Risk Factors of Pancreatitis after Percutaneous Transhepatic Biliary Stent Implantation in Patients with Distal Malignant Biliary Obstruction

Purpose Percutaneous transhepatic biliary stent implantation (PTBS) has been widely used in the treatment of malignant obstructive jaundice (MBO) (cid:0) which is an effective palliative therapy. The goal of the research was to explore the risk factors of postoperative pancreatitis in patients who had received PTBS. Materials and Methods This study retrospectively included 248 patients with distal malignant biliary obstruction (DMBO) who received PTBS from 2014 to 2019. The risk factors of postoperative pancreatitis such as clinical characteristics, laboratory and stent variables were evaluated by logistic regression analysis.

Previous studies showed that risk factors of pancreatitis after endoscopic retrograde biliary stent implantation include previous pancreatitis, sphincter of Oddi dysfunction (SOD), di cult cannulation, precut sphincterotomy, main pancreatic duct injection and so on. However, studies on the risk factors of pancreatitis after PTBS are scarce. The pathogenesis of pancreatitis following PTBS is still unclear [11,12]. A pathogenesis was reported by Andrea et al. [13] who revealed that compression of the pancreatic duct ori ce due to medial de ection of biliary stents leads to pancreatitis. According to clinical experience, patients with DMBO is more likely to develop postoperative pancreatitis than high biliary obstruction. The risk factors of pancreatitis after PTBS have not been su ciently reported. Therefore, the Page 3/16 present study was designed to investigate risk factors of acute pancreatitis after PTBS for DMBO. We hope these ndings will help with clinical treatments in future.

Patient
According to the enrollment criteria including patients who underwent PTBS successfully for distal malignant biliary stricture, 256 patients were involved in this research between January 2014 and December 2019. However, 3 of them were excluded due to preoperative pancreatitis, and 5 were excluded because of incomplete clinical data. Finally, 248 patients (137 males and 111 females; age range: 32-91 years) were included. In this study, DMBO was de ned as obstruction involving the common bile ducts. The diagnosis of MBO was based on clinical presentation, imaging ndings, serum tumor markers and bilirubin levels. All included patients were divided into two groups according to whether developing acute pancreatitis after PTBS. Risk factors of pancreatitis after PTBS were assessed by analyzing clinical characteristics, stent parameters and laboratory variables. All collected data are summarized in Table 1. . And all of them are covered stents. All surgical procedures were performed in accordance with the standard guideline. Weather to implant the stent in one-stage operation is determined by the patient's general status and liver function. External biliary drainage for 3-5 days is required in patients with two-stage operation. Whether the stent is implanted across the duodenal papilla and which type stent to choose depends on the location and length of the obstruction, as well as the physician's clinical experience. The process was shown in the Fig. 1 and Fig. 2.

Acute pancreatitis
The diagnosis of acute pancreatitis was based on the Atlanta classi cation of acute pancreatitis. The diagnostic criteria were abdominal pain, serum amylase and/or lipase levels ≥ 3 times the upper limit of normal value, and imaging manifestations consistent with acute pancreatitis. Patients who met two or more above criteria were diagnosed as pancreatitis. Using Common Terminology Criteria for Adverse Events (CTCAE), version 4 to assess the severity of acute pancreatitis [14,15].

Statistical analysis
Mann-Whitney U-test and the Chi-square test were used to compare and analyze continuous variables and categorical variables, respectively. Statistically signi cant variables in univariate logistic regression analysis (P < 0.1) were incorporated into multivariate analysis to identify independent risk factors. Variables considered to be in uential in previous studies can also be included in multivariate analysis. All statistical analysis was performed by SPSS software, version 25.0 (SPSS Inc, Chicago, IL). A P value 0.05 was considered statistically signi cant.

Patient characteristics
PTBS was successfully achieved in all patients. The condition of primary disease is as follows: 73 originate from bile duct, 131 originate from pancreas and 44 originate from other parts of ampulla. Among the 248 patients, 17 patients had in-stent restenosis, 46 patients were implanted biliary stent in one-stage surgery, 198 patients underwent biliary stent implantation across the duodenal papilla (Table 1).

Acute Pancreatitis After PTBS
The diagnosis of pancreatitis was detected within 24 hours following PTBS in all cases. In all patients with pancreatitis, the increase of amylase and lipase was detected in emergency laboratory after presenting symptoms such as abdominal pain, nausea and vomiting. The Incidence of pancreatitis after PTBS is 20%, which is higher than that in previous studies. The grade of pancreatitis was mild in all cases. The serum amylase and lipase levels restored to normal within 3-7 days following conservative therapy in all patients.

Risk factors analysis
On univariate analysis, stent type, stent implantation across the duodenal papilla (OR = 7.893, 95%CI:  (Table 2).  (Table 3). Furthermore, stent type was also an independent factor, and the incidence of pancreatitis is as follows: Cook 36%, Cordis 10%, Bard 26%, Antai 28%, respectively.  [1,9,16]. In this research, the morbidity of postoperative pancreatitis was 20%, which is higher than that in previous studies. That's probably because this research only selected patients with distal malignant biliary tract obstruction. Some scholars believe that postoperative pancreatitis is more likely to occur in patients with relatively low obstruction, which probably owing to a higher incidence of stimulation of the pancreas during the procedure.
To the best of our knowledge, studies on risk factors of acute pancreatitis after PTBS are scarce. In a previous meta-analysis, female gender, di cult cannulation, previous pancreatitis, previous post-ERCP pancreatitis (PEP), intraductal papillary mucinous neoplasm (IPMN), sphincter of Oddi dysfunction (SOD), endoscopic sphincterotomy (EST), precut sphincterotomy and main pancreatic duct injection are risk factors for post-ERCP pancreatitis [17,18]. Combined with previous reports and clinical experience, 20 variables were included in our study to analyze their in uence on pancreatitis after PTBS (Table 1). In previous studies, main pancreatic duct injection was a risk factor of postoperative pancreatitis, while the morbidity of postoperative pancreatitis in patients who had atrophy of pancreas and dilation of main pancreatic duct signi cantly decreased [19]. In present study, main pancreatic duct injection and main pancreatic duct expansion were not signi cant risk factors (P = 0.857, P = 0.296, respectively).
Univariate analysis showed that stent type, stent implantation across the duodenal papilla, preoperative white blood cell count > 10 × 10 9 /L, preoperative alkaline phosphatase level and preoperative (within 3 days) administration of PPIs were signi cant in uence factors for postoperative pancreatitis. Our results show that the OR of preoperative white blood cell count > 10 × 10 9 /L was 0.107, which suggested this variable has a protective effect. One explanation of this condition may be that clinicians usually treat patients who have leukocytosis with antibiotics. Meanwhile, some scholars believe that prophylactic use of antibiotics before surgery may reduce the risk of postoperative infection. Therefore, preoperative administration (within 3 days) of antibiotics was enrolled in the multivariate analysis instead of white blood cell count. Multivariate analysis showed that stent type, stent implantation across the duodenal papilla, preoperative administration of PPIs and preoperative administration of antibiotics have statistically signi cant difference, stent implantation across the duodenal papilla and preoperative administration of PPIs were risk factors for developing pancreatitis, while preoperative administration of antibiotics was a protective factor. The OR of alkaline phosphatase is close to 1, which reveals that it has little in uence on the occurrence of pancreatitis. In this study, the postoperative pancreatitis rate with different stent was statistically signi cant. However, bias such as the capacity for medical payment by patients in the choice of stents should not be ignored. Thus, our study could not properly assess whether pancreatitis rate is signi cantly related to stent types, and a large sample multicenter trial is necessary.
If biliary tumors involve the lower part of the common bile duct or the ampulla, it is inevitable that the stent will cross the duodenal papilla. However, stents placed across the duodenal papilla can block the discharge of pancreatic uid, leading to acute pancreatitis [20]. Generally speaking, the length of the stent to enter the intestinal tract should be controlled within 1 cm to avoid the stimulation and injury of the distal end of the stent to the contralateral duodenal wall of the ampulla. If biliary tumors are located in the middle part or above of the common bile duct, the stent should not cross the duodenal papilla as far as possible, so as to retain the important function of the duodenal papilla. At this time, the lower edge of the bracket should not be too close to the duodenal papilla. If the distance between them is < 2 cm, there may be a possibility of papilla spasm. Shunsuke Sugawara et al. [19] found that biliary stent implantation across the papilla for MBO would increase the morbidity of pancreatitis, which was consistent with the present study. Some scholars also believe that no matter whether the tumor involves the lower segment of the common bile duct and the ampulla, the stent should be placed across the duodenal papilla in order to facilitate drainage and reduce the incidence of cholangitis after stent implantation, which is still controversial and needs more clinical evidence to support [21]. When it comes to the fact that the disease indicates the need to stent across the papilla, clinicians can take preventive measures preoperatively in order to reduce the risk of pancreatitis.
PPIs combined with somatostatin are commonly used in the treatment of acute pancreatitis [22][23][24]. However, the prophylactic effect of using PPIs alone on the risk of acute pancreatitis is controversial since some studies suggest that its e cacy is limited to patients at high risk of postoperative pancreatitis. PPIs work on gastric wall cells and block acid production by inhibiting H + -K + -ATPase. However, a previous study revealed that PPIs therapy did not show bene t on alleviating systemic in ammatory response and clinical scores in severe pancreatitis patients [25]. In addition, gastric acid is an important barrier for the body to prevent abnormal colonization of bacteria in the upper digestive tract. Theoretically, inhibition of gastric acid secretion by PPIs may lead to abnormal composition and higher temporal instability of their intestinal microbiota, which may be an initiating factor for associated infections of pancreas.
The effect of PPIs consumption on the risk of pancreatitis after PTBS has not been investigated. According to a previous study reported by Carmelo Scarpignato et al. [26], PPIs did not affect the clinical course of acute pancreatitis, so they are not recommended routinely in this clinical setting [27,28]. In the present study, multivariate analysis showed an increased risk of postoperative pancreatitis was associated with preoperative administration of PPIs, and the difference was statistically signi cant. As a consequence, the management of PPIs abuse before surgery should be emphasized.
Several studies have shown that strategies such as the administration of periprocedural antibiotic could decrease the incidence of postoperative complications. Rty S et al. [29] found that antibiotic prophylaxis effectively decreases the risk of pancreatitis. Multivariate analysis in the present study revealed that preoperative administration of antibiotics has been linked to lower incidence and severity of postoperative pancreatitis, which is an independent protective factor.
There are several limitations in our study. First, this was a single-center retrospective study, in all patients with pancreatitis, the increase of amylase and lipase was detected in emergency laboratory only after presenting symptoms such as abdominal pain, nausea and vomiting. It cannot be excluded that some patients without clinical symptoms also have the increase of amylase and lipase. Thus, the authenticity of the assessed data was relatively low to some extent. Second, clinical treatment strategies, such as stent implantation position and the selection of stent, were decided by the physician. As a result, selection bias may have in ated the results. Additionally, the uneven surgical ability of different surgeons may also cause different prognosis.
In conclusion, biliary stent implantation across the duodenal papilla may signi cantly increase the risk of postoperative pancreatitis. Preoperative (within 3 days) administration of PPIs was another independent risk factor. Consequently, PPIs should be used with great caution before operation. Moreover, prophylactic administration of antibiotics before PTBS may reduce the risk of pancreatitis. Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
developed the project and helped to edit the manuscript. All authors read and approved the nal manuscript.