Purpose/Objective(s): To quantify the interfractional change in ITV volume for pancreatic malignancies due to changes in respiratory motion.

Materials/Methods: Six patients treated with SBRT (33 Gy in 5 fractions) for pancreatic adenocarcinoma (N=5) or metastasis to the pancreas (N=1) were identified for this study. Our clinical workflow at the time of treatment of these patients required a 4DCT control scan prior to each fraction for IGRT purposes. For this study, a single GI attending physician re-contoured the ITV on all 4DCT datasets obtained for each fraction for each patient, for a total of 29 extra contour sets (the 4DCT for one fraction of one case was not available). In order to evaluate how much of the variation in ITV volume is due to operator inconsistency in volume definition, an intra-user study was performed where the same dataset was contoured five times, with at least one day between each contouring session. To further assess whether ITV differences were due to differences in breathing, inter-fractional differences in 4D breathing traces and the motion of nearby surgical clips were examined for case 6 and compared with changes in the ITV volume.

Results: For the intra-user study, the average volume change was 0.1% (1.7% standard deviation) for the five recontours. Table 1 summarizes the contoured ITV volumes as well as the percentage change from the simulation volume. While the average change in ITV from the simulation volume was 0.5%, the standard deviation (SD) was a 10%, with the greatest reduction in volume being 14% and the greatest increase in volume being 20%. For one patient (case 6), the average ITV change was an increase of 15% (SD 4%). Breathing traces and nearby clip motion correlated with change in ITV volume.

Conclusion: There are non-trivial changes in ITV volume over the course of pancreas SBRT treatment. Most of this change is likely attributable to changes in respiratory motion. Given the very small ITV to PTV margins commonly used for pancreatic SBRT (3mm), our results indicate that target coverage may be compromised for some patients. We are presently investigating what margin would be required to fully cover the target as well as performing a dosimetric study to determine what degree of target coverage was delivered for these patients.

	ITV volume (cc)						ITV volume change relative to simulation (%)
Case #	Sim	FX1	FX2	FX3	FX4	FX5	FX1	FX2	FX3	FX4	FX5
1	25.6	23.2	22.0	23.0	26.0		-9%	-14%	-10%	2%
2	68.2	73.5	78.8	71.0	75.6	76.5	8%	16%	4%	11%	12%
3	66.8	62.9	63.3	60.6	64.1	68.1	-6%	-5%	-9%	-4%	2%
4	18.2	16.7	16.0	16.9	19.4	16.2	-8%	-12%	-7%	6%	-11%
5	79.4	73.0	78.1	75.5	77.7	73.5	-8%	-2%	-5%	-2%	-7%
6	41.5	47.1	47.2	49.7	48.4	45.8	13%	14%	20%	17%	10%

Table 1 : Summary of ITV volumes from each 4DCT dataset. FX= Fraction