ESTRO 2024 - Abstract Book

S3477

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2024

The final PTV-based plan is optimized using DAO for the two photon paths and the two selected electron arcs. The same plan is also robustly optimized on the clinical target volume (CTV) and without PRVs, using stochastic programming for random set-up uncertainties in all cardinal directions (equal weights for all scenarios). The set up uncertainty magnitude, σ, is calculated from the PTV margin, m, such that σ = 0.7 m. The plans are normalized with the prescription dose to the median PTV or CTV (for robust plans). The robustly optimized plans are compared to PTV-based plans on their relative photon/electron contribution and dosimetric endpoints of the nominal (no uncertainty) scenario and for random set-up uncertainties of magnitude σ in all three cardinal directions for three clinically motivated cases.

Results:

The three cases considered are described in Table 1. The relative electron contribution to the target dose was between 13.5% and 40.3%. Electrons contributed more in the robust plan than in the PTV-based plan for the brain case; the inverse is true for the pelvis and the breast case (Table 1).

Figure 1 shows the dose volume histograms for the nominal scenario (left) and for the random uncertainty scenario (right) for all cases and plans. In the nominal scenario, CTV coverage is similar for the PTV-based and robust plans for all cases but small differences in OAR sparing were observed. For the brain case, the right optic nerve and the brainstem had a PRV in the PTV-based plan. For these two small OARs, the PTV-based plan had a lower D 2% than the robust plan but the variation induced by random uncertainties were larger than for the robust plan and D98% to the CTV was 1.0 Gy higher in the robust plan than in the PTV-based plan in the random uncertainty scenario. For the pelvis case, all OARs were slightly better spared with the robust plans than with the PTV-based plan in both the nominal and the random uncertainty scenarios. However, D 98% to the CTV was 1.0 Gy lower for the robust plan compared to the PTV-based plan in the random uncertainty scenario. For the breast case, the spinal canal had a PRV in the PTV-based plan but D 2% was 0.5 Gy higher than in the robust plan. Mean dose to either lung, the contralateral breast, and the heart were similar within 0.1 Gy between the PTV-based and the robust plan in the random uncertainty scenario. Of note, the volume of ipsilateral lung (Lung L) receiving doses of 15 Gy or above were lower with the robust plan in both the nominal and the random uncertainty scenarios.