Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992), pp. 297-302
Randomized, Double-Blind, Placebo-Controlled Crossover Trial of Pirenzepine in Patients with Gastroesophageal Reflux TAKAO L. SATO, MD, MPH, WALLACE C. WU, MD, and DONALD O. CASTELL, MD
A muscarinic receptor subtype I (M1) antagonist, pirenzepine, recently has been shown to be relatively free of the usual anticholinergic side effects on esophageal smooth muscle and thus has been implicated for the treatment of gastroesophageal reflux disease (GERD). However, the effect o f pirenzepine on GERD remains to be defined. Thirteen patients who demonstrated GERD in a baseline 24-hr ambulatory intraesophageal p H monitoring study were randomized in a double-blind crossover fashion to receive pirenzepine and placebo. An ambulatory 24-hr intraesophageal p H monitor was used to assess reduction in reflux (esophageal p H < 4.0) with respect to position (upright vs supine), to total number o f reflux episodes, and to episodes >5 rain. A significant effect for pirenzepine was seen for episodes >5 min (t = 2.61, P = 0.023) and a trend towards significance was seen for total (upright and supine positions combined) percent time o f reflux (t = 2.13, P = 0.055). Although not statistically significant, pirenzepine consistently showed greater reduction in all parameters o f reflux tested. A greater reduction in percent time o f reflux in supine vs upright positions (pirenzepine: 58.9% vs 21.4%; placebo: 43.6% vs 7.3%) may be clinically important in prevention of esophageal injury due to reflux in the recumbent position. Pirenzepine may provide a unique alternative for some GERD patients who may be refractory to other therapies o f GERD. KEY WORDS: pirenzepine; gastroesophageal reflux disease.
Anticholinergic drugs are not generally recommended in the treatment of patients with gastroesophageal reflux. A major problem with the use of traditional anticholinergic agents is their diffuse Manuscript received March 7, 1990; revised manuscript received June 20, 1991; accepted July 8, 1991. From the Gastroenterology Section, Department of Medicine, The Bowman Gray School of Medicine, Wake Forest University Medical Center, 300 South Hawthorne Road, Winston-Salem, North Carolina 27103. This study was supported by a grant from Boehringer lngelheim Pharmaceutical, Inc., Ridgefield, Connecticut 06877. Dr. Sato's present address: Division of Gastroenterology, Hepatology and Nutrition, Department of Medicine, University of Florida College of Medicine, Gainesville, Florida 32610.
effect on all three muscarinic receptors (MI, Mz, and M3) as described by Goyal (1). The M1 receptor appears to participate in a stimulatory pathway to the parietal cells, so that its inhibition results in decreased acid secretion. Stimulation of the M 2 and M3 receptors, on the other hands, results in more typical muscarinic effects, ie, contraction of smooth Dr. Castell's present address: Division of Gastroenterology and Hepatology, Department of Medicine, Jefferson Medical College, 1025 Walnut, Philadelphia, Pennsylvania 19107. Address for reprint requests: Dr. T.L. Sato, Division of Gastroenterology, Hepatology and Nutrition, Department of Medicine, University of Florida College of Medicine, Box J-214 JHMHSC, Gainesville, Florida 32610.
Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
0163-2116/92/0200-0297506.50/09 1992PlenumPublishingCorporation
297
SATO ET AL muscle. T h e r e f o r e , traditional a n t i c h o l i n e r g i c agents not only suppress gastric acid secretion but also decrease esophageal contraction pressure (including the lower esophageal sphincter), thereby potentially worsening gastroesophageal reflux. Pirenzepine is a pyridobenzodiazepine compound with selective anticholinergic inhibition at the MI muscarinic receptor site. In a placebo-controlled, double-blind study of reflux esophagitis, pirenzepine showed a significant improvement in reflux symptoms and esophagitis as assessed by endoscopy (2). In addition, this drug has been shown to be relatively free o f typical anticholinergic side effects at doses that suppress gastric acid secretion (3). In the previous studies from our laboratory we have shown that pirenzepine (a) does not decrease esophageal contraction pressure to the degree occurring with the anticholinergic propantheline (4); (b) produces less inhibition o f cholinergic stimulation from bethanechol on the lower esophageal sphincter and peristaltic amplitude than that produced by propantheline (5); (c) shows these effects at doses that suppress gastric secretion at night when the patient is recumbent and gastroesophageal reflux may be most severe (6); and (d) dose not reduce salivary secretion or salivary capacity for acid neutralization (7), which confirms the previous finding by Jaup et al (8). The present study was undertaken to assess the effect o f pirenzepine in patients with documented gastroesophageal reflux b y employing an intraesophageal p H monitor for 24 hr. Specifically, this study was designed to evaluate the potential effect o f pirenzepine to reduce gastroesophageal reflux (esophageal p H < 4.0) in both upright and supine positions. Parameters tested included total percentage time of p H < 4.0, total number of reflux episodes, and reflux episodes > 5 min in a 24-hr period.
MATERIALS AND METHODS Patients. Thirteen patients with gastroesophageal reflux as defined by a baseline 24-hr intraesophageal acid exposure greater than two standard deviations above the mean normal value for our laboratory (>4.4% of monitoring time) completed the study (9). Patients with other esophageal or gastric diseases or diabetes mellitus were excluded. All patients gave informed consent and were off antacid medication including H2 blockers for a minimum of 48 hr prior to the onset of the study as well as during the study period. The study was initiated after the approval of the Institutional Review Board of the Bowman Gray School of Medicine on September 19, 1987.
298
Study Drugs. Pirenzepine (50 mg tablets) and placebo tablets matching in size, weight, and appearance were manufactured and supplied by Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut. Study Design. All patients who demonstrated gastroesophageal reflux during the baseline assessment were randomized to a single-center, double-blind, placebocontrolled, crossover study. The initial intent of the study was to complete 20 patients. However, before completion, a decision was made to terminate this study by the manufacturer of the study medication. Therefore, at the time of termination of the study, 13 patients had completed the protocol. The dose of pirenzepine (50 mg three times a day) or placebo was administered in the same manner. Two one-day drug study periods after the baseline were separated by no less than three and no more than 10 days. Patients received pirenzepine one day and placebo on the other day according to the randomization code. The medication was administered at 8 AM, 3 PM, and 11 I'M. The intraesophageal pH monitor was placed prior to the first dose of medication and continued for a 24-hr period. Patients were allowed to leave the clinic and PUrsue their normal daily activities. They were requested to eat meals as usual except for the following dietary restrictions: fruits, fruit juices, foods made with fruit sherbet, jelly, chocolate milk, and butter milk; tomatoes or foods prepared with tomatoes; sauerkraut; cream cheese; chocolate; carbonated beverages. Coffee or tea could be consumed with meals, but not between meals. The only beverage allowed between meals was water. Adverse effects of test medications were questions when patients returned for the removal of their pH monitor. No patients reported adverse effects from pirenzepine or placebo. Ambulatory Intraesophageal pH Monitoring. An ambulatory intraesophageal pH monitoring device manufactured by Sandhill Corporation was employed for all studies. The procedural details are published elsewhere from our laboratory (9). This lower esophageal sphincter (LES) was located with manometry. A small pH probe (Microelectronics, Inc.) was passed through the nose and placed 5 cm proximal to the manometrically located LES. The pH electrode was calibrated before and after the completion of study with standard pH buffer solutions at pH 4.0 and 7.0. Patients were instructed to flip a switch on the recording box to indicate supine or upright position and to press a button to indicate their symptoms or activities, ie, heartburn, regurgitation, meal, medication, and other symptoms that were recorded in their daily log book. The recorded data were analyzed by the computer program to measure to reflux events defined as pH < 4.0 with respect to upright and supine positions. Statistical Analysis. Descriptive statistics were computed, and the data were examined to ensure that variances were equal between treatment groups (baseline, placebo, pirenzepine). Repeated measures of analysis of variance was performed to assess the treatment effect. A loglo transformation of percentage time of acid exposure in the supine position was made in order to stabilize its variance. P values were adjusted for correlated measurements using the Huynh-Feldt technique to deflate the degrees of freedom of the F statistic (10) and were Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
PIRENZEPINE IN PATIENTS WITH GASTROESOPHAGEAL REFLUX considered significant if P < 0.05. Paired t tests were employed to assess pairwise differences between baseline and placebo or pirenzepine and between placebo and pirenzepine with respect to: (a) percent of time of reflux (total, upright, and supine), 0a) total episodes of reflux, and (c) number of episodes >5 min if the Huynb-Feldt test of the overall effect was at least marginally significant (P < 0.05).
22 ine 3o :epine
20 18 16 14 "E .o
12
{2_ 10
RESULTS
8
The demographic characteristics of the 13 patients are shown in Table 1. There were six males and seven females with a overall mean age of 46.6 years (range 29-65 years) and were predominantly white (76.9%). Percent time of acid exposure (pH < 4.0) for each patient during baseline study and on treatment with placebo and pirenzepine is shown in Table 2. Repeated measures of analysis of variance for treatment effect showed a trend for significance for total percent time pH < 4.0 in 24 hr (F = 3.79, P = 0.06I) and for episodes lasting >5 min in 24 hr (F = 2.94, P = 0.088). Paired t tests were carried out to further assess the improvement from baseline with respect to the treatment and possible treatment differences. For mean percent total time of reflux, pirenzepine treatment showed a trend towards a significant (t = 2.13, P = 0.055) 36.5% reduction from baseline, while a 25.5% reduction by placebo treatment was not significant (t = 1.81, P = 0.10). The difference in improvements from baseline with respect to treatment was not significant. These results are shown in Table 2 and Figure 1. For mean percent time of upright reflux, a threefold improvement in percent change from baseline was observed for pirenzepine (21.4%) over placebo treatment (7.3%). However, this difference was not significant. These results are shown in Table 2 and Figure 1. For mean percent time of supine reflux, a greater percent improvement from baseline was seen in supine than upright position, ie, 58.9% vs 21.4% for
6 4 2 0 Total
Upright
Supine
Fig 1. Mean percent time pH < 4.0 vs positions. Vertical lines indicate SEM. *P = 0.055 compared to baseline.
pirenzepine and 43.6% vs 7.3% for placebo, respectively. However, neither of these improvements from baseline nor the difference in improvements with respect to treatments were significant. These results are shown in Table 2 and Figure 1. Mean total reflux episodes -+ SEM in 24 hr was 96.5 --- 10.3 (range 55-184) for baseline, 74.6 --- 13.8 (range 21-168) for pirenzepine, and 89.0 -+ 9.9 (range 39-139) for placebo. Despite a threefold reduction by pirenzepine (22.4%) over placebo (7.7%) from baseline, this was not significant. These results are depicted in Figure 2. Mean episodes of reflux lasting >5 min ___ SEM was 7.1 ---+ 1.8 (range 2-24) for baseline, 4.1 -+ 1.6 (range 0-21) for pirenzepine, and 5.1 -+ 2.1 (range 0-28) for placebo. Pirenzepine treatment showed a significant (t = 2.61, P = 0.023) 42.9% reduction from baseline, while a 29.0% reduction by placebo was not significant (t = 1.30, P = 0.21). However, the difference in improvements with respect to treatments was not significant. These are shown in Figure 3.
120
TABLE | . DEMOGRAPHIC CHARACTERISTICSOF 13 PATIENTS Female
80,
6
7
60.
50 47.3 • 5.7 30-65
47 46.1 • 3.5 29-63
Male Sex Age Median Mean • SEM Range Race Black
100
.o ~
40. 20. O.
ill
Baseline
Placebo Pirenzepine
1
2
5
5
Fig 2. Mean number of total reflux episodes in 24 hr vs baseline
Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
299
White
and treatments. Vertical lines indicate SEM.
SATO ET AL 10
6.
~4. 2. 0 Baseline
I
Placebo
Pirenzepine
Fig 3. Mean number of reflux episodes >5 min in 24 hr vs baseline and treatments. Vertical lines indicate SEM. *P = 0.023 compared to baseline.
DISCUSSION Anticholinergics have proven their efficacy in reducing gastric acid secretion. A major drawback for the use of traditional anticholinergics is their diffuse effect on gastrointestinal tract smooth muscle. At acid-suppressing doses for peptic ulcer diseases, this anticholinergic effect is manifested by a decreased force of contraction, a disturbance in peristalsis, and a decreased LES pressure, which may worsen gastroesophageal reflux (1 I). Pirenzepine, an M1 receptor selective antimuscarinic agent with anticholinergic properties mostly restricted to the gastric mucosa (12), has been found to inhibit basal and pentagastrin-simulated acid secretion, and its ulcer healing rate (70-87%) is comparable to that of cimetidine (13-15). The rate of relapse six months after the discontinuation of therapy has been shown to be less for pirenzepine (28.5%) than that for cimetidine (47%) (16).
Because of its selectivity to gastric mucosa, pirenzepine has been used widely as an antiulcer agent in Europe, but it has not yet been marketed in the United States. Unlike classic anticholinergic agents such as atropine (17) and propantheline (18), pirenzepine has been associated with fewer side effects. However, Talley et al (19) reported a higher incidence of gastroesophageal reflux symptoms, ie, heartburn and epigastric pain in nonulcer dyspepsia patients who were treated with pirenzepine at 50 mg twice a day when compared to patients treated with cimetidine at 200 mg four times a day or with placebo. They speculated that this was due to the anticholinergic effect of pirenzepine on the LES. In previous studies from our laboratory we showed that pirenzepine did not decrease esophageal contraction pressure to the degree occurring with propantheline (4) and that pirenzepine produced less inhibition of cholinergic stimulation from bethanechol on the LES than propantheline (5). In addition, pirenzepine did not reduce salivary secretion or salivary capacity for acid neutralization (7), which may play a role in esophageal acid clearance (20). This finding was consistent with that reported by Jaup et al (8). In our present study, no patients reported adverse effects to either placebo or pirenzepine at the 50-mg three-times-a-day dose. Since this study was a single-day study for both pirenzepine and placebo, possible adverse effects of chronic doses of pirenzepine cannot be excluded. Few intraesophageal pH monitoring studies have been conducted to assess the effect of anticholinergics on gastroesophageal reflux in patients with
TABLE 2. PERCENT TIME OF ACID EXPOSURE (pH 5 min in 24 hr (F = 2.94, P = 0.088). When these were further assessed by paired t tests with a log10 transformation of percentage time of acid exposure in the supine position in order to stabilize its variance and adjusting P values for correlated measurements using the Huynh-Feldt technique to deflate the degrees of freedom of the F statistic (10), pirenzepine showed a significant reduction in episodes >5 min (t = 2.61, P = 0.023) and a trend towards significant treatment effect over placebo for a total percent reflux time during the 24-hr period (t = 2.13, P = 0.055). When the positions were further divided into upright and supine cateDigestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
gories, pirenzepine treatment showed consistently better, although not significant, improvement from baseline over placebo treatment. Pirenzepine showed a threefold improvement in percent change from baseline reflux time (21.4%) over placebo (7.3%) in the upright position. A greater magnitude of improvement from baseline in percent time of reflux was seen in the supine as well for pirenzepine (58.9%) vs placebo (43.6%). However, neither of these treatment effects were significant. A greater improvement seen in the supine position may be clinically important in prevention of esophageal injury secondary to reflux in the recumbent position. Similar treatment effects might be anticipated with respect to the total number of reflux episodes lasting >5 min, which is considered to reflect the clearance activity of the esophagus after an episode of acid reflux where significant improvement from baseline was seen for pirenzepine. Treatment of gastroesophageal reflux disease has changed greatly in the past decade due to the development of more effective acid-suppressing drugs and prokinetic agents. Unlike classic anticholinergic agents, which are not useful because of their diffuse effects over the gastrointestinal tract, including esophageal dysmotility and decreased LES pressure, pirenzepine may provide a unique alternative for some patients not satisfactorily controlled with other therapies. We recommend further large-scale studies with a longer duration of treatment to confirm the effect of pirenzepine in gastroesophageal reflux disease. ACKNOWLEDGMENTS The authors wish to thank Christine B. Dalton, Lisa Nowak, Gregg Stark, Jane W. Sinclair, and Faye Williams of the Gastroenterology Section of the Bowman Gray School of Medicine for their technical assistance. We are grateful to Timothy C. Craven of the Department of Public Health Sciences of the Bowman Gray School of Medicine for the initial statistical analysis and reanalysis of the data after the acceptance of this paper by the journal. REFERENCES 1. Goyal RK: Muscarinic receptor subtypes, physiology and clinical implications. N Engl J Med 321(15):1022-1028, 1989 2. Niemela S, Jaaskelainen T, Lehtola J, et al: Pirenzepine in the treatment of reflux oesophagitis, a placebo-controlled, double-blind study. Scand J Gastroenterol 21:1193-1199, 1986 3. Jaup BH, Abrahamsson H, Virtanen R, Iisalo E: Effective of pirenzepine compared with atropine and L-hyoscyamine on
301
SATO ET A L
4.
5.
6.
7.
8.
9.
10.
I 1.
12.
13.
14.
esophageal peristaltic activity in humans. Scand J Gastroenterol 17:233-239, 1982 Blackwell JN, Dalton CB, Castell DO: Oral pirenzepine does not affect esophageal pressure in man. Dig Dis Sci 31:230235, 1986 Hassan M, Dalton CB, Castell JA, Castell DO: Pirenzepine and propanthline effects on esophageal pressure responses to bethanechol. Gastroenterology 81(5):334-338, 1986 Castell DO: Pirenzepine: A unique 'anticholinergic' on the esophagus. In Pirenzepine: New aspects in Research and Therapy. Sept 18, 1984. Proceedings. A Bettarello (ed). Amsterdam, Excerpta Medica, 1985, pp 142-149 DeVault KR, Castell DO: Effects of antireflux therapies on salivary function in normal humans. Dig Dis Sei 32(6):603608, 1987 Jaup BH, Stockbrugger RW, Dotevall G: Comparison between the effect of pirenzepine and/-hyoscyamine in man. Scand J Gastroenterol 17(suppl 72): 119-122, 1982 Ward BW, Wu WC, Richter JE, Lui KW, Castell DO: Ambulatory 24-hr esophageal pH monitoring. J Clin Gastroenterol 8 (suppl 1):59-67, 1986 Huynh H, Feldt LS: Estimation of the box correlation for degrees of freedom from sample data in the randomized block and split plot designs. J Educ Stat 1:69-82, 1976 Bettarello A, Tuttle SG, Grossman MI: Effect of automatic drugs on gastroesophageal reflux. Gastroenterology 39:340346, 1960 Hammer R, Berrie CP, Birdsall NJM, Burgen ASV, Hulme EC: Pirenzepine distinguishes between subclasses of muscarinic receptors. Nature 283:90-92, 1980 Carmine AA, Brogden RN: Pirenzepine. A review of its pharmacodynamic and pharmacokinetic properties and therapeutic efficacy in peptic ulcer disease and other allied diseases. Drugs 30(2):85-126, 1985 Henry DA, Hawkey C, Somerville K, Burnham WR, Bell GD, Langman MJS: Pirenzapine and cimetidine in duodenal ulcer: A comparative study. In 7th World Congress of
302
15.
16.
17.
18.
19.
20.
21.
22.
Gastroenterology Proceedings. G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 214-216 Brown SG, Gomes M, Pounder RE, Rampton DS, Salmon PR, Sarner M, Vicary FR: Comparison of pirenzepine and cimetidine in gastric ulcer. In 7th World Congress of Gastroenterology Proceedings. G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 219-222 Vezzadini P, Sternini P, Valentini PL, Botti C, Lugli C, Labo G: Comparison of relapse rates in duodenal ulcer patients treated with pirenzepine or cimetidine. In 7th World Congress of Gastroenterology proceedings. G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 238-245 Texter E C, Patel GK, Navab F, et al: Effect of oral pirenzepine, a muscarinic (M1) receptor antagonist, and oral atropine on lower esophageal sphincter pressure, esophageal manometry and scintiscans, and gut peptides. Clin Res 31:766A, 1983 Toussaint J, Van der Veken J, Cremer M: A comparative double blind trial of pirenzepine and propantheline in duodenal ulcer In 7th World Congress of Gastroenterology Proceedings, G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 223-231 Talley NJ, McNeil D, Hayden A, Piper DW: Randomized, double-blind, placebo controlled crossover trial of cimetidine and pirenzepine in nonulcer dyspepsia. Gastroenterology 91:149-156, 1986 Helm JF: Effect of esophageal emptying and saliva on clearance of acid from the esophagus. N Engl J Med 310(5):284-288, 1984 Stockbrugger RW, Armbrecht U, Reul W: The effect of the Ml-selective telenzepine on esophageal acid exposure in healthy volunteers. Z Gastroenterol 26(6):297-302, 1988 Geller LI, Griaznova MV, Petrenko VF, Bessonova GA: The degree of gastroesophageal reflux during treatment of the exacerbation of duodenal peptic ulcer by various drugs. Ter Arkh 60(2):42-43, 1980
Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
Randomized, Double-Blind, Placebo-Controlled Crossover Trial of Pirenzepine in Patients with Gastroesophageal Reflux TAKAO L. SATO, MD, MPH, WALLACE C. WU, MD, and DONALD O. CASTELL, MD
A muscarinic receptor subtype I (M1) antagonist, pirenzepine, recently has been shown to be relatively free of the usual anticholinergic side effects on esophageal smooth muscle and thus has been implicated for the treatment of gastroesophageal reflux disease (GERD). However, the effect o f pirenzepine on GERD remains to be defined. Thirteen patients who demonstrated GERD in a baseline 24-hr ambulatory intraesophageal p H monitoring study were randomized in a double-blind crossover fashion to receive pirenzepine and placebo. An ambulatory 24-hr intraesophageal p H monitor was used to assess reduction in reflux (esophageal p H < 4.0) with respect to position (upright vs supine), to total number o f reflux episodes, and to episodes >5 rain. A significant effect for pirenzepine was seen for episodes >5 min (t = 2.61, P = 0.023) and a trend towards significance was seen for total (upright and supine positions combined) percent time o f reflux (t = 2.13, P = 0.055). Although not statistically significant, pirenzepine consistently showed greater reduction in all parameters o f reflux tested. A greater reduction in percent time o f reflux in supine vs upright positions (pirenzepine: 58.9% vs 21.4%; placebo: 43.6% vs 7.3%) may be clinically important in prevention of esophageal injury due to reflux in the recumbent position. Pirenzepine may provide a unique alternative for some GERD patients who may be refractory to other therapies o f GERD. KEY WORDS: pirenzepine; gastroesophageal reflux disease.
Anticholinergic drugs are not generally recommended in the treatment of patients with gastroesophageal reflux. A major problem with the use of traditional anticholinergic agents is their diffuse Manuscript received March 7, 1990; revised manuscript received June 20, 1991; accepted July 8, 1991. From the Gastroenterology Section, Department of Medicine, The Bowman Gray School of Medicine, Wake Forest University Medical Center, 300 South Hawthorne Road, Winston-Salem, North Carolina 27103. This study was supported by a grant from Boehringer lngelheim Pharmaceutical, Inc., Ridgefield, Connecticut 06877. Dr. Sato's present address: Division of Gastroenterology, Hepatology and Nutrition, Department of Medicine, University of Florida College of Medicine, Gainesville, Florida 32610.
effect on all three muscarinic receptors (MI, Mz, and M3) as described by Goyal (1). The M1 receptor appears to participate in a stimulatory pathway to the parietal cells, so that its inhibition results in decreased acid secretion. Stimulation of the M 2 and M3 receptors, on the other hands, results in more typical muscarinic effects, ie, contraction of smooth Dr. Castell's present address: Division of Gastroenterology and Hepatology, Department of Medicine, Jefferson Medical College, 1025 Walnut, Philadelphia, Pennsylvania 19107. Address for reprint requests: Dr. T.L. Sato, Division of Gastroenterology, Hepatology and Nutrition, Department of Medicine, University of Florida College of Medicine, Box J-214 JHMHSC, Gainesville, Florida 32610.
Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
0163-2116/92/0200-0297506.50/09 1992PlenumPublishingCorporation
297
SATO ET AL muscle. T h e r e f o r e , traditional a n t i c h o l i n e r g i c agents not only suppress gastric acid secretion but also decrease esophageal contraction pressure (including the lower esophageal sphincter), thereby potentially worsening gastroesophageal reflux. Pirenzepine is a pyridobenzodiazepine compound with selective anticholinergic inhibition at the MI muscarinic receptor site. In a placebo-controlled, double-blind study of reflux esophagitis, pirenzepine showed a significant improvement in reflux symptoms and esophagitis as assessed by endoscopy (2). In addition, this drug has been shown to be relatively free o f typical anticholinergic side effects at doses that suppress gastric acid secretion (3). In the previous studies from our laboratory we have shown that pirenzepine (a) does not decrease esophageal contraction pressure to the degree occurring with the anticholinergic propantheline (4); (b) produces less inhibition o f cholinergic stimulation from bethanechol on the lower esophageal sphincter and peristaltic amplitude than that produced by propantheline (5); (c) shows these effects at doses that suppress gastric secretion at night when the patient is recumbent and gastroesophageal reflux may be most severe (6); and (d) dose not reduce salivary secretion or salivary capacity for acid neutralization (7), which confirms the previous finding by Jaup et al (8). The present study was undertaken to assess the effect o f pirenzepine in patients with documented gastroesophageal reflux b y employing an intraesophageal p H monitor for 24 hr. Specifically, this study was designed to evaluate the potential effect o f pirenzepine to reduce gastroesophageal reflux (esophageal p H < 4.0) in both upright and supine positions. Parameters tested included total percentage time of p H < 4.0, total number of reflux episodes, and reflux episodes > 5 min in a 24-hr period.
MATERIALS AND METHODS Patients. Thirteen patients with gastroesophageal reflux as defined by a baseline 24-hr intraesophageal acid exposure greater than two standard deviations above the mean normal value for our laboratory (>4.4% of monitoring time) completed the study (9). Patients with other esophageal or gastric diseases or diabetes mellitus were excluded. All patients gave informed consent and were off antacid medication including H2 blockers for a minimum of 48 hr prior to the onset of the study as well as during the study period. The study was initiated after the approval of the Institutional Review Board of the Bowman Gray School of Medicine on September 19, 1987.
298
Study Drugs. Pirenzepine (50 mg tablets) and placebo tablets matching in size, weight, and appearance were manufactured and supplied by Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut. Study Design. All patients who demonstrated gastroesophageal reflux during the baseline assessment were randomized to a single-center, double-blind, placebocontrolled, crossover study. The initial intent of the study was to complete 20 patients. However, before completion, a decision was made to terminate this study by the manufacturer of the study medication. Therefore, at the time of termination of the study, 13 patients had completed the protocol. The dose of pirenzepine (50 mg three times a day) or placebo was administered in the same manner. Two one-day drug study periods after the baseline were separated by no less than three and no more than 10 days. Patients received pirenzepine one day and placebo on the other day according to the randomization code. The medication was administered at 8 AM, 3 PM, and 11 I'M. The intraesophageal pH monitor was placed prior to the first dose of medication and continued for a 24-hr period. Patients were allowed to leave the clinic and PUrsue their normal daily activities. They were requested to eat meals as usual except for the following dietary restrictions: fruits, fruit juices, foods made with fruit sherbet, jelly, chocolate milk, and butter milk; tomatoes or foods prepared with tomatoes; sauerkraut; cream cheese; chocolate; carbonated beverages. Coffee or tea could be consumed with meals, but not between meals. The only beverage allowed between meals was water. Adverse effects of test medications were questions when patients returned for the removal of their pH monitor. No patients reported adverse effects from pirenzepine or placebo. Ambulatory Intraesophageal pH Monitoring. An ambulatory intraesophageal pH monitoring device manufactured by Sandhill Corporation was employed for all studies. The procedural details are published elsewhere from our laboratory (9). This lower esophageal sphincter (LES) was located with manometry. A small pH probe (Microelectronics, Inc.) was passed through the nose and placed 5 cm proximal to the manometrically located LES. The pH electrode was calibrated before and after the completion of study with standard pH buffer solutions at pH 4.0 and 7.0. Patients were instructed to flip a switch on the recording box to indicate supine or upright position and to press a button to indicate their symptoms or activities, ie, heartburn, regurgitation, meal, medication, and other symptoms that were recorded in their daily log book. The recorded data were analyzed by the computer program to measure to reflux events defined as pH < 4.0 with respect to upright and supine positions. Statistical Analysis. Descriptive statistics were computed, and the data were examined to ensure that variances were equal between treatment groups (baseline, placebo, pirenzepine). Repeated measures of analysis of variance was performed to assess the treatment effect. A loglo transformation of percentage time of acid exposure in the supine position was made in order to stabilize its variance. P values were adjusted for correlated measurements using the Huynh-Feldt technique to deflate the degrees of freedom of the F statistic (10) and were Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
PIRENZEPINE IN PATIENTS WITH GASTROESOPHAGEAL REFLUX considered significant if P < 0.05. Paired t tests were employed to assess pairwise differences between baseline and placebo or pirenzepine and between placebo and pirenzepine with respect to: (a) percent of time of reflux (total, upright, and supine), 0a) total episodes of reflux, and (c) number of episodes >5 min if the Huynb-Feldt test of the overall effect was at least marginally significant (P < 0.05).
22 ine 3o :epine
20 18 16 14 "E .o
12
{2_ 10
RESULTS
8
The demographic characteristics of the 13 patients are shown in Table 1. There were six males and seven females with a overall mean age of 46.6 years (range 29-65 years) and were predominantly white (76.9%). Percent time of acid exposure (pH < 4.0) for each patient during baseline study and on treatment with placebo and pirenzepine is shown in Table 2. Repeated measures of analysis of variance for treatment effect showed a trend for significance for total percent time pH < 4.0 in 24 hr (F = 3.79, P = 0.06I) and for episodes lasting >5 min in 24 hr (F = 2.94, P = 0.088). Paired t tests were carried out to further assess the improvement from baseline with respect to the treatment and possible treatment differences. For mean percent total time of reflux, pirenzepine treatment showed a trend towards a significant (t = 2.13, P = 0.055) 36.5% reduction from baseline, while a 25.5% reduction by placebo treatment was not significant (t = 1.81, P = 0.10). The difference in improvements from baseline with respect to treatment was not significant. These results are shown in Table 2 and Figure 1. For mean percent time of upright reflux, a threefold improvement in percent change from baseline was observed for pirenzepine (21.4%) over placebo treatment (7.3%). However, this difference was not significant. These results are shown in Table 2 and Figure 1. For mean percent time of supine reflux, a greater percent improvement from baseline was seen in supine than upright position, ie, 58.9% vs 21.4% for
6 4 2 0 Total
Upright
Supine
Fig 1. Mean percent time pH < 4.0 vs positions. Vertical lines indicate SEM. *P = 0.055 compared to baseline.
pirenzepine and 43.6% vs 7.3% for placebo, respectively. However, neither of these improvements from baseline nor the difference in improvements with respect to treatments were significant. These results are shown in Table 2 and Figure 1. Mean total reflux episodes -+ SEM in 24 hr was 96.5 --- 10.3 (range 55-184) for baseline, 74.6 --- 13.8 (range 21-168) for pirenzepine, and 89.0 -+ 9.9 (range 39-139) for placebo. Despite a threefold reduction by pirenzepine (22.4%) over placebo (7.7%) from baseline, this was not significant. These results are depicted in Figure 2. Mean episodes of reflux lasting >5 min ___ SEM was 7.1 ---+ 1.8 (range 2-24) for baseline, 4.1 -+ 1.6 (range 0-21) for pirenzepine, and 5.1 -+ 2.1 (range 0-28) for placebo. Pirenzepine treatment showed a significant (t = 2.61, P = 0.023) 42.9% reduction from baseline, while a 29.0% reduction by placebo was not significant (t = 1.30, P = 0.21). However, the difference in improvements with respect to treatments was not significant. These are shown in Figure 3.
120
TABLE | . DEMOGRAPHIC CHARACTERISTICSOF 13 PATIENTS Female
80,
6
7
60.
50 47.3 • 5.7 30-65
47 46.1 • 3.5 29-63
Male Sex Age Median Mean • SEM Range Race Black
100
.o ~
40. 20. O.
ill
Baseline
Placebo Pirenzepine
1
2
5
5
Fig 2. Mean number of total reflux episodes in 24 hr vs baseline
Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
299
White
and treatments. Vertical lines indicate SEM.
SATO ET AL 10
6.
~4. 2. 0 Baseline
I
Placebo
Pirenzepine
Fig 3. Mean number of reflux episodes >5 min in 24 hr vs baseline and treatments. Vertical lines indicate SEM. *P = 0.023 compared to baseline.
DISCUSSION Anticholinergics have proven their efficacy in reducing gastric acid secretion. A major drawback for the use of traditional anticholinergics is their diffuse effect on gastrointestinal tract smooth muscle. At acid-suppressing doses for peptic ulcer diseases, this anticholinergic effect is manifested by a decreased force of contraction, a disturbance in peristalsis, and a decreased LES pressure, which may worsen gastroesophageal reflux (1 I). Pirenzepine, an M1 receptor selective antimuscarinic agent with anticholinergic properties mostly restricted to the gastric mucosa (12), has been found to inhibit basal and pentagastrin-simulated acid secretion, and its ulcer healing rate (70-87%) is comparable to that of cimetidine (13-15). The rate of relapse six months after the discontinuation of therapy has been shown to be less for pirenzepine (28.5%) than that for cimetidine (47%) (16).
Because of its selectivity to gastric mucosa, pirenzepine has been used widely as an antiulcer agent in Europe, but it has not yet been marketed in the United States. Unlike classic anticholinergic agents such as atropine (17) and propantheline (18), pirenzepine has been associated with fewer side effects. However, Talley et al (19) reported a higher incidence of gastroesophageal reflux symptoms, ie, heartburn and epigastric pain in nonulcer dyspepsia patients who were treated with pirenzepine at 50 mg twice a day when compared to patients treated with cimetidine at 200 mg four times a day or with placebo. They speculated that this was due to the anticholinergic effect of pirenzepine on the LES. In previous studies from our laboratory we showed that pirenzepine did not decrease esophageal contraction pressure to the degree occurring with propantheline (4) and that pirenzepine produced less inhibition of cholinergic stimulation from bethanechol on the LES than propantheline (5). In addition, pirenzepine did not reduce salivary secretion or salivary capacity for acid neutralization (7), which may play a role in esophageal acid clearance (20). This finding was consistent with that reported by Jaup et al (8). In our present study, no patients reported adverse effects to either placebo or pirenzepine at the 50-mg three-times-a-day dose. Since this study was a single-day study for both pirenzepine and placebo, possible adverse effects of chronic doses of pirenzepine cannot be excluded. Few intraesophageal pH monitoring studies have been conducted to assess the effect of anticholinergics on gastroesophageal reflux in patients with
TABLE 2. PERCENT TIME OF ACID EXPOSURE (pH 5 min in 24 hr (F = 2.94, P = 0.088). When these were further assessed by paired t tests with a log10 transformation of percentage time of acid exposure in the supine position in order to stabilize its variance and adjusting P values for correlated measurements using the Huynh-Feldt technique to deflate the degrees of freedom of the F statistic (10), pirenzepine showed a significant reduction in episodes >5 min (t = 2.61, P = 0.023) and a trend towards significant treatment effect over placebo for a total percent reflux time during the 24-hr period (t = 2.13, P = 0.055). When the positions were further divided into upright and supine cateDigestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
gories, pirenzepine treatment showed consistently better, although not significant, improvement from baseline over placebo treatment. Pirenzepine showed a threefold improvement in percent change from baseline reflux time (21.4%) over placebo (7.3%) in the upright position. A greater magnitude of improvement from baseline in percent time of reflux was seen in the supine as well for pirenzepine (58.9%) vs placebo (43.6%). However, neither of these treatment effects were significant. A greater improvement seen in the supine position may be clinically important in prevention of esophageal injury secondary to reflux in the recumbent position. Similar treatment effects might be anticipated with respect to the total number of reflux episodes lasting >5 min, which is considered to reflect the clearance activity of the esophagus after an episode of acid reflux where significant improvement from baseline was seen for pirenzepine. Treatment of gastroesophageal reflux disease has changed greatly in the past decade due to the development of more effective acid-suppressing drugs and prokinetic agents. Unlike classic anticholinergic agents, which are not useful because of their diffuse effects over the gastrointestinal tract, including esophageal dysmotility and decreased LES pressure, pirenzepine may provide a unique alternative for some patients not satisfactorily controlled with other therapies. We recommend further large-scale studies with a longer duration of treatment to confirm the effect of pirenzepine in gastroesophageal reflux disease. ACKNOWLEDGMENTS The authors wish to thank Christine B. Dalton, Lisa Nowak, Gregg Stark, Jane W. Sinclair, and Faye Williams of the Gastroenterology Section of the Bowman Gray School of Medicine for their technical assistance. We are grateful to Timothy C. Craven of the Department of Public Health Sciences of the Bowman Gray School of Medicine for the initial statistical analysis and reanalysis of the data after the acceptance of this paper by the journal. REFERENCES 1. Goyal RK: Muscarinic receptor subtypes, physiology and clinical implications. N Engl J Med 321(15):1022-1028, 1989 2. Niemela S, Jaaskelainen T, Lehtola J, et al: Pirenzepine in the treatment of reflux oesophagitis, a placebo-controlled, double-blind study. Scand J Gastroenterol 21:1193-1199, 1986 3. Jaup BH, Abrahamsson H, Virtanen R, Iisalo E: Effective of pirenzepine compared with atropine and L-hyoscyamine on
301
SATO ET A L
4.
5.
6.
7.
8.
9.
10.
I 1.
12.
13.
14.
esophageal peristaltic activity in humans. Scand J Gastroenterol 17:233-239, 1982 Blackwell JN, Dalton CB, Castell DO: Oral pirenzepine does not affect esophageal pressure in man. Dig Dis Sci 31:230235, 1986 Hassan M, Dalton CB, Castell JA, Castell DO: Pirenzepine and propanthline effects on esophageal pressure responses to bethanechol. Gastroenterology 81(5):334-338, 1986 Castell DO: Pirenzepine: A unique 'anticholinergic' on the esophagus. In Pirenzepine: New aspects in Research and Therapy. Sept 18, 1984. Proceedings. A Bettarello (ed). Amsterdam, Excerpta Medica, 1985, pp 142-149 DeVault KR, Castell DO: Effects of antireflux therapies on salivary function in normal humans. Dig Dis Sei 32(6):603608, 1987 Jaup BH, Stockbrugger RW, Dotevall G: Comparison between the effect of pirenzepine and/-hyoscyamine in man. Scand J Gastroenterol 17(suppl 72): 119-122, 1982 Ward BW, Wu WC, Richter JE, Lui KW, Castell DO: Ambulatory 24-hr esophageal pH monitoring. J Clin Gastroenterol 8 (suppl 1):59-67, 1986 Huynh H, Feldt LS: Estimation of the box correlation for degrees of freedom from sample data in the randomized block and split plot designs. J Educ Stat 1:69-82, 1976 Bettarello A, Tuttle SG, Grossman MI: Effect of automatic drugs on gastroesophageal reflux. Gastroenterology 39:340346, 1960 Hammer R, Berrie CP, Birdsall NJM, Burgen ASV, Hulme EC: Pirenzepine distinguishes between subclasses of muscarinic receptors. Nature 283:90-92, 1980 Carmine AA, Brogden RN: Pirenzepine. A review of its pharmacodynamic and pharmacokinetic properties and therapeutic efficacy in peptic ulcer disease and other allied diseases. Drugs 30(2):85-126, 1985 Henry DA, Hawkey C, Somerville K, Burnham WR, Bell GD, Langman MJS: Pirenzapine and cimetidine in duodenal ulcer: A comparative study. In 7th World Congress of
302
15.
16.
17.
18.
19.
20.
21.
22.
Gastroenterology Proceedings. G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 214-216 Brown SG, Gomes M, Pounder RE, Rampton DS, Salmon PR, Sarner M, Vicary FR: Comparison of pirenzepine and cimetidine in gastric ulcer. In 7th World Congress of Gastroenterology Proceedings. G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 219-222 Vezzadini P, Sternini P, Valentini PL, Botti C, Lugli C, Labo G: Comparison of relapse rates in duodenal ulcer patients treated with pirenzepine or cimetidine. In 7th World Congress of Gastroenterology proceedings. G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 238-245 Texter E C, Patel GK, Navab F, et al: Effect of oral pirenzepine, a muscarinic (M1) receptor antagonist, and oral atropine on lower esophageal sphincter pressure, esophageal manometry and scintiscans, and gut peptides. Clin Res 31:766A, 1983 Toussaint J, Van der Veken J, Cremer M: A comparative double blind trial of pirenzepine and propantheline in duodenal ulcer In 7th World Congress of Gastroenterology Proceedings, G Dotevall (ed). Stockholm, Sweden, Excerpta Medica, 1982, pp 223-231 Talley NJ, McNeil D, Hayden A, Piper DW: Randomized, double-blind, placebo controlled crossover trial of cimetidine and pirenzepine in nonulcer dyspepsia. Gastroenterology 91:149-156, 1986 Helm JF: Effect of esophageal emptying and saliva on clearance of acid from the esophagus. N Engl J Med 310(5):284-288, 1984 Stockbrugger RW, Armbrecht U, Reul W: The effect of the Ml-selective telenzepine on esophageal acid exposure in healthy volunteers. Z Gastroenterol 26(6):297-302, 1988 Geller LI, Griaznova MV, Petrenko VF, Bessonova GA: The degree of gastroesophageal reflux during treatment of the exacerbation of duodenal peptic ulcer by various drugs. Ter Arkh 60(2):42-43, 1980
Digestive Diseases and Sciences, Vol. 37, No. 2 (February 1992)
