Conclusions
NGAL in bile is a novel potential biomarker to help distinguish benign from malignant biliary obstruction.
Methods
Bile, urine, and serum were collected prospectively from 38 patients undergoing endoscopic retrograde cholangiopancreatography ("discovery" cohort); 22 had benign and 16 had malignant pancreatobiliary disease. Initially, label-free proteomics and immunoblotting were performed in samples from a subset of these patients. Enzyme-linked immunosorbent assay was then performed for NGAL as a potential biomarker on all samples in this cohort. The diagnostic performance of biliary NGAL was then validated in a second, independent group ("validation" cohort) of 21 patients with pancreatobiliary disease (benign n=14, malignant n=7).
Results
NGAL levels were significantly raised in bile from the malignant disease group, compared with bile from the benign disease group in the discovery cohort (median 1,556 vs. 480 ng/ml, P=0.007). Biliary NGAL levels had a receiver operating characteristic area under curve of 0.76, sensitivity 94%, specificity 55%, positive predictive value 60%, and negative predictive value 92% for distinguishing malignant from benign causes. Biliary NGAL was independent of serum biochemistry and carbohydrate antigen 19-9 (CA 19-9) in differentiating between underlying benign and malignant disease. No significant differences in serum and urine NGAL levels were found between benign and malignant disease. Combining biliary NGAL and serum CA 19-9 improved diagnostic accuracy for malignancy (sensitivity 85%, specificity 82%, positive predictive value 79%, and negative predictive value 87%). The diagnostic accuracy of biliary NGAL was confirmed in the second independent validation cohort. Conclusions: NGAL in bile is a novel potential biomarker to help distinguish benign from malignant biliary obstruction.