The secret (GTN) of successful ERCP cannulation: a prospective randomised controlled study

A. GHORI¹, M. HALLISEY², C. NWOKOLO¹, D. LOFT¹ and I. FRASER¹

¹ Department of Surgery and Gastroenterology, Walsgrave Hospital, Cliford Bridge Road, Coventry, CV2 2DX, ² Department of Surgery, Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH, UK

Introduction Patients and methods Results

Discussion References

Keywords: endoscopic retrograde cholangio-pancreatography (ERCP); glyceryl -trinitrate (GTN), cannulation

Background: A key step for success at endoscopic retrograde cholangio-pancreatography (ERCP) is cannulation of the sphincter of Oddi. This prospective randomised controlled study was conducted to assess the effect of glyceryl trinitrate (GTN) on ease of cannulation. Method: Two hundred and fifty four patients undergoing ERCP were randomised into two groups, 126 controls and 128 pre-treated with GTN. Result: There were 29 cannulation failures in the whole group (overall failure rate of 11.4%). Failure was significantly less common in the GTN group, 7.03% (9/128) as compared with 15.8% (20/126) in the control group (p=0.0002). Deep cannulation was desired but not achieved in 45/200 patients (22.5%) Of these, 18.4% (19/103) were in the GTN group versus 26.8% (26/97) in the control group. This pattern of difference applied to both trained and trainee endoscopists. No significant side effects of GTN were noticed. Conclusion: GTN appears to be safe and effective in improving ERCP success rate. We recommend its routine use

Abstract presented at BSG, published in Gut
J.R.Coll.Edinb., 47, August 2002, 634 - 637 

INTRODUCTION

A number of publications have reported that successful cannulation of the sphincter of Oddi can be achieved in 90-95% of cases of diagnostic ERCP.^1-5 Difficulty in cannulation can be caused by anatomical variations such as ampullary diverticula, differences in the site and size of the ampulla, oedema of the sphincter of Oddi as a result of repeated attempts at cannulation and pathological changes (ampullary tumour, gall stone impaction and stenosis) of the sphincter of Oddi and poor patient compliance. In these circumstances, various approaches have been tried, mechanical means such as metal and tapered tip cannulae, guide wires and needle knife pre-cut papillotomy can all be effective but are occasionally associated with bleeding or pancreatitis.^{1, 2}

Gastrointestinal hormones and their analogues (cholecystokinin, glucagon and sincalide) decrease the sphincter of Oddi base line pressure and may increase the likelihood of access to the common bile duct.^6-8

GTN has a smooth muscle relaxant effect, which has led to its wide use in coronary artery disease, hypertension and pulmonary oedema; it has been reported that it can also lead to a decrease in baseline pressure and amplitude of phasic contractions of the sphincter of Oddi.^{9, 10} Thillainavagam et al (1996) showed that GTN and glucagon speeded up successful biliary cannulation without influencing the overall success rate.¹¹ A few previous studies have even reported successful extraction of bile duct stones (6-12 mm in diameter) through the intact sphincter by using GTN relaxation.¹² This prospective randomised controlled study, therefore, was carried out to assess whether such an effect could be translated into success of cannulation at ERCP.

PATIENTS AND METHODS

A total of 305 consecutive patients undergoing ERCP during the period of one year were initially included in this study. Fifty-one patients who had undergone previous sphincterotomy, stent insertion or gastric surgery were excluded. The remaining 254 patients were randomised into two groups, 126 in the control group and 128 patients in the GTN group. Patient’s demographic distribution is shown in Table 1 and is similar in the two groups investigated.

Endoscopy staff performed randomisation by drawing sequential sealed envelopes prior to commencing the procedure. The time, success and speed of cannulation were recorded. All procedures were performed by either of the two experienced endoscopists or the two trainees. Trainees had performed over 50 ERCPs and the experienced operaters over 1000. If trainees failed to achieve cannulation after 10 minutes the procedure was taken over by the senior endoscopist. Cook diagnostic cannulae were used. All patients received 25-50 mg of pethidine, titrated doses of midazolam and 20 mg intravenous buscopan. Patients in the GTN group were given 1-2 puffs of sublingual GTN spray (400-800 mcg) on sight of the ampulla.

All patients were monitored during the procedure for oxygen saturation, pulse rate and blood pressure by using a pulse oximeter and blood pressure cuff. The statistical significance of the result was calculated using the Chi test or Mann-Whitney U test, where appropriate.

CBD: common bile duct

Table 1: Patients demographic details

RESULTS

Using either successful cholangiography or pancreatography as the end point, the following parameters were recorded: (a) overall success in cannulation; (b) deep cannulation achieved; (c) comparison between different grades of endoscopist and (d) time taken for cannulating the desired duct.

Statistically significant difference is seen in the overall reduction in the failure rate of cannulation. from 15.8% in the control group to 7.0% in the GTN group (p=0.0002), as shown in (Table 2).

Table 2: Overall success in cannulation *P=0.0002 for difference between GTN and control group

In our series, the failure of cannulation was mainly due to, ampullary diverticulum, ampullary oedema, ampullary tumour, stricture of common bile duct, poor patient compliance, floppy ampulla and, in one case, inability to identify the ampulla. The indications for failed ERCP in our study are similar to those previously described in the literature. The study has demonstrated that the overall higher success rate of cannulation can be achieved by the use of a GTN spray.

Although substantial improvement is seen in achieving deep cannulation by the use of GTN (Table 3), the results are not significant (p = 0.08).

Table 3: Deep cannulation achieved ** P=0.08

Breakdown of the effect of GTN amongst the experienced endoscopists and trainees is shown in (Table 4). Success rate of cannulation was substantially higher in the hands of experienced endoscopists, as compared with trainee endoscopists (12.4 % experienced endoscopists versus 25.7 % trainee endoscopists). The results suggest that both groups of endoscopists can benefit in achieving successful cannulation by the use of GTN (17.5 % to 7.5 % and 28.3 % to 22.9 % experienced and trainee endoscopists, respectively).

Table 4: Comparison between experienced endoscopists versus trainees in achieving successful cannulation in the two groups * P=0.003 (for experienced endoscopists) **P=0.15 (for trainees)

In all patients where the trainees failed, the experienced endoscopists took over and achieved successful cannulation in the majority of the cases. (Table 5)

Table 5: Experienced endoscopist taking over after trainee failure *P=0.009

There was no significant difference in the average time taken to achieve cannulation of the desired duct in the two groups. (Table 6) Time for cannulation was measured from the first site of the ampulla. In the GTN group, one-minute delay was usually allocated for the administration of the sublingual GTN spray and four minutes for the drug to cause a pharmacological effect of smooth muscle relaxation before embarking on cannulation. This probably explains the increased time to first duct cannulation in the GTN group. If the desired duct was obtained first no further attempts were made to cannulate; if the non-desired duct was cannulated initially then attempts were continued to cannulate the second duct.

Table 6: Time to successful cannulation of desired duct (bile or pancreatic) **P=0.02

There were no serious side effects or complications attributable to the GTN. Headache is the most common side effect of GTN therapy due to its effect on the cerebral circulation. None of our patients experienced this, possibly because all patients were sedated and this drug has rapid first pass elimination.

DISCUSSION

The sphincter of Oddi is a cluster of smooth muscle regulating the flow of bile and preventing reflux of duodenal contents,^13,14 it demonstrates a baseline muscle tone with phasic contractions superimposed. Interestingly, present data suggests that relaxation of the sphincter of Oddi is brought about by nonadrenergic noncholinergic innervation. Nitrous oxide appears to be the neurotransmitter and the enzyme nitrous oxide synthetase has been demonstrated in the nerve bundles of the ampulla in animals.^{15, 16} This enzyme activates guanylate cyclase catalysing formation of cGMP- dependant kinase, that results in calcium channel changes leading in alteration to the sphincter muscle tone.

We have studied the effect of GTN on the smooth muscle of the sphincter of Oddi, in terms of success and speed of cannulation at ERCP. GTN is cheap and is readily absorbed from the buccal mucosa reaching its peak plasma level in 2-3 minutes. Due to its marked first pass elimination through the liver its half-life in the blood stream is about 5-10 minutes.

We have found a significant reduction in failure of cannulation with the use of sublingual GTN. This applied to both the expert and trainee endoscopist. Although our overall failure rate (11.4%) was higher than in some reported large series (5-10%) this does not alter our conclusion, possibly strengthening the implications of our conclusion. We consider this rate to be a more realistic figure for clinical practice in centres involved in the teaching of trainees.

Furthermore, after randomisation at commencement of the procedure there were no exclusions. Cannulation can usually be achieved at a second procedure after pre-cut papillotomy, but these patients were excluded in the study design.

It has been suggested that GTN can be used in cases where difficulty is encountered but this was not the question addressed in our study. Our results suggest that the routine use of sublingual GTN spray can improve cannulation rates and we recommend adoption of this policy as routine practice.

REFERENCES

1. Charles F, Gholson, Dana F. Needle Knife Papillotomy in a University Referral Practice. J Clin Gastroenterol 1996; 23(3): 177-180
2. Farrell R J, Khan M I, Noonan N, Byrne K O, Keeling P. Endoscopic Papillectomy; A novel approach to dif ficult cannulation. Gut 1996; 39:36-38
3. Cotton PB, Williams CB. Endoscopic retrograde cholangiopancreatography: Practical gastrointesti nal endoscopy 1992; 3: 126-127
4. Siegel JH. Pre-cut papillotomy: a new method to improve success of ERCP and papillotomy. Endoscopy 1980; 12: 130-133
5. Shakoor T, Geenen JE, Pre-cut papillotomy. Gastrointestinal endoscopy 1992; 32:403-405
6. Allan P, Weston. Sincalide; A Cholycystokinin Agonist as an aid in Endoscopic Retrograde Cholangiopancree atography - A Prospective Assessment. J Clin Gastroen terol 1997; 24(4): 227-230
7. Staritz M, Pharmacology of the Sphincter of Oddi. Endoscopy 1988; 20: 171-174
8. Carr Locke DL, Gregg JA, Aoki TT, Effect of exogenous glucagon on pancreatic and biliary ductal and sphincteric pressure in man demonstrated on endo scopic manometry, and correlation with plasma gluca gon. Dig Dis Sci 1983; 28: 312
9. Staritz M, Poralla T, Ewe K, Meyer Z, Buschenfelde K. Effect of glyceryltrinitrate sphincter of Oddi motility and baseline pressure. Gut 1985; 26: 194-197
10. Luman W, Pryde A, Heading C R, Palmer R K. Topical glyceryltrinitrate relaxes the sphincter of Oddi Gut 1997; 40: 541-543
11. Thallainavagam AV, Mee AS. Improving speed and success of biliary cannulation at the index ERCP: A role for Glyceryl Trinitrate (GTN) and Glucagon Gastroenterolgy 1996;?: A479
12. Staritz M, Poralla T. Endoscopic removal of Common Bile Duct Stone through the intact papilla after meducal sphincter dilatation. Gastroenterology 1985; 88:1807-1811]
13. Calabuig R, Weems WA, Moody FG. Choledochoduo denal flow: effect of sphincter of Oddi in oppossums and cats. Gastroenterology 1990; 99:1641-46
14. Grace PA, Poston GJ, Williamson RCN, Biliary motil ity. Gut 1990; 31: 571-582
15. De Giorgio R, Parodi JE, Brecha NC, Brunicardi NC, Becker JM, Go VL. Nitrous oxide producing neurons in the monkey and human digestive system. J Comp Neurol 1994; 342: 619-627
16. Sand J, Arvola P, Jantri V, Oja S, Singram C, Baer G, Pasricha P J, Nordback I. The inhibitory role of nitric oxide in the control of porcine and human sphincter of Oddi activity. Gut 1997; 41: 375-80

Copyright: 7 April 2002

Correspondence: Mr A. Ghori, 3 Manor Farm Walk, Portesham, Weymouth, Dorset, DT1 3PH, UK

AWARDS AND GRANTS

Applications are invited for: 

KING JAMES IV PROFESSORSHIPS

Up to five lectureships will be awarded annually by the College, in open competition, to distinguished practitioners of surgery or dental surgery. The courtesy title of Professor will be accorded to the individuals for the duration of the College year in which their lectures are delivered. Applicants must be Fellows/Dental Fellows of the College.

The closing date for receipt of applications is Friday 6 September 2002

GENERAL RESEARCH (PUMP-PRIMING) GRANTS including one grant sponsored by the J P MORGAN FLEMING FOUNDATION

Applications are invited for research grants (up to £6,000) from surgical trainees and recently appointed consultants. The principal applicant must be a Fellow/Member of the College in good standing.

The closing date for receipt of applications is Friday 6 September 2002

TRAVELLING FELLOWSHIPS

The Cutner Travelling Fellowship in Orthopaedics The John Steyn Travelling Fellowship in Urology.

The closing date for receipt of applications is Friday 6 September 2002

Further details and application forms can be obtained from: Elizabeth Wright, Awards & Grants Secretary, The Royal College of Surgeons of Edinburgh, Nicolson Street, Edinburgh, EH8 9DW or email: e.wright@rcsed.ac.uk