Surgery After Neoadjuvant Stereotactic MRI Guided Adaptive Radiation in Pancreatic Cancer: Multi-institutional Toxicity and Survival Outcomes

Background: Favorable toxicity and survival outcomes after dose escalated stereotactic MR guided adaptive radiation therapy (SMART) have been recently published for locally advanced (LA) and borderline resectable (BR) pancreatic cancer. Perioperative morbidity and mortality are not well understood after ablative radiation therapy, which may temper enthusiasm for offering surgery. Objectives: The purpose of this study was to investigate survival and toxicity in resected pancreas cancer patients after neoadjuvant ablative SMART. Methods: In this IRB approved analysis, we retrospectively reviewed 33 consecutive patients with resectable, BR, and LA pancreatic cancer based on NCCN 2.2021 staging criteria who were treated at 2 institutions from 2017-2020 with neoadjuvant SMART 50 Gy in 5 fractions on a 0.35T MR Linac and later underwent definitive surgical resection. Overall survival (OS) and locoregional control (LRC) were evaluated by Kaplan-Meier method. Results: Median follow up was 22.4 months from diagnosis and 17.8 months from last day of RT. Most had BR (55%), otherwise initially resectable (33%) or LA (12%) pancreatic cancer. Median duration of induction chemotherapy was 3.5 (SD 1.6) months with most common regimens being FOLFIRINOX (74%), gemcitabine/abraxane (24%) and FOLFOX (3%). Performance status was ECOG 0, 1, 2 in 16 (48.5%), 12 (36.4%), and 5 (15.2%), respectively. Whipple was performed in 27 (82%) of patients, distal pancreatectomy in 4 (12%), and total pancreatectomy in 2 (6%). The median duration from SMART completion to surgery was 6.9 weeks (4.7-44.1). R0 resections were achieved in 28 (84.8%) of patients with the rest being R1, all in BR patients. Vascular resection/reconstruction was performed of the portal vein (PV) in 8 (24.2%) patients, SMV in 4 (12%), SMA in 1 (3%), and common hepatic artery in 2 (6%). Vascular resection/reconstruction was performed in all LA patients. Median OS, 1-year OS, and 2-year OS from diagnosis were 29.6 months, 93.8%, 81.5%, respectively. Median OS from RT was not yet reached; 1-year OS was 90.9%. LRC at 1 and 2 years was 97% and 93%, respectively. Radiation related acute and late grade 3+ gastrointestinal toxicity was seen in 2 (6%) and 2 (6%) patients. Post-op mortality at 30 and 90 days was seen 2 (6%) and 3 (9%) of patients with 1 death from GI bleed attributed to surgery and 1 death from hepatic ischemia related to PV resection. Conclusions: To the best of our knowledge, this is the first report suggesting that surgery for pancreas cancer after dose escalated 5-fraction SMART is feasible. Further clarification is needed with respect to ideal patient selection and timing for surgery, the safety of arterial versus venous resection/reconstruction, and histopathologic response after delivery of ablative versus non-ablative radiation dose