Long-Term Multi-Institutional Outcomes of 5-Fraction Ablative Stereotactic MR-Guided Adaptive Radiation Therapy (SMART) for Inoperable Pancreas Cancer With Median Prescribed Biologically Effective Dose of 100 Gy10

Document Type

Conference Proceeding

Publication Date


Publication Title

Int J Radiat Oncol Biol Phys


Purpose/Objective(s): Randomized trials have shown improved local control (LC) but no overall survival (OS) benefit with the addition of non-ablative radiation therapy (RT) dose compared to chemotherapy (CT) alone for pancreas cancer (PCa). Emerging data suggest that dose-escalated RT may improve LC and OS. A few studies suggest that stereotactic magnetic resonance-guided adaptive RT (SMART) can facilitate the safe delivery of ablative dose for inoperable PCa, although long-term outcomes are not well understood.

Materials/Methods: Inoperable PCa patients who received SMART were identified from the RSSearch Registry. Patients with < 3 months (mo.) follow-up after SMART were excluded. LC, progression free survival (PFS), and OS were estimated using the Kaplan-Meier method. LC was evaluated according to RECIST 1.1 criteria. Acute toxicity was considered within 90 days of SMART and evaluated by CTCAE v4 criteria.

Results: A total of 148 PCa patients were treated on a 0.35T MR LINAC across 3 institutions between 2018-2020. Median age was 68 years (range 47-91), and 73.6% had ECOG 0-1 performance status. Patients had locally advanced (57.4%), borderline resectable (29.1%), or medically inoperable (13.5%) disease. Median CA19-9 at diagnosis was 202.1 U/mL (range 0.9-21,281). Induction CT was delivered to 89.2% for a median 3.9 mo (range 0.2-11.3); FOLFIRINOX (52.7%) or gemcitabine/nab-paclitaxel (23.4%) were common. Median prescribed RT dose was 50 Gy (range 40-50) in 5 fractions, mostly in consecutive days (96.6%) and in breath hold (95.3%). Median biologically effective dose (BED10) was 100 Gy10. All patients were treated with real-time tissue tracking and automated beam gating without fiducial markers. An elective target volume was rarely used (25%). Pancreaticoduodenectomy was performed in 23% at a median 46 days (range 34-304) after SMART. Median follow-up was 16 mo. from diagnosis for all patients (range, 4-39). Median, 1-year, and 2-year LC was not reached (NR), 94.6%, and 83%, respectively. Median, 1-year, and 2-year PFS was 18 mo., 72%, and 35.9%, respectively. Median, 1-year, and 2-year OS was 26 mo., 82%, and 52.7%, respectively. Acute and late grade 3 toxicity possibly related to SMART occurred in 4.1% and 12.8%, respectively. There was no reported grade 4+ toxicity.

Conclusion: To our knowledge, this is the largest reported analysis of 5-fraction SMART for inoperable PCa. These data add to the evidence that ablative radiation dose may improve long-term outcomes including OS. Prospective evaluation of this novel approach is warranted with attention directed at optimizing patient selection, understanding the clinical significance of cumulative dose delivered across all adapted fractions, and assessing treatment response after the delivery of ablative dose.





First Page


Last Page