Click here to


Are you sure ?

Yes, do it No, cancel

Dosimetric Impact of Interfraction Anatomical Variations On Breath-Hold SBRT for Pancreatic Cancer

J Niedzielski*, S Ng, S Beddar, R Martin, E Holliday, G Smith, B Minsky, A Koong, C Taniguchi, P Das, E Koay, G Sawakuchi, UT MD Anderson Cancer Center, Houston, TX


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: To evaluate the dosimetric impact of interfraction anatomical variations on pancreas SBRT.

Methods: We evaluated 10 pancreatic cancer patients treated with SBRT (30–40 Gy in 5 fractions) using daily CT-on-rails image guidance and breath-hold motion management. Planned and delivered doses were compared for the GTV (100% coverage dose and mean dose), as well as the small bowel, duodenum, and stomach (D0.3cc, D3.0cc, V20, and V35). Delivered doses were calculated by: (i) deformably propagating each structure’s planning segmentation to each fraction’s CT-on-rails imageset (adjusted by physician as needed); (ii) re-calculating treatment plan dose on each fraction’s CT-on-rails imageset; (iii) accumulating the 5 fractional DVHs into a single DVH for comparison with the original planned dose. Paired Wilcoxon signed-rank test (p<0.05) was used to analyze differences between planned and accumulated dose metrics. Prior to each fraction, the physician and physicist shifted the couch to align to the GTV or spare OARs, based on planned isodose lines, which were superimposed on daily anatomy.

Results: Four patients had small bowel D3.0cc >20% different than planned. However, the mean difference between the planned and delivered D3.0cc was -10.2% (range:-48.1–34.3%). Seven patients received significantly higher delivered duodenum V35 (mean difference=97.5%, range:0–200%, p=0.02). Six patients received higher delivered stomach V20 (mean=69.6%, range:-146.8–200%). Eight patients had delivered OAR doses that exceeded our clinical constraints. For dose of 100% GTV coverage, delivered doses were 6.4% lower (range:-31.7–4.5%). Mean planned and delivered GTV doses were not significantly different (p=0.63). Elevated GI toxicity incidence was not discovered.

Conclusion: Delivered and planned doses can vary greatly for pancreas SBRT treatments. Thus, the use of planned dose for daily image guidance may not be dosimetrically robust for pancreas SBRT treatments. Daily online adaptation could be employed to account for anatomical variations of luminal GI anatomy and optimize SBRT treatments.


Image-guided Therapy, Targeted Radiotherapy, Organ Motion


Not Applicable / None Entered.

Contact Email