Surgery Versus Stereotactic Ablative Radiotherapy for Patients With Locally Advanced Pancreatic Cancer With Favourable Biological Response Following Induction Chemotherapy

The current evidence on the added survival benefit of surgery after induction chemotherapy is limited by strong selection bias, and international guidelines state that the longterm survival benefit of surgery in locally advanced pancreatic carcinoma (LAPC) remains uncertain. In parallel, stereotactic ablative radiotherapy (SABR), particulary MRI guided SABR, had emerged as a promising less invasive alternaitve, achieving high local control rates with limited toxicity. Therefore, this international, multicentre, open-label, phase 3 superiority randomized clinical trial assesses whethersurgery improves overall survival as compared to MRI-guided SABR. Primary objective is overall survival, secondary objectives include evaluation of failure free survival, quality of life, and serious adverse events.

• Pancreatic surgery (Procedure): Surgical exploration starts with a staging laparoscopy, unless already performed previsouly. Hereafter, exploration is typically performed via laparotomy. Although heterogeneity exist in surgical techniques, the following general surgical principles are considered state-of-the-art: 1. Artery-first approach in case of potential arterial involvement 2. Arterial divestment in case of radiological signs of (former) involvement without current signs of arterial wall ingrowth (and negative biopsy of peri-arterial tissue) during surgery to remove all (former) tumour tissue. 3. In case of vascular reconstruction that requires a patch or interposition graft, an autol-ogous (e.g., veins, peritoneal / falciform patch, splenic artery) or biological (e.g., bovine patch) graft is strongly preferred over synthetic prosthesis
• MRI guided stereotactic ablative radiotherapy (Radiation): MRI-guided SABR must be initiated within four weeks after randomization and is delivered in a hypofractioned regimen of 50 Gy in five fractions of 10 Gy, prescribed to 95% of the planning target volume (PTV). Treatment will be administered on alternative days, two or three times per week, with an intended overall treatment duration of two weeks (i.e., 14 days). All patients will undergo an MR (and CT scan at the department's discretion) for radiotherapy planning. The gross tumour volume (GTV) will be defined based on those imaging datasets and the restaging imaging. A PTV margin of 3mm is applied. Two MRI-guided SABR platforms are allowed. 1.5 Tesla MR-Linac and MRIdian system (0.35 Tesla MR-Linac). All patients will receive a proton pump inhibitor (oral) once daily during MRI-guided SABR, continued for three months afterwards to reduce gastrointestinal toxicity.

Primary outcomes

• Overall survival [Measured from the date of randomization to the date of documented death by any cause or the date of last follow up, whichever occurs first, assessed up to at 5 years of randomization.]: This will be measured in days. Participants receive systemic induction therapy before randomization, with a minimum duration of 3 months but no fixed duration. Therefore, overall survival is measured from the date of randomization.

Secondary outcomes

• Failure free survival [FFS is measured in days from date of randomization to date of first documented failure (whichever occurs first). Participants without documented failure will be censored at the date of last follow-up. FFS will be assessed up to 5yrs after randomization]: To compare the failure free survival (FFS) between surgical exploration with intention for resection and MRI guided stereotactic ablative radiotherapy. Failure is defined as any of the following: local relapse/recurrence, local progression, regional relapse/recurrence, regional progression, development of distant metastases, or occurrence of a second primary pancreatic cancer.
• Quality of life [First two years from randomization]: functioning, and symptoms scores and the summary score during the first two years from randomization, using the EORTC QLQ-C30 (version 3.0). The higher the score, means more complaints. Scale 1-4. 30 questions.
• serious adverse events [Six-months serious adverse events (SAEs) that occur from randomization]: defined following the NCI Common Terminology Criteria for Adverse Events (CTCAE) (version 5.0) (grade ≥III),71 expressed as the number of patients with serious adverse event(s), the number of events, and type of events.
• Global quality of life [First two years from randomization]: Functioning, and symptoms scores and the summary score during the first two years from randomization using the EORTC PAN26 questionnaires. The higher the score, the more complaints. Scale 1-4. 25 questions.

Other outcomes

• in hospital/90 day major morbidity [Measured from intervention until 90 days after intervention or discharge.]: defined as Clavien-Dindo classification grade ≥IIIa
• in hospital/90 day clinically relevant pancreatic surgery specific complications [Measured from surgery until 90 days after intervention or discharge.]: including post pancreatectomy hemorrhage, postoperatove pancreatic fistula, post pancreatectomy acute pancreatitis, delayed gastric emptying, chyle leak, and bile leak (grade B/C)
• R0 resection rate [Resection rate should be assessed by a pathologists within 2 weeks from pancreatic surgery]: R0 resection status will be determined based on the final histopathological examination of the pancreatic resection specimen. Defined according to the 1mm rule following th royal collega of pathologists and international collaboration on cancer reporting, except for the anterior surface where a 0mm rule will be applied, and strtified for surgical and pathological margins in accordance with the 2025 ISGPS-ISGPP statement
• Overall survival [From the date of imaging-based diagnosis until death from any cause or the end of follow-up, assessed up to 5 years after randomization]: This will be measured in days.

Sex: All, ages: 18 Years to any

Inclusion Criteria:

* Age: ≥18 years.
* Diagnosis:

Patients diagnosed with pathology proven (WHO 6-7) LAPC according to the 2025 NCCN guideline. Of note, WHO-3 is allowed for inclusion in case of evidence for neoplasia by molecular analyses if cystic lesions are absent. Chemotherapy may be started without pathologyproof after a least two attempts via fine-needle biopsy if the patient has a high radiological and clinical suspicion for pancreatic adenocarcinoma and consensus has been reached in the responsible multidisciplinary conference. Furthermore, patients must meet the following established A-B-C parameters after at least three months systemic induction therapy:

* Anatomy:
* RECIST non-progressive disease on cross-sectional imaging;
* No radiological downstaging to a borderline resectable tumour;
* Surgical resection seems potentially feasible according to the local multidisciplinary team*;
* Radiation is technically feasible according to the local multidisciplinary team;
* Biology:
* Serum CA19-9 reduction >40% with an absolute end value <300 U/mL after induction chemotherapy**.
* NB1. If non-elevated serum CA19-9 at diagnosis, metabolic response on 18F-FDG PET scan defined as ≥30% reduction of SUVmax if >4 at baseline#.
* NB2. If non-elevated serum CA19-9 at diagnosis** and on 18F-FDG PET scan an SUVmax ≤4 at baseline, non-elevated serum CEA at restaging is required (i.e., ≤5 ng/ml).
* NB3. If serum CA19-9 after chemotherapy is >150 U/ml, a negative staging laparoscopy is required prior to randomization (unless a staging laparoscopy was already performed prior to starting systemic therapy). This is based on the PREOPANC-4 cohort wherein such patients had a 23% risk of peritoneal metastases (vs. 5% with <150 U/ml).
* Condition:
* ECOG/WHO performance status 0-1 at restaging, being fit for surgery and chemotherapy

NB. If a systemic treatment switch is performed in the setting of induction therapy with the indication of (biochemical) disease progression, restaging according to the above-mentioned criteria are performed with using the A-B-C features between the first- and second-line systemic induction therapy as baseline.

* Patients who require a resection of the superior mesenteric artery (SMA) are only allowed for inclusion by centres where this is routine practice, as evidenced by a volume of on average ≥15 SMA resections for periampullary cancer over the preceding three years, considering the high risks of complication-related mortality. This also means that patients with an SMA string sign (i.e. irregularity of the arterial wall) on cross-sectional imaging after systemic induction therapycannot be randomized at centers. At baseline, serum CA19-9 is measured with a total bilirubin as low as possible and measured as short as possible before the start of chemotherapy. At restaging, serum CA19-9 is ideally measured in the same lab due to different essays with widely diverging results.

#Preferably, FDG-PET scan is performed with EARL reconstruction protocol 1 or 2.

\- Informed consent: Competent and willing to provide informed consent in accordance with the requirements of the ethics committee.

Exclusion Criteria:

\- Prior treatment: Prior systemic induction therapy combined with any type of radiother-apy. Radiotherapy is not used as induction therapy as it has not been proven beneficial and considering the potential increased risk of major morbidity, although particularly seen with external beam radiotherapy. The only other possible indication for radio-therapy that is allowed in this trial is a single fraction of 8 Gray in case of tumour bleed-ing.

\- Metastatic disease: Presence of metastatic disease

\- Comorbidities: Clinical deterioration during systemic induction therapy or increase of suspected tumour-related pain after excluding other potential causes of pain (e.g., pancreatitis, stent dysfunction).

\- History of other malignancies: A past medical history of malignancy, except under the following conditions: no evidence of disease, diagnosis more than three years ago before the diagnosis of pancreatic cancer, and life expectancy of more than five years from date of inclusion. Of note, a past medical history of non-melanoma skin cancer, pancreatic neuroendocrine tumour <2 cm, and gastrointestinal stroma tumour (GIST), or any other cancer with a life expectancy beyond five years are no exclusion criteria. Lastly, lesions on CT thorax that are too small to characterize are not considered met-astatic disease. The last exclusion criterium is tumours directly invading gastrointestinal hollow organs, except the duodenum.

• Study Protocol and Statistical Analysis Plan (PDF, 2026-05-26)

• Sterre de Vet, MD, Phd, Study director, AmsterdamUMC
• Floor Fabels, Md, Phd, Study director, EramusMC