Identifying predictive markers for survival in malignant biliary obstruction following percutaneous transhepatic biliary drainage

AIM: Malignant biliary obstructive disease is commonly managed with percutaneous transhepatic biliary drainage (PTBD). Traditionally, outcomes are evaluated collectively despite substantial variability in the underlying aetiology and extent of disease. The purpose of this study was to investigate whether variability in survival could be explained by different underlying patient and disease factors. MATERIALS AND METHODS: A single centre, retrospective study was performed looking at first-time malignant PTBD and respective survival outcomes between 2017- 2021. Survival censoring was taken as June 2022. Predictors included aetiology, age, preprocedural serum haemoglobin, and bilirubin. Subgroups were dichotomised with respect to median values. KaplaneMeier survival analysis and the log rank test were used for univariate analysis and Cox proportional hazards regression for multivariate analysis. RESULTS: One hundred fifty-six patients were identified, including 62 pancreatic, 19 ampullary/duodenal, 55 hilar cholangiocarcinoma, and 20 non hepatobiliary cancers. Median overall survival for the entire cohort was 136 days. The underlying aetiology significantly impacted median survival with non-hepatobiliary obstruction faring the worst; 53 days, and was in stark contrast to cholangiocarcinoma (347 days, p < 0.001). On multivariate analysis, we found that in addition to aetiology, patients > 70 years, preprocedural haemoglobin (< 110) and bilirubin (> 232) were all independent prognosticators and had significantly worse survival (HR 1.2, 1.8, 1.6, and 1.4, respectively, all p < 0.05). CONCLUSION: Age, underlying malignant aetiology, preprocedural haemoglobin, and bilirubin were identified as independent predictors of post-PTBD survival. Careful patient selection may therefore improve patient outcomes following PTBD. (c) 2024 The Royal College of Radiologists. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.