Anatomic-Biologic Borderline Resectability Criteria Improve Recurrence Risk Stratification in Perihilar Cholangiocarcinoma

BACKGROUND: Borderline resectability in perihilar cholangiocarcinoma (pCCA) has largely been defined anatomically, although anatomy alone may not capture oncologic risk. This study evaluated whether biologic burden improves recurrence stratification after curative-intent resection of pCCA. METHODS: Using an international, multi-institutional database, patients undergoing resection without neoadjuvant therapy for non-metastatic pCCA were identified. Literature-based borderline resectability was defined as lymph node metastasis with portal vein/hepatic artery involvement or Bismuth type IV disease. Using a 200-U/mL carbohydrate antigen (CA19-9) cutoff, a three-tier anatomic-biologic borderline resectable (AB-BR) grouping (low, intermediate, high) was derived. The outcome of interest was recurrence-free survival (RFS). RESULTS: Among 239 patients (median age, 67.0 years, interquartile range [IQR], 58.0-74.0 years), 38.9% (n = 93) of patients were anatomically borderline resectable (BR). The anatomy-only BR definition did not stratify RFS or overall survival (OS) (both p > 0.05), whereas CA19-9 ≥200 U/mL stratified 2 year RFS within both resectable (55.7% vs. 32.6%) and BR (44.5% vs. 25.3%) groups (all p < 0.05). Four-group analysis demonstrated overlap between resectable/high CA19-9 and BR/low CA19-9, supporting collapsing the categories into AB-BR tiers. The 2 year RFS was 55.7, 37.8, and 25.3% in the low-, intermediate-, and high-risk AB-BR groups, respectively (p = 0.002). The 2 year area under the curve (AUC) value was improved by AB-BR for RFS versus anatomy alone (0.612 vs 0.537). After multivariable adjustment, high-risk AB-BR remained independently associated with higher hazards of recurrence or death (adjusted hazard ratio, 1.86; 95% confidence interval, 1.15-3.00). CONCLUSION: Anatomic BR alone provided limited prognostic stratification. Incorporating CA19-9 improved recurrence prediction and identified an intermediate-risk group not captured by anatomy alone, supporting biologic resectability in pCCA. The AB-BR framework may help move surgical decision-making beyond technical resectability alone toward more oncologically informed treatment selection.