Room: Exhibit Hall | Forum 3
Purpose: Demonstrate limitations of conformity and gradient indices when used for multi-target Stereotactic Radiosurgery (SRS) plans.
Methods: Conformity Index (CI) and Gradient Index (GI) are commonly used for quantitative SRS plan evaluation and can be useful for plan optimization. However, these metrics have limitations when used for plan ranking, particularly for multi-target plans. There is no standardized method for defining cumulative CI or GI for multi-target plans, and both metrics depend strongly on target volume. In addition, GI is strongly dependent upon target conformity, and dose-bridging between multiple targets can result in nonsensical GI or CI values. We performed quantitative plan ranking for 5 clinically representative multi-target SRS cases to evaluate these indices for plan ranking. Cumulative RTOG CI and Paddick GI were calculated using simple and volume-weighted averaging over all targets, and using cumulative target volumes with mean or volume-weighted prescription doses.
Results: Calculated average CI and GI, and subsequent plan rank order, varied depending on cumulative plan calculation method. While GI should quantify dose falloff outside the target, a difference of ~40% was observed for two plans with identical 50% isodose lines due to lack of conformity of the Prescription Isodose Volume (PIV) to the target for one plan. If GI is instead calculated as PIV50%Rx/GTV, resulting GI values are nearly identical. Bridging of the 50% isodose lines of separate targets resulted in differences in GI of nearly an order of magnitude for small changes in plan quality.
Conclusion: No standardized method exists for calculating cumulative plan CI and GI for multi-target SRS plans and plan ranking depends on the calculation technique. Effects of conformity on GI can be eliminated by replacing PIV in the numerator with GTV. Overlap of isodose lines between separate targets can result in nonsensical CI and/or GI values which distort quantitative plan ranking.