Synopsis  

Complications of acute and chronic pancreatitis are varied, often complex, and potentially fatal. This chapter will attempt to summarize all of these feared developments, address their causation, and review current and potential future approaches.

The first section deals with the toxic and metabolic complications. Early death from pancreatitis most frequently follows shock, a poorly understood but dramatic occurrence seen, fortunately, in a small minority. All organ systems may be affected, in a fashion that can range from mild to very severe, resulting in renal failure, respiratory failure, disseminated intravascular coagulation, and severe gastrointestinal bleeding. Prolonged septic-like systemic inflammatory response syndrome may result in coma, and a profound catabolic state untreatable by total parenteral nutrition.

Recent clinical experience has recognized the major difference in clinical course and treatment when pancreatic necrosis complicates the early phase of acute pancreatitis. Diagnosis, treatment, and predictors of outcome are addressed.

The chapter closes with a comprehensive review of the main miscellaneous complications which are often recognized during the later phase of pancreatitis. Patients who appear to be recovering may experience fistula formation, producing acute pancreatic hydrothorax or enteric fistulas. Percutaneous or surgical drainage is a frequent cause of cutaneous fistulas, a particularly difficult and debilitating complication. Finally, dramatic and potentially fatal vascular complication can occur very abruptly, taxing the diagnostic and interventional skills of the treating team.

Throughout this review up-to-date reports of successful approaches to all of these events are addressed. We emphasize that since controlled trials of newer treatment modalities have, in general, not been directly compared to traditional surgical therapy, a multidisciplinary approach with gastroenterology, interventional radiology, and surgery is vital.

 Toxic and metabolic complications   

Few diseases can produce more varied and severe diffuse metabolic complications than acute pancreatitis. Although the organ weighs only 90 g in the adult, when inflamed, the pancreas can produce such profound systemic effects that respiratory, renal, and circulatory failure can rapidly ensue, producing multiorgan system death within a few days. In patients with severe acute pancreatitis, the likely sequence is peripancreatic leakage of proteolytic juice, activation of proinflammatory cytokines, third space loss secondary to retroperitoneal injury, hypoperfusion, and finally, rapid progression to frank pancreatic necrosis.

This sequence is variable and may appear as partial individual organ compromise or full-blown multiorgan failure. The recognition of these end organ complications served as the basis for the famous 'Ranson's criteria', which has for nearly 30 years remained the most popular clinical measure of the severity of acute pancreatitis (Fig. 1) [1].

A principal drawback to Ranson's scoring system has been the need to reassess the patient at 48 h before a final score can be calculated. In an effort to predict severity as early as possible to help triage patients to the most appropriate level of care, several other clinical scoring systems have been advocated [2–4]. The Acute Physiologic and Chronic Health Evaluation scoring system (APACHE II and III) permits a more comprehensive initial assessment of severity. Khan pointed out recently that since the natural history of pancreatitis varies considerably depending upon underlying etiology, patient responses, and presence of comorbidity, APACHE scoring, when repeated at 48 h, as in Ranson's criteria, more accurately predicts outcome [4].

The modified Glasgow Coma/Imrie score has also been used to predict severity and thus the development of complications. One direct comparison of these three scoring systems recently reported that APACHE III and modified Glasgow/Imrie had a greater magnitude of correlation with length of hospital stay as a measure of severity than did Ranson's traditional criteria [2]. However, death was equally predicted, with fatal cases uniformly having > 5 Ranson criteria, APACHE III scores > 30 at 96 h, and modified Glasgow/Imrie scores > 4.

Alternatively, in a slightly smaller but contemporary study, Ranson's criteria, particularly at 48 h, remained a valid scoring system for severity of pancreatitis when compared to the newest APACHE III scoring [5]. Because of its inherit simplicity, the authors advocated the continued use of the venerable Ranson's criteria and emphasized that elevated BUN, low calcium, base deficit, and third space loss predicted mortality.

In clinical practice currently, a CT-based measure of severity has proved valuable when added to clinical scoring [6]. Robert recently analysed all available clinical scoring systems plus a CT scan score of severity in 130 patients [7]. Multivariate analysis revealed that low serum albumin plus extra pancreatic fluid collections on initial CT scanning within the first 24 h was the best predictor of severe pancreatitis overall.

Although some concern has been raised as to the potential for renal toxicity of contrast-enhanced early CT scanning in acute pancreatitis, the general consensus suggests that the value of the information gained remains worthwhile [8].

Shock and renal failure have traditionally been the most frequent and perhaps the most feared of the potential toxic and metabolic complications of acute pancreatitis. In the mid-1960s, Frey reported a 22% rate of shock and renal failure in his series of 490 patients [9]. At that time mortality followed in nearly 80% of these complications.

Although shock and renal failure are associated in most patients, oliguria and even anuria requiring prolonged dialysis can be seen in euvolemic patients with normal blood pressures throughout their early illness. This dramatic systemic complication is attributed to profound renal cortical vasoconstriction due to poorly understood nephrotoxic vasoactive circulating factors produced by the acute inflammatory response. Permanent renal failure eventuating in renal transplantation rarely occurs.

Respiratory complications occur in a variable percentage and often in a stepwise fashion, with hypoxemia with a normal chest X-ray being the most frequently recognized. Pulmonary infiltrates, atelectasis with elevated diaphragms, and pleural effusions may then ensue and progress to full-blown acute respiratory distress syndrome (ARDS), often requiring ventilatory support. This sequence is again poorly understood but appears to be precipitated by circulating proinflammatory cytokines due to the retroperitoneal proteolytic injury, producing an alveolar capillary leak. The flood of protein-rich fluid can, at times, produce such complete consolidation that mechanical ventilation cannot compensate, resulting in a respiratory death [10].

Disseminated intravascular coagulation (DIC) is a less frequent complication of acute pancreatitis [1]. DIC may produce severe bleeding since extensive retroperitoneal injury is present. The consequence of such hemorrhage can greatly worsen the above outlined systemic complications.

A more recently described toxic reaction to pancreatitis has been termed systemic inflammatory response syndrome (SIRS) [11]. High fever, tachycardia, and delirium may persist for days or even weeks and may be difficult to distinguish from bacterial sepsis. This syndrome can produce a profoundly catabolic state that cannot be adequately reversed with total parenteral nutrition (TPN) [12–14].

In patients with severe SIRS, fine-needle radiologically guided aspiration of retroperitoneal fluid collections has proved to be valuable for distinguishing this syndrome from complicating retroperitoneal infection. To avoid potential contamination of otherwise sterile fluid, a fastidious technique to avoid traversing bowel, particularly colon, and with adequate skin preparation, is mandatory [15,16].

In patients with a less severe form of SIRS, TPN can be switched to enteral feeding to further minimize pancreatic stimulation; Takacs demonstrated that adding octreotide to enteric feedings reduces gastrointestinal hormone release (CCK, glucagon, gastrin, and glucose) to baseline [17]. Windsor has developed evidence that this change to the enteral route for nutrition can actually diminish the inflammatory response and lead to more rapid improvement compared to continuing TPN [18]. Tube placement, duodenal edema, and prolonged ileus may all hamper the initiation of enteral feeding in sicker patients, however.

Platelet activating factor (PAF) has been implicated in the pathogenesis of SIRS, possibly by amplifying mediators of inflammation. Unfortunately, an attempt to block this effect by the infusion of lexiphant, a potent inhibitor of PAF, was demonstrated to have no clinical benefit in a placebo-controlled trial [19].

Death due to acute pancreatitis follows two patterns: early (within 1, or in some series, 2 weeks) and late. Early death is secondary to these incompletely understood severe metabolic inflammatory responses leading to multiorgan system failure. Death later than 2 weeks is generally secondary to sepsis and represents about one-half of fatal cases [20]. Death rates in various recent series with variable etiologies range from 2 to 9%, which represents a marked improvement compared to the pre-TPN era [21].

Factors which may be important in the sequence of rapid onset multiorgan system failure are under active research. Ethridge et al. recently studied the impact of cyclo-oxygenase (COX) on pancreatitis severity, especially with regard to lung injury [22]. In their mouse model, inhibition of COX-2 or deletion of the COX gene profoundly decreased severity of pancreatitis and protected against ARDS.

Host response may prove to be a critical determinate of the severity of pancreatitis and the resulting complications [23,24]. A defect in interleukin-10 function has been identified in some patients with severe acute pancreatitis [25]. This defect might result in a failure to down-regulate the initial acute inflammatory reaction once they have been maximally stimulated. If confirmed, this finding would help us to understand why these seemingly unpredictable complications may be seen even after rather trivial injury, such as after simple cannulation or unremarkable sphincterotomy at ERCP.

In an attempt to prevent post-ERCP pancreatitis, an initial randomized trial of a single bolus of interleukin-10 (8 µg/kg) was given to patients 15 min prior to ERCP and compared to a placebo group [26]. Unfortunately, both groups experienced the identical rate of pancreatitis and, furthermore, the severity as measured by length of stay was the same. Nevertheless, a full understanding of the role of pro- and anti-inflammatory cytokines may be the best hope of preventing or at least attenuating the severity of acute pancreatitis.

 Pancreatic fluid collections   

In the setting of pancreatitis, fluid collections are frequent and have been used as a measure of severity to create a CT-based score as previously discussed [6]. When combined with extent of non-perfusion (equivalent to necrosis) and detection of retroperitoneal complications, CT becomes a vital tool in predicting severity [27].

Fluid collections are at first sterile, amylase rich, and unorganized. They are seen early, often within 24 h, and may persist. Unencapsulated and younger than 4 weeks from the onset of pancreatitis, they are properly watched expectantly (Fig. 2 B). If necrosis is not present, infection is rare. Whether they will persist, organize, and encapsulate beyond 4 weeks into a true pseudocyst largely depends on pancreatic ductal anatomy and the presence of necrosis. Normal ductal and sphincter anatomy predicts spontaneous resolution, whereas fistula, ductal obstruction, or disruption predict persistence and pseudocyst formation [28]. Finally, necrosis will organize slowly and an associated fluid collection will usually become a complex pseudocyst; this might be better termed an area of 'organizing necrosis' to distinguish them [29].

 Pseudocysts and abscesses   

A pseudocyst of the pancreas is a maturing fluid collection surrounded by granulation tissue which occurs as a consequence of acute or chronic leakage of pancreatic juice (Fig. 2A). In distinction, acute fluid collections may be wholly inflammatory and will frequently resolve. Alternatively, severe ductal disruption with a consequent large fistula will virtually always result in pseudocyst formation [30,31].

Prior to any decision regarding management, a pancreatic pseudocyst must be carefully distinguished from other cystic or fluid-filled collections in the retroperitoneum [32–35]. In the absence of a definite attack of pancreatitis, some pseudocysts may be exceedingly difficult to differentiate from true cystic neoplasms. [36,37]. Inadvertent endoscopic or surgical drainage of true neoplasms is ineffective and occasionally disastrous. Be aware that obstructing neoplasms, usually malignant, can produce acute or chronic pancreatitis and result in a pseudocyst. These combination cases can be extremely challenging to diagnose and treat [38]. In most cases careful clinical history, dynamic CT, and pancreatography are necessary prior to establishing a complete diagnosis. In atypical cases, endoscopic ultrasound (EUS) adds important diagnostic information, especially if a small obstructing tumour or a cystic neoplasm is present [39–42].

The evolving pseudocyst requires time to develop a complete encircling wall which, of course, lacks an epithelial lining (Fig. 2 B). By recent consensus, 4 weeks has been chosen as a minimal time from the onset of acute pancreatitis until this process is reasonably complete and the collection can be termed a pseudocyst [43]. The development of a pseudocyst in chronic pancreatitis is often more difficult to precisely age. These collections often occur more gradually as a consequence of ductal obstruction by stones or fibrotic strictures and are frequently mature upon discovery. Once a secure diagnosis of a mature or maturing pseudocyst is established, the clinician must remember that many pseudocysts will still resolve. Early authors emphasized that pseudocysts larger than 6 cm rarely resolve, but size alone does not always warrant intervention [30]. Etiology remains important since pseudocysts complicating acute pancreatitis are much more likely to resolve than those due to chronic pancreatitis [44]. Clinicians were formerly hesitant to follow pseudocysts for fear of the spontaneous development of complications, especially infection and hemorrhage. However, several authors have reported a surprisingly low incidence of adverse events during follow-ups in asymptomatic patients with stable or slowly resolving pseudocysts [45,46].

In general, patients with continued ductal leakage communicating with the pseudocyst will not stabilize and permit resumption of a diet (Fig. 3). Clinicians should be suspicious if, after initial improvement with symptom resolution on IV therapy (with or without TPN and octreotide), pain returns upon oral feeding. A serious ductal injury or leak is often present in this setting [47].

Unchecked, continued leakage with resulting expansion can produce a series of additional complications. Progressive enlargement in the usual retrogastric position will produce gastric compression resulting in early satiety, nausea, vomiting, pain, and weight loss. Frank outlet obstruction usually occurs at the pylorus or in the duodenal sweep (Fig. 4). When the ampullary area and pancreatic head are involved obstructive jaundice often occurs.

The accumulating fresh enzyme-rich pancreatic juice from a ductal leak can lead to damage and digestion of additional retroperitoneal structures such as blood vessels, or adjacent hollow organs such as the duodenum or colon, or lead to spontaneous perforation into the peritoneal cavity resulting in pancreatic ascites.

Spontaneous infection of a pseudocyst is a feared late complication. Formerly termed infected pseudocyst, the preferred term now is pancreatic abscess [43]. The source of the infecting organism is often unknown, presumably from transmigration of nearby colonic bacteria or from transient bacteremia, but may follow line sepsis or fine-needle aspiration. Gas-producing organisms produce multiple small air bubbles in the pseudocyst, but the presence of gross air within the cavity is often due to fistulization to the duodenum or colon (Fig. 5)[48].

Gaining an understanding of the pancreatic ductal anatomy earlier in the course of pseudocyst patients holds the promise of allowing earlier necessary intervention in pancreatitis patients [28,49]. Traditionally, diagnostic ERCP has been the major test for defining pancreatic ductal anatomy, but most clinicians would not perform early ERCP due to the fear of introducing potentially catastrophic infection. Magnetic resonance pancreatography (MRP) is proving useful in detecting major disruption and predicting the need for intervention without this risk. In our experience MRP does not permit reliable ductal anatomy in lesser leaks since communicating fistulas may contain very little of the static fluid upon which MRCP depends for imaging.

Once the need for intervention has become apparent, the choice of drainage has gradually yielded to endoscopic transmural and transpapillary techniques [50,51]. Clear indications for intervention in well-defined pseudocysts are summarized in Fig. 6. Radiological drainage remains popular in some centres but risks include prolonged drainage, introducing infection, and establishing a fistula when ductal obstruction or disruption is present [30,52].

The detailed techniques of endoscopic drainage are outlined completely in several recent comprehensive reviews [53,54]. To summarize, transmural endoscopic puncture begins with first identification of the appropriate site, then proceeds to needle localization with injection of contrast, puncture (Fig. 7 A, B) 10 mm balloon dilation of the tract, and, finally, multiple stent placement (Fig. 8 A, B, C, D).

Transpapillary therapy involves pancreatic stent placement with or without pancreatic sphincterotomy [55–57]. We have employed pancreatic sphincterotomy alone when the anatomy suggests that complete decompression of the duct by transecting the sphincter muscle should be adequate, thus avoiding the risks of pancreatic stent placement.

A recent long-term follow-up of such endoscopically treated pseudocysts reported excellent results with 15% (6 of 38) recurrence, all in alcohol-induced and therefore, chronic pancreatitis [58]. Complications of endoscopic drainage include perforation when the true lumen of the pseudocyst is missed or the pseudocyst is not adherent or sufficiently organized. Severe bleeding, especially from the gastric wall was a former frequent complication when simple diathermic puncture was followed by extension of the entry point using a sphincterotome [59]. A recently described technique of drainage to avoid any cautery advocated placing a guidewire through a special needle used for the initial localizing puncture. Once the guidewire is within the cavity, balloon dilation alone is used to create the endoscopic cystenterostomy drainage site [60].

Overall, the success rates for endoscopic pseudocyst drainage are summarized in Fig. 9 and approach 80%. The complication risk and recurrence rates stated in these studies are included in Fig. 10 and were 12 and 16%, respectively. Finally, recurrences could be retreated endosurgically about 50% of the time, relegating surgery to only 8 patients out of the 141 cases selected and reported.

The role of EUS pseudocyst management remains in evolution (Fig. 11). Some authors report rarely needing or using EUS but others consider EUS valuable in selecting appropriate drainage sites, excluding intervening blood vessels, and determining depth of puncture [53,61]. However, when collected as diagnostic this information may be difficult to transfer to actual therapy, which is generally done with larger-channel therapeutic endoscopes whose angle of view and therefore alignment for puncture is different from EUS [62,62]. Direct EUS therapeutic drainage but placing 7 Fr stents has been reported, but results are unlikely to equal 10 Fr drainage [63]. Finally, the development of a therapeutic 4.2 channel EUS endoscope has permitted complete therapy, including 10 Fr stent placement in pseudocyst patients [64]. As EUS becomes a more available therapeutic technique this approach may compete well with our current endoscopic transmural drainage procedure.

Finally, laparoscopically directed pseudocyst drainage has been reported but appears to be more invasive, more costly, and has not defined a superior outcome [65].

 Pancreatic necrosis   

Necrosis of pancreatic tissue complicates acute pancreatitis in a variable percentage of cases, is seen less often in acute exacerbations of chronic pancreatitis, and accounts for many of the complications and much of the mortality (Fig. 12). Etiologies may have an impact on severity with the pancreatitis caused by hypertriglyceridemia producing necrosis in perhaps the highest percentage of at-risk cases [66].

In a large single institutional review, Blum et al.[21] reported a respectably low overall mortality rate of 5% amongst 368 cases of acute pancreatitis, again with about half being earlier than 2 weeks and the remainder later. To emphasize the importance of necrosis, only 36 cases (10%) had documented necrosis but accounted for 9 of the overall 17 deaths. Thus, presence of necrosis resulted in an eventual death rate of 25%. Finally, the authors noted that late deaths in the absence of necrosis were seen in only 4 of 212 patients at risk (2%).

At present, the exact mechanism of necrosis is unknown but ischemic infarction is held as most likely. Poor perfusion secondary to rapid third space loss has been postulated but recent data suggest that the process of necrosis may be underway very rapidly before perfusion is affected. In a retrospective case analysis, patients with necrosis presented earlier but had a similar incidence of hemoconcentration compared to patients with interstitial pancreatitis [67]. Resuscitation volumes were similar retrospectively in both groups. However, patients whose hematocrits continued to rise despite large volumes of fluid resuscitation were all subsequently proven to have necrosis. A cause and effect of inadequate resuscitation could not be established.

The consequence of necrosis is a high likelihood of developing infection in the devitalized tissue, and the loss of a functioning pancreas with consequent diabetes, fistula formation, and various vascular injuries. Many of these complications result in the need for operative and, more recently, ensodcopic management.

Since pancreatic necrosis produces significant morbidity and a large proportion of the late mortality caused by acute pancreatitis, a search for necrosis using dynamic CT is generally felt justified [68].

Management of necrosis initially is conservative, with the expectation of most patients who do not develop infection eventually spontaneously resolving [69]. However, once the necrotic tissue becomes infected, intervention is almost always required. At present, the majority of these patients are still best managed with surgical debridement and drainage, almost always externally [15]. Prolonged hospitalization with multiple procedures often follows, with surgical centers favouring either closed drainage with subsequent radiologically assisted catheter drainage or open drainage with surgically placed abdominal mesh to permit planned repeated debridements [70].

A few cases of attempted retroperitoneal laparoscopic necrosectomy have been reported [71,72]. At present this experience is anecdotal and no comparative trials have yet been reported. The risk of sudden and severe bleeding, and the need for multiple repeat interventions have prevented wide adoption of the technique.

In an attempt to prevent the development of infection in the setting of necrosis, the use of broad-spectrum antibiotics, especially imipenem, has reached a consensus. All eight recently reviewed trials demonstrated benefit in the patients receiving broad-spectrum antibiotics [73]. Many questions remain as to the use of newer antibiotics, the duration of therapy, the timing of onset of use, and the need for fungal coverage [74,75].

Organizing necrosis   

As stated earlier, persistent necrotic material organizes and encapsulates into a complex collection containing a mixture of solid and semisolid debris and fluid. Simple catheter drainage will be insufficient to evacuate this material and infection will often complicate such efforts. When approaching apparent pseudocyst patients, it is of paramount importance to assess for necrosis, then plan and treat patients appropriately [76]. Endoscopic treatment of organizing necrosis is possible but demands techniques of wider drainage such as the placement of multiple stents, creation of a large cyst gastrostomy, and at times nasocystic lavage [29] (Fig. 13 A, B, C, D).

Repeated endoscopic procedures should be anticipated since cavity infections will occur in greater than 50%. When prompt reintervention is performed, these infections can usually be managed with lavage and repeat or additional stent placement. Nevertheless, a multidisciplinary approach to these cases is mandatory for optimal patient outcome. The interventional disciplines of surgery, gastroenterology, and radiology all have roles to play in specific situations [66].

 Miscellaneous complications   

Pancreatic fistulas   

These occur in both interstitial and necrotizing pancreatitis. In the presence of an intact pancreatic sphincter or a ductal stricture, the initial leak continues and, as discussed earlier, is often the etiology of pseudocyst formation. At times and for unclear reasons, some collections do not wall-off and the fistula may track throughout the retroperitoneum. Fistulous communication under the diaphragmatic cruri can result in amylase-rich pleural effusions, broncho-pleural fistulas, or even pericardial tamponade [77,78]. Cases of inguinal, scrotal, femoral, and other hernias developing with amylase-rich fluid tracking down these potential spaces have been reported.

Internal fistulas adjacent to hollow organs are perhaps the most frequently recognized. Fistulization to the duodenum may result in resolution of an otherwise expanding pseudocyst as mentioned earlier [48]. Communication between a pseudocyst and the colon will be complicated by sepsis and generally will require surgery. However, Howell et al. reported successful endoscopic treatment of two such cases without requiring surgery [79].

Perhaps, the most dramatic consequence of a pancreatic ductal fistula is pancreatic ascites. Easily diagnosed by routine testing of paracentesis fluid for amylase, these rather rare cases are often overlooked and treated mistakenly as cirrhotic ascites since liver and pancreatic disease often coexist in the alcoholic.

Finally, cutaneous pancreatic fistulas occur after attempts at external drainage are performed. Although these very severe, disabling fistulas are occasionally unavoidable, they are often a consequence of imprecise knowledge of the true diagnosis or the lack of appreciation of the importance of ductal anatomy (Fig. 14).

Currently, many of these complex fistulas can be managed endoscopically providing the duct is intact to the papilla. Various authors advocate pancreatic stent placement or nasopancreatic drainage with or without pancreatic sphincterotomy. Rapid closure of these fistulae can be expected with effective endoscopic transpapillary drainage. If no infection is present, endoscopic management is often definitive and should be attempted before external drainage establishes a cutaneous fistula [80].

Ductal disruption   

Severe ductal disruption is the rule in necrosis cases but can be seen in well-perfused interstitial pancreatitis. To define the term, disruption occurs when the main pancreatic duct has been transected by the inflammatory process of pancreatitis, most likely by direct proteolytic digestion or ischemic infarction. Ductal disruption greatly complicates the approach to treatment and worsens outcome in both acute and chronic pancreatitis. Spontaneous resolution without intervention is very unlikely to occur. External cutaneous fistulae usually follow a percutaneous or surgical drainage approach due to the presence of a viable but disconnected gland. Although the downstream pancreas can be drained and diverted endoscopically by transpapillary therapy, the upstream pancreas continues to contribute to persistence of the fistula. This so-called 'disconnected tail syndrome' often results in pseudocyst recurrence after internal transmural endoscopic or surgical internal cystgastrostomy drainage [51] (Fig. 15). A few authors have reported successful endoscopic drainage by bridging the disruption to reconnect the tail, but the long-term outcome of these efforts remains unclear. More often these patients will experience a long illness with TPN and repeated interventions until the disconnected tail eventually autolyses, atrophies due to stricturing, or is surgically resected [81].

Vascular complications   

Venous thrombosis   

A frequent vascular complication of acute pancreatitis is thrombosis of the splenic vein and, less frequently, of the portal vein [82]. The cause is an intense inflammatory response surrounding these venous structures, often with compression by the resulting edematous reaction. Stasis and activation of clotting factors then produce acute thrombosis with resulting left-sided portal hypertension. Because the obstruction to portal inflow to the liver is usually partial, esophageal varices usually do not occur. Nonetheless, bleeding from gastric varices can be severe, especially when coagulopathy coexists (Fig. 16).

During the period of convalescence, where often surgical debridement or pseudocyst drainage must be undertaken, a secondary venous thrombosis may be a major determinant in treatment selection. Furthermore, the failure to recognize this form of portal hypertension prior to such interventions can prove disastrous. Significant gastric wall varices often contraindicate endoscopic or even surgical pseudocyst gastrostomy. Helical dynamic contrast CT scanning should detect venous thrombosis and predict left-sided portal hypertension accurately (Fig. 17). EUS has proven particularly valuable in assessing for gastric varices. One or both studies should be performed near the time of any invasive intervention.

 Arterial complications   

Thrombotic arterial complications secondary to acute pancreatitis are less common, but when they occur they can be severe. Splenic artery thrombosis with resulting splenic infarction is generally survivable with splenectomy. However, superior mesenteric artery thrombosis resulting in small and, at times, large bowel infarction is accompanied by a high mortality. The middle colic artery is perhaps the most frequent artery to thrombose, often resulting in a more limited large bowel infarction which may respond to resection and temporary surgical colostomy.

A more frequent arterial complication of pancreatitis is the formation of a pseudoaneurysm resulting in hemorrhage. Various series report this serious complication in up to 10% of cases of severe acute pancreatitis and it can complicate chronic pancreatitis as well [83,84].

If the pseudoaneurysm formed in an expanding pseudocyst wall, sudden hypotension with syncope followed by intense pain has been termed 'pancreatic apoplexy. If the pseudocyst into which the pseudoaneurysm ruptures communicates with the pancreatic duct, frank gastrointestinal bleeding can be the presenting symptom. Termed 'hemosuccus pancreaticus', such bleeding is amongst the rarest causes of gastrointestinal hemorrhage [85].

Finally, the presence of a pseudoaneurysm may be silent, only to acutely rupture during any invasive intervention where the surrounding tamponade is decompressed. This can be especially devastating in endoscopic pseudocyst drainage since prompt control of bleeding in general is not possible. Delayed rupture may also occur, resulting in exsanguinating gastrointestinal bleeding if a pseudocyst enterostomy has been created or if a surgical or radiological external drain has been placed [86].

To avoid these severe bleeding complications, it is imperative that the presence of a pseudoaneurysm is carefully searched for before intervention. All drainage procedures are strictly contraindicated until such a vascular lesion can be addressed and resolved. Dynamic, arterial phase, thin-section helical CT scanning through the pancreatic region is likely the best diagnostic study [87] (Fig. 18). Doppler ultrasound can be confirmatory but does not have the comprehensive screening power of CT. MRI with an arteriography protocol has been little reported but would likely visualize these lesions [8].

Once detected, preoperative angiography with embolization of the pseudoaneurysm has become a popular approach [88] (Fig. 19). These procedures can be technically challenging if the pancreatico-duodenal artery is the affected vessel since embolization may be necessary from both the celiac trunk and the superior mesenteric artery. Pseudoaneurysm of the celiac trunk can present a nearly insurmountable problem since gallbladder, gastric, and even hepatic infarction may follow embolization. If portal vein thrombosis is also present, the risk of hepatic infarction increases dramatically. Successful treatment of hemosuccus pancreaticus radiological embolization at angiography is the preferred approach as well [89].

Once the pseudoaneurysm is thoroughly embolized and thrombosed, interventions can then be safely carried out [84]. Elton et al. reported successful endoscopic pseudoaneurysm/pseudocyst drainage following radiological embolization in three such cases [90]. In all three patients, thrombosis following embolization was documented by repeat dynamic contrast CT or Doppler ultrasound prior to endoscopic intervention. Successful endoscopic drainage of the obstructing pseudocyst, stent management of strictures, and clearance of obstructing clots within the pancreatic duct resulted in symptom resolution and avoided surgery in these cases.

Finally, massive diffuse retroperitoneal bleeding may be seen in the setting of necrotizing pancreatitis, often with coincident coagulopathy. This so-called 'hemorrhagic pancreatitis' is less often reported since better radiology more often identifies a focal arterial source. However, when true diffuse hemorrhagic pancreatitis does occur, mortality rates exceed 35%, even in the modern era [88].

 Summary   

Complications of pancreatitis vary widely, are of complex etiology, and involve multiple organ systems. Avoiding these complications remains the basic goal for all treating physicians, but once present their expert detection and appropriate management is the key to optimizing patient outcome [91]. Great progress has been made in treating these supremely ill patients but early and specific treatments to prevent complications are still lacking. Prolonged hospitalizations, TPN, dialysis, ventilatory support, antibiotic therapy, and radiological, endoscopic, and surgical treatments all have had a role in reducing mortality to less than 10% of afflicted patients. However, much needs to be discovered [92].

 Outstanding issues and future trends   

The major need in pancreatology remains a full understanding of the pathophysiology of acute pancreatitis that results in the dramatic cascade of events outlined in the chapter. Once the earliest events are identified, specific medical interventions, possibly extremely specific pharmacological agents, can be developed that can prevent progression to shock, end organ compromise, necrosis, and the other late complications outlined. More basic research is needed.

Lacking this knowledge, research will continue to look for methods of preventing the complications of pancreatitis once severe disease has been established. A major need is an effective way to prevent progression to necrosis, beyond aggressive fluid resuscitation.

Trends in the future will continue to be innovations in minimally invasive therapies. Debridement of infected necrosis, intervention prior to infection, and management of ductal disruption resulting in a disconnected tail are all areas of considerable confusion and often subjects of interdisciplinary debate. Therapeutic, endoscopic, percutaneous laparoscopic debridement and transgastric endoscopic therapy are the newest players on a seemingly crowded field.

 References  

  1 Ranson, JHC, Rifkind, KM & Roses, DF et al. Prognostic signs and the role of operative management in acute pancreatitis. Surg Gynecol Obstet 1974; 139: 69–81. PubMed

  2 Williams, M & Simms, HH. Prognostic usefulness of scoring systems in critically ill patients with severe acute pancreatitis. Crit Care Med 1999; 27: 901–7. PubMed CrossRef

  3 Halonen, KI, Pettila, V, Leppaniemi, AK, Kemppainen, EA, Puolakkainen, PA & Haapiainen, RK Multiple organ dysfunction associated with severe acute pancreatitis. Crit Care Med 2002; 30: 1274–9. PubMed CrossRef

  4 Khan, AA, Parekh, D & Cho, Y et al. Improved prediction of outcome in patients with severe acute pancreatitis by the APACHE II score at 48 hours after hospital admission compared with the APACHE II score at admission: acute physiology and chronic health evaluation. Arch Surg 2002; 137: 1136–40. PubMed CrossRef

  5 Eachempati, SR, Hydo, LJ & Barie, PS. Severity scoring for prognostication in patients with severe acute pancreatitis. Arch Surg 2002; 17: 730–6. CrossRef

  6 Balthazar, EJ & Fisher, LA. Hemorrhagic complications of pancreatitis: radiologic evaluation with emphasis on CT imaging. Pancreatology 2001; 1: 306–13. PubMed

  7 Robert, JH, Frossard, JL & Mermillod, B et al. Early prediction of acute pancreatitis: prospective study comparing computed tomography scans, Ranson, Glasgow, acute physiology and chronic health evaluation II scores, and various serum markers. World J Surg 2002; 26: 612–19. PubMed CrossRef

  8 Robinson, PJ & Sheridan, MB. Pancreatitis: computed tomography and magnetic resonance imaging. Eur Radiol 2000; 10: 401–8. PubMed CrossRef

  9 Frey, CF & Brody, GL. Relationship of azotemia and survival in bile pancreatitis in the dog. Arch Surg 1996; 93: 295–300.

 10 Salomone, T, Tosi, P & Di Battista, N et al. Impaired alveolar gas exchange in acute pancreatitis. Dig Dis Sci 2002; 47: 2025–8. PubMed CrossRef

 11 Hietaranta, A, Kemppainen, E & Puolakkainen, P et al. Extracellular phospholipases A2 in relation to systemic inflammatory response syndrome (SIRS) and systemic complications in severe acute pancreatitis. Pancreas 1999; 18: 385–91. PubMed

 12 Kingsnorth, A. The role of cytokines in the pathogenesis of acute pancreatitis. Am J Surg 1997; 40: 1–4.

 13 Norman, J. The role of cytokines in the pathogenesis of acute pancreatitis. Am J Surg 1998; 175: 76–83. PubMed CrossRef

 14 Brivet, F, Emilie, D & Galanaud, P et al. Pro- and anti-inflammatory cytokines during acute severe pancreatitis: an early and sustained response, although unpredictable of death. Parisian Study Group on Acute Pancreatitis. Crit Care Med 1999; 27: 749–55. PubMed

 15 Hartwig, W, Werner, J, Uhl, W & Buchler, MW. Management of infection in acute pancreatitis. J Hepatobiliary Pancreat Surg 2002; 9: 423–8. PubMed CrossRef

 16 Banks, PA, Gerzof, SG & Langevin, RE et al. CT-guided aspiration of suspected pancreatic infection: bacteriology and clinical outcome. Int J Pancreatol 1995; 18: 256–70.

17 Takacs, T, Hajnal, F & Nemeth, J et al. Stimulated gastrointestinal hormone release and gallbladder contraction during continuous jejunal feeding in patients with pancreatic pseudocyst is inhibited by octreotide. Inter J Pancreatol 2000; 28: 215–20.

18 Windsor, ACJ, Kanwar, S & Li, AGK et al. Compared with parenteral nutrition, enteral feeding attenuates the acute phase response and improves disease severity in acute pancreatitis. Gut 1998; 42: 431–5. PubMed

 19 Johnson, CD, Imrie, CW & McMahon, MJ et al. Double blind, randomized, placebo controlled study of a platelet activating factor antagonist, lexipafant, in the treatment and prevention of organ failure in predicted severe acute pancreatitis. Gut 2001; 48: 62–9. PubMed CrossRef

 20 Isenmann, R, Schwarz, M, Rau, B, Trautmann, M, Schobr, W & Beger, HG. Characteristics of infection with candida species in patients with necrotizing pancreatitis. World J Surg 2002; 26: 372–6. PubMed CrossRef

 21 Blum, T, Maisonneuve, P, Lowenfels, AB & Lankisch, PG. Fatal outcome in acute pancreatitis: its occurrence and early prediction. Pancreatology 2001; 1: 237–41. PubMed

 22 Ethridge, RT, Chung, DH & Slogoff, M et al. Cyclooxygenase-2 gene disruption attenuates the severity of acute pancreatitis and pancreatitis-associated lung injury. Gastroenterology 2002; 123: 1311–22. PubMed CrossRef

 23 Pezzilli, R, Billi, P, Miniero, R & Barakat, B. Serum interleukin-10 in human acute pancreatitis. Dig Dis Sci 1997; 42: 1469–72. PubMed CrossRef

 24 Chen, CC, Wang, SS, Lu, RH, Chang, FY & Lee, SP. Serum interleukin 10 and interleukin 11 in patients with acute pancreatitis. Gut 1999; 45: 895–9. PubMed

 25 Eskdale, J, Peat, J, Gallagher, GE, Imrie, CW & McKay, CJ. Fine genomic mapping implicates IL-10 as a severity gene in acute pancreatitis. Gastroenterology 2001; 120: A24.

 26 Dumot, JA, Conwell, DL & Zuccaro, C et al. A randomized, double blind study of interleukin 10 for the prevention of ERCP-induced pancreatitis. Am J Gastroenterol 2001; 96: 2098–102. PubMed CrossRef

 27 Balthazar, EJ. Staging of acute pancreatitis. Radiol Clin North Am, 2002; 40: 1199–209. PubMed

 28 Lau, ST, Simchuk, EJ, Kozarek, RA & Traverso, LW. A pancreatic ductal leak should be sought to direct treatment in patients with acute pancreatitis. Am J Surg 2001; 181: 411–15. PubMed CrossRef

 29 Baron, TH, Harewood, GC, Morgan, DE & Yates, MR. Outcome differences after endoscopic drainage of pancreatic necrosis, acute pancreatic pseudocysts, and chronic pancreatic pseudocysts. Gastrointest Endosc 2002; 56: 7–17. PubMed

 30 Bradley, EL III, Clements, JL Jr & Gonzales, AC. The natural history of pancreatic pseudocysts: a unified concept of management. Am J Surg 1979; 137: 135–41. PubMed

 31 D'Edogo, A & Schein, M. Pancreatic pseudocysts: a proposed classification and its management implications. Br J Surg 1991; 78: 981–4. PubMed

 32 Warshaw, AL, Compton, CC & Lewandrowski, K et al. Cystic tumors of the pancreas: new clinical, radiologic, and pathogenic observations in 67 patients. Am Surg 1990; 212: 432–45.

 33 Levin, MF, Vellet, AD & Bach, DB et al. Peripancreatic fluid collections: vascular structures masquerading as pseudocysts. Can Assoc Radiol J 1992; 43: 267–72. PubMed

 34 Sorgman, JA, Langevin, E & Banks, PA. Urinoma masquerading as pancreatic pseudocyst. Int J Pancreatol 1992; 11: 195–6. PubMed

 35 Sperti, C, Cappellazzo, F & Pasquali, C et al. Cystic neoplasms of the pancreas: problems in differential diagnosis. Am Surg 1993; 59: 740–5. PubMed

 36 Warshaw, AL & Rutledge, PL. Cystic tumors mistaken for pancreatic pseudocysts. Am Surg 1987; 205: 393–8.

 37 Hammond, N, Miller, FH, Sica, GT & Gore, RM. Imaging of cystic diseases of the pancreas. Radiol Clin North Am 2002; 40: 1243–62. PubMed

 38 Hsieh, CH, Tseng, JH & Huang, SF. Co-existence of a huge pseudocyst and mucinous cystadenoma: report of a case and the value of magnetic resonance imaging for differential diagnosis. Eur J Gastroenterol Hepatol 2002; 14: 191–4. PubMed

 39 Fockens, P, Johnson, TG & van Dulleman, HM et al. Endosonography is a prerequisite before endoscopic drainage of pancreatic pseudocysts. Gastrointest Endosc 1996; 43: 516 (A).

 40 Brugge, WR. Role of endoscopic ultrasound in the diagnosis of cystic lesions of the pancreas. Pancreatology 2001; 1 (6): 637–40[Review]. PubMed

 41 Kloppel, G & Kosmahl, M. Cystic lesions and neoplasms of the pancreas: the features are becoming clearer. Pancreatology 2001; 1 (6): 648–55[Review]. PubMed CrossRef

 42 Sedlack, R, Affi, A, Vazquez-Sequeiros, E, Norton, ID, Clain, JE & Wiersema, MJ. Utility of EUS in the evaluation of cystic pancreatic lesions. Gastrointest Endosc 2002; 56 (4): 543–7. PubMed CrossRef

 43 Banks, PA. Practice guidelines in acute pancreatitis. Am J Gastroenterol 1997; 3: 377–86.

 44 O'Malley, VP, Cannon, JP & Postie, RG. Pancreatic pseudocysts: cause, therapy, and results. Am J Surg 1985; 150: 680–2. PubMed

 45 Yeo, CJ, Bastidas, JA & Lynch-Nyhan, A et al. The natural history of pancreatic pseudocysts documented by computed tomography. Surg Gynecol Obstet 1990; 170: 411–17. PubMed

 46 Vitas, GJ & Sarr, MG. Selected management of pancreatic pseudocysts: operative versus expectant management. Surgery 1992; 111: 123–30. PubMed

 47 Kozarek, RA, Ball, TJ, Patterson, DJ, Freeny, PC, Ryan, JA & Traverso, LW. Endoscopic transpapillary therapy for disrupted pancreatic duct and peripancreatic fluid collections. Gastroenterology 1996; 100 (5 Part 1): 1362–70.

 48 Urakami, A, Tsunoda, T, Hayashi, J, Oka, Y & Mizuno, M. Spontaneous fistulization of a pancreatic pseudocyst into the colon and duodenum. Gastrointest Endosc 2002; 55: 949–51. PubMed

 49 Nealon, W & Walser, E. Main pancreatic ductal anatomy can direct choice of modality for treating pancreatic pseudocysts (surgery versus percutaneous drainage). Ann Surg 2002; 235: 751–8. PubMed

 50 Grace, PA & Williamson, RCN. Modern management of pancreatic pseudocysts. Br J Surg 1993; 80: 573–81. PubMed

 51 Smits, ME, Rauws, EA & Tytgat, GN et al. The efficacy of endoscopic treatment of pancreatic pseudocysts. Gastrointest Endosc 1995; 42: 202–7. PubMed

 52 VanSonnenberg, E, Wittich, GR & Casola, G. Complicated pancreatic inflammatory disease: diagnostic and therapeutic role of interventional radiology. Radiology 1985; 155: 355–40.

 53 Howell, DA, Elton, E & Parsons, WG. Endoscopic management of pseudocysts of the pancreas. Gastrointest Endosc Clin N Am 1998; 8 (1): 143–62[Review]. PubMed

 54 Parsons, WG & Howell, DA. (1998). Endoscopic management of pancreatic pseudocysts. In: ERCP and its Applications (ed. Jacobson IM), pp. 193–207. Lippincott-Raven, Philadelphia.

 55 Barthet, M, Sahel, J & Bodlou-Bertel, C et al. Endoscopic transpapillary drainage of pancreatic pseudocysts. Gastrointest Endosc 1995; 42: 208–13. PubMed

 56 Catalano, MF, Geenen, JE & Schmalz, MJ. Treatment of pancreatic pseudocysts ductal communication by transpapillary pancreatic duct endoprosthesis. Gastrointest Endosc 1995; 42: 214–18. PubMed

 57 Mallavarapu, R, Habib, TH, Elton, E & Goldberg, MJ. Resolution of mediastinal pancreatic pseudocysts with transpapillary stent placement. Gastrointest Endosc 2001; 53: 367–70. PubMed CrossRef

 58 Sharma, SS, Bhargawa, N & Govil, A. Endoscopic management of pancreatic pseuodocyst: a long-term follow-up. Endosocpy 2002; 34: 203–7.

 59 Cremer, M, Deviere, J & Engelholm, L. Endoscopic management of cysts and pseudocysts in chronic pancreatitis: long-term follow-up after 7 years of experience. Gastrointest Endosc 1989; 35: 1–9. PubMed

 60 Monkemuller, KE, Baron, TH & Morgan, DE. Transmural drainage of pancreatic fluid collections without electrocautery using the Seldinger technique. Gastrointest Endosc 1998; 48: 195–200. PubMed

 61 Chak, A. Endosonographic-guided therapy of pancreatic pseudocysts. Gastrointest Endosc 2000; 52: S23–S27. PubMed

 62 Grimm, H, Binmoeller, KE & Soehendra, N. Endosonography-guided drainage of a pancreatic pseudocyst. Gastrointest Endosc 1992; 38: 170–1. PubMed

 63 Brand, B, Penaloza-Ramirez, A & Gupta, R et al. New mechanical puncture video echoendoscope: one-step transmural drainage of a pseudocyst. Dig Liver Dis 2002; 34: 133–6. PubMed

 64 Wiersema, MJ. Endosonography-guided cystduodenostomy with a therapeutic ultrasound endoscope. Gastrointest Endosc 1996; 44: 614–17. PubMed

 65 Trias, M, Targarona, EM & Balague, C et al. Intraluminal stapled laparoscopic cystogastrostomy for treatment of pancreatic pseudocysts. Br J Surg 1995; 82: 403. PubMed

 66 Baron, TH & Morgan, DE. Endoscopic transgastric irrigation tube placement via PEG for debridement of organized pancreatic necrosis. Gastrointest Endosc 1999 50 (4): 574–7. PubMed

 67 Brown, A, Baillargeon, JD, Hughes, MD & Banks, PA. Can fluid resuscitation prevent pancreatic necrosis in severe acute pancreatitis? Pancreatology 2002; 2: 104–7. PubMed

 68 Lankisch, PG, Struckmann, K, Assmus, C, Lehnick, D, Maisonneuve, P & Lowenfels, AB. Do we need a computed tomography examination in all patients with acute pancreatitis within 72 h after admission to hospital for the detection of pancreatic necrosis? Scan J Gastroenterol 2002; 36: 432–6. CrossRef

 69 Ashley, SW, Perez, A & Pierce, EA et al. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg 2001; 234: 572–9. PubMed CrossRef

 70 Uhl, W, Warshaw, A & Imrie, C et al. IAP guidelines for the surgical management of acute pancreatitis. Pancreatology 2002; 2: 565–73. PubMed CrossRef

 71 Horvath, KD, Kao, LS, Wherry, KL, Pellegrini, CA & Sinanan, MN. A technique for laparoscopic-assisted percutaneous drainage of infected pancreatic necrosis and pancreatic abscess. Surg Endosc 2001; 15: 1221–5. PubMed CrossRef

 72 Castellanos, G, Pinero, A, Serrano, A & Parrilla, P. Infected pancreatic necrosis: translumbar approach and management with retroperitoneoscopy. Arch Surg 2002; 13: 1060–3.

 73 Bassi, C. Infections in pancreatic inflammatory disease: clinical trials for antibiotic prophylaxis. Pancreatology 2001; 1: 210–12. PubMed

 74 Isenmann, R, Rau, B & Beger, HG. Early severe acute pancreatitis: characteristics of a new subgroup. Pancreas 2001; 22: 274–8. PubMed CrossRef

 75 Howard, TJ & Temple, MB. Prophylactic antibiotics alter the bacteriology of infected necrosis in severe acute pancreatitis. J Am Coll Surg 2002; 195: 759–67. PubMed CrossRef

 76 Hariri, M, Slivka, A & Carr-Locke, DL et al. Pseudocyst drainage predisposes to infection when pancreatic necrosis is unrecognized. Am J Gastroenterol 1994; 89: 1781–4. PubMed

 77 Mahlke, R, Warnecke, B, Lankisch, PG, Elbrechtz, F & Busch, C. A sudden coughing up of foul-smelling sputum: a first sign of a pancreaticobronchial fistula, a severe pulmonary complication in acute pancreatitis. Am J Gastoenterol 2001; 96: 1952–3. CrossRef

 78 Olah, A, Jagy, AS, Racz, I & Gamal, ME. Cardiac tamponade as a complication of pseudocyst in chronic pancreatitis. Hepatogastroenterology, 2002; 49: 594–6.

 79 Howell, DA, Dy, RM, Gerstein, WH, Hanson, BL & Biber, BP. Infected pancreatic pseudocysts with colonic fistula formation successfully managed by endoscopic drainage alone: report of two cases. Am J Gastroenterol 2000; 95: 1822–3.

 80 Costamagna, G, Mutignani, M & Ingrosso, M et al. Endoscopic treatment of postsurgical external pancreatic fistulas. Endoscopy 2001; 33: 317–22. PubMed

 81 Kozarek, RA. Endoscopic therapy of complete and partial pancreatic duct disruptions. Gastrointest Endosc Clin N Am 1998; 8: 39–53[Review]. PubMed

 82 Isbicki, JR, Yekebas, EF & Strate, T et al. Extrahepatic portal hypertension in chronic pancreatitis. Ann Surg 2002; 236: 82–9. PubMed CrossRef

 83 Sawlani, V, Phadke, RV, Baijal, SS, Kumar, S, Roy, S, Gupta, P & Gujral, RB. Arterial complications of pancreatitis and their radiological management. Australas Radiol 1996; 40 (4): 381–6[Review]. PubMed

 84 Marshall, GT, Howell, DA & Hansen, BL et al. Multidisciplinary approach to pseudoaneurysms complicating pancreatic pseudocysts: impact of pretreatment diagnosis. Arch Surg 1996; 131: 278–83. PubMed

 85 Koizumi, J, Inoue, S, Yonekawa, H & Kunieda, T. Hemosuccus pancreaticus: diagnosis with CT and MRI and treatment with transcatheter embolization. Abdom Imaging 2002; 27: 77–81. PubMed CrossRef

 86 Born, LJ, Madura, JA & Lehman, GA. Endoscopic diagnosis of a pancreatic pseudoaneurysm after lateral pancreatiticojejunostomy. Gastrointest Endosc 1999; 49: 382–3. PubMed

 87 Balthazar, EJ. Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 2002; 223: 603–13. PubMed

 88 Flati, G, Andren-Sandberg, A, La Pinta, M, Porowska, B & Carboni, M. Potentially fatal bleeding in acute pancreatitis: pathophysiology, prevention, and treatment. Pancreas 2003; 26: 8–14. PubMed CrossRef

 89 Dasgupta, R, Davies, MJ, Williamson, RC & Jackson, JE. Haemosuccus pancreaticus: treatment by arterial embolization. Clin Radiol 2002; 57: 1021–7. PubMed CrossRef

 90 Elton, EDA, Howell, SM & Amberson Dykes, TA. Combined angiographic and endoscopic management of bleeding pancreatic pseudoaneurysms. Gastrointest Endosc 1997; 46: 544–9. PubMed

 91 Beger, HG, Rau, B & Isenmann, R. Prevention of severe change in acute pancreatitis: prediction and prevention. J Hepatobiliary Pancreat Surg 2001; 8: 140–7. PubMed

 92 Bank, S, Singh, P, Pooran, N & Stark, B. Evaluation of factors that have reduced mortality from acute pancreatitis over the past 20 years. J Clin Gastroenterol 2002; 35: 50–60. PubMed

Copyright © Blackwell Publishing, 2005