Nicotine decreases diarrhoea and pain in ulcerative colitis without reducing inflammation.
Nicotine decreases diarrhoea and pain in ulcerative colitis without reducing inflammation.
(i) To evaluate the effect of ulcerative proctosigmoiditis on motor functions of an uninflamed segment of descending colon; and (ii) to assess nicotine’s effects on colonic motor functions in patients and healthy subjects.
In healthy subjects (n=30) and patients with ulcerative colitis (13; 11 active, two quiescent colitis), we studied the effects of intravenous nicotine on colonic transit of solid residue by scintigraphy (healthy subjects) and on colonic motility in healthy subjects and 11 patients.
In ulcerative colitis, fasting colonic motility was increased, whereas motor response to a meal was significantly reduced; compliance was unchanged. In healthy subjects, high-dose nicotine induced transient high amplitude propagated contractions and relaxation of the descending colon followed by decreased phasic contractions. This dose also accelerated colonic transit. Low-dose nicotine (mimicking a transdermal nicotine patch) reduced colonic compliance in healthy subjects, but did not affect motor function in ulcerative colitis.
Ulcerative proctosigmoiditis increases fasting colonic motility and reduces tone response to a meal in the descending colon without affecting colonic compliance, suggesting changes in physiological responses but not intrinsic wall properties. Nicotine has dose-dependent effects on colonic motor activity in healthy subjects.
Patients with ulcerative colitis experience diarrhoea, tenesmus, or urgent defecation. The pathogenesis of these symptoms is incompletely understood. While decreased absorption and/or excessive secretion of fluids and electrolytes may partly explain liquid diarrhoea, there is often a mismatch between severity of symptoms and the degree of inflammation demonstrable endoscopically.1–5 This discrepancy suggests that symptoms may in part be attributable to other mechanisms, including disturbances of colorectal motor or sensory functions. However, the results of previous studies have been inconsistent; for example, different studies have reported either a reduced or an exaggerated postprandial colonic phasic pressure response.6–10
The rectum is poorly compliant in patients with active proctitis;11, 12 however, previous studies have not evaluated colonic tone or compliance in ulcerative colitis. The radiological appearance of ‘lead-piping’ is suggestive of impaired colonic compliance. It is conceivable that reduced colonic capacitance, or accommodation of a load of chyme delivered from the small bowel into the proximal colon may contribute to diarrhoea or urgency. Viscoelastic properties of the colon in ulcerative colitis could be altered either as a result of mucosal and submucosal inflammation, or by reflex changes in the muscularis layer. Indeed, Collins et al. have shown that mucosal inflammation alters nerve and muscle function in in vitro models of colitis.13
The prevalence of ulcerative colitis in lifetime non-smokers is five times greater than in current smokers.14–17 A causal relationship between smoking and ulcerative colitis is supported by the recurrence of colitis after smoking cessation and remission with smoking commencement in non-smokers or resumption among smokers.16, 18, 19 Clinical trials in patients with active ulcerative colitis indicate that transdermal nicotine reduces diarrhoea and urgency, but nicotine does not significantly impact the endoscopic features of inflammation.20–22 These observations suggest that alternative mechanisms, unrelated to anti-inflammatory actions, may be responsible for the beneficial clinical effects of nicotine. Nicotine receptors are abundantly present on colonic intrinsic and extrinsic nerves and in pre- and paravertebral ganglia. It is unknown whether nicotine influences colonic motility, compliance or tone in ulcerative colitis, or whether the beneficial effects of nicotine on symptoms are attributable to alterations in motor function.
We hypothesized that: (i) ulcerative colitis is associated with increased colonic tone, and reduced colonic compliance; and (ii) nicotine relaxes the colon thereby ameliorating symptoms. To test these hypotheses, we conducted a series of studies: first, to assess the effects of ulcerative colitis on colonic compliance, fasting and postprandial motility and tone; and second, to evaluate the acute effects of nicotine on colonic motor activity and tone in healthy subjects and ulcerative colitis patients. To preliminarily assess possible mechanisms of action of nicotine, we also measured plasma levels of catecholamines (dopamine, adrenaline, noradrenaline), and substance P as markers of inhibitory and excitatory innervation of the gut, respectively.
Thirteen patients with mild to moderate active ulcerative colitis, aged 27–73 years (mean ± S.E.M., 48 ± 3.7; nine male and four female), were recruited from the Inflammatory Bowel Disease Clinic at Mayo. Thirty healthy volunteers, aged 20–50 years (mean ± S.E.M., 32 ± 2.2; 20 male and 10 female), were recruited by public advertisement. None of the participants had previously undergone abdominal surgery (other than appendectomy and/or cholecystectomy). All subjects were non-smokers or had been ex-smokers for >3 months. A clinical interview and physical examination were performed; none of the subjects were taking medications that might influence gastrointestinal motility or autonomic function. Severity of ulcerative colitis was assessed endoscopically using the criteria of Baron et al.5
All participants signed informed consent to participate in the studies which were approved by the Institutional Review Board at the Mayo Clinic.
All 13 patients with ulcerative colitis underwent intraluminal assessment of colonic motor activity because our major objective was to compare colonic motor function in controls and patients with ulcerative colitis. We also assessed the effects of nicotine on fasting colonic motility and compliance in three patients with ulcerative colitis.
Sixteen healthy subjects underwent intubated colonic motility studies; eight received placebo, four received low dose nicotine, and four received high dose nicotine. The effect of nicotine on colonic transit was measured in a separate group of 14 healthy subjects: eight subjects were randomized to receive nicotine (three high dose nicotine; five low dose nicotine) and six were randomized to receive placebo.
Due to human safety considerations, the Institutional Review Board suggested that the subject, but not the investigator conducting the studies (BC), be blinded to whether they were randomized to intravenous nicotine or placebo. Physiological (0.9%) saline served as the placebo and was administered as a ‘bolus’, followed by an infusion. The ‘bolus’, continuous infusion volume and rate of administration were identical for the nicotine and saline.
Identical dosing schedules were employed for the colonic motility and transit studies. Two different doses of nicotine (Sigma, St Louis, MO) were prepared by the Mayo Clinic Pharmacy and were administered as an intravenous bolus infusion (0.5 μg.min/kg over 30 min [low dose] or 2.5 μg.min/kg over 15 min [high dose]) followed by continuous infusion (0.133 μg.min/kg over 150 min following the low dose, or 0.7 μg.min/kg over 165 min following the high dose). The infusion started 30 min after assessment of fasting colonic tone. The lower dose of nicotine infusion mimics the maximal plasma concentrations (Cmax) reached during application of a 22 mg nicotine dermal patch.23, 24 Both doses were tested in the healthy control group (n=9 for low dose; n=7 for high dose); however, the patients with ulcerative colitis only received the lower dose (n=3). Because intravenous nicotine induces emesis frequently when administered with a meal, study subjects who received either dose of nicotine did not receive a meal and, hence, no postprandial data are available from studies in which nicotine was administered.
We continuously monitored arterial oxygen saturation and blood pressure using a pulse oximeter (CO2SMO, Novametrix Medical Systems, Inc., Wallingford, CT), and Finapres sphygmomanometry (Ohmeda, Madison, WI).
On the evening prior to the study, all healthy subjects who participated in the intubated colonic motility arm of the study were admitted to the General Clinical Research Centre for bowel preparation. This consisted of 1.5–2 L of polyethylene glycol and electrolyte solution (OCL; Abbott Laboratories, Chicago, IL) which was ingested until the faecal effluent became a clear liquid. Patients with ulcerative colitis underwent a colonoscopy on the morning of the study without prior admission to the General Clinical Research Centre. Their bowel preparation was identical to that of the healthy subjects.
After an overnight fast, colonoscopy was performed with sedation and analgesia; sedation was induced by midazolam maleate 2–3 mg administered intravenously (Versed, Roche Pharmaceuticals, Nutley, NJ), and meperidine hydrochloride 25–75 mg was administered intravenously (Demerol, Sanofi Pharmaceuticals, NY) for pain relief. After examination of the entire colon in the patients and the left colon in healthy controls, a 4 m Teflon-coated guidewire (Microvasive, Hobbs Medical, Stafford Springs, CT) was placed with its tip at the splenic flexure and the colon was deflated as the colonoscope was withdrawn. The balloon assembly was introduced into the colon over the guidewire and positioned under fluoroscopic control with the balloon in the mid-descending colon. At the end of the endoscopic procedure, the benzodiazepine antagonist, flumazenil 0.4 mg intravenous (Romazicon, Roche Pharmaceuticals, Nutley, NJ) was administered to all subjects.
A multi-lumen polyethylene tube assembly incorporating a polyethylene balloon and several manometric point sensors was placed into the prepared colon as described in previous studies.25, 26 Tonic and phasic contractile activity of the colon were measured as in previous studies, using an infinitely compliant 10 cm long balloon with a maximum volume of 600 cc (Hefty Baggies, Mobil Chemical Co., Pittsford, NY) linked to an electronic barostat (Mayo Barostat, Rochester, MN) which has a rigid piston.27, 28 The manometric portion of the tube comprised six water-perfused (0.4 cc/min) ports located in the upper descending (sensor numbers 1–2), mid-descending and sigmoid colon (sensor numbers 5–6); manometric sensors were 5 cm apart.3, 4 The first and second sensors were 5 cm orad and caudad to the balloon, respectively (Figure 1).
The intraballoon pressure at which respiratory excursions were regularly recorded as changes in barostat volume was defined as the ‘minimum distending pressure’. The ‘operating pressure’ was set 2 mmHg above the minimum distending pressure (median pressure 10 mmHg, range 6–14 mmHg). Intraballoon volumes and manometric pressure changes in response to wall contractions and relaxations were monitored continuously throughout the study. A pneumobelt was placed around the abdomen at the level of the lower costal margin to help exclude artefact induced during movement and coughing.
The barostat balloon was inflated to the operating pressure after a 30 min equilibration period. After another 30 min equilibration period, we assessed colonic compliance by increasing the balloon volume as the balloon was distended in 2 mmHg increments at 60 s intervals from 2 to 36 mmHg. The intrinsic compliance of the rigid piston in the barostat used for this study is nearly zero. After measuring colonic compliance, we recorded fasting colonic tone for 30 min before and for 90 min after drug administration; a second colonic compliance test was then performed. Finally, in patients and healthy volunteers who did not receive intravenous nicotine, a 1000 kcal liquid meal (23% carbohydrate, 53% fat, 19% protein) was ingested and colonic motility was recorded for a further 90 min.
Fourteen healthy subjects participated in the colonic transit study. They were admitted to the General Clinical Research Centre on the morning of the study after an overnight fast. Women of childbearing potential underwent a plasma β-human chorionic gonadotrophin pregnancy test within 48 h of the start of transit studies. Transit in the unprepared colon was measured by means of scintigraphy, using a pH-sensitive, methacrylate-coated capsule containing 111In-labelled activated charcoal.29, 30 Labelling of the activated charcoal and preparation of the methacrylate, pH-sensitive capsule have been described previously.31 The pH-sensitive capsule dissolves at an alkaline pH in the terminal ileum.
After the capsule was ingested, anterior gamma camera scans were taken at regular intervals obtained with the subject standing to assess progression of the capsule. The colonic transit study protocol was initiated when at least 75% of the total isotope counts was in the cecum or ascending colon (AC), which was defined as time 0. Administration of drug or placebo started at time 0. Dual (anterior and posterior) gamma camera scans were then obtained with the subject standing at time 0 (just before start drug/placebo), 15, 30, 45, 60, 120, 180, and 240 min after starting the administration of drug or placebo. The imaging window for 111In was set at 247 ± 20 keV.
During the intubated colonic study and colonic transit study in healthy subjects, blood samples were taken at 10 min before and after the start of drug/placebo administration for the measurement of blood levels of nicotine or metabolites and serum catecholamines (dopamine, noradrenaline, adrenaline) and substance P.32, 33
In analysing the compliance curve, we averaged the barostat balloon volume over a 60 s interval at each pressure setting. We have previously shown that colonic compliance in humans is non-linear.27 Therefore a linear model may not detect differences between different segments of the compliance curve. We used a non-linear, power exponential model to approximate colonic pressure–volume relationships wherein the volume (V) at any given pressure (P) is defined as:27
where relative pressure (Rel P)=(1/P), Vmax is the maximum volume, Pmax is the maximum pressure in the compliance assessment, and r is the ratio of the minimum to the maximum volume in the compliance curve. The parameter β reflects the overall shape of the curve while κ is essentially the change in volume as a function of 1/P at any given point, and represents the slope of the compliance curve.
Only the data from the second compliance curves performed 90 min after the start of infusions, were used for comparison between nicotine and placebo.
Phasic manometric pressure activity and changes in both pressure and volume of the barostat balloon were sampled as analogue signals at 8 Hz and converted to digital signals before entry into a computer. As in previous studies, movement and respiratory artefact were filtered out using a modified VAX LAB filtering program (Digital Equipment Corporation, Boston, MA).25 All data were analysed by a customized computer program which assessed phasic contractions and changes in the volume within the barostatically controlled balloon.25 On-line continuous recordings of all other parameters were similarly collected independent of the investigator.
Previous studies showed that phasic volume peaks recorded by the barostat balloon occur at a frequency of three per minute;28 the waveform was filtered to remove frequencies > 6 cpm. The barostat balloon volume was then computer analysed to separate baseline balloon volume, representing colonic tone, from phasic volume deflections of > 10 mL.
For the manometric record, the waveform was initially smoothed using a high frequency filter to remove instrumentation artefact. All contractions with a pressure change > 10 mmHg amplitude above baseline, duration from 1.5 to 60 s and an interpeak interval of > 1.5 s were selected. These parameters are based on previous data showing that > 98% of all colonic myoelectric activity has a frequency range of 1–11 per minute.34 Phasic pressure activity was calculated early (0–30 min after start of infusion) and later (60–90 min after start of infusion), and expressed as the motility index per hour:
as in previous studies for the upper descending colon (sensors 1–2), lower descending colon (sensors 3–4) and sigmoid colon (sensors 5–6).35 High amplitude propagated contractions were identified as contractions with amplitude > 75 mmHg, propagation over at least 20 cm, and propagation velocity of 0.5–2.2 cm/s.
For an analysis of transit data, a region of interest program was used to estimate the amount of radioactivity in the ascending colon (AC), transverse colon (TC), descending colon (DC), and the rectosigmoid (RS). Counts were corrected for radionuclide decay. The geometric mean of anterior and posterior counts was calculated (square root of product of anterior and posterior counts), and for each scan the proportion of isotope was estimated by dividing the counts in that region by the total counts of radioactivity in the entire colon. Ascending colonic transit was also summarized as percentage isotope retained in the ascending colon 30 and 60 min after the start of nicotine or saline infusion.
Differences in colonic tone response to a meal, colonic compliance parameters, and motility index between healthy volunteers and ulcerative colitis patients were compared by Wilcoxon signed rank test or unpaired t-test depending on the distribution of the data. The effects of nicotine and placebo on colonic basal tone, compliance and transit, and plasma values of measured hormones in healthy subjects, were compared using the same statistical methods. The level of significance was 0.05. Results are expressed as mean ± S.E.M.
The clinical and endoscopic characteristics of the patients with ulcerative colitis are summarized in Table 1. None of the patients had evidence of ‘lead piping’ of the descending colon on recent radiography or endoscopy. In nine out of 13 patients, the extent of endoscopic disease activity was limited to the proctosigmoid region; in two, there was active disease reaching into the proximal descending colon; the two patients with pancolitis had mild or quiescent disease.
Ulcerative colitis had a significant effect on fasting but not postprandial phasic pressure activity and the tonic component of the colonic response to a meal when compared to healthy controls. Fasting phasic pressure activity in the upper descending colon (sensors 1–3) was significantly greater in ulcerative colitis patients (MI=11.4 ± 0.8 vs. 9 ± 0.6 in controls; P=0.02). Phasic pressure activity in the lower descending and sigmoid colon (sensors 4–6) were not significantly different in ulcerative colitis patients and healthy controls. The postprandial reduction in barostat balloon volume, representing an increase in tone, was significantly lower in ulcerative colitis patients compared to controls (ΔV=20 ± 12 mL vs. 60 ± 10 mL; P < 0.02; Figure 2). Colonic compliance (Table 2) and postprandial phasic activity (data not shown) did not differ between the two groups.
As shown in Figure 3(A and B), intravenous administration of high dose nicotine was accompanied by changes in both phasic and tonic pressure activity in the descending colon. Three out of four healthy subjects who received high dose nicotine developed high amplitude propagated contractions [high amplitude propagated contractions, mean number per 30 min=3 ± 1 vs. 0 (saline); P < 0.05] within 30 min after starting the infusion. Thereafter, nicotine suppressed phasic motor activity (P < 0.05) in the upper and lower descending colon regions compared to placebo (Figure 4). In contrast, low dose nicotine did not induce high amplitude propagated contractions or alter phasic motor activity (data not shown).
High dose nicotine significantly increased baseline balloon volume (Δ post – pre-baseline volume=32 ± 16 mL, P < 0.05 vs. placebo), indicating a reduction in fasting colonic tone or relaxation. Although low dose nicotine also decreased colonic tone (Δ post – pre-baseline=– 19 ± 8 mL), this effect did not reach statistical significance.
Low and high dose nicotine altered colonic compliance (Figure 5), however, whereas high dose nicotine shifted the pressure–volume curve to higher volumes (increased compliance) compared to the control group, low dose nicotine decreased compliance: Δκ=3.5 ± 1.3 for high dose nicotine; Δκ=−1.8 ± 1.4 for low dose nicotine (P < 0.05 vs. placebo for both; Figure 5).
The nicotine infusion was commenced after ≥ 75% of 111In had emptied from the ileum into the ascending colon. High dose nicotine significantly accelerated emptying of isotope from the ascending colon at 30 min (30 ± 10% vs. 3 ± 3% placebo; P < 0.05; Figure 6); no significant effect was detected with low dose nicotine (data not shown).
High dose nicotine increased plasma adrenaline (Δ post – pre-infusion=216 ± 122 pg/mL nicotine vs. 0.3 ± 1.9 pg/mL placebo, P < 0.02). Low dose nicotine did not affect plasma adrenaline levels. Noradrenaline, dopamine, and substance P levels were not significantly altered by either dose of nicotine relative to placebo.
Known side-effects of nicotine such as excessive sweating, nausea, vomiting, light-headedness, and increase in blood pressure and heart rate were invariably present in the healthy subjects receiving a high dose of nicotine. Low dose nicotine did not result in any clinically significant side-effects.
The present data demonstrate that fasting phasic pressure activity (but not tone or compliance) in the upper descending colon is significantly greater in ulcerative colitis patients compared to healthy controls. In contrast, the colonic contractile response to a meal is significantly reduced in patients with ulcerative colitis compared to that of healthy controls.
Intravenous nicotine has dose-dependent effects on colonic motor activity in healthy subjects. Low dose nicotine infusions, at concentrations which mimic a 22 mg nicotine dermal patch, do not affect phasic contractility or tone of the colon, but they do reduce colonic compliance. A high dose of nicotine induces an initial burst of high amplitude propagated contractions while relaxing the descending colon. Colonic transit is accelerated soon after beginning a nicotine infusion. Thereafter, colonic relaxation persists and high amplitude propagated contractions and non-propagated phasic contractile activity subside during a continuous infusion of nicotine.
Our data show that ulcerative colitis, primarily affecting the proctosigmoid region and unassociated with ‘lead piping’, does not alter compliance in the descending colon. In contrast to our data, others have shown decreased rectal compliance in patients with active ulcerative proctitis.11, 12 This discrepancy can be explained by a difference in the colonic region studied and the severity of inflammation of that region. In nine out of 13 patients in our study, the disease was limited to the rectosigmoid, and compliance was assessed in the endoscopically normal descending colon, in contrast to the previous studies of rectal compliance in patients with active proctitis.11, 12
An impaired colonic tone response to a meal was evident in patients with either mild or no inflammation in the descending colon. This raises the possibility that neuromuscular responsiveness to physiological stimuli may be altered at sites distant from the region of active mucosal inflammation. Supporting this concept are observations by Collins et al. in experimental models of mucosal inflammation induced by Trichinella spiralis infection that demonstrate abnormal enteric nerve and muscle functions even in uninflamed tissue.13 However, the pathogenesis of these disturbances in uninflamed tissue is unclear because alterations in ion channels that regulate contractility have only been reported in inflamed segments.36, 37
Given the presence of nicotine receptors on intrinsic and extrinsic neurones innervating colonic smooth muscle, and the discrepancy between therapeutic effects on inflammation and symptoms in ulcerative colitis, we hypothesized that motor effects of nicotine would, at least partly, explain its beneficial effects on symptoms. Previous studies provided some support for this hypothesis, in view of the report of a dose-dependent reduction in colonic transit time with transdermal nicotine in healthy non-smokers, a reduction in colonic tone with intrarectal nicotine, and a biphasic effect of smoking on distal colonic phasic pressure activity in chronic smokers.38–40
Our study expands on these observations by assessing the dose-dependent effects of nicotine on fasting colonic tone and compliance and the postprandial colonic response. While low dose nicotine did not affect fasting phasic pressure activity, colonic tone or transit, high dose nicotine induced a series of alterations in motor activity. The effects of high dose nicotine include an initial surge of high amplitude propagated contractions, followed by a pronounced, persistent inhibition of colonic tone and phasic pressure activity. We suspect that these events, in part, reflect the several known neurohumoral effects of nicotine accompanied by rapid tachyphylaxis at nicotine receptors.41 Thus, high amplitude propagated contractions may be induced by activation of nicotine receptors on cholinergic or neurokinergic (substance P, neurokinins A and B) motor neurones.42, 43 The relatively short burst of high amplitude propagated contractions may be explained either by rapid desensitization and tachyphylaxis at nicotine receptors, or by the pronounced increase in plasma adrenaline levels, reflecting stimulation of nicotine receptors in the adrenal medulla.44, 45 Increased adrenaline levels may inhibit high amplitude propagated contractions and reduce colonic tone and phasic pressure activity. Adrenaline evokes a relaxing effect on longitudinal and circular muscles of the proximal colon from cattle; these effects were mediated by beta and alpha-1 receptors, respectively.46
Low dose nicotine, at concentrations designed to mimic the nicotine patch, did not influence colonic transit in healthy subjects or colonic motility in patients with ulcerative colitis. These observations do not support the hypothesis that the beneficial therapeutic effects of nicotine in ulcerative colitis are consequent to its motor effects. We acknowledge that the pharmacokinetics of intravenous and topical nicotine may differ. While the intravenous route mimics the maximum and the steady state concentrations achieved with the transdermal nicotine patch, other pharmacokinetic parameters (e.g. time to reach maximal plasma concentration) may differ between the two methods of administration.23, 24 Hence, we cannot completely exclude an effect of transdermal nicotine on colonic motor functions. Future studies are necessary to assess the effects of long-term transdermal nicotine on colonic transit and motility in healthy participants and in patients with ulcerative colitis, to dissect the effects of nicotine on inflammation and motility.
This study was sponsored in part by use of the Physiology Core and Immunochemistry Core Laboratories of the General Clinical Research Center (NIH grant #RR00585). Dr Camilleri is supported by grant #R01-DK54681-03, #K24-DK02638-02 from the National Institutes of Health. We thank Ms Cindy Stanislav for secretarial support.