Recurrent chronic obstructive pancreatitis is attributed to pancreatic flow obstruction caused by PD stenosis, pancreatolithiasis, pancreas divisum, etc., and the consequent increased internal pressure in the caudal PD [1-3]. Methods of PD decompression include surgical procedures, such as pancreatectomy and pancreaticojejunostomy, as well as endoscopic PD stenting combined with extracorporeal shock wave lithotripsy (ESWL). Previous reports showed favorable results of PD stenting, with success rates of 72-100% and symptom relief rates of 65–87% [4-11]. At present, minimally invasive endoscopic therapy is widely employed [12]. However, prospective randomized controlled trials comparing endoscopic therapy and surgical drainage have been conducted and shown surgical drainage to be more effective [13,14]. Further studies are needed for optimal selection of therapeutic methods.

Endoscopic PD stenting combined with ESWL is itself a minimally invasive procedure and is in widespread use. However, there are problems with diagnosing whether a stenotic lesion of PD is benign or malignant, deciding on whether the procedure is indicated, techniques for management of severe stenosis, types of stents to be implanted (diameter and form), duration of implantation, and complications. After a stent is implanted especially, in more than a few cases it cannot be removed and must repeatedly be replaced with a new one. If such cases can be predicted before stenting, surgical drainage might be applied as an initial procedure. In this study, we retrospectively examined the current status of endoscopic therapy for recurrent CP, especially endoscopic PD stenting, at our department. Moreover, we aimed to analyze patient background databy dividing patients undergoing PD stenting into a group in which the stent was removed and a group in which the stent could not be removed,remained implanted, and thushad to repeatedly be replaced with a new one.

Patients

Between April 2006 and April 2012, endoscopic drainage was performed in 86 patients with recurrent CP (436 procedures). Among them, 66 patients who received some form of treatment for PD were selected as the study population (410 procedures; 57 men (86%) and 9 women (14%) with a mean age of 59 ± 14 years [range, 34-80 years]); causes of recurrent CP were alcohol consumption in 58 patients (87%), pancreas divisum in 1 (2%), postoperative anastomotic stenosis in 1 (2%), and idiopathic in 6 (9%) (Table 1). The breakdown of the 20 excluded patients was as follows: 11 undergoing only endoscopic retrograde cholangiopancreatography (ERCP) and cytology, 6 undergoing only endoscopic bile duct stenting, and 3 in whom cannulation of PD was not performed (due to post-pancreatectomy in 1, groove pancreatitis in 1, and only choleducholithotripsy in 1).

Table 1: Patient Characteristics (n=66).

Diagnosis of chronic pancreatitis

RecurrentCP was diagnosed based on images obtained by extracorporeal and endoscopic ultrasound (EUS), computed tomography (CT), magnetic resonance imaging (MRI), and ERCP in addition to biochemical data. In all patients, the imaging findings of PD met the definition of severe pancreatitis according to the Cambridge criteria [15].

Indications for endoscopic pancreatic duct drainage

At our hospital, endoscopic PD drainage is indicated for (1) symptomatic patients, (2) asymptomatic patients whose pancreatic function may be preserved, and (3) patients with alcoholic pancreatitis who can abstain from drinking.

ESWL

When the disease was attributable to pancreatolithiasis and calcification, ESWL was performed as needed. An electromagnetic lithotripter (Dornier Lithotripter S, Dornier MedTech, Wessling, Germany) was used for 41 patients. The locations of the stones were placed in the shock-wave focus employing an X-ray focusing system. ESWL was performed as needed according to the size and number of stones and followed by endoscopic therapy.

Classification based on initial ERCP findings

Based on the initial images of PD, recurrent CP was classified into 3 types: pancreatolithiasis (Stone) type in 7 patients (11%) (Figure 1a), PD stenosis (Stenosis) type in 18 (27%) (Figure 1b), and Stone+Stenosis type in 41 (62%) (Figure 1c). There were 59 patients (89%) with PD stenosis (Table 2).

Figure 1: Classification based on initial pancreatic duct images.
a. Pancreatolithiasis (Stone) type. b. Pancreatic duct stenosis (Stenosis) type. c. Stone+Stenosis type.

Table 2: Initial ERCP findings.

For all patients with the Stone type, ESWL was performed as needed, and stones were removed with a basket or a balloon catheter. A PD stent (5- or 7-Fr) was used only for prevention of impaction. Unless there are residual stones, it would be unnecessary to leave the stent implanted. For patients with the Stenosis type, a PD stent was implanted to drain pancreatic juice and to dilate PD. Because it is especially important to differentiate theStenosis type from pancreatic cancer, cytology and biopsy were performed. For 34 patients with the Stone+Stenosis type, stones were fragmented by ESWL and removed. A PD stent was implanted to drain pancreatic juice and dilate PD. Because removal of pancreatic stones is often difficult in the presence of PD stenosis, lithectomy was performed as soon as possible. Even when there were residual stones, a stent was implanted. Removal of residual stones was postponed until the next procedure.

Crossing the stenosis with a guidewire

ERCP was performed with a JF-240, a JF-260V or a TJF-260V (Olympus Medical Systems Corp., Tokyo, Japan) under conscious sedation with diazepam and pethidine. We attempted to cross the stenotic lesion with a guidewire (JagwireTM High Performance Guidewire, Boston Scientific Corp., Natick MA, USA). Upon successfully crossing the stenosis, Intra Ductal Ultra Sonography (IDUS) after the initial procedure, brushing cytology, pancreatic juice cytology, and biopsy were performed to exclude cancer. When crossing of a stenotic lesion was difficult due to the severe stenosis, an approach via the minor papilla was attempted.

Dilation of stenosis

After crossing the sites of stenosis, the stenotic lesions were dilated with a dilation catheter (6-,7-, or 9-Fr, Soehendra Biliary Dilation Catheter, Cook Medical., Bloomington, IN, USA) or a dilation balloon catheter (Hurricane TM RX Balloon Dilation Catheter, 6 mm in diameter and 2 cm in length, Boston Scientific Corp., Natick MA, USA). In case of difficultywith dilation, a Soehendra Stent Retriever (5- or 7-Fr, Cook Medical., Bloomington, IN, USA) was used to attempt dilation of stenotic lesions.

Pancreatic duct stenting

After dilation, a straight (GeenenPancreatic Stent Sets, Cook Medical., Bloomington, IN, USA) or S-ShapedPancreatic Stent (Olympus Medical Systems Corp., Tokyo, Japan) 5-, 7-, 8.5-, or 10- Fr polyethylene stent with multiple side holes was implanted in the stenotic lesion.

Lithectomy

The stones in PD were removed with a basket catheter (Flower Basket or TetraCatch, Olympus Medical Systems Corp., Tokyo, Japan) or a balloon catheter (Offset Balloon Catheter, Zeon Medical Inc., Tokyo, Japan) during ERCP. In order to facilitate lithectomy, papillotomy was applied to the bile duct orifice and the PD orifice.

Treatment protocol

Every 3 months, a stent was removed, and pancreatography was performed for assessment. In patients with residual stenosis, a stent was implanted again. At this point, a stent with a larger diameter was used if possible. Finally, the stent should be removed within 1 year. When symptoms of pancreatitis were caused by stent occlusion, the stent was replaced. When there were residual stones, they were removed as needed at the time of replacing the stent.

Study items

The current status of endoscopic PD drainage at our hospital was retrospectively examined by comparing patients according to the above-described classification based on the images of PD. The study items were patient characteristics, outcomes after therapy, complications, and follow-upresults. Out of the 66 patients undergoing endoscopic therapy, 7 patients with Stone type, 5 with treatment failure, 1 undergoing only bile duct stenting, and 1 undergoing only dilation of PD were excluded; and the remaining 52 were divided into the stent-removed group of 24 and the stent-maintained group of 28. In the stent-maintained group, the following patients were excluded: 5 with less than 3 months of stenting, 1 with failure of continuation of stenting, 2 deaths from intercurrent disease, and 1 drop-out case. The remaining 19 patients were ultimately included in the stentmaintained group and compared to the 24 patients in the stentremoved group (Figure 2).

Figure 2: Study items.

Results were expressed as means (SD) or as a percentage of the total number of patients. A χ-squared analysis or a two-tailed Fisher’s exact test was used to compare differences between the two groups. A P value of less than 0.05 wasconsidered to be significant. All analyses were performed using statistical software (Stat View Ver.5.0, SAS Institute, Cary, NC, United States).

Success rate of the procedure

The procedure was successful in 61 patients (92%), unsuccessful in 5. In 4 of the 5 patients, although the procedure was not completed due to large stones or PDbeing full of stones, PD dilation decreased. In other words, PD dilatation improved in 65 patients (98%). The breakdown of the procedures was concomitant ESWL in 41 patients (62%), PD stenting in 51 (77%), bile duct stenting in 14 (21%), and an approach via the minor papilla in 14 (21%) (Table 3).

Table 3: Procedural success rates and breakdown of treatment methods.

Outcomes after therapy

Complete pain relief with no requirementfor analgesics was achieved in 60 patients (91%). Partial pain relief with reduction in the doses of analgesics was achieved in 5 patients (7%). There was no improvement in 1 patient (2%). Thus, effects, to some degree, were observed in 65 patients (98%) (Table 4).

Table 4: Outcomes after Therapy.

Complications

We examined a total of 407 procedures including 229 PD stenting procedures. The early complications were ERCP-induced pancreatitis in 11 patients (2.7%, all mild in severity) and hemorrhage, basket impaction, and rupture of PD in 1 patient (0.2%) each. The late complications were pancreatic ductitis in 3 patients (0.7%), stent displacement in 2 (0.5%), PD stent migration in 4 (1%, retrieved in all patients), bile duct stent migration in 2 (0.5%, retrieved in all patients), and tear during removal of PD stent in 4 (1%, retrieved in 3 patients) (Table 5).

Table 5: Complications (total procedures: 407, ERPD: 229).

The pancreatolithiasis (Stone) type vs. the types including pancreatic duct stenosis

Comparison between Stone type and the types including pancreatic duct stenosis (Stenosis type or Stone+Stenosis type) revealed no significant differences in sex, age, or causes. In Stone type, the mean number of procedures was significantly smaller (P=0.0133), and there were significantly fewer patients who required at least 1 year of treatment (P=0.0043). No difference was observed in the incidence of complications (Table 6).

Table 6: Stone type vs. the types including pancreatic duct stenosis.

The pancreatolithiasis+pancreatic duct stenosis (Stone+Stenosis type)type vs. the pancreatic duct stenosis (Stenosis) type

Comparison between Stone +Stenosis type and Stenosis type revealed no significant differences in sex, age, causes, the number of patients with failed PD stenting, or the number of patients in whom a PD stent was removed. In Stenosis type, the mean number of procedures was significantly smaller (P=0.0423), and the incidence of complications was also significantly lower (P=0.0366). However, there was no difference in the number of patients who required at least 1 year of treatment (Table 7).

Table 7: Stone+Stenosis type vs. Stenosis type.

Stent-removed group vs. Stent-maintained group

Comparison between the stent-removed and stent-maintained groups revealed no significant differences in sex, age, causes, concomitant ESWL, the mean number of procedures, implantation of a large stent with a diameter (thicker than 8.5-Fr), or the incidence of complications. In the stent-removed group, the number of patients who required at least 1 year of treatment was significantly smaller (P= 0.0285) (Table 8).

Table 8: Stent-removed group vs. Stent-maintained group.

In this study, male patients with alcoholic recurrent chronic obstructive pancreatitis accounted for approximately 90% of the subjects. Although this figure is consistent with those in previous reports, the number of female patients with alcoholic chronic pancreatitis has also been increasing in recent years [1-12]. When alcohol consumption is a cause of pancreatitis, an important condition for treatment is whether patients can abstain from drinking. Although it can be assumed that drinking will often be resumed after alleviation of symptoms, such as pain,in many cases, it is difficult to assess the degree of compliance with abstinence.

Moreover, 90% of patients with recurrent chronic obstructive pancreatitis had stenotic lesions. Approximately 80% of the patients required PD stenting, and 60% required concomitant ESWL. There are also reports describing favorable outcomes as having been achieved with only lithotripsy, such as ESWL, without stenting [16-18]. However, because our study included many patients with stones combined with stenosis, we considered lithotripsy alone to be insufficient treatment. Because the basis of managing recurrent chronic obstructive pancreatitis is treatment for stenosis, it was suggested that a combination of pancreatic duct stenting and ESWL is necessary.

Moreover, endoscopic stenting combined with ESWL was effective for pain relief, was associated with a low incidence of complications of endoscopic therapy, and could be performed safely and effectively. This result is consistent with the favorable 84% success rate of drainage/pain relief with stenting reported by Costamagna et al. [19]. Endoscopic therapy combined with ESWL was demonstrated to be an extremely effective therapeutic method in the short term.

Based on the comparison of patients classified according to the initial images of PD, those with Stone type had a significantly smaller number of procedures and significantly shorter treatment duration than patients with pancreatic duct stenosis (Stone+Stenosis type or Stenosis type). Because therapeutic effects and outcomes are favorable in patients with only pancreatolithiasis, they may be good candidates for endoscopic therapy combined with ESWL. Moreover, compared to patients with Stenosis type, those with Stone+Stenosis type included more patients who required ESWL, and they tended to undergo a significantly larger number of procedures and had a significantly higher incidence of complications. There were also refractory cases and those in which a stent could not be removed. Treatments, including surgical drainage, for patients with Stone+Stenosis type, who accounted for 62% in this study, may need to be reconsidered.

In this study, the success rates and outcomes of endoscopic PD drainage were favorable and appeared to be satisfactory as short-term effects. In terms of long-term effects, however, the inability to remove a stent would require regular replacement of stents and consequently reduce quality of life for patients. We believe that surgical drainage should have been selected for patients whose stent could not be removed. It would be better if patient in whom stent removal would be difficult could be predicted beforehand. In this regard, further studies are needed.

In hopes of achieving dilation of a stenotic lesion by stenting, we have attempted the implantation ofPD stents with the largest possible diameter (8.5- or 10-Fr). However, comparison between the stentremoved and stent-maintained groups suggested that stents with a larger diameter might not contribute to dilation of stenotic lesions. In recent years, implantation of a covered metallic stent based on this premise of removal has been reported [20,21]. Because a dilating effect can be expected, this procedure may be useful in cases with pancreatolithiasis in the caudal pancreatic duct. However, there are concerns regarding the duration of implantation and complications, such as migration, displacement, and difficultyremoving the stent. Studies focusing on the safety of this strategy are needed.

In the stent-removed group, the duration of implantation tended to be shorter than that in the stent-maintained group. In caseswith solitary pancreatic duct stenosis or pancreatolithiasis, early treatment may prevent progression of pathological conditions. For patients who would require long-term implantation of a stent, it may be better to consider surgical drainage given thepossible complications associated with stent replacement and adverse effects of long-term stenting, as well as medical costs.

In order to achieve favorable outcomes with endoscopic therapy, selection of patients appears to be extremely important. Farnbacher et al. reported that the only index predicting long-term remission of pain after endoscopic drainage is short disease duration [22]. Future studies should focus on how to determine whether endoscopic therapy or surgical drainage should be selected as the initial treatment.