Room: Stars at Night Ballroom 4
Purpose: To explore the feasibility and dosimetric effects of applying various setup margin recipes to Gamma Knife frameless treatment planning.
Methods: Measured values of systematic (Σ) and random/residual (σ) uncertainty were input into previously published setup margin recipes: ICRU, VHMF, Parker, and Zhang. For the VHMF, the effective values of Σ and σ were utilized which allows it to be fractionation dependent. The four resulting setup margins were implemented for 30 different patients and new treatment plans were developed (IRB 2017-0266). Treatment plans were scored as pass/fail based on whether normal tissue goals were met. Additionally, an “optimal score� was applied to each plan factoring in mathematical true target coverage ranking relative to other recipes and how well normal tissue goals were met. Bivariate statistics were utilized to determine whether various patient and target characteristics or the use of margin could predict plan failure or higher optimal scores.
Results: Fraction number showed a moderate negative correlation with plan failure (r = -0.398, p < 0.001) and moderate positive correlation with optimal score (r = 0.362, p<0.001). For single fraction plans, largest GTV size showed the strongest correlation with plan failure (r = 0.550, p < 0.001) and lower optimal score (r = -0.675, p <0.001). For multi-fraction plans, only the Zhang formula resulted in failing plans while the VHMF, Parker, and Zhang resulted in statistically significantly higher optimal scores from the original plans.
Conclusion: Fractionation may be highly conducive to increasing the potential to not only apply setup margins, but also better meet normal tissue goals, especially those predictive of radionecrosis. For single fraction plans, target size may be more important for meeting normal tissue goals than use of margin or not. For multi-fraction, use of VHMF or Parker may be ideal for rational application while maintaining normal tissue goals.