2054 - Stereotactic Body Radiotherapy after Whipple Resection: Feasibility, Toxicity and Local Control

03:00pm - 04:00pm ET

Poster Hall - Exhibit Hall A

Screen: 31

POSTER

Presenter(s)

Sarah Finton, BA - David Geffen School of Medicine, Los angeles, CA

S. Ebrahimi¹, S. C. Finton², A. Raldow¹, and T. C. Wu¹; ¹Department of Radiation Oncology, University of California, Los Angeles, Los Angeles, CA, ²David Geffen School of Medicine at UCLA, Los Angeles, CA

Purpose/Objective(s):

Locoregional recurrence following pancreaticoduodenectomy remains common despite optimal adjuvant therapy. SBRT in the post-Whipple setting is approached with caution due to altered anatomy, which complicates adherence to OAR constraints and risks high-grade gastrointestinal toxicity. We hypothesized that stereotactic MR-guided adaptive radiation therapy (SMART) could mitigate these challenges and enable the safe delivery of SBRT in this population. We report local control and toxicity outcomes among patients treated with SMART following Whipple resection.

Materials/Methods:

We retrospectively reviewed patients treated with SMART after Whipple (2016–2025) at a single institution. Daily online adaptive replanning was performed as needed for deviation from predicted OAR constraints or target coverage. Outcomes were estimated by Kaplan–Meier. Analyses were performed in R.

Results:

19 patients with locoregional recurrence after curative-intent pancreaticoduodenectomy for pancreatic adenocarcinoma (89.5%) or cholangiocarcinoma (10.5%) underwent salvage SMART. Median age was 68.2 years (60.5–75.3); 57.9% were female and 84.2% were White. ECOG performance status 0–1 was observed in 94.7%. R0 resection was achieved in 84.2%. Node-positive disease occurred in 44.4% (median 5 nodes). Most tumors were pT2 (52.6%); pN0, pN1, and pN2 comprised 52.6%, 26.3%, and 21.1%, respectively.

Median time from Whipple to SMART was 21.9 months (IQR 13.1–48.1); among re-irradiated patients (15.8%; all with prior chemoradiation), the interval from prior RT was 44.6 months (13.1–99.2). Median GTV/CTV and PTV volumes were 46.9 cc (18.9–52.0) and 68.5 cc (34.4–78.6), respectively. The median dose was 50 Gy (40–50) in 5 fractions. Targets included gross tumor within the postoperative bed (78.9%) and relapsed nodal disease (21.1%). Adaptive planning was used in 60% of fractions. Acute toxicity was grade 0–1 in 73.7% and grade 2 in 21.1% - mostly commonly fatigue, nausea, and pain. One patient experienced a grade 3 biliary stricture requiring hospitalization, unrelated to radiation; no grade 4 events occurred.

Median follow-up was 17.3 months (9.6–NR). Four patients developed out-of-field regional failure (three with synchronous distant disease). Seven developed distant metastases, all involving the liver, with lung (n=3) and bone (n=2) also observed. Median progression-free survival (PFS) was 6.6 months (3.2–NR); 6-month and 1-year PFS were 55% (36–83.9%) and 30.9% (13.7–69.7%), respectively. There were no infield recurrences. One- and two-year competing risk–adjusted locoregional control (LC) rates were 58.1% (95% CI 30.9–85.2%) and 38.8% (10.8–67.8%).

Conclusion:

Salvage SMART for locoregional recurrence following pancreaticoduodenectomy was delivered safely, with low rates of clinically significant toxicity and no radiation-attributable grade =3 events. Treatment was feasible in this high-risk population with acceptable LC.