Room: Exhibit Hall
Purpose: To evaluate the feasibility of using IMPT to treat locally advanced pancreatic cancer (LAPC) patients.
Methods: We analyzed 8 LAPC patients who have been previously treated using hypofractionated IMRT (67.5 Gy in 15 fractions). IMPT plans were generated for each patient using an analytical dose calculation algorithm from a commercial treatment planning system (TPS). Four or five fields, typically composed of posterior-oblique and right-oblique beams, but in some cases also anterior-oblique, were used. Fields were robustly optimized using multi-field optimization with 0.3 cm spot spacing, and 0.5 cm isocenter-shift or 3.5% range uncertainty in combination (8 perturbations) to cover 100% of the GTV with 67.5 Gy and meet OAR constraints. The plansâ€™ DICOM files were exported into a fast dose calculation (FDC) engine. The FDC uses a tracking-repeating algorithm, which better accounts for density heterogeneities and provides more accurate dose estimations than the TPS. The FDC was used to re-calculate the doses using constant RBE (RBE=1.1) and two different variable RBE models (RBE1 and RBE2).
Results: GTV coverage was satisfactory for IMPT versus IMRT (90% ranging [84.0-99.7%] versus 90% [73.2-97.3%], respectively). The re-calculated dose distributions using FDC agreed with TPS (gamma 2%/2 mm of 98.7% [97.4-99.8%]). The variable RBE models also provided satisfactory coverage (GTV mean dose: 69.8 [67.6-71.8] Gy for IMRT; 68.6 [67.6-69.5] Gy for TPS; 67.9 [66.9-68.4] Gy for FDC-RBE=1.1; 66.8 [66.1-67.8] Gy for FDC-RBE1; and 69.4 [68.4-70.8] Gy for FDC-RBE2). OAR doses were much lower for IMPT versus IMRT â€“ even when variable RBE models were employed (e.g. 1 cm3 stomach: 25.3 [0.9-36.5] Gy for FDC-RBE=1.1 versus 41.6 [34.0-47.0] Gy for IMRT).
Conclusion: Our results suggest that IMPT has the potential to lower the doses to OAR in LAPC compared to IMRT. This indicates that proton therapy could be used to escalate the doses even more.