ESTRO 38 Abstract book

S507 ESTRO 38

clinical as well as the dose fall-off plans, PTV and OAR dose parameters (rectum, anal canal, bladder, femoral heads) were evaluated to check if OAR sparing improved. In analogy to knowledge-based approaches, consistency of the the rectum sparing was analysed by plotting the mean dose against the fraction of rectum volume overlapping the PTV [1] for both the clinical and dose fall-off treatment plans. Results For all patients, the new optimisation approach resulted in clinically acceptable plans with a reduction of the dose in the rectum, anal canal and bladder (fig. 1), while maintaining similar PTV coverage and conformity and only a slight increase of the dose in the femoral heads. The largest improvement in OAR sparing was observed for the rectum and anal canal (an average reduction of the mean dose of 19.1 and 16.3% respectively). This was mostly achieved by a reduction of the intermediate dose to these OAR (< 45 Gy). Fig. 2 shows the normalized mean rectum dose as a function of the fraction of the rectum volume overlapping the PTV. Over the whole range of overlap a reduction of the mean rectal dose was achieved, with the largest gains at smaller overlaps fractions. For these patients, the dose constraints can be easily achieved, and larger manual adjustments are needed to achieve optimal sparing. The dose fall-off based optimisations furthermore resulted in a more consistent OAR sparing, reflected by a reduction of the spread of the rectum dose as a function of the overlap fraction.

In Figure 2 the dose-volume histograms of a patient plan with and without the modulation is presented. It shows an underdosage for the PTV (red) and an overdose in the distal tissue (here: the trachea in blue). The results from the phantom study show an increase in the dose difference as the distance in lung increases and the volume decreases. Underdoses from a few percent up to 12% for distances up to 15cm in lung were found in a conservative approach. For patients plans, the PTV underdosage ranges between 1% and 5% in comparison to the plan calculated with the treatment planning system. Conclusion We are able to analyse the effects of Bragg curve degradation due to lung parenchyma in the treatment planning process of lung cancer patients. As the inclusion of this Bragg peak degradation cannot be easily implemented in treatment planning routines, this study gives a conservative approximation for the underdose in the PTV, when it is not accounted for. [1] Baumann K-S, Witt M, Weber U, Engenhart-Cabillic R and Zink K 2017 An efficient method to predict and include Bragg curve degradation due to lung-equivalent materials in Monte Carlo codes by applying a density modulation Phys. Med. Biol. 62 3997-4016 [2] Perl J, Shin J, Schuemann J, Faddegon B and Paganetti H 2012 TOPAS: an innovative proton Monte Carlo platform for research and clinical applications Med. Phys. 39 6818– 37 PO-0941 Can dose gradient-based plan optimisations compete with autoplanning for optimal prostate plans? D. Schuring 1 , A. Van Nunen 1 , F. Van Aarle 1 , T. Jongsma- van Nunen 1 , T. Budiharto 1 1 Catharina Hospital, Department of Radiation Oncology, Eindhoven, The Netherlands Purpose or Objective Multiple studies have shown that autoplanning improves consistency and optimal OAR sparing for individual patients in treatment planning [1]. These inconsistencies in OAR sparing are often caused by the use of mean dose or DVH objectives in the inverse optimisation, which are sensitive to anatomical variations in patients and require adjustments by the treatment planner. By optimising on the dose gradients inside the OAR, more consistent results can be obtained as these objectives are less sensitive to these variations. This was tested by replanning clinical prostate patients using these gradient-based objectives and comparing the resulting OAR doses to the clinical plans using the “traditional” optimisation approach. Material and Methods 27 patients with prostate cancer receiving a total dose of 76 Gy were randomly selected from the clinic. These patients were then replanned with a new class solution that uses dose fall-off objectives on the OAR in the RayStation TPS. During re-optimisation no manual adjustments to the objectives were allowed. For the

Conclusion When using a new class solution with dose-gradient objectives for the OAR, a significant improvement of rectum, anal canal and bladder sparing for prostate patients was achieved without manual adjustments of the planning objectives. Similar to autoplanning approaches, this results in more consistent sparing of the OAR with