Evaluation of the Possibility of Dose Realignment Adaptation by Shifting the Isocenter in Proton Beam Therapy for Pancreatic Cancer

Abstract

In pancreatic cancer, a tumor is surrounded by the gastrointestinal (GI) tract, which is subject to changes in location, shape, and contents. Due to these inter-fractional changes, proton beam therapy (PBT) for pancreatic cancer may result in unintentionally high doses to the GI tract. Daily adaptive re-planning can solve this problem, but is not yet established with PBT due to its resource intensive characteristics. This study aims to evaluate the GI tract dose using weekly computed tomography (CTw) and the possibility of dose realignment adaptation by shifting the isocenter (IC) of the PBT plan, which does not require re-planning. We retrospectively analyzed 6 consecutive patients with unresectable pancreatic cancer treated with real-time-image gated PBT using a fiducial marker. The planning CT was scanned at the natural expiration of respiration and a PBT plan of 60 GyE in 25 fractions (baseline plan, PLANbase) was created. The CTw images were acquired the day before start of PBT and once a week during the PBT course thereafter. The PLANbase was rigidly transferred to the CTw based on the relationship between the three-dimensional coordinates of the fiducial marker and those of the IC in the PLANbase. The PLANeval was created by recalculating the PLANbase on the CTw. We evaluated the doses to the stomach, duodenum, and intestines in the PLANeval according to the following criteria: Dmax of the stomach < 60 GyE, duodenum and intestines < 55 GyE, and D1cc of the stomach < 55 GyE, duodenum and intestines < 54 GyE. In addition, we investigated the GI tract dose realignment adaptation for the PLANeval with its IC shifted 2mm, 4mm, and 6mm in each of 6 directions (right, left, ventral, dorsal, cranial, and caudal), respectively. A total of 35 PLANeval were created for the CTw. In the PLANbase of the 6 patients, the average of Dmax and D1cc of the stomach, duodenum and intestines were 50.7 GyE (range, 46.7-53.6) and 50.0 GyE (45.0-53.2), 49.2 GyE (44.3-51.7) and 48.8 GyE (43.7-51.5), and 49.2 GyE (44.8-52.0) and 48.9 GyE (44.6-51.8), respectively. In the PLANeval, the average of Dmax and D1cc of the stomach, duodenum, and intestines were 53.3GyE (43.8-61.4) and 52.8 GyE (43.2-61.1), 51.0 GyE (36.1-60.0) and 50.3 GyE (35.4-59.8), and 52.5 GyE (36.6-61.0) and 51.9 (34.4-60.9) GyE, respectively. Twenty-two of the 35 PLANeval (63 %) did not meet at least one of the GI tract dose criteria. In 11 of 22 PLANeval with higher doses to the GI tract, the IC shift resulted in GI dose reductions and all dose criteria were met. The minimum amount of the IC shift required to meet the criteria was 2 mm for 8 plans and 4 mm for 3 plans. The remaining 11 PLANeval did not meet the criteria using dose realignment adaptation by shifting the isocenter. Adaptive replanning is necessary for PBT for pancreatic cancers due to excessive GI tract doses in more than 60% of the plans. Dose realignment adaptation by shifting the IC, which does not require re-planning, may be an option in adaptive treatment strategies.