Room: Exhibit Hall | Forum 7
Purpose: To study the sensitivity of IMRT delivery parameters in non-small-cell lung cancer using Energy-based inverse optimization.
Methods: Twelve lung cases were retrospectively optimized using dose-volume (DVH) and Energy optimization. DVH optimization was used as a baseline since it is the most common clinically used optimization scheme. In both optimization approaches nine equispaced, split beams were used for step-and-shoot deliverable IMRT. Changes to two parameters were investigated: the number of IMRT segments (5 and 10 per beam) and the minimum allowed segment area (2 and 6 cm2). Optimization was performed by iteratively lowering doses to organs at risk until the standard deviation across the PTV was ~4.0%. The plans were normalized such that 95% of the PTV received the same dose. The sensitivity metric used was based on dose indices (DIs), which represent the dose delivered to a certain structure volume. The DIs included1%_cord, 33%_heart, 20% and 30%_both-lungs, and 50%_esophagus. Changes of the DIs with changing delivery parameters were tallied.
Results: The results indicate larger changes in the Energy-based plans as compared to DVH-based plans. The Energy-based plans showed statistically significant differences when increasing the segment area and keeping 5 segments/beam for 1%_cord, 30%_both-lungs and 50%_esophagus. Also, changes were significant for 1%_cord when increasing the segment area with 10 segments/beam and when the number of segments increased keeping the segment area 6 cm2. The DVH plans showed significant difference in 20%_both-lungs when the number of segments increased while keeping the segment area 6 cm2.
Conclusion: This study showed that for NSCLC Energy-based optimization plans exhibit larger sensitivity than DVH-based plans to changing IMRT delivery parameters such as segment area and number of segments. The changes were more pronounced when increasing the segment area rather than when allowing more segments per beam.