Session Item

August 29
16:45 - 17:45
Room 1
Proffered papers 25: Upper GI
Karin Haustermans, Belgium;
Thomas Brunner, Austria
Proffered papers
17:25 - 17:35
Long-term outcomes of MR-guided SABR & on-table adaptive replanning for unresectable pancreas cancer
Michael Chuong, USA


Long-term outcomes of MR-guided SABR & on-table adaptive replanning for unresectable pancreas cancer

Michael Chuong1, Kathryn Mittauer1, Roberto Herrera1, Tino Romaguera2, DIane Alvarez1, Rupesh Kotecha1, Matthew Hall1, Adeel Kaiser1, John Bryant3, James McCulloch1, Antonio Ucar4, Fernando De Zarraga4, Santiago Aparo4, Horacio Asbun5, Ramon Jimenez5, Govindarajan Narayanan6, Sarah Joseph4, Alonso Gutierrez1, Muni Rubens7

1Miami Cancer Institute, Radiation Oncology, Miami, USA; 2Miami Cancer institute, Radiation Oncology, Miami, USA; 3Moffitt Cancer Center, Radiation Oncology, Tampa, USA; 4Miami Cancer Institute, Medical Oncology, Miami, USA; 5Miami Cancer Institute, Surgical Oncology, Miami, USA; 6Miami Cancer Institute, Interventional Oncology, Miami, USA; 7Miami Cancer Institute, Office of Clinical Research, Miami, USA

Show Affiliations
Purpose or Objective

Radiation therapy (RT) dose escalation may improve clinical outcomes for patients with inoperable pancreas cancer (PCa), although can potentially cause severe toxicity due to the proximity of gastrointestinal luminal organs.  A few studies have suggested that significant dose escalation may be safely achieved using magnetic resonance image (MRI) guidance with on-table adaptive replanning although long-term outcomes are not well understood

Material and Methods

A single institution retrospective analysis was performed of 50 consecutive patients with inoperable non-metastatic pancreas adenocarcinoma treated on a 0.35T-MR Linac without fiducial markers from 2018-2020.  Median age was 70 years (range 35-91).  ECOG performance status was mostly 0-1 (94%).  Locally advanced disease was most common (82%), otherwise borderline resectable (8%) or medically inoperable (10%).  Median CA19-9 at diagnosis was 122.2 U/mL (range, 0.9-12,868.6).  Induction chemotherapy (CT) was routine (94%), usually FOLFIRINOX (58%) or gemcitabine/nab-paclitaxel (26%), for a median 4.2 months (range, 0.2-14.5).  Local control (LC) was defined using RECIST 1.1 criteria.  Acute (<3 months from SABR) and late toxicity were assessed as per CTCAE v5 criteria.  Outcomes were assessed from the date of PCa diagnosis.


Median prescribed dose was 50 Gy (range 40-50), median biologically effective dose (BED10) was 100 Gy10.  The median PTV volume was 111.0 cm3 (range 11.3-374.33).  Treatment in breath hold was common (90%).  On-table adaptive replanning was performed for a median 5 fractions (range 1-5) in 96% of patients.  Post-SABR therapy included pancreaticoduodenectomy (14%), irreversible electroporation (10%), and/or CT (52%).  Negative margins were achieved in 6/7 patients who had surgery.  Median follow-up was 18 months (range 7-39 months) from diagnosis.  Median, 1-yr, and 2-yr estimated LC was not reached, 97.8%, and 88.9%, respectively.  Median, 1-yr, and 2-yr estimated progression free survival (PFS) was 16 months, 67.7%, and 25.7%, respectively.  Median, 1-yr, and 2-yr estimated overall survival (OS) was 21 months, 87.9%, and 50%, respectively.  Induction CT duration >3 months was the only predictor of improved PFS (hazard ratio [HR] 0.358; 95% confidence interval [CI] 0.141-0.913; P=0.031) and OS (HR 0.288; 95% CI 0.104-0.799; P=0.017) on multivariate analysis.  Acute and late grade 3+ toxicity rates were 2% and 10%, respectively.


To our knowledge, this is the largest series of 5-fraction MRI-guided stereotactic ablative body radiation therapy (SABR) with on-table adaptive replanning for inoperable PCa.  Our 2-yr estimated OS of 50% is significantly higher compared to historical outcomes of ~20% from CT +/- non-ablative RT.  Additional study is needed to improve our understanding about optimal patient selection, treatment response assessment, and the need for additional therapy after SABR.