Room: AAPM ePoster Library
Purpose:
To assess the dosimetric feasibility of HN SBRT on a MR-Linac, Elekta Unity.
Methods:
10 HN SBRT cases were retrospectively planned using Monaco for MR-Linac. Reference clinical plan was optimized for Truebeam STx (2.5 mm HDMLC) using VMAT with options of non-coplanar beams and collimator rotation. Elekta Unity (7 mm MLC leaf width) limited MR-Linac planning to coplanar step-and-shoot IMRT with 90-degree collimator rotation only. For comparison, target coverage of MR-Linac was scaled to match reference clinical plan. Relevant plan quality metrics for the targets; such as target conformity index (CI, V100/TV), gradient index (GI, V50/V100) and homogeneity index (HI, Dmax/Dmin), (V100, V50 = 100% and 50% isodose volume respectively, Dmax, Dmin = maximum and minimum dose respectively, and TV = target volume), and for organ-at-risks (OARs) such as Dmax and mean dose were collected.
Results:
MR-Linac plans were inferior in conformity and dose gradient in 9 cases. The average percentage difference of CI and GI were 8.4% (±9.6) and 22.4% (±16.8). These translated to higher OAR dose in the MR-Linac plans. For the neck and base of tongue cases, blood vessels, larynx, mucosal avoid, oral cavity and mandible were common OARs close to the targets, with spinal cord and brainstem further away. MR-Linac plans met all clinical goals in these cases. Nasopharynx and skull bases cases posed significant challenges for MR-Linac planning with additional nerve and brain OARs and higher demands on dose drop-off around multiple OARs. For these cases, MR-Linac plan was not able to meet all the constraints.
Conclusion:
Initial MR-Linac plan quality may not be optimal when compared to current standard of care. Further study is warranted to examine the effect of online plan adaptation, especially for sites (neck region) where soft tissue visualization assumes greater importance and favorable anatomy may facilitate improved plan quality.
Not Applicable / None Entered.
Not Applicable / None Entered.