Stimulation of colonic motility by oral PEG electrolyte bowel preparation assessed by MRI: comparison of split vs single dose

L Marciani; K C Garsed; C L Hoad; A Fields; I Fordham; S E Pritchard; E Placidi; K Murray; G Chaddock; C Costigan; C Lam; J Jalanka-Tuovinen; W M De Vos; P A Gowland; R C Spiller

doi:10.1111/nmo.12403

. 2014 Jul 24;26(10):1426–1436. doi: 10.1111/nmo.12403

Stimulation of colonic motility by oral PEG electrolyte bowel preparation assessed by MRI: comparison of split vs single dose

L Marciani ^*,^#, K C Garsed ^*,^#, C L Hoad ^†, A Fields ^*, I Fordham ^*, S E Pritchard ^†, E Placidi ^†, K Murray ^†, G Chaddock ^†, C Costigan ^†, C Lam ^*, J Jalanka-Tuovinen ^‡, W M De Vos ^‡,^§, P A Gowland ^†, R C Spiller ^*

PMCID: PMC4321061 PMID: 25060551

Abstract

Background

Most methods of assessing colonic motility are poorly acceptable to patients. Magnetic resonance imaging (MRI) can monitor gastrointestinal motility and fluid distributions. We predicted that a dose of oral polyethylene glycol (PEG) and electrolyte solution would increase ileo-colonic inflow and stimulate colonic motility. We aimed to investigate the colonic response to distension by oral PEG electrolyte in healthy volunteers (HVs) and to evaluate the effect of single 2 L vs split (2 × 1 L) dosing.

Methods

Twelve HVs received a split dose (1 L the evening before and 1 L on the study day) and another 12 HVs a single dose (2 L on the main study day) of PEG electrolyte. They underwent MRI scans, completed symptom questionnaires, and provided stool samples. Outcomes included small bowel water content, ascending colon motility index, and regional colonic volumes.

Key Results

Small bowel water content increased fourfold from baseline after ingesting both split (p = 0.0010) and single dose (p = 0.0005). The total colonic volume increase from baseline was smaller for the split dose at 35 ± 8% than for the single dose at 102 ± 27%, p = 0.0332. The ascending colon motility index after treatment was twofold higher for the single dose group (p = 0.0103).

Conclusions & Inferences

Ingestion of 1 and 2 L PEG electrolyte solution caused a rapid increase in the small bowel and colonic volumes and a robust rise in colonic motility. The increase in both volumes and motility was dose dependent. Such a challenge, being well-tolerated, could be a useful way of assessing colonic motility in future studies.

Keywords: colon, intestinal fluid, motility, small bowel, symptoms

Key Messages.

Current methods of assessing colonic motility have significant limitations of expense, availability and patient acceptability.
We aimed to investigate the use of oral polyethylene glycol (PEG) electrolyte as standardized colonic stimulation and magnetic resonance imaging (MRI) to monitor fluid distributions and gastrointestinal motility noninvasively.
Standard MRI, now widely available, showed that ingestion of PEG electrolyte solution increased the small bowel water content fourfold.
The size of the colon doubled after dosing and the ascending colon motility index rose markedly, the increase being dose dependent.
Such a challenge was well tolerated and could be a useful way of assessing colonic motility in future studies.

INTRODUCTION

Most current methods of assessing colonic motility are invasive, involving intestinal intubation which is arduous for the patient and require a highly skilled investigator.¹ As such they are limited in their use to highly selected patients. While ambulatory manometry using a radio-pill does allow non-invasive measurement of motility, because its transit is uncontrolled it is not possible to reliably assess any particular area of colon or its response to stimulation.² Furthermore without a standard provocation, colonic motor activity shows marked diurnal variation¹^,³ and repeated transit measures show poor reproducibility with 27–31% variation,⁴ suggesting variables like diet and activity alter colonic motility. Magnetic resonance imaging (MRI) can non-invasively monitor gastrointestinal motility and fluid distributions⁵^–¹² and may be suitable for assessment of colonic function after an appropriate stimulus.

Oral polyethylene glycol (PEG) formulations have been shown to be effective colonic stimulants at a 4 L dose.¹³ Smaller volumes (2 L) of PEG-based solutions containing ascorbic acid and electrolyte ions have comparable efficacy and may be better tolerated.¹⁴^–¹⁸ PEG is administered as a hyper-osmotic 402 mOsmol/L non-nutrient solution. When given with the poorly absorbed SO₄²⁻ ion, it would therefore be expected to generate a net inflow of fluid into the highly permeable upper small intestine.¹⁹ As such, the solution would be predicted to pass unabsorbed through the small bowel causing a substantial increase in its fluid content leading to an increased inflow into the colon distending the lumen and thereby stimulating motility. Until recently there was no way of imaging serially and non-invasively gastrointestinal fluid volumes, but recent developments in MRI of gastrointestinal function show that it is now feasible to provide functional information on the gastrointestinal tract with both excellent spatial and temporal characteristics.²⁰ With specific reference to gastrointestinal fluids, MRI has been recently used to investigate fasting and postprandial fluid distribution in health²¹^,²² and disease⁸ and also to investigate the mode of action of drugs such as loperamide²³ and a 5-HT₃ antagonist.²⁴

Our hypothesis was that PEG electrolyte would increase small bowel and colonic water content and stimulate colonic motility which would be clearly detectable with MRI imaging.

MATERIALS AND METHODS

Study design

This was an open-label, parallel group design study in healthy volunteers (HVs) using MRI to compare the effects of a split, 2 × 1 L doses (separated by an overnight fast) vs one single 2 L dose of PEG (Macrogol 3550) electrolyte solution (MOVIPREP®; Norgine Pharmaceuticals Ltd, Harefield, UK) given on the morning of the treatment day. This trial was registered on the EU Clinical Trials Register with EudraCT Number 2010-021879-85. There were no changes to the protocol. It was approved by the National Research Ethics Service (approval 10/H0906/50), the NHS Trust R&D (approval 10GA018), and by the Medicines and Healthcare products Regulatory Agency (MHRA Clinical Trial Authorization CTA 03057/0045/001-0001, protocol 10050). All volunteers gave informed written consent before participating and the studies were carried out to Good Clinical Practice standards. Study product supplies and dispensing was organized by the Clinical Trials Pharmacy of Nottingham University Hospitals.

Group 1 received 1 L of PEG electrolyte solution at 1 pm on Day −1 and 1 L at 8 am the next morning on ‘Study day’. On each occasion, they were allowed 1 h to drink the 1 L solution. Group 2 was given a single dose of 2 L PEG electrolyte solution at 8 am on ‘Study day’. They were allowed 2 h to drink the 2 L solution. A range of gastrointestinal MRI parameters and stool characteristics were assessed at baseline and at intervals as shown in Fig.1. This schematic diagram shows the timeline of each stool collection, transit measurement, and MRI scans. It also shows the timing of the split and single dosing. Whole gut transit was assessed by giving the subjects inert MRI marker pills to ingest 24 h before undergoing a MRI transit scan. This was performed at Day −8 and then again at Days 14 and 28. The subjects were asked to provide samples of the first stool of the day at Day −1. The split dose group later on the same Day −1 received the first PEG electrolyte split dose, but the stool was provided before treatment. Subjects underwent MRI scanning at intervals during the study day. They attended again for morning scans at days 1, 3, 14, and 28 after purgation. They provided stool samples at Day −1, on the ‘Study day’, after dosing when the stool had become liquid and at Day 14 and at Day 28. All baseline MRI scans were taken on the study day after an overnight fast and before ingesting the PEG electrolyte. Throughout the study, subjects completed a daily Bristol Stool Diary,²⁵ documenting bowel symptoms, stool frequency, and consistency. The average of values from the stool diary from Day −8 to Day −1 were taken to be ‘Pre-study week average’ values.

Schematic diagram of the study protocol.

Subjects

Twenty-four HVs were recruited from the general campus population by advertisement. The inclusion criteria were male or female, 18–65 years of age, body mass index (BMI) between 18 and 28 kg/m². The exclusion criteria were pregnancy, any history of serious acute or chronic illness especially gastrointestinal, use of medication known to affect gastrointestinal transit, such as opiates and constipating drugs, substance abuse, and unsuitability for MRI scanning (such as having any contra-indicated metal implants or pacemakers). Subjects were asked to avoid alcohol for 24 h before the study day. They were divided into two age–sex-matched groups of 12 subjects. At enrollment, they all filled in a Hospital Anxiety and Depression Scale,²⁶ Perceived Stress²⁷ and a Patient Health Questionnaire 15.²⁸ The demographic data are given in Table S1.

Magnetic resonance imaging

The MRI scanning was carried out on a 1.5 T Philips Achieva scanner (Philips Healthcare, Best, The Netherlands). Each volunteer was positioned supine in the scanner with a 4-element, parallel imaging body coil wrapped around the abdomen. After acquiring a scout scan to locate the position of the abdominal organs, a range of MRI scans were taken. A balanced gradient echo (balanced turbo field echo, BTFE) sequence with echo time TE = 1.19 ms and repetition time TR = 2.4 ms was used to assess gastric emptying. This comprised 20 transverse images with an in-plane resolution of 1.56 mm × 1.56 mm and slice thickness of 10 mm, with no gap between slices. A single shot, fast spin echo sequence (similar to that used for MR cholangiopancreatography [MRCP]) with effective TE = 320 ms was used to assess small bowel water content.¹⁰ This comprised 24 coronal images with in-plane resolution interpolated to 0.78 mm × 0.78 mm and a slice thickness of 7 mm, with no gap between slices (acquired voxel size = 1.56 × 2.83 × 7 mm³). A dual gradient echo imaging sequence (dual-echo fast field echo, FFE) with TE1 = 2.3 ms, TE2 = 4.6 ms, and TR = 158 ms was used to visualize the anatomy. This comprised 24 coronal plane and 45 transverse images with in-plane resolution 1.76 mm × 1.76 mm and a slice thickness of 7 mm, with no gap between slices. Each image set was acquired on a single expiration breath-hold, the duration of which varied between 13 and 24 s depending on the sequence. Including set-up and imaging the volunteers spent ∼15 min in the scanner for every time point, spending the rest of the time sitting upright in an adjacent room.

Ascending colon motility index

The motility of the ascending colon was assessed from a dynamic ‘cine’ MRI scan. This acquired a single, sagittal slice through the ascending colon every second for 2 min, with the subjects simply resting and breathing gently. The images were then used to measure the ‘motility index score’ which was calculated by dividing the ascending colon into three sections and calculating the sum of the duration of contractions in minutes observed multiplied by the number of sections of the ascending colon that showed motility. Thus, if motility was observed involving the whole ascending colon for 15 s and later involving only one-third of the ascending colon for 20 s, the corresponding motility index would be = (15 × 3) + (20 × 1) = 65 in units of segment × s.

Ascending colon chyme heterogeneity scores

The apparent heterogeneity of the ascending colon chyme was graded by an experienced operator on high resolution images of the ascending colon.²³ The grading consisted in assigning a score to the appearance of the chyme based on a predetermined score card. The scores ranged between 1 (all dark) and 5 (all bright) with 3 having a mixture of dark and bright patches. This chyme heterogeneity scoring system, which was developed in-house using a mannitol model of diarrhea, has good inter-observer (weighted kappa 0.82) and intra-observer (weighted kappa 0.84) variability.²⁹

Ascending colon chyme relaxation times T1

The longitudinal relaxation times (T1) of the ascending colon chyme were measured from a single, sagittal slice through the ascending colon using an Inversion Recovery balanced Turbo Field Echo sequence with the following parameters: 1 sagittal slice, TR/TE = 3.0/1.5 ms, field of view = 400 × 400 mm, matrix size 256 × 256, slice thickness = 10 mm, and eight different inversion times (TI) ranging from 100 to 5000 ms. Each image was acquired during a breath-hold with 15 s of free breathing between each different TI to allow for full relaxation of the MRI signal. T1 reflects water proton mobility with cerebrospinal fluid having (at 1.5 T) a T1 of the order of 2400 ms and liver tissue a T1 of the order of 500 ms. The appearance of the chyme was mostly relatively uniform throughout the region but in two of the subjects who took the single dose, the colonic contents appeared to be divided between ‘brighter/wetter’ and ‘darker/drier’ contents, suggesting a wash-in of watery ileal fluid which had not mixed well with the thicker colonic contents. In this case, the relaxation data reported are from the brighter/wetter region.

Abdominal symptoms

On the study day, the subjects were asked to complete 100 mm VAS scales assessing their bowel sensation every time they came out of the MRI scanner. The VASs assessed bloating, distension, pain/discomfort, and nausea/sickness.

Whole gut transit time

The transit scans were carried out at three time points throughout the study: 8 days before dosing (defined as Day −8), 14 days after dosing, and 28 days after dosing. For each of these occasions, the volunteers swallowed three inert MRI transit marker capsules, with as much water as they desired, 24 h before the scan and under the supervision of one of the study investigators. The inert polyoxymethylene (POM) capsules (prescription pill shaped, 2.5 cm long with 1.0 cm diameter) were filled with water, traces of Gadolinium contrast agent (0.2 mL of 15 μM Gd), and perfluorooctyl bromide (PFOB, 0.6 mL) so that they would be visible both on T1-weighted ¹H proton MRI scans and on ¹⁹F fluorine MRI scans. The dual-label transit marker pills were developed and tested under separate Ethics approval G/05/2010. On the transit measurement day, the volunteers attended the study center having fasted overnight and a transit scan was performed. A series of proton (¹H) and fluorine (¹⁹F) images were acquired on a 3T Philips Achieva scanner (Philips Healthcare) equipped with a multinuclear channel and an abdominal fluorine MRI coil (PulseTeq, Chobham, Surrey, UK) to localize the position of the capsules in the gut. On the Days 14 and 28 of the protocol, the transit measurements were performed first on the 3 T scanner and immediately after that, all the other abdominal scans were performed on the 1.5 T scanner. The transit was assessed by scoring the position of the capsules according to previous gamma scintigraphy methods with score 1 = in the small bowel, 2 = lower ascending colon, 3 = upper ascending colon, 4 = right transverse colon, 5 = left transverse colon, 6 = descending colon, 7 = recto-sigmoid, and score 8 = expelled from the rectum. The mean position was calculated to give a single final score. A recent study using improved, smaller capsules showed that this measure is moderately reproducible with a correlation between measures repeated 1 week apart being 0.69,³⁰ though there are outliers who show up to 25 h differences in transit between assessments as other have also found,³¹ variability which may in part reflect genuine changes in transit with diet and other factors rather than error in measurement. The method correlates well with the standard radio-opaque marker technique of Metcalf, r = 0.85.³²

Stool pH

The subjects were asked to provide samples of the first stool of the day at Day −1 and at Day 14 and Day 28. They were also asked to provide a sample on the main study day, after treatment when the stool had become liquid. A small amount of the stool sample was placed in a laboratory tube containing 3 mL distilled water with pH = 7. The mixture was then shaken to break up the stool using a bench vortex mixer (Fisher Scientific, Loughborough, UK) at a medium speed, for around 10–15 s, or until there were no visible solid pieces. The sample's pH was then determined as an average of two repeated measurement made using pH indicator strips (Fisher Scientific).

Image analysis

Measurements of the volume of the gastric contents were carried out on the axial images by manually tracing a region of interest around the meal within the stomach on each image slice using Analyze6 software (Biomedical Imaging Resource, Mayo Foundation, Rochester, MN, USA) and summing across the slices to determine the total volume at the different time points. The surrounding organs and gastrointestinal gas were easily discriminated and excluded from the region of interest.

The small bowel water content images were analyzed as previously described.¹⁰ Briefly, regions of interest were drawn manually around the small bowel on each of the MRCP images excluding regions such as the stomach, gall bladder, and blood vessels using in-house software on an IDL platform (IDL, Exelis Visual Information Solutions, Boulder, CO, USA). The software then calculated the number of voxels (a 3-dimensional dataset) in each of the regions of interest with an intensity above a threshold calibrated using the corresponding subject's cerebrospinal fluid.¹⁰

Colonic volumes were measured by drawing manually on each relevant slice of the coronal dual-echo fast field echo images on Analyze9 software (Biomedical Imaging Resource, Mayo Foundation). Separate regions of interest were drawn on the ascending, transverse, and descending colon as described previously and then summed across each dataset, yielding also the total colonic volume as the sum of the three regions.³³

Power and statistical analysis

The primary endpoint was the small bowel water content with colonic water content, motility, and associated symptoms as secondary endpoints. Previous work using mannitol and glucose⁸ showed that the small bowel water content was 47 mL (SD 15 mL) at 40 min postprandially after ingesting 300 mL glucose solution. We calculated that we could detect an increase of 20 mL in excess of this (which is very much less than that observed after the mannitol challenge) with 90% power using n = 12 subjects.

The data are presented as mean ± SEM. Normality of the data was tested using Shapiro–Wilk's test. The data were log transformed to achieve normal distribution after which two-way repeated measures anova was used. Normal distribution of the data was not achieved even after log transform for the ascending colon motility index, chyme heterogeneity scores, and T1, or for the abdominal symptoms on the MRI day. The temporal averages for the two treatments on the study day were compared by calculating the AUC from end of treatment to 4 h (AUC4h) using the trapezoidal method. Comparisons between groups were performed using two tailed unpaired t-test or Mann–Whitney's U-test, respectively. Correlations were assessed using Pearson's or Spearman's test, respectively. Statistical analysis was carried out using GraphPad Prism 5 (GraphPad Software Inc, La Jolla, CA, USA). Differences were considered significant when p < 0.05.

RESULTS

The two treatment groups were well matched for age, gender, BMI, and psychological assessments (Table S1). The study procedures were generally well-tolerated with only one withdrawal from Group 1.

The MRI images showed clearly that as hypothesized the ingestion of the PEG electrolyte solution caused a rapid total increase in small bowel and colonic water content (Fig.2).

Representative anatomical images (upper panel) and bowel water images (lower panel) for a subject that drank a single 2 L PEG electrolyte dose. (A) and (B) were acquired at fasting baseline before dosing. In (B), a small amount of fluid can be seen in the small bowel, the transverse colon appears very dark, and the stomach shows the presence of limited resting fluid. The other images (C–F) were taken immediately after the subject completed the 2 L PEG electrolyte dose (2 h after starting the drink). (D) It shows the small bowel full of bright fluid and a marked distension of the cecum and transverse colon, both full of bright fluid. (F) is from a more posterior image plane and shows the marked distension of the ascending and descending colon, also both full of bright fluid.

Small bowel water content

There was a significant effect of treatment on small bowel water content (p = 0.0127, Table S2). Upon dosing, the small bowel water content (shown in Fig.3) significantly increased fourfold from the fasting baseline value for both the split dose (p = 0.0010) and for the single dose (p = 0.0005). Peak small bowel water content volumes ranged between 233 to 792 mL with no difference between treatments (p < 0.8). Small bowel water volumes returned to baseline values the following day.

Colon volumes

Colonic volumes increased markedly upon dosing (Fig.4). As expected after treatment, the increase in total colonic volumes from baseline values was smaller for the 1 L split dose at 35 ± 8% (range 0–81%) than for the single 2 L dose at 102 ± 27% (range 9–289%) p = 0.0332. The largest changes were found in the ascending and transverse colon with little change in the descending colon, although most subjects defecated during the study showing the solution had passed through the descending colon. The ascending colon volumes were significantly greater with the 2 L dose, p = 0.0099 (Table S2). The volumes gradually returned to baseline values toward the end of the study day.

Regional anatomical volumes of the (A) ascending, (B) transverse, and (C) descending colon for the group of 11 healthy volunteers who took the split 2 × 1 L dose of PEG electrolyte and the separate group of 12 healthy volunteers who took the single 2 L dose of PEG electrolyte. (D) The corresponding total colonic volume. Values are mean volume (mL) ±SEM.