Hepatolithiasis, stones in the intrahepatic bile ducts, can cause cholangitis, liver abscess, or obstructive jaundice, and guidelines recommend treatment including endoscopic retrograde cholangiopancreatography (ERCP), percutaneous transhepatic biliary drainage (PTBD), or surgery [
1,
2] PTBD was widely performed as a first-line non-surgical treatment for hepatolithiasis [
3,
4] but there was a concern in impairing activity of daily life and risk of early adverse events including bleeding [
5]. A recent study suggested comparative effectiveness of ERCP and PTBD [
6], and ERCP is now a treatment option for hepatolithiasis in clinical practice.
ERCP in patients with surgically altered anatomy (SAA), however, is technically challenging. Balloon endoscopy-assisted ERCP (BE-ERCP) is an emerging procedure for this population and previous studies reported its effectiveness and safety [
7‐
9]. However, the clinical evidence is still limited to date regarding feasibility of BE-ERCP for hepatolithiasis in patients with SAA [
5,
10]. Considering a low clearance rate of hepatolithiasis via ERCP even in patients with normal anatomy [
6], effectiveness and safety of BE-ERCP needs to be validated in this cohort. Furthermore, clinical outcomes of BE-ERCP could be different between patients who underwent surgery in their childhood and adulthood due to different etiology (congenital biliary dilation or malignancy), as well as gastrointestinal reconstruction (Roux-en-Y or Billroth-II).
Therefore, we conducted the current study to elucidate clinical outcomes and risk factors for failed complete stone removal via BE-ERCP for hepatolithiasis after hepaticojejunostomy (HJS).
Methods
Study design
The current study was a single-center retrospective study to evaluate the effectiveness and safety of BE-ERCP for hepatolithiasis after HJS. This study was conducted according to the guidelines in the Declaration of Helsinki and was approved by the ethics committee of the University of Tokyo. Written informed consent for procedure was obtained from each patient before the procedure. Consent for use of data was obtained on a basis of the opt-out consent.
Patients
Consecutive patients who underwent BE-ERCP for hepatolithiasis after HJS during the study period were identified through our ERCP database. Hepatolithiasis were defined as stones in the intrahepatic bile duct. Patients with choledochojejunostomy were not included in the analysis. Exclusion criteria were as follows: (1) patients who underwent transpapillary procedure, (2) patients who underwent BE-ERCP for benign biliary stricture alone (without hepatolithiasis), and (3) patients with malignant biliary obstruction. For patients who underwent repeated BE-ERCP for hepatolithiasis during the study period, only the first session was included for the analyses. The primary outcome was a rate of complete stone removal via BE-ERCP. The secondary outcomes included a rate of final complete stone removal, early adverse events associated with BE-ERCP, and a cumulative incidence of recurrent hepatolithiasis after complete stone removal. For the evaluation of the recurrent hepatolithiasis, patients with failed stone removal via BE-ERCP, patients who underwent subsequent biliary stent placement for concomitant biliary stricture, patients without follow up > 30 days were excluded from analyses.
Endoscopic procedures
A short type double-balloon endoscope (EC-450BI5/EI-530B with a 2.8-mm-wide working channel or EI-580BT with a 3.2-mm-wide working channel; Fujifilm, Tokyo, Japan) was utilized for BE-ERCP. EI-580BT was consecutively used from July 2015 [
11]. Most standard ERCP devices are applicable to the scope other than mother-baby cholangioscopy due to the small working channel. All patients were admitted for BE-ERCP procedures. The details of BE-ERCP procedures in our institution were reported elsewhere [
12,
13].
The algorithm of removal of hepatolithiasis is shown in Supplementary Table 1. Balloon dilation was performed in cases with concomitant biliary stricture. A diameter of balloon was selected not to exceed a diameter of the bile duct just above the stricture. Basket or balloon catheter was used for removal of hepatolithiasis. Endoscopic mechanical lithotripsy (EML) was performed if stones were larger than the diameter of HJS. For patients with failed stone fragmentation via EML, electrohydraulic lithotripsy (EHL) under direct cholangioscopy [
14] or extracorporeal shock wave lithotripsy (ESWL) after placement of endoscopic nasobiliary drainage were performed. For patients with failed complete stone removal via BE-ERCP, PTBD, endoscopic ultrasonography-guided hepaticogastrostomy (EUS-HGS), surgery, or conservative treatment was selected according to each patient’s condition [
15]. Generally, patients with hepatolithiasis in the right lobe underwent PTBD and those in the left lobe underwent EUS-HGS. Patients with persistent biliary stricture after balloon dilation underwent plastic or metal stent placement across the stricture after stone removal [
16].
Definitions of outcome variables and follow-up strategy
Early adverse events and their severities were defined according to the lexicon by American Society of Gastrointestinal Endoscopy [
17]. The severity of early adverse event was graded as follows: mild, requiring an unplanned prolongation of hospital stay for ≤ 3 days; moderate, requiring an unplanned prolongation of hospital stay for 4 to 10 days, endoscopy, interventional radiology, or admission to the intensive care unit for 1 night; and severe, requiring an unplanned prolongation of hospital stay for > 10 days, admission to the intensive care unit for > 1 night, or surgical intervention. Complete stone removal was defined as absence of bile duct stones via balloon occluded cholangiography during BE-ERCP or computed tomography (CT).
After successful stone removal, patients were followed regularly in the outpatient clinic at least every three months. In each visit, laboratory test including liver enzymes and inflammatory markers were evaluated and imaging studies with abdominal ultrasound, CT, or magnetic resonance cholangiopancreatography were performed every six months.
Statistical analysis
Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. Continuous variables were compared using the Wilcoxon rank-sum test.
In subgroup analyses, patients were divided into two groups by a period from surgery of HJS to BE-ERCP. Briefly, patients who had undergone the index surgery being over 10 years prior to the index BE-ERCP were categorized in the past HJS group, compared with who underwent surgery within 10 years were in the recent HJS group. Baseline characteristics and clinical outcomes were compared between the past and the recent HJS groups.
Uni- and multi-variable logistic regression models were used to estimate factors associated with failed stone removal via BE-ERCP. The multivariable model included variables with their P value < 0.10 in the univariable model. Cumulative incidences of recurrent hepatolithiasis were estimated using the Kaplan–Meier method and compared using the log-rank test. The lost follow-up or death were dealt as censored.
For all analyses, a two-sided
P value < 0.05 was used to denote statistical significance. All statistical analyses were performed using the EZR software (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for the R software (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.1) [
18].
Discussion
The current study demonstrated the effectiveness and safety of BE-ERCP for hepatolithiasis in patients with HJS. Complete stone removal via BE-ERCP was achieved in 73%. The rate of early adverse events was 9.9%, but there were no severe adverse events. Multivariable logistic regression model showed the past HJS group was the only factor associated with failed stone removal via BE-ERCP.
Before introduction of BE-ERCP, PTBD has been the first-line non-surgical treatment for hepatolithiasis in patients with SAA. By using percutaneous transhepatic cholangioscopy-guided lithotomy, complete stone clearance rate was as high as 85.3%, but the major adverse event rate was 1.6% and the mortality rate was 0.8% [
4]. In general, PTBD and its related procedures are associated with higher risk of early adverse events compared with endoscopic procedure [
5,
19], and it developed 26% in our cohort. Meanwhile, the low rate of early adverse events in our study suggested safety of endoscopic removal of hepatolithiasis even in patients with SAA. Ishihara et al. also reported a similarly high complete stone removal rate with a low-adverse event rate in the management of hepatolithiasis via BE-ERCP [
20], but evidences are still lacking to date on comparison between PTBD and BE-ERCP in this setting. Given its safety profile, BE-ERCP could be utilized as a first-line treatment for hepatolithiasis in patients with HJS, but further comparative studies are mandatory.
Our study revealed different clinical outcomes according to the timing of surgery. Contrary to the satisfactory high complete stone removal rate of 94% in the recent HJS group, it was as low as 50% in the past HJS group. There were two major reasons for technical failure in the past HJS group. One was failed scope insertion and the other was failed guidewire or device insertion to an intended bile duct. At first, 20% of patients failed scope insertion in the past HJS group. Compared with adult surgery, pediatric surgery was reported to have a higher risk of post-operative adhesive small bowel obstruction [
21,
22]. Furthermore, one retrospective study reported technical difficulty of BE-ERCP in patients undergoing pediatric surgery [
23]. In lines with previous studies, adhesion was the most frequent cause of failed scope insertion in the past HJS group in our study. Although CT findings might predict failed non-surgical management of adhesive small bowel obstruction [
24], further investigation is needed whether failed scope insertion can be predictable by the imaging studies prior to BE-ERCP. Secondly, about 50% of patients had undergone HJS for congenital biliary dilation in the past HJS group. Hepatolithiasis after HJS in patients with congenital biliary dilation was reportedly observed in about 10% [
25,
26]. Furthermore, type IV-A in Todani’s classification with dilated intrahepatic bile ducts is frequently associated with hepatolithiasis after surgery [
27,
28]. Due to the narrowed orifice of dilated peripheral bile ducts, insertion of guidewire or devices for stone removal is often technically difficult in these patients.
The rate of complete stone removal raised up to 90% by rescue PTBD or EUS-HGS after failed BE-ERCP in this study. EUS-HGS is a potential salvage treatment for hepatolithiasis in patients with SAA [
15,
29], and there are some case reports for successful removal of hepatolithiasis using HGS route [
30]. Furthermore, direct cholangioscopy through the HGS route and subsequent electrohydraulic lithotripsy under a direct visualization could be an option for large stones [
31]. However, data on the safety and effectiveness of EUS-HGS for hepatolithiasis are still limited to date, and furthermore, removal of hepatolithiasis in the right lobe is challenging through HGS route [
32]. Considering the low complete stone removal rate via BE-ERCP in the past HJS group, EUS-HGS might be a good alternative first-line treatment in this cohort. However, further a large-scale study is needed to clarify the effectiveness and safety of EUS-HGS for hepatolithiasis in patients with SAA, in comparison with BE-ERCP or PTBD.
The cumulative incidence of recurrent hepatolithiasis was 17% in the first year, and it seemed to be higher compared with a previous study [
20]. Complete stone removal was confirmed by cholangiography in this study, but it was sometimes difficult to distinguish remnant stone and pneumobilia. Direct cholangioscopy through an overtube was reported to be effective for confirming remnant bile duct during BE-ERCP, but this technique could be hardly available in patients with Roux-en-Y reconstruction [
20]. In our study cohort, 71% of patients had Roux-en-Y reconstruction and we did not routinely perform direct cholangioscopy to confirm complete stone removal. Furthermore, 32% of our cohort was congenital biliary dilation, and these patients were likely to recur hepatolithiasis due to their morphology of the bile duct [
28]. The high rate of recurrent hepatolithiasis in our study justified careful follow-up with imaging studies even after complete stone removal.
Consecutively performed BE-ERCP data and a relatively large sample size were the major strength of our study. To the best of our knowledge, the current study included the largest number of patients who underwent BE-ERCP for hepatolithiasis. Our study had several limitations. At first, the current study was a single-center retrospective design and could contain a risk of selection bias. Second, patients who underwent PTBD or surgery as a first line treatment were not included in the analyses. Third, the rate of gastrointestinal perforation was 3.1% in the current study, which was comparable to that in a large-scale prospective study [
8] but higher than that in the standard ERCP. Careful periprocedure monitoring is necessary for this potentially severe adverse events in cases undergoing BE-ERCP. Fourth, technical difficulty of BE-ERCP could be different by the afferent limb length, which could be different by surgeons, institutions, countries, and time periods. The current study had a single center design and needed external validation before generalizing our results. Finally, only 59 patients were included in the long-term analyses, and this small sample size was the limitation of our study.
In conclusion, BE-ERCP was effective and safe for hepatolithiasis, but the complete stone removal rate was low in the past HJS group. Rescue EUS-HGS or PTBD was helpful to achieve completer stone removal after failed BE-ERCP. Recurrent hepatolithiasis was common and careful follow up study is needed even after complete stone removal.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.