Author: Ivan L. Cordrey, Dharmin D. Desai 👨🔬
Affiliation: Cumberland Medical Center - Cancer Center, Varian Advanced Oncology Solutions 🌍
Purpose: For larger PTVs, VMAT optimizations performed within Eclipse often generate SRS plans with widely scattered hotspots distributed around the inner edge of the PTV – the “ring of fire.” This project presents an optimization approach for Eclipse using nested substructures to guide the dose to the central region of the PTV.
Methods: HyperArc SRS plans were created for 5 PTVs (2 spherical, 3 irregular) in the volume range 7.90 cm3 to 20.34 cm3 to represent the range of volumes where the “ring of fire” is expected. A standard Eclipse optimization was performed on each of the 5 PTVs. A guided optimization was then performed on the PTVs by creating 2 nested substructures within each PTV and entering substructure-specific parameters into the Eclipse optimization engine. The larger substructure (A’) was created by copying and uniformly shrinking the original PTV (A) to approximately 25% of its volume [A’ ≈ 0.25(A)]. Similarly, the innermost substructure (A’’) was created from A’ [A’’ ≈ 0.25(A’) ≈ 0.063(A)]. The resulting standard and guided optimized plans were compared via hotspot location and plan conformality (GI, CIRTOG, and CIPaddick).
Results: The guided optimization approach reliably shifted the high dose volume to the center of all 5 SRS PTVs planned with Eclipse. Additionally, the use of nested substructures A’ and A’’ did not adversely affect plan conformality (Standard optimization: GIavg = 2.43±0.07, CIRTOG-avg = 1.03±0.02, CIPaddick-avg = 0.92±0.02; Guided optimization: GIavg = 2.76±0.22, CIRTOG-avg = 1.05±0.04, CIPaddick-avg = 0.91±0.03).
Conclusion: When used within Eclipse, the guided optimization approach utilizing nested substructures and substructure-specific optimization parameters is effective at centrally focusing the dose for large SRS PTVs with scattered, non-focused hotspots (“ring of fire”), while preserving plan quality.