Purpose: To directly compare the plan quality of Gamma Knife (GK)- vs linear accelerator (LINAC)-based delivery techniques for fractionated stereotactic radiotherapy (fSRT) of large brain metastases.
Methods: For 15 patients with >10cc clinical target volumes (CTVs), comparative plans were run for 9Gy×3 fractions utilizing Eclipse vs Leksell GammaPlan (LGP) treatment planning system (TPS). Each GK plan was first developed using LGP’s automatic planning initially, followed by manual adjustment/refinement. The same MRI and structures (CTV+OARs) were then DICOM-transferred to Eclipse TPS, and used to run Volumetric Modulated Arc Therapy (VMAT) plans and Dynamic Conformal Arc (DCA) plans for a Truebeam with high-definition MLCs, using a single isocenter and 3 non-coplanar arcs. No PTV margins were added and no heterogeneity correction was used for either TPS. GK plans were prescribed to the 50% isodose line, and Eclipse VMAT and DCA plans allowed the maximal dose up to 170% and ~125%, respectively. Gradient index (GI), selectivity, Paddick Conformity Index (CI), V20GyRind, and V4GyRind of all three techniques were calculated and compared.
Results: Median CTV volume was 13.9cc (range 10.3cc–39.1cc). Mean±standard deviation of Paddick CI were 1.18±0.05, 1.11±0.03, and 1.37±0.16 for GK, VMAT and DCA plans, respectively. They were respectively 2.64±0.17, 2.47±0.17, and 2.76±0.40 for GI; 13.9±5.8, 9.8±3.29 and 17.80±6.6 for V20GyRind; and 280.1±88.4, 305.4±84.9, and 368.53± 86.8 for V4GyRind. The differences were all statistically significant with p<0.01 for most cases.
Conclusion: A direct comparison using the same MRI scan and contours was performed to evaluate the plan quality of various FSRT delivery techniques for CTV>10cc. LINAC VMAT provided the best dosimetric outcome in regards to CI, GI, and V20GyRind. GK outcomes were similar while DCA outcomes were significantly worse. Even though GK has a smaller collimator penumbra, VMAT outperformed GK in this study due to enhanced penumbra sharpening and better beam optimization.