Pancreatic Exocrine Dysfunction

In Western countries, the number of patients suffering from pancreatic exocrine insufficiency has increased. The most common cause is an increase in alcohol consumption with a concomitant increase of CP (FitzSimmons, 1993). The main goal of therapy for pancreatic exocrine dysfunction is to avoid fat maldigestion. Reasons for earlier and more severe impairment of fat digestion, compared with protein and carbohydrate digestion, in patients with pancreatic insufficiency are that 1) impairment of pancreatic lipase synthesis and secretion occurs earlier; 2) more rapid and complete inactivation of lipase occurs in the acidic duodenum as a result of impaired bicarbonate output; 3) proteolytic degradation of lipase occurs earlier during aboral transit than with amylase and proteases; 4) impairment of pancreatic bicarbonate secretion decreases duodenal pH, resulting in precipitation of glycine-conjugated bile acids and further deterioration of fat digestion; and 5) extrapancreatic sources of lipase are unable to compensate for loss of pancreatic lipase activity.

For all of the reasons listed, steatorrhea is the leading symptom in patients with pancreatic exocrine insufficiency.

Pancreatic Exocrine Enzyme Supplementation

When weight loss or steatorrhea (≥15 g/day) or both develop, supplementation of pancreatic enzymes is indicated. Dyspepsia, diarrhea, meteorism, and malabsorption of proteins and carbohydrates also have been cited as indications. Another interesting indication for pancreatic enzyme supplementation, although not formally studied, is in the treatment of pain (discussed in the next section). The main goal of the treatment of pancreatic exocrine dysfunction is to ensure that optimal amounts of lipase reach the duodenum with the delivered food. With the currently available pancreatic enzyme supplement preparations, azotorrhea (protein malabsorption) can be eliminated (Brady et al, 1991), whereas steatorrhea usually can be reduced but not totally corrected.

Four different types of pancreatic exocrine enzyme preparations are available. Uncoated preparations show only poor effects because of their inactivation by gastric acid. Huge doses of enzymes are required to have any effect on fat malabsorption. These preparations should be used only in patients with pancreatic exocrine insufficiency and hypochlorhydria or achlorhydria. The use of enteric-coated tablets is strongly discouraged, because these preparations are ineffective for decreasing fat excretion owing to erratic enzyme release. The superiority of enteric-coated microsphere preparations over conventional enzyme preparations with regard to decrease in stool fat excretion has been firmly established (Layer & Holtmann, 1994). Pancreatic enzymes in these preparations are protected at low pH by a special polymer coating. The release of the enzymes occurs only at a pH of at least 4.5. Simultaneous administration of antacids, H2-receptor antagonists, or proton pump inhibitors is unnecessary. Unfortunately, microencapsulation results in a considerable increase in costs. Another problem concerning enteric-coated microspheres might be considerable differences in the diameter of the microspheres and in their physical nature as a result of differences in the manufacturing process; however, clinical differences in the various enteric-coated microsphere preparations have not yet been proofed.

The use of enzyme preparations in combination with acid-reducing compounds also is not justified for many reasons, although additional administration of proton pump inhibitors or H2-receptor antagonists reduces fecal nutrient loss. These preparations are expensive, there is a variability of the effects owing to the acid-reducing therapy, and especially in children, the safety of long-term administration has not been firmly established (Lebenthal et al, 1994).

Certain bacterial and fungal lipases have shown no benefit in the treatment of pancreatic steatorrhea, although they are resistant against acid denaturation and proteolytic digestion. Decreased lipolytic activity of fungal lipase in the presence of bile acids could explain the fact that no superiority over enteric-coated microsphere preparations has been demonstrated to date (Griffin et al, 1989; Zentler-Munro et al, 1992).

Although rare, possible side effects of pancreatic exocrine enzyme supplementation are sourness of the mouth, perianal irritation, abdominal pain, diarrhea, constipation in infants, allergic reactions to pork proteins, hypersensitivity reactions after inhalation, and fibrosing colonopathy in cystic fibrosis patients (Lebenthal et al, 1994; Loser & Folsch, 1995). In regard to fibrosing colonopathy, it has been recommended that the maximum dose of lipase should not exceed 10,000 IU/kg body weight per day, or 2500 IU/kg body weight per meal.

Substitution Therapy in Chronic Pancreatitis

Approximately 80% of patients with CP can be managed by dietary means and pancreatic enzyme supplements. Some patients (10% to 15%) need oral supplements (polymeric or semielemental), 5% need enteral tube feeding, and approximately 1% need total parenteral nutrition (TPN). Reduction of steatorrhea and supplementation of calories are the main goals of nutritional therapy in CP. Treatment of exocrine insufficiency starts with dietary therapy and pancreatic enzyme supplementation (DiMagno, 1979).

Total abstinence from alcohol and partaking of frequent meals are basic dietary recommendations. The diet should be rich in proteins (1.0 to 1.5 g/kg body weight/day) and carbohydrates, although carbohydrates must be limited when diabetes mellitus is present. In addition, 30% to 40% of calories should be in the form of fats. Medium-chain triglycerides may be used to increase fat absorption, because they are absorbed directly across the small bowel into the portal vein, even in absence of lipase, co-lipase, and bile salts. Vitamins should be supplemented if serum levels indicate a deficiency (Havala et al, 1989). Patients with protein maldigestion and steatorrhea must be supplemented individually with exogenous pancreatic enzymes. Weight control, symptomatic relief of diarrhea, and a decrease in 72-hour fecal fat excretion are practical end points of therapy.

Enteral nutrition (EN) is indicated when a patient is unable to eat, has progressive weight loss despite an adequate dietary regimen, or develops acute complications; EN may also be indicated preoperatively and postoperatively. Only rarely is TPN indicated in patients with CP, such as when gastric emptying is blocked, when the patient needs gastric decompression, when the feeding tube cannot be introduced into the jejunum, or when a complicated fistula is present. In most cases, surgical intervention can solve these problems.

Conservative Treatment of Pain

Pain significantly reduces patients’ quality of life, and pain relief should be one of the main goals of conservative treatment in CP. Pain also may be treated interventionally or surgically, but medical treatment is generally the first-line therapy in patients with painful CP.

Different medical treatment options and therapeutic interventions are available, and these must be integrated into an individualized treatment plan. The pathogenesis of pain especially influences the therapeutic procedure. Two mechanisms are suggested for the generation of pain in the absence of local complications: 1) inflammatory changes of pancreatic parenchyma with intrapancreatic and peripancreatic neural alterations (Bockman et al, 1988; Ceyhan et al, 2009a, 2009b; Di Sebastiano et al, 1997), and 2) ductal and intraparenchymal hypertension (Di Sebastiano et al, 2003; Ebbehoj et al, 1990).

As for the second mechanism, options that reduce the intrapancreatic pressure may lead to a significant reduction of pancreatic pain. In the case of inflammation, a significant number of inflammatory mediators in pancreatic parenchyma and pancreatic nerves are detected in patients with pain (Ceyhan et al, 2009b; Di Sebastiano et al, 2000). Several medical, analgesic, and antiinflammatory treatment options are available, which may be combined with or supported by interventional methods.

Interventional Procedures to Treat Pancreatic Pain

Of patients with pancreatic pain, 40% to 70% seem to benefit from medical treatment (Ammann et al, 1984). This percentage may increase by combining medical therapy and interventional procedures (van der Hule et al, 1994). Pancreatic pain may also be improved by the combination of interventional endoscopy and lithotripsy, although a recent RCT demonstrated increased costs but no outcome benefit when systematic endoscopy was added to lithotripsy (Dumonceau et al, 2007). Some studies showed a significant difference in improvement of pain among patients with successful or unsuccessful lithotripsy (Delhaye et al, 1992; Costamagna et al, 1997; Johanns et al, 1996; Ohara et al, 1996; Sauerbruch et al, 1992; Schneider et al, 1994; Smits et al, 1996). Other studies did not report this difference (Adamek et al, 1999).

Celiac plexus neurolysis and celiac block involve injecting an agent at the celiac axis, with the aim of either selectively destroying the celiac plexus or temporarily blocking visceral afferent nociceptors. Agents most commonly used for this purpose include alcohol or phenol for neurolysis and bupivacaine and triamcinolone for a temporary block. Methods to administer such agents to the celiac ganglion include CT imaging, percutaneous ultrasound, fluoroscopy, or endoscopic ultrasound (EUS). Whereas the EUS-guided technique may be superior compared with the other methods, response rates in general are relatively low (Gress et al, 1999; Noble & Gress, 2006; Santosh et al, 2009).

Autoimmune Pancreatitis

A type of CP that might be caused by an autoimmune mechanism was first described by Sarles and collegues (1961) and was termed primary inflammatory sclerosis of the pancreas. Since then, a growing body of evidence suggests that autoimmune pancreatitis (AIP) is distinct from obstructive or calcifying forms of CP. Two clinical and histologic patterns of AIP can be distinguished (Park et al, 2009). Type 1 is characterized by the histologic key features of dense periductal lymphoplasmacytic infiltrate, swirling or storiform fibrosis, and obliterative venulitis; serum levels of the immunoglobulin G4 (IgG4) subclass of IgG are typically increased (Hamano et al, 2001). Type 1 AIP fits the classic description of the disease reported in Japan, also known as lymphoplasmacytic sclerosing pancreatitis. Type 1 AIP appears to be the pancreatic manifestation of a systemic disease that may also affect other organs, including the bile duct, retroperitoneum, kidney, lymph nodes, and salivary glands. Type 2 AIP is typically characterized by a duct-destructive pathology with infiltrating neutrophils, lymphocytes, and plasma cells. Typical serologic markers seen in type 1 AIP are not elevated in type 2 AIP (Park et al, 2009). Typical imaging features—a diffuse, sausage-shaped, enlarged pancreas with delayed and peripheral enhancement—only show up in about 50% of patients. Furthermore, classic features of lymphoplasmacytic sclerosing pancreatitis are found in just 20% of core biopsies, and false-positive IgG4 immunostaining has also been reported in the setting of cancer. Thus the preoperative differentiation from pancreatic carcinoma is a diagnostic challenge. In highly suspected or histologically confirmed cases of AIP, steroid therapy (prednisolone, initially 30 to 40 mg/day) may ameliorate symptoms and imaging findings within two weeks. Importantly, the similarity to pancreatic carcinoma and the absence of specific diagnostic parameters may result in surgical treatment of this disease. Because reports about patients misdiagnosed as having AIP are increasingly encountered, surgery should always be considered in doubtful cases (Gardner et al, 2009).

Decompression of Pancreatic Pseudocysts

Pancreatic pseudocysts are a late complication of CP. The main techniques for decompression of pancreatic pseudocysts include transgastric and transduodenal approaches (Sahel, 1990; Dohmoto & Rupp, 1992); these require a clear bulging of the cyst into the gastric or duodenal cavity to ensure a short distance between the cystic wall and the intestinal tract. In this context, endosonography has been shown to reduce the risk of cyst perforation and hemorrhage (Binmoeller et al, 1994; Etzkorn et al, 1995; Grimm et al, 1992).

The technical success rate appears to be better for cystoduodenostomy than for cystogastrostomy owing to a higher complication rate (Binmoeller et al, 1994; Sahel, 1991). In more recent articles, the mortality appears to be almost zero, whereas morbidity rates are reported between 3% and 11% (Binmoeller et al, 1994; Dohmoto & Rupp, 1992; Etzkorn et al, 1995; Grimm et al, 1992; Sahel, 1990, 1991). A more recent approach to access the cystic cavity is indirectly via the papilla and the ductal system (Pinkas et al, 1994). For this reason, a selective endoscopic pancreatic sphincterotomy and a retrograde pancreatography must be performed to introduce a guidewire, over which the dilation of the cannulation can be performed. A double-pigtail endoprothesis should be inserted through the papilla into the cyst. No deaths (Dohmoto & Rupp, 1992; Pinkas et al, 1994; Sahel, 1990) and very low morbidity (2% to 7%) have been reported. A major problem of this approach is its nonfeasibility in patients with strictures or stones located between the papilla and the pseudocyst. Because these procedures call for a highly specialized endoscopist, cases with poor anatomic conditions or with contraindications determined by preliminary endosonography should be referred to a specialized center.

Common Bile Duct Stenting

Common bile duct stenosis is a mechanical complication of CP. A team approach that includes endoscopists and surgeons should guide the treatment of common bile duct stenosis in patients with CP. The decision for either interventional or surgical treatment depends on the patient’s age, comorbidities, the course of CP, and the cause of the stricture. A subgroup of patients has been reported to benefit from endoscopic stenting, with permanent regression of the stenosis reported after stent removal. Endoscopic stenting should be performed as first-line therapy in those patients with CP-induced acute cholestasis, when malignancy can be excluded. Especially for patients who refuse surgery or those who have significant comorbidities, the endoscopic approach might be the therapy of choice. Kahl and associates (2003) reported in a prospective follow-up study that endoscopic drainage of biliary obstruction provides excellent results in the short term but offers only moderate long-term results. Patients without calcifications of the pancreatic head benefit from biliary stenting; patients with calcifications were identified to have an increased risk of failure within a 12-month course of endoscopic stenting. Major limitations, such as plastic stent clogging, migration, and cholangitis, frequently occur in the long-term follow-up. Alternatively, self-expanding Wallstents may improve the results, and metal stents that cannot migrate offer a stent effectiveness greater than 80% (Ng & Huibregtse, 1998). At present, long-term follow-up data to clarify the role of endoscopic approaches and to identify the preferred stent for the management of cholestasis in CP are unavailable. The placement of Wallstents used to be irreversible, but results with the latest generation of stents that are covered, and therefore removable, are awaited. However, for young patients with CP and no comorbidity, adequate and definitive therapy for recurrent cholestasis is provided by a surgical drainage procedure or pancreatic head resection (Eickhoff et al, 2001).

Drainage Procedures

Pancreatic duct sphincterotomy was one of the first surgical procedures proposed for patients with CP and stenosis at the papilla of Vater. This procedure was recognized as a dangerous approach, however, with lower success rates for amelioration of pain, indicating that a stenosis at the papilla Vater is not the cause of pain in CP. The original Puestow procedure, and its modification by Partington and Rochelle (1960), proved more successful in patients with CP and a dilated pancreatic duct. The procedure includes resection of the tail of the pancreas, followed by a longitudinal incision of the pancreatic duct along the body of the pancreas and an anastomosis with a Roux-en-Y loop of jejunum. The modification by Partington and Rochelle abandons the resection of the pancreatic tail. Preservation of tissue and reduction of mortality to less than 1% and of morbidity to less than 10% are the advantages of this operation (Evans et al, 1997; Izbicki et al, 1999; Prinz & Greenlee, 1981; Proctor et al, 1979). Although these ductal drainage operations have good primary success rates (Duval, 1956; Partington, 1952), their long-term outcome is poor (Markowitz et al, 1994). Moreover, these procedures are only promising if the duct is substantially dilated (>7 mm), which is the case in about 25% of patients (Büchler & Warshaw, 2008), and in patients who do not show a dominant mass in the head of the pancreas. For these reasons, pure drainage procedures have been replaced by techniques that combine resection and drainage for the majority of patients.

In patients with isolated pancreatic pseudocysts, and often in those with a history of a severe episode of acute pancreatitis, a drainage procedure in the form of a cystojejunostomy with Roux-en-Y reconstruction (Fig. 55B.1) is still the surgical procedure of choice.

FIGURE 55B.1 Drainage procedures in chronic pancreatitis. A, If the cyst wall is thick enough, a pancreatic pseudocyst can be safely and effectively treated by drainage with a cystojejunostomy and Roux-en-Y reconstruction. B, In rare patients with a dilation of the pancreatic duct >7 mm and no inflammatory mass, a laterolateral pancreatojejunostomy (Partington-Rochelle procedure) may be performed.

Resective Procedures

The vast majority of patients are seen with a ductal obstruction in the pancreatic head, frequently associated with an inflammatory mass. In these patients pancreatic head resection is the procedure of choice; the available techniques are shown in Figure 55B.2. The partial pancreatoduodenectomy, or Kausch-Whipple procedure, in its classic or pylorus-preserving variant, has been the procedure of choice for pancreatic head resection in CP for many years (Jimenez et al, 2000, 2003). The duodenum-preserving pancreatic head resections and its variants—the Beger (Beger et al, 1985), Frey (Frey & Smith, 1987), and Bern procedures (Gloor et al, 2001)—represent less invasive, organ-sparing techniques with equivalent long-term results (see Table 55B.2).

FIGURE 55B.2 Techniques of pancreatic head resection for chronic pancreatitis. Resections are shown on the left side, reconstructions on the right side. A, Partial pancreatoduodenectomy; the pylorus-preserving procedure is shown. The pancreatic head and duodenum are removed, reconstruction is performed by pancreatojejunostomy, hepaticojejunostomy, and a duodenojejunostomy. B, Duodenum-preserving pancreatic head resection (DPPHR)-Beger procedure. The pancreas is dissected on the level of the portal vein, the pancreatic head is excavated, and the duodenum is preserved with a thin layer of pancreatic tissue. If the bile duct is obstructed, it can be opened, and an internal anastomosis with the excavated pancreatic head can be performed (not shown). The reconstruction is performed with a Roux-en-Y jejunal loop, including two anastomoses, one to the pancreatic tail remnant and one to the excavated pancreatic head. C, DPPHR-Frey procedure. A circumscript excision in the pancreatic head is combined with longitudinal dissection of the pancreatic duct toward the tail. Reconstruction is performed with an anastomosis to a Roux-en-Y jejunal loop. Compared with the Beger procedure, pancreatic head resection is less extensive; reconstruction is easier, because only one anastomosis to the pancreas is required. D, DPPHR-Bern modification. A technical simplification of the Beger procedure, the extent of resection of the pancreatic head is comparable, except the pancreas is not dissected on the level of the portal vein. Thus reconstruction can be performed with a single anastomosis of the pancreas to a Roux-en-Y jejunal loop. The bile duct may be opened if necessary, with an internal anastomosis to the loop (shown). The pancreatic duct toward the tail is probed to rule out distal stenosis.

Only very few patients come to medical attention with small duct disease (diameter of the pancreatic duct <3 mm) and no mass in the pancreatic head. In these cases, a V-shaped excision of the anterior aspect of the pancreas is a safe approach and offers effective pain management (Yekebas et al, 2006). In the rare case of a patient seen with segmental CP in the pancreatic body or tail, such as that seen as a result of posttraumatic ductal stenosis, a middle segment pancreatectomy or a pancreatic left resection may be the best approach. The most frequently applied resection techniques are detailed here.

Evidence-Based Surgery in Chronic Pancreatitis: Pancreatoduodenectomy and DPPHR

Irrespective of the technique, if carried out by experienced personnel, pancreatic head resection is a safe and effective therapy with good short- and long-term results in patients with CP and an inflammatory mass in the head of the pancreas. Various techniques of pancreatic head resection were compared in several RCTs (Table 55B.3; Büchler et al, 1995; Izbicki et al, 1995, 1997, 1998; Klempa et al, 1995; Köninger et al, 2008; Müller et al, 2008a; Strate et al, 2003, 2008), in which their safety and efficacy were confirmed. The RCTs that compared PD and DPPHR (Büchler et al, 1995; Izbicki et al, 1998; Klempa et al, 1995; Müller et al, 2008a; Strate et al, 2008), as well as a recent meta-analysis (Diener et al, 2008), demonstrated comparable mortality and efficacy rates in terms of pain relief and endocrine insufficiency. The less invasive DPPHR was superior in terms of duration of hospital stay, exocrine insufficiency, weight gain, and quality of life in medium-term follow-up. In recent studies with long-term follow-up, these metabolic advantages appear to decrease over time, and long-term results of PD and DPPHR are equal in terms of pain management, quality of life, and endocrine and exocrine function (compare results reported in the initial trials and the latest follow-ups in Table 55B.3). Of interest, the resection techniques remain effective in terms of pain relief and quality of life but may not stop the progress of exocrine and endocrine insufficiency in the long term (Müller et al, 2008a; Strate et al, 2008). One explanation may be that resection effectively treats duct obstruction and hypertension but does not completely stop continued cellular damage and parenchymal loss (Büchler & Warshaw, 2008). Certainly these observations stress the importance of continued postoperative therapy, including the reduction of etiologic factors, substitution of exocrine and endocrine insufficiency, and possibly antioxidant therapy (Bhardwaj et al, 2009).

The RCTs that compared the various techniques of DPPHR demonstrated equal outcomes regarding pain control, quality of life, and metabolic parameters after the Beger procedure versus the Frey and Bern procedures (see Table 55B.3; Izbicki et al, 1995, 1997; Köninger et al, 2008; Strate et al, 2005). However, the trial by Köninger and colleagues (2008) demonstrated that the Bern modification of the DPPHR represents a technically simpler alternative, as reflected by a significantly shorter operative time (by 46 minutes) and a significantly shorter hospital stay (11 vs. 15 days). Thus the Bern technique represents a modification of the DPPHR that will find a broader acceptance in the future because of technical and economic advantages.