Anatomic and biologic criteria improve recurrence predictions post-surgery for perihilar cholangiocarcinoma, enhancing patient selection.

• Anatomic borderline resectability alone failed to predict recurrence; adding CA19-9 biomarker separated patients into three distinct risk levels for recurrence-free survival.
• Two-year recurrence-free survival rates were 55.7%, 37.8%, and 25.3% for low-, intermediate-, and high-risk groups respectively.

This approach shifts decision-making towards a more comprehensive assessment of oncologic risk rather than relying solely on anatomy.