Impact of Stereotactic MR-Guided Adaptive Radiation Therapy on Early Clinical and Dosimetric Outcomes in Patients With Pancreatic Cancer.

PURPOSE/OBJECTIVE(S) Dose escalated stereotactic body radiotherapy (SBRT) may improve both local control and overall survival for patients with pancreatic cancer. However, escalated dose cannot be routinely offered because of the risk of injury to adjacent normal organs. Stereotactic magnetic resonance (MR)-guided adaptive radiation therapy (SMART) represents an innovation that may achieve safer delivery of ablative doses and improve outcomes. The goal of the present study is to identify specific dosimetric gains that can be achieved with MR-guided adaptive pancreatic SBRT and describe acute clinical outcomes. MATERIALS/METHODS We performed a single institution IRB approved retrospective analysis of 37 consecutive pancreatic cancer patients treated with 40 Gy in 5 fractions with SMART on a MR-linear accelerator using a light inspiration breath hold. The adaptive workflow starts with an "original" plan which is then adapted daily based on anatomical variation for each fraction. "Predicted" plan dosimetry is generated for each fraction by re-calculating the original plan on the re-contoured pre-treatment anatomy, estimating what would have been delivered without adaptive capabilities. A "re-optimized" plan is developed based on the anatomy of the day to improve target coverage and meet critical organ-at-risk (OAR) metrics, including stomach, duodenum, large bowel and small bowel V33 < 1cc. Acute GI toxicity rates were assessed according to CTCAE v5. RESULTS The median age of our cohort was 71 (range 35-87) with 23/37 (62%) patients having locally advanced disease and 14/37 (38%) having borderline resectable disease. Adaptive planning was used in all cases with a prescription goal of 40 Gy in 5 fractions prescribed to 90% of the target volume. The median CTV was 80.5 cc (range, 10-180cc) and the PTV was a 3mm expansion. Adaptive planning was required in 156/185 fractions to meet either the stomach (63.8%), duodenum (33.5%), large bowel (12.4%) or small bowel (11.4%) V33 < 1cc metric. Of the plans adapted for OAR sparing, 90 plans required adaptation for one gastrointestinal organ, 50 plans for two, 10 plans for three and 1 plan for four. For predicted plans exceeding the V33 < 1cc metric, the median V33 for duodenum was 2.475cc (range: 1.01-10.68cc), stomach was 2.67cc (range: 1.01-8.21cc), small bowel was 2.71cc (range: 1.14-9.54cc) and large bowel was 1.93cc (range: 1.05-8.36cc). All re-optimized plans met the V33 < 1cc metric. No local failure or acute grade 3 toxicities were observed with a median follow-up of 6 months (range 3-15). CONCLUSION Online adaptive re-optimization was necessary for the majority of fractions and improved plan dosimetry was achieved specifically for duodenum, stomach, large and small bowel, achieving low rates of acute toxicity. The clinical benefits of MR-guided adaptive pancreatic SBRT for dose escalation warrant further investigation.