A 49-year-old female had mildly icteric sclera since 1986, her ultrasonic examination showed a small gallbladder stone. Intermittently she had choleretic and litholytic herbal medicine in the following 9 years. In 1995, her ultrasound result showed both gallbladder stones and polyps. Then cholecystectomy and T-tube choledochotomy was performed, which no stones and strictures were found in biliary duct. Four years after cholecystectomy, this patient had intermittent nausea and dull epigastric pain. With the pain radiated to the back and nausea worsened after eating, she lost 7.5 kg in 2 months. The serum amylase level remained normal but urine amylase increased to 844u/L (Somogyi). Abdominal computed tomography (CT) scan revealed a dilated proximal biliary system, which demonstrated the presence of biliary infection and chronic pancreatitis (CP). After 2 months of conservative treatment, percutaneous transhepatic cholangial drainage (PTCD) and anti-infective therapy, her worsened symptoms had no obvious relief. Therefore, RYHJ had to be performed in 1999. The Roux limb is prepared by transecting the jejunum and the mesentery 15 cm distal to the ligament of Treitz. End-to-side anastomosis was performed between expanded common hepatic duct (CHD) and the afferent intestinal limb at 5 cm from the sutured intestinal end. Then end-to-side anastomosis was performed between the proximal and distal jejunum (40 cm from the biliary-intestinal anastomosis). After operation her recurrent episodes of pain and nausea were alleviated. And long-term oral administration of digestive enzymes was applied continually.

However, the patient has had 3 times of recurrent acute pancreatitis (AP) from 2002 to 2011. The second time was the most serious episode that she suffered from AP combined with acute suppurative cholangitis (ASC). Each time endoscopic sphincterotomy (EST) was performed, endoscopic retrograde biliary drainage (ERBD) tube was installed, and the tube was removed 3 days after the procedures. During the third time of AP in 2011, her abdominal CT showed choledocholithiasis, dilated CBD stump and AP (Supplementary Fig. 1A). ERCP showed filling defect in the CBD remnant. The stone extracted from the bile duct stump was grayish-white in color, rough and horny in contour and very hard (Supplementary Fig. 1B). After operation her symptoms were relieved.

The 79-year-old woman was admitted again to our hospital with a 2-month history of upper abdominal pain spreading through to the back in November 2016. Physical examination revealed no abdominal tenderness and icteric sclera. Laboratory blood tests, including blood cell count, liver function tests, and so on, were in normal ranges. Abdominal CT scan and magnetic resonance cholangiopancreatography (MRCP) revealed filling defects in CBD stump, CP, and dilatation of CBD stump, and main pancreatic duct (MPD) (Fig. 1A–B). Therefore, an elective ERCP exploration was performed.

During the ERCP, a balloon catheter was pulled through the major duodenal papilla (MDP) after EST was performed, and no stone was found, then pancreatic stent was placed for the stricture in the opening of the MPD in the following procedures. Next, the balloon catheter was pulled through the CBD, and white crushed stones were removed from the CBD remnant into the duodenum (Fig. 1C–D). It looks like pancreatic duct stones that have white grain of sand and hard texture. Then endoscopic nasobiliary drainage (ENBD) tube placement was performed after clearance of CBD stump stones. Elemental analysis and infrared spectrophotometry demonstrated that the main constituent of the calculi showed in Supplementary Figure 1B and Figure 1D was calcium carbonate, identical to the common chemical composition of pancreatic duct stone. As CBD has been transected for RYHJ, how did this stone be created? We thought pancreatobiliary reflux induced by several ESTs was the main cause of choledocholithiasis, but we found other reasons for the occurrence of CBD stump stones during the latest ERCP.

After the papilla was identified, cannulation and Iopromide injection sailed through for cholangiopancreatography. There were filling defects that considered to be crush stones in the distal CBD stump, dilation of CBD stump, and stricture in the opening of MPD with proximal duct dilation on fluoroscopy. And there are 2 separate exit holes for the bile duct and pancreatic duct after several ESTs. Surprisingly, cannulation proceeded easily from MPD to CBD through a variant pancreatic duct (Fig. 1E), which did not display clear by reviewing the MRCP images. As far as we know, this sort of pancreatic duct variant which confluences with CBD as an incidental finding has never been reported before. What is the type and cause of this sort of stones? We nominated it as “pancreatogenic choledocholithiasis,” and pancreatobiliary reflux caused by a variant pancreatic duct may be the main cause. After operation her symptoms were relieved, and a 9-month follow-up indicated no remaining stones or lithiasis relapse. The chronicle of disease course and treatment was shown in Figure 2. However, pancreatobiliary reflux persists as the ERCP procedure could not ligate or plug up the variant pancreatic duct. Therefore, the treatment of pancreatobiliary reflux needs further investigation.