ESTRO 2024 - Abstract Book

S3453

Physics - Dose calculation algorithms

ESTRO 2024

The expected result was a plan optimised for AXB-Dm,m with the same dose distribution characteristics as the first one. We recalculated this second plan in terms of Dw,w using AAA algorithm, while maintaining consistent treatment parameters. We carried out this assessment to gauge the impact of transitioning to AXB-Dm,m. We calculated dose differences as percentages for D0,1cc to the skull, as well as for D2% and D98% to the CTV. Additionally, we determined differences in the ICRU 83 Homogeneity Index (HI) to the CTV. Our focus was on the CTV rather than the PTV to assess potential clinical implications. To evaluate the significance of these discrepancies, we employed the two-tailed Wilcoxon signed-rank test at α=0.05.

Results:

Plans optimized with AXB-Dm,m met clinical criteria very similarly to AAA-optimized plans, as expected. However, when recalculating AXB-Dm,m optimized plans with AAA, all metrics increased compared to AAA-optimized plans. Specifically, D0,1cc to the skull, D2% to the CTV and D98% to the CTV were higher by 5.5% (p=0.005), 4.0% (p=0.047) and 1.0% (p=0.005) respectively. HI increased from 0.022 to 0.053 (p=0.005). Figure 1 shows an example of the resulting dose distributions and dose-volume histograms. The observed differences for D0,1cc to the skull and D2% to the CTV are consistent with variations in dose to bone tissue between Dw,w and Dm,m. The difference in D98% to the CTV can be attributed to the soft tissue composition within the majority of the CTV, as dose-to-muscle from AXB-Dm,m is 1% lower than dose-to-water from AAA. Disparities in HI are the consequence of the calculated differences in D98% and, more significantly, D2% to the CTV. Compared to AAA-optimized plans, which constitute the foundation of most clinical knowledge, the higher dose to the bone tissue when planning with AXB-Dm,m may result in an increased Normal Tissue Complication Probability (NTCP). An indirect consequence in this scenario is the potential enhancement in the Tumor Control Probability (TCP). Furthermore, when reevaluating AXB-Dm,m plans in terms of Dw,w, ICRU homogeneity criteria for the CTV, and consequently for the PTV, are not met.