Author: Megan E. Daly, Ryan D. Hernandez, Abriel J. Horak, Soo Kyoung Kim, Peter C. Park, Dwaine O. Spence, Payton H. Stone, Cari L. Wright 👨🔬
Affiliation: UC Davis Cancer Center 🌍
Purpose: To evaluate a fallback planning workflow used for cross-modality treatment planning and delivery between helical Tomotherapy and volumetric modulated arc therapy (VMAT) using conventional L-shaped linear accelerators.
Methods: A total of 91 base/fallback plan pairs were selected, covering 4 different anatomical sites: prostate and nodes (29), prostate (25), pelvis (23), and head and neck (14). All plans were delivered clinically and converted mid-course from Tomotherapy to conventional L-shaped linear accelerator to account for machine downtime. The dose volume histograms were sampled for planning target volumes and organ at risk structures in both plan types at commonly used, clinically significant dose levels.
Results: Comparison between the base and fallback plans reveals that target structure metrics undergo only slight deviation throughout the fallback planning process. Target coverage was slightly lower (PTV V100: -0.66% avg), while conformality and homogeneity indices were slightly improved (CI: +0.83% avg, HI: +0.25% avg). Variance in normal structure dose constraints was highest in head and neck plans, ranging from -5.6% to +4.87% normalized to prescribed dose per fraction. The differences in dose to normal structures for prostate and nodes, prostate, and pelvis plans ranged from -2.36% to 1.85%, -2.73% to 1.08%, and -0.23% to 3.95% respectively.
Conclusion: Normal structure doses remained within acceptable ranges for most anatomical sites, with prostate and pelvis plans showing the least variability. However, head and neck plans exhibited greater deviations in organ-at-risk dose constraints, underscoring the need for careful consideration in complex treatment sites.