Background: The sensitivity of transpapillary biliary forceps biopsy for malignancy has been reported as varying from 43–81%. Therefore, there are false negatives in more than 20% of patients, which makes it difficult to diagnose benign biliary stricture as benignancy in a clinical setting.
Methods: To clarify the number of tissue samples that should be obtained during transpapillary forceps biopsy to distinguish benign biliary stricture from malignancy, patients undergoing transpapillary biliary forceps biopsy at our institute were examined retrospectively in this study.
Results: Seventy-two biliary forceps biopsy procedures were performed on 61 patients. The final diagnoses were malignant biliary stricture in 34 patients and benign stricture in 27 patients. The overall sensitivity and specificity for malignancy in this study were 76.5% and 100%, respectively. There were zero out of 10 (0%) false-negative patients when three or more tissue samples were obtained. In contrast, when four or more tissue samples were obtained, eight out of eight (100%) patients had negative diagnoses for malignancy and were finally diagnosed with benignancy.
Conclusion: We suggest that three or more tissue samples are recommended for the diagnosis of biliary malignant stricture.
Due to the advances in imaging modalities such as multidetector computed tomography (MDCT)1 and magnetic resonance cholangiopancreatography (MRCP), many biliary strictures can be detected without the presence of jaundice. However, in some patients with biliary stricture, it is difficult to diagnose malignancy based on only imaging studies. Although the usefulness of transpapillary bile duct forceps biopsy for the diagnosis of patients with biliary strictures has been reported,2,3 the sensitivity for malignancy detection ranges 43–81%.4,5 This variability indicates that there are false negatives in more than 20% of biopsied patients, which makes it difficult to diagnose benign biliary stricture as benignancy in a clinical setting. It has been suggested that the sensitivity of forceps biopsy can be improved by increasing the number of tissue samples that are obtained. To clarify the number of tissue samples obtained by transpapillary forceps biopsy required to distinguish benign biliary stricture from malignancy, patients undergoing transpapillary biliary forceps biopsy (TBFB) at our institute were examined retrospectively in this study.
This retrospective study included 61 patients with biliary stricture or irregularities of the bile duct wall who underwent TBFB in Nagoya University Hospital. The TBFB were performed after evaluating the biliary stricture or irregularity of the bile duct wall by endoscopic retrograde cholangiography (ERC) and intraductal ultrasonography (IDUS).6,7 The forceps were inserted into the bile duct without endoscopic sphincterotomy (EST).2 One to three tissue samples were typically obtained for each patient in a single endoscopic procedure for differential diagnosis, and tissue samples were numbered according to their order of acquisition. Pathologists at our institute histopathologically diagnosed the samples as malignancy, suspected malignancy, atypical epithelium or benignancy. In this study, malignancy was defined as a positive finding of histopathological diagnosis, and the other diagnoses were defined as negative findings. When the histopathological diagnosis differed from the clinical findings, the TBFB was repeated. The final diagnosis was based on the surgical specimen or was made after more than 6 months of follow up. To clarify how many tissue samples were needed to distinguish benign biliary stricture from malignancy, we determined the number of tissue samples required to increase the sensitivity of biliary forceps biopsy for malignancy to 100%. The ERC procedures were performed with a backward oblique-viewing duodenoscope (JF-260V; Olympus Optical, Tokyo, Japan). The forceps used were Radial Jaw 3 (outer diameter 1.8 mm, with needle; Boston Scientific, Natick, MA, USA).
Seventy-two endoscopic biliary forceps biopsy procedures were performed on 61 patients to obtain a total of 174 tissue samples (1–9 samples/patients). The final diagnoses were malignant biliary stricture in 34 patients (cholangiocarcinoma, 33 patients; gallbladder carcinoma, one patient) and benign biliary strictures in 27 patients. The overall sensitivity and specificity of TBFB for malignant biliary stricture in this study were 76.5% and 100%, respectively. Mild post-ERC pancreatitis occurred in six of 72 (8.3%) patients.
Among the 34 patients with final diagnoses of malignancy, there were six of 15 (40%) false-negative patients when only one tissue sample was obtained, whereas there were zero of 10 (0%) false-negative patients when three or more tissue samples were obtained (Table 1). In contrast, when four or more tissue samples were obtained, eight of eight (100%) patients had initially negative diagnoses and were finally diagnosed with benignancy.
Table 1. Sensitivity of biliary forceps biopsy according to the number of tissue samples
No. of tissue samples
Final diagnosis malignancy patients
False- negative patients
Sensitivity of biopsy
A 29 year old man was referred to our hospital presenting distal bile duct stricture and jaundice (serum total bilirubin, 34.5 mg/dL). His jaundice persisted despite endoscopic biliary stenting (EBS). The ERC findings showed localized distal bile duct stricture (Fig. 1a); three tissue samples were obtained by TBFB (Fig. 1b), and a 7 Fr endoscopic nasobiliary drainage (ENBD) catheter was inserted.8 All three tissue samples were diagnosed as atypical epithelium, and the patient's jaundice gradually improved. The results of a reevaluation after 1 month were that the biliary stricture was improved (Fig. 2), and three more tissue samples obtained in a subsequent ERC were also diagnosed as atypical epithelium. The ENBD catheter was removed, and the patient was observed conservatively. The patient's serum total bilirubin value decreased to within normal limits after 2 months. After 18 months, the MRCP findings showed that the biliary stricture had improved (Fig. 3), and the biliary stricture was finally diagnosed as a benignancy.
A 66 year old man was referred to our hospital with distal bile duct wall irregularity. The patient had previously undergone EST and biliary stenting. The ERC findings showed that the irregularity of the bile duct wall had improved compared with ERC findings of a previous hospital 2 months earlier (Fig. 4), and IDUS findings showed a smooth-regular wall thickening spreading in the direction of the long axis (Fig. 5). Although the ERC and IDUS findings did not indicate malignancy, one of three tissue samples obtained by TBFB was diagnosed as adenocarcinoma. Therefore, an additional ERC for a peroral cholangioscopy (POCS) and a re-biopsy were performed. The POCS findings showed only a lower homogeneous papillogranular surface pattern and no evidence of malignancy (Fig. 6).9,10 Additionally, two tissue samples obtained at additional ERC were histopathologically diagnosed as atypical epithelium and not malignancy. Three months later, a subsequent ERC was performed. The irregularity of the distal common bile duct showed no changes, but the bifurcation of the right posterior bile duct had become stenotic (Fig. 7). Finally, this patient was diagnosed with a cholangiocarcinoma. We thought that the first detected distal bile duct irregularity was due to inflammation but coexisted with cholangiocarcinoma spreading in the submucosa; therefore, the distal bile duct irregularity improved and POCS could not detect malignant epithelial change.
It is sometimes difficult to diagnose biliary stricture as benignancy based solely on imaging studies, such as IDUS and POCS. Because the surgical procedures for cholangiocarcinoma are usually invasive (e.g. pancreatoduodenectomy or extended hepatectomy),11 not only is a certain differential diagnosis required, but a diagnosis of the lateral extension is also needed to determine the correct surgical procedure. Brush cytology and biliary forceps biopsy during an endoscopic procedure are the most common techniques for providing a definitive diagnosis of biliary stricture. Because the sensitivity of brush cytology is inferior to that of forceps biopsy,12 and lateral tumor extension must be assessed, biliary forceps biopsy is usually used to provide a definitive diagnosis of biliary stricture at our institute. To improve sensitivity for diagnosing cholangiocarcinoma, it has been recommended that two or more examinations are combined.13 But it is suggested that obtaining a greater number of tissue samples by TBFB can contribute to improving diagnostic sensitivity in this study. Other methods, such as immunohistological examination, can also improve the sensitivity of forceps biopsy.14 In our institute, there were no false-positive diagnoses of biliary stricture using TBFB. Although in case 2, as described above, we initially suspected that the first patient presented a false positive, the final diagnosis was malignancy. Endoscopic ultrasound-guided fine- needle aspiration may be useful for case of bile duct wall thickening without obvious epithelial abnormality, such as case 2.15
In this study, when four or more tissue samples were obtained, all patients had initially benign diagnoses and were finally diagnosed with benignancy. Additionally, there were no false-negative patients when three or more tissue samples were obtained. In conclusion, we suggest that three or more tissue samples are recommended for the diagnosis of biliary malignant stricture.