Biliary complications may occur in up to 20% of patients after deceased donor liver transplantation (LT).[1-6] Endoscopic retrograde cholangiopancreatography (ERCP) is considered a first-line therapeutic approach for patients with biliary complications who have undergone a duct-to-duct anastomosis because it confirms the diagnosis and allows therapy with success rates that have ranged from 80% to 100% in most series.[1, 4, 6] That said, the procedure has its own limitations because it cannot directly visualize the bile duct or specifically evaluate the nature and characteristics of a stricture or filling defects. Direct visualization of the bile duct via cholangioscopy with the mother-baby scope was first described in 1976.[7-10] However, the technique has failed to become a routine examination because of the fragility of the scope, the long procedural time, the high cost, and the need for 2 operators. The recent introduction of a single-operator cholangioscopy (SOC) system using the SpyGlass direct visualization system (Boston Scientific Corp., Natick, MA) has overcome many of these problems, mainly because it can be used by a single operator. To date, SOC is mainly used for the evaluation of indeterminate biliary strictures and the treatment of large common bile duct (CBD) stones. Recent studies have indicated that SOC has a high procedure success rate and high accuracy in teasing apart benign and malignant lesions.[12-17] There are very limited data on the role of cholangioscopy in the evaluation of biliary anastomotic strictures (ASs) or other biliary complications in LT recipients requiring ERCP. Most of the data available come from case reports[18-20] and 2 small retrospective case series.[21, 22] Therefore, the specific mucosal abnormalities of ASs have not been defined. In addition, it is unknown whether specific findings at the time of cholangioscopy and/or histological findings of strictures may help in determining the responses to and outcomes of endoscopic therapy. Thus the aim of this prospective study was to describe both the cholangioscopic and histological findings of biliary lesions in liver transplant recipients using the SOC-SpyGlass direct visualization system.
In this descriptive study, we examined the role of single-operator cholangioscopy (SOC) in the evaluation of biliary complications after liver transplantation (LT). We prospectively included adult recipients of deceased donor LT who were referred for endoscopic retrograde cholangiopancreatography between June 2009 and July 2011. All patients underwent SOC with biopsy of the biliary anastomosis. Sixteen patients were included: 12 with biliary anastomotic strictures (ASs), 2 with common bile duct stones, 1 with a bile leak, and 1 with sphincter of Oddi dysfunction. Patients with ASs displayed 1 of 2 patterns: (A) mild erythema (n = 9) or (B) edema, ulceration, and sloughing (n = 3). Those without ASs displayed a pale mucosa with mild edema at the anastomosis. Patients with ASs and pattern B required a longer period of stenting than patients with pattern A (457 versus 167 days, P = 0.01). In addition, patients with pattern A had a better response and better resolution of their strictures with endoscopic therapy than those with pattern B (88.9% versus 33.4%, P = 0.13). Histological examinations of ASs showed nonspecific intraepithelial inflammation in patients with patterns A and B. Biopsy samples from patients without ASs showed normal columnar epithelial bile duct cells. The total cholangioscopy time for all procedures was 26.8 ± 10.1 minutes. In conclusion, SOC in LT recipients is feasible and allows adequate visualization and tissue sampling of ASs and bile ducts. Two distinct visual patterns that are easily identified with SOC may help to predict the outcomes of endoscopic therapy in patients with biliary complications after LT. Liver Transpl 19:199-206, 2013. © 2012 AASLD.
alcoholic liver disease
acute liver failure
common bile duct
endoscopic retrograde cholangiopancreatography
hepatitis B virus
hepatitis C virus
sphincter of Oddi dysfunction.
PATIENTS AND METHODS
This prospective, single-center, descriptive study was conducted at the Hospital Clinic in Barcelona, a tertiary care hospital at which approximately 80 LT procedures are performed yearly and more than 450 ERCP procedures are performed annually. The study protocol was approved by the institutional review board of the hospital. All patients gave informed consent for the procedure.
We included adult recipients (age > 18 years) of deceased donor LT with biliary complications who were referred for ERCP between June 2009 and July 2011. The exclusion criteria included a refusal to participate in the study, an inability to provide informed consent, pregnancy, living donor LT, previous Roux-en-Y hepaticojejunostomy, a confirmed malignancy of the biliary tree, advanced liver failure, coagulopathy, hemodynamic instability, and sepsis.
All procedures were performed with an Olympus TJF 160R duodenoscope. The procedures were performed by 1 of 3 experienced endoscopists (A.C., D.B., and J.M.B.). All patients included in the study underwent ERCP and cholangioscopy with the SpyGlass direct visualization system. The system included a main tower with a light source, a water pump, and a video monitor as well as 3 disposable devices: a SpyGlass fiber optical probe, a SpyScope access and delivery cholangioscopy catheter, and a SpyBite biopsy forceps (Boston Scientific). If it had not been done previously, biliary sphincterotomy was performed in all cases. Patients underwent deep sedation with propofol and remifentanil administered by an anesthesiologist. All patients received intravenous prophylactic antibiotics (piperacillin/tazobactam or ciprofloxacin) 1 to 2 hours before the procedure. Antibiotics were continued in patients with prior evidence of infection. All patients remained in the hospital for at least 24 hours after the procedure.
In patients with ASs, a guide wire (0.035; Jagwire, Boston Scientific) was placed in one of the intrahepatic bile ducts, and SOC with a cholangioscopy catheter and its probe was performed as previously described.[12-15] The following characteristics of ASs and mucosa were evaluated: borders, ulcers, and concentricity versus eccentricity. According to previous data from Siddique et al., the description was based on 2 patterns of findings in ASs: one with scarring and minimal inflammatory changes and the other with edema, ulceration, and severe inflammatory changes. Afterwards, 2 to 4 biopsy samples were obtained with biopsy forceps under direct visualization from the area of each AS. After the dilation of the AS with a 4- to 8-mm Hurricane balloon (Boston Scientific), the cholangioscope was advanced to the proximal bile duct. Afterwards, the AS was managed with the placement of 1 or more plastic 10-Fr stents according to a standard protocol.[1-4, 23-25] The amount of stenosis required for stent placement was defined when ASs were dominant with a short narrowing at the anastomotic site without free or effective passage of contrast material to the proximal bile duct. In patients without ASs who had other complications [CBD stones, a bile leak, and sphincter of Oddi dysfunction (SOD)], descriptive findings of the anastomosis and bile duct were recorded, and this was followed by 2 to 4 biopsies of the area of the anastomosis. Finally, endoscopic therapy was performed for the specific biliary complication (ie, stone removal). The cholangioscopy procedures were video-recorded in their entirety for all patients. Images were thoroughly evaluated by 3 examiners.
The samples obtained with the previously described procedure were routinely processed at the histopathology laboratory, where they were fixed in buffered formalin (10%) and paraffin-embedded. Cut sections (5 μm) were stained with hematoxylin-eosin. Microscopic histopathological features were recorded in all cases.
Outcomes and Definitions
The main outcome measure was the feasibility of the procedure in LT recipients with adequate visualization of ASs, biliary anastomoses, and bile duct mucosa as well as the ability to obtain biopsy samples for the prespecified areas described previously. The secondary outcomes included the impact on endoscopic therapy, the incidence of adverse events, the total cholangioscopy time, and other diagnostic findings.
The procedure was considered successful if the cholangioscope was advanced to the bile duct and adequate visualization and sampling of the anastomosis was obtained. The total cholangioscopy time was measured from the moment at which the cholangioscope was inserted into the duodenoscope channel to the time at which it was withdrawn from the duodenoscope. This included both the visualization and sampling periods. Tissue sampling was considered adequate if the sample was easily identified macroscopically and was adequate for histological evaluation. Data were prospectively collected for each procedure and entered into a preset electronic database designed for the study. Clinical and biochemical data for patients with ASs were obtained before the procedure and then every 3 months after ERCP. In patients with ASs, the duration of stenting plus balloon dilation was analyzed until the resolution of their strictures. Plastic stents were removed every 3 months. At the time of each removal, the stricture was evaluated after the stent or stents were extracted. In accordance with published data and guidelines, when an AS was still present, further therapy with balloon dilation and stent placement was performed with the goal of using the maximum number of stents for the resolution of the stricture.[26, 27] If there was no evidence of stenosis on cholangiography and there was free passage of an extraction balloon, the stricture was considered to be relieved, and no further stents were placed. ERCP therapy was defined as failing when there was an indication for percutaneous transhepatic cholangiography or surgery during the follow-up. Adverse events during or after the procedure were prospectively collected and were defined and graded according to the Cotton classification.
Continuous variables were summarized as means and standard deviations or as medians and ranges according to their homogeneity. Categorical variables were compared with the χ test or Fisher's exact test as appropriate. Continuous variables were compared with the Student 2-tailed t test or the Mann-Whitney U test. Associations were specified as odds ratios with confidence intervals established at 95%. A 2-sided probability value < 0.05 was considered to be significant. Statistical analysis was performed with the SPSS 18.0 statistical package (SPSS, Inc., Chicago, IL).
During the study period, 23 patients who met the inclusion criteria were evaluated for inclusion in this study. Seven patients were excluded for the following reasons: hemodynamic instability (n = 2), a concomitant bacterial infection (n = 2), a concomitant and severe cytomegalovirus infection (n = 1), recent T-tube extraction (n = 1), and previous postsphincterotomy bleeding (n = 1). The baseline characteristics of the 16 patients who were included are summarized in Table 1. Twelve patients had ASs, 2 had CBD stones, 1 had a bile leak, and 1 had SOD. Complete SOC was successful in 15 of the 16 patients (93.8%). In 1 patient, the procedure could not be completed because the cholangioscope could not be advanced across the AS area despite previous dilation of the stricture with a 6-mm balloon and several attempts to advance the cholangioscope catheter. Five patients had a T-tube placed previously; in none of the cases was SOC performed with a T-tube in place. The total cholangioscopy time was 26.8 ± 10.1 minutes.
|Variable||Patients With ASs (n = 12)||Patients Without ASs (n = 4)a||P Value|
|Age (years)b||55.2 ± 8.7||52.1 ± 14.3||0.61|
|Male sex [n (%)]||10 (83.3)||1 (25)||0.06|
|Etiology [n (%)]|
|HCV-related||11 (91.7)||2 (50)||0.26|
|Other||1 (8.3)||2 (50)||0.26|
|Time from LT to ERCP (months)c||13.2 (6.3–29.9)||79.4 (2.7–210.9)||0.27|
|T-tube use [n (%)]||3 (25)||2 (50)||0.64|
|T-tube in place (days)d||102||104||0.81|
|Hepatic artery complications [n (%)]||1 (8.3)||0 (0)||0.99|
The area of the anastomosis or biliary AS was successfully identified in all patients with ASs. The intrahepatic biliary tree was evaluated in 15 (93.8%) of the 16 patients (supporting video 1). The findings are described in Table 2. Two different patterns of cholangioscopic findings were observed in patients with ASs: (A) the presence of mild erythema and scarring of the AS (n = 9; Fig. 1A and supporting video 2) and (B) the presence of severe edema and erythema plus ulceration with sloughing at the AS (n = 3; Fig. 1B and supporting video 3). Under direct visualization, the anastomoses of patients without ASs showed a slightly pale mucosa with an area of mild edema without any stenosis of the lumen (Fig. 1C). There were no significant findings or abnormalities in the biliary epithelium of the native or graft bile duct outside the biliary anastomosis in any of the patients. Cholangioscopy showed additional CBD stones (Fig. 1D) that were not seen in the initial cholangiograms for 5 patients (31%; Table 2).
|Patient||Sex/Age (Years)||Etiology||Hepatic Artery Thrombosis||Time From LT to ERCP (Months)||Biliary Complication||AS Type||Borders of Anastomosis||AS Pattern||Additional Findings||Stenting Period (Days)||ERCP Procedures With Stenting (n)||Need for Surgery|
The mean number of biopsy samples per patient was 4.8 (range = 3–8). The samples consisted of 1 or 2 tissue fragments that were approximately 2 mm in size. Adequate tissue for histological examinations was obtained from 81% of the patients. Most of the specimens consisted of superficial strips of mucosal biliary cells detached from the subepithelial connective tissue. Adequate samples for subepithelial tissue were obtained in 4 cases. The samples were insufficient for fully and adequately evaluating the subepithelial architecture. Samples from ASs in patients with cholangioscopy pattern A showed variable findings. The biliary epithelium was disrupted and showed moderate distortion with nuclear pseudostratification, prominent nucleoli, focal mucinous metaplasia, and focal intraepithelial inflammatory cells (mostly neutrophils; Fig. 2A). Subepithelial mucinous biliary glands were observed in 1 case and were associated with a chronic inflammatory infiltrate (Fig. 2B). Fibrosis was present in 1 case. In patients with pattern B, the histological findings showed changes of acute cholangitis. Tissue from stenoses in patients with pattern B showed a background formed by marked fibrinous material with scattered neutrophilic aggregates (Fig. 3A). Mild nuclear pseudostratification and abundant intraepithelial PMNs were observed in epithelial strips (Fig. 3B). Histopathology examinations of samples from patients without ASs showed normal columnar epithelial biliary cells with basal nuclei. There were no findings compatible with graft rejection, fungal infection, cytomegalovirus infection, or posttransplant lymphoproliferative disorder.
All patients with ASs were followed until their ASs were resolved or surgery or percutaneous transhepatic cholangiography was needed. Patients with ASs and pattern B required a longer period of stenting than patients with pattern A (457 versus 167 days, P = 0.01; Table 3). The maximum number of 10-Fr stents placed in any given patient was 4 (mean = 2–4). In addition, patients with pattern A had a better response and better resolution of their strictures with endoscopic therapy than patients with pattern B (88.9% versus 33.4%, P = 0.13). One patient (6.2%) developed cholangitis after cholangioscopy, was successfully treated with intravenous antibiotics for 5 days, and then was discharged on oral antibiotics. There were no other complications such as pancreatitis or bleeding related to the procedure. The main characteristics at the baseline and after ERCP for the patients with ASs (divided by patterns) are described in Table 4. Three patients with ASs underwent hepaticojejunostomy because of an inability to traverse the stricture with a guide wire in one case, the development of a liver abscess (unrelated to the endoscopic procedure) in another case, and a lack of a response despite multiple sessions in the last case (Table 2). There was no restenosis in any of the cases during the follow-up period. There were no significant differences in age, sex, post-LT evolution, or time from LT to ERCP between AS patients with pattern A and AS patients with pattern B.
|Characteristic||Pattern A (n = 9)||Pattern B (n = 3)||P Value|
|Duration of stenting (days)a||167 ± 87||457 ± 257||0.01|
|ERCP procedures with stentinga||2.7 ± 1.2||5.3 ± 2.5||0.03|
|Success of endoscopic therapy [n (%)]||8 (88.9)||1 (33.4)||0.13|
|Variable||Pattern A (n = 9)||Pattern B (n = 3)||P Value|
|Age (years)a||55.6 ± 8.4||53.9 ± 11.4||0.81|
|Male sex [n (%)]||8 (88.9)||2 (66.7)||0.45|
|HCV etiology [n (%)]||8 (88.9)||3 (100)||>0.99|
|Bile leak [n (%)]||2 (22.2)||0 (0)||>0.99|
|T-tube use [n (%)]||3 (33.3)||0 (0)||0.51|
|Acute cellular rejection [n (%)]||1 (11.1)||0 (0)||>0.99|
|Cytomegalovirus infection [n (%)]||2 (22.2)||0 (0)||>0.99|
|Hepatic artery thrombosis [n (%)]||1 (11.1)||0 (0)||>0.99|
|Immunosuppression at ERCP [n (%)]|
|Tacrolimus||3 (33.3)||2 (66.7)||0.53|
|Cyclosporine||3 (33.3)||0 (0)||0.51|
|Mammalian target of rapamycin inhibitorb||2 (22.2)||1 (33.3)||>0.99|
|Prednisone||4 (44.4)||0 (0)||0.49|
|Mycophenolate mofetil||2 (22.2)||0 (0)||>0.99|
|Time from LT to ERCP (months)c||14.9 (9.5)||24.7 (15.1)||0.24|
|Time from LT to ERCP > 12 months [n (%)]||6 (66.7)||2 (66.7)||>0.99|
Biliary complications after LT continue to haunt LT teams, and they are still considered the Achilles' heel of the posttransplant period.[2-5] Nonetheless, significant advances in both surgical techniques and the endoscopic and radiological management of this common problem after LT have improved the outcomes of patients who develop not only ASs but other complications such as bile leaks, nonanastomotic strictures, stones, and bile duct cast syndrome.[2-5, 29] A limited number of case reports of LT recipients indicate that SOC with the Spyglass cholangioscopy system is useful for obtaining biopsy samples and for directly visualizing and aiding in the passage of a guide wire in difficult strictures that cannot be traversed with fluoroscopy.[18, 19, 22] However, there is a lack of adequately collected data describing the specific cholangioscopic findings in LT recipients with ASs and other biliary complications. In this prospective study, we examined the cholangioscopic and histological findings of patients with biliary complications after LT. The main findings of our study indicate that (1) ERCP-guided SOC with the SpyGlass system is feasible and can successfully be performed in LT recipients with biliary complications, (2) 2 different cholangioscopic AS patterns can be easily identified and may help to predict responses to therapy, and (3) histological findings in ASs show nonspecific inflammatory changes.
To date, the main indications for SOC are the evaluation of indeterminate strictures of the pancreaticobiliary tract and therapy for large and difficult-to-remove CBD stones. However, the role of SOC in the evaluation of biliary complications after LT has not been adequately studied. In 2 small retrospective studies, cholangioscopy (in one with the mother-baby scope and in the other with the SpyGlass system) was used to evaluate the bile ducts of LT recipients.[21, 22] In a study that included 14 patients with ASs after LT, cholangioscopy showed abnormal findings for ASs in 8 patients, and 2 visual patterns of ASs were described. One pattern was defined as ulceration and severe inflammatory changes, and the other was defined as scarring and mild inflammatory changes; however, biopsy samples were not taken. The visual patterns described in the study by Siddique et al. are similar to the findings described in the present study. However, a major difference is that we identified the 2 distinct patterns in all cases, and biopsy samples were obtained from the ASs. Our findings not only confirm the visual patterns described previously but also detail the visual appearance of the anastomoses in patients with ASs and patients without ASs and demonstrate that biopsy samples from ASs can be obtained. In addition, both patterns were different than the findings in patients without ASs, and this reinforces the notion that both patterns are associated with ASs.
The histological findings of cholangioscopy-guided biopsy showed nonspecific changes that did not differentiate the patterns. This was likely due to the limitations of the small amount of tissue obtained with the biopsy forceps. The previously described nonspecific findings likely correlated with the visual impressions of the ASs in the patients with patterns A and B. However, these histological results did not translate into significant changes that altered the management or outcomes of the patients. The findings described in this study reflected only superficial changes; deeper abnormalities of the bile duct wall could not be adequately evaluated with the amount of tissue obtained via luminal biopsy with this system.
An interesting and novel finding of this study is that responses to endoscopic treatment in LT recipients with ASs may differ according to the cholangioscopy pattern. Patients with pattern A responded better than patients with pattern B and required 290 fewer days of stenting (mean) to achieve a final response to endoscopic therapy. In addition, patients with ASs and pattern B needed more ERCP sessions with stenting than those with pattern A. This finding, we believe, may be useful in clinical practice because patients with pattern B require prolonged therapy and thus should be considered for early surgery if there is no good response after 1 or 2 sessions; this will prevent a prolonged course of ERCP and the potential complications associated with the procedure. We thus propose a treatment algorithm based on the findings of SOC in patients with ASs (Fig. 4). Although there was no significant difference in the clinical or histological characteristics of patients with ASs and pattern A versus patients with ASs and pattern B, the number of patients was low for each pattern (9 and 3, respectively), so factors that predispose patients to the development of one pattern versus the other need to be evaluated in a larger number of patients. This would require a large prospective, multicenter study that could evaluate predetermined criteria based on patient characteristics, surgical characteristics of transplants, radiographic and cholangiographic criteria, and visual characteristics under SOC as well as correlations of specific endpoints with the outcomes of endoscopic therapy, the need for surgical reinterventions, and clinical outcomes.
Some limitations of this study should be noted. First, the number of analyzed patients was low, so statistical comparisons certainly must be treated with caution. Second, this was a single-center study, and the procedures were performed by physicians with experience in the management of biliary complications after LT. Thus, these results may not be applicable to all centers. Third, there is a learning curve with this procedure, and although we feel that with time our technique, interpretation of images, and biopsy collection have improved, we cannot appropriately evaluate this point because of the small number of patients in the study. Finally, we did not include patients who underwent living donor LT or recipients of transplants from donors after cardiac death, who have a higher incidence of ASs versus recipients of cadaveric donors.
In conclusion, ERCP-guided SOC with the SpyGlass system is feasible and can be successfully performed in LT recipients with biliary complications. Cholangioscopic findings of ASs may predict the response to ERCP therapy; this means that patients with significant ulceration and sloughing of the mucosa are more difficult to treat and may require longer periods of therapy. Further prospective studies comparing ERCP alone to ERCP plus SOC are needed to confirm these results and firmly establish the role of cholangioscopy in the response to endoscopic therapy in patients with ASs after LT.