This study aims to evaluate the safety and preliminary efficacy of NALIRIFOX combined with adebrelimab and PULSAR as first-line treatment for locally advanced unresectable or metastatic pancreatic ductal adenocarcinoma (PDAC). Additionally, it will explore potential predictive and efficacy-related biomarkers.

After confirmation of eligibility, enrolled patients will undergo radiation CT simulation and planning per standard of care. IV contrast will be administered with CT simulation at the treating physician's discretion though is not required.

Inclusion Criteria:

• Age: 18-75 years, regardless of gender.
• Histologically confirmed pancreatic ductal adenocarcinoma (PDAC).
• Previously untreated, locally advanced unresectable or metastatic PDAC, with at least one measurable lesion (RECIST v1.1) not previously irradiated.
• ECOG Performance Status (PS): 0-1.
• Expected survival ≥ 3 months.
• Willing and able to comply with study procedures, treatment, and follow-up.
• No contraindications to radiotherapy.
• Adequate organ function: WBC ≥ 2.5×10⁹/L, ANC ≥ 1.5×10⁹/L; Platelets ≥ 75×10⁹/L; Hemoglobin (HGB) ≥ 90 g/L (no transfusion or EPO dependence within 7 days); Total bilirubin (Tbil) ≤ 1.5×ULN; ALT/AST ≤ 5×ULN;Albumin ≥ 30 g/L; INR ≤ 1.5×ULN; Serum creatinine (Cr) ≤ 1.5×ULN Urine protein ≤ 1+
• HBsAg-positive patients must have HBV-DNA ≤ 1×10³ IU/mL (copies/mL). If HBV-DNA ≥ 1×10³ IU/mL, patients may still be eligible if chronic HBV is stable and not expected to increase risk, per investigator assessment.
• Voluntary participation with signed informed consent form.

Exclusion Criteria:

• History of severe hypersensitivity to chimeric, human(ized) antibodies, or fusion proteins.
• Pregnant or breastfeeding women, or men/women of childbearing potential unwilling/unable to use effective contraception during the study.
• Other malignancies within 5 years, except: Malignancies treated with curative intent and no known active disease for ≥5 years with low recurrence risk; Adequately treated non-melanoma skin cancer or lentigo maligna without disease evidence; Adequately treated carcinoma in situ (e.g., cervical, breast) with no current disease.
• Symptomatic moderate/severe pleural effusion or ascites.
• Active bleeding or coagulopathy (PT >16s, APTT >43s, INR >1.5×ULN), bleeding tendency, or current use of thrombolytics/anticoagulants/antiplatelets.
• GI bleeding within 6 months or high bleeding risk (e.g., active ulcer with occult blood++). If occult blood+ persists, endoscopy required.
• High-risk esophageal/gastric varices needing intervention.
• History of drug abuse, psychiatric disorder, or inability to abstain.
• Solid organ/bone marrow transplant, or active autoimmune disease requiring systemic treatment within 2 years.
• Immunodeficiency or HIV infection.
• Objective evidence of pulmonary fibrosis, interstitial lung disease, pneumoconiosis, radiation-/drug-induced pneumonitis, or severely impaired pulmonary function.
• Major surgery within 4 weeks or minor surgery within 1 week (e.g., tooth extraction).
• Vaccination within 30 days before the first dose.
• Abdominal fistula, GI perforation, or abscess within 4 weeks.
• Any clinically significant abnormality affecting safety per investigator, including: Active infection requiring systemic therapy; Uncontrolled diabetes/hypertension (BP >140/90 mmHg despite ≤2 antihypertensives); Myocardial infarction within 6 months; Thyroid dysfunction (>NCI CTCAE v4.0 Grade 1).
• Other conditions deemed ineligible by the investigator.

IPD will not be shared in order to protect participant confidentiality, in compliance with local data protection regulations and ethical committee requirements. Additionally, IPD sharing is restricted under contractual agreements with study sponsors/partners, who retain data ownership for independent analyses. Given the complexity and size of the dataset (e.g., genomic/imaging data), anonymized IPD sharing is technically unfeasible without risking data integrity. Furthermore, to safeguard intellectual property rights and permit ongoing secondary analyses by the research team, IPD will be retained internally.