| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original Studies |
Psychophysiology Unit, Department of Surgery (G.St., G.St.-J., U.W.), and Departments of Biomedical Engineering and Physics (H.B.) and Nuclear Medicine (H.B.), University of Vienna, A-1090 Vienna; Department of Medicine I, Krankenanstalt Rudolfstiftung (G.Sc., H.-P.K.) A-1030 Vienna; Ludwig Boltzmann Institute of Nuclear Medicine (H.B.); A-1090 Vienna; and Rehabilitationszentrum der PVA (M.F.), A-2356 Alland, Austria
Address all correspondence and requests for reprints to: Prof. Georg Stacher, M.D., Psychophysiologisches Laboratorium, Waehringer Guertel 18–20, A-1090 Wien, Austria. E-mail: georg.stacher{at}akh-wien.ac.at
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The failure of cisapride to enhance gastric emptying in the majority of the long term studies referred to may be attributable to various factors. High blood glucose levels as well as the presence of autonomic neuropathy in the patients studied may have precluded stimulatory effects. Up until now, however, the impact of neither of these two factors on the effects of cisapride has been studied. Even the relationship between cardiovascular autonomic neuropathy (CANP) and delayed gastric emptying is unclear. Whereas some investigators reported that emptying was slower with more severe CANP (17, 18, 19, 20) or delayed more often in patients with than without CANP (21, 22, 23, 24), others reported that disordered gastric and small intestinal motility was not significantly associated with CANP (25, 26, 27). Finally, the rate of emptying prevailing in individual patients studied may have blunted the effects of cisapride; there is evidence that cisapride has marked accelerative effects on delayed but only minor effects on normal emptying (28).
The present study therefore was aimed at determining, in patients with IDDM and unstable glycemia, 1) the effects of 8-week oral administration of cisapride vs. placebo on glycemic control and gastric emptying under random, cross-over, double blind conditions; and 2) the effects of cisapride, with respect to the prevailing blood glucose levels, on the presence of CANP and on delayed vs. nondelayed gastric emptying (for definition, see Materials and Methods).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The study was designed as a randomized, cross-over, double blind trial. The design accounted for two strata: patients with delayed and patients with nondelayed emptying (for definition, see below). Half of the patients in each stratum were to receive oral doses of 10 mg cisapride four times daily (QID) for 8 weeks, then placebo QID during a 4-week washout period, and finally placebo QID for another 8 weeks; the other half of the patients were to receive placebo in the first 8 weeks, as well as in the 4-week washout period and cisapride in the second 8 weeks. The randomization plan was generated by Janssen Research Foundation. The code was kept there and broken when all data were listed in Case Record Forms, in which no treatment assignments were entered. The medication was provided as identically appearing cisapride and placebo tablets. For each patient, the investigators received a sealed envelope to be opened in case of emergency only, which contained the individual treatment code. The study was monitored throughout by Janssen.
For the detection of a difference between the treatment effects on the primary outcome measure, i.e. the SD of the within-patient mean blood glucose level over an 8-week treatment period, the target sample size was projected assuming a between-subject SD of glycemic variability of 11% in placebo-treated patients (Slama, G., personal communication). As the intraindividual SD can be expected to be about half as high as the interindividual SD, it was estimated that six patients were required for each treatment sequence in each stratum to detect a difference between the treatment effects as indicated by an intraindividual SD of 0.555 mmol/L blood glucose or more with a power of at least 80% and an error probability of less than 5% using a two-tailed test.
The study was conducted in accordance with the Declaration of Helsinki as revised in 1996. It was approved by the joint ethical committee of the Faculty of Medicine, University of Vienna, and the General Hospital of the City of Vienna. All patients gave written informed consent to participate.
Inclusion criteria and randomization
Patients with classical IDDM were eligible for inclusion if they were C peptide negative, 19–70 yr old, were receiving intensified insulin treatment (Basis Bolus Insulin Therapy), had a history of unexplained unstable glycemia, and could be expected to be compliant with drug intake, glucose monitoring, and the completion of a diary. Further, patients were eligible if hematological, biochemical, and urine analyses showed no gross abnormalities; serum creatinine levels were less than 221 µmol/L; there was no significant organic impairment; and there was no illness apart from IDDM. Patients were not eligible if they had used an investigational drug 30 days before the initial assessment; were taking medications potentially interfering with the aims of the study, i.e. cholinergics, anticholinergics, spasmolytics, antiemetics, dopamine antagonists, opioids, and gastroprokinetic agents including macrolide antibiotics; had undergone major gastrointestinal surgery; and were known or suspected to use illicit drugs or to abuse alcohol. Females of child-bearing age had to be using adequate contraceptive control and not pregnant or lactating. Patients were included in the study if they had, during a 4-week phase (basal phase), a SD of the mean blood glucose of more than 2.78 mmol/L, i.e. moderate to severe glycemic instability.
Procedure
Eligible patients underwent evaluations of gastric emptying and
CANP as described below and completed a symptom questionnaire. They
were trained to make capillary blood glucose measurements using a
monitor with automatic data storage (One Touch II, Life Scan, Ortho
Diagnostic Systems, Milpitas, CA). When this was accomplished, the
patients were asked to make at least four blood glucose measurements
per day over a 4-week period. Half of the measurements were requested
to be made just before breakfast, lunch, and supper, and the other half
to be made 1.5–2 h postprandially. The mean blood glucose and the
number of biological hypoglycemic (
2.78 mmol/L) and hyperglycemic
(
13.88 mmol/L) events per week were determined. Patients were
provided with a diary to record the incidence and severity of clinical
hypoglycemic events and other relevant information, such as missed
meals, sports, insulin doses, and illnesses. A clinical hypoglycemic
event was classified as mild if the patient had symptoms but was able
to help her/himself and as severe if she/he required assistance or coma
was induced.
At the end of the basal 4 weeks, the patients were asked to register blood glucose every 3 h for at least 15 h while maintaining their daily routine. On the subsequent day and after an overnight fast, venous glucose levels were determined before as well as 30, 60, and 120 min after a standard breakfast. Patients completed a quality of life questionnaire and their hemoglobin A1c (HbA1c) levels were determined. During the treatment phase, patients were required to take tablets containing 10 mg cisapride QID, i.e. 30 min before breakfast, lunch, and supper as well as at bedtime. Further, they were instructed to make at least four glucose measurements per day, i.e. one just before each main meal (breakfast, lunch, and supper) and one at bedtime, and to complete their diaries daily. After each of the two treatment periods, venous glucose levels in response to a standard breakfast were evaluated again, and the patients completed a symptom and a quality of life questionnaire. Further, gastric emptying and HbA1c were recorded, and hematological, biochemical, and urine analyses were performed to evaluate drug safety.
Patients
Twenty-eight out-patients entered the basal phase; 4 of them dropped out during that phase because they found the study protocol too demanding. Of the 24 patients included in the treatment phase, 1 dropped out after a short time for the same reason. Twelve females (7 postmenopausal) and 11 males completed the study; their median age was 45 yr (interquartile range, 32–53 yr), their median body mass index (BMI) was 22.8 kg/m2 (21.2–25.4 kg/m2), their median IDDM duration was 20 yr (12–27 yr), and their median HbA1c at inclusion was 8.4% (7.5–8.9%). Nineteen patients injected ultralente insulin twice daily, i.e. between 0600–0800 h as well as between 1000–1100 h, and 4 injected ultralente insulin in the morning and neutral protamine Hagedorn insulin at bedtime. Before breakfast as well as at lunch and supper, all patients injected a standardized bolus of regular insulin. Seven patients had nonproliferative retinopathy, 4 had mild nephropathy (albumin excretion, 30–300 mg/24 h); 14 had delayed and 9 had nondelayed gastric emptying (for definitions, see below).
Assessment of gastrointestinal symptoms and quality of life
Dysphagia, heartburn, chest pain, regurgitation, early satiety, postprandial epigastric fullness, anorexia, nausea, vomiting, and abdominal discomfort were evaluated using a standard questionnaire at the start of the basal phase as well as after each of the two treatment periods. The frequency of symptoms was scored as follows: 0, absent; 1, present rarely; 2, present frequently; 3, present very often; and 4, always present. The mean frequency was evaluated as the total symptom score. The number of bowel movements per week and day and the consistency of stools were also noted. The quality of life was assessed using the questionnaire employed in the Diabetes Control and Complications Trial (29).
Evaluation of CANP
CANP was assessed as described previously (20) by four tests, i.e. the resting heart rate variation, the maximum minus minimum heart rate difference, the heart rate response to standing up, and the heart rate response to a Valsalva maneuver. Further, the blood pressure response to standing up was measured. The result of each of the tests was classified as normal; borderline, i.e. between the 90–95% confidence limit for healthy subjects; or abnormal, i.e. beyond the 95% confidence limit. The overall classification was as follows: 1, normal; 2, borderline CANP, i.e. 1 test result abnormal or 2 borderline; 3, definite CANP, i.e. 2 or 3 results abnormal; and 4, severe CANP, i.e. 4 or 5 results abnormal.
Measurement of gastric emptying
The emptying of a semisolid 1168-kJ meal was recorded
scintigraphically (30). The ingredients were 250 ml milk (9.2 g fat,
8.5 g protein, and 12.3 g carbohydrates), 15 g sugar,
14 g maize starch (11.9 g carbohydrates), and, for flavoring,
cinnamon; its osmolality was 558 mmol/kg. The meal was cooked under
continuous stirring until a semisolid consistency was reached and,
after cooling to a temperature at which it could be eaten, was mixed
with 60 MBq [99mTc]phylate in isotonic saline. The
99mTc radiolabel is bound reliably to the meal (30).
Recording began at the end of ingestion and lasted 120 min. The
patients sat in an armchair tilted 30° backward to avoid
overprojection of stomach and small intestine. A posteriorly located,
single headed 
-camera (GCA901A, Toshiba Medical Systems, Tokyo,
Japan) was used. Serial images over 1 min each were acquired in the
frame mode. A region of interest including the stomach was drawn.
Counts in this region were corrected for the variation in tissue
attenuation caused by the changing intragastric position of the meal
and the use of a single headed camera (31) as well as for background
activity and radionuclide decay. An activity-time curve, expressed as
percentages of counts in the region immediately after ingestion, was
derived. The percentage of the meal remaining in the stomach at 120 min
(residual radioactivity, RRA120) and the area under the
emptying curve (AUC120) were taken as overall measures of
emptying. The shape of the emptying curves was analyzed using the power
exponential and the parameter ß (32). A ß more than 1 characterizes
an initially slow emptying followed by a more rapid emptying, whereas a
ß less than 1 describes a fast initial and subsequent slower
emptying. The radiation burden was 1.7–5.4 milli-Sievert,
i.e. less than from a barium contrast investigation. The
capillary blood glucose concentration was measured just before meal
ingestion as well as 60 and 120 min after the termination of the
meal.
According to the results of the study performed in the basal phase, patients were classified as having delayed (n = 13) or nondelayed (n = 10) emptying. The criterion for delayed emptying was a residual radioactivity at 50 min (RRA50) of more than 73.2%, i.e. more than the mean RRA50 + 1.5 SD (53.5 + 19.7%) of 48 healthy, symptom-free young men (median age, 25 yr (interquartile range, 23–27 yr); and a median BMI of 22.2 kg/m2 (20.3–23.3 kg/m2). Ten of the 13 patients with delayed emptying had an RRA50 greater than the RRA50 ± 2 SD of the healthy subjects.
Statistical analysis
Effects on the SD of the mean within-patient blood glucose level were examined by ANOVAs accounting for the influences of the random factor patient (no. 1–23), the fixed between-subject factors stratum (delayed emptying, nondelayed emptying) and CANP (normal/borderline, definite/severe), and the fixed within-subject factors treatment (cisapride, placebo) and sequence (cisapride-placebo, placebo-cisapride).
For the residual radioactivity determinations and the blood glucose
levels in the course of the emptying studies, ANOVAs for repeated
measures were carried out. They accounted for the random factor
patient; the fixed between-subject factors sex, illness duration, BMI,
stratum, and CANP; and the fixed within-subjects factors treatment,
sequence, and period (residual radioactivity at 40, 80, and 120 min;
blood glucose before meal ingestion as well as at 60 and 120 min). In
addition, comparisons were performed using Student’s t test
for paired data and correlation analyses for continuous variables, and
the 
2 test for categorical variables. For the rejection
of the null hypothesis, an error probability of P <
0.05 (two-tailed) was considered significant. Data are expressed as the
mean ± SEM. To investigate the degree to which the
factors patient, sex, illness duration, BMI, stratum, CANP, treatment,
sequence, blood glucose before meal ingestion, and blood glucose rise
over fasting levels predicted the RRA120, a stepwise
multiple linear regression analysis was carried out.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Glycemic control was unaffected by cisapride; the
SD of the within-patient mean capillary blood glucose, the
mean blood glucose, as well as the numbers of biological hypoglycemic,
clinically mild and severe hypoglycemic, and biological hyperglycemic
events did not differ from the values recorded during placebo treatment
(Table 1
). There was also no difference
between the values recorded during cisapride and placebo
treatments, respectively, and those recorded in the basal phase. The
plasma blood glucose profiles in response to the standard breakfast
after cisapride did not differ significantly from those
after placebo and in the basal phase (Table 2). HbA1c after
cisapride differed neither from the HbA1c
after placebo nor from that in the basal phase (Table 1). After
cisapride as well as after placebo, patients with delayed
emptying did not differ in HbA1c (8.5 ± 0.3%
vs. 8.6 ± 0.3%) from patients with nondelayed
emptying (8.2 ± 0.3% vs. 8.4 ± 0.2%), and
patients with CANP had HbA1c levels (8.3 ± 0.4%
vs. 8.6 ± 0.4%) not different from those in subjects
without CANP (8.3 ± 0.2% vs. 8.5 ± 0.2%).
|
|
After cisapride, emptying did not differ
significantly from that after placebo [treatment factor, F(1, 20)
< 1; Fig. 1]. The treatment effects
differed neither in patients with delayed nor in those with nondelayed
emptying [interaction treatment-stratum, F(1, 20) < 1]. The
meal was emptied faster after cisapride than after placebo
in 8 of the 14 patients with delayed and in 7 of the 9 with nondelayed
emptying [2 (1) = 1.70; P =
NS]. The shape of the emptying curves, as described by the parameter
ß, did not differ between cisapride and placebo (Table 3) or between patients with delayed and
those with nondelayed emptying.
|
|
2 (1) = 3.65;
P < 0.1 (NS)]. Patients with CANP were older than
those without [48.3 ± 4.2 vs. 39.7 ± 2.7 yr;
t (21) = 5.96; P < 0.0001] and had
a longer illness duration (24.6 ± 4.1 vs. 17.6 ±
2.1 yr; P = NS). In the 15 patients with no CANP,
emptying was somewhat faster after cisapride than after
placebo [area under the curve at 120 min, 6858 ± 335
vs. 7662 ± 335 U; t (14) = 1.95;
P < 0.07 (NS)], whereas in the 8 patients with CANP,
the emptying after cisapride was virtually identical to
that after placebo (area under the emptying curve at 120 min
(AUC120), 7589 ± 698 vs. 7358
± 825 U; Fig. 2). This was reflected by
the F ratios for the interactions treatment-CANP [F(1, 20) =
2.11; P = NS] and period-CANP [F(2, 20) = 3.52;
P < 0.04]. The shapes of the emptying curves did not
differ between patients with and without CANP. Sex, illness duration,
BMI, and treatment sequence had no influence on emptying.
|
Gastrointestinal symptoms and quality of life
The number of patients with symptoms as well as the reported
nature of the symptoms and the total symptom score (Table 1
) after
cisapride treatment did not differ from the respective
data after placebo and at the initial assessment. The number of stools
per week was greater after cisapride than after placebo in
two patients, less in two, and the same as after placebo in 17. The
total quality of life score after cisapride differed
neither from that after placebo nor from that in the basal phase (Table 1).
Adverse effects
No potentially drug-related adverse effects occurred. One patient developed a slight facial paresis after 8 weeks of placebo treatment. She continued to participate in the study, and the paresis faded away within 3 weeks.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous cross-over studies showed that the rate of gastric emptying is determined in part by the prevailing blood glucose level (4, 5, 6). Fraser et al. (4) reported that 8 of 10 IDDM patients emptied solid and 9 of 10 patients emptied liquid meal components slower with blood glucose stabilized using a clamp technique at 16–20 mmol/L than at 8 mmol/L. Concordant findings were obtained by others (6). At 8 mmol/L, a glucose level slightly lower than the ones prevailing in the present study, the emptying of liquids as well as solids in 9 IDDM patients was slower than that at 4 mmol/L (5). Thus, hyperglycemia could have accounted for the failure of cisapride treatment to accelerate gastric emptying in the present investigation. However, in contrast to the above-mentioned studies, emptying was only slightly slower with higher preprandial blood glucose levels in our patients. This finding is consistent with earlier cross-sectional observations in patients with long standing IDDM, in whom no correlation was found between the emptying rate, on the one hand, and the glucose level before ingestion of the meal, of which the emptying was assessed (3, 36). That there is no close relationship between the blood glucose concentration during the assessment of emptying and the rate of emptying is suggested also by another cross-sectional study in 86 diabetic patients; although the emptying of liquid was slower in patients with a mean blood glucose level greater than 15 mmol/L than in those with a mean glucose level less than 15 mmol/L, the solid meal components were not emptied differently (2). The direct relationship found in the present investigation between the rate at which the semisolid meal was emptied and the postprandial increase in blood glucose above fasting levels corresponds to earlier observations made after the administration of liquid meals (1, 2, 3): the faster the emptying, the faster the rise in blood glucose. By contrast, the administration of a human amylin analog, which yields a slowing of emptying (37), resulted in a reduction of postprandial (38) as well as 24-h plasma glucose levels (39).
The finding that cisapride had no effect different from those of placebo on the frequency of occurrence and the nature of gastrointestinal symptoms is consonant with earlier observations. In two long term, double blind, cross-over trials (13, 14) cisapride failed to reduce the frequency and/or severity of symptoms more than did placebo. Beneficial effects of cisapride, by contrast, were noted in open studies (10, 11, 40).
It can be concluded that cisapride, in the number of patients studied and the dosage assessed, 1) did not affect glycemic control; 2) had no effect on gastric emptying, which differed significantly from the effect of placebo; 3) did not act differently in patients with delayed and those with nondelayed emptying; 4) slightly accelerated emptying in patients with no or only borderline CANP, but not in those with definite CANP; and 5) did not affect the frequency of occurrence or the nature of gastrointestinal symptoms or the self-rated quality of life. Further, the rate of gastric emptying was not related to the blood glucose levels before and during the emptying studies, whereas the blood glucose rise over fasting levels was directly related to the rate of gastric emptying. In the dosage investigated, cisapride does not seem suited as a means to accelerate slow gastric emptying and thereby to improve glycemic control in patients with IDDM. It remains to be investigated whether a higher dosage, i.e. 20 mg cisapride QID, would be more efficacious.
|
Footnotes |
|---|
Received February 3, 1999.
Revised March 26, 1999.
Accepted April 7, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. J. Meier, G. Kemmeries, J. J. Holst, and M. A. Nauck Erythromycin Antagonizes the Deceleration of Gastric Emptying by Glucagon-Like Peptide 1 and Unmasks Its Insulinotropic Effect in Healthy Subjects Diabetes, July 1, 2005; 54(7): 2212 - 2218. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N J Talley, M Verlinden, D J Geenen, R B Hogan, D Riff, R W McCallum, and R J Mack Effects of a motilin receptor agonist (ABT-229) on upper gastrointestinal symptoms in type 1 diabetes mellitus: a randomised, double blind, placebo controlled trial Gut, September 1, 2001; 49(3): 395 - 401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Rayner, M. Samsom, K. L. Jones, and M. Horowitz Relationships of Upper Gastrointestinal Motor and Sensory Function With Glycemic Control Diabetes Care, February 1, 2001; 24(2): 371 - 381. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
