iDREM-BTC: Dynamic Biomarker Monitoring Enables Real-Time Prognosis in Advanced Biliary Tract Cancer: Journal of Hepatology | 2026

Introduction:

First-line immunochemotherapy is now standard for advanced biliary tract cancer (BTC), but treatment responses vary widely. Most prognostic models rely on baseline information, ignoring how patients evolve during therapy. This multicenter study developed iDREM-BTC, a dynamic prediction model that continuously updates survival estimates using routinely available biomarkers during treatment.

Why was this study needed?

Advanced BTC has marked clinical and molecular heterogeneity.

Baseline prognostic models cannot capture changing treatment response.

Serial biomarkers may provide an early signal of progression or treatment benefit.

Dynamic risk assessment could support more personalized treatment decisions.

Results:

The model was developed and validated across 2,314 patients from eight centers, including prospective external validation.

Combining baseline characteristics with serial CA19-9, CRP, and bilirubin substantially improved prognostic accuracy as treatment progressed.

An enhanced iDREM(Pro)-BTC incorporating genomic and immunohistochemical biomarkers improved early risk discrimination.

The models identified evolving high-risk patients, potentially providing an opportunity for earlier reassessment of treatment strategy.

Clinical Impact:

iDREM-BTC introduces an important shift from static to dynamic precision oncology. Rather than predicting prognosis only at treatment initiation, routinely measured biomarkers can continuously update an individual patient's risk during immunochemotherapy. This could eventually help clinicians identify treatment failure earlier and guide timely imaging, treatment modification, clinical trial referral, or supportive care. Prospective interventional validation is still required before using the model to direct treatment.

Bottom Line:

In advanced biliary tract cancer, prognosis is dynamic rather than fixed. iDREM-BTC uses serial CA19-9, CRP, and bilirubin measurements to continuously refine survival prediction, offering a promising framework for real-time, adaptive cancer management.