BACKGROUND AND STUDY AIMS: Endoscopic biopsies have a low sensitivity for diagnosing malignant bile duct strictures. Tumor markers detected by mucin staining and immunohistochemistry may help to determine the malignancy of a biopsy specimen where histologic evaluation alone is nondiagnostic. PATIENTS AND METHODS: 61 patients who underwent forceps biopsies were retrospectively identified, yielding 49 and 40 biopsy specimens for strictures finally diagnosed as benign and malignant, respectively. Biopsy specimens were histologically evaluated and stained for p53, Ki-67, carcinoembryonic antigen (CEA), CA19-9, CAM5.2, and presence of intracytoplasmic lumina (ICL). Sensitivity, specificity, positive and negative predictive values (PPV and NPV), and positive and negative likelihood ratios (PLR and NLR) were calculated to evaluate the performance of each test. RESULTS: Histology alone provided sensitivity and specificity of 53 % and 100 %. Addition of ICL or CAM5.2 increased sensitivity to 73 % or 60 %, respectively, and provided excellent specificity, PPV, and PLR (ICL, 98 %, 97 %, and 36; CAM5.2, 100 %, 100 %, and infinite). Both stains in combination increased the sensitivity to 75 %. Staining for Ki-67, p53, CEA, and CA19-9 increased the sensitivity to detect malignancy (range 60 % to 83 %), but significantly reduced the specificity, PPV and PLR (ranges 73 % to 90 %, 72 % to 86 %, and 3 to 7, respectively). Markers in all combinations performed poorly as a negative test (NPV 69 % to 87 %, and NLR 0.19 to 0.55). CONCLUSIONS: Staining for tumor markers ICL and CAM5.2 can improve the diagnostic value of endoscopic biopsies, and may change the course of management for patients with indeterminate histological findings. Georg Thieme Verlag KG Stuttgart. New York.
BACKGROUND AND STUDY AIMS: Endoscopic biopsies have a low sensitivity for diagnosing malignant bile duct strictures. Tumor markers detected by mucin staining and immunohistochemistry may help to determine the malignancy of a biopsy specimen where histologic evaluation alone is nondiagnostic. PATIENTS AND METHODS: 61 patients who underwent forceps biopsies were retrospectively identified, yielding 49 and 40 biopsy specimens for strictures finally diagnosed as benign and malignant, respectively. Biopsy specimens were histologically evaluated and stained for p53, Ki-67, carcinoembryonic antigen (CEA), CA19-9, CAM5.2, and presence of intracytoplasmic lumina (ICL). Sensitivity, specificity, positive and negative predictive values (PPV and NPV), and positive and negative likelihood ratios (PLR and NLR) were calculated to evaluate the performance of each test. RESULTS: Histology alone provided sensitivity and specificity of 53 % and 100 %. Addition of ICL or CAM5.2 increased sensitivity to 73 % or 60 %, respectively, and provided excellent specificity, PPV, and PLR (ICL, 98 %, 97 %, and 36; CAM5.2, 100 %, 100 %, and infinite). Both stains in combination increased the sensitivity to 75 %. Staining for Ki-67, p53, CEA, and CA19-9 increased the sensitivity to detect malignancy (range 60 % to 83 %), but significantly reduced the specificity, PPV and PLR (ranges 73 % to 90 %, 72 % to 86 %, and 3 to 7, respectively). Markers in all combinations performed poorly as a negative test (NPV 69 % to 87 %, and NLR 0.19 to 0.55). CONCLUSIONS: Staining for tumor markers ICL and CAM5.2 can improve the diagnostic value of endoscopic biopsies, and may change the course of management for patients with indeterminate histological findings. Georg Thieme Verlag KG Stuttgart. New York.
Authors: Tae Ho Kim; Jae Hyuck Chang; Hee Jin Lee; Jean A Kim; Yeon Soo Lim; Chang Whan Kim; Sok Won Han Journal: Medicine (Baltimore) Date: 2016-07 Impact factor: 1.889