Serum Bilirubin Levels Can Predict Pancreatic and Biliary Malignancies in Patients with Obstructive Jaundice and Non-conclusive Cytology

Abstract

Background: Elevated serum bilirubin has been shown to be a reliable predictor of pancreatic and biliary malignancy but the relationship between serum bilirubin and inadequate (C1), benign (C2) and indeterminate (C3) cellular samples has not been explored. The aim of this study is to determine the relationship between serum bilirubin and pancreatic, biliary or ampullary malignancy in the context of non-confirmatory cytology. Methods: This is a retrospective analysis of patients with obstructive jaundice undergoing investigation for possible pancreatic, peri-ampullary or biliary malignancy between 2009 and 2013. Results: 135 patients were included; 84 had a malignant diagnosis and 51 benign. All patients with C4 or C5 cytology (n=49) had confirmed malignancy. 35 out of 86 C1 – C3 samples were falsely negative. ROC curve analysis demonstrated a strong association (AUC 0.912) between elevated serum bilirubin and malignancy; serum bilirubin ≥ 100 µmol/L had a sensitivity of 86% and a specificity of 88%. In the C1-C3 subgroup, this association was maintained (AUC 0.905). Serum bilirubin ≥ 100 µmol/L had a sensitivity of 80% and specificity of 88%. Using this cut-off highlighted 28 out of 35 of the malignancies missed by cytology (p = 0.003). Conclusion: Our study demonstrates that a serum bilirubin ≥ 100 µmol/L is associated with malignancy and this relationship is maintained in C1-C3 cytology. When faced with non-confirmatory cytology in the absence of a benign aetiology and an elevated serum bilirubin ≥ 100 µmol/L, we advocate more aggressive investigation to avoid missing an occult malignancy.