Sucient Tissue Acquisition Rate of Peroral Cholangioscopy-guided Forceps Biopsy

Background: Peroral cholangioscopy (POCS)-guided forceps biopsy is a method for diagnosing indeterminate biliary strictures and for the preoperative identication of the exact perihilar and distal margins of biliary tract cancer (BTC). However, POCS-guided forceps biopsy may result in an insucient amount of specimen at times. Aims: We evaluated the sucient tissue acquisition rate and the factors affecting the sucient tissue acquisition of POCS-guided forceps biopsy for the biliary tract. Methods: Patients who underwent POCS-guided forceps biopsy for biliary disease between September 2016 and October 2018 at our hospital were enrolled retrospectively. We evaluated the sucient tissue acquisition rate of POCS-guided forceps biopsy for the biliary lesion and that for non-stenotic bile duct. In addition, the factors affecting the sucient tissue acquisition rate of POCS-guided forceps biopsy were evaluated. Results and Conclusions : We enrolled 47 patients with the biliary disease and performed POCS-guided forceps biopsy for biliary lesion and POCS-guided forceps mapping biopsy for non-stenotic bile duct in 40 and 36 patients, respectively. The sucient tissue acquisition rates of POCS-guided forceps biopsy for biliary lesions and that for non-stenotic bile duct were 86.4%, and 68.9%, respectively. In the multivariate logistic regression analyses, age and previous biliary stenting before POCS were factors affecting the sucient tissue acquisition rate of POCS-guided forceps biopsy for the biliary lesion. For non-stenotic bile duct, the location of the biliary lesion, endoscopic sphincterotomy, and procedure time of POCS were factors affecting the sucient tissue acquisition rate of POCS-guided forceps mapping biopsy.


Introduction
Biliary tract cancer (BTC), including intrahepatic cholangiocarcinoma (ICC), extrahepatic cholangiocarcinoma (ECC), gallbladder carcinoma (GC), and ampullary carcinoma (AC), is a disease with a poor prognosis. The 5-year survival rate and median survival time of BTC was 18.7% and 8.5 months, respectively [1]. Although the early diagnosis of BTC might improve the prognosis of that, BTC is often diagnosed at an advanced stage and is often unresectable because it is di cult to differentiate between BTC and benign biliary diseases, such as primary sclerosing cholangitis, immunoglobulin G subclass 4 (IgG4)-associated sclerosing cholangitis, xanthogranulomatous cholecystitis, and Mirizzi syndrome [2,3].
It is important to distinguish BTC from benign biliary disease, as the treatment strategies and prognoses differ.
Endoscopic retrograde cholangiopancreatography (ERCP)-related tissue acquisition, such as bile aspiration cytology, biliary brush cytology, and forceps biopsy, is commonly used to diagnose indeterminate biliary lesions. Although the speci city of the pathological examination of ERCP-related tissue acquisition for indeterminate biliary strictures is almost 100%, the sensitivity of that is unsatisfactory, with a range of 6-72% [4,5].
BTC has a clinical feature of super cial intraductal spread, in which epithelium extends continuously from the main lesion [9]. The frequency of super cial intraductal spread in cholangiocarcinoma is reported to be 14.6% [10]. The presence of super cial intraductal spread is related to non-curative resection for BTC so that it is essential to identify the exact perihilar and distal margins of the preoperative BTC. The usefulness of POCS-guided forceps mapping biopsy for preoperative identi cation of the longitudinal extension of BTC has been reported [11][12][13]. On the other hand, the diagnostic accuracy of the POCS visual ndings for indeterminate biliary lesions was higher than that of the POCSguided forceps biopsy ndings [14].
Although POCS plays important roles for diagnosis of indeterminate biliary strictures and for preoperative identi cation of the longitudinal extension of BTC, the diagnostic performance of POCS-guided forceps biopsy is not enough. The working channel of the cholangioscope is narrow, so only a mini-forceps with a 1.0-mm diameter cup could be used for POCS-guided forceps biopsy. Therefore, only a small amount of specimen can be obtained by POCS-guided biopsy. This is a possible reason for the inability to determine a pathological diagnosis and to evaluate the longitudinal tumor extent by POCS-guided forceps biopsy. In this study, we examined the su cient tissue acquisition rate of POCS-guided forceps biopsy for biliary lesions and that of POCS-guided forceps mapping biopsy for non-stenotic bile duct. We also evaluated the diagnostic performance of POCS-guided forceps biopsy for BTC and adverse events resulting from its use.

Study Population
Patients who underwent ERCP-related tissue acquisition with POCS for biliary disease between September 2016 and October 2018 at our hospital were retrospectively enrolled. Inclusion criteria were as follows: (1) Patients who underwent POCS-guided forceps biopsy for indeterminate biliary lesion; (2) Patients who underwent POCS-guided forceps mapping biopsy for BTC to identify longitudinal extension; (3) Patients aged 20 years or older when the endoscopic procedures were performed. Exclusion criteria were as follows: (1) Patients who did not provide consent; and (2) Patients who have been receiving chemotherapy for malignant diseases within one month prior to the acquisition of the pathological specimens.
Forty-seven patients with biliary disease were enrolled in the study. Participants included 32 men and 15 women aged 26-88 years (median age, 73 years). Twenty-eight patients had BTC and 19 had benign biliary lesions (Table 1). We de ned Cohort 1 as patients who received POCS-guided forceps biopsy for biliary lesions. Cohort 2 was de ned as the group of patients who received POCS-guided forceps biopsy for non-stenotic bile duct. Stenotic bile duct and non-stenotic bile duct was de ned by uoroscopy nding of ERCP. The biliary lesion sites with caliber changes in the biliary tract diameter were de ned as stenotic bile ducts, and other sites were de ned as non-stenotic bile duct. This study was performed according to the guidelines described in the Helsinki Declaration for biomedical research involving human participants.
This study was performed according to the guidelines described in the World Medical Association Declaration of Helsinki statement of ethical principles for medical research involving human subjects. The study protocol was also approved by the institutional review board of Tottori University (No.18A178).
In this retrospective study, informed consent was obtained from all participants using an opt-out approach.

Diagnostic Criteria and De nition of Su cient tissue
The diagnosis of BTC was based on the pathological diagnosis of bile aspiration cytology, and transpapillary forceps biopsy, endoscopic ultrasonography-guided ne needle aspiration (EUS-FNA), or surgical specimen. Patients with benign disease had a nal diagnosis based on clinical and radiological follow-up data. Biopsy specimens were stained with hematoxylin and eosin and, if necessary, immunostaining, including p53 and Ki-67, was also performed. Su cient tissue was de ned as a specimen that allowed the evaluation of the histological nding of the biliary epithelium. A specimen that contained only interstitial tissue was deemed insu cient ( Figure 2). Malignancy or suspected malignancy was considered positive in histological ndings.
We evaluated the su cient tissue acquisition rate of POCS-guided forceps biopsy for biliary strictures in Cohort 1 and that of POCS-guided forceps mapping biopsy for non-stenotic bile duct in Cohort 2. The factors affecting su cient tissue acquisition of POCS-guided forceps biopsy in each cohort were also evaluated. We evaluated the pathological diagnostic performance of the POCS-guided forceps biopsy for BTC. We also evaluated adverse events of POCS-guided forceps biopsy for biliary disease.

Statistical Analysis
Statistical analysis was performed using StatFlex ver. 6.0 for Windows (Artec Corp., Osaka, Japan).
Categorical variables were compared using the chi-squared test, and continuous variables were compared using the Mann-Whitney U-test. All values are expressed as mean ± standard deviation or median with interquartile range. Subgroup analyses of age (<65 or ≥65 years), sex, malignant biliary disease, the location of the biliary lesion or stricture (distal or non-distal), length of biliary stricture (<15 or ≥15 mm), macroscopic types of BTC ( at type or non-at type composed of nodular and papillary type), presence of acute cholangitis, level of serum T-Bil (<1.5 or ≥1.5 mg/dl), level of CEA ( <5.0 or ≥5.0 ng/ml), level of CA19-9 (<35 or ≥35 U/ml), procedure time (≤75 minutes or >75 minutes), EST, and previous biliary stenting before POCS-guided forceps biopsy were assessed to determine the su cient tissue acquisition rate of POCS-guided forceps biopsy. In Cohort 2, biopsy site (bifurcation of the bile duct or not, intrahepatic bile duct or extrahepatic bile duct, right side or left side) was also included in the subgroup analysis. We carried out univariate analyses to assess the su cient tissue acquisition rate, and factors with P < 0.1 were included in the multivariate logistic regression analyses. The signi cance level was set at P < 0.05. Table 1 showed the characteristics of patients with biliary disease. In this study, the malignant group included 12 patients with perihilar cholangiocarcinoma, 14 patients with distal cholangiocarcinoma, one patient with intrahepatic cholangiocarcinoma and one patient with cystic ductal carcinoma. Macroscopic types of BTC included four papillary-type, 16 nodular-type, and eight at-type. The benign group included nine patients with benign biliary strictures, three patients with IgG4-associated sclerosing cholangitis, three patients with primary sclerosing cholangitis, two patients with drug-induced cholangitis, one with intraductal papillary neoplasm of the bile duct (IPNB), and one with a peribiliary cyst. This study group included 24 patients with perihilar bile duct stricture, 19 patients with distal bile duct stricture, two patients with cystic duct stricture, and one patient with intrahepatic bile duct stricture. One patient did not have any biliary stricture, and he received POCS-guided forceps biopsy for distal bile duct near the ampulla of Vater in order to evaluate the hypertrophic extrahepatic bile duct. The median length of stricture in this study group was 16.6 mm (range, 0-46.0 mm). The median levels of serum total bilirubin (T-Bil), carcinoembryonic antigen (CEA), and carbohydrate antigen 19-9 (CA19-9) were 1.3 mg/dl (range, 0.3-14.8 mg/dl), 2.4 ng/ml (range, 0.8-10.1 ng/ml), and 34.6 U/ml (range, 0.8-4430.0 U/ml), respectively ( Table 1). The median procedure time of POCS-guided forceps biopsy was 77 minutes (range, 26-170 minutes). The nal clinical diagnosis was derived from surgical pathology in 15 patients.

Patient's Characteristics and Baseline Evaluation
There were 18 patients with a nal diagnosis based on clinical and radiological follow-up data (median follow up period, 7 months; range, 1-22 months). We performed both the POCS-guided forceps biopsy for biliary lesions and the POCS-guided forceps mapping biopsy for 29 patients with biliary disease. Eleven patients received only the POCS-guided forceps biopsy for indeterminate biliary lesions, and Seven patients received only the POCS-guided forceps mapping biopsy for non-stenotic bile duct to determine the longitudinal extension of BTC. Therefore, Cohort 1 and Cohort 2 included 40 and 36 patients, respectively ( Figure 3).

Diagnostic
Performance and Tissue Acquisition Rate of POCS-guided Forceps Biopsy for Biliary Tract Cancer.
The total and the median number of biopsies for indeterminate biliary lesions were 140 and 3 (range, 2-8) in Cohort 1, respectively. The total number of POCS-guided forceps mapping biopsies was 196 in Cohort 2. By the location in the bile duct, 93 specimens were obtained from the perihilar bile ducts of 25 patients (41 specimens from the con uence of the hepatic ducts of 23 patients), 31 specimens from the distal bile ducts of 15 patients, 22 specimens from the junction of the cystic ducts of 11 patients, 27 specimens from the rst order branch of the intrahepatic ducts of 12 patients, and 23 specimens from the second order branch of the intrahepatic ducts of 12 patients. Eleven patients received biliary stenting before POCS-guided forceps biopsy. In Cohort 1, 21 patients were with BTC, and 19 patients were with benign. Cohort 1 included 22 patients with perihilar bile duct stricture, 15 patients with distal bile duct stricture, 2 patients with cystic duct stricture, and 1 patient with extrahepatic bile duct hypertrophy. Table  2 summarizes the diagnostic performance of POCS-guided forceps biopsy to differentiate BTC from benign biliary disease. The values for sensitivity, speci city, positive predictive value (PPV), negative predictive value (NPV), and the accuracy of POCS-guided forceps biopsy were 76.2%, 94.7%, 94.1%, 78.3%, and 85.0%, respectively ( Table 2). The su cient tissue acquisition rate of POCS-guided forceps biopsy for biliary lesions was 86.4% (121/140).

Factors
Affecting the Su cient Tissue Acquisition of POCS-guided Forceps Biopsy for Biliary Lesion. Table 3 summarizes the result of analyses for the su cient tissue acquisition in POCS-guided forceps biopsy for biliary lesions. In the univariate analyses, age, sex, the location of the biliary lesion, and previous biliary stenting before POCS were factors with P < 0.1 and were included in the multivariate logistic regression analyses. Age <65 years old (odds ratio 0.170, 95% con dence interval [CI] 0.044-0.649, P = 0.004) and previous biliary stenting before POCS (odds ratio 0.199, 95% CI 0.053-0.756, P = 0.017) were the signi cant factors affecting the su cient tissue acquisition rate in the multivariate analysis.

Tissue Acquisition Rate of POCS-guided Forceps
Biopsy for Non-Stenotic Bile Duct. Cohort 2 included 17 patients with perihilar bile duct stricture, 16 patients with distal bile duct stricture, two patients with cystic duct stricture, and one patient with intrahepatic bile duct lesion. The su cient tissue acquisition rate of POCS-guided forceps mapping biopsy for non-stenotic bile duct was 68.9% (135/196), which was signi cantly lower than that of POCS-guided forceps biopsy for biliary lesion (P < 0.001). The su cient tissue acquisition rates of POCS-guided forceps mapping biopsy were calculated by the location in the bile duct: the distal bile duct (58.1%), the junction of the cystic duct (86.4%), the perihilar bile duct (77.4%; including the con uence of the hepatic duct 82.9%), the rst order branch of the intrahepatic bile duct (37.0%), and the second order branch of the intrahepatic bile duct (69.6%) ( Table 4).

Factors
Affecting the Su cient Tissue Acquisition of POCS-guided Forceps Biopsy for Non-Stenotic Bile Duct. Table 5 summarizes the results of the analyses for the su cient tissue acquisition in POCS-guided forceps mapping biopsy for a non-stenotic bile duct. In the univariate analyses, age, malignant lesion, the location of the biliary lesion, the length of stricture, the biopsy site in the intrahepatic bile duct, procedure time, and EST were factors with P < 0.1, and were included in the multivariate logistic regression analyses. The biliary lesion in the distal bile duct (odds ratio 0.322, 95% CI 0.139-0.744, P = 0.008), procedure time of 75 min or less (odds ratio 3.012, 95% CI 1.092-8.312, P = 0.033), and EST (odds ratio 7.041, 95% CI 2.117-23.421, P = 0.002) were the signi cant factors affecting the su cient tissue acquisition rate in the multivariate analyses.

Adverse Events
Adverse events following POCS-guided forceps biopsy occurred in eight patients (17.0%), with four patients (8.5%) developing acute pancreatitis, including one case of severe pancreatitis; and three patients (6.4%) developing cholangitis. Severe hemorrhage related to the precut papillotomy occurred in one patient (2.1%). No perforations were observed, all cases were resolved with conservative treatment, and there was no procedure-related mortality (Table 6).

Discussion
In recent years, POCS has become a widely-used method for the diagnosis of indeterminate biliary stricture. POCS allows optical viewing of the biliary system, as well as targeted forceps biopsies under direct vision. Some studies reported that the use of POCS can improve the diagnostic accuracy of indeterminate biliary lesions [6,12]. On the other hands, the sensitivities of POCS-guided forceps biopsy for malignant biliary stricture (60.1%) and cholangiocarcinoma (66.2%) were still insu cient [8] and were similar for uoroscopic tissue acquisition [7,15]. Moreover, the sensitivity and accuracy of POCS-guided forceps biopsy for indeterminate biliary lesions were lower than those of visual ndings for POCS in past study [13]. Because of the narrow working channel of the cholangioscope, only a mini-forceps with a 1.0-mm diameter cup was used for the diagnosis in POCS-guided forceps biopsy so that the specimen obtained by POCS is relatively small. A small amount of tissue sampling by POCS-guided forceps biopsy may be related to the insu cient pathological diagnostic ability of that for indeterminate biliary lesions. In the tissue amount, less tissue was obtained from POCS-guided forceps biopsy than uoroscopy-guided biopsy [16]. Indeed, POCS-guided tissue acquisition may only result in an insu cient amount of tissue. In our study, the su cient tissue acquisition rate of POCS-guided forceps biopsy for biliary lesions was acceptable (86.4%), and there were only 3 patients with biliary lesions in which a su cient amount of tissue could not be obtained. Meanwhile, the sensitivity of POCS-guided forceps biopsy for BTC was not acceptable (76.2%), and it was unsatisfactory compared to the su cient tissue acquisition rate. A working channel with a larger diameter for larger capacity forceps might improve the diagnostic utility for the indeterminate biliary lesion [17]. In the multivariate logistic regression analyses, we found that the previous biliary stenting before POCS was a factor affecting low tissue acquisition rate. In any case with POCS-guided forceps biopsy for biliary lesion, it may be better to avoid previous biliary stenting as much as possible except for patients with acute obstructive suppurative cholangitis. If it was impossible, multiple biopsies or combination of other tissue acquisition methods, such as biliary brush cytology, and forceps biopsy in uoroscopy, might be needed.
Super cial intraductal spread, which is a feature of BTC, is present in 14.6% of patients with BTC in previous studies [3,10]. The presence of super cial intraductal spread is related to positive resection margins, local recurrence, and poor prognosis after surgery. Therefore, it is essential to identify the exact perihilar and distal margins of the preoperative BTC. Hijioka et al. reported the usefulness of uoroscopic mapping biopsy procedures to distinguish between benign and malignant foci [18]. However, that method did not allow targeted forceps biopsies under direct vision of the biliary system, and it was uncertain whether the biopsy site was true or not. On the other hand, POCS-guided forceps mapping biopsy allowed targeted forceps biopsy under direct and uoroscopic vision of the biliary system. Although recent studies reported the usefulness of POCS-guided forceps biopsy in the preoperative assessment of the longitudinal extension of BTC, the sensitivity of that was not necessarily su cient [12,13,19]. In our study, the su cient tissue acquisition rate of POCS-guided forceps biopsy for non-stenotic bile duct was also insu cient (68.9%). Reasons for this may be not only the small amount of specimen obtained by POCS-guided forceps biopsy but also the technical di culty of the limited exion of the cholangioscope tip, especially in lesions above the biliary stricture as it may be inadequate to target the desired biopsy site [20]. Furthermore, the desired biopsy site above the curvature of the biliary tract might be a negative factor in the su cient tissue acquisition. Therefore, we think it is important to evaluate the factors affecting the su cient tissue acquisition of POCS-guided forceps biopsy.
In previous studies, there were a few reports about the factors affecting the accuracy or sensitivity of ERCP-related tissue acquisition for BTC or malignant biliary stricture [21][22][23]. However, no study reported on the factors affecting the su cient tissue acquisition of ERCP-related tissue acquisition, including POCS-guided forceps biopsy, for BTC or malignant biliary stricture. Therefore, we selected some candidate factors related to the su cient tissue acquisition from the factors affecting the accuracy or sensitivity of ERCP-related tissue acquisition for BTC or malignant biliary stricture.
In multivariate logistic regression analyses, we found that the presence of the lesion in the distal bile duct and the POCS-guided forceps mapping biopsy without EST were factors affecting a low tissue acquisition rate. The reason for those might be the limited exion of the cholangioscope tip making it di cult to target the desired biopsy site. Multiple biopsies might be needed to improve the su cient tissue acquisition of POCS-guided forceps mapping biopsy for patients with distal bile duct lesions. It is uncertain why the procedure time of 75 minutes or less was a factor affecting a high tissue acquisition rate of POCS-guided forceps mapping biopsy. Speculation includes that the biliary epithelium may be detached from the biliary tract as the procedure time of POCS increases and that the specimen which can be evaluated for a histological nding of the biliary epithelium may be lost. In any case, it is di cult to shorten the procedure time of POCS-guided forceps biopsy and increase the sampling. A more exible cholangioscope with a wide working channel and a larger capacity forceps might help to improve the su cient tissue acquisition rate.
This study has some limitations. First, this was a retrospective analysis of a small number of cases in single-center. Second, patients who were diagnosed through clinical follow-up were also included in this study. Therefore, it was unclear whether the cases with negative pathological ndings were truly benign biliary disease or not. Third, the accuracy of POCS-guided forceps mapping biopsy to diagnose tumor longitudinal extension was not uncertain because few patients underwent surgery and high rate of positive resection margins in patients underwent surgery in the study. A prospective long-term study including a larger number of patients is required to strengthen the investigation.
In conclusions, previous biliary stenting was a factor affecting a low tissue acquisition rate of POCSguided forceps biopsy for the biliary lesion. In the POCS-guided forceps mapping biopsy for non-stenotic bile duct, the presence of the lesion in the distal bile duct, the POCS-guided forceps mapping biopsy without EST, and the long procedure time of POCS were factors affecting low tissue acquisition rates.

Declarations
Ethics approval and consent to participate: The study protocol was also approved by the institutional review board of Tottori University (No.18A178).
Competing interests: This research was retrospective study. Informed consent was obtained from all participants using an opt-out approach.
Funding: This research received no external funding.
Con icts of Interest: The authors declare no con ict of interest.
Authors' contributions: TO conceived and designed the experiments; TO, YT, SK, HK, HK, TY, WH, YS, and KM contributed to data acquisition; TO analyzed the data; TO wrote the original paper; HI reviewed and edited the paper; and HI supervised the experiment.