Session Item

Several studies have shown that dual energy CT (DECT) can lead to an improved accuracy for proton range estimation, and a few proton centers have introduced DECT-based proton treatment planning along with a reduced range uncertainty margin. In this study, we investigated the clinical benefit of applying a reduced range uncertainty in the robust optimization in proton treatment planning for neuro-oncological patients.

In total, 27 neuro-oncological patients were included. All patients had a DECT scan, acquired during the first week of treatment in either dual spiral or dual source mode (Siemens Confidence or Drive). For dual spiral, an 80 and 140 kVp scan were acquired sequentially, while for dual source an 80 and Sn140 kVp scan were acquired simultaneously (Sn: 0.4 mm extra tin filtration). Commercial software was applied to create stopping power ratio (SPR) map based on the DECT scan, and a one-to-one curve was used for SPR conversion in the treatment planning system. Two plans with different range uncertainties were optimized on the SPR map, keeping the beam and plan settings identical to the clinical plan (optimized on a single energy CT (SECT) scan). In the first plan, the clinically used range uncertainty of 3% was applied in the robust optimization and robust evaluation (3%-plan). In the second plan, a range uncertainty of 2% was used (2%-plan). Full optimization was performed for both plans, and changes to the optimization objectives were allowed between the two plans in an attempt to fully exploit the reduced range uncertainty to lower the dose to the organs-at-risk (OARs). The dose-volume-histogram (DVH) parameters were compared between the two plans. Two experienced radiation oncologists determined the relevant dose difference for each OAR. Moreover, the toxicity levels were determined using the ROCOCO performance scoring system (in ’t Ven et al. 2021).

All created plans were clinically acceptable; an example of dose distribution and the dose difference is seen in Figure 1. For all but three patients, a relevant dose difference (classified as a dose difference above 0.5 or 1 Gy depending on the OAR) was seen in one or more OARs in favor of the 2%-plan (Table 1). For 37% (10/27) of the patients the maximum dose to the brainstem was decreased below the relevant limit (0.5 Gy), and for 22% (6/27) hippocampus D40% was decreased below the limit. Furthermore, almost half of the patients (12/27) had a reduction in the toxicity level for one or two OARs, showing an actual clinical benefit for the patient.

Robust treatment plan optimization with reduced range uncertainty is feasible in clinical practice and allows for reduction of the OAR dose and toxicity level, which provides a rationale for further clinical implementation.

Based on these results, we have clinically introduced DECT-based proton treatment planning for our neuro-oncological patients, accompanied with a reduced range uncertainty of 2%.