Simultaneous Measurement of Basal Pancreatic, Gastric Acid Secretion, Plasma Gastrin, and Secretin During Smoking

GASTROENTEROLOGY 73:758-761, 1977 Copyright © 1977 by the American Gastroenterological Association

Vol. 73, No. 4, Part 1 Printed in U.S A.

SIMULTANEOUS MEASUREMENT OF BASAL PANCREATIC, GASTRIC ACID SECRETION, PLASMA GASTRIN, AND SECRETIN DURING SMOKING S. N. S. MuRTHY, PH.D., V. P. DrNoso, JR., M.D., H. R. CLEARFIELD, M.D., AND Y. CHEY, M.D.

w.

Division of Gastroenterology, Hahnemann Medical College and Hospital , Philadelphia, Pennsylvania, and Isaac Gordon Institute for Gastrointestinal R esearch , Rochester, New York

The effect of smoking one unfiltered cigarette every 15 min for 1 hr on basal gastric acid and pancreatic secretion was studied in 10 subjects with a history of duodenal ulcer and 10 without duodenal ulcer. Smoking induced a transient rise of basal acid output followed by a slight decrease. This effect was more pronounced in the duodenal ulcer group. Smoking markedly inhibited fluid and bicarbonate secretion during the smoking period. The bicarbonate and fluid secretion returned to control levels within 30 to 60 min in the ulcer group and in 60 to 90 min in the nonulcer group. There was no difference in the degree of inhibition of pancreatic secretion between the two groups. Immunoreactive gastrin and secretin in the peripheral venous blood did not change significantly during smoking. The degree of inhibition of basal pancreatic secretion correlated well with the plasma concentrations of nicotine. These observations indicate that smoking in the fasting state induces alterations of basal gastric and pancreatic secretions which are not related to changes in plasma gastrin and secretin concentration, but to plasma concentrations of nicotine. The implication of this study in relation to a possible causal association between cigarette smoking and peptic ulcer disease is discussed. A causal relationship between cigarette smoking and peptic ulcer disease has been suggested and disputed. 1-4 Studies on the effect of cigarette smoking on gastric acid secretion in man have also been variable.s-8 Similarly, there is no universal agreement on the effect of systemic nicotine on pancreatic exocrine function. 9-11 Because the studies cited above were done under a variety of experimental conditions, using rabbits, dogs, and man, discrepancies of results and conclusions may be related, in part, to differences in protocols. In this report, we studied a specifically defined population whose clinical, secretory, and smoking characteristics were known or determined. We also limited our study to basal gastric acid and pancreatic secretion to eliminate other variables such as acid loads in the duodenum, dose of exogenous secretin, and background stimulation. 9-11 Finally, we measured simultaneously concentrations of gastrin, secretin, and nicotine in the plasma to ascertain the influence(s) of these substances on the changes of basal gastric acid and pancreatic secretion during smoking.

Materials and Methods The protocol for this study was approved by the Committee on Human Studies and Research ofHahnemann Medical ColReceived August 12, 1976. Accepted April 11, 1977. Address requests for reprints to: Vicente P. Dinoso, Jr. , M.D., Division of Gastroenterology, Hahnemann Medical College and Hospital, 230 North Broad Street, Philadelphia, Pennsylvania 19102. This study was supported by research grants from the Department of Medicine, Hahnemann Medical College and Hospital and, in part, from a grant from the Eberhard Foundation. 758

lege and Hospital on March 12 and 23, 1976, respectively. We studied 20 subjects, 17 males and 3 females, whose ages ranged from 24 to 63 years. Ten subjects had a clinical history of duodenal ulcer disease, proven radiologically and/or endoscopically. Their ulcers were in remission at the time of study. The 10 other subjects had no history of peptic ulcer or other gastrointestinal disorder. All subjects in the ulcer group and 5 in the nonulcer group smoked an average of one pack of cigarettes per day. Five subjects in the nonulcer group did not smoke. The range and mean of the 1-hr basal acid output in milliequivalents for the three groups were as follows: duodenal ulcer group, 4.00 to 17.98 with a mean of 10.58; nonduodenal ulcer, nonsmoker, 0.68 to 9.40 with a mean of 5.29; nonduodenal ulcer, smoker, 0.26 to 1.14 with a mean of 0.49. The subjects refrained from smoking 24 hr before the study. After an overnight fast, a triple lumen tube (Dreiling's tube with a polyethylene tubing added) was introduced orally so that the distal segment laid astride the duodenal loop, the proximal segment along the greater curvature of the stomach, and the polyethylene tubing in the midduodenum. The position of the tube was checked fluoroscopically before and after each study. All subjects were in the semirecumbent position during the study. Polyethylene glycol (PEG 4000, A. H. Thomas, Company, Philadelphia, Pa.; concentration, 5 g per liter; pH adjusted to 7.0) was infused at 1.15 ml per min using a Harvard infusion pump (Harvard Apparatus Company, Cambridge, Mass.). The PEG solution was allowed to equilibrate for 1 hr. A steady state was usually achieved within 45 min. Gastric and duodenal contents were collected every 15 min using an automatic intermittent suction pump.12 The completeness of duodenal aspiration was computed using the formula of Go et al. 13 The study periods were divided into three 1-hr periods: control, smoking, and postsmoking periods. In the subjects in whom the volume of duodenal and gastric aspirates have not

October 1977

759

SMOKING AND GASTRIC AND PANCREATIC SECRETION

returned to presmoking levels after 1 hr, the study was extended until the volumes collected per 15 min returned to control levels. Gastric and duodenal contents and peripheral venous blood were collected every 15 min. During the smoking period, the subject smoked an unfiltered cigarette with a nicotine content of 1.5 mg at the rate of 1 cigarette per 15 min. The subjects were instructed to inhale the smoke. The room was aerated after the smoking period. The volume, pH, and HCl concentration of the gastric aspirate were determined. The acid output was expressed in milliequivalents per 15 min. The volume and bicarbonate concentration were determined in the duodenal aspirate. Peripheral venous blood was collected in heparinized test tubes. Plasma for quantitation of immunoreactive gastrin and secretin was separated immediately. To each milliliter of plasma, 500 units of trasylol (FBA Pharmaceuticals, New York, N.Y.) were added. The concentration of plasma gastrin and secretin was quantitated by the radioimmunoassay method of Rehfeld and Stadil 14 and Tai et al., 15 respectively. Standard curves for both gastrin and secretin were set up in concentrations of 1, 2.5, 5, 10, 25, 50, 100, and 200 pg per ml of incubate. Standard curves were obtained by plotting percentage bound against concentration. The sensitivity of our assay system was 10 pg per ml for both gastrin and secretin. Nicotine was extracted from the plasma samples by the modification oflsaac and Rand.16 We used 50 ng ofO-toluidene per ml of plasma as internal standard. Nicotine was quantitated by gas-liquid chromatography using FM Scientific Model 402 (Hewlett-Packard Company, Avondale, Pa.). Nicotine was separated by using a glass column (8 feet by 3 mm) packed with 80/100 mesh Chromosorb W, coated with 10% 20 M Carbowax (PEG 20,000, Union Carbide Corporation, New York, N.Y.) and 10% polyphenyl ether (OS-138). The injector, detector, and column temperatures were 230°, 235° and 190°C, respectively. Gas flow rates of helium, air, and H 2 were maintained at 35, 300, and 40 ml per min. Under these conditions the retention time for nicotine was 14 min, whereas, for 0toluidene it was 6.7 min. The minimum detectable concentration of nicotine was 1.3 ng. Polyethylene glycol was analyzed by the method of HydenY Because the standard deviations during the control and postsmoking periods were large for all three groups of subjects, for clarity, the acid and pancreatic secretion are plotted as a flat rate per 15 min. The Student's t-test for paired data and the analysis of variance were used to test the significance during the smoking and postsmoking periods compared to the presmoking values.18 Results

Effect of smoking on basal gastric acid secretion. The effect of smoking on basal acid secretion is shown in figure 1. Smoking increased the acid output in the first 15 min in 7 of the 10 subjects in the duodenal ulcer group. The increase for the group as a whole was statistically significant (P < 0.05, using analysis of variance). In the nonulcer group, smoking increased the acid output later. Among the nonsmokers in this group smoking increased the 15-min acid output in the second 15 min period after initiation of smoking. Among the smokers, smoking increased the 15-min acid output in the third and fourth 15-min periods after the start of smoking. However, the increases were not statistically significant. The initial rise in acid secretion was followed by a slight and transient decrease for both ulcer and nonul-

cer groups, but the decreases were not statistically significant. The increase in basal acid output for all groups was attributable to increased HCl concentration and volume. The acid output returned to control levels promptly after cessation of smoking.

Effect of smoking on basal secretion of bicarbonate and fluid by the pancreas. The effect of smoking on bicarbonate and fluid secretion by the pancreas is shown in figures 2 and 3. There was a prompt and marked decrease of fluid and bicarbonate secretion after st art ing smoking. The changes which were statistically significant during smoking are indicated in the figures . The onset and degree of inhibit ion in the ulcer and nonulcer groups were similar. However, there was a difference in the length of time it took bicarbonat e and fluid secretion to return to control levels. The valu es returned to control levels within 30 to 60 min in the ulcer group, but it took 60 to 90 min for the nonulcer group. There was no difference in t he degree of inhibi4

c: E3 It) .....

ou-s

...

::J

....::Ja.

NOU-NS

01

l -.,.. _,_ _Jr ·

"0

~

_JLL~ NDU-S

u

< 0

60

120

110

Time in min FIG. 1. Effect of smoking on acid output expressed in milliequivalents per 15 min. DU-S, duodenal ulcer subjects, smokers ( n = 10); NDU-NS , nonduodenal ulcer subjects, nonsmokers ( n = 5); N DU-S, nonduodenal ulcer subjects, smokers ( n = 5); *, P < 0.05 by analysis of variance; t - t , period of smoking. ·4

• •

~

16

c:

14

E.3

12~

It)

..... r:T w

10

E.2 41

a 3

Cll

6 ...

... c:

i: 3CD

(JI

0

-e .1

3

4 ::J -·

Cll

u

iij

2 0 0

60

in min FIG. 2. Effect of smoking on bicarbonate and fluid secretion in the

duodenal ulcer group. *1, P < 0.005; *2, P < 0.001; *3, P < 0.005; *4 , P < 0.025; *6, P < 0.005; *7, P < 0.10; (Student's t-test for paired data).

760

MURTHY ET AL.

Vol. 73,No . 4,Partl

Discussion We elected to study basal gastric and pancreatic se14 c: cretion simultaneously for two reasons: (1) most studies ·e .3 12 ~ done previously in experimental animals and in man5-Io c: measured changes in gastric secretion or pancreatic ~ 10 ~ CT exocrine secretion independently. Because most of these w E .2 studies attempted to establish a link between inhibition Gl of pancreatic exocrine secretion and peptic ulcer dis111 c: ease, it is important to determine the changes in gastric 0 ~ .1 acid secretion while measuring pancreatic function; (2) 111 u it is common among smokers to smoke heavily between 2 iii meals. 0 ~o~~~~~~~1~~~~~0~~60~~~~~o~~~~~~~o Our findings on the effect of nicotine on pancreatic Time in min secretion of fluid and bicarbonate are similar to those FIG. 3. Effect of smoking on bicarbonate and fluid secretion in the observed by Konturek and his co-workers in dogs, 9 ' 10 to nonduodenal ulcer group. *I ,P < 0.001; *2,P < 0.01; *3,P < 0.01; *4, that by Solomon et al. in rabbits, 19 and to those of p < 0.005; *6, p < 0.010; *7, p < 0.010; *8, p < 0.001; *9, p < 0.010 Bocheneck and Koronczewski2° and Bynum et al. 21 in ~Student's t-test). man, but differ from that reported by Boden et al. 11 in dogs. 10 t Smoking t Our study, however, brings new information complementary to these previous studies. We found that smokE 10 ing at a rate common among heavy smokers profoundly ..... inhibited basal fluid and bicarbonate secretion by the ~60 pancreas for periods which ranged from 60 to 90 min .5 while the basal acid secretion returned promptly to control levels after cessation of smoking. This differing effect of nicotine on basal gastric and pancreatic secretion may be important in the hypothesis linking cigarette smoking to peptic ulcer disease. Indeed, preliminary studies in our laboratory indicate that a pH of less than 3.5 is maintained in the proximal duodenum dur60 Gl ing smoking in subjects with high basal acid secretory c 22 The observation of a more rapid return of bicarrates. 300 bonate secretion in the ulcer group is in keeping with .~ z the observation of Isenberg et al., who showed that 0 basal and acid-stimulated secretion of HC0 3 is greater in min in doudenal ulcer subjects compared to normal subFIG. 4. Plasma concentrations of gastrin, secretin, and nicotine jects, 23 and suggests that duodenal ulcer subjects have a before, during, and after smoking. 0 - 0 , gastrin; e-e, secregreater ability to rebound from inhibition by nicotine. tin. Rhodes et al. have recently established immunoreactive concentrations of secretin in the plasma in man tion and the duration of the return to control values during infusion of various doses of exogenous secretin between the smoker and nonsmokers in the nonulcer during intraduodenal infusion of HCl, and after a test group. meal. 24 Their studies indicate that intravenous adminisPlasma concentrations of gastrin, secretin, and nico- tration of exogenous secretin (GIH, Karolinska Institutine. The mean concentrations of gastrin, secretin and tet, Stockholm, Sweden) and intraduodenal infusions of nicotine are shown in figure 4. There was a trend to- HCl in doses and rates commonly employed in biological ward a rise in immunoreactive gastrin during the smok- experiments in man and various experimental animals ing period. The increases, however, failed to reach sta- exceed that which is physiologically attainable, e.g., tistical significance. There was essentially no change in after a test meal. Thus, the effect of nicotine on panthe serial concentrations of basal plasma secretin. creatic secretion stimulated by 1 clinical unit per kg per The plasma concentrations of nicotine are shown in hr or more of exogenous secreting...11 may not be applicafigure 4. The highest concentrations coincided with the ble to actual situations in man. periods of maximum inhibition of fluid and bicarbonate It now appears clear that inhibition of pancreatic secretion by the pancreas. The correlation coefficients at secretion by nicotine does not involve secretin, as pro15, 30, 45, and 60 min during the smoking period were posed earlier by Konturek and his co-workers. 10 Boden 0.290, 0.864, 0.925, and 0.451, respectively. The return of and his co-workers, in fact, found that nicotine, 100 p.,g fluid
-

9oe g,

October 1977

SMOKING AND GASTRIC AND PANCREATIC SECRETION

761

chronic duodenal ulcer. Lancet 1:808-809, 1934 opposite to the findings of an impressive array of stud3. Doll R, Avery Jones F , Pygott F: Effect of smoking on the ies. 9' 10• 19• 20 In this study we demonstrated marked inhiproduction and maintenance of gastric and duodenal ulcers. bition of bicarbonate and fluid secretion but plasma Lancet 1:657-662, 1958 immunoreactive concentrations of secretin remained es4. Monson RR: Cigarette smoking and body form in peptic ulcer. sentially unchanged. Compelling evidence that nicotine Gastroenterology 58:337-344, 1970 inhibits pancreatic secretion by releasing catechol- 5. Steigman F, Dolehide RH, Kaminski L: Effect of smoking toamines in rabbits has been presented by Solomon and bacco on gastric acidity and motility of hospital controls and his group. 19 patients with peptic ulcer. Am J Gastroenterol 22:299-409, 1954 Although most published data on smoking and pan6. Piper DW, Paine J M: Effect of smoking on gastric secretion. Lancet 1:696-698, 1959 creatic exocrine secretion claim that the amount of nicotine infused to produce inhibition of pancreatic secretion 7. Cooper P , Knight JB: Effect of cigarette smoking on gastric secretion of patients with duodenal ulcer. N Eng! J Med 255:17falls within the range of blood levels attained during 21, 1956 smoking, blood concentrations of nicotine during intra8. Schnedorf JG, Ivy AC: Effect of tobacco smoking on alimentary venous administration were never measured. To our tract. Experimental study .of man and animals. JAMA 112:898knowledge, our study is the first one to correlate 904, 1939 changes of pancreatic exocrine secretion and blood con- 9. Konturek SJ, Solomon TE, McCreight WG, et al: Effects of centrations of nicotine. Our data indicate that the denicotine on gastrointestinal secretions. Gastroenterology gree of inhibition of basal secretion ofHC03 and fluid by 60:1098-1105, 1971 the pancreas correlated well with the blood concentra- 10. Konturek SJ, Dale J, Jacobson ED, et al: Mechanism of nicotineinduced inhibition of pancreatic secretion of bicarbonat e in the tions of nicotine. dog. Gastroenterology 62:425-529, 1972 The biphasic basal acid response during smoking was unexpected. As discussed in the introduction, previous 11. Boden G, Shore IS, Koumar-Essa N , et al: Effect of nicotine on serum secretin and exocrine pancreatic function. Am J Dig Dis studies of the effect of smoking on acid secretion are 1976 conflicting. All of the three groups of subjects demon- 12. 21:974-977, Chey WY, Shay H, Nielsen OF, et al: Evaluation of tests of strated this effect. Differences in the magnitude and pancreatic function in chronic pancreatic disease. JAMA onset of the initial increase was observed within the 201:347-350, 1967 three groups. However, there were only 5 subjects in 13. Go VLW, Hofmann AF, Summerskill WHJ: Simultaneous meaeach subgroup of nonulcer subjects. Therefore, a definisurements of total pancreatic, biliary and gastric outputs in man tive conclusion on the significance of these observations using a perfusion technique. Gastroenterology 58:320-328, 1970 could not be drawn. Nevertheless, it appears reasonable 14. Stadil F, Rehfeld JF: Preparation of 1251-labelled synthetic h uman gastrin I for radioimmunoanalysis. Scand J Clin Lab Invest to suggest that this differing acid response may in part, 30:361-368, 1972 account for the conflicting data on the effect of smoking 25 on gastric acid secretion in the literature. The greater 15. Tai HH, Korsch B, Chey WY: Prepara tion of ' 1-labelled secretin of high specific radioactivity. Anal Biochem 69:34-42, 1975 stimulation of HCl secretion in the ulcer group may be 16. Isaac PF, Rand MJ: Cigarette smoking and plasma levels of explained by the greater reactivity of the parietal cells nicotine. Nature 236:308-310, 1972 to exogenous stimulatory substances in duodenal ulcer 17. Hyden S: A turbidometric method of determination of higher subjects, as was demonstrated recently with pentagaspolyethylene glycols in biological mater ials. Ann R Agric Coli trin by Isenberg et al. 25 Sweden 22:139-145, 1955 The secretory changes we have described suggest a 18. Snedecor GW, Cochran WG: Statistical Methods. Sixth edition. Ames, Iowa, Iowa State University Press, 1967 mechanism whereby smoking may be linked to duodenal ulcer disease. The transient increase of basal acid 19. Solomon TE, Solomon N , Shanbour LL, et al: Direct and indirect effects of nicotine on rabbit pancreatic secretion. Gastroenterolsecretion and its prompt return to basal levels associogy 67:276-283, 1974 ated with a marked and prolonged inhibition of pan20. Bocheneck WJ, Koronczewski R: Effects of cigarette smoking creatic bicarbonate secretion may create a highly acidic and volume of duodenal contents. Am J Dig Dis 18:729-733, 1973 pH in the duodenal bulb for long periods of time. Initial 21. Bynum TE, Solomon TE, Johnson LR, et al: Inhibition of panstudies in our laboratory indicate that this actually creatic secretion in man by cigarette smoking. Gut 13:361-365, occurs, especially in subjects with elevated basal acid 1972 secretion. 22 Inasmuch as smoking incessantly for sev- 22. Murthy SNS, Dinoso VP, Clearfield HR: Serial pH changes in eral hours at a time is common among heavy smokers, the proximal duodenum during smoking (a bstr). Gastroenterology 72:1106, 1977 and because nicotine inhaled accumulates in the blood, 16 it is highly likely that even more prolonged 23. Isenberg JI, Cano R, Bloom SR: Effect of graded amounts of acid instilled into the duodenum on pancreatic bicarbonate secretion inhibition of pancreatic secretion occurs than what we and plasma secretion in duodenal ulcer patients and normal have demonstrated here with just 1 hr of smoking.

REFERENCES 1. Barnett CW: Tobacco smoking as a factor in the production of peptic ulcer and gastric neuroses. Boston Med Surg J 197:457459, 1927 2. Trowel OA: The relation of tobacco smoking to the incidence of

subjects. Gastroenterology 72:6-8, 1977 24. Rhodes RA, Tai HH, Chey WY: Observations on plasma secretin levels by radioimmunoassay in response to duodenal acidification and to a meat meal in humans. Am J Dig Dis 21:873- 879, 1976 25. Isenberg JI, Grossman MI, Maxwell V, et al: Increased sensitivity to stimulation of acid secretion by pentagastrin in duodenal ulcer. J Clin Invest 55:330-337, 1975