ESTRO 2024 - Abstract Book

S4685

Physics - Optimisation, algorithms and applications for ion beam treatment planning

ESTR0 2024

For the mean DVHs and DVH parameters, of both targets and organs at risk, no differences between pCT and sCT were found for the VMAT and IMPT plans. The DVH parameters calculated on sCT were within 1.5 % of the corresponding parameters for the pCT for all patients, and outliers could be explained by anatomical differences. Gamma pass rates were lower for IMPT plans compared to VMAT plans, with an average 2%/2mm gamma pass rate of 95.3 % and 98.3 % for IMPT and VMAT plans, respectively. All major deviations could be identified as results of deviating image registrations between sCT and pCT caused by outer patient contour differences, relevant for both photon and proton dose calculation differences, and/or misalignment of bony anatomy, relevant for proton dose calculation differences. Gamma index analysis for a representative case with good anatomical agreement for the image registration, is shown in Figure 2. Common for these cases was that gamma index >1 was located close to the CTV. Patients with worse anatomical agreement had a high gamma index in the same regions, however larger and sometimes extending into the CTV. When visually studying line doses for each field separately, it was apparent that the dose discrepancy was dominating in the dose falloff region of the proton beam.

Figure 2: Representative example of a global gamma index analysis for 2%/2mm for the proton plan of patient with a good registration. The absolute dose difference is shown (top). The dose around the CTV is commonly larger for the sCT compared to the pCT (blue shade). Global gamma analysis with dose cutoff 10 % (bottom) not including points where the contour size was different. Gamma pass rate in this example is 98.6 %. Worse pass rate is seen in the vicinity outside the CTV where the absolute dose difference was the largest.

Conclusion:

The dosimetric evaluation of sCT versus pCT for proton therapy of prostate cancer revealed no differences and results were similar as for photon RT. Hence, the MRI-only workflow for prostate cancer seems to be acceptable in a clinical setting also for proton RT, at least for the field setup investigated. Minor dose discrepancies could only be seen in the gamma index analysis, where deviations were located outside delineated structures. Large dose deviations for photon or proton RT plans could exclusively be explained by registration issues between pCT and sCT.