- Top of page
Acute pancreatitis is characterized by the sudden onset of abdominal pain associated with elevated blood amylase and/or lipase activities. Among the various aetiologies about two-thirds of the cases can be attributed to biliary tract disease or ethanol abuse.1, 2 Sphincter of Oddi dysfunction, defined by a sphincter baseline pressure ≥ 40 mmHg, was found at endoscopic sphincter of Oddi manometry in 15–72% of patients with acute recurrent pancreatitis of otherwise unknown origin (so called ‘idiopathic’ recurrent pancreatitis).3–13 Perfusion manometry of the pancreatic duct, however, is associated with a substantial risk to induce acute pancreatitis (up to 29%) especially when patients with presumed sphincter of Oddi are studied.12, 14–17 Only after the introduction of modified techniques, reducing the risk of procedure-related pancreatitis, has manometric evaluation of the pancreatic sphincter segment become acceptable.17–20 The only proof that sphincter of Oddi is the cause of recurrent pancreatitis in these patients is cure from further episodes after sphincter ablation. A favourable outcome had been reported in about 73–94%, after both surgical and endoscopic sphincterotomy.4, 5, 8, 10, 11, 21, 22 Post-interventional pancreatitis, however, was reported in up to 22% of the cases after endoscopic pancreatic sphincterotomy in this situation.8, 10, 11, 23–25
Botulinum toxin, a potent inhibitor of acetycholine release from pre-synaptic, cholinergic nerve endings, significantly reduces the basal sphincter of Oddi pressure after local injection into the papilla of Vater.26 In a recent prospective study we found that botulinum toxin injection into the papilla was safe and effective in 22 patients with biliary sphincter of Oddi.27 This investigation evaluates whether this also applies to the patients with pancreatic sphincter of Oddi (who are at a higher risk for procedure-related pancreatitis) and whether botulinum toxin injection may serve as a pilot therapy identifying those pancreatic sphincter of Oddi-patients who will benefit from second line endoscopic pancreatic sphincterotomy.
- Top of page
Fifteen consecutive patients (six male, nine female, mean age 38.1 ± 12.3 years, range 26–55 years) with at least two attacks of acute pancreatitis (abdominal pain associated with elevation of the serum amylase and lipase activities > 3-fold above normal) during the last 6 months were enrolled. CT-scans performed in nine of them during their pancreatitis episodes had revealed oedematous pancreatitis in all. In this group of referred patients, pancreatitis was neither attributed to gallstone disease nor to alcohol. Pancreatic sphincter of Oddi dysfunction (basal sphincter of Oddi pressure at least in the pancreatic sphincter portion ≥ 40 mmHg) was proven by manometry in all of them. Additional biliary sphincter hypertension was documented in eight out of 15 patients. The detailed clinical and manometric data of the 15 patients studied are given by Table 1.
Table 1. Clinical and manometric data of 15 patients with acute recurrent pancreatitis due to pancreatic sphincter of Oddi dysfunction
These 15 patients were a selection from 42 patients who were referred to our department for evaluation of suspected pancreatic sphincter of Oddi dysfunction during a 5 year period. The 15 patients actually enrolled suffered from recurrent pancreatitis episodes with a median of four attacks of acute pancreatitis during the last 6 months prior to study entry (Table 1). Furthermore, seven of the 15 patients also experienced recurrent epigastric pain throughout their pancreatitis-free intervals.
All patients had normal liver tests and a normal bile duct diameter at endoscopic retrograde cholangiography (ERC). Common bile duct stones had been routinely excluded by ERC and additionally by intra-ductal ultrasonography (20 MHz) and light microscopic bile sample analysis (for detection of microlithiasis) in six out of 15 patients of whom four had been formerly cholecystectomized because of gallstones. None of the patients had undergone sphincterotomy or other surgical interventions at the bilio-pancreatic tract (except cholecystectomy) before study entry. ERP revealed a normal pancreatography in 11 patients but showed a slight dilation of the pancreatic duct in the remaining four cases (Table 1), without any further evidence for chronic pancreatitis (no pancreatic duct irregularities). Pancreas divisum had been excluded by ERP in all. None of the patients had evidence for either exocrine (normal stool weight, faecal fat and elastase concentrations during 3-day analysis, n=12, normal pancreolauryl test, n=5) or endocrine (oral glucose tolerance test) pancreatic insufficiency. Additional imaging procedures (abdominal ultrasonography, computed tomography, endoscopic ultrasound, n=4, magnetic resonance tomography, n=3) did not demonstrate any further pathology.
Endoscopic botulinum toxin injection
After an overnight fast, a standard side-viewing duodenoscope (JF 100 or TJF 130/140; Olympus Optical, Hamburg, Germany) was inserted under adequate sedation with i.v. propofol or benzodiazepines.28 After positioning the endoscope in front of the papilla of Vater, 100 mouse units of botulinum toxin (Botox, Pharma Merz, Frankfurt/Main, Germany), reconstituted in 1 mL of isotonic saline, were injected into the intra-duodenal sphincter of Oddi segment along the general direction of a virtual pancreatic sphincterotomy. For botulinum toxin administration the needle was inserted into the upper margin of the papillary orifice at the one o’clock position, injecting botulinum toxin as a single deposit through a standard sclerotherapy needle (0.5 mm diameter, 5 mm needle length). Because of the small injection volume, we do not split the total botulinum toxin dose into multiple deposits.27 Cannulation of the papilla was not intended and the procedure was performed on an out-patient basis with clinical monitoring at the endoscopy unit for 4 h after the procedure. The patients remained without oral intake overnight and they were routinely interviewed for side-effects on the evening of botulinum toxin injection and on the following morning, by telephone calls. If abdominal symptoms occurred during the 4 h observation period, the patients were admitted to our hospital and clinically followed overnight with blood samples (blood cell count, liver enzymes and serum amylase and lipase) taken 24 h after the endoscopic botulinum toxin injection. A post-botulinum toxin injection pancreatitis was defined according to published criteria.29
Clinical and manometric follow-up after botulinum toxin treatment
After botulinum toxin injection the patients were clinically re-visited at monthly intervals. In case of any abdominal pain after botulinum toxin injection, the patients were clinically re-investigated and at least serum amylase and lipase activities were measured.
If recurrence of acute pancreatitis was documented during a 3-month period after botulinum toxin treatment (botulinum toxin non-responder) or during the later course (botulinum toxin responder), endoscopic sphincter of Oddi manometry was repeated (the manometric recordings were analysed blindly with respect to the former manometric data and the duration of the symptomatic response to botulinum toxin) and selective pancreatic sphincterotomy was performed over a guide-wire, irrespective of the manometric finding.11, 17 Further, a straight 7 French pancreatic stent was inserted for several days in order to reduce the risk of post-sphincterotomy pancreatitis.30 If biliary sphincter hypertension was known from the index manometry before botulinum toxin treatment, a biliary sphincterotomy was added in the same session.
The presence of abdominal pain alone (without laboratory abnormalities) was not used to define the response to botulinum toxin treatment.
After sphincterotomy, all patients underwent further prospective clinical follow-up with symptom evaluation, physical examination and laboratory data at bi-monthly intervals. The patients were instructed to contact us immediately if severe abdominal pain recurred. Recurrence of acute pancreatitis during follow-up after sphincterotomy served as the clinical end-point of the study.
All data are given as median and ranges and/or as mean ± s.d. Statistical analysis was performed with the Wilcoxon signed rank test or the exact Fisher-test whenever appropriate. A P < 0.05 was considered significant.
The study protocol had been approved by the Ethics Committee of the University of Frankfurt and written informed consent was obtained from all patients.
- Top of page
Our prospective clinical trial evaluated both technical feasibility and safety, as well as efficacy of endoscopic botulinum toxin injection for pancreatic sphincter of Oddi dysfunction in patients with recurrent pancreatitis.
Evidence for a key role of sphincter of Oddi in the aetiology of idiopathic recurrent pancreatitis is increasing and fostered by different methodological approaches. Studies with either real time ultrasonography or magnetic resonance cholangio-pancreatography showed an abnormal increase in the pancreatic duct diameter after i.v. secretin infusion (indicating outlet obstruction) in patients with idiopathic recurrent pancreatitis when compared with controls.31–35 Further, a positive secretin test was associated with a favourable clinical outcome after surgical sphincteroplasty.31 By direct measurement of sphincter of Oddi motor function with endoscopic manometry, sphincter of Oddi was found in 15–72% of such patients.3–13 Recurrent pancreatitis can result from dysfunction of the biliary sphincter segment, the pancreatic sphincter zone, or both.6, 7, 11–13, 17, 18 Most manometric studies recorded sphincter pressure in one duct only (biliary or pancreatic); unfortunately some do not report which duct was addressed. Those series evaluating both ducts, however, report higher frequencies (50–72%) of basal sphincter tone abnormalities of at least one duct in idiopathic pancreatitis patients.6, 7, 11, 12 Unfortunately, standard perfusion manometry of the pancreatic duct sphincter segment is hampered with a substantial incidence (26–29%) of procedure-related pancreatitis, especially in patients with suspected sphincter of Oddi.14–16 It has been demonstrated recently that the use of a special perfusion catheter with an aspiration port, as well as pressure measurement with an electronic micro-transducer system can reduce the pancreatitis rate by avoiding the volume load of the pancreatic duct.19, 20 However, manometric techniques requires sophisticated equipment and experience and are therefore not widely used. With non-invasive techniques for the detection of pancreatic sphincter dysfunction (secretin stimulation during either ultrasonography or magnetic resonance cholangio-pancreatography), discrepant results have been reported if correlated with manometric data.33–35 Therefore, some authorities do not accept or recommend these tests when making therapeutic decisions.17 Further, endoscopic transpapillary stent placement in the pancreatic duct has been used to select patients with idiopathic pancreatitis for sphincter ablation but this technique requires cannulation of the pancreatic duct (with its potential risk to induce pancreatitis) and ductal irregularities may result from pancreatic stent placement.36 Due to these circumstances, in routine clinical practice, a valid diagnosis of sphincter of Oddi is difficult to establish for patients with ‘idiopathic’ recurrent pancreatitis if manometry is not available. The performance of endoscopic pancreatic sphincterotomy based solely on the clinical suspicion of pancreatic duct obstruction (without manometry) also does not seem advisable since pancreatic sphincterotomy in this patient population is associated with a substantial risk for pancreatitis (21% in this series) and the clinical benefit is uncertain (the frequency of sphincter of Oddi in patients with recurrent pancreatitis is about 50%, and if sphincter of Oddi is documented the response to sphincter ablation is about 80%).8, 10, 11, 23–25
In this scenario, the use of botulinum toxin may be a more feasible practical alternative. Cholinergic stimulation of the sphincter of Oddi increases the sphincter motor activity in animal models.18 Furthermore, it has been shown that excessive cholinergic stimulation with an acetylcholine agonist can induce acute pancreatitis.37 Recent data suggest that acute pancreatitis due to poisoning with scorpion toxin is caused by a stimulation of acetylcholine release from cholinergic nerve endings.38 Local injection of botulinum toxin, which is a potent inhibitor of acetylcholine release from cholinergic nerves, into the papilla of Vater was first shown in 1994 to reduce the sphincter of Oddi baseline tone and to accelerate the scintigraphically determined bile duct clearance in two patients with biliary sphincter of Oddi.26 In a larger series of 22 patients treated with botulinum toxin injection for biliary sphincter of Oddi, we found this method safe and effective.27 Furthermore, it has been shown only recently that four of five patients with recurrent pancreatitis due to pancreas divisum respond to botulinum toxin injection (50 mouse units) into the minor papilla.39 Along with a number of studies addressing the efficacy of botulinum toxin injection into other gastrointestinal sphincters, we have observed that the effect of botulinum toxin on the sphincter of Oddi as well as on the minor sphincter fade with time.27, 39–45 In this situation both studies showed that endoscopic ablation of the respective sphincters may then offer definite cure for the patients.27, 39
The present investigation expands our knowledge to the treatment of pancreatic sphincter of Oddi with botulinum toxin. We show for the first time that local injection of botulinum toxin into the papilla of Vater (at the one o’clock position) may not only affect the targeted pancreatic duct sphincter but also reduces the tone in the biliary sphincter segment (Tables 1 and 2). This effect could be documented although we repeated manometry after botulinum toxin injection only in the case of a clinical relapse. Therefore, a positive clinical response to botulinum toxin injection was also obtained in six out of eight patients with proven sphincter hypertension of both sphincter segments.
Clinical remission after botulinum toxin injection for pancreatic sphincter of Oddi was restricted to a mean period of 5 ± 2 months. Therefore, repeated botulinum toxin injections would be necessary to induce relief from pancreatitis over a longer time span. However, we did not perform repetitive botulinum toxin injections and also would not recommend this because the effects of multiple injections into the papilla are, up to now, totally unclear. Morphological alterations may be anticipated because in patients who had undergone repeated botulinum toxin injections into the lower oesophageal sphincter for treatment of achalasia, scarring of the sphincter muscle with resulting difficulties in sphincter preparation during surgery has been reported.46 After a single botulinum toxin injection we found no macroscopic evidence for papillary damage in this study, nor in the preceding investigation for biliary sphincter of Oddi.27 All sphincterotomies performed on botulinum toxin treated patients were uneventful with regard to duodenal perforation. However, since our botulinum toxin responders gained long-term benefit from endoscopic sphincter ablation this remains the gold standard for definite therapy of patients with manometrically proven pancreatic sphincter of Oddi.
Based on our results (botulinum toxin as a highly effective treatment option for sphincter of Oddi without relevant side-effects, but with only short duration of action), botulinum toxin injection appears as an attractive pilot treatment modality. The use of botulinum toxin injection in patients with suspected pancreatic (or biliary) sphincter of Oddi may help to identify those patients who will benefit from sphincterotomy, especially in a setting, where manometry is not at hand. Botulinum toxin injection is technically much simpler and easier to perform than sphincter of Oddi manometry and may reduce the risk for procedure-related pancreatitis. In our experience of > 100 botulinum toxin injections for bilio-pancreatic sphincter of Oddi, procedure-related pancreatitis occurred in only one case and this was one of our first patients treated, in whom papillary swelling was observed after injection. For reasons of feasibility, safety, and efficacy, the probative use of botulinum toxin for establishing the diagnosis of (mainly biliary) sphincter of Oddi is therefore nowadays widely performed in several European countries.
For the treatment of pancreatic sphincter of Oddi, however, one has to keep in mind that our study was performed on a highly selected sub-group of patients with idiopathic pancreatitis, namely patients with frequent relapses of acute pancreatitis (with a median of four attacks during the last 6 months). In patients with more rare pancreatitis episodes, the short-term response to botulinum toxin may not be helpful as an indicator.
Further, the documented 80% response rate to botulinum toxin must be counterbalanced against an (ill-defined) placebo response rate of approximately 40–60%.
Up to now we have performed botulinum toxin treatment in patients with manometrically proven sphincter hypertension only, and our definition of response to botulinum toxin treatment in this study (clinical remission for at least 3 months) may be considered somewhat arbitrary because it is based solely on our experiences with the clinical response to botulinum toxin injection in biliary sphincter of Oddi patients.27 Furthermore, the number of patients without botulinum toxin response in this study is too small to determine a precise predictive value of the response to botulinum toxin injection for the patient outcome after sphincterotomy. Therefore, a randomized, controlled trial for patients with suspected pancreatic sphincter of Oddi is needed in order to determine the predictive value of probative botulinum toxin injection into the papilla of Vater.
In summary, our trial demonstrates, for the first time, that endoscopically guided injection of botulinum toxin into the papilla is a quick and safe procedure and may provide short-term relief (5 months) for about 80% of patients with acute recurrent pancreatitis due to pancreatic sphincter of Oddi dysfunction. Furthermore, botulinum toxin treatment may be indicative of prolonged response after sphincterotomy in botulinum toxin responders.