Chronic pancreatitis (CP), typically associated with ethanol abuse, is associated with a broad spectrum of clinical and pathologic features, all primarily related to the permanent alterations in the structure and function of the pancreas itself caused by this disease. Structural abnormalities include irregular dilated main pancreatic ducts; areas of stricture and mild dilatation in a nondilated duct; fibrosis in the gland, which may result in a large mass; and extraglandular abnormalities, such as pseudocyst formation. Functional derangements are also well described. Endocrine dysfunction, such as elevated postprandial fasting glucose advancing to frank insulin dependence, occurs in approximately 25% of patients. Exocrine insufficiency manifesting as steatorrhea is slightly more common, noted in approximately 40% of patients, and is more likely as the duration of disease advances. The most common symptom by far associated with this disease is chronic abdominal pain. ^1–5 This pain is characteristically midepigastric with referral to the back, particularly the left side of the upper back. In the most common manifestation, this pain is chronic and unrelenting. The chronic pain typically results in narcotic dependence. Narcotic dependence in a population with substance abuse tendencies renders proper management of this disease challenging. Narcotic dependence and the impact of disease on quality of life serve as measures of the severity of CP. In addition, visual analog scales have been applied to the management of this disease. The only interventions that have been recognized as effective in treating the chronic unrelenting pain have been surgical procedures. Although several nerve-ablation procedures have been described, their success rates have been much lower, and even the successes have not produced long-term efficacy. ¹ The surgical procedures generally recognized as achieving success are resectional procedures such as the standard Whipple resection, pylorus-preserving pancreaticoduodenectomy, distal pancreatectomy or so-called left resection of the pancreas, and a duodenum-preserving pancreatic head resection. ^4,5 The alternative surgical approach is drainage of the dilated main pancreatic duct, known commonly as a modified Puestow procedure, as described by Partington and Rochelle. ⁶ Recent modifications have added an element of resection to these drainage procedures. These include the so-called Frey procedure, in which a generous resection of a dominant mass in the head of the pancreas is accompanied by the longitudinal pancreaticojejunostomy. ⁷ Finally, Izbicki et al ⁴ recently described excavation along the body of the pancreas when the main pancreatic duct is nondilated so that a canyon of pancreatic parenchyma is surgically bypassed to a limb of jejunum. Although the results of individual series have varied, the overall success rates for relieving the chronic unrelenting pain of CP has been on the order of 75% to 95%. ^{1,3–5,8–16}

Although the primary indication for surgery in this disease is chronic daily unrelenting abdominal pain associated with chronic narcotic dependence, all pancreatic surgeons recognize a subset of patients who have recurrent exacerbations during the course of their disease. These exacerbations often require hospital admission and at times occur with increasing frequency to the point that the patient is spending many months in a year in the hospital and often has significant nutritional deficits. It is likely that every busy pancreatic surgeon has used surgical interventions to treat patients with this pattern of disease. Despite this fact, there are few data to help the surgeon advise the patient of the likelihood of success in abolishing such recurrent attacks. We have prospectively followed the course of disease in patients with CP since 1985 and previously reported data on these patients. ^14,15 Throughout our studies, we have observed several variables related to the natural course of this disease and specifically have monitored the frequency of patients with the pattern of recurrent exacerbations. Three subsets have evolved. There are patients with chronic unrelenting abdominal pain as their only manifestation of disease. There are patients who have chronic unrelenting abdominal pain and intermittent acute exacerbations that require hospital admission and are quite distinct from the daily unrelenting pain. Finally, there are patients whose manifestation of disease is recurrent exacerbations alone. It is the purpose of this report to evaluate success rates using various surgical procedures in achieving pain relief in the presence of chronic unrelenting abdominal pain, but more importantly to evaluate the success rates in ameliorating or abolishing recurrent exacerbations of CP. We believe that the clinical impact of this subset may have more pertinence to the question of the economic impact of this disease because the patients in this subset so often require hospital admission and nutritional support for extended periods.

In 1985 a clinic was created at The University of Texas Medical Branch and John Sealy Hospitals to evaluate and follow up patients with complex diseases of the pancreas, including carcinoma of the pancreas, cystic lesions of the pancreas, and CP. Each patient undergoes an initial evaluation that includes considerable demographic and clinical information and several important measures of function. After the initial evaluation, patients are followed at an interval of approximately 12 to 14 months, although many are evaluated more frequently as the clinical circumstances dictate. Appropriate measures are instituted to maximize the medical management of these patients. Attention is at all times specifically directed toward maximizing the nutritional status of all patients.

The specific clinical data obtained in all patients include the presence or absence of ethanol abuse, the presence of persistent unrelenting pain, the frequency of acute exacerbations of CP, hospital admissions required for such events, use of narcotic analgesics, and daily activity measures, such as maintaining employment status and general function.

Persistent pain is defined as upper abdominal pain radiating to the back with or without nausea or vomiting, and postprandial or unrelated to meals. Acute exacerbations are defined as episodes clearly distinct from the chronic daily pain either in magnitude or in character. For the purposes of this report, we defined acute exacerbation as one requiring hospital admission or cessation of oral intake for greater than 48 hours. Thus, recurrent attacks were characterized for the frequency of hospital admission during a 12-month period, the duration of each attack, and the impact on nutritional status, primarily reflected in weight loss caused by recurrent attacks after surgery and in follow-up. In addition, recurrent attacks were scrutinized to determine their relation to acute alcohol ingestion compared with spontaneous acute exacerbations.

Table 1. SURGICAL PROCEDURES

In terms of recurrent exacerbations, the indications for surgery included greater than three hospital admissions during 1 year or at least one episode of exacerbation requiring a hospital stay of more than 30 days in 1 year. Contributory indications for surgery included ongoing weight loss or need for nutritional supplementation, such as total parenteral nutrition or enteral feeding.

There were 104 patients in group 1, 71 in group 2, and 84 in group 3. Surgery was performed in 86 of the 104 patients in group 1, 64 of the 71 patients in group 2, and 49 of the 84 patients in group 3. Among the surgical procedures performed were 124 LPJ procedures, 46 pancreatic head resections, and 29 left resections (see Table 1). Of the 86 patients in group 1 who underwent surgery, 60 had LPJ, 15 had pancreatic head resection, and 11 had left resection. Of the 64 patients in group 2 who underwent surgery, 33 had LPJ, 19 had pancreatic head resection, and 12 had left resection. Of the 49 patients in group 3 who underwent surgery, 31 had LPJ, 12 had pancreatic head resection, and 6 had left resection.

Table 2. PAIN RELIEF

Table 3. NO ACUTE EXACERBATIONS SINCE SURGERY

Table 4. MEAN YEARLY RATE OF ACUTE EXACERBATIONS

The overall mean length of hospital stay before surgery in groups 2 and 3 was 17.3 ± 3.1 days (15.6 ± 1.7 for group 2, 19.2 ± 3.9 days for group 3). The overall mean length of hospital stay after surgery for acute episodes in groups 2 and 3 combined was 5.1 ± 1.1 days (4.2 ± 1.3 for group 2, 6.2 ± 2.1 for group 3). These data were calculated as a percentage of patients who had attacks and not as a percentage of all patients in groups 2 or group 3. The differences again achieved statistical significance.

Seventy percent of our patients had lost at least 10 lb before surgery. Eighty-four of the 165 patients who underwent surgery with weight loss regained their weight to predisease levels in follow-up.

Chronic pancreatitis associated with chronic unremitting abdominal pain has long been recognized as the primary indication for surgical intervention. ^1,2 Various surgical interventions have been described to address this clinical imperative. Many advocate a variety of resection procedures, most commonly the Whipple type of pancreaticoduodenectomy. ^{1,4,5,8–10,16} The criteria for choosing resectional therapy are the standard ones: patients with a “dominant mass” in the head of the pancreas, with or without ductal dilatation; patients with the “small duct variant” of CP; patients with suspected malignancy; and patients with failure of LPJ. Drainage procedures such as the one characterized by Rochelle and Partington’s modification of the Puestow procedure are also a well-accepted treatment for this chronic unremitting pain. ^{1,5,6,8,9,11–15} The indications for drainage procedure included a dilated main pancreatic duct (>6 mm) or the presence of ductal stones, both in the absence of a “dominant mass” in the head of the pancreas. Nerve ablation procedures, both in the thoracic sympathetic chain and in the parasympathetic celiac glands, have also been applied. ¹ Only rarely has an author identified the subset of patients with recurrent acute exacerbations of abdominal pain, and no uniform data exist regarding the success rates for each of these procedures in ameliorating or abolishing such exacerbations. A report by Warren ¹⁷ evaluated the outcome of surgery for “chronic relapsing pancreatitis.” His results offer some insight into this entity, but 28% of the population he reviewed had gallstone-related pancreatitis. We defined three distinct populations of patients with CP. Group 1 patients were the standard candidates for surgery in the world literature. Group 2 patients have probably been included in previous reports because they have chronic abdominal pain; however, a successful outcome in these patients must also identify improvements in the recurrent attacks in CP. A recent report by Beger et al ¹⁶ identified this subset of patients, but no segregation of the population was made to identify individual groups (i.e., patients with acute exacerbations vs. those with pain alone). Our data strongly suggest that all surgical procedures applied to patients with CP achieve improvement in the rate of recurrent acute exacerbations. We found the LPJ procedure to produce unusually favorable outcomes. Although the success rate of left resection was not impressive, the total number of patients undergoing this procedure was small. The indication for this procedure is segmental disease to the left of the spine. The procedure’s success rate for ameliorating or abolishing these exacerbations was not as high as its rate for achieving pain relief (overall 83% success for pain relief vs. 73% for abolition of recurrent attacks).

These data considerably improve the clarity with which a clinician can offer patients surgical alternatives. Although surgery achieves less impressive overall complete freedom from recurrent exacerbations, the total number of episodes per year and the duration of hospital stay for patients after surgery were consistently and significantly reduced compared with preoperative levels. Thus, although complete abolition of these events may not be achieved at as high a rate as for unremitting abdominal pain, there appears to be an advantage achieved with regard to the total number of episodes and the duration of episodes. Thus, one can speculate that the total economic burden created by recurrent exacerbations may be greatly reduced by surgical intervention. Unlike chronic unremitting abdominal pain, which rarely requires hospital admission, the economic impact of this intervention on this subset of patients reflects a significant contribution.

The question must arise regarding the possibility that these recurrent acute exacerbations simply reflect ongoing ethanol abuse. Our data suggest that only a small percentage (16/113 patients in groups 2 and 3) had their recurrent exacerbations caused by continued ethanol abuse. Further, our data regarding nutritional status suggest that additional benefits may be gained from this intervention beyond those reflected simply by the decreased frequency and magnitude of recurrent acute exacerbations.

The report by Beger et al ¹⁶ examined 504 patients with the diagnosis of CP who underwent surgery. This excellent review had a death rate of 0.8%; 79% of patients were rendered completely pain-free, and 93% had significant resolution of pain. This report comes closest to any report reviewed to addressing the question of acute exacerbations. The authors specifically evaluated the percentage of patients requiring emergency admission in the preoperative period. Sixty-nine percent of patients evaluated had a significant preoperative admission rate. Although the criteria for admission and even the definition of multiple readmissions were not clearly stated, it would seem likely that this population represents patients who have recurrent attacks combined with those whose disease is characterized by chronic unremitting pain. The postoperative readmission rate for pain in that report was 9%. This group was composed of patients undergoing duodenum-preserving pancreatic head resection. ¹⁶ In that report, the success rate for preventing readmission appeared to be better than the success rate for achieving complete pain relief.

The available options for a surgeon managing a patient with CP and chronic unremitting abdominal pain are well chronicled. Multiple exhaustive reviews have established a reasonably stable success rate for these interventions. ^1,2,4,5 It is thought that the advantages of drainage procedures in treating CP are the reduced rates of death and complications and the preservation of pancreatic tissue, in view of the fact that this disease at its end stage is associated with functional deficits. The range of success rates for ameliorating or abolishing pain using a pancreaticojejunostomy range from 77% to 93% in collected series. ^{1,4–6,8,11–15} Many studies have included lengthy follow-ups, and many have had a death rate of 0%. Although some of these studies defined resolution of pain as being pain-free and free of narcotic use, others used a visual analog scale to determine improvement in abdominal pain. Unfortunately, this may confuse the interpretation of success rates for pain relief.

Surgical resection has existed as a modality in the management of CP for decades. ^{1,4,5,8–10,16} Long ago, the concept of a 95% resection of the pancreas was popularized, but few practitioners use this modality today. Significant functional deficits were the rule, and outcomes were variable. The more common resectional therapy applied to CP had been the classic Whipple resection, replaced more recently by the pylorus-preserving pancreatic head resection. ^{1,3–5,8–10} A multitude of studies in the surgical literature has established the safety and efficacy of the procedure. In general, the success rates for pain relief have ranged from 52% to 94%. ^{1,3–5,8–10} In recent years, the death rate has been less than 5%. In the many large series that include a variety of surgical procedures, the choice of resection versus drainage varies widely. Our study involved a considerable preponderance of drainage procedures over resections. Interpretations of this variation may point to differences in the presenting disease in various geographic regions: it appears that a dominant mass is far more common in Germany, and small duct disease is far more common in Britain. ^1,3

More recently, further modifications of the pancreaticoduodenectomy have been proposed. Beger is credited with developing the duodenum-preserving resection of the pancreatic head. Over more than 20 years, his experience has confirmed the efficacy and safety of this procedure, and several studies have compared that procedure with the standard Whipple and the pylorus-preserving pancreatic head resection. A more recent variation, the so-called Frey procedure, has been described in which a longitudinal incision in the main pancreatic duct is accompanied by an excavation of the dominant mass in the head of the pancreas without dividing the neck of the pancreas. Once again, the aim is to preserve as much pancreatic tissue as possible. The outcomes for this procedure have not been significantly different from those for a standard resection or a drainage procedure. ^4,7 Finally, most recently, Izbicki et al ⁴ have developed a procedure for patients with small duct disease in which a V-shaped incision is made along the body of the pancreas down to the main pancreatic duct as an alternative to the resection of the head. The success rate for this procedure, relatively early, suggests that an acceptable outcome can be achieved in patients with small duct disease who would otherwise be treated by a major resection. Once again, with all of these studies, we have found none that specifically addresses the question of acute exacerbations.

In conclusion, our data can help clinicians estimate the likelihood of ameliorating or abolishing recurrent acute exacerbations in patients with CP. Certainly, some of these patients are affected by continued ethanol abuse, but our data suggest that this by no means represents a majority of patients with this complaint. The precise mechanism involved in the recurrent acute attacks superimposed on chronic abdominal pain in CP has never been fully elucidated. They are often unassociated with enzyme elevations and thus do not appear to reflect episodes of acute pancreatis superimposed on CP. We advocate surgical therapy for patients with CP whose manifestations of disease are either recurrent exacerbations alone or accompanying chronic intermittent abdominal pain.

Dr. David B. Adams (Charleston, South Carolina): Thank you very much, Dr. Baker. Surgical treatment of chronic pancreatitis is a palliative treatment, and if the patients are not readmitted to the hospital with recurrent pancreatitis and pain or complications associated with diabetes, they frequently suffer diseases associated with alcohol and cigarette abuse, lung cancer, esophageal cancer, cardiomyopathy, or traumatic deaths. The remarkable feature of this study is that it is a single-surgeon experience and it represents an experience over a long period of time with outstanding follow-up that really is unparalleled in the literature. It has all the assets of a prospective study.

I have a few questions related to Dr. Nealon’s interest in this issue. One has to do with the issue of pancreatic burnout. We have all been taught that eventually chronic pancreatitis will burn out, and you don’t have to operate on these patients. Many of us are still waiting to see this patient. The other issue relates to improving exocrine and endocrine insufficiency associated with chronic pancreatitis with early surgery. Dr. Nealon has an interest in this, and I would appreciate his comments on that.

Another point relates to quality-of-life issues related to chronic narcotic use. We have found that many patients continue to view their quality of life as very good, despite the continued use of narcotics. Would you comment on this as an issue related to surgery for chronic pancreatitis?

In relation to technical details of the operation, would you give your views of the Beger procedure in terms of treating disease on the right side of the pancreas as well as the Frey procedure combining that with the LPJ? And then, finally, would you just take a moment to prognosticate into the future for us and comment on the use of a laparoscopic lateral pancreaticojejunostomy and, as well, perhaps earlier intervention in this disease with total pancreatectomy with auto-islet transplantation? Thank you for your attention.

Dr. Charles J. Yeo (Baltimore, Maryland): I, too, would like to rise to congratulate Dr. Nealon for an interesting and fairly novel analysis focusing on a little-studied group of patients. These are pancreatitis patients with acute exacerbations of pain often requiring hospitalization. Using his prospectively collected database, the authors have analyzed the outcomes after various operative interventions, and they have observed the best outcomes following longitudinal pancreaticojejunostomy, followed by pancreatic head resection, and, lastly, by distal pancreatectomy. I have four questions for Dr. Nealon.

What do you think are the physiologic mechanisms of these acute exacerbations in chronic pancreatitis? Are we trying to treat inflammation mediated by cytokines? Are we treating ductal obstruction? Are we treating neural stimulation from various nerve growth factors, or are there other processes going on?

Two, your database would seem to offer the opportunity to formally assess quality of life, as Dr. Adams mentioned: functional outcomes, employment status, endocrine dysfunction, etc. Are you planning to address any of these issues? Because I think you have the opportunity to do that in a wonderful database.

Third, your data analyses have lumped together pancreaticoduodenectomy and various duodenum-sparing pancreatic head resections under the term “pancreatic head resection.” If you analyze these two different operative procedures separately, do you generate any differences in outcome?

And, last, I have a question about the threshold or indication for operation, comparing between groups 2 and 3. In group 2, defined as chronic, unrelenting pain and acute exacerbations with six events per year, the chance of being operated on was 90%. In contrast, in group 3, defined as acute exacerbations only with 7.8 events per year, the chance of being operated on was only 58%: much lower. This suggests to me that different criteria perhaps applied to group 2 and 3 patients. How do you explain these differences? Are they reflected by differences in microscopic findings at surgery or microscopic findings at the time of resection?

I very much enjoyed this paper, and I really think it does have a particular niche to fill in an area where patients have not been well studied and well reported. My congratulations.

Dr. F. Charles Brunicardi (Houston, Texas): Dr. Baker, Secretary Townsend. I’d like to thank Dr. Nealon for the opportunity to discuss this excellent study and providing in advance a copy of the manuscript. Dr. Nealon is to be congratulated for thoughtfully presenting results of essentially a prospective trial with data gathered from 259 patients with chronic pancreatitis over a 14-year period. He is able to present such data because he had the foresight to establish a pancreas clinic in 1985 at the University of Texas Medical Branch and John Seeley Hospitals, in coordination with his mentors, Dr. Thompson and Dr. Townsend. Dr. Nealon has become the complete pancreatic surgeon, mastering diagnostic and therapeutic ERCP as well as a host of pancreatic operations.

My first question is, how many of these patients had preoperative ERCP by you, and did this influence the choice of your operation? Dr. Nealon demonstrated overall postoperative pain relief was 83%, with 86% for pancreaticojejunostomy and 91% for pancreatic head resections. Would you comment on your outstanding results with the modified Puestow procedure, and how do you choose between the Puestow procedure and pancreatic head resection? 92% of your patients have alcohol-induced pancreatitis. It is sad to note that 43% of these patients relapse into alcohol abuse following their operation. Would you comment on this relapse rate and whether this influences your willingness to operate? Furthermore, are the acute exacerbations related to recurrent alcohol abuse, and are you reporting an epiphenomenon?

I would like to thank the Association for the privilege of the floor.

Dr. William H. Nealon (Galveston, Texas): I’d like to thank the discussants for their kind comments and interesting questions.

To start with, Dr. Adams mentions what I think irks almost all pancreatologists, particularly surgeons involved in the management of chronic pancreatitis. Because if you agree with Rudi Aman from Zurich, if we just hold our knives and wait long enough, everybody gets better. It is my finding, as it has been many other students of disease finding, that in fact burnout has been relatively rare. And I do believe I have even come upon a flaw in Dr. Aman’s data which I’m not going to share with you today, but I think really well explains why he has a slightly different outcome in his subset of patients.

Dr. Adams asked about functional changes and mentioned what, unfortunately, many have misinterpreted in my early data, which is that we get improvement in function after operation, and my data have only shown that we have actually a delayed loss of function and that, over time, certainly many of the patients having operation have lost function, which I am certain some of you surgeons have found as well. I will say that we continue to find that the rate of loss of function is considerably delayed after a ductal decompression. And we continue to believe that is a meaningful measure.

Quality of life was mentioned by Dr. Adams and by Dr. Yeo. And you may know that a paper with Dr. Sohn as well from the SSAT a year and a half ago looked very carefully at quality of life. We actually have developed our own version of a quality-of-life measure, and I have not included it in the manuscript, but very strong data to show—it includes three primary categories, which is employment status, functionality within the family, and what we have called functionality within the community. And data suggests that all of these have had at least favorable outcomes after operation.

Dr. Adams and Dr. Yeo mentioned the Beger procedure, the so-called duodenum-preserving pancreatic head resection. My experience with head resections here includes only 14 of these procedures. My findings are not different from those in other institutions where, especially in Europe, this has been compared to the pancreatic head resection, where the differences—there have been essentially no differences. So the effectiveness of the duodenum-preserving pancreatic head resection, I think, is reliable but it is not, clearly, any different. I have not performed any Frey procedures.

Dr. Adams mentioned laparoscopic approach, and I have to admit that each day I have to further admit the likelihood that bigger and better operations can be done laparoscopically. With the exception of pseudocysts, I have not yet been sufficiently intrepid to consider further, more advanced laparoscopic procedures on the pancreas, but many of you know these have been performed, including a Whipple. So that I guess anything is possible. I can’t easily determine from any data I have as to how far they will go.

Dr. Adams asked about early operation. I am actually an advocate of early operation if in fact we believe, and our data support, that preservation of function or at least delayed loss of function is a possible goal. Sutherland from Minnesota has looked at total pancreatectomy and islet cell transplantation. I have no experience with that.

Dr. Yeo asked an interesting question about what do I suspect is the mechanism of acute exacerbation. My sense, honestly, is that I have always had trouble explaining the mechanism of chronic pain. I think it is a little easier for us to imagine that partially obstructed secondary ductals have a stimulation that may result in acute exacerbation. It’s a little easier for me to imagine that than why do they have pain every single minute of every single day, particularly when the ducts are not completely obstructed. We know that microscopic evaluation has shown that there seems to be a clustering of inflammatory cells around nerves. There seems to be something specific about the disease and inflammation around nerves that could explain some of the pain associated, but if that is true, why does decompression help?

Reeber, of course, has presented the question of the compartment syndrome and showed that both intracellular and intraductal and intraparenchymal pressures that are elevated in his experimental model and seem to result in an acidosis or an anaerobic metabolism going on in the pancreas that may explain the acute inflammation and has shown in nice studies with cats that he has done actually a Puestow procedure in the cat and shown that all these effects are reversed, particularly the intraparenchymal pressures are reversed. I would only borrow from other people’s data to answer that question, however.

I have already mentioned quality of life and the pancreatic head resection.

Your question, Charlie, sort of explains why I undertook this study. You mentioned that I operated on 90% of group 2, which have the conventional indication of chronic abdominal pain. I have only operated on 50% of group 3, which I believe takes you outside the edge until now, if you read the world literature. Because the world literature has not said what do you do when someone has purely intermittent acute exacerbations. I am now more aggressive, so maybe that other 42% is in trouble.

I performed ERCP on 247. They obviously are helpful in establishing a diagnosis. I perform a head resection on a patient with a dominant mass in the head of the pancreas, the very occasional question of whether it is or isn’t a carcinoma and failed Puestow procedure.

Continued ethanol abuse of 43% is actually—I sometimes lightly mention that the detox program such as you see at the famous centers quotes a long-term success rate of 33% for reducing ethanol abuse. Apparently, chronic pancreatitis is considerably more effective than those measures at reducing ethanol abuse. And I think it is an important question, are these acute exacerbations simply a reflection of a binge? And our data suggests that that is the case sometimes, but by no means most of the time, in this subset of patients. It seems that there is a subset who have this pattern of the disease.

I thank the Association for the privilege of the floor.