Benign or Malignant? Predicting the Nature of Indeterminate Biliary Strictures

Abstract

Purpose: Gastroenterologists are often faced with questions regarding the benign or malignant nature of biliary strictures. We sought to develop a model based on demographic characteristics and laboratory tests to determine the likelihood of malignancy in patients with indeterminate biliary strictures. Methods: Tumor markers (Ca 19-9, CEA, and α-fetoprotein), liver function tests and demographic characteristics (age and sex) of consecutive patients with indeterminate biliary strictures (defined as stricture involving the extrahepatic biliary tree without convincing radiological or clinical signs of malignancy) were evaluated in a prospective cohort study (ClinicalTrials.gov number NCT00894348). Univariable and multivariable analyses were performed using logistic regression to assess differences between subjects with benign and malignant biliary strictures. Receiver operating characteristics (ROC) analysis was used to evaluate the utility of tumor markers, liver function tests, and demographic characteristics alone or in combination in discriminating stricture type. The final diagnosis of malignancy was based on surgical pathology or brush biopsy revealing adenocarcinoma, or long-term follow-up showing obvious cancer progression. Results: Seventy-five patients were evaluated. Fifty-four (72%) had a final diagnosis of malignancy. All patients were followed up for at least 1 year or until death if it occurred sooner. Compared with the other predictor variables, Ca 19-9 provided the best differentiation between benign and malignant biliary strictures (AUC=0.847). A combination of age and bilirubin levels (Z = -5.8255 + 0.0859 *age (yrs) + 0.2044 *Bilirubin) produced the most accurate and parsimonious prediction of malignancy, with an AUC of 0.894. When converted to a 0-100 score model (100*[exp(Z)/(1+exp(Z)]), a value of ≥74.5 provided a sensitivity and specificity of 81% and 90%, respectively. Conclusion: Ca 19-9 is the best single marker for differentiating between malignant and benign biliary strictures. However, a model combining age and bilirubin produced similar discrimination as CA19-9. Prior to use in clinical settings, this model has to be verified in larger studies.