Room: Track 5
Multi-target SRS necessitates high dosimetric accuracy, which may be affected by beam commissioning as a single Dosimetric Leaf Gap(DLG) must be selected and uniformly applied across a given field. We recently proposed CAVMAT, a VMAT technique with improved conformity of MLCs to targets, reduced plan complexity, and improved healthy tissue sparing. MLC motion is simplified by maximizing collimation between targets and by preventing targets from sharing the same leaf pair. We hypothesize that the simplified leaf motion and reduced complexity of CAVMAT may reduce sensitivity to commissioning and treatment delivery uncertainty.
10 VMAT plans were replanned with CAVMAT and all plans were re-calculated with fixed MU using DLG values of 0.4mm, 0.8mm, and 1.2mm. Change in V6Gy, V12Gy, V16Gy, and target dose was quantified. The plans were delivered to a 3D Phantom (Delta4), and dose agreement was quantified using Gamma Index with 3%/1mm, 2%/1mm, and 1%/1mm criteria. Trajectory files were analyzed to quantify MLC positioning errors during delivery. 16 targets were selected and delivered to SRS MapCHECK to evaluate dose differences per DLG.
For varying DLG, VMAT’s average V6Gy, V12Gy, and V16Gy sensitivity was 35.83±9.48%/mm, 34.12±6.61%/mm, and 39.22±8.41%/mm, compared to 23.18±4.53%/mm, 22.45±4.28%/mm and 24.88±4.91%/mm for CAVMAT. Similarly, VMAT was roughly twice as sensitive to changes in target dose. For the strictest criteria of 1%/1mm, VMAT’s average passing rate was 94.53±4.42% compared to CAVMAT’s 99.28 ± 1.74%. For SRS MapCHECK, a DLG of 1.2mm yielded an average absolute dose difference of 5.59±1.93cGy for VMAT and 2.76±1.74cGy for CAVMAT. Log file analysis demonstrated a minimal difference in MLC positional errors.
CAVMAT is more robust to variations in DLG setting in the planning system than VMAT and has better agreement with patient specific QA. The reduced sensitivity and superior dose agreement make CAVMAT a promising technique for multi-target SRS.