Wall stent expansion in the pancreatic duct following endoscopic insertion. All stents were expanded fully. The apparent expansion beyond 10 mm was due to magnification on the abdominal radiograph without an appropriate reference for measurement correction.
Radiographs of the wall stent at the time of endoscopic placement (A) and on the morning of operation (B, arrow).
Self-assessed pain score of the patients based on average daily symptoms described during follow-up interviews.
Madura JA, Canal DF, Lehman GA. Wall Stent–Enhanced Lateral Pancreaticojejunostomy for Small-Duct Pancreatitis. Arch Surg. 2003;138(6):644-650. doi:10.1001/archsurg.138.6.644
The purpose of this study was to see if a small (<7 mm) pancreatic duct could be dilated to an acceptable diameter, allowing lateral pancreaticojejunostomy to decompress the pancreatic duct and relieve pain, while preserving pancreatic endocrine and exocrine function.
Patients with chronic pancreatitis who had a small main pancreatic duct underwent progressive trans-ampullary dilation of the duct and subsequent placement of an expandable metallic wall stent (wallstent; Boston Scientific Microvasive Division, Natick, Mass). Approximately 14 days later, a lateral pancreaticojejunostomy was done.
A 400-bed university referral center hospital in an urban setting.
Thirty-five patients were selected from a large group with chronic pancreatitis. Thirty-one had pancreas divisum. All patients had undergone transendoscopic sphincterotomies and stenting before being accepted into the study. All had endoscopic retrograde cholangiopancreatography–proven chronic pancreatitis, and all ducts were observed to be 7 mm or smaller.
Patients were selected after endoscopic sphincterotomy and stenting failed. Progressive transendoscopic duct dilation with plastic stents was carried out, and a 10-mm expandable metallic wall stent was placed prior to surgical decompression. Lateral pancreaticojejunostomy was performed.
Main Outcome Measures
Patients were observed for pain relief, postoperative symptoms, analgesic use, glucose intolerance, and quality of life. All patients were seen or contacted by telephone, and their results were recorded.
There were no operative deaths, but 26% of patients had complications. Seventy-one percent of patients reported that their pain was better than preoperatively. Three patients had subsequent pancreatic surgery. No new cases of diabetes occurred except in the 2 patients who underwent total pancreatectomy.
In general, most patients feel that their lives were improved by the procedure. A quarter of the patients no longer take narcotics, and many have been able to resume a relatively normal lifestyle. Although this procedure is not a panacea for all patients with chronic pancreatitis and a nondilated duct, it is a reasonable alternative to resection.
A MAJOR ROLE of surgery in the treatment of chronic pancreatitis is relief of pain, which is based on the assumption that the pain is due to ductal hypertension. Procedures can be divided into the categories of duct-enteric drainage and resection of part or all of the pancreas. More recently, duodenal-preserving proximal resection combined with ductal drainage has been proposed.1,2 With better perioperative care, mortality and morbidity rates for all of these procedures have substantially improved. Unfortunately, preservation of endocrine and exocrine function has been foiled by the continued abuse of alcohol by many patients in the reported studies.
Most surgeons make operative decisions based on the anatomic configuration of the pancreatic ductal system. With proximal stricture and head-dominant disease, a dilated duct will usually lend itself to lateral pancreaticojejunostomy. If the disease is located primarily in the head of the duct, many surgeons will use proximal resection to give a better chance of long-term pain relief despite probable loss of pancreatic function. However, in patients with a nondilated duct less than 6 or 7 mm in diameter, lateral pancreaticojejunostomy has not proved to provide long-term anastomotic patency, and most surgeons would avoid a drainage procedure in that situation.
Because resection of a portion of the pancreas in patients with chronic disease may sacrifice an actively functioning gland, endocrine and exocrine dysfunction may be accelerated. An attempt to preserve the functional pancreas in patients with a small duct using progressive dilation of the duct and subsequent lateral pancreaticojejunostomy was proposed in a group of patients, selected from a large number of individuals with pancreas divisum and normal ductal anatomy, who had chronic pancreatitis and a small duct. Additionally, these patients had all received some measure of relief from ductal stenting via endoscopic retrograde cholangiopancreatography (ERCP). Dilating the duct to at least 1 cm allowed mucosa-to-mucosa anastomosis and anticipated long-term patency. This article details the procedure and its outcome.
Between 1994 and 2001, more than 4000 patients underwent diagnostic ERCP at the Indiana University Medical Center (Indianapolis). Pancreas divisum was the diagnosis in 1572 patients, and 646 patients underwent minor papilla sphincterotomy. During this period, 2523 patients had ERCP for chronic pancreatitis and its sequelae. From this pool of patients with residual pain following endoscopic diagnosis and treatment for pain from chronic pancreatitis using endoscopic sphincterotomy and ductal stenting with plastic stents, we selected 35 individuals.
The major selection criteria were a small (<7 mm) main pancreatic duct and transient but significant relief from pain while the plastic stents were patent. There were 23 women and 12 men, with a mean ± SD age of 45.8 ± 11.0 years. Thirty-one patients had pancreas divisum with stenotic or strictured dorsal ducts at the minor papilla, whereas 4 patients had the more common ductal anatomy with stenotic major papillae. Of this latter group, only 1 patient had suspected heavy alcohol use. Thirty-three patients were white, 1 was Asian, and 1 was Hispanic. The average length of symptoms was 84.3 months, and these included abdominal pain (100%), back pain (66%), nausea (77%), vomiting (69%), diarrhea (40%), and weight loss (31%). All patients were regularly using narcotics as well as antiemetics (37%), pancreatic enzymes (60%), and various psychotropic agents (46%). All of the patients had previously undergone diagnostic and therapeutic ERCP, and all patients had had 2 or more endoscopic sphincterotomies and transendoscopic insertion of plastic stents through the major and/or minor pancreatic ductal papillae. These patients were observed to get substantial but transient pain relief from initial stent placement.
After recurrence of pain following several stent placements, these patients agreed to undergo ductal expansion and lateral pancreaticojejunostomy. The process of sequential dilation of the main pancreatic duct was begun by initial insertion of size 5F to 7F catheters followed by progressively larger endoscopically placed stents. The stents were changed every 1 to 2 months or when they appeared to have become occluded by the occurrence of increased pain or pancreatitis. The last temporary stent placement consisted of two 5F to 7F plastic catheters. Finally, 2 weeks before the intended surgical procedure, a 10-mm × 68- to 80-mm-long expandable metallic wall stent (WallStent; Boston Scientific Microvasive Division, Natick, Mass) was placed with approximately 5 mm remaining in the duodenum, and the patient was scheduled for surgery. The plastic stents were well tolerated with only a few patients requiring 1- to 2-day hospitalizations for increased pain or pancreatitis. No patient developed significant pancreatitis or pain as a result of the metallic stent placement. Radiographs of the abdomen were obtained at the time of stent placement and on the morning of operation to assess the adequacy of ductal expansion.
Perioperative antibiotic prophylaxis was given to all patients. Eleven patients who still had gallbladders underwent cholecystectomy, and 3 patients underwent placement of a Stamm3 gastrostomy tube. The pancreatic duct containing the metallic stent was located with simple palpation and needle aspiration or more commonly using intraoperative ultrasonography and subsequent needle aspiration. The duct was then opened longitudinally from the tail of the pancreas to within a few millimeters of the duodenal wall. The stents were easily removed in toto in all patients except for those remaining in place longer than 20 days. These had to be removed piecemeal but were successfully extracted in their entirety in all patients. A biopsy of the pancreas in mid body was done in 23 patients. Careful hemostasis was carried out and included suture ligation of the pancreaticoduodenal arteries on the anterior surface of the head of the pancreas. Next a 45-cm Roux-en-Y limb was created 45 cm from the ligament of Treitz and placed through the transverse mesocolon. A 2-layer isoperistaltic anastomosis between the jejunal mucosa and pancreatic ductal mucosa was done using absorbable sutures on the mucosal layer and 3-0 silk sutures on the serosa with a modification of the Puestow procedure developed by Partington and Rochelle.4 Drains were placed in the lesser sac and were used to monitor the patient for anastomotic leaks. In 1 patient who had previously undergone splenectomy and caudal pancreatic drainage, the caudal pancreaticoenteric anastomosis was totally occluded. The Roux-en-Y limb was easily converted to a lateral pancreaticojejunal anastomosis.
Although postoperative care was routine, we observed the patients for pancreatitis, intra-abdominal leaks, and sepsis. No patients were given octreotide postoperatively. After discharge, the patients had follow-up visits at 1 month, 3 months, and 1 year as well as additional visits for the investigation of postoperative complaints. Postoperative ERCP was not done routinely but was performed when indicated for recurrent pain or documented pancreatitis.
Thirty-five patients underwent wall stent–enhanced lateral pancreaticojejunostomy between 1996 and 2002. All wall stents expanded to full diameter by the time of surgery (Figure 1 and Figure 2). There were no perioperative deaths, and thus far there has been no long-term mortality. Significant perioperative complications occurred in 9 patients, the most serious in 2 patients with bleeding from branches of the splenic and pancreaticoduodenal arteries into the pancreatic duct, necessitating reoperation. Two patients developed a leak from the pancreatic-enteric anastomosis, and 1 developed subsequent multisystem organ failure. One of the patients with intraductal bleeding also developed multisystem organ failure and adult respiratory distress syndrome. Both of these patients recovered after long stays in the intensive care unit. Other complications included bleeding from a drainage site in the subcutaneous tissue (1), abdominal abscess (1), superior mesenteric artery compression of the duodenum (1), and atelectasis (1) (Table 1).
All patients were seen in the follow-up clinic, and their symptoms and complaints were recorded. Each available patient was contacted by telephone and responded to a questionnaire regarding pancreas-related symptoms, general state of health, current use of medications, and a self-assessment of current condition. Continued use of 1 or more narcotics was reported by 74% of patients (Table 2). Twelve patients are currently employed at their previous occupations, 3 are retired, and 10 are on disability. The remaining patients were not employed prior to surgery. Patient self-assessment revealed that 71% feel better than before their surgery, 13% are the same, and 16% are worse (Figure 3). Fifteen patients have had repeated ERCP, and anastomotic patency was demonstrated in all of these patients. Three patients have had further pancreatic surgery because of persistent or recurrent pain: 2 patients had a total pancreatectomy and islet cell transplantation, and 2 patients underwent pylorus-preserving pancreaticoduodenectomy, 1 of whom later had a total pancreatectomy and islet cell transplantation. One patient had thoracoscopic splanchnicectomy, 1 had several incisional hernial repairs, and 1 had adhesiolysis with insertion of a jejunal feeding tube at another hospital. Additionally, several patients underwent placement of enteral feeding tubes because of postprandial pain and inability to maintain their weight (Table 3).
In the last 7 patients, a double-pigtailed stent was used to prevent stricture and stenosis in the most proximal pancreatic duct, which had been seen in several patients earlier in the study. Two of these stents were removed because of unexplained recurrent pain, but no improvement in symptoms was observed. Two patients were diabetic (1 was insulin dependent) before surgery and remain so postoperatively. The 2 patients who had total pancreatectomy now require insulin. No other patient in this group has developed glucose intolerance. The use of pancreatic enzyme replacement is more difficult to assess; 21 patients were receiving enzyme replacement preoperatively, and 16 are currently. Diarrhea is present in 43% of patients, and enzyme replacement does not seem to alleviate it.
Overall assessment would categorize 71% of patients as achieving excellent to good results, 14% as fair, and 16% with a poor outcome, including the 3 patients who required further pancreatic surgery. Cessation of narcotic use in a third of the patients suggests a beneficial outcome in this difficult-to-treat group.
The primary goals of surgical therapy in patients with chronic pancreatitis are relief of pain and return to as normal a lifestyle as possible. Therapeutic options can be divided into 2 main groups: resection (distal, proximal, or total), or drainage of an obstructed pancreatic duct. The latter group of procedures includes ablation of the pancreatic papillary sphincter mechanism, distal retrograde drainage, and lateral pancreaticojejunostomy. Pancreatic duct-enteric anastomosis has generally been reserved for patients with a duct diameter of 7 mm or greater. Most of the literature reports poor results when anastomotic attempts are carried out in smaller ducts. Rios et al5 described a group of patients with small ducts similar to those in our study. No attempt was made to dilate the main pancreatic duct preoperatively, and 76% of patients reported that their pain was the same as or worse than before the procedure. Although only a few reports describe follow-up ERCP in patients with drainage procedures, in the study by Kugelberg et al,6 caudal panceatostomy resulted in anastomotic stricture in most patients, whereas the anastomoses in patients who had undergone lateral pancreaticojejunostomy remained patent. One patient in our series had previously undergone caudal retrograde pancreatic drainage, and that anastomosis was completely obliterated.
Metallic stents have commonly been used in patients with malignant strictures of the biliary tree, but the short survival in these patients makes long-term patency assessment impossible. The use of stents in benign conditions of the pancreatic duct is more limited. Eisendrath and Deviere7 reported the use of both covered and uncovered metallic mesh stents in patients with benign strictures of the main pancreatic duct. Follow-up investigations demonstrated progressive endothelial ingrowth in all patients, especially at the sites of ductal strictures. The conclusion of their study was that this therapy should not be applied to benign strictures at this time. In our study, retention of the ductal metallic stent longer than 21 days resulted in ductal ingrowth into the stent and difficulty in its removal.
In this study, a mucosa-to-mucosa anastomosis was performed uniformly. Little information about anastomotic techniques is detailed in the literature, and no prospective trials have been done. Nonetheless, careful mucosal approximation seems to be a more secure method to prevent anastomotic leaks and assure long-term patency.
Our patients were younger than those usually described in chronic pancreatitis series, and most had minor papillary obstruction in the setting of pancreas divisum. The large number of patients with pancreas divisum reflects the pattern of referral to the ERCP service at our institution. The lack of ductal dilation may have been due to parenchymal fibrosis and inflammation, preventing ductal dilation or short-term restenosis. This suggests that in the face of significant pancreatic parenchymal disease, perhaps ductal decompression alone is insufficient to render all patients pain-free.
Grodsinsky et al8 reported a series of 63 patients of whom 30 did not have a dilated duct seen on preoperative ERCP; these individuals all underwent subtotal distal pancreatectomy. Izbicki et al9 described a series of 13 patients with small-duct chronic pancreatitis in whom longitudinal V-shaped excision of the ventral pancreas with anastomosis to a Roux-en-Y limb was carried out. Because the duct size was demonstrated to be less than 2 mm in diameter, duct-enteric anastomosis was not possible. At a median follow-up period of 30 months, 92% of patients reported complete relief of pain. The authors attributed the success of this procedure to a wide excision of the pancreas, allowing direct drainage of the secondary and tertiary ducts of the pancreas into the jejunal limb. The advantages of pancreatic preservation would appear to be improvement or preservation of endocrine and exocrine function in drainage procedures, in contrast to proximal or distal resective therapy. However, Bradley and Nasrallah10 investigated fat absorption before and after duct-enteric anastomosis and found no improvement with surgical decompression. Nealon et al11 studied patients preoperatively and after longitudinal pancreaticojejunostomy and found a delay in the progression of the disease when compared with controls who did not undergo the operation. An extensive review of more than 150 articles from the available literature (Table 4 and Table 5) confirms the advantage of pancreatic preservation by ductal drainage. Although long-term relief from pain and rehabilitation outcomes are relatively similar for ductal drainage and resection, postoperative endocrine and exocrine function seem to be better maintained when all pancreatic tissue is preserved (Table 6, Table 7, and Table 8).
Not all of the patients in our study were rendered pain-free, but a quarter of them were able to function without narcotic pain medication and with fewer trips to the hospital or emergency department. No new glucose intolerance was produced, and no clinically apparent change in fat malabsorption was seen.
In patients with major or minor papillary restenosis and a nondilated pancreatic duct, ductal expansion by means of an expandable metallic wall stent is relatively safe and well tolerated. It allows these patients to undergo lateral pancreaticojejunostomy rather than pancreaticoduodenectomy or subtotal pancreatectomy.
For symptomatic patients with evidence of terminal pancreatic duct restenosis (major or minor papillae), treatment choices include sphincterotomy, sphincteroplasty, stricture zone resection, or surgical decompressive procedures that bypass the obstruction. Most patients who undergo endoscopic or surgical sphincter therapy have a stricture that extends several millimeters up into the pancreatic duct and makes repeated sphincter therapy unlikely to succeed. Our study suggests that most of these patients benefit from lateral pancreaticojejunostomy. Long-term pain relief following this procedure appears to be comparable with that achieved after usual pancreaticojejunostomy with a dilated duct or more extensive parenchymal resection, although this does not preclude further resective therapy in case of therapeutic failure. This procedure has not resulted in glucose intolerance or exocrine insufficiency.
Randomized prospective comparative trials will be required to better define the optimal therapy for this difficult-to-treat group. We caution readers to consider this therapy for patients with a clear-cut therapeutic response to prior sphincter ablation and stenting and who have equally clear endoscopic or manometric evidence of restenosis. Additionally, experimental studies with membrane-coated removable metallic stents are being conducted, and these stents may allow maximal dilation of pancreatic ductal strictures with subsequent removal. These stents may also be easier to remove during lateral pancreaticojejunostomy procedures.
Corresponding author and reprints: James A. Madura, MD, 9525 Copley Dr, Indianapolis, IN 46260 (e-mail: email@example.com).
Accepted for publication February 21, 2003.
This study was presented at the Western Surgical Association Annual Scientific Meeting, Vancouver, British Columbia, November 18, 2002, and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.
Michael B. Farnell, MD, Rochester, Minn: The authors' hypothesis is that the patency of lateral pancreaticojejunostomy requires a precise mucosa-to-mucosa anastomosis. For small-duct chronic pancreatitis, this is generally not possible. However, the authors have proposed a creative solution to this dilemma. In 35 patients with chronic pancreatitis and intractable pain secondary to pancreas divisum or sphincter of Oddi dysfunction, they progressively dilated the pancreatic duct with transendoscopic stents and ultimately deployed a 10-mm expandable metallic stent 14 days preoperatively. At operation the expandable stent was removed, and the resulting dilation allowed a 2-layer mucosa-to-mucosa anastomosis.
The patients were assessed postoperatively at 1, 3, and 12 months by questionnaire relative to pain relief, symptoms, analgesia use, and quality of life. In 15 of the 35 patients, postoperative ERCP was performed and confirmed anastomotic patency in each of those 15. There were no perioperative deaths and to date no late mortalities. The serious morbidity rate was 26%. The authors report that 71% of the patients were improved, 13% were the same, and 16% were worse. The authors also concluded that long-term relief of pain with their approach appears to be comparable to that achieved with lateral pancreaticojejunostomy in patients with dilated ducts or more extensive pancreatic resection in patients with small-duct disease. Moreover, they point out that their proposed operation does not preclude further resective therapy and therapeutic failures and to date has not resulted in glucose intolerance or exocrine insufficiency.
I want to congratulate Dr Madura for a novel approach to a very difficult subset of patients; namely, those with pancreas divisum and chronic abdominal pain. It is laudable that he and his colleagues have attempted to design an operation that both relieves pain and preserves pancreatic function. That said, I do have a number of questions and concerns that should be addressed before this approach can be endorsed for patients with small-duct chronic pancreatitis and intractable pain.
While the Mayo Clinic is a midwestern medical center, as is Indiana University, we must be located outside the pancreas divisum belt. In a series of 105 patients on whom we performed pancreaticoduodenectomy for small-duct current pancreatitis, only 1 patient had pancreas divisum as the etiology. At Indiana University in an 8-year period, no less than 646 patients underwent therapeutic endoscopic procedures for pancreas divisum or sphincter of Oddi dysfunction. The 35 patients in this series were culled from that experience.
This leads me to my first question. Did the 31 patients with pancreas divisum have chronic pancreatitis as a result of their anatomic anomaly, which we know is present in 7% of the population, or is it the result of repetitive endoscopic interventions? How many patients had documented elevated amylase levels prior to therapeutic endoscopy? What are your endoscopists' criteria for performing therapeutic intervention in patients with pancreas divisum and chronic abdominal pain?
The experience in the literature relative to lateral pancreaticojejunostomy for small-duct disease has been mixed at best. Izbicki and colleagues have reported excellent results with the longitudinal V-shaped excision of the ventral pancreas for small-duct disease in severe chronic pancreatitis based on the hypothesis that a compartment syndrome is responsible for the pain. Alternative hypotheses include perineural information and ductal hypertension. Which pain hypothesis, Dr Madura, do you subscribe to, and on which is your operation based?
The authors report that the majority of patients are improved and that their results are comparable to results for both resective and drainage procedures, as reported in the literature. While 71% of the patients reported that they were improved, I remind the audience that 69% of these patients were still taking narcotics, and of the 25 patients who were employed at the time of operation or prior to operation, fully 40% remain medically disabled.
Is your operation durable; that is, will it stand the test of time? In the manuscript I must have missed the duration of follow-up, but I was pleased to note the mean follow-up in your presentation today. My other question had to do with biopsy of the pancreas, and I think you told us that 20 of 23 had severe fibrosis.
I found this paper to be both provocative and stimulating. Dr Madura and colleagues are to be congratulated for attempting to provide a surgical solution for a subset of patients that the pancreatic surgical community find particularly vexing: patients with pancreas divisum and chronic abdominal pain.
It is worth repeating that the vast majority of patients coming to operation with chronic pancreatitis and intractable pain have etiologies dissimilar to the group reported in this series. Resection for small-duct disease and drainage for patients with dilated ducts provide the mainstay of surgical treatment.
I encourage Dr Madura and colleagues to continue to examine this highly select group of patients in order to more precisely define the pathophysiology of chronic pancreatitis in patients with pancreas divisum.
Richard A. Prinz, MD, Chicago, Ill: Most surgeons do a resective procedure for small-duct disease in chronic pancreatitis, and I would imagine that you have done some resections for this problem before you were doing this current approach. Have you compared the results of your pancreatectomies with wall stent dilation of the duct and subsequent lateral pancreaticojejunostomy?
With lateral pancreaticojejunostomy for alcoholic pancreatitis, a number of gastroenterologists have proposed placing a stent to see if there is a favorable response in terms of pain relief and then going ahead with your operation in those patients who did improve. Did you find such a response in your patients, and did you use that as a determinant for operative intervention?
Most of the patients in your series did not have alcohol as an etiology of their disease. I wondered if there is a difference in the outcome in those patients who had alcohol as an etiology vs those who did not. Have you studied any of these patients subsequent to your operation to see if the pancreaticojejunostomy remains open?
Finally, in this era when we are all concerned about costs, stents are extremely expensive, and you are using a number of them. Do you have any idea about the cost of preparing the patient for your operation?
Theodore X. O'Connell, MD, Los Angeles, Calif: Dr Madura, I just have 1 question. Do you have any data at all to say what the advantage is of adding the operation to the wallstent or simply leaving the wallstent in place without doing the Puestow operation?
Merrill T. Dayton, MD, Salt Lake City, Utah: Probably none of us really understand the pathophysiology of the pain that patients with pancreas divisum have, but I have always assumed that it was related to stricturing of the duct of Santorini. Dr Madura, did your dilatation procedures ever involve dilating up the minor duct and opening that up as well?
Dr Madura: I would like to thank Dr Farnell and all of the discussants for their insightful comments and questions.
As you can see, this group is very highly selected from a huge group of individuals studied for abdominal pain thought to be from pancreas divisum and/or sphincter of Oddi dysfunction. One of the concerns that we had about presenting this study was that surgeons would start using this technique without better long-term follow-up, and I assured Dr Farnell that those who do not have a skilled endoscopist like Dr Lehman doing ERCPs and placing stents would not be able to perform this procedure very successfully.
Dr Farnell, these patients did not come to us with acute pancreatitis. Rather, they came to us with chronic abdominal pain and were seen initially by the ERCP service, who started out by doing sphincterotomies and pressure measurements across the minor papilla.
Dr Dayton, I think that also answers your question. There was only 1 patient whom we could determine was an alcohol abuser, but he also had pancreas divisum. The way these patients were finally selected was if they had relief from the initial stent that Dr Lehman placed—which was usually a 5F plastic stent—if they got relief for several months, then it was felt that probably the etiology of the disease was a result of proximal obstruction and ductal hypertension. The factor that we could never determine is how much effect the inflamed and fibrotic pancreas has on the continued symptoms in many of our patients. Dr Farnell asked whether this disease process occurred congenitally or if it is produced iatrogenically. Well, I keep accusing our endoscopists of causing these problems by doing sphincterotomies and placing occluding stents in the main pancreatic duct in the head of the pancreas. But these patients had pain and pancreatitis prior to the initial studies. One of our colleagues, Dr Stuart Sherman, demonstrated that if you put stents in the ducts of the pancreas in dogs, they get significant fibrosing pancreatitis. So in fact, we may be treating a disease we have created.
The Izbicki series that Dr Farnell mentioned is a small one from Germany, about 30 patients in whom a ventral wedge of pancreas was resected followed by an onlay pancreaticojejunostomy. Dr Izbicki reports that 90% of the people are pain-free after follow-up of about 3 years, and I think that is interesting. I can't explain his results either, especially since he did not do a duct-to-mucosa anastomosis. Our average follow-up is 4 years, as you saw. Some patients had an immediate response, which lasted months, and in several patients it was a year or 2, but then their pain recurred. The narcotic issue, I think, is a difficult one. These patients have been on narcotics for 6 or 7 years, and some of them might be expected to be on narcotics forever no matter what procedure you did, but nonetheless we take continued narcotic use as somewhat of a failure.
One of the patients who had a subsequent Whipple resection still got no pain relief and then went on to a total pancreatectomy. Fewer patients are on tranquilizers and antiemetics, but still we have a narcotic problem. We have not had any evidence of anastomotic failure. Fifteen patients have had an ERCP postoperatively, and all of the anastomoses are patent. These were not done prospectively but in patients who had complaints, and it was thought that perhaps an anastomotic stricture might be the problem.
The intraoperative biopsy in 23 patients demonstrated 20 with severe pancreatitis and 3 in whom the pancreas appeared normal. As I mentioned in the talk, there was no difference in how they fared when compared by histopathologic appearance.
Dr Prinz, I answered some of your questions. The outcome of the patient who abused alcohol was no better than for the other patients, and we did study the ERCPs carefully preoperatively. I have no idea what these stents cost, I must admit. I am sure they are expensive, particularly the expandable wallstent and the expenses associated with multiple ERCP sessions.
Dr O'Connell asked about leaving a wallstent in place as the only therapy. There is only 1 study I can find in the literature, by Eisendrath, who put in metallic wall stents with and without a superficial coating, and all of them had ingrowth of ductal epithelium into the stent within a few months. This was done experimentally in animals, and as a result he did not do this in patients and concluded that this was probably not a good long-term solution in benign disease. The stents become very difficult to remove, as we found out in several of our patients who had them in longer than 20 days.
We do not have similar outcome data on the numerous patients in whom we have performed resection for chronic pancreatitis. We feel a randomized prospective study on these difficult patients is appropriate to see which is a better treatment.