Gastrointestinal Motility-Induced Interplay in Pancreas Proton Therapy: Motion Simulation and Dosimetric Impact.

Abstract

To simulate motion and evaluate its dosimetric interplay from gastrointestinal (GI) motility in pancreas pencil-beam scanning (PBS) proton therapy, and to assess how fractionation and motility amplitude affect target coverage and organ-at-risk (OAR) doses. A novel, physiology-informed GI motion simulator was developed to quantify GI motion by generating time-resolved motion modes (peristalsis, rhythmic segmentations, and high-amplitude propagated contractions (HAPC)) parameterized along organ centerlines (stomach, duodenum, and small bowel). The simulator was integrated with an in-house 4D dose accumulation tool to evaluate dose interplay between GI motion and PBS delivery. Seven clinically treated pancreatic adenocarcinoma patients receiving proton PBS were studied to assess correlations between motion characteristics and the resulting dosimetric interplay effects. Two motion settings were studied: MODERATE (small magnitude) and MAX (large magnitude). Endpoints were dosimetric differences (Δ) relative to static plan (no GI motion), e.g., CTV D95% and OAR D2%. Single-fraction (1-fx) and full-treatment with multiple fractions (full-tx) deliveries were compared. Dosimetric Δ and dose range (defined as the difference between the maximum and minimum values across motion scenarios) are reported as percentages relative to the prescribed dose (mean [min, max] in %). GI motility alone produced measurable interplay effect that degraded the target coverage and increased critical OAR dose. For 1 MAX/1-fx (1 fraction with MAX motion), CTV D95% changed by -1.34% [-5.52%, +0.08%] across prescription doses ranging from 30 to 59.4 Gy. For OARs, relative dose ranges were 6.90% [4.48%, 13.45%] for duodenum, 6.74% [0.43%, 25.57%] for small bowel, and 8.47% [3.08%, 12.63%] for stomach. MAX motion yielded larger Δ and wider dose ranges than MODERATE. For full treatment, conventional fractionation substantially reduced the dose interplay effect and pulled Δ toward zero, whereas hypofractionation provided only partial averaging. GI motility is a dosimetrically relevant, geometry-contingent source of interplay in pancreas PBS. We present the first plan-specific replay of clinical PBS deliveries on motility-only time-resolved anatomies. Fractionation mitigates this interplay effect.