.