Figure 54-1. Three-hit hypothesis for pathophysiology of chronic pancreatitis. NGF, nerve growth factor. Adapted from Whitcomb DC. Genetic risk factors for pancreatic disorders. Gastroenterology 2013;144:1292–1302.
The pathophysiology of CP is not well elucidated. A theory known as the “three-hit hypothesis” holds that (1) a stochastic event occurs resulting in (2) inappropriate trypsin activation causing acute pancreatitis. The patient then has (3) an unfavorable immunologic response to the inflammation resulting in fibrosis and CP (Fig. 54-1).20 Environmental factors (e.g., alcohol and tobacco) are implicated as the inciting events to this cascade, with potential key modulating factors including genetics and the histologic milieu.
Insights into pathophysiology of pancreatitis have occurred through advances in cellular basic science. PSCs are causative in pancreatic fibrogenesis. PSCs are residents of the healthy pancreas that become activated by inflammatory cytokines during pancreatitis to become myofibroblast-like cells, producing extracellular matrix in the interstitial space.21 In addition, several matrix metalloproteinases are implicated in altering extracellular matrix remodeling and collagen degradation, enhancing fibrogenesis and irreversibly altering the organ architecture to a diseased, fibrotic pancreas.22
The pathophysiology of the CP pain syndrome is not well delineated and is likely multifactorial. Classically, pancreatic ductal obstruction due to fibrosis and resulting in elevated intraductal pressures has been implicated as a primary cause for pain. Additionally, pancreatic capsular and parenchymal fibrosis resulting in intracapsular hypertension and ischemia, create a “pancreatic compartment syndrome,” and has been theorized to result in pain. More recent theories have focused on peripancreatic neuropathy. On a histologic level, peripancreatic neuronal hypertrophy as well as infiltration of periaxonal tissue with inflammatory cells is evident.23,24 Increased presence of the “pain neurotransmitters” is identified including calcitonin gene-related peptide and substance P, stimulated by nerve growth factor.25 These changes in the peripancreatic neuronal milieu may result in peripheral and central neural sensitization and undoubtedly contribute to the CP pain syndrome. The pancreas has a uniquely villainous role in abdominal pain syndromes. No other visceral organ can match it in terms of pain severity. Neural remodeling precipitated by pancreas-synthesized tachykinins may lead to centralization of pain that is precipitated by extra pancreatic stimuli. The “phantom pancreatitis” pain that occurs after total pancreatectomy (TP) is related to centralization of pain pathways and is a reminder of the vast intersecting neuronal web of the pancreas and the foregut.
2 The primary presentation of CP is severe, intractable epigastric abdominal pain that radiates into the back. The classic pain is daily and constant with periods of exacerbation. Patients typically describe pancreatitis pain as if someone is slowing twisting a knife into the epigastrium and interscapular region. Patients often have associated gut dysfunction, with nausea, emesis, and difficulty tolerating a diet consistently, particularly during pain episodes. Pain and nausea are frequently precipitated by ingestion of a fatty meal or less commonly a high protein meal. Patients may develop endocrine failure (pancreatogenic diabetes, type 3c) and exocrine pancreatic insufficiency (EPI) over time. Less often, patients may present with an acute complication of CP such as biliary or duodenal obstruction, a pancreatic pseudocyst, pancreatic ascites, mesenteric venous thrombosis, or mesenteric arterial pseudoaneurysm (Table 54-1).
On physical examination, patients may generally appear malnourished and underweight. Abdominal tenderness in the epigastrum may be elicited. Significant laboratory values to be examined include chemistries to evaluate for dehydration and acidosis. A hepatic panel may reveal elevated alkaline phosphatase or direct bilirubin, indicating biliary obstruction, or hypoalbuminemia from chronic malnutrition. Serum amylase and lipase may be elevated or may be normal during pain exacerbations in advanced disease.
Table 54-1 Complications of Chronic Pancreatitis
Figure 54-2. Contrasted CT scan in venous phase, coronal images showing (A), dilated main pancreatic duct with intraductal stones and (B), significant fibrotic disease burden in the head of the pancreas with intraparenchymal and intraductal calcifications.
Radiography is useful in both diagnosis and treatment planning in CP. The most relevant imaging modalities include contrast-enhanced computed tomography (CT), secretin-stimulated magnetic resonance imaging with cholangiopancreatography (MRCP), endoscopic retrograde cholangiopancreatography (ERCP), and endoscopic ultrasound (EUS).
Abdominal CT can show pancreatic parenchymal changes including edema, fibrosis, or atrophy. Pancreatic ductal dilation may be evident, as well as intraparenchymal and intraductal calcifications (Fig. 54-2). CT can also be helpful in recognizing intra-abdominal complications of pancreatitis such as biliary or duodenal obstruction, pancreatic pseudocysts or ascites, or thrombosis or pseudoaneurysms of the mesenteric vasculature. Similarly, MRCP can be very useful in showing parenchymal changes in enhancement and secretion (T1-weighted images) and ductal anatomy (T2-weighted images), particularly with the addition of secretin stimulation (Fig. 54-3). With the increased capabilities of MR technology over the past couple of decades, MR imaging has largely replaced ERCP for diagnostic imaging in CP. ERCP is the classic imaging modality for CP. The ERCP Cambridge classification system, derived from an international consensus, remains the gold standard of CP staging (Table 54-2).26 In the modern era, ERCP is primarily utilized as a therapeutic modality. EUS is useful for evaluation of pancreatic parenchyma and ductal anatomy, while being less invasive than ERCP. EUS also has a grading system to objectify and document pancreatitis disease severity, although the modality still maintains interobserver variability (Table 54-3). With the addition of fine needle aspiration, EUS can be helpful in the differentiation of pancreatic neoplasms from CP.27
Figure 54-3. T2-weighted magnetic resonance imaging with secretin stimulation demonstrate a pathologically dilated pancreatic duct in the body of the pancreas (A), and cystic changes and dilated ducts in a fibrotic and inflammatory head of the pancreas (B).
CONSIDERATIONS IN MANAGEMENT
3 Frontline management of CP includes risk factor modification, such as alcohol and tobacco cessation. Primary medical interventions entail pain management, including adjunctive behavioral therapy, and nutritional optimization, including pancreatic enzyme replacement.
Pancreatic enzymes are the mainstay of medical management of CP despite controversy about their efficacy. Enzyme replacement is presumed to improve pain by feedback inhibition of cholecystokinin (CCK) release from the duodenum, leading to decreased pancreatic exocrine secretions. A meta-analysis of six randomized, placebo-controlled trials did not reveal a significant benefit for supplemental pancreatic enzyme therapy for pain relief.28 Antioxidant therapy has been proposed as a treatment for CP, based on the theory that antioxidants will reduce oxygen free radicals and ameliorate oxidative stress and pancreatic acinar cell injury. Prospective randomized trials have shown conflicting results in antioxidant therapy for CP.29
Table 54-3 EUS Criteria for Chronic Pancreatitis
Because CP pain is associated with extrapancreatic neuronal remodeling, neuromodulation with anticonvulsant or antidepressant medications similar to gabapentin have been utilized for CP pain management with limited success in patients with ongoing pancreatic inflammation.
Avoidance of narcotic analgesia in the management of CP pain is uniformly recommended. Analgesic management with nonsteroidal anti-inflammatory medication, acetaminophen, and tramadol is typically recommended. However, because the severity of pancreatic pain is so incapacitating, most patients who are evaluated for endoscopic or surgical management have been treated with narcotic analgesics and have developed physiologic narcotic dependence.
Behavioral therapy is a keystone of therapy in CP, particularly since many patients who are debilitated by CP are young and have previously been in good health. Behavioral modification is effective in modulating pain perceptions in many chronic disorders. In addition, behavioral therapy is important in patients at risk for opioid misuse.30
Patients who have failed medical management, have continued debilitating pain or nutritional failure, and are physiologically fit are candidates for therapeutic interventions.
ERCP is the primary endoscopic modality for therapy in CP. The principle goal with ERCP is to relieve any obstructive process. Potential maneuvers include sphincterotomy, stone extraction, stricture dilation, and stenting.
EUS can be utilized for endoscopic pseudocyst drainage procedures and for celiac plexus neurolysis. Percutaneous and endoscopic-guided celiac nerve blockade has been utilized to manage CP pain. Meta-analysis studies of EUS-guided celiac plexus blockade found short-term pain relief in about one-half of the subjects.31
In general, the endoscopic approach is undertaken prior to consideration of surgery, given the perceived advantages of lower morbidity with a less invasive approach. Two prospective, randomized, controlled trials have compared endoscopy and surgery in patients with obstructive CP. Dite and colleagues randomized 72 patients with CP, pancreatic duct obstruction and pain to endoscopic or surgical intervention. Endoscopic therapy consisted of ERCP with 52% undergoing sphincterotomy and stenting and 23% stone removal. Operative management was 20% drainage procedure and 80% resection. At 5-year follow-up, the surgical group had a greater proportion of patients that were pain free (34% vs. 15%).32 In another study Cahen and colleagues from randomized 39 patients with dilated duct CP and pain to endoscopic or surgical management. Endoscopic treatment was ERCP with sphincterotomy and stenting, and operative therapy was a drainage procedure (longitudinal pancreaticojejunostomy). At 5-year follow-up, the surgical group had a greater proportion of patients that had pain relief (80% vs. 38%, p = 0.001), had larger improvements in quality of life, and underwent fewer procedures, despite equivalent morbidity, length of stay, and preserved pancreatic function.33,34
4 Approximately two-thirds of patients with debilitating pain from CP fail medical and endoscopic managements and are candidates for consideration for operative therapy. The primary indication for surgical intervention in CP is intractable pain, and the goals of surgery are to effectively relieve pain while minimizing morbidity, including minimizing perioperative complications and preserving pancreatic parenchyma. As the underlying cause of CP pain is not well understood, operative decision-making can be difficult. The pancreatic anatomy is the primary determinant in surgical planning. Patients with a dilated (greater than 6 to 7 mm diameter) main pancreatic duct are assumed to have obstructive pathology and are candidates for a drainage-type procedure (lateral pancreaticojejunostomy, Frey procedure). In patients with a small-diameter main pancreatic duct, resection of fibrotic and poorly drained parenchyma is undertaken. Patients with head-predominant or tail-centered disease can undergo a directed partial resection. In patients with diffuse parenchymal involvement a TP with islet autotransplantation may be considered (Algorithm 54-1).
Parenchymal fibrosis associated with CP may involve adjacent organs and lead to complications requiring operative management. Other indications for surgical management of CP include terminal biliary stenosis, duodenal stenosis, gastric variceal hemorrhage due to splenic vein thrombosis, stenosis of the transverse colon, and symptomatic pancreatic pseudocysts. These complications are managed by a variety of bypass or resection procedures depending on the underlying pancreatic ductal disorder. Uncomplicated biliary stenosis is managed with biliary bypass with choledochoduodenostomy or Roux-en-Y hepaticojejunostomy.35 When associated with an inflammatory mass in the head of the pancreas, pancreatic head resection may be indicated. When biliary stenosis is associated with CP and a pseudocyst in the region of the pancreatic head, pseudocyst drainage should be undertaken prior to performing biliary bypass as this may lead to resolution of the obstruction. Duodenal stenosis is usually associated with biliary stenosis and an inflammatory pancreatic head mass and is best managed with resection of the head of the pancreas.36 When patient factors make resection unsafe, a double bypass is undertaken with gastrojejunostomy and biliary bypass. Gastric varices due to splenic vein occlusion are not an indication for operation unless associated with hemorrhage. When indicated for gastric variceal bleeding complications, splenectomy is indicated. Preoperative splenic artery embolization or balloon occlusion may diminish intraoperative blood loss when splenomegaly and fibrosis in the region of the pancreatic tail make operative control of the splenic artery problematic.37 Fibrosing stenosis of the transverse colon, a rare complication of CP, is managed with colonic resection and anastomosis or colostomy, depending on the condition of the patient and the condition of the pancreas.38 Pancreatic pseudocysts associated with CP and ductal obstruction are managed by addressing the underlying ductal disorder with resection or drainage procedures.39
Retrograde pancreatic drainage for relapsing pancreatitis was described by Puestow and Gillesby in 1958.40 A modification of this original drainage procedure that more closely resembles the modern-day technique of the lateral pancreaticojejunostomy (LPJ) was reported by Partington and Rochelle in 1960.41 LPJ is the classic operation for pancreatic drainage and entails opening the pancreatic duct anteriorly along its length, medially to the level of the gastroduodenal artery and laterally into the tail. The opened pancreatic duct is then cleared of stones, including into the head, and anastomosed to a Roux-en-Y jejunal limb for drainage.
Procedure-specific complications of note include intraoperative hemorrhage (due to splenic vein or gastroduodenal artery injury), postoperative hemorrhage (often from the gastroduodenal artery), and anastomotic leak (seen in 10% of cases).
Multiple retrospective single-institution case series have been reported while evaluating outcomes with LPJ, with pain-relief rates of 48% to 91%.42–49 Morbidity rates are low (20% on average) and endocrine and exocrine function is often preserved.50 LPJ is an effective and safe procedure for pain relief in many patients with dilated duct pancreatitis. Recurrent pain does occur after LPJ, however, likely due to disease in the head of the pancreas. Intraductal stone disease in the head of the pancreas can be cleared with intraoperative pancreatoscopy and electrohydraulic lithotripsy, which has been shown to improve outcomes (reduced readmissions, increased pain-relief rates).51 Alternatively, combining a localized head resection with LPJ can help to reduce recidivism.
Localized Pancreatic Head Resection with Lateral Pancreaticojejunostomy
In 1987, Frey and colleagues described a localized pancreatic head resection with a lateral pancreaticojejunostomy (LR-LPJ) with the goal of achieving pancreatic ductal drainage, resection of damaged and poorly drained parenchyma in the head of the pancreas, and preservation of the duodenum to minimize postoperative gastrointestinal dysfunction. The Frey procedure combines a classic longitudinal ductotomy of the neck, body and tail of the pancreas with unroofing of the pancreatic ducts in the head and uncinate process of the pancreas with a “coring” out of the overlying ductal tissue and preservation of the pancreas parenchyma along the posterior and lateral margin of the pancreas. Frey reported initially on 50 patients, describing a morbidity of 22% and a pain-relief rate of 84%.52 His outcomes have been validated in modern series, both at his own institution and at others, with pain-relief rates of 62% to 88% and morbidity of 20% to 30% reported.53–56
Algorithm 54-1. Algorithm for operative decision-making in chronic pancreatitis. LPJ, lateral pancreaticojejunostomy; LR-LPJ, local pancreatic head resection with lateral pancreaticojejunostomy; PD, pancreatoduodenectomy; DPPHR, duodenal-preserving pancreatic head resection; DP, distal pancreatectomy; TPIAT, total pancreatectomy with islet autotransplantation.