Aliment Pharmacol Ther 1996 ; 10 : 441–446.

Sucralfate gel versus placebo in patients with non-erosive

gastro-oesophageal reflux disease
B. S I M O N*, G.-P. R A V E L L I† & H. G O F F I N‡
* Department of Internal Medicine, Kreiskrankenhaus Schwetzingen, Germany, † Zyma SA, Nyon, Switzerland and ‡ ECR
Eucam Clinical Research Ltd, Basel, Switzerland
Accepted for publication 4 January 1996

significantly higher for the patients treated with

SUMMARY
sucralfate (71 %) than for the placebo patients (29 %)
Background : Gastro-oesophageal reflux disease, a term (P ! 0.0001, Fisher’s exact test). The overall response
used to refer to chest symptoms that result from reflux of the non-ulcer dyspepsia and gastro-oesophageal
of gastric acid into the oesophagus, occur at least daily reflux disease symptoms was better for sucralfate gel
in 7 % and every 3 days in 33 % of the population. than for placebo with 45 % of patients treated with
Methods : One hundred and forty-one patients with sucralfate gel being considered as having a ‘ good ’ or
moderate to severe gastro-oesophageal reflux ‘ excellent ’ overall response compared with 22 % of the
symptoms occurring at least three times per week (but patients who received placebo (P ! 0.0001, Wilcoxon
no oesophageal erosions or ulcers at endoscopy) were test). Only a few adverse experiences were reported by
treated in this randomized, double-blind, placebo- 10 % of sucralfate patients and 7 % of placebo patients.
controlled study at six trial centres. Treatment was Conclusion : In this trial, we demonstrated a statistically
given for 6 weeks and consisted of daily doses of either significant superiority of sucralfate gel at a dosage of
1 g sucralfate gel b.d. or placebo. The evaluation of 1 g b.d. compared to placebo in the treatment of
efficacy was based on data of 139 patients. patients with gastro-oesophageal reflux disease.
Results : The responder rate at Day 42 was statistically Sucralfate gel was well tolerated.

States has heartburn at least once monthly." About 18

INTRODUCTION
million of these individuals, or 13 % of the adult popu-
Non-ulcer dyspepsia is a disease of still unknown patho- lation, take some type of indigestion aid at least twice a
genesis which afflicts about 30 % of the population and is week.
characterized by a combination of acute, intermittent or Sucralfate is a complex of sucrose sulphate with
chronic abdominal symptoms, such as bloating, feeling of aluminium hydroxide. Sucralfate gel, which is present in
fullness, belching, acid regurgitation, abdominal or epi- the Citogel suspension, is an original and patented
gastric pain, and nausea. physical form of sucralfate. This new formulation is
Gastro-oesophageal reflux disease (GERD) is the term presently registered and sold in Italy under the brand-
used to refer to chest symptoms and}or mucosal damage names Citogel (Zyma SpA), Gastrogel (Bracco SpA) and
that result from reflux of gastric acid into the oesophagus. Sucrate (Lisapharma SpA). In a pharmacodynamic trial,#
GERD symptoms occur at least daily in 7 % and every 3 the stomach coating achieved with sucralfate gel, at half
days in 33 % of the population. A recent Gallup poll the dosage of conventional sucralfate (Ulcogant sus-
revealed that 44 % of the adult population of the United pension), tended to be more extensive owing to a finer
Correspondence to : Prof. Dr B. Simon, Department of Internal Medicine,
particle size combined with a higher viscosity of the gel
Kreiskrankenhaus Schwetzingen, Germany. formulation. The therapeutic activity of sucralfate is

# 1996 Blackwell Science Ltd 441

442 B. S I M ON et al.

Table 1. Patient characteristics (number (%) of patients) pancreatitis, inflammatory bowel disease, diverticulitis,
Sucralfate Placebo
etc., were to be excluded from the study. Patients who
(n ¯ 70) (n ¯ 71) had been treated on a regular basis (more than 3 days)
with either H -blockers, proton-pump inhibitors, sucral-
Sex #
fate, colloidal bismuth, high-dose antacids or cisapride
Male 50 (71.4) 47 (66.2)
Female 20 (28.6) 24 (33.8) in the 2 weeks prior to endoscopy were also excluded.
Age (years) After the patients had been informed about the nature
Mean 45.4 46.1 and purpose of the study, and had given their written
s.d. 12.5 12.4 consent to participate, they were assigned at random to
Range 17–72 20–71 one of the following medication groups : (1) sucralfate
Smoking
No smoking 50 (71) 67 (94)
gel, one sachet (1 g) b.d. ; and (2) placebo, one sachet
Smoking 20 (29) 4 (6) (1 g) b.d. The recommended intake was one sachet of
Alcohol consumption either sucralfate or placebo to be taken 1 h before
No drinking 33 (47) 44 (62) breakfast and at bedtime. Each intake of the trial
Moderate drinking 37 (53) 27 (38) medication was to be followed immediately by a sip of
water or of another fluid.
determined by the protection of the mucosa from diges- Patients were asked to record their non-ulcer dyspepsia
tive juices. Absorption of sucralfate from the gastro- and GERD symptoms and well being (on a four-point
intestinal tract is minimal, and only trace amounts are scale) daily on a diary card, as well as their test
excreted in the urine. The healing effect of sucralfate is medication intake. After 2 and 4 weeks of double-blind
brought about by a local effect and not through systemic treatment the patients returned for a control visit, where
action. their non-ulcer dyspepsia and GERD symptomatology,
The present trial was a multicentre, randomized, including well being, adverse experiences, vital signs and
double-blind, placebo-controlled, parallel-group trial, in compliance to treatment, were recorded by the investi-
which the efficacy of Citogel was evaluated in patients gator. The patients were also issued trial medication
suffering from non-ulcer dyspepsia with non-erosive for the next 2 weeks. After 6 weeks of double-blind
GERD. treatment a full clinical examination was repeated and
the non-ulcer dyspepsia and GERD symptomatology was
recorded. The investigator’s assessment of response was
PATIENTS AND METHODS
recorded at the final visit.
The study involved out-patients of both sexes, between Six trial centres were involved, each with a minimum of
18 and 80 years of age, who had complained for at least seven patients and a maximum of 40 patients. The
1 month of gastroenterological symptoms indicative of primary efficacy parameter was the response rate to
non-ulcer dyspepsia with GERD, namely reflux symptoms treatment, a responder being defined according to the
such as heartburn at a frequency of at least 3 times per investigator’s judgement of the improvement of reflux
week. At baseline, the severity of the daytime and}or symptoms at the final visit compared to baseline. Sec-
night-time reflux symptoms (assessed on a four-point ondary efficacy parameters were the presence and
scale) had to be scored moderate or severe during the severity of the GERD symptoms and the other non-ulcer
week prior to baseline. Other non-ulcer dyspepsia symp- dyspepsia symptoms (bloating}feeling of fullness ;
toms, such as bloating, feeling of fullness, nausea, nausea ; abdominal}epigastric pain) and general well
abdominal or epigastric pain (assessed on a four-point being. The overall response of the non-ulcer dyspepsia
scale), could also be present but were not an entry and GERD symptoms was defined as : excellent (all five
requirement. Endoscopy of the oesophagus had to reveal symptoms had improved to ‘ none ’), good (all five
a normal mucosa or only a reddening (Savary–Miller symptoms had improved to ‘ none ’ or ‘ mild ’), insufficient
Grade 0 or 1), but no erosions or ulcers.$ (no worsening of any symptom, at least one symptom
Patients with concomitant erosive reflux oesophagitis, had improved at the end of treatment, but it was not a
gastric or duodenal ulcers, gastric erosions, gastric outlet ‘ good ’ or ‘ excellent ’ response), no change (all five
obstruction or any other organic gastroenterological symptoms had not changed) or failure (not ‘ excellent ’,
disease, such as gastric malignancy, cholecysto-lithiasis, ‘ good ’ nor ‘ insufficient). Based on the five non-ulcer

# 1996 Blackwell Science Ltd, Aliment Pharmacol Ther 10, 441–446

 S U C R A L F A T E V S. P L A C E B O I N P A T I E N T S W I T H G E R D 443

dyspepsia and GERD symptoms (severity of daytime and Table 3. Response to treatment, overall summary
night-time GERD symptoms, bloating}feeling of fullness, Fisher’s
nausea and abdominal}epigastric pain) the sumscore Sucralfate Placebo exact test
(sum of scores) was also calculated. Population n (%) n (%) P value
The primary efficacy parameter, response rate, was
Intent-to-treat 69 70 ! 0.0001
tested by the 2¬2 table and Fisher’s exact probability in Responder (%) 49 (71) 20 (29)
a confirmatory sense. A Wilcoxon–Mann–Whitney U- Non-responder (%) 20 (29) 50 (71)
test for the overall symptom response was performed for
exploratory purposes. Both an intent-to-treat and a per-
protocol analysis were performed ; in this paper the
results of the intent-to-treat analysis will be presented. to insufficient efficacy. One placebo patient was a protocol
violator, as he was previously treated with ranitidine
(Zantac) 300 mg daily for 12 days and which was
RESULTS
stopped only 5 days prior to trial start. Two patients in
the sucralfate group discontinued the trial prematurely
Patients
owing to other reasons.
A total of 157 patients, 79 patients in the sucralfate One hundred and forty-one patients (sucralfate group,
group and 78 patients in the placebo group, were enrolled 70 patients ; placebo group, 71 patients) could be
in six trial centres. Sixteen patients (sucralfate group, evaluated for safety, and 139 patients (sucralfate group,
nine patients ; placebo group, seven patients) could not 69 patients ; placebo group, 70 patients) were included in
be included in any safety and}or efficacy analysis as no the intent-to-treat efficacy analysis. Two patients, one
follow-up was available after the trial started. sucralfate patient and one placebo patient, retained for
Of the 141 evaluable and treated patients, 116 patients the safety analysis, however, had to be excluded from the
(82 %) completed the trial and 25 patients (18 %) efficacy analysis as a result of the lack of follow-up
withdrew from the trial prematurely. Six patients, three information.
in each group, discontinued prematurely owing to The patient characteristics and anamnestic data are
adverse experiences. Sixteen patients, eight patients in summarized in Table 1.
each group, withdrew from the trial prematurely owing Details on the anamnestic data relating to the GERD
symptoms can be found in Table 2. The average weekly
Table 2. Anamnestic data : GERD symptoms
frequency of the GERD symptoms over the previous 4
weeks was 5.9 times in the sucralfate group and 5.6
Sucralfate Placebo times in the placebo group and therefore well above the
(n ¯ 70) (n ¯ 71) minimum requirement of three or more occurrences per
Time since first week.
onset (months) About half of the patients in both groups had previously
Mean (³s.d.) 14.7 (³28.3) 14.9 (³28.9) been treated for GERD symptoms. The majority of patients
Average frequency
had been treated with H -receptor blockers or antacid
over the previous #
4 weeks preparations. In addition, antiemetics, alone or in com-
Mean (³s.d.) 5.9 (³4.1) 5.6 (³3.9) bination, and unknown self-medication had been taken
No. of days over by a few patients.
the previous week The most frequent endoscopic findings were non-erosive
Mean (³s.d.) 5.9 (³1.6) 5.6 (³1.8) gastritis (sucralfate, 56 % ; placebo, 55 %) and oeso-
Frequency daytime
Mean (³s.d.) 3.3 (³1.8) 3.2 (³1.7)
phagitis (sucralfate, 51 % ; placebo, 55 %), followed by
Frequency night-time non-erosive duodenitis (sucralfate, 40 % ; placebo, 38 %).
Mean (³s.d.) 1.6 (³1.2) 1.5 (³0.9) For the patients with oesophagitis in the sucralfate gel
Sleep disturbed group the Savary–Miller grade was 1 in 81 % of patients
Yes 53 (76 %) 59 (83 %) and 0 in 7 % of patients. In the placebo group the
Previously treated
Savary–Miller grade was 1 in 77 % of patients and 0 in
Yes 32 (46 %) 35 (49 %)
23 % of patients.

# 1996 Blackwell Science Ltd, Aliment Pharmacol Ther 10, 441–446

444 B. S I M ON et al.

GERD symptoms
The number of days with GERD symptoms per week
decreased from 5.9 to 1.1 days for the sucralfate group
and from 5.6 to 3.9 days for the placebo group at the end
of treatment.
Results on the daytime GERD symptoms (i.e. average
frequency and maximum severity) are displayed graphi-
cally in Figures 1 and 2.
An improvement of the maximum severity of the
daytime and night-time GERD symptoms at the end of the
treatment was found in 77 and 67 % of sucralfate patients
and in 48 and 51 % of placebo patients, respectively. As
regards the disturbance of sleep, because of GERD
Figure 1. Average frequency of daytime GERD symptoms symptoms 58 % of sucralfate patients and only 31 % of
placebo patients reported an improvement of this symp-
tom at the end of the trial period.

Non-ulcer dyspepsia symptoms

Seventy-five per cent of sucralfate patients and 39 % of
placebo patients reported an improvement of bloating}
feeling of fullness at the end of the trial period. Nausea
improved in 67 % of sucralfate patients and in 44 % of
placebo patients at the end of the trial period. Eighty
per cent of sucralfate patients, but only 46 % of placebo
patients, reported an improvement of abdominal}
epigastric pain at the end of the trial period.

Overall response of the non-ulcer dyspepsia and GERD

Figure 2. Maximum severity of daytime GERD symptoms.
symptoms
This was better for sucralfate gel than for placebo, with
45 % of patients treated with sucralfate gel considered as
Response to treatment having a ‘ good ’ or ‘ excellent ’ overall response compared
with 22 % of the patients who received placebo (P !
The responder rate at Day 42 was statistically signi- 0.0001, Wilcoxon test).
ficantly higher for sucralfate gel (71 %) than for placebo
(29 %) (P ! 0.0001) (Table 3).
Sum of scores
The changes in the sum of scores from baseline to the Day
Response rate in patients with and without gastritis and}or 42 examination support these results, with the mean
oesophagitis reduction from baseline in the sum of scores (indicating
an improvement) being higher with sucralfate gel (®6.8)
The responder rate in the subgroup of patients with compared to placebo (®2.9) (see Figure 3).
gastritis and}or oesophagitis was higher with sucralfate
(77 %) than with placebo (18 %) (P ! 0.0001, Fisher’s
Well being
exact test). In patients without gastritis and}or oeso-
phagitis there was only a small, non-significant effect in As regards well being, 74 % of the sucralfate patients
favour of sucralfate. reported an improvement of their well being compared

# 1996 Blackwell Science Ltd, Aliment Pharmacol Ther 10, 441–446

 S U C R A L F A T E V S. P L A C E B O I N P A T I E N T S W I T H G E R D 445

parison to a H -receptor blocker( and vs. placebo.)

#
Sucralfate has also been shown to be effective in the
treatment of reflux oesophagitis,* non-ulcer dyspepsia,"!
as well as bile reflux gastritis."" In an open study of 18
patients with endoscopically and histologically verified
oesophagitis, 12 weeks of treatment with sucralfate
resulted in a significant reduction of gastro-oesophageal
reflux and the elimination of mucosal damage."#
Hence we considered it warranted to also investigate
the efficacy of sucralfate gel in the treatment of non-ulcer
dyspepsia, mainly of the reflux type. The advantage of the
new sucralfate formulation is a more intensive coating of
the mucosa owing to a finer particle size, combined with
Figure 3. Sum of non-ulcer dyspepsia and GERD symptom scores.
a higher viscosity. Sucralfate, as a specific anti-ulcer
drug, selectively binds to oesophageal and gastroduo-
denal mucosa. It blocks acid ions by absorption, inhibits
with 36 % of placebo patients. The mean score for pepsin at the site of lesions and binds, particularly at low
patient’s well being is higher for patients treated with pH, bile acids.
sucralfate gel than for placebo patients at each time The results of the present study suggest that sucralfate
point, indicating a better general well being in that group gel is superior to placebo in patients with non-ulcer
of patients. dyspepsia of the reflux type : the responder rate at Day 42
was statistically significantly higher for the patients
treated with sucralfate gel (71 %) than for placebo
Safety results
patients (29 %).
Only a few adverse experiences were reported by seven Three other trials have been performed comparing
sucralfate patients (10 %) and four placebo patients (6 %) sucralfate with placebo for patients with severe reflux
during the double-blind treatment period. Three suc- oesophagitis. In one trial"$ of 66 patients, 72 % of
ralfate patients and three placebo patients discontinued sucralfate patients but only 40 % of placebo patients were
therapy prematurely owing to adverse experiences. These symptom-free after 12 weeks of treatment, and 14 and
were mainly nausea and constipation. There were no 20 %, respectively, showed improvement, a difference
relevant changes from baseline with respect to blood which reached statistical significance (P ! 0.05).
pressure and heart rate in either treatment group. Another trial"% of 138 patients, again with severe reflux
oesophagitis, showed a statistically non-significant trend
in favour of sucralfate in healing oesophageal lesions
DISCUSSION
after 6 and 12 weeks of treatment. A placebo-controlled
Non-ulcer dyspepsia is a very common complaint in multicentre trial"& including a high number of patients
general practice.%, & The identity of symptoms in many with oesophageal lesions in the sucralfate group failed to
functional and organic gastrointestinal disorders makes show any statistical significant treatment difference.
definite diagnosis difficult.' Thus, the diagnosis can only The results of these trials are only of limited value for
be made by exclusion. Heartburn and acid regurgitation comparison with our trial, insofar as they have been
are common non-ulcer dyspepsia symptoms. It is believed performed on patients with severe reflux oesophagitis,
that reflux of gastric juice into the distal part of the even presenting oesophageal ulcerations which were
oesophagus is responsible for these complaints. A rational excluded from our study. The degree of symptom im-
therapeutic approach is, therefore, neutralization of provement in our study resembles closely that seen with
gastric acid with antacids, or its suppression by system- the H -receptor antagonist ranitidine."' In 1995, Pace et
#
ically working H -receptor blockers or proton-pump al."( published a meta-analysis on the effect of placebo on
#
inhibitors. the outcome of medically treated reflux oesophagitis,
In previous studies, sucralfate has been recognized as including 22 double-blind studies performed between
being efficacious in treating duodenal ulcers in com- 1976 and 1990 with various active medications, such as

# 1996 Blackwell Science Ltd, Aliment Pharmacol Ther 10, 441–446

446 B. S I M ON et al.

H -receptor antagonists, proton-pump inhibitors and 5 Mead GM, Morris A, Webster GK, Langman MJS. Uses of
# barium meal examination in dyspeptic patients under 50. BMJ
sucralfate. The results showed the active drugs to be
1977 ; 1 : 1460–1.
significantly more effective than placebo, with a pooled
6 Lindberg G. On the similarity of symptoms in peptic ulcer
mean healing rate of 47 vs. 27 % after 4–8 weeks of disease (PUD), non-ulcer dyspepsia (NUD) and the irritable
therapy. bowel syndroms (IBS). Dig Dis Sci 1986 ; 31 (Suppl.) : 5.
We, therefore, conclude that both acid inhibition and (Abstract.)
topical cytoprotection by sucralfate coating the mucosa 7 Glise H, Carling I, Hallerba$ ck B, et al. Treatment of peptic
are effective therapeutic manoeuvres alleviating reflux ulcers,Acid reduction or cytoprotection ? Scand J Gastroenterol
1987 ; 22 (Suppl. 140) : 39–47.
symptoms in non-ulcer dyspepsia patients. Summarizing, 8 Bell NJV, Bailey RJ, Cohen L, et al. Treatment of acute duodenal
sucralfate in its new gel form is a safe, locally active agent and pyloric channel ulcers with sucralfate suspension
with minimal side-effects. It can be used for the treatment (2 g}10 mL, twice daily) : efficacy in smoking patients. Eur J
of patients with non-ulcer dyspepsia, especially of the Gastroenterol Hepatol 1993 ; 5 : 287–91.
reflux type. 9 Laitinen S, Stahlberg M, Kairaluoma MI, et al. Sucralfate and
alginate}antacid in reflux esophagitis. Scand J Gastroenterol
1985 ; 20 : 229–32.
A C K N O W L E D G E M E N TS 10 Matti I, Kairaluoma MD, et al. Sucralfate versus placebo in
treatment of non-ulcer dyspepsia. Am J Med 1987 ; 83 (Suppl.
Thanks go to the following investigators : Dr W. Grebe 3B) : 51–5.
(Frankenberg, Germany), Dr J. Hagel (Schwabach, 11 Stahlberg M, Mokka R, Laitinen S, et al. Bile acid binding
Germany), Dr P. Krupp (Bad Zwischenahn, Germany), capacity of sucralfate, cholestyramine and antacids. Scand J
Prof. Dr B. Simon (Schwetzingen, Germany), Dr G. Topf Gastroenterol 1983 ; 18 (Suppl. 86) : 86. (Abstract.)
(Ansbach, Germany) and Dr R. Vogt (Mannheim, 12 Elsborg L, Beck B, Stubgaard M. Effect of sucralfate on
gastroesophageal reflux in esophagitis. Hepatogastro-
Germany) ; and to the following statistician : Dr V. W. enterology 1985 ; 32 : 181–84.
Rahlfs (A. Schwenck, Munich, Germany). 13 Weiss W, Brunner H, Bu$ ttner GR, et al. Therapie der
Refluxo$ sophagitis mit Sucralfat. Dtsch Med Wochenschr 1983 ;
REFERENCES 108 : 1706–11.
14 Carling L, Cronstedt J, Engqvist A, et al. Sucralfate versus
1 Gallup. Gallup Survey on Heartburn Across America. placebo in reflux esophagitis. A double-blind multicenter study.
Princeton, NJ : The Gallup Organization, 24 March 1988. Scand J Gastroenterol 1988 ; 23 : 1117–24.
2 Hardy JG, Hooper G, Ravelli GP, Steed KP, Wilding IR. A 15 Williams RM, Orlando RC, Bozymski EM, et al. Multicenter trial
comparison of the gastric retention of a sucralfate gel and a of sucralfate suspension for the treatment of reflux esophagitis.
sucralfate suspension. Eur J Pharmacol Biopharmacol 1993 ; Am J Med 1987 ; 83 (Suppl. 3B) : 61.
39 (2) : 70–4. 16 Mu$ ller P, Hotz J, Franz E, Simon B. Ranitidine in the treatment
3 Savary M, Miller G. Der Oesophagus. Lehrbuch und Endo- of non-ulcer dyspepsia. Arzneimittelforschung}Drug Res
skopischer Atlas. Solothurn : Verlag Gassmann AG, 1977 : 1994 ; 10 : 1130–2.
135. 17 Pace F, Maconi G, Molteni M, Minguzzi M, Bianchi Porro G.
4 Nyren O, Adami HO, Bates S, et al. Absence of therapeutic Meta-analysis of the effect of placebo on the outcome of
benefit from antacids or cimetidine in non-ulcer dyspepsia. N medically treated reflux esophagitis. Scand J Gastroenterol
Engl J Med 1986 ; 314 : 339–43. 1995 ; 30 : 101–5.

# 1996 Blackwell Science Ltd, Aliment Pharmacol Ther 10, 441–446