Sphincter of Oddi Dysfunction: Cut and Inject, but Don’t Measure the Pressure?

Sphincter of Oddi dysfunction (SOD) is a poorly defined clinical entity that presents with recurrent right-sided pain (presumed to be biliary SOD) or epigastric pain (presumed to be pancreatic), caused by a fixed stenosis or dyskinesia of the tiny sphincter located in the papilla of Vater. In most cases, the precise origin of the suspected sphincter of Oddi stenosis (e. g., repeated stone passage or inflammation) or dyskinesia (e. g., autonomic neuropathy or postoperative nerve damage) cannot be evaluated using the diagnostic methods currently available [1] [2] [3] [4]. Beginning in the late 1970s, it was shown by endoscopic sphincter of Oddi perfusion manometry that an elevated baseline pressure is the most consistent finding in such patients [1] [2] [3]. On the basis of sparse published data from healthy individuals [1] [5], an elevation of the baseline pressure by more than three standard deviations above the mean (i. e., ≥ 40 mmHg) was classified as abnormal, and sphincter of Oddi hypertension was defined as an indicator of the presence of sphincter of Oddi dysfunction [1] [2] [3] [6]. Geenen’s group in Milwaukee first demonstrated that the probability of sphincter of Oddi hypertension correlates with the presence of laboratory findings (enzymatic cholestasis or elevated pancreatic enzymes) and ERCP findings (dilated duct, delayed duct drainage) - leading to what became known as the Milwaukee classification [6]. As it was shown in some trials that patients in biliary group I uniformly have both sphincter of Oddi hypertension and also notable clinical benefit after biliary sphincterotomy [7] [8], most guidelines suggest performing endoscopic sphincterotomy in these cases without prior manometry [9]. However, it must be borne in mind that there have been no larger trials with long-term follow-up periods and no randomized controlled studies on this topic.

There is substantial controversy regarding the management of the remaining groups of patients. Whereas a diagnosis of sphincter of Oddi hypertension by manometry has been widely accepted in the United States, sphincter of Oddi manometry is only used by a very small group of enthusiastic investigators in European countries and in most Asian countries; even several of the European and Asian endoscopic centers with the best worldwide reputations do not carry out sphincter of Oddi motility tests [10]. One reason for this may have been an earlier fear of the risk of pancreatitis resulting from sphincter of Oddi manometry (although it has in the meantime been recognized that it is not the manometry procedure itself, but rather the specific SOD patient that presents the most notable risk factor for pancreatitis). Another reason may be the very low reimbursement rates for this sophisticated technique outside the USA (see below). On the other hand, randomized controlled studies demonstrating that the response to endoscopic sphincterotomy depends on the presence of sphincter of Oddi hypertension have only been conducted in the biliary type II group [11] [12]. In biliary type III patients, the reported response rates to biliary endoscopic sphincterotomy vary substantially even if sphincter of Oddi hypertension is diagnosed [13] [14] [15], and the numbers of patients enrolled in published studies dealing with pancreatic SOD have been very low [16] [17] [18] [19] [20]. In pancreatic type III patients in particular, there has not yet been a single full publication on the outcome of pancreatic or dual sphincterotomy.

The two papers published in this issue of Endoscopy address two important topics. The first trial, by Elta’s group [21], investigated whether the injection of botulinum toxin (BTX) reduces the risk of pancreatitis after the performance of manometry and biliary endoscopic sphincterotomy in biliary SOD patients. Neurotoxin A from Clostridium botulinum was shown to reduce the sphincter of Oddi baseline pressure in two patients with biliary SOD [22], and it was later shown in one patient that a BTX injection can also lower the pancreatic sphincter of Oddi baseline pressure [23]. In two prospective series, our own group was also able to demonstrate that in patients with biliary type III [24] and pancreatic types I and II [25] with manometrically confirmed sphincter of Oddi hypertension, a clinical response to BTX injection identifies a subgroup of patients who will genuinely benefit from sphincter ablation. Since it has been shown that residual pancreatic sphincter hypertension that is not eliminated by routine biliary sphincterotomy may be present in biliary SOD patients and that this constitutes a risk factor for pancreatitis after biliary endoscopic sphincterotomy [26], it has been hypothesized that the pancreatic sphincter zone should be an additional target for the endoscopist in biliary SOD patients. Pancreatic duct stenting after biliary sphincterotomy has been shown to reduce the risk of pancreatitis significantly [27], while the performance of dual sphincterotomy may be associated with a better clinical outcome [28]. Since both techniques - dual sphincterotomy and pancreatic duct stenting - are invasive and cumbersome, simple injection of BTX appears to be an attractive alternative.

At first glance, the study by Gorelick et al. [21] appears to show that it is possible to reduce the pancreatitis rate after biliary sphincterotomy by carrying out an additional BTX injection in SOD patients: the pancreatitis frequencies were six of 14 in the sham arm (43 %) and three of 12 in the BTX arm (25 %; P = 0.34). However, several questions remained unanswered.

Firstly, there was a surprisingly high frequency of pancreatitis in the control group (42 %); this is the highest incidence so far reported in the literature. This high incidence of pancreatitis led to premature termination of the study. Although this is not specifically addressed by the authors in their report, the high frequency of pancreatitis is likely to have been due to the fact that a large number of patients with documented pancreatic sphincter hypertension were included in the study, as patients with confirmed normal pancreatic manometry (but with biliary sphincter of Oddi hypertension) had been excluded.

Secondly, post-sphincterotomy pancreatitis develops during the first 4 h after manometry and/or sphincterotomy. The effect of BTX therefore has to start immediately after injection. In an abstract publication by Okolo et al. [29], BTX injection was reported to reduce the sphincter of Oddi baseline pressure in several dogs immediately after administration. However, in a much larger full publication comparing the effects of BTX with bile duct stenting, it was shown that the efficacy of BTX increases significantly up to 14 days after injection [30]. This finding corresponds with the clinical observation that the effects of BTX on biliary symptoms in SOD patients, as well as on dysphagia in achalasia patients, are only recognizable several days after treatment (personal observation). This delayed efficacy is also observed after BTX delivery into the eye or neck musculature; clinically evident effects were not observed until the day after injection (W. Jost, personal communication). In pharmacodynamic in-vitro studies, the first electrophysiological effects after BTX injection were detectable 2 - 3 h after administration [31] [32], and in an animal model, effects on antral motor function were evident after 6 h [33]. In view of this, fast-acting drugs such as nitro derivatives might be of interest; in human experiments, these have been shown to lower the sphincter pressure during the first few minutes after application [34] [35], with a significant effect on the incidence of post-ERCP pancreatitis in one study [36]. Nitro derivatives are also much more inexpensive than BTX.

Thirdly, the technique of BTX injection appears to be quite simple only on first inspection. The authors injected two aliquots of 0.25 ml (each containing 25 mouse units of BTX) at the 2-o’clock and 6-o’clock positions. However, the filling volume of a standard sclerotherapy needle is about 0.4 - 1.0 ml (personal observation). To ensure that the complete dosage will reach the sphincter, one therefore has to preload the needle with the injection volume (the precise technique used is not described in the report), as it is difficult to guarantee that such a small volume will persist in the needle. In addition, even just a single air bubble remaining in the sclerotherapy needle will prevent the precise delivery of the 0.25 ml of BTX. Since BTX is able to migrate several millimeters into the targeted muscle [32], it might be easier (although still not perfect) to inject the complete volume (0.5ml) in a single shot.

In general, on the basis of the lack of any significant effect of BTX documented in the trial by Gorelick et al. [21], the conclusion drawn by the authors in their earlier abstract publication [37] that ”BTX injection into the residual pancreatic sphincter after biliary sphincterotomy was technically feasible and safe, but failed to decrease the post-ERCP pancreatitis rate” appears to be justified.

In the other publication on the topic in this issue of Endoscopy, by Arguedas et al. [38], a decision analysis was carried out with regard to the management of patients with suspected biliary type II SOD. The authors evaluated a scenario in which ERCP with sphincter of Oddi manometry was carried out, with sphincterotomy only being carried out if sphincter of Oddi hypertension was present, versus ERCP with ”empirical” sphincterotomy in all patients but without manometry [38]. In their hypothetical cohort of 100 patients, 55 % improved with the first strategy (with manometry), but 60 % benefited from the strategy without manometry [38]. In addition, ”empirical” sphincterotomy without manometry (at US $ 2244 per patient) proved to be less expensive than performing sphincterotomy guided by the manometric results (US $ 2790 per patient). Several assumptions had to be made in order to develop the mathematical decision model, and Arguedas et al. [38] derived these mainly from two previously published randomized and controlled studies on the outcome of biliary endoscopic sphincterotomy in manometrically defined biliary type II SOD patients [11] [12]. The reimbursement rates usual in the USA were used to calculate the cost-effectiveness.

The most critical hypothesis in this model, which is the main cause of the observed superiority of ”empirical” biliary sphincterotomy, is the assumed 42 % response rate to endoscopic sphincterotomy in patients with normal manometry. It must be borne in mind that the randomized studies mentioned [11] [12], from which the data were derived, used very strict clinical criteria for study entry. Since the a priori patient selection (suspected biliary type II) is not based on manometric results, the clinical criteria for patient selection then become all-important (e. g., strict definition of biliary colic, all laboratory abnormalities present, ERCP pathology featuring drainage times, esophagogastroduodenoscopy, colonoscopy and computed tomography carried out to rule out other sources of pain, etc.). However, the use of such strict criteria may not represent the real-world situation (e. g., ERCP is now not usually performed before the indication for sphincter of Oddi manometry is discussed). For example, in the well-conducted study by Toouli et al. [12], all patients completed a diary to document their symptoms for 3 months before sphincter of Oddi manometry was carried out - an approach that is not feasible in everyday practice. It is obvious that the criteria outlined in the model would likely have resulted in a much lower prevalence of real SOD and therefore a much lower response rate to ”empirical” sphincterotomy if such strict selection criteria had not used.

In the study by Geenen et al. [11], the response rate to biliary endoscopic sphincterotomy was 33 % in patients with normal manometry. If the response rate falls below this value (e. g., to 30 %), patients would benefit from the performance of sphincter of Oddi manometry rather than from ”empirical” treatment, according to the two-way sensitivity analysis used by Arguedas et al. (Figure 2 in the paper) [38]. There has so far only been one published study reporting data from ”empirical” endoscopic sphincterotomy in biliary type II patients, showing that patients with enzymatic cholestasis have a higher probability of responding to endoscopic sphincterotomy than patients with a dilated bile duct [39]. This result [39] underscores the concern regarding the selection criteria applied. For a clinician who is able to carry out sphincter of Oddi manometry, the significantly better response rates to endoscopic sphincterotomy confirmed in patients with documented sphincter of Oddi hypertension may outweigh the calculated benefit of ”empirical” endoscopic sphincterotomy in the mathematical model presented by Arguedas et al. [38] - particularly in the light of the potential risks of endoscopic sphincterotomy in patients with SOD. In addition, the model presented by Arguedas et al. [38] assumes that carrying out ERCP (which has its own inherent complication rate) is necessary in all patients with presumed biliary SOD. Due to improvements in other noninvasive techniques (such as endoscopic ultrasonography and magnetic resonance cholangiopancreatography), this is obviously not the case, and the performance of sphincter of Oddi manometry without ERCP may be associated with a reduced risk for pancreatitis.

However, this paper provides the first theoretical support for the vast majority of endoscopists around the world who have been performing endoscopic sphincterotomy in biliary SOD without manometry for many decades. A randomized study comparing these two strategies thus appears to be necessary.

With regard to cost-effectiveness, the results presented by Arguedas et al. [38] are only applicable in the USA, where the additional performance of sphincter of Oddi manometry adds about US $ 300 - 400 to the total costs of ERCP. However, these reimbursement rates are much lower in most European countries (in Germany, for example, the reimbursement rate for sphincter of Oddi manometry is equivalent to approximately US $ 40). In these conditions, cost calculations would probably lead to completely different results. The use of extremely rigorous exclusion evaluations prior to ERCP in order to obtain the best results from ”empirical” endoscopic sphincterotomy may also result in a substantial increase in total costs.

What we can learn from these two studies published in this issue of Endoscopy? Firstly, residual pancreatic sphincter hypertension in patients with biliary SOD is an important risk factor for the development of postprocedural pancreatitis (e. g., after ERCP, manometry, and/or sphincterotomy), as confirmed by the dramatically high rate of pancreatitis (nearly half of the patients) in the sham arm of the study by Gorelick et al. [21]. Endoscopists should therefore direct their attention to this part of the sphincter if they have carried out manipulations at the papilla in patients with presumed SOD. Although there have been no randomized studies supporting this, it may be wise to carry out a dual sphincterotomy with protective pancreatic stenting not only in patients with pancreatic SOD, but also in patients with biliary SOD. If the decision is taken not to carry out additional pancreatic sphincterotomy, one should at least leave a pancreatic stent in place. The use of drugs (e. g., BTX, nitro derivatives) cannot at present be recommended, but this issue should be evaluated in future trials.

Secondly, sphincterotomy without manometry is justified in patients with suspected biliary SOD type I and may also be effective in very well selected cases of biliary type II. However, it appears to be more rational to measure the pressure, if this is possible, before cutting the sphincter. The cutting-edge procedure in relation to cutting the sphincter is thus manometry.

References

• 1 Wehrmann T, Seifert H, Hagenmüller F. et al . Endoscopic sphincter of Oddi manometry - does skepticism vanish?.  Z Gastroenterol. 1996;  34 17-25
• 2 Allescher H D. Clinical impact of sphincter of Oddi dyskinesia.  Endoscopy. 1998;  30 (Suppl 2) 231-236
• 3 Tzovaras G, Rowlands B J. Diagnosis and treatment of sphincter of Oddi dysfunction.  Br J Surg. 1998;  85 588-595
• 4 Wehrmann T, Stergiou N, Riphaus A, Lembcke B. Correlation between sphincter of Oddi manometry and intraductal ultrasound morphology in patients with suspected sphincter of Oddi dysfunction.  Endoscopy. 2001;  33 773-777
• 5 Guelrud M, Mendoza S, Rossiter G, Villegas M I. Sphincter of Oddi manometry in healthy volunteers.  Dig Dis Sci. 1990;  35 38-46
• 6 Lans J L, Parikh N P, Geenen J E. Application of sphincter of Oddi manometry in routine clinical investigations.  Endoscopy. 1991;  23 139-143
• 7 Rolny P, Geenen J E, Hogan W. Post-cholecystectomy patients with objective signs of partial bile outflow obstruction: clinical characteristics, sphincter of Oddi manometry findings, and results of therapy.  Gastrointest Endosc. 1993;  39 778-781
• 8 Wehrmann T, Wiemer K, Lembcke B, Jung M. Effect of endoscopic sphincterotomy on sphincter of Oddi manometry results in patients with or without papillary stenosis.  Z Gastroenterol. 1995;  33 662-668
• 9 American Motility Society. Position paper: sphincter of Oddi manometry.  Gastrointest Endosc. 1997;  45 342-348
• 10 Ladas S D, Tassios P S, Katsogridakis J. et al . The clinical need for sphincter of Oddi manometry in gastrointestinal endoscopy units.  Endoscopy. 1993;  25 387-391
• 11 Geenen J E, Hogan W J, Dodds W J. et al . The efficacy of endoscopic sphincterotomy after cholecystectomy in patients with sphincter of Oddi dysfunction.  N Engl J Med. 1989;  320 82-87
• 12 Toouli J, Roberts-Thomson I C, Kellow J. et al . Manometry based randomised trial of endoscopic sphincterotomy for sphincter of Oddi dysfunction.  Gut. 2000;  46 98-102
• 13 Botoman V A, Kozarek R A, Novell L A. et al . Long term outcome after endoscopic sphincterotomy in patients with biliary colic and suspected sphincter of Oddi dysfunction.  Gastrointest Endosc. 1994;  40 165-170
• 14 Bozkurt T, Orth K H, Butsch B, Lux G. Long term clinical outcome of post-cholecystectomy patients with biliary type pain: results of manometry, non-invasive techniques and endoscopic sphincterotomy.  Eur J Gastroenterol Hepatol. 1996;  8 245-249
• 15 Wehrmann T, Wiemer K, Lembcke B. et al . Do patients with sphincter of Oddi dysfunction benefit from endoscopic sphincterotomy? A five year prospective trial.  Eur J Gastroenterol Hepatol. 1996;  8 251-256
• 16 Chen J WC, Saccone G TP, Toouli J. Sphincter of Oddi dysfunction and acute pancreatitis.  Gut. 1998;  43 305-308
• 17 Geenen J E, Nash J A. The role of sphincter of Oddi manometry and biliary microscopy in evaluating idiopathic recurrent pancreatitis.  Endoscopy. 1998;  30 (Suppl 2) 237-241
• 18 Guelrud M, Plaz J, Mendoza S. et al . Endoscopic treatment of type II pancreatic sphincter dysfunction [abstract].  Gastrointest Endosc. 1995;  41 398
• 19 Lehman G A, Sherman S. Hypertensive pancreatic sphincter.  Can J Gastroenterol. 1998;  12 333-337
• 20 Wehrmann T, Zipf A, Caspary W F, Jung M. Sphincter of Oddi dysfunction in patients with idiopathic recurrent pancreatitis.  Dtsch Med Wochenschr. 1996;  121 781-787
• 21 Gorelick A, Barnett J, Chey W. et al . Botulinum toxin after biliary sphincterotomy.  Endoscopy. 2004;  35 170-173
• 22 Pasricha P J, Miskovsky E P, Kalloo A N. Intrasphincteric injection of botulinum toxin for suspected sphincter of Oddi dysfunction.  Gut. 1994;  35 1319-1321
• 23 Muehldorfer S M, Hahn E G, Ell C. Botulinum toxin as a diagnostic tool of verification of sphincter of Oddi dysfunction causing recurrent pancreatitis.  Endoscopy. 1997;  29 120-124
• 24 Wehrmann T, Seifert H, Seipp M, Lembcke B, Caspary W F. Endoscopic injection of botulinum toxin for biliary sphincter of Oddi dysfunction.  Endoscopy. 1998;  30 702-707
• 25 Wehrmann T, Schmitt T, Arndt A. et al . Endoscopic botulinum toxin injection for treatment of idiopathic recurrent pancreatitis due to sphincter of Oddi dysfunction.  Aliment Pharmacol Ther. 2000;  14 1469-1477
• 26 Tarnasky P, Cunningham J, Cotton P. et al . Pancreatic sphincter hypertension increases the risk of post ERCP-pancreatitis.  Endoscopy. 1997;  29 252-257
• 27 Eversman D, Fogel E L, Rusche M. et al . Frequency of abnormal pancreatic and biliary sphincter manometry compared with clinical suspicion of sphincter of Oddi dysfunction.  Gastrointest Endosc. 1999;  50 637-641
• 28 Tarnasky P R, Palesch Y Y, Cunningham J T. et al . Pancreatic stenting prevents pancreatitis after biliary sphincterotomy in patients with sphincter of Oddi dysfunction.  Gastroenterology. 1998;  115 1518-1524
• 29 Okolo P I, Wadwa K, Magee C A. et al . Immediate reduction of pancreatic SO pressure after endoscopic injection of botulinum toxin (BoTox): implications for prevention of ERCP-induced pancreatitis [abstract].  Gastroenterology. 1998;  114 A535
• 30 Marks J M, Bower A L, Goormastic M. et al . A comparison of common bile duct pressures after botulinum toxin injection into the sphincter of Oddi versus biliary stenting in a canine model.  Am J Surg. 2001;  181 60-64
• 31 MacKenzie J, Burnstock J, Dolly J O. The effects of purified botulinum neurotoxin type A on cholinergic, adrenergic and non-adrenergic, atropin-resistant autonomic neuromuscular transmission.  Neuroscience. 1982;  7 997-1006
• 32 Montecucco C, Schiavo G. Mechanisms of action of tetanus and botulinum neurotoxins.  Molec Microbiol. 1994;  13 1-8
• 33 James A N, Ryan J P, Parkman H P. Inhibitory effects of botulinum toxin on pyloric and antral smooth muscle.  Am J Physiol. 2003;  285 G291-G297
• 34 Luman W, Pryde A, Heading R C, Palmer K R. Topical glyceryl trinitrate relaxes the sphincter of Oddi.  Gut. 1997;  40 541-543
• 35 Wehrmann T, Schmitt T, Stergiou N. et al . Topical application of nitrates onto the papilla of Vater: manometric and clinical results.  Endoscopy. 2001;  33 323-328
• 36 Sudhindran S, Bromwich E, Edwards P R. Prospective randomized double-blind placebo-controlled trial of glyceryl trinitrate in endoscopic retrograde cholangiopancreatography-induced pancreatitis.  Br J Surg. 2001;  88 1178-1182
• 37 Gorelick A B, Barnett J L, Chey W D. et al . A study of botulinum toxin injection into the residual pancreatic sphincter after biliary sphincterotomy in patients with sphincter of Oddi dysfunction [abstract].  Gastrointest Endosc. 2001;  53 AB94
• 38 Arguedas M R, Linder J D, Wilcox C M. Suspected sphincter of Oddi dysfunction type II: empiric biliary sphincterotomy or manometric-guided therapy.  Endoscopy. 2004;  35 174-178
• 39 Lin O S, Soetinho R M, Young H S. The utility of liver function test abnormalities concomitant with biliary symptoms in predicting a favorable response to endoscopic sphincterotomy in patients with presumed sphincter of Oddi dysfunction.  Am J Gastroenterol. 1998;  93 1833-1836