Clinical decision making regarding the indication and type of procedure for palliation of malignant biliary obstruction (MBO) in patients with advanced cancer is burdened by lack of reliable prediction tools assessing periprocedural risk and mortality. Consequently, frail patients may sometimes be exposed to unnecessary invasive procedures with no clinical benefit. We aimed to assess the predictive value of a simple prognostic model validated for advanced cancer patients (the PROMAC score) in identifying patients at risk for early post-procedure death in this particular clinical setting.

We conducted a prospective observational cohort study including adult patients diagnosed with MBO referred for palliative drainage in a tertiary endoscopy center. All patients underwent clinical and biochemical assessment prior to the index drainage procedure and were followed up with scheduled clinical or telephone visits at 14 days after the intervention. The PROMAC score was calculated based on clinical and paraclinical data from the patients’ chart, including patient demographics, number of organ systems with metastasis, serum albumin and total white cell counts and the Palliative Performance Scale, PPS V.2 and the Edmonton Symptom Assessment System, ESASr as previously described¹.  Based on this prognostic score, patients were stratified as low-, medium- and high-risk for early post-procedure death at 14 days. Biliary drainage was conducted according to local protocols, with a primary endoscopy-based approach (ERCP or EUS) and salvage percutaneous approaches available in cases where ERCP or EUS were not technically feasible. 

Seventy-five consecutive patients (44 male, median age 66 years) with MBO admitted for palliative drainage at Colentina Clinical Hospital, Bucharest between March and September 2025 were included in the study. Perihilar cholangiocarcinoma (31) and pancreatic adenocarcinoma (15) were the most common primary tumours. ERCP represented the primary drainage modality in 64 cases, with EUS-guided and percutaneous approaches used in 8 and 3 cases respectively.

Follow-up data was available for 73/75 patients, with a total 14-day mortality rate of 12% in our cohort. Based on the PROMAC risk model, patients in the high-risk group had a significantly higher mortality rate compared to medium- and low-risk groups (45.5% vs 7.9% vs. 4.5%, p=0.004). Bilirubin levels dropped by a mean of 2.2 mg/dl at 14 days after the index procedure (from 12.6 to 10.4mg/dl, p<0.001). Early reintervention due to technical failure or postprocedural complications was required in 12(16%) patients during the index admission.

Our findings suggest that a simple clinical prediction tool can be used to identify patients at high risk of early post-procedure death following minimally invasive treatment of MBO. While technically feasible in most cases, drainage procedures seem to have a limited clinical impact for this subgroup of patients with a high disease burden and limited survival. Future studies should focus on incorporating such clinical tools to help tailor individualized treatment plans and avoid potentially futile and risky procedures in patients with advanced cancer and MBO.