- Top of page
- Pathophysiology of pain in chronic pancreatitis
- Management of chronic pancreatitis pain
Background Pain in chronic pancreatitis chronic pancreatitis is a frustrating and challenging symptom for both the patient and clinician. It is the most frequent and most significant symptom. Many patients fail the currently available conservative options and require opiates or endoscopic/surgical therapy.
Aim To highlight the pathophysiology and management of chronic pancreatitis pain, with an emphasis on recent developments and future directions.
Methods Expert review, utilizing in addition a comprehensive search of PubMed utilizing the search terms chronic pancreatitis and pain, treatment or management and a manual search of recent conference abstracts for articles describing pain and chronic pancreatitis.
Results Pancreatic pain is heterogenous in its manifestations and pathophysiology. First-line medical options include abstinence from alcohol and tobacco, pancreatic enzymes, adjunctive agents, antioxidants, and non-opiate or low potency opiate analgesics. Failure of these options is not unusual. More potent opiates, neurolysis and endoscopic and surgical options can be considered in selected patients, but this requires appropriate expertise. New and better options are needed. Future options could include new types of pancreatic enzymes, novel antinociceptive agents nerve growth factors, mast cell-directed therapy, treatments to limit fibrinogenesis and therapies directed at the central component of pain.
Conclusions Chronic pancreatitis pain remains difficult to treat. An approach utilizing conservative medical therapies is appropriate, with more invasive therapies reserved for failure of this conservative approach. Treatment options will continue to improve with new and novel therapies on the horizon.
- Top of page
- Pathophysiology of pain in chronic pancreatitis
- Management of chronic pancreatitis pain
Chronic pancreatitis is a heterogeneous and complex disease. While we traditionally define acute pancreatitis as an event from which the pancreas recovers completely and chronic pancreatitis as a condition characterized by permanent and irreversible damage, this distinction is not accurate and the two conditions seem to be part of a single spectrum. Chronic pancreatitis often (perhaps always) evolves from episodes (clinical or subclinical) of acute pancreatic injury and the transition from acute to chronic pancreatitis may be subtle or even unrecognizable.
Pain may initially be episodic early in this process and the evolution and character of pain are highly variable. Pain may be intermittent, constant or continuous with superimposed acute flares. In some patients with chronic pancreatitis, pain can occur in discrete episodes with or without identifiable pancreatic inflammation (elevations in amylase and/or lipase with evidence of pancreatic or peripancreatic inflammation on abdominal computed tomography), with complete interval resolution. Others experience a chronic, daily, dull, epigastric ache, punctuated by severe exacerbations that may require hospitalization even in the absence of identifiable inflammation.
Pain may manifest in a multiplicity of forms, ranging from those patients with little or no pain to those with overwhelming, continuous and severe pain. The classic character of pain is a constant, severe, dull ache in the mid-epigastrium worsened by high-fat foods, which radiates between the shoulder blades and lasts for days at a time. This classic pain pattern is not universal, and the location, character and quality of pain can be quite variable. Generally, chronic pancreatitis is a condition that evolves slowly over time and pain may be a prominent early feature; at a time when the gland appears relatively normal on imaging studies and neither exocrine nor endocrine insufficiency has developed. Patients in this situation may have rather severe pain, but lack the features of pancreatic damage (structural or functional) that allow one to make a diagnosis of chronic pancreatitis using currently available diagnostic tests. In other words, patients lie on a spectrum ranging from very early painful disease (so called ‘minimal change’, or ‘small duct’ CP) with relatively preserved physiology to end stage disease with very little endocrine or exocrine function. In natural history studies, the progression from this early stage chronic pancreatitis to end stage disease with steatorrhoea and diabetes is about 20 years.1 The pain of chronic pancreatitis may remain stable, worsen or improve over this prolonged period of time. This may lead to a tendency during this time, on the part of patients and clinicians, towards complacency. This complacency may have been propagated by some studies that showed that the pain of chronic pancreatitis has a tendency to lessen with time;2–4hence, the adage that pain will probably ‘burn out’. Not all studies demonstrate this phenomenon, however, and it is not possible to predict in an individual patient if pain will diminish, remain stable or intensify over time. Most current theories of the pathophysiology of chronic pancreatitis postulate that repeated episodes of inflammation and pancreatic injury drive the process within the gland towards irreversible injury. This process is linked to peripheral and central nociceptive nerve sensitization, which play an important role in producing a chronic pain state that is self-perpetuating. Once the disease has advanced and these pathophysiological processes are firmly established, the management of pain becomes much less effective. It is therefore important to intervene as early as possible to prevent this progression, if possible, and complacency in management under the assumption that pain will ‘burn-out’ on its own is misguided.
The causes of chronic pancreatitis are diverse. While that is not the subject of this review, recognition of the cause of chronic pancreatitis in an individual is an important step in identifying possible interventions to reduce the progressive pancreatic damage and in that way mitigate, as much as possible, the pain of chronic pancreatitis. In the most common cause of chronic pancreatitis in the Western World, alcohol abuse, which accounts for 70% of cases, cessation of alcohol can slow the progression of disease5 and have some beneficial effect on pain.6–8 In chronic pancreatitis due to hypertriglyceridaemia, maintenance of triglycerides in a more normal range would be expected to reduce the chance of repeated painful attacks and ultimately the chance of chronic pancreatitis. In postnecrotic chronic pancreatitis due to severe gallstone pancreatitis, prevention of recurrent choledocholithiasis could minimize further damage to the pancreas. In all patients with recurrent pancreatitis, pancreatic duct strictures may occasionally develop and cause progressive injury to the upstream gland; recognition and management of this condition could mitigate the severity of pain. Patients with hypercalcaemia-induced pancreatitis due to hyperparathyroidism should be appropriately treated. In some causes of chronic pancreatitis, such as that due to genetic mutations, no specific therapy directed at the underlying cause may currently be possible. In all forms of chronic pancreatitis, however, it appears that smoking worsens the injury to the pancreas and may be enough, in and of itself, to cause chronic pancreatitis.9 Avoidance of tobacco would be expected therefore to have some long-term benefit on the pancreas outside the more obvious benefits of avoiding lung injury and lung cancer.
All these interventions have two important goals. The first is to reduce the recurrent injury to the pancreas. An acute episode of inflammation (acute pancreatitis) may or may not eventually lead to chronic pancreatitis. With repeated episodes, it becomes increasingly likely that the inflammatory milieu within the pancreas will shift towards chronic inflammation, with the activation of pancreatic stellate cells, fibrinogenesis and irreversible pancreatic damage.10–12 The second goal is one that seems very likely, but is not yet well established is that this reduction in pancreatic damage will ultimately be translated into a slowing of the natural history of disease, less exocrine and endocrine insufficiency and most importantly less severe abdominal pain.
Management of chronic pancreatitis pain
- Top of page
- Pathophysiology of pain in chronic pancreatitis
- Management of chronic pancreatitis pain
The approach to chronic pancreatitis pain first concentrates on making the correct diagnosis. This is not difficult in many patients with longstanding or advanced disease, as they usually have obvious structural damage to the pancreas visible on widely available tests such as computerized tomography (CT) or magnetic resonance imaging with magnetic resonance cholangiopancreatography. Making a correct diagnosis may be much more challenging and difficult in those with ‘small-duct’ chronic pancreatitis who by definition lack these obvious changes. In this setting, tests such as endoscopic ultrasonography and direct pancreatic function testing (e.g. secretin test) are usually used. These tests are more likely to be able to detect abnormalities in pancreatic structure or function at an earlier time point, but they may not be available to many clinicians. Although not the subject of this review, making a correct diagnosis is critical (Table 2). In our pancreas clinic, we frequently see patients with a chronic pain syndrome who have been incorrectly labelled as having chronic pancreatitis, who after further testing turn out to have a nonpancreatic source of pain. Treating this type of patient with therapies that may carry risk (chronic opiates, endoscopic or surgical therapy) is obviously not appropriate in the absence of a secure diagnosis.
Table 2. Diagnosis of chronic pancreatitis
|Diagnostic test||Possible findings in ‘big-duct’ disease||Findings in ‘small-duct disease’|
|Faecal elastase||Usually low (<100 μ/g of stool)||Usually normal|
|Serum trypsin||Usually low (<20 ng/mL)||Usually normal|
|Abdominal ultrasonography||Pancreatic atrophy, pancreatic duct dilation, pancreatic calcifications, pseudocyst||Usually normal|
|Computerized tomography||Pancreatic atrophy, pancreatic duct dilation, pancreatic calcifications, pseudocyst||Usually normal or equivocal|
|MRI with MRCP||Pancreatic atrophy; pancreatic duct dilation, irregularity or stricture; pancreatic calcifications, pseudocyst||Usually normal or equivocal|
|Endoscopic ultrasonography||Abnormal (>4 features of chronic pancreatitis)||May be abnormal|
|ERCP||Abnormal||Normal or minimally abnormal|
|Direct hormonal stimulation test (e.g. secretin test)||Abnormal||Usually abnormal|
Once the diagnosis of chronic pancreatitis has been made, most patients are initiated on a trial of medical therapy. There are several important components of medical therapy. In those who drink alcohol, it is important to counsel against its use. Even if the disease is not due to alcohol, it is an appropriate recommendation to make as limiting any further damage to the gland is an important goal of medical therapy. In those with alcohol-induced chronic pancreatitis, cessation of alcohol slows the progression of disease and prolongs survival5 and relieves pain in some patients.6, 7 In addition to stopping alcohol, we strongly encourage patients to stop tobacco as there is mounting evidence that smoking is an important co-factor for chronic pancreatitis.11, 12
Most patients will require some analgesics for pain control. There is a risk of narcotic addiction in these patients, but the focus should be on pain improvement and not solely on the risk of dependence. We start with the least potent agent (usually tramadol or propoxyphene) and establish at the outset that the goal is significant improvement in pain but not necessarily complete absence of pain. Many patients will require more potent narcotics, although there is evidence that when used appropriately, tramadol can be equivalent to more potent agents with fewer negative effects on gut motility.57, 58 In those who require more potent narcotics, it is prudent to consider starting an adjunctive agent. These adjunctive agents have as their goal minimizing the dosage of narcotics required for pain control. They include tricyclic antidepressants, selective serotonin reuptake inhibitors, combined serotonin and norepinephrine reuptake inhibitors (duloxetine) and α2δ inhibitors (gabapentin or pregabalin). These agents have not been sufficiently studied to be able to determine if one or the other is superior in chronic pancreatitis, but they have all been used in a variety of chronic painful conditions and are often worth a therapeutic trial (Table 3).
Table 3. Medical treatment of pain
|Agent||Typical starting dosage||Comments|
|Propoxyphene with acetaminophen (100 and 650 mg respectively)||1–2 po q 8 h||Maximum dose should not exceed 4 g of acetaminophen daily|
|Tramadol (50 mg)||1–2 po q 8 h||Do not exceed 400 mg daily|
|Antioxidants||Typical total daily dose might contain 500–1000 mg vitamin C, 250–300 IU vitamin E, 500–800 μg selenium, 2 gm methionine and 9000–10 000 IU β-carotene|| |
|Tricyclic antidepressants||Amitryptyline: initiate at 25 mg po nocte, increase 25 mg/week as tolerated||Usual maximum of 100–150 mg nocte. Variety of gastrointestinal, cardiac, neurological side effects at high dosage|
|Selective serotonin reuptake inhibitors||Citalopram, fluoxetine, sertraline, paroxetine and others||Initial dose and side effects vary with agent|
|Combined serotonin and norepinephrine re-uptake inhibitors||Duolexitine beginning at 20–40 mg daily, increase to 60 mg daily as tolerated||Side effects rare, may cause nausea, vomiting and constipation|
|Pancreatic enzymes||Non-enteric-coated enzymes only (e.g. Viokase-16 or equivalent at 4 pills with each meal and at night)||Co-treat with H2-receptor antagonist or proton pump inhibitor to prevent degradation of enzymes by gastric acid|
|Octreotide||50–100 μ SQ t.d.s., can increase to 200 μ SQ t.d.s.||Data supporting effectiveness are very limited. May cause abdominal pain, gallstones|
Medical therapy also often involves pancreatic enzyme therapy. The use of pancreatic enzymes to control pain has been the subject of several randomized trials and a meta-analysis. The basis of their use is the ability to degrade CCK-releasing factor in the duodenum and by doing so, lowering CCK levels and through this mechanism, reducing pain. Only non-enteric coated (i.e. tablet) formulations have this activity and studies using this type of enzymes have demonstrated improvement in pain.59, 60 In contrast, studies using enteric coated preparations (which are not active in the duodenum and hence cannot degrade CCK-releasing factor) have not shown similar results.61–63 One study that did show pain relief of enteric coated enzymes used a measurement of pain that included malabsorptive (such as bloating, gas or cramping), rather than more classic pancreatic abdominal pain.64 A meta-analysis65 combining these studies saw no effect of enzymes, although combining the two types of formulations in this analysis is probably not appropriate given the proposed mechanism of action. In our experience, enzyme therapy is most likely to be effective in those with ‘small-duct’ or ‘minimal change’ CP. In these patients particularly, we routinely attempt a trial with a high potency enzyme in tablet form (non-enteric-coated), along with a proton pump inhibitor to ensure that the enzymes survive into the duodenum. The dose is the equivalent of Viokase-16 (Axcan Pharma US, Inc., Birmingham, AL, USA), four tablets with meals and snacks, or q.d.s., if meals are skipped. A therapeutic trial should last at least 6 weeks. In patients with chronic relapsing pain (with discrete episodes of more severe pain), we also adjust the dose of enzymes during painful attacks – the equivalent of one tablet of Viokase-16 every 2 h while awake. This approach appears to allow some of our patients to prevent a flare from requiring hospitalization.
There are additional medical therapies for pain, but they are not well studied and their effectiveness is not known. One of these is antioxidants. The hypothesis of their use is the observation that free radicals may play a role in pancreatic injury. Numerous studies in animal models seem to support the potential efficacy of antioxidant therapy, but the results of human studies are inconclusive.66–70 Nonetheless, they appear to be risk-free and a trial of these agents is not unreasonable. A typical mixture would contain selenium, vitamin C, β-carotene, vitamin E and methionine.
A second medical therapy with insufficient data supporting its use is the agent octreotide, which is a synthetic somatostatin analogue. It is felt to work by suppressing CCK and secretin and hence by suppressing pancreatic secretion, although it may have other effects as well. In experimental models, octreotide has been shown to have anti-inflammatory properties, alter the cytokine milieu and protect pancreatic cells. Octreotide is available in two forms, short acting and depot-form given subcutaneously once a month. Several small studies have reached differing conclusions on its effectiveness in painful chronic pancreatitis. While pilot studies showed a trend in pain relief in advanced chronic pancreatitis71, 72 with an apparent dose response curve,73 other studies have not reached similar conclusions.74 In addition, most of these studies used octreotide in patients with ‘big-duct’ disease, so it is not known if the effect would be similar in ‘small duct’ chronic pancreatitis. We use the drug after other therapies have failed and in patients with few or no other options. In that setting, the short acting form is started first and titrated, at doses of 50–100 μg SQ thrice daily, escalating the dosage up to 200 μg t.i.d. If the patient achieves significant pain relief on the short acting form, we will convert to the long acting form. The long acting form only reaches steady state after 3 months, so the short acting form should be continued for at least 6 weeks (or somewhat after the second injection). The use of octreotide in chronic pancreatitis is controversial. It can also suppress the function of the islets of Langerhans. Depending on the patient, it can lead to either hypoglycaemia or lead to poorer glucose control in diabetics. In addition, it predisposes to biliary stasis. We maintain special vigilance for biliary stones in patients on octreotide. We attempt to reduce this risk with periodic holidays or with periodic use of pro-contractile agents such as macrolides.
In the patient who has failed medical management, one can consider options, which include nerve block, neurolysis, or endoscopic or surgical therapy. Endoscopic and surgical techniques are primarily used in those with a dilated pancreatic duct. These patients tend to be those with longstanding disease and most commonly alcohol-induced chronic pancreatitis. However, neurolysis can be performed irrespective of the size of the main pancreatic duct.
The most common target for nerve block or neurolysis is celiac ganglion. This can be performed with the intent of temporarily relieving pain (a nerve block injecting a combination of steroids and an anaesthetic) or with the intent of permanent neurolysis (injecting absolute alcohol). The procedure can be performed via CT-guided techniques, but techniques using EUS-guidance are safer and more effective. Celiac plexus nerve block by endoscopic ultrasound guidance may also last somewhat longer than percutaneous techniques.75 Although this technique is quite helpful in patients with pancreatic carcinoma, its use in chronic pancreatitis has been less utilized due to the often temporary effect of nerve block. Even when neurolysis is attempted with absolute alcohol, pain may not be permanently relieved, although experience with neurolysis in chronic pancreatitis is limited. In patients with painful chronic pancreatitis, based on small and uncontrolled studies, about half of patients undergoing nerve block will get significant pain relief. Unfortunately, pain relief is often transient, on average 2–4 months.76 Although the effect is transient, EUS-guided celiac plexus block may be used to provide a temporary respite from pain for patients who might benefit from this, and often to allow some temporary downward titration of narcotic analgesics. EUS-guided celiac plexus blocks are quite safe, but there are reports of paraplegia with repeated CT-guided celiac plexus neurolysis.
Another technique to achieve neurolysis is via sectioning of the splanchnic nerves. This can be performed with thoracosocopic, minimally invasive approaches. This can be performed unilaterally or bilaterally and the surgeon can usually gain access to the splanchnic nerve roots from T5 to T10. In short-term follow-up, pain is relieved in up to 80% of patients, but this falls to <50% with even short-term follow-up.77–79 The explanation for this decrease in efficacy may be because the splanchnic nerve roots may extend over up to 10 spinal segments (ganglia), which is far more than can be accessed by thoracoscopy. In addition, central mechanisms of pain may be present and cause pain recurrence by central rather than pancreatic mechanisms. The relatively poor long-term efficacy has led some centres to utilize this approach rarely, if at all.
Endoscopic therapy, in addition to EUS-guided celiac plexus block, is considered when pain continues despite medical therapy and in appropriately selected patients. Endoscopic therapy includes stenting or dilation of pancreatic duct strictures and removal of obstructing pancreatic duct stones. Proper selection of patients requires a detailed understanding of the individual patient’s pancreatic ductal anatomy. Stones can be removed if they are not too large (1 cm is probably the upper limit of size), are not tightly impacted, are relatively close to the tip of the working side-viewing duodenoscope (found in the head of the pancreas) and are not upstream from a stricture (i.e. there is not a stricture between the endoscope and the stone). Adding extracorporeal or intraductal lithotripsy to standard endoscopic techniques allows larger stones to be fragmented into pieces which can be extracted. These endoscopic techniques require substantial expertise.
In one large case series of 1018 patients treated with endoscopic techniques at eight different centres, pain improvement was seen in 65% at 2–12 years of follow-up.80 Over this same time period, one-quarter required surgery for failure of endoscopic therapy. These patients were selected based on an appropriate ductal anatomy, with a mixture of strictures and stones. Therapy required a median of 3 ERCPs, with a range of 1–25.
While lithotripsy may be required to fragment stones to the point that they can be extracted, there are some data that extracorporeal shock wave lithotripsy (ESWL) alone (without subsequent attempts at endoscopic stone extraction) may actually improve pain. In one recent randomized trial, 55 patients were randomized to ESWL alone or combined with endoscopy and stone extraction.81 After 2 years of follow-up, pain relief was equivalent. The explanation for these results is not clear; perhaps ESWL may have some effect on nociceptive neurons separate from its effect on stones.
Surgical therapy is also considered in those patients who fail medical therapy. The modified Puestow or lateral pancreaticojejunostomy is the most commonly performed procedure and is a relatively simple procedure with low morbidity and mortality. In this operation, the pancreatic duct is filleted from the anterior surface, ductal strictures are incised and pancreatic duct stones are removed. The longitudinal incision into the pancreatic duct is carried out from the body of the pancreas to as close to the duodenum as is easily accessible, and this is overlaid with a defunctionalized jejunal limb. This operation conserves the maximum amount of pancreatic tissue. This type of operation requires that the pancreatic duct is dilated enough that it is identifiable at surgery (usually a lower limit of 5–6 mm). Pain relief after this operation is very good, in the range of 80%.82–84 With long-term follow-up, this drops to about 50%.85–88
More complex operations for chronic pancreatitis pain usually involve some resection of the gland, in addition to ductal drainage. A number of variations have been developed, including the traditional pancreaticoduodenectomy (Whipple operation), a limited resection of the pancreatic head (Beger operation), a combined opening of the pancreatic duct and excavation of some of the pancreatic head (Frey operation) and variations of the above.89, 90 These operations involve some resection of the pancreatic head, often in addition to drainage of the pancreatic duct. There is a common although unsupported belief amongst surgeons that the pancreatic head may be the ‘pacemaker’ of the disease. In addition, in certain situations, resection of the head may be useful to treat concomitant duodenal or bile duct obstruction caused by enlargement and/or fibrosis in the head of the pancreas. These operations involving resection seem to have equivalent short-term pain relief, compared to a Puestow operation, but better long-term pain relief.91 These operations do have somewhat higher morbidity as well compared with simple drainage procedures. They are used as the primary surgical approach in some countries, but tend to be less utilized in the US. This is primarily because most US surgeons have not been trained in these techniques.
There have now been some randomized comparisons of endoscopic and surgical therapy for painful chronic pancreatitis. In the first study,92 72 patients were randomized to surgical or endoscopic therapy. Patients with pancreatic duct strictures or obstructing stones were included. Of 140 eligible patients, only 72 agreed to participate. The analysis of pain relief in the randomized groups showed the two approaches to be equivalent at 1 year of follow-up. At 5 years, pain relief was significantly better in the group randomized to surgery: 34% of the surgical group experienced complete absence of pain vs. 14% of the endoscopic group. Partial relief of pain was seen in 52% of the surgery group compared to 46% of the endoscopic group. The surgical group was also noted to have a statistically better outcome in body weight at the end of both 1 year and 5 years, without an increase in diabetes. This study has been criticized in that the endoscopic techniques were less than might be considered optimal (only half received a pancreatic stent, stone extraction was carried out in 23%) and the surgical therapy was perhaps more aggressive than might be typical (80% underwent resection including pancreatic head resections and only 20% had a simple drainage operation). Nonetheless, this study seemed to demonstrate superior durability for surgery. A second randomized trial reported on 39 patients randomized to endoscopic or surgical therapy.93 The endoscopic therapy in this study was more complex and included ESWL if needed and the use of larger calibre pancreatic duct stents. The surgical therapy was a lateral pancreaticojejunostomy (modified Puestow operation). The study was ended early after an interim analysis revealed the superiority of surgery over endoscopic therapy. At 24 months of follow-up, the surgical group had better quality of life and less pain (complete or partial pain relief in 75% of surgical patients, compared to 32% of endoscopic patients). This study has also been criticized due to the relatively short follow-up. While endoscopic therapy may still be preferable for many patients who wish to avoid surgery, we now notify patients of these study results and discuss the options in light of these studies. Many patients still choose endoscopic therapy first.
One surgical therapy which has not yet been discussed is total pancreatectomy, usually coupled with harvesting of islet cells and autotransplantation. This is a particularly attractive option in patients who have a high risk of pancreatic cancer, such as those with hereditary pancreatitis. Total pancreatectomy is available at only a few highly specialized pancreatic centres in the US. Pain relief is not universal with these procedures94 and up to 50% may still require narcotics at 1 year after these procedures. Proponents point out that having half of the most intractable chronic pancreatitis patients off of narcotics should be considered a clinically significant success. The risk of diabetes is largely dependent on the number of islets that can be harvested. Some patients do not require insulin therapy after total pancreatectomy with islet cell autotransplantation, but diabetes will occur in many of these patients and may be brittle and difficult to manage. The yield of islet cells available for autotransplantation may be improved if the patient has not previously undergone pancreatic surgery.95, 96 It is the rare patient with chronic pancreatitis, outside of those with cancer risk, who has such severe and intractable symptoms to consider this option.
In addition to these management options for pain, it is important to not forget that in a patient with chronic pancreatitis, pain may occur from a complication of the disease (pseudocyst, duodenal obstruction, superimposed carcinoma), an associated condition (in particular gastroparesis or narcotic bowel) or an unrelated coexistent disease. Several of these have specific therapy and are worth considering in all patients with chronic pancreatitis experiencing pain. In addition, one should not neglect the management of exocrine or endocrine insufficiency in these patients while focusing too intently on pain. The foregoing discussion also serves to illustrate that current options for managing pain are imperfect. In the future, new therapies may be able to take advantage in advances in our understanding of nociception and target molecules, receptors and mechanisms involved in pain perception.