Endoscopic transpapillary drainage in disconnected pancreatic duct syndrome after acute pancreatitis and trauma: long-term outcomes in 31 patients

Disconnected pancreatic duct syndrome (DPDS) most commonly occurs as an adverse effect of acute necrotizing pancreatitis (ANP) or secondary to abdominal trauma [1]. The incidence of DPDS with ANP in our center was less than 2%. Chronic pancreatitis, pancreatic malignancies, and abdominal surgery can also cause DPDS. DPDS causes pancreatic ductal (PD) leakage owing to a complete discontinuity of the main pancreatic duct(MPD) and isolation of the entire upstream pancreatic segment (e.g., body or tail), which does not communicate with the duodenal papilla. Because of persistent pancreatic leak, diagnosis of DPDS is then considered, and clinically DPDS most commonly presents with a nonhealing external pancreatic fistula (EPF) with high amylase activity or peripancreatic fluid collections (PFC) refractory to conservative management. These conditions cause further complications including intra-abdominal sepsis, hemorrhage from peripancreatic blood vessels, and exocrine or endocrine pancreatic insufficiency [2‐4]. Thus, early diagnosis and treatment are important, which unfortunately are not possible in clinical practice.

Pancreatography (magnetic resonance cholangiopancreatography [MRCP] or endoscopic retrograde cholangiopancreatography [ERCP]) is the most reliable modality to accurately diagnose DPDS. An abrupt discontinuity in the MPD observed during ERCP is diagnostic of DPDS with or without contrast extravasation [5, 6].

Potential treatment for DPDS includes conservative, endoscopic, and surgical measures or the use of interventional radiology. Primary percutaneous drainage is being replaced by endoscopic drainage, which prevents EPF formation [6]. Although surgical and endoscopic procedures are equally effective in draining non-resolving symptomatic pancreatic pseudocysts, endoscopy is a cost-effective option associated with a shorter hospital stay [7]. Surgical management has demonstrated a mortality rate of 6–9% and a recurrence rate of 23–60% [8‐11]. To date, there is no consensus regarding the optimal endoscopic treatment for DPDS. Long-term indwelling transluminal stents are currently recommended, albeit with inadequate supporting evidence because this is a rare condition [5, 12‐14]. Endoscopic transpapillary drainage (ETD) was considered safe and effective treatment in cases of partial PD disruption where a bridging endoprosthesis could be used; however, it was shown to be ineffective to treat DPDS, with a low reported success rate of 26–43% [5, 12, 15, 16]. However, most of these studies were performed at a few expert centers and included a small cohort of patients with DPDS without long-term follow-up.

We report 31 consecutive patients with DPDS who underwent a diagnostic and therapeutic ERCP with long-term follow-up. Moreover, we propose that transpapillary stent insertion improves clinical outcomes in patients with DPDS when the site of disruption is not bridged.

We performed a retrospective analysis of a prospectively obtained ERCP database. We investigated every patient who developed DPDS after ANP or abdominal trauma between September 2008 and January 2016. Most of them were treated in local hospitals, they came to our center for endoscopic treatment of this delayed complication when the condition of ANP or abdominal trauma was stable. Electronic medical records were reviewed. Variables included patient demographics, clinical manifestations, etiology, therapy performed (operation, percutaneous or transmural drainage), ERCP findings and intervention, other endoscopic management, adverse events, clinical course and outcomes (success, failure, and number of procedures performed). The study protocol was approved by the Changhai Hospital Ethics Committee and it conforms to the provisions of the Declaration of Helsinki. All patients provided written informed consent. Pre-procedure evaluation of all disconnected viable pancreatic tissue was performed using cross-sectional imaging (contrast-enhanced computed tomography [CT] or MRCP). We included patients with ANP or trauma with an ERCP-proven diagnosis of DPDS.

An accurate preoperative diagnosis of DPDS requires both, cross-sectional imaging (CT or MRI) and pancreatography. Although secretin-enhanced MRCP is proposed as an alternative to ERCP to diagnose DPDS, its sensitivity in demonstrating the site of the ductal disconnection is lower than that of ERCP [17]. ERCP is the gold standard to identify ductal disruption, which can be further defined as partial (opacification of the PD upstream to the site of disruption) or complete (nonopacification of the PD upstream to the leak) [12, 18‐20]. Unlike MRCP, ERCP also serves as an endotherapeutic modality in selected patients with pancreatic injury. In this study, all patients routinely underwent MRCP or CT before ERCP and were diagnosed with DPDS using ERCP. An MRCP suggested DPDS in only 2 of 31 patients (Fig. 3). The median interval of DPDS in our study was 6.6 months, and much of this delay is attributable to a lack of awareness regarding the disease [21‐23].

In our study, the ANP group showed a higher rate of CT-guided percutaneous drainage than the trauma group (P = 0.006). Pancreatic necrosis-induced infection and ductal leakage-induced PFCs usually necessitate percutaneous drainage. In many of these cases, DPDS was overlooked initially, and the patients were only diagnosed with DPDS once they develop an EPF after percutaneous drainage. In this condition, percutaneous drainage alone should be avoided for patients with DPDS because of the inevitable development of an EPF in these patients. Dual modality drainage is a good treatment option for these patients if the cyst is properly located. It involves the placement of percutaneous drains followed immediately by endoscopically placed transmural stents into the PFC. After transmural drainage is completed, the percutaneous drain is also opened. Once the fluid collection resolves, the percutaneous drain is removed [24]. We also observed that PD disruption was significantly associated with percutaneous drain placement [25, 26]. However, a causal relationship between them remains unclear. Gallstones were the most common cause of ANP (10/15, 67%), although any cause of pancreatitis can produce ductal leakage. Additionally, in concurrence with previously published articles [5, 25], DPDS occurred predominantly in the head or neck of pancreas (22/31, 71%) owing to the unique pancreatic anatomy that increases the susceptibility of this region. This group showed a higher tendency to develop pancreatic atrophy than that in the group with distal PD disconnections (P = 0.009). This is perhaps because pathology of head/neck duct affects the pancreatic parenchyma to a greater extent than disruptions in the body-tail area, where a small amount of viable pancreatic parenchyma exists upstream to the ductal disruption. Two of the 10 patients with pancreatic atrophy underwent pancreatojejunostomy during follow-up. The two patients had neither ETD nor sphincterotomy. Both of them had no recurrence of pancreatic fistula or pseudocyst after operation, but one patient still has abdominal pain.

Conventional treatment such as percutaneous drainage and medical treatments show a low success rate and high morbidity; thus, these are no longer recommended for DPDS. Surgical treatment including resection and derived procedures is accepted as a standard modality, and most physicians recommend derivd techniques, which were used in 8 patients as the first-line surgical treatment for DPDS [10, 11, 22, 23, 27, 28]. However, this approach demonstrates a higher morbidity and mortality than a nonsurgical approach [8‐11, 29]. We attempted to identify predictors of successful outcomes following ETD to treat MPD disruption and observed that the best results were obtained in patients with partial PD disruption that can be bridged. However, in this study, 7 patients with DPDS could not undergo bridge treatment using stents but showed resolution of PFCs and clinical symptoms [12]. If these 7 patients can be considered clinically successful cases, the success rate of ETD in treating 23 patients with DPDS increases from 26 to 56.5%. It should be mentioned that it was impossibility to make bridging therapy in all cases in this study with the experienced doctors. Up to date, there was only a successful bridging in our center, which was reported in 2018 [30].

ETD was associated with clinical success in 87% (13/15) of patients with DPDS. This route of drainage is physiological because it uses the normal anatomical route of drainage of pancreatic secretions and does not involve the creation of an alternative route of drainage as is observed with transmural drainage. Thus, we attempted ETD before surgery, although 8 patients showed clinical failure and underwent surgery with complete cure without adverse events or recurrence during follow-up. ERCP can be associated with serious adverse events such as post-procedure pancreatitis and the risk of infection in sterile PFCs [31‐33]. This study shows that ETD is safe and effective; however, postoperative infection was reported in 1 patient with PFC, which could be initially treated with transmural drainage. The diameter of the MPD limits the number and the size of the endoprosthesis that can be placed for drainage. The optimal duration of stent therapy remains unclear [12, 18]. In this study, stents remained in place over 3 months and no adverse events such as stent occlusion or stent-induced ductal changes were observed.

To our knowledge, the number of patients who met the diagnostic criteria of DPDS and were treated with transpapillary drainage was less than 31 [1, 10, 12, 13, 21, 34]. Limitations of our study: 1) this was a retrospectively designed study, 2) a small number of cases were included. Because of the low incidence of DPDS, it is difficult to study large patient cohorts. Nevertheless, our sample size is larger than those in previous reports, and notably, the follow-up duration is long enough, 3) a potential selection bias toward endoscopic transpapillary management vs. transmural or surgical management cannot be ignored, which is attributable in part, to our large and experienced ERCP practice and, 4) other drawbacks include that 16.1% of the patients were lost to follow-up. Evidence regarding DPDS is limited and this group of patients included in our study might represent a selected patient population with DPDS who underwent conventional multidisciplinary treatment.

We present our experience with DPDS and emphasize that endoscopic management for DPDS is safe and is associated with favorable long-term outcomes. Additionally, the incidence of adverse events is low, and the failure rate was 13.3% (2/15). Thus, primary surgery is not required for all patients with DPDS. Systematic reviews and prospective, randomized multicenter trials are warranted in the future to validate the data regarding DPDS.