2185 - Do Uniform Fasting Instructions Lead to Reproducible Stomach Volumes that Decrease the Proximity of the Stomach to the Target for Pancreatic Cancer Patients?
Presenter(s)
E. Nguyen1, D. Rusu1, J. L. Dolan2, and P. J. Parikh3; 1Henry Ford Health System, Detroit, MI, 2Department of Radiation Oncology, Henry Ford Health, Detroit, MI, 3Henry Ford Health - Cancer, Detroit, MI
Purpose/Objective(s): Conventional wisdom is to recommend fasting prior to simulation and treatment of patients with pancreatic cancer to improve stomach volume reproducibility and decrease the amount of stomach near the target. This study hypothesizes that uniform fasting instructions result in limited variation in stomach volume throughout treatment and therefore results in consistent stomach volume within the treatment region.
Materials/Methods: 74 pancreatic cancer patients with a median age of 72, were treated with 50 Gy in 5 fractions using daily MR-guided adaptive radiation therapy. Patients were given instructions to fast for 2 hours before treatment at several time points; consultation, simulation, with reminders during treatment. The stomach was contoured at simulation and each of the five fractions. The volume of the overall stomach (stomach), as well as the volume of the stomach within 3cm of the PTV (stomach_3cm) were collected. Patients were grouped into three clusters using SimpleKMeans clustering based on their standard deviation in stomach volume across treatments. The three clusters represented patients with successful, variable, and unsuccessful fasting. The significance of the simulation volumes and standard deviations between each cluster was calculated using the Mann–Whitney U test. The correlation between the stomach_3cm and full stomach volumes was calculated on a total of 432 fractions that had clinically available 3 cm structures.
Results: There was a total of 443 full stomach volumes. Of the 74 patients, 42% were classified as successful at fasting with a significantly smaller standard deviation than the other two groups (P < 0.00001). The median stomach volume and standard deviations of each group can be seen in the Table, (* showing significance). The volume at simulation of the successful fasting cluster was significantly different from the variable and unsuccessful clusters (P < 0.001). There was a total of 432 fractions that had stomach_3cm volumes available; due to a few patients not having the stomach close to the pancreas. The correlation coefficient of the stomach and stomach_3cm was only 0.28, (R2 = 0.08).
Conclusion: Uniform fasting instructions did not result in uniform stomach volumes for the majority of patients. Patients with a simulation stomach volume of greater than 183 ± 24 cc, are likely to be unsuccessful at reproducible stomach volumes during treatment. Moreover, the total stomach volume was not highly correlated with the amount of the stomach within 3cm of the target. The conventional wisdom of instructing patients to fast before simulation and treatment may need revisiting.
Abstract 2185 - Table 1| Fasting Cluster | % of patients | Median SIM volume (cc) | Median St. Dev (cc) |
| Successful | 42 | 183* | 24* |
| Variable | 43 | 255 | 74* |
| Unsuccessful | 15 | 238 | 168* |