Session Item

Proton pencil beam therapy for selected breast cancer patients provides an alternative to photon therapy, as the dose to heart and lungs can be reduced without compromising target coverage. Robustness towards inter-fractional changes, however, pose an issue for proton therapy. Changes due to the breast mobility or physiological swelling or shrinkage challenge the treatment quality. We investigate the anatomical robustness of the initial 20 patients with early breast cancer treated at the Danish Centre for Particle Therapy (DCPT). The patients are pilots to the Danish Breast Cancer Group (DBCG) proton trial (NCT04291378) and selected based on high photon dose to the heart or lung.

The DCPT treatment strategy is free-breathing, 2-3 en face fields combined with single field- and robust optimization, a 5 cm range shifter and a 5 mm distal margin to all CTVs optimized with low priority. The CTV consists of: CTVp breast/chest wall, CTVn levels 2-4, interpectoral nodes, internal mammary nodes (IMN) and for 13/20 patients CTVn level 1. The prescribed dose is 50 Gy(RBE) in 25 fractions (12/20) or 40 Gy(RBE) in 15 fractions (8/20). 5/20 received simultaneously integrated boost. The planning objectives are V95%>98% for CTVp and V90%>98% for CTVn. A robust evaluation is performed using 14 combined scenarios (table 1b) requiring a worst-case V95%>95% for CTVp and V90%>95% CTVn. 2-3 scenarios with lower CTVn IMN coverage are allowed in order to minimize the heart dose. A final anatomical evaluation is performed using 4 artificial CTs created by adding/removing 3 mm and 5 mm tissue to/from the patient outline, figure 1a.

The target coverage, normal tissue dose, robust and anatomical evaluations are shown in table 1. All plans meet the nominal and robust requirements. The anatomical evaluations show that 3 mm shrinkage approximately doubles the mean heart dose and 5 mm swelling reduces the nominal CTVn IMN coverage below the constraints. Figure 1b illustrates the steep dose fall-off towards the heart. These results are used to create a patient specific variable "outer limit" tolerance structure to be used during daily, online CBCT evaluations. Based on the weekly CT scans, a total of 15 plan adaptions were performed: 4/15 due to shrinkage and 9/15 due to extra tissue. Figure 1c shows an example of a plan adaption due to 5 mm swelling.

Robustness towards anatomical changes must be evaluated separately from range and setup uncertainties. Small amounts of extra tissue deteriorate the target coverage, while shrinkage increases the normal tissue dose. Using this planning strategy, the CTVns are in general robust to 3 mm and the CTVp to 5 mm extra tissue. Increasing it further will have consequences for the heart and lung dose. At DCPT, the weekly CT scans are now replaced by an adaptive re-scanning strategy based purely on daily CBCT evaluations using the variable outer limit as tolerance combined with a 3 mm tolerance on shrinkage.