ESTRO 2024 - Abstract Book

S4911

Physics - Quality assurance and auditing

ESTRO 2024

contribute to a large data base, which can be retrospectively analyzed to evaluate applied dosimetric and plan specific parameters and tolerances.

Material/Methods:

A total of 1197 clinically accepted plan evaluations with ClearCheck were carried out in Eclipse treatment planning system V15.6 and V18.0 between April and October 2023. Plans were created for both Varian Halcyon and Elekta VersaHD machines using VMAT, IMRT and 3D conformal techniques. For an initial retrospective dosimetric analysis, 72 cases of whole breast irradiation (15 fractions and 2.67 Gy dose per fraction normalized at D 50% for the PTV) were selected and compared to our department’s internal constraints. In evaluating breast treatments, the dosimetric parameters D 98% and D 2% for PTV excluding the outer 5 mm of the body (PTV-Skin) as well as the mean values for the heart, ipsi- and contralateral lung were investigated. To analyze plan-specific parameters, we conducted comparisons of the dose grid size with the CT slice thickness across all 1197 plan evaluations.

Results:

The results for the considered breast cases are presented in Fig. 1, in which our department’s specific constraints are depicted by horizontally dashed lines. The D98% constraint for PTV-Skin could be achieved in only approximately 25% of the cases. It became evident that large deviations occurred due to the steep dose gradient towards the adjacent lung, particularly considering the small total volume of the breast (e.g. blue circle in Fig. 1). Overall, this constraint appears to be overly stringent. The D2% constraint was exceeded once. The mean dose constraints for both lungs and heart were clearly overachieved, with three cases exhibiting inadvertent swapping of ipsi- and contralateral lung constraints (red circle in Fig. 1). For further daily clinical analysis, it is imperative to differentiate between left- and right-sided breast cases and reevaluate the respective constraints and tolerances accordingly. Furthermore, it turned out that for the plan-specific parameter considered as an example here in 70 cases – i.e. 5.8% – the dose grid size exceeded the respective CT slice thickness. This observation highlights the potential for including it as an automated check parameter in future assessments.