It is widely accepted that patients with a history of gastrectomy have an increased incidence of CCL and increased morbidities requiring surgery [9]. The exact mechanisms for these observations remain unclear. According to the literature, the complex interaction between sectioning of the nerve supply to the gallbladder and the change in cholecystokinin secretion plays an important role [10, 11]. In gastric operations, reconstruction of the digestive tract may decrease passage of food through the duodenum, which probably decreases cholecystokinin secretion and reduces gallbladder motility, facilitating gallstone formation [12]. In addition, the hepatic branch of the vagus nerve is unavoidably damaged during surgical operations for gastric cancer because of the need for extended lymphadenectomy; moreover, the absence of or damage to the hepatic branch may cause dysregulation of gallbladder emptying, which may in turn contribute to gallstone formation [13]. Therefore, individualized and appropriate treatment strategies for CCL in patients with a history of gastrectomy are very valuable because such patients are often encountered in clinical practice.

Historically, the treatment of CCL required open laparotomy and CBDE [14]. After the introduction of ERCP and endoscopic sphincterotomy in the 1970s, ERCP+LC provided a less invasive option for treating CCL and has largely replaced CBDE in the management of CCL in the last two decades [15, 16]. With the development of laparoscopic equipment and technology, LCBDE has been widely used in clinical practice since its first introduction in 1991 [17–19]. In contrast to ERCP+LC, which is generally performed in two stages, LCBDE+LC for treatment of CCL is generally performed in a single stage. Additionally, this procedure appears to have a shorter hospital stay and similar stone clearance rate, relative cost-effectiveness while preserving the function of the sphincter of Oddi, and fewer ERCP-related complications [20, 21]. Although many studies have proven that LCBDE+LC is both feasible and effective in the management of CCL [22, 23], one retrospective cohort study performed in the United States showed that the overall use of ERCP+LC for treatment of CCL increased from 52.8% of admissions in 1998 to 85.7% in 2013 and that the percentage of patients with CCL undergoing CBDE (including open CBDE and laparoscopic CBDE) decreased from 39.8 to 8.5% in the same period. These results indicate that despite the potential benefits of LCBDE+LC over ERCP+LC for managing CCL, the current trends in CCL management continue, and CBDE may be at risk of disappearing from the surgical armamentarium [24]. Although the results of various studies strongly support this view, which treatment strategy is more beneficial to patients with CCL, especially those with a history of gastrectomy, still needs further investigation.

Most surgical specialists believe that either ERCP+LC or LCBDE+LC should be specifically chosen to treat patients with CCL in clinical practice based on the size and quantity of CBD stones, whether the CBD stones are combined with gallstones, the location and severity of the obstruction, and especially the level of the surgeon’s experience in ERCP or LCBDE at individual treatment centers [25, 26]. In the two hospitals of the present study, the ERCP technique was introduced in 2001 and has been applied to clinical practice for almost 20 years. Our endoscopists have accumulated abundant experience in performing ERCP to treat CCL, even in patients with a history of gastrectomy, and ERCP is typically the first-line treatment for CCL in both hospitals. We reviewed the clinical data of patients with CCL who underwent ERCP from 2001 to 2018 in our two medical centers and found that a history of gastrectomy was the most common cause of ERCP treatment failure (38.8%), followed by compact CBD stones (21.3%) and duodenal papilla hemorrhage (14.2%). However, the success rate of ERCP in patients with CCL with a history of gastrectomy still reached 67.9%. Previous randomized trials and meta-analyses have demonstrated the safety and efficacy of ERCP management for CCL with a success rate of 61.7 to 94.6% [27, 28]. Compared with our observation, the success rate of ERCP in patients with CCL and a history of gastrectomy is in accordance with the average level. These results indicate that a history of gastrectomy may be an important reason for failure of ERCP, but not a contraindication. Identification of the risk factors for ERCP failure in patients with CCL and a history of gastrectomy is important and was the major aim of our study. We selected gastroenteric anastomosis as the candidate risk factor and conducted a retrospective cohort study from May 2010 to March 2018, and we found that patients with a history of Billroth I gastrectomy have a higher success rate of ERCP for clearance of CBD stones and that ERCP might therefore be the first choice to treat choledocholithiasis in these patients.

Regardless of whether ERCP is performed successfully, all patients with CCL will inevitably undergo second-step LC or one-stage CBDE+cholecystectomy [28]. Previous upper abdominal surgery, especially gastrectomy, is a relative contraindication for laparoscopic surgery [29]. In one study, all surgical failures were attributable to adhesions, which included adhesions to the anterior abdominal wall at the site of insertion of the initial trocar and adhesions around the gallbladder and CBD [30]. In our two hospitals, LC and LCBDE were first introduced in 1996 and 2008, respectively, and gained widespread clinical acceptance even in patients with a history of upper abdominal surgery. We adopted a preoperative intra-abdominal adhesion evaluation procedure in 2011 to anticipate the severity of intra-abdominal adhesions, and this evaluation procedure significantly reduced the conversion rate of LC in these patients as shown in our previous research [7, 8]. In the present study, we chose the preoperative intra-abdominal adhesion evaluation score as a risk factor for conversion of LCBDE or performance of second-step LC in patients with CCL and a history of gastrectomy. We found that the success rates of laparoscopic surgery in patients with CCL and a history of gastrectomy are different when the preoperative intra-abdominal adhesion evaluation scores vary. When the evaluation score is > 3 points in an individual patient, extensive intra-abdominal adhesions are suspected or present, and safe peritoneal access is therefore needed. Open laparoscopy is the most recommended method in these patients with Hasson cannula [31, 32]. The peritoneal access technique is not difficult, but it is essential to increase the success rate of initial trocar insertion. Therefore, comprehensive and accurate preoperative evaluation of the severity of adhesions is important, and application of this procedure to clinical treatment would help to reduce the conversion rate of laparoscopic surgery in these patients.

At last, we collected the clinical information of the patients who were performed ERCP+LC or LCBDE+LC successfully, and we found that the postoperative complications have no differences between two groups, but the hospital costs and length of hospital stay were reduced in LCBDE+LC group. These results accord with the previous research conclusions.

Although majority of carried out studies confirmed that the incidence of CCL after gastric resection is increased compared with the people without gastrectomy history, but performing prophylactic cholecystectomy during gastric cancer surgery is still being debated [33]. In the fact, according to the data from the available published literature, the incidence of gallstone formation and symptomatic cholecystolithiasis requiring cholecystectomy after gastrectomy is low [34]. Based on these observations, we believed that routine prophylactic cholecystectomy may not be necessary for all patients undergoing gastrectomy, but identify the risk factors which contribute to gallstone formation and subsequent cholecystectomy is really mattered and which can help surgeons to make their rational surgical treatment strategies and avoid subsequent surgery or surgical overtreatment.

Limitations and possible biases in this study are the lack of randomization, which may have caused some selection bias, and the small number of patients, making the detection of small differences between the study groups unreliable.