Session Item

Since 2019 mediastinal lymphoma (ML) patients in the Netherlands are treated with intensity modulated proton therapy (IMPT) using model-based selection (MBS). MBS was performed according to the Dutch national indication protocol for proton therapy. For ML this is based on the difference in normal tissue complication probability (ΔNTCP) for lifetime acute coronary events (ACE). The objective of this study was to evaluate dose advantages of both qualifying and non-qualifying patients. We evaluated the feasibility of MBS and the limitations of this approach by evaluating doses to other organs at risk (OARs) relevant for induction of secondary tumours of which currently no NTCP models are available.

All 61 ML patients referred for plan comparisons (VMAT vs IMPT) to one of the three Dutch proton centres were included in this study. In each patient, the estimated clinical benefit in terms of ΔNTCP is derived from both treatment plans using the ACE NTCP model which includes mean heart dose (MHD), age, gender and risk factors for ACE as predictors. The ΔNTCP threshold for MBS is a 2% reduction for ACE. The differences in MHD, lungs (MLD) and oesophagus (MED) were evaluated. Finally, the reduction in mean breast dose (MBD) was investigated in female patients.
Robust IMPT plans were created using two to eleven beams depending on the target volume and a 3D robustly optimised planning technique was used. Radiotherapy consisted of 20-36 Gy in 2 Gy fractions and chemotherapy consisted of 3-4 cycles of A(B)VD or escBEACOPP for Hodgkin lymphoma or 6x R-CHOP for B-cell lymphoma.

Thirty-two out of 61 patients (52%) referred for plan comparisons met the predefined ΔNTCP threshold and were eligible for IMPT (Table). In patients qualifying for IMPT, IMPT resulted in a significantly lower MHD (8.9 Gy vs. 5.3 Gy; p<0.01), MLD (8.3 Gy vs. 6.0 Gy; p<0.01), MED (15.3 Gy vs. 11.2 Gy; p<0.01), MBD-right (2.9 Gy vs. 2.1 Gy; p=0.001) and MBD-left (3.6 Gy vs. 2.3 Gy; p=0.004). Using the NTCP model for ACE, the mean NTCP for VMAT was 12.1% (range 2.1 – 30.4%) compared to 8.6% (range 0.1– 27.4%) for IMPT, an average ΔNTCP with IMPT for ACE of 3.4% (range 2.0 – 9.8; p<0.01). In patients that did not qualify for IMPT, the dose to all OARs was also significantly reduced; MHD (7.3 Gy vs. 5.8 Gy; p<0.01), MLD (8.3 Gy vs. 6.2 Gy; p<0.01), MED (12.5 Gy vs. 10.5 Gy; p=0.004), MBD-right (3.3 Gy vs. 2.2 Gy; p=0.012) and MBD-left (4.6 Gy vs. 3.6 Gy; p=0.014) (Figure).

Model-based selection of ML patients for proton therapy is feasible. IMPT resulted in a significant MHD reduction and a subsequent 3.4% reduction in estimated lifetime risk of ACE. However, patients not qualifying for proton therapy could benefit from proton therapy as we found a significant reduction in dose to several OARs relevant for the induction of cardiac complications and secondary tumour induction. Tools are needed to further optimise patient selection.