Article Text

Sphincter of Oddi dysfunction and acute pancreatitis
  1. J W C CHEN,
  1. Gastrointestinal Surgical Unit,
  2. Department of Surgery,
  3. Flinders Medical Centre,
  4. Bedford Park, Adelaide,
  5. SA 5042, Australia
  1. Professor Toouli (email:Jim.Toouli{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Fifty years ago Lium et al stated that... “Acute pancreatitis is the result of ductal obstruction in an actively secreting pancreas”. A number of mechanisms for pancreatic ductal obstruction were reviewed and the role of the sphincter of Oddi (SO) in producing “obstruction” was discussed.1 The studies by Archibald were reported and suggested a possible role of SO “spasm” in producing ductal obstruction. It was suggested that the cause of biliary pancreatitis in patients with gallstones who did not have a stone impacted at the ampulla may be secondary to SO “spasm” and oedema.

Sphincter of Oddi dysfunction

Normal human SO motility has been characterised by a number of studies and normal manometric parameters have been established using standardised manometry, as has SO dysfunction. The clinical presentation of patients with SO dysfunction may be divided into two groups: biliary SO dysfunction presenting with biliary type pain; and recurrent pancreatitis.2 In 1995, the following definition of SO dysfunction was developed: “partial obstruction of the SO biliary segment giving rise to intermittent, episodic upper abdominal pain, deranged liver function tests, dilatation or delayed drainage of injected contrast from the common bile duct. Likewise, similar condition of the pancreatic segment can give rise to pancreatitis or episodic pain suggesting a pancreatic origin”.3 The main consequence of SO dysfunction is impedance of bile and pancreatic juice flow, either through a structural stenosis or functional stenosis from hypertonia. In some instances, however, hypotonia as a result of dyskinesia may also occur and reflect SO dysfunction.

The term SO dysfunction includes structural stenosis of the SO and functional stenosis secondary to hypertonia.4 It also encompasses dysmotility which may lead to intermittent or transient impedance of bile or pancreatic juice flow. The frequency of SO dysfunction in the community is not known. It is also uncertain whether transient SO dysfunction secondary to food, drugs or systemic infection is common. The definition of SO dysfunction to date has not included the dimension of time. It is uncertain whether transient SO dysfunction occurs and whether it results in symptoms or biochemical abnormality.

Biliary sphincter of Oddi dysfunction

Much of the initial understanding of SO dysfunction was related to patients presenting with recurrent biliary type pain following cholecystectomy. In a series of studies reported in the 1980s, manometric abnormalities have been defined and assessed in determining which patients will respond to treatment by sphincterotomy. We have subdivided the patients with biliary SO dysfunction into two groups according to SO manometry: patients with manometric SO stenosis and patients with SO dyskinesia.4

Patients with SO stenosis characteristically have an abnormally increased SO basal pressure (>40 mm Hg) recorded on manometry (fig 1). This manometric abnormality is reproducible, and does not relax with administered pharmacological agents.4 A prospective randomised study which evaluated the relation between manometry, SO dysfunction and clinical outcome following endoscopic sphincterotomy has shown that this group of patients will respond well to sphincterotomy and experience long term cure or a noticeable reduction in their symptoms.5 Other non-invasive investigations do not correlate as well as manometry with a successful treatment outcome, consequently manometric diagnosis remains the gold standard in selecting patients for treatment (table 1).

Figure 1

Sphincter of Oddi manometry is recorded using a triple lumen catheter introduced via the biopsy channel of a duodenoscope. A normal pressure profile is illustrated in the upper tracing, whereas the lower tracing illustrates a recording from a patient with sphincter of Oddi stenosis. CBD, common bile duct; PD, pancreatic duct.

Table 1

Sphincter of Oddi pressures

Patients with SO dyskinesia are characterised by a number of manometric findings: an excess of retrograde contractions (>50%), rapid contraction frequency (>7/min) and a paradoxical response to administration of cholecystokinin-octapeptide (CCK-8).4 5These manometric findings are not reproduced well by repeated study and may reflect the intermittent nature and the methodological difficulty of SO manometry in that it only “samples” a short period of SO motility. Treatment of this group of patients by either sphincterotomy or pharmaceuticals has not been associated with lasting good results, and thus awaits further investigation.

Acute pancreatitis

The most common causes of acute pancreatitis are gallstones and alcohol. Other causes are less frequent and include hypercalcaemia, hyperlipidaemia, viral infections, drugs, and the more exotic aetiologies such as scorpion envenomisation and organophosphate poisoning. Recent studies suggest that the SO may be involved in the development of pancreatitis. However, the underlying mechanisms have not been defined. Recently reported clinical studies suggest that SO dysfunction may be an aetiological factor in recurrent pancreatitis, hence strengthening the hypothesis that SO motility is involved in its pathogenesis.5

The sphincter of Oddi and pancreatitis

Ever since Opie first described the association of impacted gallstones with acute pancreatitis in 1901, the trigger mechanism remains a matter for debate.6 It is accepted that passage of gallstones causes acute pancreatitis and that obstruction is the likely trigger. SO “spasm” resulting in pancreatic duct obstruction in gallstone pancreatitis was first suggested by Archibald in 1913.1 However, it is not known why “spasm” occurs in some patients whereas in others, stones may pass without causing ill effects.

In one study, unlike in the controls, the sphincter of Oddi of patients with biliary pancreatitis did not relax following injection of the CCK analogue, ceruletide. It was postulated that this group of patients had SO dysfunction predisposing them to gallstone pancreatitis.7 In another study, T tube cholangiography revealed SO stenosis/obstruction in patients with biliary pancreatitis in the absence of common bile duct stones, suggesting that oedema or SO dysfunction may be contributing to biliary pancreatitis.8

Idiopathic recurrent acute pancreatitis

Evidence that the SO is involved in the aetiology of human pancreatitis is lacking. However, indirect evidence of an association between SO dysfunction and pancreatitis is increasing. In patients undergoing surgery for idiopathic recurrent pancreatitis, surgeons have shown using small probes that the SO narrows at the opening of the pancreatic duct.9 The morphine prostigmine test has been used to show an association between SO dysfunction and abdominal pain associated with pancreatitis.10 Nardi and Acosta postulated that there is an association between SO motor abnormalities and the development of pancreatitis.11

In a manometric study, we showed an association between SO dysfunction and a proportion of patients with idiopathic recurrent pancreatitis.5 In this series of 28 patients, one or several SO manometric disorders were detected in 25, the most common of which was an abnormally raised SO basal pressure, reflecting SO stenosis. However, in addition, the patients also had SO dyskinesia, characterised by excessive retrograde contractions, rapid contraction frequency and paradoxical response to CCK-8. The manometric findings in patients with idiopathic recurrent pancreatitis suggested that impedance of flow of pancreatic secretions may produce pancreatitis.

Real time ultrasonographic studies of the diameter of the pancreatic duct after induction of pancreatic secretion by secretin support this observation.12 After infusion of secretin (1 unit/kg), the diameter of the pancreatic duct was monitored: 83% of patients with SO stenosis and 72% with stenosis of the accessory papilla showed pancreatic duct dilatation of ⩾1 mm compared with controls. This dilatation response was abolished after surgical sphincteroplasty. A positive secretin test was associated with good operative outcome in 90% of patients and was thought to be of good predictive value.

The results of these studies suggesting pancreatic SO stenosis have led to treatment aimed at relieving obstruction. Surgical division of the SO by transduodenal sphincteroplasty and septoplasty has been reported.9 13 Early studies were associated with mixed results mainly as a result of the types of patients selected for treatment. Although some patients were cured after sphincter division, a high proportion continued to have symptoms and episodes of pancreatitis. These treatment failures were largely attributed to the inhomogeneity of the patient population, which included patients with alcoholic pancreatitis and those who were heavy narcotic users. In a more recent study which used SO manometry to select patients for treatment, those without a history of alcohol or narcotic drug addiction and manometric SO stenosis were treated by open sphincteroplasty and septoplasty.5 Medium to long term follow up has shown that in over 90% of these patients, episodes of recurrent pancreatitis have stopped. This finding led to the conclusion that transduodenal sphincteroplasty and septoplasty should be offered to patients with recurrent idiopathic pancreatitis if SO manometry reveals manometric stenosis.

Similar results have been found in patients with the congenital anomaly of pancreas divisum. The real time ultrasound investigation of pancreatic duct size was used to select patients for surgical treatment. A high correlation was found between an abnormal investigation and clinical outcome.14

Endoscopic stenting of the pancreatic duct has also been used either to treat or to select patients for surgical treatment.15 16Results of long term outcome are not available. It has been suggested that complete division of the SO may be achieved via an endoscopic rather than an open surgical approach. The potential difficulty of the endoscopic approach is the safety of an adequate sphincterotomy. In open surgery, in order to achieve successful ablation of the pancreatic SO, division is extended beyond the duodenal wall. This opening is subsequently closed by suture apposition, hence avoiding the complication of “duodenal perforation”. In order to achieve a similarly effective sphincteroplasty and septoplasty via the endoscope approach, one would need to determine how perforation can be avoided.

Endoscopic retrograde cholangiopancreatography (ERCP) and SO manometry have been associated with production of acute pancreatitis. In many instances, iatrogenic or technical factors such as excessive injection of contrast medium or oedema produced by multiple attempted cannulations are the cause of pancreatitis. However, in some patients pancreatitis seems to relate to a susceptibility to production of SO “spasm” or dysmotility. This conclusion is supported by circumstantial evidence which shows a higher incidence of pancreatitis in patients with idiopathic recurrent pancreatitis undergoing either ERCP or SO manometry.17

SO dysfunction has been implicated as a possible factor in chronic pancreatitis with several studies showing increased SO basal and pancreatic duct pressure in patients with this condition. There is no evidence to suggest whether these SO manometric abnormalities were a cause or result of chronic pancreatitis. The results of sphincterotomy to correct these abnormal SO pressures have been mixed.

Cholinergic stimulation of the pancreas and the SO results in both increased pancreatic secretion and increased SO activity in animal models.18 It has been shown that excessive cholinergic stimulation using an acetylcholine agonist can result in acute pancreatitis. Organophosphate (Diazinon) used as an insecticide irreversibly inhibits cholinesterase resulting in delayed breakdown of synaptic acetylcholine and has been noted to cause acute pancreatitis in humans.19 In animal models Diazinon results in acute pancreatitis associated with raised pancreatic duct pressure.20 This is thought to be secondary to “obstruction” at the SO level coupled with cholinergic stimulation of pancreatic secretion.

Scorpion envenomisation is a known cause of acute pancreatitis.21 Evidence suggests that scorpion toxin releases acetylcholine from cholinergic nerves, leading to stimulation of the pancreas and SO, resulting in a secretion–obstruction block similar to organophosphate poisoning.22 SO dysfunction secondary to excessive cholinergic stimulation is likely to impede increased pancreatic secretion, thereby causing acute pancreatitis.

Hypercalcaemia has been shown to stimulate pancreatic secretion in animal models23 and is also a well known cause of acute pancreatitis. The pathophysiological mechanism underlying hypercalcaemia induced pancreatitis is not known. The presence of high extracellular calcium has been shown to stimulate smooth muscle and it is possible that abnormal calcium regulation in SO smooth muscle may play a role in this type of acute pancreatitis.

Octreotide, a somatostatin analogue, is used in various pancreatic disorders to shut down pancreatic exocrine secretion. Recently, the effect of octreotide on SO activity has been studied in both humans and animals. This has shown a stimulatory effect resulting in increased SO motility and impaired pancreatic flow into the duodenum.24 Indeed, cases of acute pancreatitis following octreotide administration have been reported and these are postulated to be secondary to SO dysfunction caused by octreotide.25 26

Hyperlipidaemia is thought to be associated with acute pancreatitis. However, the role of the SO has not been studied. A recent case report suggests a possible association with hyperlipidaemia and the nitric oxide pathway causing SO dysfunction and pancreatitis.27This is supported by a study showing failure of SO relaxation in hypercholesterolaemic rabbits.28


The role of the SO in the production of pancreatitis is supported by some direct evidence, but mainly by a large body of circumstantial evidence. Abnormal motility of SO (SO dysfunction), either as a primary event or secondary to triggering factors, seems to be the mechanism underlying the production of obstruction–secretion induced pancreatitis and theoretically if repeated over a number of years, may result in permanent dysfunction of the sphincter. Manometric studies can be used to identify patients who will respond to treatment requiring total ablation of the SO. Division of the SO via transduodenal sphincteroplasty and septoplasty is associated with excellent results in patients with recurrent pancreatitis who have been shown to have SO manometric stenosis and no evidence of alcoholic pancreatitis or addiction to opiates. Laboratory and human studies, currently underway, may determine the underlying mechanism for SO dysfunction and may lead to treatment which can influence the induction of pancreatitis.