Dosimetric Impact of Gastrointestinal Tract Variability during Online Adaptive MR-Guided Hypofractionated Stereotactic Ablative Radiotherapy (SABR) to the Pancreas

Abstract

<h3>Purpose/Objective(s)</h3> To evaluate the impact of intra-fraction gastrointestinal (GI) tract motion on dose delivered during online adaptive MR-guided stereotactic ablative radiotherapy (SABR) to the pancreas using a 0.35 T MR-Linac. <h3>Materials/Methods</h3> Twelve patients receiving a dose of 40 Gy in 5 fractions with online daily adaptations were included in this analysis. Eight of 12 had tumor in the head of the pancreas. The average PTV volume was 66.5 cc (11.1-115 cc). Treatment was delivered in breath-hold with real-time tracking and gating. High resolution 3D images were acquired at the beginning of each fraction (MRI-pre) and following treatment delivery (MRI-post). Impact of intrafraction GI tract (stomach, duodenum, small bowel, and large bowel) motion was evaluated by propagating the pre-treatment baseline plan to MRI-pre following contour editing (non-adapted plan). A new adapted plan is generated on MRI-pre based on the new contours (adapted-pre plan). Subsequently, the adapted plan is propagated onto MRI-post and contours were edited accordingly (adapted-post plan). PTV coverage and GI tract doses were evaluated on MRI-pre and MRI-post. Wilcoxon signed-rank test was used to analyses the data. <h3>Results</h3> Without daily adaptation, 73% of pre-treatment plans did not meet GI tract dose constraints. With adaptation, GI tract dose tolerance violations were observed in 0% and 37% of cases on adapted-pre and adapted-post plans respectively. There were statistically significant reductions in GI tract receiving 36 Gy and 33 Gy between non-adapted vs adapted-post plans (V36 Gy - 0.19 vs 0.08 cc, V33 Gy – 0.70 vs 0.39 cc, p=0.0067 and 0.0072 respectively) and adapted-pre vs adapted-post plans (V36 Gy – 0.00 vs 0.08 cc, V33 Gy – 0.07 vs 0.39 cc, p=0.0004 and p=0.0002 respectively). There was no statistically significant difference in PTV V100 coverage between pre-treatment, non-adapted on adapted-pre and adapted-post plans. <h3>Conclusion</h3> Due to changes in anatomy during treatment, GI tract dose violations were observed in about a third of patients on the post treatment MRI. However, the magnitude of the dose violations was smaller than the benefit gained from performing daily online adaptive treatment. Longer term outcome data is required to assess the impact of these on GI toxicity. In view of this, caution needs to be taken in dose escalation trials.