Pancreatic Disease

The degree of severity and outcomes of a patient with acute pancreatitis was described by Ranson et al. (25,26) and included 11 clinical criteria, 5 of which were assessed on admission and 6 after 48 hours (Table 128.2). They are well correlated with morbidity, length of ICU stay, and mortality. The presence of less than 3 of these 11 criteria correlates with a benign course of disease, with 3% mortality rate. The presence of three or more of these parameters on admission or within 48 hours implies a more severe form of pancreatitis, and is associated with high risk of death and major complication. With a single exception, the patients’ age, these criteria are the result of the statistical analysis of 43 parameters. These parameters were gathered retrospectively from 3 overlapping series comprising 450 patients (25,26). In this series, however, only 21% had acute pancreatitis definitively confirmed by surgery or postmortem examination. Nevertheless, the Ranson scale has been largely used since the 1980s in virtually all studies relating to acute pancreatitis.

A more detailed system, the Acute Physiology and Chronic Health Evaluation II (APACHE II) scoring system, has also been extensively used and is easily obtained by web-based calculators (Table 128.2). Values equal or greater than 8 are indicators of severe acute pancreatitis. Although cumbersome, it can be measured in a daily basis as opposed to Ranson criteria. The introduction of the Bedside Index for Severity in Acute Pancreatitis (BISAP) score has allowed a simpler prediction of mortality within the first 24 hours of presentation (Table 128.2) (27). Scores greater than 3 are associated with higher rates of mortality and organ failure. BISAP has the disadvantage that it cannot easily distinguish transient from persistent organ failure, however its prognostic accuracy is similar to those of the other scoring systems (28). Recent evidence suggests that measurement of IL-6 and CRP may play some role in prediction and assessment of the severity of acute pancreatitis (29). Further scores taking into account these values are forthcoming.

TABLE 128.2 Scoring Systems for Prediction of Outcomes in Acute Pancreatitis

Several scoring systems exist based on CT findings as predictors of disease severity. The Modified CT Scan Index is associated with better correlation of severity and clinical outcomes (Table 128.2) (30). This score classifies disease as mild, moderate, and severe with scores 0 to 2, 4 to 6, and 8 to 10, respectively (30). In the clinical setting, not all patients undergo immediate CT scan on presentation, thereby making these scores limited in early management.

TREATMENT OF ACUTE PANCREATITIS

The rationale approach to the therapy for acute pancreatitis includes placing the pancreas “at rest,” supporting the patient’s nutritional and metabolic needs, correcting the acute causes of mortality (i.e., cardiovascular collapse, respiratory insufficiency, and renal failure), and detecting complications of disease that require intervention.

Gland Suppression and Nutrition

Suppression of the secretory function of the pancreas has been attempted by elimination of oral fluids, suppression of acid secretion with various H₂ and antacids and use of anticholinergics and proteolytic enzyme inhibitors. Calcitonin and somatostatin, which are potent inhibitors of pancreatic enzyme secretion, have also been investigated. Although theoretically beneficial, randomized controlled trials (RCTs) have not shown significant improvement.

Historically, early feeding was associated with increased severity of the disease. Bowel rest, with or without parenteral nutrition, had become the standard of care. It is known that acute pancreatitis results in a hypermetabolic, hyperdynamic SIRS that results in a catabolic state (31). Retrospective evidence suggests that enteral nutrition is not only feasible but may be desirable in such patients. Lack of enteral feeding results in atrophy of the gastrointestinal mucosa, bacterial overgrowth, increased gut permeability, and translocation of bacteria or its products into the circulation (32). Enteral feeding also stimulates lysosomal movement to the cell surface, minimizing the intracellular release of pancreatic enzymes and may, in fact, be therapeutic in acute pancreatitis (31). In addition, enteral nutrition reduces the production of proinflammatory mediators that may also have therapeutic potential in such patients. Enteral nutrition is associated with fewer complications by preserving gut integrity and is more cost effective as compared to total parenteral nutrition (TPN). A meta-analysis of 8 randomized trials involving 348 patients confirmed that enteric tube feeding, as compared with TPN, reduced the rate of infections and mortality among patients with severe acute pancreatitis (33).

The data regarding the timing to initiate enteral nutrition have been heavily investigated, however are nowhere near a consensus. Several nonrandomized studies have shown that nasoenteric tube feeding started within 48 hours of admission, as compared after 48 hours, significantly reduced the rate of major infection and mortality rates (34–37). However, a recently published large multicenter RCT including 208 patients with severe pancreatitis, the PYTHON trial, has shown no significant differences between early feeding (within 24 hours) and oral diet initiated 72 hours after presentation in the rates of major infection or mortality rates (38). These findings challenge the concept of the gut mucosa-preserving of early feeding during acute pancreatitis and question results of prior nonrandomized series.

The route of administration has also been studied. Enteral nutrition can be administered either via gastric or jejunal routes. Although the latter is theoretically preferred as it might result in lesser increase in enzyme, bicarbonate, and volume output from the pancreas, this premise has not been proven. Eatock et al. (39) showed no exacerbation of the disease with the nasogastric route as compared to nasojejunal. Though, the jejunal route is advised in patients with gastric atony or duodenal obstruction secondary to pancreatitis and as a result will not tolerate being fed in the stomach.

Cardiovascular Collapse, Renal Failure, and Respiratory Insufficiency

The cornerstone of management in early pancreatitis is fluid resuscitation and close monitoring for early manifestations of organ dysfunction. Although a variety of scoring systems, biomarkers, and radiologic findings can help identify patients at risk of organ dysfunction, these do not substitute for frequent clinical assessment and monitoring. Clinical findings including thirst, poor urine output, persistent tachycardia, tachypnea, hypoxemia, altered mental status, and lack of improvement within the first 48 hours are warning signs of impending severe disease (8).

In addition to the frequent assessment of vital signs, the intravascular volume status should be monitored by physical examination and urinary output. Hypovolemia is secondary to increased capillary permeability, relative lymphatic and splanchnic venous obstruction. This accounts for sequestration of up to 40% of circulatory plasma volume in just a few hours, thereby resulting in renal insufficiency. If cardiac output remains low despite adequate filling pressures, inotropic agents should be initiated to improve hemodynamic status.

Early identification of hypoxemia via either pulse oximetry or arterial blood gas measurement is also paramount. Respiratory insufficiency associated with severe pancreatitis is resulted from a decrease in functional residual capacity and shunting, which may be related to elevated paralyzed hemidiaphragms, basilar atelectasis, pleural effusion, empyema, pneumonia, micropulmonary emboli, or alveolar collapse. The latter occurs due to the decrease in pulmonary surfactant, which is degraded by circulating pancreatic enzymes. Positive end-expiratory pressure (PEEP) is required for maintenance of adequate minute ventilation and oxygenation.

Patients with severe acute pancreatitis with hemodynamic instability should be admitted to the ICU, as well as those patients who are at high risk of rapid deterioration such as the elderly (40), patients requiring high-volume resuscitation, renal failure, respiratory compromise, and patients with evidence of substantial pancreatic necrosis (>30%) (16). Obese patients are also at increased risk for developing complications secondary to severe acute pancreatitis (41).

Therapeutic Peritoneal Lavage

A meta-analysis of eight randomized clinical trials evaluating the use of continuous peritoneal lavage in patients with pancreatitis indicated that this modality was not associated with any significant improvement in morbidity or mortality (42). This is partially explained by the presence of large volumes of fluid within the peritoneal cavity that may degrade the peritoneal defense mechanisms due to the inability to localize the source of contamination through local fibrinous adhesions between omentum, loops of bowel, and the abdominal wall. The lavage may, in addition, enhance the absorption of inflammatory mediators into the systemic circulation and remove important local inflammatory mediators, thereby impairing peritoneal defense mechanisms (43,44). The peritoneal mesothelial cells are usually lost in association with peritonitis and their regeneration may be important in the resolution of the inflammation (42). Lavage, with either crystalloid or peritoneal dialysis solutions, inhibits the rate of mesothelial healing (45). Despite a number of early enthusiastic reports, the use of continuous lavage in patients with acute pancreatitis is not supported by the current available evidence.

Antibiotics

The natural course of severe acute pancreatitis progresses in two phases (46). The first 14 days are characterized by the release of multiple inflammatory mediators leading to SIRS. The second phase, 2 weeks after the onset of disease, is dominated by an anti-inflammatory state with inhibition of immune system leading to sepsis-related complications and possibly, infected pancreatic necrosis. It occurs in up to 70% of patients with necrotizing pancreatitis and has become the most important risk factor for death. In patients with necrotizing pancreatitis, the proportion of patients with proven infected necrosis at the time of surgery increased from 22% to 24% after the first week, from 36% to 55% after the second week, and up to 72% after the third week. The extent of pancreatic necrosis may also be associated with increased infection rates in patients with more than 50% necrosis of the pancreas (47).

Mortality rates in sterile necrotizing pancreatitis range from 10% to 15% (48), while infected pancreatic necrosis carries mortality rates between 30% and 50% (48). The extent of necrosis in patients with sterile lesions correlates with the presence of organ failure, whereas infected necrosis is associated with organ failure regardless of the extent of necrosis (49).

The bacterial spectrum of infection in acute necrotizing pancreatitis has been described as Enterococci, gram-negative bacteria, and fungi (50). Bacterial translocation from the gut has been demonstrated to play a vital role in necrotizing pancreatitis (51). After the administration of antibiotics with primary efficacy against gram-negative and anaerobic bacteria, the spectrum changes to gram-positive infection (46). These infections do not originate in the gut, but rather are nosocomial infections acquired via venous catheters, urinary catheters, or endotracheal tubes (52). Hospital-acquired infections tend to occur much later (typically after 20 days) while infections with gram-negative organisms occur sooner, usually within 2 weeks of admission (46).

The use of prophylactic in necrotizing pancreatitis has been extensively studied. Most studies utilized a carbapenem, imipenem, as it has high pancreatic tissue concentration and the highest bactericidal activity against most of the organisms in pancreatic infections. The data are anecdotal with reduced mortality and infection rates reported by some studies (53,54) while others have found no significant clinical benefit (55,56). The results vary on inclusion and exclusion criteria, and the focus of each series. A recent meta-analysis demonstrated that the timing of administration may determine the clinical outcome (57). By including only RCTs in which antibiotics were administered early (within 72 hours from the onset of symptoms or hospital admission), the mortality rates (7.4% vs. 14.4%) and incidence of infected pancreatic necrosis (16% vs. 25%) were significantly reduced in patients receiving prophylactic antibiotics. It was also shown that antibiotic prophylaxis does not result in increased incidence of fungal infections (58). The possibility of concomitant existence of cholangitis in addition to pancreatitis has to be entertained when dealing with biliary obstruction. In such cases, antibiotics have an essential role in the therapy. Our recommendation at this time pending definitive studies is to use carbapenem-based antibiotics early in the disease process.

SURGICAL MANAGEMENT

Most episodes of acute pancreatitis are mild and resolve spontaneously within 3 to 5 days. The mortality rate in these patients is less than 1% and they do not routinely require intensive care or surgical management. However, operative intervention is needed in certain cases to prevent recurrences and manage complications (59,60).

Biliary Pancreatitis

Occasionally, a common bile duct (CBD) stone may become impacted at the ampulla of Vater, leading to biliary pancreatitis. Typically, the highest serum amylase levels are seen initially, but they return relatively rapidly to normal, as do clinical signs and symptoms. The treatment of gallstone pancreatitis should be guided by the severity of disease. Patients presenting with cholangitis, severe acute gallstone pancreatitis, and obstructive jaundice should undergo urgent ERCP and, if choledocholithiasis is confirmed, endoscopic sphincterotomy should be performed (16). ERCP significantly reduces both the overall complication and mortality rates in patients with severe biliary pancreatitis (54,61). However, ERCP and endoscopic sphincterotomy have no influence on the outcome of mild biliary pancreatitis (61). Should ERCP be considered in a critically ill patient, then it has to be done early in the disease process. ERCP performed after 72 hours of the onset of symptoms will unlikely improve the outcomes.

If there is failure to retrieve the stones during ERCP in a stable patient, bile duct exploration with stone removal needs to be completed at the time of cholecystectomy, and is safely performed laparoscopically in nearly all cases. If ERCP is not previously attempted, CBD stones identified during laparoscopic cholecystectomy by intraoperative cholangiogram can be retrieved by postoperative ERCP. This alternative is preferred if the surgeon is not comfortable in safely performing laparoscopic transcystic CBD exploration. Open cholecystectomy with supraduodenal bile duct exploration is also an option, but should be avoided due to high complication rates (59).

Recurrence of acute pancreatitis in patients with cholelithiasis has been reported in 29% to 63% after hospital discharge without a definitive treatment. The rationale for cholecystectomy and clearance of the CBD is to prevent recurrent biliary pancreatitis. In mild gallstone pancreatitis, cholecystectomy should be performed during the same hospital stay after recovery from the acute event. Patients with mild gallstone pancreatitis should be admitted directly to a surgical service instead to a medical team. Evidence suggests that this results in decreased time for surgery (44 vs. 80 hours), shortened hospital stay (3 vs. 5 days) and lower hospital costs ($11,492 vs. $16,183) (62). In severe gallstone pancreatitis, cholecystectomy should be performed once the inflammatory process has subsided to make the procedure technically feasible and safe. If ERCP and endoscopic sphincterotomy is performed, cholecystectomy should be performed within 6 weeks (63).

MRCP can be used to select patients that might still have CBD stones prior to cholecystectomy. In a study by our group, we found that MRCP had a sensitivity of 100% and an accuracy of 92% in detecting CBD stones in patients recovering from gallstone pancreatitis (64). We advocate that patients with a negative MRCP will not need a preoperative ERCP. Patients with positive MRCP will need a preoperative ERCP to clear the duct. In our opinion, this approach will cut on unnecessary ERCP and will avoid intraoperative issues when dealing with a positive intraoperative cholangiogram (64).

Vascular Complications

The systemic vascular effects of acute pancreatitis are related to the release of pancreatic proteases, such as trypsin, which activate complement C5a and precipitate the coagulation cascade, granulocyte aggregation and leukoembolization of vital tissues, such as the lung, kidney, and splanchnic and systemic vascular beds. These events lead to the respiratory distress syndromes, including pulmonary emboli, renal insufficiency, and splanchnic venous thrombosis (65).

Also, there are local arterial and venous complications of pancreatitis. Bleeding from pancreatic pseudocysts and ruptured pseudoaneurysms carries mortality rates of 25% to 40% (66). Bleeding may present as melena from erosion into the proximal gastrointestinal tract or as hypovolemia and abdominal pain if there is rupture into the peritoneal cavity. Diagnosis is usually late or only at postmortem examination. Most patients with gastrointestinal tract bleeding secondary to acute or chronic pancreatitis are alcoholics. Other causes of gastrointestinal bleeding in this population include peptic ulcer disease, gastritis, varices, and Mallory–Weiss tear. The development of pseudoaneurysms is probably related to severe inflammation and enzymatic autodigestion of the pancreatic and peripancreatic arteries. With growth and expansion, pseudoaneurysms may rupture into pseudocysts, adjacent viscera, peritoneal cavity, or pancreatic duct.

The most common vessel involved in splanchnic pseudoaneurysms related to pancreatitis is the splenic artery, followed by the gastroduodenal and the inferior pancreaticoduodenal (67). Pseudoaneurysms may be frequent in up to 10% of patients with chronic pancreatitis, but bleeding rarely occurs unless are associated with pseudocysts. The diagnosis of ruptured pseudoaneurysms requires a high index of suspicion. Diagnostic and therapeutic tools include emergency upper endoscopy, selective visceral angiography, US, and CT scanning. Currently, the best method of assessing a relatively stable patient is CT angiogram with or without a formal angiogram. This provides an accurate diagnosis but also allows, if required, radiologic intervention by embolization of the feeding vessel or the pseudoaneurysm itself (Fig. 128.1) (68). If visceral ischemia distal to the pseudoaneurysm is a concern, a stent may be placed (69). Control can be rendered by either selective arterial vasopressin infusion, angioembolization with Gelfoam, detachable intravascular balloons, Gianturco coils, or polymerizing adhesives. Surgical control is only indicated for immediate life-threatening bleeding or failure of interventional radiology approach or need for management of other coexisting abdominal pathologies (68). However, it is emphasized that control of bleeding from a ruptured visceral aneurysm in the setting of pancreatitis is rarely achieved in the operating room due to the severe inflammatory process. In most cases, the only option available to the surgeon is packing the area and immediate transfer to the radiology suite for embolization.

FIGURE 128.1 Selective angiogram showing embolization of a giant pseudoaneurysm of the gastroduodenal artery in a patient with acute pancreatitis.

Hemoductal pancreatitis (or hemosuccus pancreatitis) occurs when a pseudoaneurysm ruptures into the pancreatic duct. It usually encompasses the triad of gastrointestinal bleeding, pancreatitis with epigastric pain, and partial CBD obstruction (70). The diagnosis can be confirmed by selective visceral angiography or ERCP. The treatment of this rare complication requires ligation of the pseudoaneurysm and possible pancreatic resection.

Venous thrombosis of the portal vein is a potential complication of acute or chronic pancreatitis. The clinical course is complicated by acute decompensation, hypotension with sequestration in the vascular bed, acidosis, hepatic enzyme elevation, coagulopathy, and venous infarction of the bowel. Patients who survive this insult develop portal hypertension, and some present months to years later with bleeding esophageal varices. Selective splenic venous thrombosis occurs more frequently, and patients usually present with splenomegaly, unexplained blood loss, pain in the left upper quadrant and subscapular area, and possibly, cardiovascular collapse secondary to the subscapular hematoma. The treatment is splenectomy with preoperative vascular control by angiographic techniques.

During drainage procedures for pancreatic pseudocysts in the presence of associated splenic venous thrombosis, the transgastric approach should be avoided in order to decrease postoperative bleeding from the rich submucosal plexus of high-pressure veins. In the absence of bleeding gastric varices, one may elect to leave the spleen in situ even with splenic vein thrombosis, because not all patients develop bleeding from gastric varices.

Pancreatic Pseudocyst

Peripancreatic fluid collections can occur as a result of acute pancreatitis, chronic pancreatitis, surgery (either pancreatic or other abdominal surgery), trauma or neoplasia. With the exception of a cystic neoplasm, peripancreatic fluid collections form either as a result of a disrupted pancreatic ductal system with subsequent fluid leakage or maturation of peripancreatic necrosis. An acute peripancreatic fluid collection (APFC) is located in or near the pancreas; occurs early in the course of acute pancreatitis and always lacks a wall of granulation or fibrous tissue (Table 128.1) (34). APFC is common in patients with severe acute pancreatitis, and more than half of these lesions regress spontaneously.

Pseudocyst is defined as a collection of pancreatic juice that is enclosed by a nonepithelialized wall composed of either fibrous or granulation tissue. It is a frequent complication in acute pancreatitis (10% to 20% of cases) and chronic pancreatitis (20% to 40% of cases) (71). Formation of an acute pseudocyst requires 4 or more weeks from the onset of acute pancreatitis (Table 128.1). It is critical to distinguish a pseudocyst from a pancreatic fluid collection prior to intervention.

The patient usually presents with persistent abdominal or back pain, bloating, early satiety, nausea, vomiting or failure to thrive after an episode of pancreatitis. In few cases, biliary obstruction may ensue if pseudocyst is located in the head of the pancreas. Not uncommonly, a pseudocyst may be asymptomatic and discovered incidentally on imaging studies. The contents are usually rich in pancreatic enzymes and are most often sterile. Bacteria may be present in pseudocysts but often are of no clinical significance. When pus is present, the lesion is termed pancreatic abscess. The distinction between pancreatic abscess and infected necrosis is critical for two reasons: the mortality risk for infected necrosis is double than for pancreatic abscess and specific therapy for each condition may be remarkably different.

Pseudocysts are usually diagnosed by imaging studies, either contrast-enhanced CT scan or US. On CT scan, a pseudocyst appears as a well-circumscribed, rounded lesion with homogeneous fluid density and thickened hyperdense capsule (Fig. 128.2). The initial step in planning intervention is to define the underlying pancreatic ductal anatomy with MRCP (or ERCP) and its relationship to surrounding structures. Failure to address pancreatic ductal involvement may result in postoperative failure and recurrence. Distinguishing a pseudocyst from other pancreatic cystic neoplasms is also essential and can be made by initial imaging studies and a detailed history of the patient. The traditional management of pancreatic pseudocyst has been based for decades on a report by Bradley et al. (72) who found spontaneous resolution of the pseudocyst in 24 of 54 patients. This spontaneous resolution occurred most frequently in collections less than 6 cm and/or those resolved within 6 weeks of follow-up. Thus, historically, intervention has been advisable in pseudocysts larger than 6 cm that persist beyond 6 weeks. However, recent experience showed that conservative approach even in these circumstances might be occasionally acceptable.