Chapter 56 Exocrine Pancreas
The average pancreas weighs between 75 and 125 g and measures 10 to 20 cm. It lies in the retroperitoneum just anterior to the first lumbar vertebrae and is anatomically divided into four portions, the head, neck, body and tail. The head lies to the right of midline within the C loop of the duodenum, immediately anterior to the vena cava at the confluence of the renal veins. The uncinate process extends from the head of the pancreas behind the superior mesenteric vein (SMV) and terminates adjacent to the superior mesenteric artery (SMA). The neck is the short segment of pancreas that immediately overlies the SMV. The body and tail of the pancreas then extend across the midline, anterior to Gerota’s fascia and slightly cephalad, terminating within the splenic hilum (Fig. 56-1).
FIGURE 56-1 Anatomy.
(Netter illustration from www.netterimages.com. © Elsevier Inc. All rights reserved.)
The pancreas is supplied by a complex arterial network arising from the celiac trunk and SMA. The head and uncinate process are supplied by the pancreaticoduodenal arteries (anterior and posterior), which arise from the hepatic artery via the gastroduodenal artery (GDA) superiorly and the SMA inferiorly. The neck, body, and tail receive arterial supply from the splenic arterial system. Several small branches originate from the length of the splenic artery, supplying arterial blood flow to the superior portion of the organ. The dorsal pancreatic artery arises from the splenic artery and courses posterior to the body of the gland to become the inferior pancreatic artery. The inferior pancreatic artery then runs along the inferior border of the pancreas, terminating at its tail.
The venous drainage mimics the arterial supply, with blood flow from the head of the pancreas draining into the anterior and posterior pancreaticoduodenal veins. The posterior superior pancreaticoduodenal vein enters the SMV laterally at the superior border of the neck of the pancreas. The anterosuperior pancreaticoduodenal vein enters the right gastroepiploic vein just prior to its confluence with the SMV at the inferior border of the pancreas. The anterior and posteroinferior pancreaticoduodenal veins enter the SMV along the inferior border of the uncinate process. The remaining body and tail are drained via the splenic venous system.
The exocrine pancreas begins development during the fourth week of gestation. Pluripotent pancreatic epithelial stem cells give rise to exocrine and endocrine cell lines, as well as the intricate pancreatic ductal network. Initially, dorsal and ventral buds appear from the primitive duodenal endoderm (Fig. 56-2A). The dorsal bud typically appears first and ultimately develops into the superior head, neck, body, and tail of the mature pancreas. The ventral bud develops as part of the hepatic diverticulum and maintains communication with the biliary tree throughout development. The ventral bud will become the inferior part of the head and uncinate process of the gland. Between the fourth and eighth week, the ventral bud rotates posteriorly in a clockwise fashion to fuse with the dorsal bud (see Fig. 56-2B). At approximately 8 weeks’ gestation, the dorsal and ventral buds are fused (see Fig. 56-2C).
The initiation of pancreas bud formation and differentiation of the ventral bud from the hepatic-biliary fates is dependent on the expression of pancreatic duodenal homeobox 1 (PDX1) protein and pancreas-specific transcription factor 1 (PTF1). In the absence of PDX1 expression in mice, pancreatic agenesis occurs, indicating its importance in the early phases of organogenesis. PTF1 expression is first detectable shortly after PDX1 in cells of the early endoderm, which will become the dorsal and ventral pancreas. By lineage analysis, 95% of acinar cells express PTF1. In PTF1 null mice, acini do not form. The notch signaling pathway is also critical to duct and acinar differentiation. In the absence of notch signaling, embryonic cells commit to endocrine lineage, suggesting that notch signaling is vital to exocrine differentiation. In addition to PDX1, PTF1, and notch signaling, complex interactions between mesenchymal growth factors such as transforming growth factor-β (TGF-β) and other signaling pathways, including hedgehog and Wnt, seem to play critical roles in pancreas development.1 The precise interactions that lead to normal organogenesis continue to be defined. Table 56-1 summarizes the factors and pathways that affect pancreas development.
|PDX1||Critical role in exocrine differentiation; knockout mice develop primitive pancreatic buds, but agenesis of the organ.|
|PTF1||Coexpression with PDX1 determines progenitor cells to pancreatic fate.|
|Notch signaling pathway||Suppresses endocrine differentiation, promoting exocrine development.|
|Hedgehog signaling pathway||Inhibition of hedgehog in PDX1-positive cells leads to initiation of endoderm differentiation into pancreas lineage.|
|Wnt signaling pathway||Complex Wnt signaling is important in all aspects of pancreas development; lack of Wnt signaling results in absence of acinar tissue.|
During normal organogenesis, the dorsal and ventral buds most commonly fuse to form a common duct, which enters the duodenum along with the common bile duct via the ampulla of Vater. Failure of the dorsal and ventral ducts to fuse during embryogenesis leads to pancreas divisum, a condition identified by a ventral pancreatic duct and common bile duct, which enter the duodenum via a major papilla, whereas a dorsal pancreatic duct enters through a minor papilla, which is slightly proximal (Fig. 56-3). Because most pancreatic exocrine secretions exit via the dorsal duct, pancreas divisum can lead to a condition of partial obstruction caused by a small minor papilla, leading to chronic backpressure in the duct. This relative outflow obstruction has been implicated in the development of relapsing acute or chronic pancreatitis. Although 10% of the population is affected by pancreas divisum, only rarely do affected individuals develop pancreatitis.
Annular pancreas results from aberrant migration of the ventral pancreas bud, which leads to circumferential or near-circumferential pancreas tissue surrounding the second portion of the duodenum. This abnormality may be associated with other congenital defects, including Down syndrome, malrotation, intestinal atresia, and cardiac malformations. If symptoms of obstruction occur, surgical bypass via duodenojejunostomy is performed.
Ectopic pancreas may arise anywhere along the primitive foregut, but is most common in the stomach, duodenum, and Meckel’s diverticulum. Clinically, ectopic nodules may result in bowel obstruction caused by intussusception, bleeding, or ulceration. They can sometimes be found incidentally as firm yellow nodules that arise from the submucosa. Although there have been rare case reports of adenocarcinoma arising in ectopic pancreas tissue, resection is not necessary unless symptoms occur.
The human pancreas is a complex gland, with endocrine and exocrine functions. It is mainly composed of acinar cells (85% of the gland) and islets cells (2%) embedded in a complex extracellular matrix, which comprises 10% of the gland. The remaining 3% to 4% of the gland is comprised of the epithelial duct system and blood vessels.
The main function of the exocrine pancreas is to provide the vast majority of the enzymes needed for alimentary digestion. Acinar cells synthesize many enzymes (proteases) that digest food proteins such as trypsin, chymotrypsin, carboxypeptidase, and elastase. Under physiologic conditions, acinar cells synthesize these proteases as inactive proenzymes that are stored as intracellular zymogen granules. With stimulation of the pancreas, these proenzymes are secreted into the pancreatic duct and eventually the duodenal lumen. The duodenal mucosa synthesizes and secretes enterokinase, which is the critical enzyme in the enzymatic activation of trypsin from trypsinogen.2 Trypsin also plays an important role in protein digestion by propagating pancreatic enzyme activation through autoactivation of trypsinogen and other proenzymes, such as chymotrypsinogen, procarboxypeptidase, and proelastase. Figure 56-4 summarizes the mechanisms of pancreatic exocrine secretion.
FIGURE 56-4 Physiology of the secretion of pancreatic enzymes. The presence of peptides and fatty acids from food triggers the release of CCK. CCK induces the release of pancreatic enzymes into the duodenal lumen. Conversely, S cells located in the duodenum release secretin in response to the acidification of the duodenum. Secretin induces the secretion of HCO3− from pancreatic cells into the duodenum.
In addition to protease production, acinar cells also produce pancreatic amylase and lipase, also known as glycerol ester hydrolase, as active enzymes. With the exception of cellulose, pancreatic amylase hydrolyzes major polysaccharides into small oligosaccharides, which can be further digested by the oligosaccharidases present in the duodenal and jejunal epithelium. Pancreatic lipase hydrolyzes ingested fats into free fatty acids and 2-monoglycerides. In addition to pancreatic lipase, acinar cells produce other enzymes that digest fat, but they are secreted as proenzymes, like the proteases previously mentioned. These include colipase, cholesterol ester hydrolase, and phospholipase A2. The main function of colipase is to increase the activity of pancreatic lipase. Cholesterol esters are cleaved by cholesterol ester hydrolase into free cholesterol and one fatty acid, phospholipase A2 hydrolyzes phospholipids, and pancreatic acinar cells also secrete deoxyribonuclease and ribonuclease. These are enzymes required for the hydrolysis of DNA and RNA, respectively.
Pancreatic enzymes are inactive inside acinar cells because they are synthesized and stored as inactive enzymes. In addition to this autoprotective mechanism, acinar cells synthesize pancreatic secretory trypsin inhibitor, which also protects acinar cells from autodigestion because it counteracts premature activation of trypsinogen inside acinar cells. Pancreatic secretory trypsin inhibitor is encoded by serine protease inhibitor Kazal type 1 (SPINK-1) gene. SPINK-1 gene mutations are associated with the development of chronic pancreatitis, especially in childhood.
The primary function of pancreatic duct cells is to provide the water and electrolytes required to dilute and deliver the enzymes synthesized by acinar cells. Although the concentrations of sodium and potassium are similar to their respective concentration in plasma, the concentrations of bicarbonate and chloride vary significantly, according to the secretion phase.
The mechanism responsible of the secretion of bicarbonate was first described in 1988 based on in vitro studies. According to this model, extracellular CO2 diffuses across the basolateral membrane of ductal cells. Once CO2 is inside pancreatic duct cells, it is hydrated by intracellular carbonic anhydrase; as a result of this reaction, HCO3− and H+ are generated. The apical membrane of pancreatic duct cells contains an anion exchanger that secretes intracellular HCO3− into the lumen of the cell and favors the exchange of luminal Cl− inside the ductal epithelium. Recent studies have shown that this exchanger interacts with the cystic fibrosis transmembrane conductance regulator (CFTR). This may correlate with the inability of patients with cystic fibrosis to secrete water and bicarbonate. Although the nature of this exchanger has not been completely elucidated, it is possible that this anion exchanger is an SLC26 family member. This family contains different anion exchangers that transport monovalent and divalent anions, such as Cl− and HCO3−. Some of these exchangers are known to interact with CFTR.
In addition to HCO3−, CO2 hydration also generates H+ ions, which are secreted by Na+ and H+ exchangers present in the basolateral membrane of ductal cells. These exchangers belong to the SLC9 gene family. The main function of these exchangers is to maintain the intracellular pH within a physiologic range. In addition, the basolateral membrane of duct cells contains multiple Na+,K+-ATPases that provide the primary force that drives HCO3− secretion; the Na+,K+-ATPase maintains the Na+ gradient used to extrude H+ as well. Finally, K+ channels present in the basolateral membrane of acinar cells maintain the membrane potential to allow recirculation of K+ ions brought by the Na+,K+ pump inside the cell. Figure 56-5 illustrates HCO3− secretion inside pancreatic duct cells.
(From Steward MC, Ishiguro H, Case RM: Mechanisms of bicarbonate secretion in the pancreatic duct. Annu Rev Physiol 67:377–409, 2005.)
Once the HCO3− secreted by pancreatic duct cells reaches the duodenal lumen, it neutralizes the hydrochloric acid secreted by parietal cells. Pancreatic enzymes are inactivated at a low pH; therefore, pancreatic bicarbonate provides an optimal pH for acinar cell enzyme function. The optimal pH for the function of chymotrypsin and trypsin is from 8.0 to 9.0, for amylase the optimal pH is 7.0, and for lipase it is from 7.0 to 9.0.
Pancreatic exocrine secretion occurs during the interdigestive state and after the ingestion of food, which is also known as the digestive state. The same phases of secretion that have been identified in the stomach during the digestive state have been also described in pancreatic secretion. The first phase is the cephalic phase, in which the pancreas is stimulated by the vagus nerve in response to the sight, smell, or taste of food. This phase is generally mediated by the release of acetylcholine at the terminal endings of postganglionic fibers. The main effect of acetylcholine is to induce acinar cell secretion of enzymes. This phase accounts for 20% to 25% of the daily secretion of pancreatic juice.
The second phase of pancreatic secretion is known as the gastric phase. It is mediated by vagovagal reflexes triggered by gastric distention after the ingestion of food. These reflexes induce acinar cell secretion. It accounts for 10% of the pancreatic juice produced daily.
The most important phase of pancreatic secretion is the intestinal phase, which accounts for 65% to 70% of the total secretion of pancreatic juice. It is mediated by secretin and cholecystokinin (CCK). Acidification of the duodenal lumen induces the release of secretin by S cells. Secretin was the first polypeptide hormone identified more than 100 years ago. It is the most important mediator of the secretion of water, bicarbonate, and other electrolytes into the duodenum. Secretin receptors are located in the basolateral membrane of all pancreatic duct cells but cannot be identified in other pancreatic components, such as islet cells, blood vessels, or extracellular matrix. Secretin receptors are members of the G protein–coupled receptor superfamily. The most important effect of secretin stimulation is an increase of intracellular cyclic adenosine monophosphate (cAMP), which activates the HCO3−-Cl− anion exchanger in the apical membrane of pancreatic duct cells. It also increases the activity of the enzyme carbonic anhydrase, the excretion of H+ outside the duct cell, and the activity of the CFTR.
The presence of lipid, protein, and carbohydrates inside the duodenum induces the secretion of CCK-releasing factor and monitor peptide. Both peptides induce release of CCK by I cells present in the duodenal mucosa. Whereas secretin is the main mediator of the secretion of water and bicarbonate in the intestinal phase, CCK is the main mediator of the secretion of pancreatic enzymes. CCK exerts a number of effects:
The incidence of acute pancreatitis (AP) has increased during the past 20 years. AP is responsible for more than 300,000 hospital admissions annually in the United States. Most patients develop a mild and self-limited course; however, 10% to 20% of patients have a rapidly progressive inflammatory response associated with prolonged length of hospital stay and significant morbidity and mortality. Patients with mild pancreatitis have a mortality rate of less than 1% but, in severe pancreatitis, this increases up to 10% to 30%.3 The most common cause of death in this group of patients is multiorgan dysfunction syndrome. Mortality in pancreatitis has a bimodal distribution; in the first 2 weeks, also known as the early phase, the multiorgan dysfunction syndrome is the final result of an intense inflammatory cascade triggered initially by pancreatic inflammation. Mortality after 2 weeks, also known as the late period, is often caused by septic complications.4
The exact mechanism whereby predisposing factors such as ethanol and gallstones produce pancreatitis is not completely known. Most researchers believe that AP is the final result of abnormal pancreatic enzyme activation inside acinar cells. Immunolocalization studies have shown that after 15 minutes of pancreatic injury, both zymogen granules and lysosomes colocalize inside the acinar cells. The fact that zymogen and lysosome colocalization occurs before amylase level elevation, pancreatic edema, and other markers of pancreatitis are evident suggests that colocalization is an early step in the pathophysiology and not a consequence of pancreatitis. In addition, the inflammatory response seen in AP can be prevented if acinar cells are pretreated with cathepsin B inhibitors. In vivo studies have also shown that cathepsin B knockout mice have a significant decrease in the severity of pancreatitis.2
Intra-acinar pancreatic enzyme activation induces autodigestion of normal pancreatic parenchyma. In response to this initial insult, acinar cells release proinflammatory cytokines, such as tumor necrosis factor-α (TNF-α), interleukins (IL)-1, -2, and -6, and anti-inflammatory mediators such as IL-10 and IL-1 receptor antagonist. These mediators do not initiate pancreatic injury but propagate the response locally and systemically. As a result, TNF-α, IL-1, and IL-7, neutrophils, and macrophages are recruited into the pancreatic parenchyma and cause the release of more TNF- α, IL-1, IL-6, reactive oxygen metabolites, prostaglandins, platelet-activating factor, and leukotrienes. The local inflammatory response further aggravates the pancreatitis because it increases the permeability and damages the microcirculation of the pancreas. In severe cases, the inflammatory response causes local hemorrhage and pancreatic necrosis. In addition, some of the inflammatory mediators released by neutrophils aggravate the pancreatic injury because they cause pancreatic enzyme activation.5
The inflammatory cascade is self-limited in approximately 80% to 90% of patients. However, in the remaining patients, a vicious cycle of recurring pancreatic injury and local and systemic inflammatory reaction persists. In a small number of patients, there is a massive release of inflammatory mediators to the systemic circulation. Active neutrophils mediate acute lung injury and induce the adult distress respiratory syndrome frequently seen in patients with severe pancreatitis. The mortality seen in the early phase of pancreatitis is the result of this persistent inflammatory response. A summary of the inflammatory cascade seen in AP is shown in Figure 56-6.
FIGURE 56-6 Pathophysiology of severe acute pancreatitis. The local injury induces the release of TNF-α and IL-1. Both cytokines produce further pancreatic injury and amplify the inflammatory response by inducing the release of other inflammatory mediators, which cause distant organ injury. This abnormal inflammatory response is responsible for the mortality seen during the early phase of acute pancreatitis.
Gallstones and ethanol abuse account for 70% to 80% of AP cases. In pediatric patients, abdominal blunt trauma and systemic diseases are the two most common conditions that lead to pancreatitis. Autoimmune and drug-induced pancreatitis should be a differential diagnosis in patients with rheumatologic conditions such as systemic lupus erythematosus and Sjögren’s syndrome.
Gallstone pancreatitis is the most common cause of AP in the West. It accounts for 40% of U.S. cases. The overall incidence of AP in patients with symptomatic gallstone disease is 3% to 8%. It is seen more frequently in women between 50 and 70 years of age. The exact mechanism that triggers pancreatic injury has not been completely understood, but two theories have been proposed.6 In the obstructive theory, pancreatic injury is the result of excessive pressure inside the pancreatic duct. This increased intraductal pressure is the result of continuous secretion of pancreatic juice in the presence of pancreatic duct obstruction. The second, or reflux, theory proposes that stones become impacted in the ampulla of Vater and form a common channel that allows bile salt reflux into the pancreas. Animal models have shown that bile salts cause direct acinar cell necrosis because they increase the concentration of calcium in the cytoplasm; however, this has never been proven in humans.2
Excessive ethanol consumption is the second most common cause of AP worldwide. It accounts for 35% of cases and is more prevalent in young men (30 to 45 years of age) than in women. However, only 5% to 10% of patients who drink alcohol develop AP. Factors that contribute to ethanol-induces pancreatitis include heavy ethanol abuse (>100 g/day for at least 5 years), smoking, and genetic predisposition. As compared with nonsmokers, the relative risk of alcohol-induced pancreatitis in smokers is 4.9.7
Alcohol has a number of deleterious effects in the pancreas. It triggers proinflammatory pathways such as nuclear factor κB (NF-κB), which increase the production of TNF-α and IL-1. It also increases the expression and activity of caspases. Caspases are proteases that mediate apoptosis. In addition, alcohol decreases pancreatic perfusion, induces sphincter of Oddi spasm, and obstructs pancreatic ducts through the precipitation of proteins inside the ducts.8
Pancreas divisum is an anatomic variation present in 10% of the population. Its association with AP is controversial. Patients with this variation have a 5% to 10% lifetime risk of developing AP caused by relative outflow obstruction through the minor papilla. Endoscopic retrograde cholangiopancreatography (ERCP) with minor papillotomy and stenting may be beneficial for such patients.
Infrequent anatomic obstructions that have been associated with AP include Ascaris lumbricoides infection and annular pancreas. Although pancreatic cancer is not uncommon, patients with pancreatic cancer usually do not develop AP.
AP is the most common complication after ERCP, occurring in up to 5% of patients. AP occurs more frequently in patients who have undergone therapeutic procedures as compared with diagnostic procedures. It is also more common in patients who have had multiple attempts of cannulation, sphincter of Oddi dysfunction, and abnormal visualization of the secondary pancreatic ducts after contrast injection. The clinical course is mild in 90% to 95% of patients.8
Up to 2% of AP cases are caused by medications. The most common agents include sulfonamides, metronidazole, erythromycin, tetracyclines, didanosine, thiazides, furosemide, 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitors (statins), azathioprine, 6-mercaptopurine, 5-aminosalicylic acid, sulfasalazine, valproic acid, and acetaminophen. More recently, antiretroviral agents used for the treatment of AIDS have been implicated in AP.
Hypertriglyceridemia and hypercalcemia can also lead to pancreatic damage. Direct pancreatic injury can be induced by triglyceride metabolites. It is more common in patients with type I, II, or V hyperlipidemia. It should be suspected in patients with a triglyceride level higher than 1000 mg/dL. A triglyceride level higher than 2000 mg/dL confirms the diagnosis. Hypertriglyceridemia secondary to hypothyroidism, diabetes mellitus, and alcohol does not typically induce AP.
Hypercalcemia is postulated to induce pancreatic injury through the activation of trypsinogen to trypsin and intraductal precipitation of calcium, leading to ductal obstruction and subsequent attacks of pancreatitis. Approximately 1.5% to 13% of patients with primary hyperparathyroidism develop AP.8
Blunt and penetrating abdominal trauma can be associated with AP in 0.2% and 1% of cases, respectively. Prolonged intraoperative hypotension and excessive pancreatic manipulation during abdominal surgery can also result in AP. Pancreatic ischemia in association with acute pancreatic inflammation can develop after splenic artery embolization. Other rare causes include scorpion venom stings and perforated duodenal ulcers.
The cardinal symptom of AP is epigastric and/or periumbilical pain that radiates to the back. Up to 90% of patients have nausea and/or vomiting that typically does not relieve the pain. The nature of the pain is constant; therefore, if the pain disappears or decreases, another diagnosis should be considered.
The physical examination of the abdomen varies according to the severity of the disease. With mild pancreatitis, the physical examination of the abdomen may be normal or reveal only mild epigastric tenderness. Significant abdominal distention, associated with generalized rebound and abdominal rigidity, is present in severe pancreatitis. Note the nature of the pain described by the patient may not correlate with the physical examination or the degree of pancreatic inflammation.
Rare findings include flank and periumbilical ecchymosis (Grey Turner and Cullen’s signs, respectively). Both are indicative of retroperitoneal bleeding associated with severe pancreatitis. Patients with concomitant choledocholithiasis or significant edema in the head of the pancreas that compresses the intrapancreatic portion of the common bile duct can present with jaundice. Dullness to percussion and decreased breathing sounds in the left or, less commonly, in the right hemithorax suggest pleural effusion secondary to AP.
The cornerstone of the diagnosis of AP are the clinical findings plus an elevation of pancreatic enzyme levels in the plasma. A threefold or higher elevation of amylase and lipase levels confirms the diagnosis. Amylase’s serum half-life is shorter as compared with lipase. In patients who do not present to the emergency department within the first 24 or 48 hours after the onset of symptoms, determination of lipase levels is a more sensitive indicator to establish the diagnosis. Lipase is also a more specific marker of AP because serum amylase levels can be elevated in a number of conditions, such as such as peptic ulcer disease, mesenteric ischemia, salpingitis, and macroamylasemia.
Patients with AP are typically hyperglycemic; they can also have leukocytosis and abnormal elevation of liver enzyme levels. The elevation of alanine aminotransferase levels in the serum in the context of AP confirmed by high pancreatic enzyme levels has a positive predictive value of 95% in the diagnosis of acute biliary pancreatitis.6
Although simple abdominal radiographs are not useful to diagnose pancreatitis, they can help rule out other conditions, such as perforated ulcer disease. Nonspecific findings in patients with AP include air-fluid levels suggestive of ileus, cutoff colon sign as a result of colonic spasm at the splenic flexure, and widening of the duodenal C loop caused by severe pancreatic head edema.
The usefulness of ultrasound to diagnose pancreatitis is limited by intra-abdominal fat and increased intestinal gas as a result of the ileus. Nevertheless, this test should always be ordered in patients with AP because of its high sensitivity (95%) in diagnosing gallstones. Combined elevation of liver transaminase and pancreatic enzyme levels, and the presence of gallstones on ultrasound, have an even higher sensitivity (97%) and specificity (100%) for diagnosing acute biliary pancreatitis.
Contrast-enhanced computed tomography (CT) is currently the best modality to evaluate the pancreas, especially if the study is performed using a multidetector CT scanner. The most valuable contrast phase to evaluate the pancreatic parenchyma is the portal venous phase (65 to 70 seconds after contrast injection), which allows evaluation of the viability of the pancreatic parenchyma amount of peripancreatic inflammation and presence of intra-abdominal free air or fluid collections. Noncontrast CT scanning may also be of value in the setting of renal failure by identifying fluid collections and/or extraluminal air.9
Abdominal magnetic resonance imaging (MRI) is also useful to evaluate the extent of necrosis, inflammation, and presence of free fluid. However, its cost and availability, and the fact that patients requiring imaging are critically ill and need to be in intensive care units, limit its applicability in the acute phase. Although magnetic resonance cholangiopancreatography (MRCP) is not indicated in the acute setting of AP, it has an important role in the evaluation of patients with unexplained or recurrent pancreatitis because it allows complete visualization of the biliary and pancreatic duct anatomy. In addition, IV administration of secretin increases pancreatic duct secretion, which causes a transient distention of the pancreatic duct. For example, secretin MRCP is useful in patients with AP and no evidence of a predisposing condition to rule out pancreas divisum, intraductal papillary mucinous neoplasm (IPMN), or the presence of a small tumor in the pancreatic duct.9
In the setting of gallstone pancreatitis, endoscopic ultrasound (EUS) may play an important role in the evaluation of persistent choledocholithiasis. Several studies have shown that routine ERCP for suspected gallstone pancreatitis reveals no evidence of persistent obstruction in most cases and may actually worsen symptoms because of manipulation of the gland. EUS has been proven to be sensitive for identifying choledocholithiasis; it allows for examination of the biliary tree and pancreas with no risk of worsening the pancreatitis. In patients in whom persistent choledocholithiasis is confirmed by EUS, ERCP can be used selectively as a therapeutic measure.
The earliest scoring system designed to evaluate the severity of AP was introduced by Ranson and colleagues in 1974.10 It predicts the severity of the disease based on 11 parameters obtained at the time of admission and/or 48 hours later. The mortality rate of AP directly correlates with the number of parameters that are positive. Severe pancreatitis is diagnosed if three or more of the Ranson criteria are fulfilled. The main disadvantage is that it does not predict the severity of disease at the time of the admission because six parameters are only assessed after 48 hours of admission. Ranson’s score has a low positive predictive value (50%) and high negative predictive value (90%). Therefore, it is mainly used to rule out severe pancreatitis or predict the risk of mortality.11 The original scoring symptom designed to predict the severity of the disease and its modification for acute biliary pancreatitis are shown in Boxes 56-1 and 56-2.
Box 56-1 Ranson’s Prognostic Criteria for Nongallstone Pancreatitis
Box 56-2 Ranson’s Prognostic Criteria for Gallstone Pancreatitis
AP severity can also be addressed using the Acute Physiology and Chronic Health Evaluation (APACHE II) score. Based on the patient’s age, previous health status, and 12 routine physiologic measurements, APACHE II provides a general measure of the severity of disease. An APACHE II score of 8 or higher defines severe pancreatitis. The main advantage is that it can be used on admission and repeated at any time. However, it is complex, not specific for AP, and based on the patient’s age, which easily upgrades the AP severity score. APACHE II has a positive predictive value of 43% and a negative predictive value of 89%.11
Using imaging characteristics, Balthazar and associates12 have established the CT severity index. This index correlates CT findings with the patient’s outcome. The CT severity index is shown in Table 56-2.
|Focal or diffuse pancreatic enlargement||1|
|Intrinsic pancreatic alterations with peripancreatic fat inflammatory changes||2|
|Single fluid collection/or phlegmon||3|
|Two or more fluid collections or gas, in or adjacent to the pancreas||4|
CTSI, 0-3, mortality 3%, morbidity 8%; CTSI, 4-6, mortality 6%, morbidity 35%; CTSI, 7-10, mortality 17%, morbidity 92%.
In 1992, the International Symposium on Acute Pancreatitis defined severe pancreatitis as the presence of local pancreatic complications (necrosis, abscess, or pseudocyst) or any evidence of organ failure. Severe pancreatitis is diagnosed if there is any evidence of organ failure or a local pancreatic complication (Box 56-3).
Box 56-3 Atlanta’s Criteria for Acute Pancreatitis
C-reactive protein (CRP) is an inflammatory marker that peaks 48 to 72 hours after the onset of pancreatitis and correlates with the severity of the disease. A CRP level 150 mg/mL or higher defines severe pancreatitis. The major limitation is that it cannot be used on admission; the sensitivity of the assay decreases if CRP levels are measured within 48 hours after the onset of symptoms. In addition to CRP, a number of studies have shown other biochemical markers (e.g., serum levels of procalcitonin, IL-6, IL-1, elastase) that correlate with the severity of the disease. However, their main limitation is their cost and that they are not widely available.
Regardless of the cause or the severity of the disease, the cornerstone of the treatment of chronic pancreatitis is aggressive fluid resuscitation using isotonic crystalloid solution. The rate of administration should be individualized and adjusted based on age, comorbidities, vital signs, mental status, skin turgor, and urine output. Patients who do not respond to initial fluid resuscitation or have significant renal, cardiac or respiratory comorbidities often require invasive monitoring with central venous access and a Foley catheter.
In addition to fluid resuscitation, patients with AP require continuous pulse oximetry because one of the most common systemic complications of AP is hypoxemia caused by the acute lung injury associated with this disease. Patients should receive supplementary oxygen to maintain arterial saturation above 95%.
It is also essential to provide effective analgesia. Narcotics are usually preferred, especially morphine. One of the physiologic effects described after systemic administration of morphine is an increase in tone in the sphincter of Oddi; however, there is no evidence that narcotics exert a negative impact in the outcome of patients with AP.
Nutritional support is vital in the treatment of AP. Oral feeding may be impossible because of persistent ileus, pain, or intubation. In addition, 20% of patients with severe AP develop recurrent pain shortly after the oral route has been restarted. The main options to provide this nutritional support are enteral feeding and total parenteral nutrition (TPN). Although there is no difference in the mortality rate between both types of nutrition, enteral nutrition is associated with less infectious complications and reduces the need for pancreatic surgery. Although TPN provides most nutritional requirements, it is associated with mucosal atrophy, decreased intestinal blood flow, increased risk of bacterial overgrowth in the small bowel, antegrade colonization with colonic bacteria, and increased bacterial translocation. In addition, patients with TPN have more central line infections and metabolic complications (e.g., hyperglycemia, electrolyte imbalance). Whenever possible, enteral nutrition should be used, rather than TPN.
Given the significant increase in mortality associated with septic complications in severe pancreatitis, a number of physicians advocated the use of prophylactic antibiotics in the 1970s. Recent meta-analyses and systematic reviews that have evaluated multiple randomized control trials have proven that prophylactic antibiotics do not decrease the frequency of surgical intervention, infected necrosis, or mortality in patients with severe pancreatitis. In addition, they are associated with gram-positive cocci infection such as by Staphylococcus aureus, and Candida infection, which is seen in 5% to 15% of patients.13
Early ERCP, with or without sphincterotomy, was initially advocated to reduce the severity of pancreatitis because the obstructive theory of AP states that pancreatic injury is the result of pancreatic duct obstruction. However, three randomized trials have demonstrated that ERCP is only beneficial for patients with severe acute biliary pancreatitis. Routine use of ERCP is not indicated for patients with mild pancreatitis because the bile duct obstruction is usually transient and resolves within 48 hours after the onset of symptoms. In addition to severe acute biliary pancreatitis, ERCP is indicated for patients who develop cholangitis and those with persistent bile duct obstruction demonstrated by other imaging modalities, such as EUS. Finally, in older patients with poor performance status or severe comorbidities that preclude surgery, ERCP with sphincterotomy is a safe alternative to prevent recurrent biliary pancreatitis.
In the absence of definitive treatment, 30% of patients with acute biliary pancreatitis will have recurrent disease. With the exception of older patients and those with poor performance status, laparoscopic cholecystectomy is indicated for all patients with mild acute biliary pancreatitis.3 Studies have shown that early laparoscopic cholecystectomy, defined as laparoscopic cholecystectomy during the initial admission to the hospital, is a safe procedure that decreases recurrence of the disease.6 Choledocolithiasis can be excluded via intraoperative cholangiography, ERCP, or laparoscopic common bile duct exploration.
For patients with severe pancreatitis, early surgery may increase the morbidity and length of stay.14 Current recommendations suggest conservative treatment for at least 6 weeks before laparoscopic cholecystectomy is attempted in this setting. This approach has significantly decreased morbidity.6
The presence of acute abdominal fluid during an episode of AP has been described in 30% to 57% of patients.3 In contrast to pseudocysts and cystic neoplasias of the pancreas, fluid collections are not surrounded or encased by epithelium or fibrotic capsule. Treatment is supportive because most fluid collections will be spontaneously reabsorbed by the peritoneum. The presence of fever, elevated white blood cell (WBC) count, and abdominal pain suggest infection of this fluid and percutaneous aspiration is confirmatory. Percutaneous drainage and IV administration of antibiotics should be instituted if infection is present.
Pancreatic necrosis is the presence of nonviable pancreatic parenchyma or peripancreatic fat; it can present as a focal area or diffuse involvement of the gland. Contrast-enhanced CT is the most reliable technique to diagnose pancreatic necrosis. It is typically seen as areas of low attenuation (<40 to 50 HU) after the injection of IV contrast. Normal parenchyma usually has a density of 100 to 150 HU.9 Up to 20% of patients with AP develop pancreatic necrosis. It is important to identify and provide proper treatment of this complication because most patients who develop multiorgan failure have necrotizing pancreatitis; pancreatic necrosis has been documented in up to 80% of the autopsies of patients who died after an episode of AP.4
The main complication of pancreatic necrosis is infection. The risk is directly related to the amount of necrosis; in patients with pancreatic necrosis involving less than 30% of the gland, the risk of infection is 22%. The risk is 37% for patients with pancreatic necrosis that involves 30% to 50% of the gland and up to 46% if more than 70% of the gland is affected.4 This complication is associated with bacterial translocation usually involving enteric flora, such as gram-negative rods (e.g., Escherichia coli, Klebsiella and Pseudomonas spp.) and Enterococcus spp.
Infected pancreatic necrosis should be suspected in patients with prolonged fever, elevated WBC count, or progressive clinical deterioration. Evidence of air within the pancreatic necrosis seen on a CT scan confirms the diagnosis but is a rare finding. If infected necrosis is suspected, fine-needle aspiration (FNA) should be performed. A positive Gram stain and/or culture establish the diagnosis. Although positive cultures are confirmatory, a recent review has demonstrated that despite negative preoperative cultures, 42% of patients with so-called persistent unwellness will have infected necrosis.15 Figure 56-7 illustrates the pathophysiology of pancreatic necrosis infection.
FIGURE 56-7 Pathophysiology of pancreatic necrosis infection. The acute inflammatory injury that occurs during the first 48 to 72 hours causes mucosal ischemia and reperfusion injury. Both effects favor bacterial overgrowth because they alter local immunity. Mucosal ischemia also produces an increase in the permeability of intestinal cells, which is initiated 72 hours after the acute episode but typically peaks 1 week later. These transient episodes of bacteremia are associated with pancreatic necrosis infection. Less frequently, distant sources of infection such as pneumonia, vascular, or urinary tract infection associated with central lines and catheters are associated with bacteremia and pancreatic necrosis. Finally, local contamination after surgery or interventional procedures such as ERCP is responsible for necrosis infection.
Once infection has been demonstrated, IV antibiotics should be given. Because of their penetration into the pancreas and spectrum coverage, carbapenems are the first option of treatment. Alternative therapy includes quinolones, metronidazole, third-generation cephalosporins, and piperacillin.
Definitive treatment for infected pancreatic necrosis is surgical débridement with necrosectomy, closed continuous irrigation, and open packaging (Fig. 56-8). The overall mortality rate after open necrosectomy is 25% to 30%.15 Outcomes are time-dependent, with patients who undergo surgery in the first 14 days having a mortality rate of 75%; those who undergo surgery between 15 and 29 days and after 30 days have mortality rates of 45% and 8%, respectively.16 As a result of the elevated morbidity and mortality rates with open débridement, endoscopic and laparoscopic techniques are being used more often. Both may ultimately provide similar outcomes, with hopes of reducing perioperative morbidity and mortality, although level 1 data are lacking.
FIGURE 56-8 Infected pancreatic necrosis. This 45-year-old man had severe ethanol-induced pancreatitis. Four weeks after the initial episode, the patient developed fever (39.5° C [103° F]), hypotension, and leukocytosis (19,000 cells/mm3). The CT scan documented pancreatic necrosis involving 35% of the gland. After FNA, Gram staining documented the presence of gram-negative rods. The exploratory laparotomy indicated pancreatic necrosis involving mainly the body of the gland (arrow). The patient was treated with necrosectomy, closed drainage, and IV meropenem. Final culture documented the presence of Escherichia coli. The patient was discharged home 56 days after the initial episode.
Pancreatic pseudocysts occur in 5% to 15% of patients who have peripancreatic fluid collections after AP. By definition, the capsule of a pseudocyst is composed of collagen and granulation tissue and it is not lined by epithelium.17 The fibrotic reaction typically requires at least 4 to 8 weeks to develop. Figure 56-9 shows CT scans of a large pseudocyst arising in the tail of the pancreas.
Up to 50% of patients with pancreatic pseudocysts will develop symptoms. The presence of persistent pain, early satiety, nausea, weight loss, and elevated pancreatic enzyme levels in plasma suggest this diagnosis. The diagnosis is corroborated with by CT or MRI. EUS with FNA is indicated for patients in whom the diagnosis of pancreatic pseudocyst is not clear. Characteristic features of pancreatic pseudocysts include high amylase levels associated with the absence of mucin and low carcinoembryonic antigen (CEA) levels.
Observation is indicated for asymptomatic patients because spontaneous regression has been documented in up to 70% of cases; this is particularly true for patients with pseudocysts smaller than 4 cm in diameter, located in the tail, and no evidence of pancreatic duct obstruction or communication with the main pancreatic duct.17 Invasive therapies are indicated for symptomatic patients or when the differentiation between a cystic neoplasm and pseudocyst is not possible. Because most patients are treated with decompressive procedures and not with resection, it is imperative to have a pathologic diagnosis. Surgical drainage has been the traditional approach for pancreatic pseudocysts. However, there is increasing evidence that transgastric and transduodenal endoscopic drainage are safe and effective approaches for patients with pancreatic pseudocysts in close contact (defined as <1 cm) with the stomach and duodenum, respectively. In addition, transpapillary drainage can be attempted in pancreatic pseudocysts communicating with the main pancreatic duct. For patients in whom a pancreatic duct stricture is associated with a pancreatic pseudocyst, endoscopic dilation and stent placement are indicated.
Surgical drainage is indicated for patients with pancreatic pseudocysts that cannot be treated with endoscopic techniques and patients who fail endoscopic treatment. Definitive treatment depends on the location of the cyst. Pancreatic pseudocysts closely attached to the stomach should be treated with a cystgastrostomy. In this procedure, an anterior gastrostomy is performed. Once the pseudocyst is located, it is drained through the posterior wall of the stomach using a linear stapler. The defect in the anterior wall of the stomach is closed in two layers. Pancreatic pseudocysts located in the head of the pancreas that are in close contact with the duodenum are treated with a cystoduodenostomy. Finally, some pseudocysts are not in contact with the stomach or duodenum. The surgical treatment for these patients is a Roux-en-Y cystojejunostomy. Surgical cyst enterostomy is successful in achieving immediate cyst drainage in over 90% of cases. Following initial resolution, recurrent pseudocyst formation may occur in up to 12% of cases during long-term follow-up, depending on the location of the cyst and underlying cause of the disease.17
Complications of pancreatic pseudocysts include bleeding and pancreaticopleural fistula secondary to vascular and pleural erosion, respectively, bile duct and duodenal obstruction, rupture into the abdominal cavity, and infection. Percutaneous drainage is only indicated for septic patients secondary to pseudocyst infection because it has a high incidence of external fistula.
Although very rare, complete disruption of the pancreatic duct can lead to significant accumulation of, fluid. This condition should be suspected in patients who have an episode of AP, develop significant abdominal distention, and have free intra-abdominal fluid. Diagnostic paracentesis typically demonstrates elevated amylase and lipase levels. Treatment consists of abdominal drainage combined with endoscopic placement of a pancreatic stent across the disruption. Failure of this therapy requires surgical treatment; it consists of distal resection and closure of the proximal stump.
Posterior pancreatic duct disruption into the pleural space has been described rarely. Symptoms that suggest this condition include dyspnea, abdominal pain, cough, and chest pain. The diagnosis is confirmed with chest x-ray, thoracentesis, and CT scan. Figure 56-10 demonstrates a large, left-sided pleural effusion caused by a pancreatic-pleural fistula. Amylase levels above 50,000 IU in the pleural fluid confirm the diagnosis. It is more common after alcoholic pancreatitis and, in 70% of patients, is associated with pancreatic pseudocysts. Initial treatment requires chest drainage, parenteral nutritional support, and administration of octreotide. Up to 60% of patients respond to this therapy. Persistent drainage should also be treated with endoscopic sphincterotomy and stent placement. Patients who do not respond to these measures require surgical treatment, similar to that described for pancreatic ascites.