Gallstones is a common disease with a morbidity of approximately 15% [8]. Even after cholecystectomy, CBD stones would relapse. With anatomical alteration, cholecystectomy might result in biliary strictures and abdominal adhesions, which increased the difficulty of further clinical approach. In this study, 5 cases had biliary strictures in ERCP group, while abdominal adhesion was found in all cases of LCBDE group. The aims of this study was to investigate the clinical efficacy and superiority of ERCP and LCBDE for CBD stones in post-cholecystectomy patients.

In previous studies, CBD clearance rate of ERCP was 95%-97% [9, 10]^{9, 10}in experienced hands [9, 10]. No difference existed for preoperative ERCP or postoperative ERCP procedure [11]. Similarly, our results demonstrated that success rate was a bit higher by ERCP than LCBDE (97.7% VS 87.0%, p=0.03). For ERCP, procedural failure was usually due to large and impacted stones, postsurgical gastrointestinal anatomic variations, duodenal diverticulum, or CBD strictures, which was also shown in this study. One failed cases was diagnosed with large stones in the common bile duct, which is commonly considered as difficult bile duct stones and limits safe extraction [12]. For large or difficult CBD stones, endoscopic biliary stenting is an effective strategy, which could facilitate bile drainage and prevent stone impaction or cholangitis before the surgical intervention or a second ERCP attempt [13]. Instead of stone extraction, patient’s symptoms relieved after implantation with two biliary stents. In another case unexpected retained stones were found by nasobiliary cholangiography 5 days later. Campagnacci R et al. [14] and Naumowicz E et al. [15] reported the retained stones percentage after ERCP was 9% and 13.5% respectively. In our cohort, retained stones appeared only in 1.1% (1/87) of patients after first ERCP, lower than these studies. Retained stone was removed easily by the second ERCP.

According to randomized controlled trials, successful laparoscopic CBD stone clearance was 75–100 % [4, 5, 16, 17]. The rate was also consistent in our study (80.0%). Common failure reasons were impacted stones or retained stones. Tinoco et al [18] showed 1.5% retained stones were found in 481 LCBDE, lower than our results (7.4%, 4/54). Most of the cases in our study had no intraoperative cholangiography in LCBDE, which might have influenced the detection for stones. It’s difficult for LCBDE to deal with the impacted stone, especially in distal bile duct. For ERCP, sphincterotomy could be applied to extract impacted stones from the bile duct. All 3 casee with impacted stones were successfully treated by ERCP. Choledochoscope combined with micro-blasting lithotripsy via T-tube was another potential approach for impacted stones and retained stones after failed LCBDE.

In our study, therapeutic efficacy was better by ERCP. Previous random control trials reported that the clearance rate between ERCP+LC with LCBDE +LC was similar [4, 5, 16, 17]. LCBDE+LC is one stage procedure, while the ERCP+LC is commonly applied in the two stages. ERCP is performed within 4 weeks after LC. Our study focused on the patients after cholecystectomy, and most cases were treated by ERCP or LCBDE at one year after cholecystectomy. Post-surgical anatomic alteration might hinder clinical performance of LCBDE. As aforementioned, biliary stricture was found for both groups, and abdominal adhesion was common in all LCBDE cases. All these increased the difficulty of extracting stones. Procedural duration was longer for LCBDE due to adhesion dissection (102.9±40.1minutes), while ERCP took 50.2±15.8 minutes as usual. All cases with biliary benign stricture were cured by balloon dilatation under ERCP [19].

In our study, there was no significantly difference for complications, while ERCP group looked lower than LCBDE group (3.4% vs 11.1 %). The ERCP complications include pancreatitis, hemorrhage, cholangitis, duodenal perforation and mortality [20–22]. Due to the cohort limit, ERCP complication incidence was lower than previous retrospective studies 7.92% ∼ 11% [17, 20]. Experienced endoscopists were also important for preventing undesired complications. Only 3 patients had post-ERCP pancreatitis and were managed conservatively.

For LCBDE, the complication rate was similar as previous reports (9.5%) [21]. postoperative infection happened in three cases. Two patients appeared with biliary benign strictures two months after LCBDE, then were treated by ERCP as recommended [19]. One postoperative intestinal obstruction was reported due to postoperative gastrointestinal tract dysfunction, which was common in surgical patients. No bile leakage was observed in this study, which was common for LCBDE [21, 23]. Compared with ERCP, LCBDE patients need to carry with T-Tube for 75.6±32.3 days, which might greatly affect their life quality.

Postoperative hospitalization and cost was also similar for both groups. ERCP and LCBDE group respectively took 5.5±2.6 and 5.9±2.3 days. It was mainly influenced by the complications. The longest case was 12 days in ERCP group due to postoperative pancreatitis, while 13 days for postoperative cholangitis in LCBDE group. The total cost was approximate in our study. Cost was affected by many factors, such as hospital style, operative time, postoperative hospitalization and complications. This study just included patients from our single hospital. Treatment cost might vary in different regions, which might influence the choice of patients and health care providers.

Follow-up results indicated that recurrence rate was similar in the two groups. There were 6 (6.9%) patients in ERCP group and 4 (7.4%) patients in LCBDE group. Previous reported showed that the recurrent CBD stone occurred in 7.9% the patients after LCBDE [24]. The median time of recurrence was 11.3 months in ERCP group and 11.8months in LCBDE group after the operation. Most cases (70%, 7/10) were treated by ERCP, due to its minimally invasive and repeatable characteristics.

In conclusion, our results confirmed that for patients with CBD stones after cholecystectomy, ERCP was recommended due to its clinical efficiency. There was no difference for postoperative complication, hospitalization and recurrence. Without T-tube, ERCP patients also experienced the better life. As the retrospective study, lack of randomization might cause undesired selection bias, though exclusion criterion was set to reduce the bias. And sample size limits the final conclusion in this study. The randomized controlled trials with large cohort will be designed for further validating our analysis results.