Sublingual glyceryl trinitrate during colonoscopy and terminal ileal intubation: a randomized controlled trial

Introduction

Glyceryl trinitrate (GTN) relaxes gastrointestinal smooth muscle. It exerts its effect by being converted to nitric oxide by mitochondrial aldehyde dehydrogenase. Nitric oxide increases the level of cyclic guanosine monophosphate (cGMP) within the cell, which in turn activates myosin light chain phosphatase via a cGMP-dependent protein kinase. The end result is the inhibition of smooth muscle contraction.^1,2

GTN is more widely known as a pharmacological agent for its application in vasodilation and the treatment of angina pectoris. However, it is effective in relaxing smooth muscle elsewhere in the body such as the sphincter of Oddi, where it has been used in the past to aid cannulation during endoscopic retrograde cholangiopancreatography (ERCP).^3–5 Both topical and sublingual GTN have been used for this purpose with mixed success.^3,5 We have shown previously that topical GTN can relax the smooth muscle within the ileocolic sphincter. ⁶ A role for sublingual GTN to aid terminal ileal intubation (TII) has not been previously explored.

TII is regarded by many as the gold standard for evidence of a completed colonoscopy.^7,8 In addition, there is increasing evidence that it adds diagnostic value, particularly in patients suffering from right lower quadrant abdominal pain. ⁹ The procedure is not widely popular since it is technically very demanding. Different techniques, including altering the patient’s position, have been described that increase TII rate and reduce time taken. ¹⁰ However, there are no studies that describe the use of pharmacological aids for TII.

The aim of this study was to investigate the effect of sublingual GTN on TII rate and time during colonoscopy.

Method

A triple-blind (patient, endoscopist and data analyst) randomized controlled trial comparing sublingual GTN (800 µg) vs. placebo (saline) in relation to TII during colonoscopy was performed.

A total of 110 patients presenting to a routine colonoscopy list in Monklands Hospital Endoscopy Unit between 2004 and 2009 were recruited. The colonoscopies were performed by three experienced consultants. On admission, all patients fitting the inclusion criteria were asked to participate in the trial by a specialist nurse overseeing the study. The format of the study was explained to the patients and written consent was obtained.

All patients who fulfilled the criteria were included. The inclusion criteria were: patients receiving a colonoscopy aged between 18 and 75 years old. The exclusion criteria were: (1) unable to give consent, (2) had taken buscopan or metoclopramide that day, (3) all patients on cardiac/anti-hypertensive medicines (4) patient refusal. Patients whose colonoscopy failed to reach the caecum were automatically excluded, since TII was not feasible (these last numbers were small as all endoscopists worked within the current BSG guidelines (2011) for colonoscopy: 90% overall completion rate).

Once the caecum was intubated, randomisation took place. The specialist nurse drew from a box an opaque sealed envelope which contained an A4 piece of paper with either ‘Drug A’ or ‘Drug B’ written on it. Concealed allocation was ensured by an individual who was not involved in the trial. Opaque envelopes had been pre-prepared and labelled; half of them as ‘Drug A’ and half as ‘Drug B’. Envelopes were mixed and placed in a box in batches of 50.

Patients either received sublingual GTN (800 µg) or sublingual normal saline. Each solution was discharged from one of two identical white containers, pre-prepared by our pharmacy department. Since the containers were identical, the patient and the colonoscopist were not aware whether Drug A or B was administered and therefore into which arm the patient was allocated. This ensured blinding of the patient and colonoscopist.

Following the sublingual administration of GTN/normal saline, the colonoscopist waited for 2 min before initiating intubation of the terminal ileum. This ensured adequate systemic distribution of GTN. The time taken for TII was recorded. If the colonoscopist failed to intubate the terminal ileum after 15 min, the procedure was abandoned and recorded as a failed intubation.

Standard demographic data were recorded and throughout the procedure the heart rate, oxygen saturation and blood pressure were monitored.

All data were collected prospectively and collated in Microsoft Excel 2010™. The data analyst was blinded to the true identity of Drug A and Drug B.

Intubation rates for each group were analysed using Fisher’s exact test. Intubation times were non parametric and are therefore expressed as median (range) and compared using a Mann–Whitney test. Microsoft Excel and IBM SPSS version 22 (SPSS Inc., Chicago, IL, USA) were used for the statistical analysis. Failed intubations were given the maximum time of 900 s.

This study was approved by the Regional Ethics Committee (Lanarkshire Health Board) and local research and development department. Written informed consent was obtained by all patients participating in this study.

Results

During this five-year period (2004–2009), 110 patients were successfully randomised for this study. Randomisation was performed as per protocol. Fifty-four patients were allocated to the GTN group and 56 to the placebo group. There was no difference in the median age of the two groups (GTN vs. placebo; 57 (18–75) vs. 58 (18–74) years).

The intubation rate of the GTN group was 100% (54/54). In the placebo group, the intubation rate was 94.6% (53/56). This difference was not significant (p = 0.243, Fisher’s exact test) (see Table 1).

OPEN IN VIEWER

Table 1.

Intubation rate and times of GTN vs. Placebo.

	GTN	Placebo	p value
Intubation rate	54/54 (100%)	53/56 (94.6%)	0.243, Fisher’s exact test
Median (range) intubation time in seconds	72.5 (7–900)	125 (5–900)	0.150, Mann–Whitney test

The median time of intubation in the GTN group was 72.5 (7–900) s, and in the placebo group was 120 (5–900) s. When the failed intubations were included at 900 s, the figures were 72.5 (7–900) s in the GTN group and 125 (5–900) s in the placebo group. There was a non-significant trend to faster TII in the GTN group (p = 0.150, Mann–Whitney test) (see Table 1 and Figure 1). OPEN IN VIEWER

There was no abnormality in oxygen saturation, blood pressure and heart rate while patients were having their terminal ileum intubated. Only one patient suffered from any adverse effect: after receiving GTN, the patient became transiently hypotensive and bradycardic. The patient responded to intravenous fluids and was not excluded from the data analysis.

Discussion

GTN is known to relax smooth muscle throughout the body, including that within the ileocolic sphincter. ⁶ We hypothesised that by relaxing the smooth muscle at the ileocaecal valve, TII would become easier.

Our results showed no statistically significant difference in intubation rate or timing in the two trial arms. Therefore, the null hypothesis, that there is no difference between intubation rate and time in patients receiving sublingual GTN or placebo, is not rejected. This could be due to a variety of reasons.

In the context of the study, sublingual GTN administration did not significantly improve the TII time. This could be attributed to GTN’s short duration of action as suggested in a similar trial reporting the role of topical GTN in ERCP cannulation. ¹ It has been shown that the effects of GTN wane after 2–3 min in the sphincter of Oddi.^4,11 Alternatively, the smooth muscle relaxation effect of GTN on the ileocaecal valve may simply not make this already technically difficult procedure any easier.

Another possible explanation is that the consultants performing the TIIs were all very experienced colonoscopists. Their TII rates and timings were already very high, as demonstrated by the placebo group. Therefore, room for improvement is limited and relaxing the ileocaecal valve's smooth muscle with GTN is unlikely to make an already very well-performed procedure any more successful or faster. The colonoscopists TII rate was 94.6% and median TII time 2 min in the placebo group. In the literature, TII rates have been described anywhere between 74% and 96%,^9,11–13 and a mean TII time of 6.4 min has been reported. ¹³ A different result may have occurred if the colonoscopists were still on the learning curve for TII. The effects of GTN on TII may be more pronounced if the colonoscopists had not yet ‘mastered’ the TII technique, and repeating the trial with more junior colonoscopists might demonstrate a significant difference in favour of sublingual GTN.

We need to acknowledge the possibility of a type 2 statistical error, as only 110 patients were recruited over a five-year period. At the end of 2009, our ethics approval expired and any additional patients would have required an altered protocol under the new national guidelines. ¹⁴ A power calculation was not possible as there is no available reference material.

It is worth noting that the participants’ median age is not representative of the average patient receiving a colonoscopy since patients over 75 years of age were excluded from the trial.

As far as we can ascertain, this is the first study examining the role of sublingual GTN on TII rates. Other RCTs have looked at the role of GTN in other procedures^1,3,5 or the role of other variables on TII (e.g. patient position, ¹⁰ age, BMI, height and pre-existing surgery ¹³ ).

Acknowledging the limitations of this study, this trial suggests that in the hands of experienced colonoscopists (as witnessed by the high success rate in the placebo arm), the routine use of sublingual GTN does not improve TII time or intubation success rates. However, considering the visible trend towards improved intubation times in the GTN group, it is possible that GTN might be a helpful aid for the more challenging TIIs or for less-experienced endoscopists. One suggestion could be that rather than routinely using GTN for TII, GTN is reserved for cases where TII has not been completed after 3 to 5 min. In order to establish whether GTN truly has a role in TII, larger trials are required, perhaps with colonoscopists still on the learning curve of this challenging procedure.