Room: Room 209
Purpose: Some concerns exist about critical organ dose and delivery accuracy for VMAT plans. The purpose of this study is to retrospectively compare clinically treated VMAT and ssIMRT spine SBRT plans in terms of dosimetric endpoints and pre-treatment quality assurance (QA).
Methods: 98 plans of 18 Gy single fraction spine SBRT that were delivered from 2014 to 2016 in our institution were selected from a board reviewed registry, including 34 ssIMRT and 29 VMAT for cervical/thoracic spines, and 19 ssIMRT and 16 VMAT for lumbar spines. The conformality index (CI) and homogeneity index (HI) were calculated, and phantom QA records (gamma passing rate) were compared for each plan. For the spinal cords, the maximum dose in 0.03 cc (D0.03cc) and volume received 10 Gy (V10Gy) were recorded. Five ssIMRT plans with large and complex target volumes were selected to be replanned with VMAT and then delivered to phantom for head-to-head comparison. Statistical significance was tested with the Mannâ€“Whitney U test.
Results: For the cervical/thoracic ssIMRT and VMAT plans, the median cord V10Gy were 0.11 cc and 0.05 cc (p = 0.006); the median cord D0.03cc were 10.85 Gy and 10.10 Gy (p = 0.032); the median CI were 1.28 and 1.08 (p = 0.009); and the median HI were comparable. For lumbar spines, no significant dosimetric endpoint differences were observed. Two modalities were comparable in delivery accuracy. For the five selective cases with head-to-head comparison between VMAT and ssIMRT, the observed differences in plan quality were confirmed. However, at least one beam from each ssIMRT plan had passing rate <90% with 2% and 2 mm gamma criteria, while all VMAT plans had >95% passing rate with the same criteria.
Conclusion: VMAT provides better dosimetric quality without compromising delivery accuracy compared to ssIMRT for single fraction spine SBRT.