Author: Ivan L. Cordrey, Dharmin D. Desai, Ellis Lee Johnson 👨🔬
Affiliation: Cumberland Medical Center - Cancer Center, University of Kentucky, Varian Advanced Oncology Solutions 🌍
Purpose: HyperArc (HA) is known to produce a high quality SRS plan. Adding Surface Area Informed Optimization (SAIO) to HA has been shown to improve plan quality metrics R50% and the Conformality Index (CI). However, the Efficiency Index (η50%) incorporates Mean Dose (D̅) and is specifically designed for the comparison of overall SRS plan quality. This project is a direct head-to-head comparison of the Efficiency Index for HA vs HA+SAIO for single target and single isocenter multiple targets (SIMT) SRS.
Methods: The Efficiency Index (η50%) proposed by Dimitriadis and Paddick (2018) is a metric of the “useful energy” / ”total energy” delivered in an SRS treatment and is represented as the ratio of mean doses (D̅) and volumes (V) as: η50% = (D̅TV x VTV) / (D̅PIV50% x VPIV50%), where TV = Target Volume and PIV50% = 50% of Prescription Isodose Volume. These quantities are easily obtained from most commercial treatment planning systems (TPS). Three single target and three SIMT SRS cases were planned with both HA and HA+SAIO in the Eclipse TPS. A publicly available script was used to create the SAIO optimization structures and constraints. Using data extracted from Eclipse, η50% was calculated for 20 individual PTVs and compared by Wilcoxson Signed Rank Test to assess statistical relevance.
Results: The HA median η50% = 32.4% vs the HA+SAIO median η50% = 38.8%. Wilcoxon Signed Rank Test verifies HA+SAIO returns higher η50% values than HA alone; p < 0.0001 meaning the observed improvement in the Efficiency Index for HA+SAIO is statistically significant at the 0.05 confidence level.
Conclusion: Aggregate results show HA+SAIO improvement over HA alone in Efficiency Index (η50%) for the 20 individual PTVs. This means a higher “useful energy” is delivered when SAIO parameters are added to HyperArc optimization.