REVIEW

Management of Severe Acute Pancreatitis A Surgical Odyssey Edward L. Bradley III, MD,* and Nadine D. Dexter, MLS, AHIP† Abstract: Throughout much of history, surgery of the pancreas was restricted to drainage of abscesses and treatment of traumatic wounds. At the turn of the 20th century under the impetus of anesthesia, such surgical stalwarts as Mayo Robson, Mickulicz, and Moynihan began to deploy laparotomy and gauze drainage in an effort to salvage patients afflicted with severe acute pancreatitis (SAP). Over the next thirty years, surgical intervention in SAP became the therapy for choice, despite surgical mortality rates that often exceeded 50%. When the discovery of the serum test for amylase revealed that clinically milder forms of acute pancreatitis existed that could respond to nonoperative therapy, a wave of conservatism emerged, and, for the next quarter century, surgical intervention for SAP was rarely practiced. However, by the 1960s, conservative mortality rates for SAP were reported to be as high as 60% to 80%, leading surgeons to not only refine the indications for surgery in SAP, but also to consider new approaches. Extensive pancreatic resections for SAP became the vogue in continental surgical centers in the 1960s and 1970s, but often resulted in high mortality rates and inadvertent removal of viable tissue. Accurate diagnosis of pancreatic necrosis by dynamic CT led to new approaches for management. Some surgeons recommended restricting intervention to those with documented infected necrosis, and proposed delayed exploration employing sequestrectomy and open-packing. Others advocated debridement early in the course of the disease for all patients with necrotizing pancreatitis, regardless of the status of infection. In the 1990s, however, a series of prospective studies emerged proving that nonoperative management of patients with sterile pancreatic necrosis was superior to surgical intervention, and that delayed intervention provided improved surgical mortality rates. The surgical odyssey in managing the necrotizing form of SAP, from simple drainage, to resection, to debridement, to sequestrectomy, although somewhat tortuous, is nevertheless an notable example of how evidence-based knowledge leads to improvement in patient care. Today’s 10% to 20% surgical mortality rates reflect not only considerable advances in surgical management, but also highlight concomitant improvements in fluid therapy, antibiotics, and intensive care. Although history documents the important contributions that surgical practitioners have made to acute pancreatitis and its complications, surgeons are rarely complacent, and the recent emergence of minimally invasive techniques holds future promise for patients afflicted with this “. . . most formidable of catastrophes.” (Ann Surg 2010;251: 6 –17)

f Acute pancreatitis is the most terrible of all the calamities that occur in connection to the abdominal viscera. The suddenness of its onset, the illimitable agony which accompanies it, and the mortality attendant upon it, render it the most formidable of catastrophes. —B. Moynihan, 19251 From the *Department of Clinical Sciences (Surgery), Florida State University College of Medicine, Tallahassee, FL; and †Reference Library, Charlotte Edwards McGuire Medical Library, Florida State University College of Medicine, Tallahassee, FL. Nadine D. Dexter is currently at Harriet F. Ginsberg Health Sciences Library, University of Central Florida College of Medicine, Orlando, FL. Reprints: Edward L. Bradley, III, MD, 1600 Baywood Way, Sarasota, FL. E-mail: [email protected]. Copyright © 2009 by Lippincott Williams & Wilkins ISSN: 0003-4932/10/25101-0006 DOI: 10.1097/SLA.0b013e3181c72b79

6 | www.annalsofsurgery.com

S

urgical intervention in necrotizing pancreatitis is a relatively recent phenomenon, encompassing only the last 6 generations of surgeons. However, the most significant advances in diagnosis and surgical management have occurred within the lifetime of contemporaries. The delay in application of operative principles to acute necrotizing pancreatitis can be attributed to a combination of factors; delayed recognition of necrotizing pancreatitis as a disease entity, the remote retroperitoneal location of the gland, difficulties in differential diagnosis, confusion in terminology, and incomplete understanding of the metabolic needs of seriously ill patients. Nevertheless, surgical mortality rates in necrotizing pancreatitis have been dramatically reduced, from 60% to 80% in the early 20th century, to the 10% to 20% rates commonly reported today. This is the history of that progress.

THE PAST f The past is only the present become invisible and mute —M. Webb, Foreword to Precious Bane Beginning with the time of Galen (129 AD–217 AD), and continuing well into the Middle Ages, such diverse diseases as pancreatic malignancy, chronic pancreatitis, and various forms of acute pancreatitis and its complications were all lumped together under the heterogeneous classification of “scirrhi” (hard) conditions. The longevity of ancient medical dogma was considerably abetted by the edict from the medieval Catholic Church prohibiting human dissection, as well as to the unwillingness of scholars to challenge the classic opinions of Galen. As a result, clinicopathologic differentiation among human diseases did not begin to emerge until late in the fifteenth century, when human postmortem dissection began to be practiced in Bologna, Padua, and other European university centers. The first recorded description of necrotizing pancreatitis is apparently that of Nikolaus Tulp, a Dutch physician and anatomist, who in 1652 performed a postmortem examination upon a young man afflicted by an apocalyptic attack of abdominal pain that proved to be fatal after 5 days.2 At autopsy, the gland was found to be enlarged, purulent, and “rotten.” No explanation was offered for the observed changes in the pancreas. Similar anatomic studies were infrequent, however, and acute inflammation of the pancreas continued to be considered a rare condition. Failure to recognize acute pancreatitis as a separate disease entity was partly because symptoms were often attributed to other, more well-known diseases, and partly due to the relatively small number of postmortem examinations being performed. It was not until 1842 that acute pancreatitis was proposed as a distinct clinical entity. Heinrich Claessen, a physician in Cologne, collected a series of 6 fatal cases of severe acute pancreatitis from the literature of the times.3 He noted a common clinical and pathologic presentation among these cases, and suggested that this commonality in presentation might be useful for diagnosis. Over the next 50 years, increasing numbers of autopsy-based studies of necrotizing pancreatitis by such eminent pathologists as Annals of Surgery • Volume 251, Number 1, January 2010

Annals of Surgery • Volume 251, Number 1, January 2010

Rokitansky, Klebs, Friedrich, and Chiari, confirmed Claessen’s initial clinical observations.4 –7 Theodor Klebs, a student of Virchow, and professor of pathologic anatomy and bacteriology in Berne, proposed that hemorrhagic pancreatitis was related to necrotizing pancreatitis, and to subsequent suppuration of the gland. Moreover, he first postulated that the destruction of the pancreas was caused by the well-known “corroding qualities” of the gland secretions. Nikolaus Friedreich, also a student of Virchow, who later rose to become professor of pathology in Wurzberg, suggested a postmortem classification of acute pancreatitis in 1882 consisting of hemorrhage, gangrene, and abscess formation. His classification was admittedly restricted as it was based upon but 4 cases, only one of which was personally observed by the author. A year later, Hans Chiari, professor of pathology in Prague, published his findings concerning the pathophysiology of acute pancreatitis, attributing the destruction of the gland to the “autodigestive” properties of pancreatic ferments. In 1889, Reginald Fitz, the Shattuck Professor of Pathoanatomy at Harvard, former pupil of Rokitansky, Virchow, and Billroth, and already well known for his clinicopathologic description of appendicitis, published a Hunterian Lecture in which a pathology-based classification system for acute pancreatitis was proposed (Fig. 1). As a result of Fitz’ classification system, clinicians were enabled to make an antemortem diagnosis of acute pancreatitis.8 Although this autopsy-based classification system, including hemorrhagic, gangrenous, and suppurative forms, was superficially similar to the earlier one proposed by Friedrich, Fitz’s

FIGURE 1. Reginald Huber Fitz (1843–1913). American pathologist who studied under Rudolf Virchow, and brought the European concept of the importance of the microscopic examination of tissues back to America. As the Shattuck Professor of Pathoanatomy at Harvard, he established the first clinicopathologic classification of acute appendicitis. His subsequent pathology based classification of acute pancreatitis enabled clinicians to make an antemortem diagnosis of severe acute pancreatitis. © 2009 Lippincott Williams & Wilkins

Management of Severe Acute Pancreatitis

classification system involved a melding of clinical symptomatology and pathologic findings derived from a painstaking collective review of prior cases in the literature, with a few cases added from the Boston area. Even though Fitz noted that the hemorrhagic, suppurative, and gangrenous forms of acute pancreatitis were often combined with disseminated fat necrosis, he was seemingly unaware that fat necrosis was a different manifestation of the same disease process. During this period of awakening interest in the pancreas, surgeons were not idle. According to Hollender,9 Wandeleben had incised and drained a pancreatic abscess as early as 1845, although the result of this effort is unknown. In 1882, Rosenbach marsupialized an abscess of the pancreas. Unfortunately, the patient died, reportedly in “schock” 6 hours after surgery.10 Hirschberg performed an exploratory laparotomy in 1887 in an obese male who had suffered a “collapse” 4 days after the onset of hemorrhagic pancreatitis, but within 5 hours after exploration, his patient also expired.11 These initially unsuccessful forays into the surgical management of necrotizing pancreatitis did not deter surgical interest. In 1886, Nikolas Senn (Fig. 2) published extensive animal experiments describing surgery of the gland.12 Furthermore, he offered his own clinical classification of acute pancreatitis, embodying gangrene, abscess, and hemorrhage of the gland, 3 years prior to the classification proposed by Fitz. Senn wrote that “One of the terminations of acute inflammation of the pancreas is gangrene. It would seem plausible that timely removal of the necrosed organ by surgical interference would add to the chances of recovery. Consequently, we shall add gangrene as one of the diseases of the pancreas which should be treated by operative measures” (p 175). Of note,

FIGURE 2. Nicholas Senn (1844 –1908). A childhood immigrant from Switzerland, Senn graduated from Chicago Medical School in 1868. After working at Cook County Hospital for several years, he journeyed to Munich for postgraduate studies under Professor Johann von Nussbaum. In 1878, he joined the faculty at Rush Medical College as Professor of Surgery. While at Rush, he carried out numerous animal experiments involving surgery of the pancreas, thereby providing the investigative basis for future generations of surgical scientists. He left his position in 1898 to become Chief Surgeon of the Army Medical Corps in Cuba during the Spanish-American War. www.annalsofsurgery.com | 7

Bradley and Dexter

Annals of Surgery • Volume 251, Number 1, January 2010

however, although Senn remained a staunch advocate for surgical intervention in acute pancreatitis for his entire career, there does not seem to be any record of whether he actually performed such operations. Nevertheless, his influence on the so-called “Chicago School of Surgery” was profound and long-lasting. Senn’s application of the scientific method to a clinical problem ignited the interest of many other surgeons to employ surgical techniques in human acute pancreatitis. Many students of surgical history are aware that 3 years later in 1889, Fitz was adamant that surgery would not benefit patients with acute pancreatitis, and that fatalities could not be prevented by operative “meddling.” Less well known, however, is the observation that Fitz appeared to modify his antisurgical stance in a subsequent publication in 1903, in which he allowed that some surgical successes had been reported in these desperately ill patients.13 The initial operative success in the management of necrotizing pancreatitis was reported by Werner Koerte (Fig. 3) in 1894. He described the successful drainage of a large pancreatic abscess via a left flank incision in a 48-year-old obese woman 1 month after the onset of severe pancreatitis.14 Iodoform gauze-wrapped drains were placed into the retroperitoneum, and postoperative management consisted of repeated changes of the gauze. After wound discharge of a considerable amount of necrotic pancreas and fatty tissue, and the formation of a pancreatic fistula, the patient was discharged completely healed 5 months later. Unfortunately, 2 other cases of postnecrotic abscess in his series that were treated similarly expired

due to recurrent infection. Koerte advocated delayed exploration of pancreatic infections: “In the acute stage, surgical treatment is not recommended, wherein patients have the propensity for cardiovascular collapse. If pancreatic apoplexy occurs, surgical treatment cannot help. Later, when we can prove that a purulent collection is arising from the gland, surgery is indicated” (p 739). These thoughts antedate the contemporary approach to surgery for necrotizing pancreatitis by more than 90 years. One year later, in 1895, Thayer from the Johns Hopkins Hospital reported a similar case of necrotizing pancreatitis with secondary infection that had been operated upon by JMT Finney, using debridement and closed drainage. The patient recovered and was discharged 4 months after the initial exploration.15 Six years later, from the same institution, Opie generalized from a single autopsy case that hemorrhagic pancreatitis resulted from unremitting obstruction of the pancreatic and bile ducts by biliary calculi.16 Given these initial successes, surgeons flocked to the banner of operative intervention in acute pancreatitis.17–19 At the turn of the twentieth century, such surgical stalwarts as Mayo Robson and Mickulicz also reported successful surgical interventions in necrotizing pancreatitis. In the second of 3 scholarly Hunterian Lectures given before the Royal College of Surgeons of England in 1904, Mayo Robson (Fig. 4) described 2 survivors in 4 early operations for necrotizing pancreatitis, and 5 survivors in 6 operations for pancreatic abscess.20 He noted that in contrast to previous beliefs “… inflammatory afflictions of the pancreas are very much more com-

FIGURE 3. Werner Koerte (1853–1937). Born in Berlin, the son of a famous physician, he received his MD from the University of Strasburg. In 1889, he was appointed chief surgeon of the Urban Hospital in Berlin, where he achieved international acclaim as a surgical technician and excellent teacher. Among his many teachings was the admonition that it was more important to know when not to perform surgery. He became President of both the Berlin and German Surgical Societies, and was honored by the German government with the issue of a stamp bearing his likeness.

FIGURE 4. Arthur W. Mayo Robson (1853–1933). Born in Filey in the United Kingdom, the son of a chemist, he trained in medicine at Leeds, at graduation becoming lecturer in Anatomy at the University. In 1890, he was appointed Professor of Surgery at the Leeds Infirmary, where he achieved recognition as a pioneer in surgery of the pancreas and bile ducts. He also undertook removal of knee cartilages before orthopedics became a specialty. Mayo Robson was widely known for his speed and skill as a surgeon, and later moved to London to further his career, where he was knighted in 1908. In the First World War, he served in France with a field ambulance, followed by distinguished service in other war theaters. His many honors included KBE, CB, CVO, MRCS Eng, FRCS Eng, and DSc (Hon) Leeds.

8 | www.annalsofsurgery.com

© 2009 Lippincott Williams & Wilkins

Annals of Surgery • Volume 251, Number 1, January 2010

mon than is generally supposed” (p 846). Johann von Mikulicz, Professor of Surgery in Breslau, and also a former student of Billroth, advocated early exploration in necrotizing pancreatitis, but favored pancreatic marsupialization by “pancreatostomy,” rather than by gauze drainage.21 In rapid succession, other surgeons reported successful procedures in patients with necrotizing pancreatitis and its complications. Woolsey, a Professor of Surgery at Columbia, described 3 cases of successful early exploration in necrotizing pancreatitis employing debridement and gauze drainage.22 Similarly successful anecdotal experiences with this approach were reported by Bunge23 and Villar.24 The favored technique consisted of surgical intervention early in the course of acute hemorrhagic/necrotizing pancreatitis, incisions into the pancreatic “capsule,” and gauze drainage of the peripancreatic region. In contrast, delayed intervention for pancreatic abscesses was preferred, employing marsupialization and gauze packing, with or without a counter-incision in the flank. As a result of these and other successful reports, surgical intervention in necrotizing pancreatitis became established policy for the next quarter century, even though surgeons continued to dispute the optimal timing for surgical intervention, as well as the specific surgical techniques to be employed. In an undeniable display of surgical zeal, Hoffman in 1911 attempted to perform total pancreatectomy in a patient with hemorrhagic pancreatitis, but was forced to withdraw because of heavy bleeding.25 This romance with surgical intervention in clinically diagnosable severe acute pancreatitis continued well into the 1920s, perhaps reaching its zenith with Moynihan’s statement in 1925 expressing doubt that survival from necrotizing pancreatitis could ever be expected without early surgical intervention: “… recovery from this disease, apart from operation, is so rare that no case should be left untreated.”1 Lord Moynihan, Professor of Surgery at Leeds in England, had adopted Professor Koerte’s approach of wound closure over gauze wrapped drains, but because of the large number of patients in his clinic with associated biliary disease, he also added cholecystostomy to the technique. In 1927, Professor Viktor Schmieden of the University of Frankfurt reported a series of 1510 patients with necrotizing pancreatitis collected from 124 international clinics over the preceding 8-year period.26 Of the 1278 cases who had undergone surgery, the overall mortality was 51%. The authors registered surprise that this mortality rate was not appreciably different from the 60% mortality rate experienced at the turn of the century. Mortality rates were noted to be highest in hemorrhagic (60%) and necrotizing pancreatitis (65%), and lowest in acute edematous pancreatitis (24%). They advocated early surgical intervention within days of onset, longitudinal incisions into the gland for “decompression,” and tampon drainage for 6 to 8 days. “The (abdominal) incision should be kept wide open for many weeks because of the persistent pancreatic enzyme secretion, and also because of the extremely long time required for the discharge of the gland sequestra” (p 751). Of passing interest in this report is Schmieden’s proposal that the mechanism for necrotizing pancreatitis associated with gallstones might be due to “the passage of concretions with their temporary fixation of the ampulla of Vater …” Schmieden’s theory of pathophysiology of biliary pancreatitis preceded by almost 50 years the report of Acosta and his associates confirming transient obstruction of the pancreatic duct at the ampulla as the responsible mechanism.27 Despite the prevailing opinion that only surgical intervention in acute pancreatitis was capable of preventing death, practical difficulties in the performance of surgical therapy and postoperative care were widely appreciated. As Okinczyc wrote in 1933, “Go right to the target, expose the gland, drain, and hope!”28 © 2009 Lippincott Williams & Wilkins

Management of Severe Acute Pancreatitis

Continuing difficulties in the clinical diagnosis of acute pancreatitis no doubt accounted for a significant portion of this initial surgical mortality, as only the most florid and severe examples of pancreatitis were easily diagnosed. In essence, the measure of severity of an episode of acute pancreatitis during these times was its very ability to be diagnosed. All that changed in 1929, when Elman et al proposed that a serum test for amylase could reliably diagnose acute pancreatitis.29 At the time of this epochal paper, Elman was a surgical resident working under Professor Evarts Graham at the Barnes Hospital in St. Louis. In one definitive stroke, Elman et al enabled acute pancreatitis to be differentiated from other abdominal conditions without the necessity for surgical intervention. Serum amylase became the simple diagnostic test for acute pancreatitis for which Koerte and other clinicians had longed several generations earlier. As information regarding serum amylase began to accumulate, however, it soon became apparent that most of patients with acute pancreatitis had a much milder form of the disease. As a result, physicians began to question the dominant opinion that surgery in acute pancreatitis was necessary for survival. In 1929, Peter Walzel, a Viennese surgeon, was the first to point out that conservative treatment of the more common form of the disease was demonstrably superior to mortality rates from surgical intervention, which at that time ranged between 50% to 78%.30 Moreover, he stated that acute edematous pancreatitis and necrotizing pancreatitis were 2 different diseases; thereby contradicting the widely held view that edematous pancreatitis was a necessary precursor to necrotizing pancreatitis. Nonoperative management of acute pancreatitis was rapidly championed by Mikelson, Demel, and Nordmann in Europe,31–33 and by Trasoff, Pratt, Fallis, and Lewison in North America.34 –37 From the early 1930s until the late 1950s, surgical intervention in acute pancreatitis was rarely practiced. However, conservative therapy proved not to be the pancreatitis panacea than had been initially hoped. In 1948, Paxton and Payne from the Los Angeles County General Hospital reported a consecutive series of 307 patients with acute pancreatitis and an overall mortality of 33%. Upon subanalysis, surgical mortality was noted to be 45%, while nonoperative cases experienced a mortality of 28%.38 Although the comparatively lower conservative mortality was considered by the authors to be support for the nonsurgical approach, it also had the undesired effect of demonstrating that conservatively treated patients continued to expire at a high rate. In a 1959 review of 100 consecutive cases of acute pancreatitis from the University of Leeds, Pollock re-emphasized that patients with severe acute pancreatitis continued to expire despite the best efforts of conservative management.39 In this series, a number of patients with necrotizing pancreatitis underwent transduodenal pancreatography, with the frequent finding of a disruption of the main pancreatic duct. These initial observations of the necrotizing process extending to include the pancreatic duct were confirmed 30 years later by our group using endoscopic retrograde cholangiopancreatography in patients following recovery from necrotizing pancreatitis.40 In a subsequent report from Naples, Uomo and his associates documented the frequency of duct disruptions in more than 30% of patients with necrotizing pancreatitis.41 Even today, the clinical significance of necrosis-induced duct disruption in acute pancreatitis is not known with certainty. A few surgeons have called for resection of the involved section of the gland whenever duct disruption has been demonstrated,42,43 but since the majority of patients with proved duct disruption recover without consequence,41 a conservative approach seems justified. Concerns for the high mortality rate in severe acute pancreatitis treated conservatively were also voiced by Foster and Ziffren www.annalsofsurgery.com | 9

Bradley and Dexter

in 1962. They noted mortality rates for conservative therapy to exceed 80% in some instances.44 Subsequently, other workers reported mortality rates with persistent medical treatment of severe acute pancreatitis in the range of 50% to 90%.45,46 Fuel for the selective surgical fires, indeed! Although from today’s evidencebased perspective, these historical mortality rates for conservatively treated severe acute pancreatitis would be considered uncontrolled and questionable, they were considered authoritative at the time. As a result, such articles served to establish a high-level risk for conservative treatment of the clinically severe form of acute pancreatitis. Accordingly, surgeons began to wonder once again whether a more selective indication for surgical intervention in severe acute pancreatitis might further improve mortality. For many years, surgeons had considered it axiomatic that the presence of necrotic tissue was sufficient to mandate its removal. Sporadic attempts at extirpation of the gland for necrotizing pancreatitis had been undertaken following the initial unsuccessful attempt by Hoffman in 1911. In 1945, Dargent of Lyon was able to complete an emergency total pancreatectomy for necrotizing pancreatitis, but the patient did not survive.47 Chau et al reported the first successful distal resection of the pancreas for necrotizing pancreatitis in 1959.48 Four years later, George Watts, a consultant surgeon at the Birmingham General Hospital in England, performed the first successful total pancreatectomy for a patient in shock with “fulminant” acute pancreatitis.49 While covering less than 1 page in The Lancet, this report generated widespread interest in pancreatic resection for patients with severe acute pancreatitis. Using “failure to respond to supportive therapy” as the primary indication for surgical intervention in severe acute pancreatitis, a number of continental surgeons rapidly developed experience with extensive ablative surgery in this high-risk group.50 – 63 Surgical intervention was usually initiated within 48 hours of onset of the acute episode, and consisted of left resections and occasional total pancreatectomies. Mortality rates ranged from 28% for distal resections59 to 60% for duodenopancreatectomies.60 However, multiple reoperations were necessary in more than three quarters of patients with these ablative approaches, primarily for postoperative infection and hemorrhage. The largest experience with ablative surgery during this period was that of Professor Louis Hollender et al of Strasburg (Fig. 5) Beginning in 1967, Hollender et al employed capsular incisions in the gland and early resection of hemorrhagic segments.64 Later, these resections were combined with necrosectomies and lavage. Overall, they were able to achieve a mortality rate of 26% in 82 patients.64 – 67 Hollender’s statement that, “Conservative treatment may delay a fatal outcome, but very seldom prevents death.”67 was representative of the opinions of the ablative surgeons during this time. The comparative success of pancreatic resection for the necrohemorrhagic form of acute pancreatitis exhibited during the 1960s and the 1970s was in sharp contrast with the 50% to 70% mortality rates for resection experienced during the first quarter of the century. In many ways, however, this improvement in the mortality of resection was due to major breakthroughs in the understanding of surgical metabolism and fluid requirements that had been generated by the combined works of Francis Moore,68 Tom Shires,69 and Shoemaker and Walker.70 As a surgical house officer in the early 1960s, the senior author can recall being advised to give vasoconstrictors to patients in shock with severe acute pancreatitis! At that time, pharmacologic support of the blood pressure was considered state-of-the-art supportive therapy, despite the emergence of an extensive parallel experience in renal failure. In a series of remarkable insights into human fluid homeostasis, this trio of surgical investigators had not only dramatically improved the post10 | www.annalsofsurgery.com

Annals of Surgery • Volume 251, Number 1, January 2010

FIGURE 5. Louis Francois Hollender (1922–Contemporary). Born in Strasbourg, he completed his MD in Freiburg and Paris, and surgical training under Professor Weiss at the University of Strasbourg. He rose to become Professor of Surgery at the Louis Pasteur University in 1970, and spent several sabbatical years in surgical departments in America and Europe. He achieved the unofficial status of “Surgical Ambassador to the World” through his many visiting professorships on 5 continents. Professor Hollender is widely published, with many texts and articles covering the spectrum of general surgery. He has been honored as a Commander of the Legion of Honor, and also with the French Cross. operative care of contemporary surgical patients, but also had provided the framework for modern conservative management. Even though ablative surgery was widely practiced in European surgical centers, extended resections never really caught on elsewhere, the exception being the report of Norton and Eiseman from America detailing near total pancreatectomies in 4 patients with necrotizing pancreatitis.71 Although overall mortality rates for early pancreatic resection in the necrohemorrhagic form of acute pancreatitis ranged from 30% to 50% for left resections, to 80% for total pancreatectomies, surgeons at that time believed that these mortality results were superior to those historical mortality rates from supportive therapy in similarly severe cases. During the time that pancreatic resection held sway on the Continent, more conservative operative techniques were being practiced in the United States. In 1963, Altemeier and Alexander from the University of Cincinnati reported the largest series up to that time of proved pancreatic abscesses from a single institution.72 Eighteen of the 21 patients survived that were treated with incision and passive drainage with Penrose drains placed through flank incisions, whereas all 11 patients treated conservatively died. Debridement and closed passive drainage came to be known as the “conventional” approach to pancreatic infections. © 2009 Lippincott Williams & Wilkins

Annals of Surgery • Volume 251, Number 1, January 2010

Less radical surgery for acute pancreatitis and its complications became more effective with a 1968 report from Louisville of an improved form of sump drainage, developed during animal experiments by Waterman et al.73 All 10 patients undergoing early operation in the necrohemorrhagic stage survived with the use of sump drainage for extended postoperative periods. Cholecystostomy, gastrostomy, and feeding jejunostomy (“triple ostomy”) were added to sump drainage in 1970 by Lawson et al from the Massachusetts General Hospital.74 Of 15 patients with necrohemorrhagic pancreatitis treated by early surgical intervention using this method, 11 survived. The 4 deaths were due to infection in the later stages of the disease, causing the authors to state that, “Late sepsis, however, remains a difficult clinical problem” (p 614). In a follow-up study 4 years later of 38 patients from the same institution using the triple ostomy approach, Warshaw et al reported a mortality rate of 34%, employing early surgical intervention within the first 48 hours of the attack.75 White and Heimbach were equally reluctant to employ major pancreatic resections for necrotizing pancreatitis. In 1976, they reported an observed mortality of 20% in 30 patients with refractory hemorrhagic pancreatitis treated with sump drainage, intravenous hyperalimentation, and the triple ostomy approach.76 Half of their patients required a secondary procedure, most commonly for postoperative infection. The late mortality problem with sepsis experienced by the MGH group was improved in this latter series by reoperation “on demand,” and removal of infected necrotic tissues. As the European experience with resections of the pancreas for acute pancreatitis increased in number, a great deal of pancreatic tissue from patients with the clinically severe forms of the disease became available for histologic examination. From this wealth of anatomic information, 2 principal observations were made. First, necrotizing pancreatitis was present in the overwhelming majority of SAP cases, thereby suggesting that the necrotizing process was related to the severity of clinical manifestations. But extensive resections were frequently found to include areas of viable pancreatic tissue within the operative specimen, demonstrating that necrosis could be superficial or patchy in nature, and that estimations of the extent of necrosis at surgery were unreliable. These observations suggested that many of the ablative surgical procedures had been more extensive than necessary.77,78 Given the observed high mortality rates for extensive pancreatic resections in the necrohemorrhagic phase of the disease, as well as the inability to detect viable pancreatic tissue within the areas proposed for resection, many surgeons sought alternatives to early intervention and extensive resections in necrotizing pancreatitis.

THE PRESENT The modern era for surgery of SAP can be said to have begun with attempts to predict the severity of an attack of acute pancreatitis. It was hoped that establishing severity could identify those patients with SAP who might benefit from surgery, or indicate those requiring more extensive supportive therapy. The initial attempt to predict SAP was provided by Ranson et al in 1974 who described a combination of clinical observations and easily obtainable laboratory tests that were highly correlated with a severe clinical course.79 Shortly thereafter, a number of Continental surgeons proposed that the presence and extent of pancreatic necrosis was the principal determinant of the severity of acute pancreatitis. Today, we know that the presence of organ failure is more highly correlated with severity than is the presence of necrosis.80 Establishing these clinical and pathologic parameters of severity, however, generated a series of important clinical questions: How can the necrotizing form of acute pancreatitis be diagnosed © 2009 Lippincott Williams & Wilkins

Management of Severe Acute Pancreatitis

clinically? What are the indications for surgical intervention? What is the optimal surgical technique, and when should it be done?

Diagnosis The answer to the first question was provided in 1984 by Leena Kivisaari, then a radiology resident at Helsinki University Hospital, who observed that patients with necrotizing pancreatitis failed to exhibit opacification of the necrotic portion of the pancreas during intravenous contrast computed tomography.81 For some time, it had been known that organ opacification was a normal consequence of the leak of the small contrast molecule from capillaries into the interstitial space. Failure to opacify an organ was, therefore, a priori evidence that the contrast material was not present in the capillaries. In this instance, the absence of contrast material represented capillary thrombosis. This important observation was subsequently confirmed by experimental and clinical studies from Europe82 and America,83,84 establishing the presence of pancreatic necrosis in more than 90% of cases failing to exhibit opacification. Concerns from animal experiments that contrast-enhanced computed tomography might lead to an exacerbation of severity of pancreatitis, or to an increased incidence of renal insufficiency,85 proved to be unfounded in humans.86

Indications for Surgical Intervention Sterile Necrosis Despite the lack of objective evidence, a considerable body of opinion had arisen which held that removal of necrotic pancreatic tissue was mandatory. The rationale for this opinion was based on 3 apparently reasonable assumptions: (1) removal of dead pancreas would prevent release of “toxic substances” causing the development of organ failure, (2) secondary infection of the necrotic tissues would not occur, and (3) surgical mortality rates would improve. Hans Beger (Fig. 6) et al from Ulm were the principal advocates of programmatic removal of pancreatic necrosis. These interventions were carried out early in the course of SAP, and necessarily involved patients with sterile pancreatic necrosis. Rather than resection, they began to employ debridement of necrosis for patients in the necrohemorrhagic phase of acute pancreatitis. Drainage was affected by prolonged postoperative lesser sac catheter lavage.87 In their initial report, the correlation between necrosis and clinical severity was emphasized, as was the increased mortality risk for the development of infected pancreatic necrosis. While the successes of retroperitoneal and lesser sac lavage, as previously demonstrated by Hollender et al65 and Pederzoli and coworkers,88 were well known, the substitution of debridement of pancreatic necrosis in place of extensive resection represented a major step forward. However, indications for surgery in Ulm included all patients with necrotizing pancreatitis documented by contrast-enhanced computed tomography who were “not responding within 48 hours of intensive medical therapy,” regardless of the presence or absence of infection. In keeping with the European tradition, surgical intervention was urged early within the course of necrotizing pancreatitis, averaging 5.1 days from the onset of symptoms. As might be expected from the fact that secondary infection of pancreatic necrosis is primarily a delayed phenomenon most often developing several weeks after onset, adherence to this indication for surgery resulted in a predominance of patients with sterile pancreatic necrosis undergoing debridement. In a 1988 report,89 the Ulm group demonstrated an overall mortality rate of 8.1% using these techniques in patients with sterile pancreatic necrosis. In addition, the presence of active trypsin and phospholipase A were noted in the effluent from the lavage drains for as long as 12 to 14 days after operative intervention. This finding was interpreted to mean that the necrotizing process was a continuing one in the postoperative period, thereby providing the www.annalsofsurgery.com | 11

Bradley and Dexter

FIGURE 6. Hans Gunter Beger (1936 –Contemporary). Born in Dresden, he attended medical school in Bonn, and surgical training under Professor Emil Buecherl at the Charlottenberg Clinic of the Free University of Berlin. He worked in the surgical clinics of Dr. McDermott, Dr. Eiseman, and Dr. Starzl in America, and rose to the rank of Professor in Berlin in 1978. Called to the chairmanship at the University of Ulm in 1982, he proceeded to establish a world-famous center for pancreatic surgery. Following the traditions of Billroth and Buecherl, translational research was brought to the bedside, and a new school of academic surgeons was founded. His honors include Fellowship in the American College of Surgeons, Presidencies of the European Pancreatic Club and the International Association of Pancreatology, and multiple scientific awards from 4 continents. basis for prolonged postoperative drainage. By 1995, the group had modified their original approach from programmatic debridement for all patients with documented pancreatic necrosis, to one agreeing with medical management of some patients with sterile pancreatic necrosis, providing they improved with initial conservative therapy.90 In a 2005 summary of their collective experience, the mortality rate for the 107 operated patients consisting of both sterile and infected necrosis was 13.1%, while the conservatively treated patients exhibited a mortality rate of 6.2%.91 Moreover, they were able to reduce the operative mortality rate for infected necrosis to 21%. However, to achieve these commendable results with the technique of debridement and lesser sac lavage, the necessity for reoperations was required in more than one-quarter of their cases. Postoperative hemorrhage and recurrent sepsis were the principal indications for re-exploration. During this same period, debridement of sterile pancreatic necrosis was also advocated by Warshaw and his colleagues from Boston. They described an overall mortality rate of 25% in a mixed population of patients with both sterile and infected necrosis, using the conventional surgical approach of debridement combined with 12 | www.annalsofsurgery.com

Annals of Surgery • Volume 251, Number 1, January 2010

drainage of the lesser sac with a Penrose and sump drains.92 Their reoperation rate approached 20%, but the necessity for reoperation for recurrent or persistent infection in this series was subsequently reduced by the extensive use of guided transcutaneous drainage. In contrast to Beger’s studies, they did not find a significant difference in postoperative mortality rates between patients operated upon for sterile necrotizing pancreatitis and those with infected pancreatic necrosis. However, the patients in the MGH study were not directly comparable to those from Ulm, as the duration from onset to surgical intervention was 3 to 4 weeks in Boston as a result of local referral patterns, as opposed to 5 days in the German series. Not commented upon by the Boston authors, but calculable from the data presented, was the observation that almost 40% of the patients operated upon with initially sterile pancreatic necrosis became secondarily infected in the postoperative period. Those unfortunate patients with surgically induced infections experienced a significantly higher mortality rate (50%), than did patients remaining persistently sterile (3%). In a 1998 report from this group composed of 64 patients divided between sterile and infected pancreatic necrosis, they again found that the overall mortality rate of 6% was not significantly different between sterile and infected pancreatic necrosis.93 As we have noted, the rationale for surgical intervention in necrotizing pancreatitis had always been based upon suppositions that mortality rates would be improved by surgery, and that complications of necrotizing pancreatitis, such as secondary infections and organ failure, could be eliminated or reduced. However, no credible data had ever appeared demonstrating a procedure-related reduction in organ failure, or reductions in any other complication of sterile pancreatic necrosis.94 Although surgically-induced reductions in mortality and morbidity were widely claimed, none of the reported surgical experiences for sterile pancreatic necrosis had ever included the necessary control group, unoperated patients with sterile necrosis, against whom the surgical mortality could be compared. In 1986, we entered patients into a prospective, longitudinal study designed to provide the missing control group, by addressing the fate of patients with severe acute pancreatitis and sterile pancreatic necrosis who were not subjected to surgery. From a total group of 194 cases admitted for severe acute pancreatitis over a 3 year period, 11 patients with were identified with pancreatic necrosis involving 30% to 60% of the gland by CECT, most with accompanying organ failure. Fine needle aspiration cultures were persistently negative in each of these 11 patients. Moreover, each patient was successfully managed with persistent medical therapy.95 From these observations, we concluded that neither the presence of sterile pancreatic necrosis, nor organ failure, per se, was an indication for surgical intervention in these patients. As these observations challenged widely held surgical dogma, considerable discussion was engendered. Other supportive prospective studies followed rapidly, however, demonstrating mortality rates for nonoperative management of patients with sterile necrotizing pancreatitis that were equal or superior to surgical mortality rates.96 –101 These studies have provided the rationale for acceptance of the conservative approach to the management of patients with sterile pancreatic necrosis in practice today. Even though the shift away from surgical intervention and toward conservative management of sterile pancreatic necrosis has resulted in a lower mortality rate for most of these patients, a small number of patients with sterile necrosis remain who can benefit from operative intervention. Postnecrotic disruption of the main pancreatic duct, variably referred to as “persistent pancreatitis,”102 “refeeding pancreatitis,”40 or “disconnected duct syndrome,”103 can result in recurrent abdominal pain and hyperamylasemia following attempts at reinstitution of oral alimentation after clinical recovery © 2009 Lippincott Williams & Wilkins

Annals of Surgery • Volume 251, Number 1, January 2010

from necrotizing pancreatitis. Longer standing postnecrotic strictures may result in “upstream” chronic pancreatitis,103 enteric obstruction due to ischemia,104 or common duct obstruction,105 each of which responds favorably to surgical therapy. These complications of necrotizing pancreatitis are relatively uncommon, occurring in perhaps 2% to 3% of cases,105 but may become more evident as the experience with conservative management of sterile necrosis increases.

Infected Necrosis In contrast to the controversies concerned with whether or not surgical intervention was necessary in patients with sterile necrotizing pancreatitis, until very recently, there was almost universal agreement that surgical intervention was mandatory in patients with infected pancreatic necrosis. Whatever discussions existed revolved around other issues, such as the preferred surgical technique for infected pancreatic necrosis and the timing of surgical intervention. Differentiation of a systemic inflammatory response from secondary infection of pancreatic necrosis was markedly facilitated by the studies of Gerzof et al from the Brigham Hospital when they demonstrated the diagnostic precision and safety of CT-guided fine needle aspiration of the pancreatic area, with gram staining and culture of the aspirate.106 Even into the 1980s, however, operative mortality rates for patients with infected pancreatic necrosis using the conventional approach of debridement and closed Penrose drainage had often remained in the 30% to 60% range.107 Many of the deaths were caused by delays in postoperative re-exploration for recurrent infection, or by retroperitoneal hemorrhage, both accounting for necessary reoperation in 1 patient in 3 undergoing conventional drainage. In an effort to address these problems and improve these mortality statistics, in 1976 we embarked upon a different surgical direction. Employing a technique that consisted of delayed sequestrectomy and planned serial re-explorations, marked improvements in historical mortality rates were observed.108 Healing of the abdominal cavity was initially permitted to occur by secondary intention. At the time of our initial report in 1981, we were unaware of Koerte’s recommendation almost 100 years earlier for prolonged drainage by keeping the abdominal wound open. However, we were aware of the efforts of Bolooki et al in treating 20 patients with suture-created marsupialization, with a resultant mortality of 50%.109 Improvements to our original technique were added over the course of years, including; placement of a nonadherent gauze over the exposed stomach and colon to prevent inadvertent debridement of the intestinal tract and subsequent intestinal fistulization, routine placement of a feeding jejunostomy as first recommended by Kiekens et al,110 keeping the packs continually moist with a bioticide, and secondary closure of the abdominal wall.111 Secondary closure was accompanied by the placement of 2 large sump drains in the retroperitoneal area for continuous lavage. At the time of our last report, our mortality experience with what came to be known as “open packing,” was 14% in 102 patients with documented infected pancreatic necrosis.112 Today, however, the senior author would recommend reserving the open drainage technique for patients with large collections extending bilaterally across the retroperitoneum and into the retrocolic spaces, and in those situations in which bleeding from contiguous vessels is difficult to control with standard measures. Frey et al from the University of California in Sacramento were early proponents of the open packing technique for infected pancreatic necrosis, reporting a series of 83 patients, with an overall mortality of 17%.113 In the Mayo Clinic experience with infected necrosis summarized by Sarr et al, the mortality rate was also found to be 17%, with a minimal incidence of recurrent infection.114 In a series of patients undergoing open packing for infected pancreatic © 2009 Lippincott Williams & Wilkins

Management of Severe Acute Pancreatitis

necrosis collected from world surgical centers in 1998, the combined overall mortality rate was also noted to be 17%.112 Howard has pointed out the important essential differences between infected pancreatic necrosis and infected peripancreatic fat necrosis, with the latter experiencing a significantly lower mortality rate of 3%.115,116 Other surgeons continued to advocate conventional techniques in patients with infected pancreatic necrosis, using either closed surgical drainage, or lesser sac lavage. Fernandez and his group from the MGH reported a mortality rate of 8% in 36 selected patients with infected necrosis using Penrose drainage, in combination with guided percutaneous drainage of recurrent abscesses and a low threshold for surgical re-exploration.93 The collected experience of Beger and his associates in Germany with debridement and lesser sac lavage for infected necrosis had produced a mortality rate of 21%, which included a reoperation rate of 40%.91 However, comparisons between the 3 alternative surgical approaches to infected necrosis were made difficult by numerous variations between the patient populations, the inclusion of lesser risk patients with sterile necrosis in some series, differences in timing of surgical intervention, and variations in operative technique. In the absence of evidence-based studies comparing the 3 techniques, superiority could not be assigned to any one approach. Since the results from each surgical approach in expert hands seemed to represent a considerable improvement over previous mortality rates using the traditional approach, choosing the technique with which one felt most comfortable seemed to be the most reasonable conclusion.

Nonsurgical Management of Infected Necrosis Anecdotal reports of patients with documented infected pancreatic necrosis surviving without surgical intervention have recently appeared.117,118 The majority of these reports have been composed of patients who had either refused surgery, or were considered to be excessive surgical risks. Prolonged antibiotic administration, assisted in some cases by guided percutaneous drainage and markedly delayed surgical “rescue,” led to successful conclusions. These reports have called into question the surgical dictum of mandatory surgical drainage for all patients with infected pancreatic necrosis. However, the number of patients who might be successfully managed by this nonoperative approach is not known. Given the excessive mortality with nonsurgical therapy for infected pancreatic necrosis demonstrated in the past, randomization between operative debridement and persistent antibiotics is unlikely to ever be done. Furthermore, since fatalities with the nonoperative approach to infected necrosis are probably underreported, nonoperative management may best be reserved for patients refusing surgery, or for those exhibiting severe comorbidities.

Timing of Surgical Intervention Throughout the 1980s and 1990s, the timing of surgical intervention in acute necrotizing pancreatitis re-emerged as an issue, between those favoring “early” intervention (within the first week of illness),50 – 67,89 and those advocating “delayed” procedures (2– 4 weeks after onset).111,113–115 In one sense, this controversy was an artificial one, since those favoring early intervention were primarily surgeons advocating debridement of sterile pancreatic necrosis in the early necrohemorrhagic phase, while proponents of the delayed approach were surgeons restricting intervention to patients until the later appearance of infected necrosis. With the indications for surgery in necrotizing pancreatitis shifting away from intervention in patients with sterile pancreatic necrosis, the advantages of delayed intervention soon became apparent. Those advantages included clearer demarcation between necrotic and viable tissues over time, technical ease of debridement at a later time, clearer CT depictions of the extent of the infectious process, and better control of the metabolic and organ consequences of necrotizing pancreatitis. The www.annalsofsurgery.com | 13

Annals of Surgery • Volume 251, Number 1, January 2010

Bradley and Dexter

TABLE 1. Milestones in the Surgical History of Necrotizing Pancreatitis Senior Author

Year

Contribution

Tulp2

1652

Klebs5

1870

Chiari7

1883

Senn12

1886

Fitz13

1889

Koerte14

1894

MayoRobson20

1904

Elman et al29

1929

Nordmann33

1938

Watts49

1963

Altemeier and Alexander72 Lawson et al74

1963

Hollender67

1979

Bradley and Davidson108 Kivisaari et al81

1983

Initial post-mortem description of necrotizing pancreatitis. Relationship between hemorrhagic, necrotizing, and suppurative pancreatitis. Proposed “autodigestion” as the mechanism for pancreatic necrosis. Applied scientific method of animal experimentation to surgery of the gland. Established clinicopathologic correlation for necrotizing pancreatitis. First successful operation for infected pancreatic necrosis. Advocated early exploration and drainage in necro-hemorrhagic stage. Established serum amylase in diagnosis of acute pancreatitis. Recommended conservative therapy for all forms of acute pancreatitis. First successful total pancreatectomy for “fulminant” acute pancreatitis. Advocated debridement and closed drainage for pancreatic “abscess.” Introduced “triple ostomy” and drainage for necro-hemorrhagic phase. Largest experience with early resection in the necro-hemorrhagic phase. Delayed sequestrectomy and planned re-explorations for infected necrosis. Diagnosis and extent of necrotizing pancreatitis by contrast-enhanced CT. Debridement and lesser sac lavage for all forms of necrotizing pancreatitis. Diagnosis of infected pancreatic necrosis by guided fine needle aspiration. Non-operative management for most forms of sterile necrotizing pancreatitis. Clinically based definitions for necrotizing pancreatitis, and proposed treatment algorithms. Minimal access debridement and drainage of infected pancreatic necrosis.

Beger et al

89

1970

1984 1985

Gerzof et al106

1987

Bradley and Allen95

1991

Atlanta Symposium121

1993

Carter et al126

2000

issue of the timing of surgical intervention was finally resolved with the appearance of a randomized, controlled study by Mier et al in 1997, demonstrating a mortality advantage to the delayed approach.119 It was necessary for them to stop the trial prior to its scheduled conclusion due to excessive mortality in the early intervention patients.

Terminology By the 1980s, the terminology for acute pancreatitis and its complications was in severe disarray. The authors of many articles dealing with acute pancreatitis and its complications had seemingly adopted the mandate of Humpty Dumpty from Alice in Wonderland, “When I use a word it means just what I choose it to mean – neither more nor less.”120 The term “pancreatic abscess” was a case in 14 | www.annalsofsurgery.com

point. In a review of more than 1100 patients labeled as “pancreatic abscess” collected from 45 articles published between 1966 and 1987, only 11 authors had actually provided a definition of “pancreatic abscess,” and no 2 of the definitions were the same!107 Similarly, the term “pancreatic phlegmon” came to be used to describe tissue that, depending on the author, could be either sterile or infected, or composed of either edematous or necrotic tissues. A pancreatic Tower of Babel had developed, and communications between workers in acute pancreatitis and progress in the management of its complications were being held hostage by imprecise terminology. The need for standardization was apparent. In 1992, a group of 40 internationally renowned workers in acute pancreatitis was assembled in Atlanta for the expressed purposes of devising a clinical classification system for acute pancreatitis and its complications, and suggesting treatment algorithms. The symposium attendants came from 15 countries and represented 6 medical disciplines. At the completion of the meeting that came to be known as the Atlanta Symposium, the participants had unanimously agreed upon a series of clinical definitions which are in use today, 17 years after their proposal.121 However, increasing knowledge and experience over the intervening years have made some aspects of the symposium less useful, and a contemporary update can be anticipated. In Table 1 are listed a number of significant historical contributions that have led to the formulation of our current approach to the management of necrotizing pancreatitis. The reader will note that the majority of these formative events are the result of translational clinical research conducted by surgeons.

PROLOGUE f It is difficult to predict, especially the future! —Y. Berra, Former Manager, New York Yankees Although admittedly imprecise, predictions for the future of surgical intervention in necrotizing pancreatitis may be becoming less difficult. Rapid advances in the technology and application of minimally invasive surgery have provided significant patient advantages over traditional operative methods, albeit not without attendant problems and increased costs. Even though image-guided catheterization procedures have proved to be a valuable adjunct to therapy for some pancreatic infections,122,123 guided transcutaneous drainage has left much to be desired when used as primary therapy for drainage of infected pancreatic necrosis. Incomplete resolution of infected collections has been common, being attributed to the inability of small-size drainage catheters to adequately remove thick, infected, necrotic material.124 In 1999, Baron, a gastroenterologist from the University of Alabama, and his associates reported removing pancreatic necrosis with an endoscope via a transgastric access to the collection.125 Endoscopic drainage and removal of infected pancreatic necrosis have now been reported by several groups. Carter et al from Glasgow have used video-assisted sinus tract endoscopy to access and remove the necrotic material.126 Other techniques for endoscopic necrosectomy include direct retroperitoneal access,127 and transperitoneal laparoscopic necrosectomy.128 Collected success rates in approximately 200 patients undergoing endoscopic necrosectomy have approached 60% to 70% with low mortality, even though multiple “sessions” are required, and surgical “rescue” is currently necessary in about one-third of cases.129 Currently, endoscopic necrosectomy techniques cannot be considered to satisfy evidence-based criteria for complete acceptance. However, the future for minimally invasive techniques in the management of patients with infected pancreatic necrosis appears promising, © 2009 Lippincott Williams & Wilkins

Annals of Surgery • Volume 251, Number 1, January 2010

particularly as experience increases, and the inevitable technical modifications occur.

RECOMMENDED READING For those individuals seeking extensive historical information on a broad range of pancreatic diseases, the scholarly work, History of the Pancreas: Mysteries of a Hidden Organ, by John Howard and Walter Hess, Kluwer, New York, 2002 is highly recommended. REFERENCES 1. Moynihan B. Acute pancreatitis. Ann Surg. 1925;81:132–142. 2. Tulp N. Observationum Medicarum, Editio Nova et Aucta. 2nd ed. Amsterdam, The Netherland; 1652:345. Book 4. Cited by Howard JM, Hess W History of the Pancreas, New York: Kluwer; 2002. 3. Claessen H. Krankheiten der Bauchspeicheldruse. Cologne, Germany: Dumont-Shaumberg; 1842. 4. Rokitansky K. Lehrbuch der Pathologischen Anatomie. Vol. 3.Vienna, Austria: W Braumuller; 1842:313. 5. Klebs A. Handbuch der Pathologischen Anatomie. Vol. 3. Berlin, Germany: Hirschwald; 1870:529 –553. 6. Friedreich N. Diseases of the pancreas. In: von Ziemssen HW, ed. Cyclopedia of the Practice of Medicine. New York, NY: W Wood & Co; 1878:551– 630. 7. Chiari H. Ueber die selbstverdauung des menschlichen pankreas. Z Heilk Prague. 1896;17:1–35. 8. Fitz RH. Acute pancreatitis: a consideration of pancreatic hemorrhage, hemorrhagic, suppurative, and gangrenous pancreatitis. Boston Med J. 1889;70:181–235. 9. Wandeleben. Cited by Hollender LF, Pieper HJ. Chirurgische Therapie der Akuten Pankcretitis. In: Pankreaschirurgie, Heidelberg, Germany: SpringerVerlag; 1988. 10. Rosenbach. Centralblatt fur Chirurgie. 1882:66. Cited by Senn (12). 11. Hirschberg. Berlin Klin Wochensch. 1887;34:287.Cited by Senn (12). 12. Senn N. Surgery of the pancreas as based upon experiments and clinical researches. Trans Am Surg Assoc. 1886;4:99 –232. 13. Fitz RH. The Symptomology and Diagnosis of Pancreatitis. Washington, DC: Medical Congress; 1903. 14. Koerte W. Zur chirurgischen behandlung der pancreas-eiterung und pancreas-nekrose. Arch Klin Chir. 1894;48:721–752. 15. Thayer WS. Acute pancreatitis–parapancreatic abscess– disseminated fat necrosis. Am J Med Sci. 1895;110:396 – 405. 16. Opie EL. The etiology of acute hemorrhagic pancreatitis. Bull Johns Hopkins Hosp. 1901;12:82– 88. 17. Mayo Robson AW. La chirurgie du pancreas. In: 13me Congress International de Medecine; 1900:140; Paris, France. 18. Ceccherelli A. La chirurgie du pancreas. In: 13me Congress International de Medecine; 1900:159; Paris, France. 19. Boeckel J. La chirurgie du pancreas. In: 13me Congress International de Medecine; 1900:207; Paris, France. 20. Mayo Robson AW. Hunterian lecture II–inflammatory affections of the pancreas. Lancet. 1904;846 – 854. 21. von Mickulicz-Radecki J. Surgery of the pancreas: with special consideration of trauma and inflammatory processes. Ann Surg. 1903;38:1–29. 22. Woolsey G. The diagnosis and treatment of acute pancreatitis. Med Surg Rep Presby Hosp. 1902:47– 61. 23. Bunge 1903: Cited by Hollender LF, Peiper HJ. See reference 9. 24. Villar F. La chirurgie du pancreas. In: 18me Congress Francais de la Chirurgie; 1905; Paris, France. 25. Hoffman 1911: Cited by Hollender LF, Peiper HJ. Op Cit. (13). See reference 9 and 14. 26. Schmieden V, Sebening W. Surgery of the pancreas: with especial consideration of acute pancreatic necrosis. Surg Gynecol Obstet. 1928;46:735–751. 27. Acosta JM, Ledesma CL. Gallstone migration as a cause for acute pancreatitis. N Eng J Med. 1974;190:484 – 487. 28. Ockinczyc P. Technique Operatoire du Pancreas et de la Rate. Paris, France: Doin; 1933. 29. Elman R, Arneson N, Graham EA. Value of blood amylase estimations in the diagnosis of pancreatic disease: a clinical study. Arch Surg. 1929;19: 943–967.

© 2009 Lippincott Williams & Wilkins

Management of Severe Acute Pancreatitis

30. Walzel P. Zur diagnose und therapie der akuten pankreasnekrose. Beitr Klin Chir. 1929;147:3–15. 31. Mikkelsen O. Pancreatitis akuta: schwere Falle, besonders hinsichtlich ihrer konservativen Behandlung. Acta Chirurg Scand. 1934;75:373– 415. 32. Demel R. Umstrittene fragen bei akuter pankreasnekrose: aktuelles zur aetiologie, diagnose und Behandlung der akuten pankreasnekrose. Wien Klin Wochenschr. 1936;49:1309 –1312. 33. Nordmann O. Neue anschauungen uber die acute pankreasnekrose und ihre Behandlung. Arch Klin Chir. 1938;193:370 –382. 34. Trasoff A, Scarf M. Acute pancreatitis: medical problem. Am J Med Sci. 1937;194:470 – 474. 35. Pratt JH. Diseases of the pancreas. Oxford Med. 1939;3:473. 36. Fallis LS. Acute pancreatitis. Am J Surg. 1939;46:593–599. 37. Lewison EF. Acute pancreatitis. Arch Surg. 1940;41:1008 –1037. 38. Paxton JR, Payne JH. Acute pancreatitis: a statistical review of 307 established cases. Surg Gyneco1 Obstet. 1948;86:69 –75. 39. Pollock AV. Acute pancreatitis: analysis of 100 patients. Br Med J. 1959; 1:6 –10. 40. Bradley EL III. A fifteen year experience with open drainage for infected pancreatic necrosis. Surg Gynecol Obstet. 1993;177:215–222. 41. Uomo G, Molino D, Visconti M, et al. The incidence of main pancreatic duct disruption in severe biliary pancreatitis. Am J Surg. 1998;176:49 –52. 42. Neoptolemos JP, London NJ, Carr-Locke DL. Assessment of main pancreatic duct integrity by ERCP in patients with acute pancreatitis. Br J Surg. 1993;80:94 –99. 43. Lau ST, Simchuk EJ, Kozarek RA, et al. A pancreatic duct leak should be sought to direct treatment in patients with acute pancreatitis. Am J Surg. 2001;181:411– 415. 44. Foster PD, Ziffren SE. Severe acute pancreatitis. Arch Surg. 1962;85:252– 259. 45. Nugent FW, Zuberi S. Treatment of acute pancreatitis. Surg Clin North Am. 1968;48:595–599. 46. Jordan GL, Spjut HJ. Hemorrhagic pancreatitis. Arch Surg. 1972;104:489 – 493. 47. Leger L, Brehant J. Chirurgie du Pancreas. Paris, France: Masson; 1956. 48. Chau A, Grier WN, Pfeffer RB, et al. Pancreatic apoplexy: report of a case treated by partial pancreatectomy, with recovery. Am J Surg. 1959;97:789 – 792. 49. Watts GT. Total pancreatectomy for fulminant pancreatitis. Lancet. 1963: 364. 50. Rives J, Stoppa R, Lardennois B. Traitement des pancreatitis necrotique et hemorrhagiques par la pancreatectomie gauche. Acad Chir. 1969;95:346 – 350. 51. Rettori R, Grenier J. Traitement chirurgical et evolucion precoce des pancratitis aigues. In: 72me Congres Francais de Chirurgie. Vol. 1. Paris, France: Masson; 1970. 52. Devic G, Sarde J, Ge1ain J. Traitement des pancreatitis aigues par pancreatectomie. Lyon Chir. 1970;66:438 – 443. 53. Boutelier P, Edelmann G. Tactique chirurgicale dans les pancreatites aigues necrosantes: plaidoyer en faveur des sequestrectomies. Ann Chir. 1972;26: 249 –259. 54. Fagniez PL, Julien M, Veller M, et al. Sur le traitement chirurgical des pancreatites aigues necrosantes: a propos de 47 cas. Chirurgie. 1974;100: 816 – 817. 55. Edelman G, Boutelier P. Le traitement des pancreatites aigues necrosantes par l’ablation chirurgical precoce des portions necrosees. Chirurgie. 1974; 100:155–167. 56. Rettori R, Clot JC, Tordeanu N, et al. Resultat de traitement surgical, elements du prognostic dans les pancreatitis aigues avec necrose. Chirurgie. 1974;100:168 –179. 57. Kuemmerle F, Neher M, Schonborn H, et al. Vorzeitige Operation bei akuter hamorrhagisch-nekrotisierender Pankreatitis. Dtsch Med Wochenschr. 1975; 100:2241–2245. 58. Schonborn H, Pross E, Olbermann M. Neuere Vorstellungen zur konservativen und operativen Therapie der akuten Pankreatitis. Internist. 1975;16: 108 –115. 59. Autio V, Juusela E, Lauslahti K, et al. Resection of the pancreas for acute hemorrhagic and necrotizing pancreatitis. World J Surg. 1979;3:631– 639. 60. Alexandre JH, Guerri MT. Role of total pancreatectomy in the treatment of necrotizing pancreatitis. World J Surg. 1981;5:369 –377.

www.annalsofsurgery.com | 15

Bradley and Dexter

61. Kivilaakso E, Fraki 0, Nikki P. Resection of the pancreas for acute fulminant pancreatitis. Surg Gyneco1 Obstet. 1981;152:493– 498. 62. Aldridge MC, Ornstein M, Glazer G. Pancreatic resection for severe acute pancreatitis. Br J Surg. 1985;72:369 –377. 63. Mercadier M. Discussion en course. Conclusion de discussion sur la pancreatique aigue. Chirurgie. 1977;103:928 –929. 64. Hollender LF, Gillet M, Sava G. La pancreatectomie d’urgence dans le pancreatitis aigues. Ann Chir. 1970;24:647– 653. 65. Hollender LF, Kohler JJ, Klein A. Zur chirurgischen Behandlung der akuten nekrotischen pankreatitis. Der Chirurg. 1972;43:256 –261. 66. Hollender LF, Starlinger M, Meyer C. Die Chirurgie der akuten Pancreatitis. Aktuelle Chir. 1977;12:43–70. 67. Hollender LF. Resection of the pancreas for acute hemorrhagic and necrotizing pancreatitis. World J Surg. 1979;3:637. 68. Moore F. Metabolic Care of the Surgical Patient. Philadelphia, PA: WB Saunders; 1959:135–261. 69. Shires T, Cohn D, Carrico J. Fluid therapy in hemorrhagic shock. Arch Surg. 1964;88:688 – 693. 70. Shoemaker W, Walker W. Fluid Electrolyte Therapy in Acute Illness. Chicago, IL: Year Book Medical Publishers; 1970. 71. Norton LW, Eiseman B. Near total pancreatectomy for hemorrhagic pancreatitis. Am J Surg. 1974;127:191–195. 72. Altemeier WA, Alexander JW. Pancreatic abscess: a study of 32 cases. Arch Surg. 1963;87:80 – 89. 73. Waterman NG, Walsky R, Kasdan ML, et al. The treatment of acute hemorrhagic pancreatitis by sump drainage. Surg Gynecol Obstet. 1968;126: 963–971. 74. Lawson DW, Daggett WM, Civetta JM, et al. Surgical treatment of acute necrotizing pancreatitis. Ann Surg. 1970;172:605– 615. 75. Warshaw AL, Imbembo A, Civetta JM, et al. Surgical intervention in acute necrotizing pancreatitis. Am J Surg. 1974;127:484 – 490. 76. White TT, Heimbach DM. Sequestrectomy and hyperalimentation in the treatment of hemorrhagic pancreatitis. Am J Surg. 1976;132:270 –275. 77. Leger L, Chiche B, Louve A. La necrose dans les pancreatites aigues. Confrontations operatoires et anatamo-pathologiques. Nouv Presse Med. 1977;5:337–345. 78. Smadja C, Bismuth H. Pancreatic debridement in acute necrotizing pancreatitis: an obsolete procedure? Br J Surg. 1980;23:408 – 410. 79. Ranson JH, Rifkind KM, Roses DF, et al. Prognostic signs and the role of operative management in acute pancreatitis. Surg Gynecol Obstet. 1974;139: 69 – 81. 80. Tenner S, Sica G, Hughes M, et al. Relationship of necrosis to organ failure in severe acute pancreatitis. Gastroenterology. 1997;113:899 –903. 81. Kivisaari L, Somer CG, Standerstokjold-Nordenstam T, et al. Early detection of acute fulminant pancreatitis by contrast enhanced computed tomography. Scand J Gastroenterol. 1983;8:18 –39. 82. Block S, Maier W, Bittner R, et al. Identification of pancreas necrosis in severe acute pancreatitis: imaging procedures versus clinical staging. Gut. 1986;27:1035–1042. 83. Bradley EL III, Murphy M, Ferguson C. Prediction of pancreatic necrosis by dynamic pancreatography. Ann Surg. 1989;210:495–504. 84. Johnson CD, Stephens DH, Sarr MG. CT of acute pancreatitis: correlation between lack of contrast enhancement and pancreatic necrosis. Am J Radiol. 1991;156:93–95. 85. Foitzik T, Bassi DG, Fernandea-del Castillo C, et al. Intravenous contrast medium impairs oxygenation of the pancreas in acute necrotizing pancreatitis in the rat. Arch Surg. 1994;129:706 –711. 86. Hwang TL, Chang KY, Ho YP. Contrast enhanced dynamic computed tomography does not aggravate the clinical severity of patients with severe acute pancreatitis. Arch Surg. 2000;135:287–290. 87. Beger HG, Krautzberger W, Bittner R. Results of surgical treatment of necrotizing pancreatitis. World J Surg. 1985;6:972–979. 88. Pederzoli P, Pilon T, Vantini I, et al. Early surgical treatment of acute hemorrhagic pancreatitis: a pilot study in proteases-antiproteases in clinical practice. Exerpta Med. 1980;62– 65. 89. Beger HG, Buchler M, Bittner R. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis. World J Surg. 1988;120: 255–262. 90. Rau B, Pralle U, Uhl W. Operative vs. non-operative management in sterile necrotizing pancreatitis. J Am Coll Surg. 1995;181:279 –288. 91. Rau B, Bothe A, Beger HG. Surgical treatment of necrotizing pancreatitis

16 | www.annalsofsurgery.com

Annals of Surgery • Volume 251, Number 1, January 2010

and closed lavage: changing patient characteristics and outcome in a 19year, single-center series. Surgery. 2005;138:28 –39. 92. Rattner DW, Legermate DA, Lee MG, et al. Early debridement of symptomatic pancreatic necrosis is beneficial irrespective of infection. Am J Surg. 1992;163:105–110. 93. Fernandez-del Castillo C, Rattner DW, Makary MA, et al. Debridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg. 1998;228:676 – 684. 94. Teerenhovi O, Nordback I, Isolauri J. Influence of pancreatic resection on systemic complications in acute necrotizing pancreatitis. Br J Surg. 1988; 76:93–95. 95. Bradley EL III, Allen K. A prospective longitudinal study of observation vs. surgical intervention in the management of necrotizing pancreatitis. Am J Surg. 1991;161:19 –25. 96. Guillaumes S, Blanco I, Clave P, et al. Nonoperative management of necrosis in acute pancreatitis. Pancreas. 1992;7:740. 97. Uomo G, Visconti M, Manes G, et al. Nonsurgical treatment of acute necrotizing pancreatitis. Pancreas. 1996;12:142–148. 98. Yamamoto M, Saitoh Y. Severe acute pancreatitis in Japan. JHPB Surg. 1996;3:203–209. 99. Buechler MW, Gloor B, Muller CA, et al. Acute necrotizing pancreatitis treatment strategy according to the status of infection. Ann Surg. 2000; 232:627– 629. 100. Ashley SW, Perez A, Pierce E. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg. 2001;234:572–580. 101. Andersson B, Olin H, Eckerwall G, et al. Severe acute pancreatitis: outcome following a primarily non-surgical regime. Pancreatol. 2006;6:536 –541. 102. Rutledge PL, Warshaw AL. Persistent acute pancreatitis: a variant treated by pancreatoduodenectomy. Arch Surg. 1988;123:597– 600. 103. Howard TJ, Moore SA, Saxena R, et al. Pancreatic duct strictures are a common cause of recurrent pancreatitis after successful management of pancreatic necrosis. Surgery. 2004;136:909 –916. 104. Bradley EL III. Enteropathies. In: Complications of Pancreatitis: Medical and Surgical Management. Philadelphia, PA: Saunders; 1982:274:–287. 105. Kyosola K, Fock G. Complications in acute pancreatitis. Ann Chir Gynaecol. 1976;65S:7–12. 106 .Gerzof SG, Banks PA, Robbins AH, et al. Early diagnosis of pancreatic infection by computed tomography-guided aspiration. Gastroenterology. 1987;93:1315–1320. 107. Lumsden A, Bradley EL III. Secondary pancreatic infections. Surg Gynecol Obstet. 1990;170:459 – 467. 108. Davidson E, Bradley EL III. Marsupialization in the treatment of pancreatic abscess. Surg Gynecol Obstet. 1981;193:638 – 648. 109. Bolooki H, Jaffe B, Gliedman ML. Pancreatic abscesses and lesser sac collections. Surg Gynecol Obstet. 1968;126:1303–1308. 110. Kiekens R, Kinneart P, Govearts JP. La jejunostomie d’alimentation dans le traitment de la pancreatite aigue grave. Acta Chir Belg. 1967;66:45–57. 111. Bradley EL III, Fulenwider JT. Open treatment of pancreatic abscess. Surg Gynecol Obstet. 1984;159:509 –513. 112. Bradley EL III. Management of infected pancreatic necrosis by open packing. In: Beger HG, Warshaw AL, Buchler MW, eds. The Pancreas. London: Blackwell; 1998:1439 –1442. 113. Ho HS, Frey CF. The role of antibiotic prophylaxsis in severe acute pancreatitis. Arch Surg. 1997;132:487– 493. 114. Sarr MG, Nagorney DM, Mucha P, et al. Acute necrotizing pancreatitis: management by planned, staged pancreatic necrosectomy and delayed primary wound closure over drains. Br J Surg. 1991;78:576 –581. 115. Howard JM. Studies of acute pancreatitis with retroperitoneal necrosis–the “suet syndrome”: improvement in patient survival. J Hepatobiliary Pancreat Surg. 1996;3:185–202. 116. Howard JM. Delayed surgical debridement of massive peripancreatic necrosis associated with acute necrotizing pancreatitis. Surg Gynecol Obstet. 1989;168:25–29. 117. Dubner H, Steinberg W, Hill M, et al. Infected pancreatic necrosis and peripancreatic fluid collections: serendipitous response to antibiotics and medical therapy in three patients. Pancreas. 1996;12:298 –302. 118. Adler DG, Chiari ST, Dahl TJ, et al. Conservative management of infected necrosis complicating severe acute pancreatitis. Am J Gastroenterol. 2003; 1:98 –103. 119. Mier J, Leon EL, Castillo A, et al. Early versus late necrosectomy in severe necrotizing pancreatitis. Am J Surg. 1997;173:71–75.

© 2009 Lippincott Williams & Wilkins

Annals of Surgery • Volume 251, Number 1, January 2010

120. Carroll L. Alice’s Adventures in Wonderland: Through the Looking Glass. London: Puffin; 1962:274. 121. Bradley EL III. A clinically based classification system for acute pancreatitis: summary of the international symposium on acute pancreatitis, Atlanta, Georgia, September 11–13, 1992. Arch Surg. 1993;128:586 –590. 122. vanSonnenberg E, Wing VW, Casola G, et al. Temporizing effect of percutaneous drainage of complicated abscesses in critically ill patients. Am J Radiol. 1984;142:821– 826. 123. Carpenter KD, Freeney PC. Imaging and imaging-guided therapy in severe acute pancreatitis. J Hepatobiliary Pancreat Surg. 1996;3:253–264. 124. Brolin RE, Flancbaum L, Ercoli FR, et al. Limitations of percutaneous catheter drainage of abdominal abscesses. Surg Gynecol Obstet. 1991;173: 203–210. 125. Baron T. Organized pancreatic necrosis: endoscopic, radiologic, and patho-

© 2009 Lippincott Williams & Wilkins

Management of Severe Acute Pancreatitis

logic features of a distinct clinical entity. Pancreas. 1999;1:105–108. 126. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: An initial experience. Ann Surg. 2000;232:175–180. 127. Chang YC, Tsai HM, Lin XZ, et al. No debridement is necessary for symptomatic or infected acute pancreatitis: delayed, mini-retroperitoneal drainage for acute necrotizing pancreatitis without debridement and irrigation. Dig Dis Sci. 2006;51:1388 –1395. 128. Parekh D. Laparoscopic-assisted pancreatic necrosectomy: a new surgical option for treatment of severe necrotizing pancreatitis. Arch Surg. 2006;141: 895–903. 129. Bradley EL III, Howard TJ, vanSonnenberg E, et al. Intervention in necrotizing pancreatitis: an evidence-based review of surgical and percutaneous alternatives. J Gastrointest Surg. 2008;12:634 – 639.

www.annalsofsurgery.com | 17