Cigarette Smoking Inhibits Acid-Stimulated Duodenal Mucosal Bicarbonate Secretion

Methods

Participants

Six healthy nonsmoking volunteers (one woman and five men; mean age, 45 years; range, 35 to 60 years) with no previous history of peptic ulcer disease and seven healthy chronic smokers (all men; mean age, 41 years; range, 27 to 56 years) with no previous history of peptic ulcer disease participated in these studies. All smokers smoked at least one pack of cigarettes per day. The protocol was approved by the University of California at San Diego Human Subjects Committee; all participants gave written informed consent.

On the first test day a double-lumen nasogastric tube was fluoroscopically positioned so that the tip was in the middle of the gastric antrum. Basal gastric acid secretion and the response to pentagastrin, 6 g/kg body weight administered subcutaneously (Ayerst Laboratories; New York, New York), were determined. This procedure ensured that each participant had normal gastric secretory function and also introduced the participants to intubation methods. Basal and maximal acid outputs in smokers and nonsmokers were within normal limits and not significantly different between groups: 5.2 mmol/h (95% CI, 2.5 to 7.9 mmol/h) and 40.3 mmol/h (CI, 35.4 to 45.2 mmol/h) compared with 7.2 mmol/h (CI, 0.8 to 13.6 mmol/h) and 41.2 mmol/h (CI, 28.0 to 54.0 mmol/h), respectively.

Experimental Design

The nonsmokers participated in two studies and the smokers participated in four separate studies. The nonsmokers did not participate in the smoking experiments. All experiments were performed after a 12-hour fast and, in the smokers, after abstention from smoking for 12 hours. The order of experiments was randomized and only a single test was done on a single day. After two 15-minute basal collections in each study, participants puffed on either an unlit cigarette (sham smoking) or smoked one cigarette (Marlboro [18 mg tar, 1.1 mg nicotine [33], Philip Morris Inc.; Richmond, Virginia) per 15-minute period for a total of 60 minutes.

The methods used to isolate the proximal duodenal test segment free from gastric, pancreaticobiliary, and intestinal secretions have been described and validated previously [34-36]. In brief, a 4-cm segment of the proximal duodenum (the duodenal bulb) was isolated between two balloons. The isolated test segment was perfused with 154 mmol/L NaCl (2 mL/min, 37 C) containing Carbon-14-polyethylene glycol as a nonabsorbable marker. The effluent was collected in 15-minute intervals; pH and PCO ₂ were measured (Micro 13, Instrument Laboratories; Lexington, Massachusetts) in triplicate anaerobically and (HCO₃) was determined by using the Henderson-Hasselbalch equation. Bicarbonate outputs were calculated in 15-minute intervals and are expressed as mole/cm x hour. To ensure that the isolated proximal duodenal test segment remained uncontaminated with gastric and pancreaticobiliary secretions, markers of gastric (phenol red infused into the stomach) and pancreaticobiliary secretions (bilirubin and trypsin) were measured at 15-minute intervals. Experiments during which phenol red staining or trypsin greater than 10 U/mL or bile staining occurred were discarded. Almost all duodenal collections were free from contamination. The trypsin content in the remaining samples was only 2.3 U/mL (range, 1.8 to 2.7 U/mL), compared with a trypsin concentration of approximately 110 U/mL distal to the segment [34]. Further, the trypsin content in the isolated test segment was unrelated to bicarbonate secretion [34].

Basal bicarbonate secretion was measured for two 15-minute periods. The participants then either began sham smoking or actual smoking for two more 15-minute periods (periods 1 and 2) during which time the isolated duodenal test segment was still perfused with 154 mmol/L NaCl. To determine the effect of intraluminal acidification, the isolated duodenal segment was perfused for 5 minutes with 100 mmol/L HCl (iso-osmolar by addition of NaCl) at 37 C, equivalent to 2 mmol H+ [34]. After acidification, the duodenum was flushed with 154 mmol/L NaCl and acid-stimulated bicarbonate secretion was measured for 1 hour (periods 3 to 6).

The response to PGE₂ was also studied after measuring basal bicarbonate. Natural PGE₂ (Upjohn, Kalamazoo, Michigan) was infused at 8.33 g/mL in 154 mmol/L NaCl into the isolated test segment for 30 minutes (periods 3 and 4), equivalent to 1.4 mol of PGE₂. Bicarbonate secretion was measured for two more 15-minute periods thereafter (periods 5 and 6).

Calculations and Statistical Analysis

Corrected bicarbonate outputs were determined at 15-minute intervals using standard equations based on the recovery of Carbon-14-polyethylene glycol [37]. Results are expressed as means and 95% confidence intervals. All results were calculated from individual data. Net bicarbonate outputs indicate responses after the basal secretion was subtracted. Two-way analysis of variance and the Student t-test for paired and unpaired data were used in the statistical analysis, which was done with the computer program SPIDA (Statistical Package for Interactive Data Analysis; Statistical Laboratory, Macquarie University, New South Wales, Australia). Differences were considered significant if P < 0.05.

Results

Acid-Stimulated Bicarbonate Secretion

In the sham-smoking studies, mean basal duodenal mucosal bicarbonate outputs in nonsmokers and smokers were comparable: 185 mol/cm per hour (CI, 113 to 257 mol/cm per hour) and 178 mol/cm per hour (CI, 134 to 222 mol/cm per hour), respectively. These values are comparable to those reported previously in this and other laboratories [34-36]. In nonsmokers sham smoking had no significant effect on unstimulated bicarbonate secretion (Figure 1, top), whereas in smokers sham-smoking caused a modest increase. During sham smoking, bicarbonate secretion increased significantly in response to luminal acidification in both smokers and nonsmokers (Figure 1, top). The increases above basal were comparable: 242 mol/cm per hour (CI, 41 to 443 mol/cm per hour) and 220 mol/cm per hour (CI, 97 to 343 mol/cm per hour), respectively (145% and 120% above basal measurements, respectively).

View larger version:

• In this window
• In a new window

Figure 1. The effect of sham smoking in seven smokers () and six nonsmokers () (top panel) and effect of smoking () or sham smoking () in 7 smokers ( ) on mean ( 95% CIs) net proximal duodenal mucosal bicarbonate secretion in response to acid (2 mmol HCl [100 mmol/L] infused intraduodenally during a period of 5 minutes). * < 0.05 compared with basal. < 0.05 compared with sham smoking. Effect of smoking on acid-stimulated duodenal mucosal bicarbonate secretion.bottomPP

The mean basal duodenal mucosal bicarbonate output of smokers before smoking was comparable to the sham-smoking studies: 201 mol/cm per hour (CI, 152 to 250 mol/cm per hour) and 178 mol/cm per hour [CI, 134 to 222 mol/cm per hour], respectively. Compared with sham smoking, actual smoking failed to alter basal bicarbonate secretion (Figure 1, bottom). However, whereas luminal acidification resulted in a significant increase in bicarbonate secretion during sham smoking, smoking almost completely abolished the acid-stimulated response (Figure 1, bottom). The peak increase in response to luminal acid during smoking (53 mol/cm per hour [CI, 33 to 139 mol/cm per hour], 26% above basal) was only approximately 20% of the response observed during sham smoking (242 mol/cm per hour [CI, 41 to 443 mol/cm per hour], 145% above basal measurements). Further, the net total 30-minute bicarbonate outputs (periods 2 to 4) during sham smoking was 315 mol/30 min (CI, 26 to 604 mol/30 min) compared with only 45 mol/30 min (CI, 107 to 197 mol/30 min) during smoking (P < 0.01). Compared with sham smoking, cigarette smoking reduced the peak bicarbonate output in response to luminal acidification in each of the participants studied.

Prostaglandin-E₂-Stimulated Bicarbonate Secretion

During sham smoking, the basal bicarbonate secretion of the smokers did not differ significantly from the nonsmokers: 183 mol/cm per hour (CI, 132 to 234 mol/cm per hour) and 160 mol/cm per hour (CI, 103 to 217 mol/cm per hour), respectively. Further, during sham smoking PGE₂ significantly increased bicarbonate secretion in smokers and nonsmokers (Figure 2, top); the peak increases above basal measurements were comparable in the two groups: 126 mol/cm per hour (CI, 62 to 190 mol/cm per hour) and 121 mol/cm per hour (CI, 21 to 221 mol/cm per hour), respectively (73% and 114% above basal measurements, respectively).

View larger version:

• In this window
• In a new window

Figure 2. Effect of sham smoking in seven smokers () and six nonsmokers () (top panel) and effect of smoking () or sham smoking () in seven smokers (bottom panel) on the mean ( 95% CIs) net proximal duodenal mucosal bicarbonate secretion in response to PGE (1.4 mol infused intraduodenally during a period of 30 minutes). No significant differences were found at any time between smoking and sham smoking. * < 0.05 compared with basal measurements. Effect of smoking on prostaglandin-E₂-stimulated duodenal mucosal bicarbonate secretion.2P

Compared with sham smoking, smoking failed to alter basal and PGE₂-stimulated mucosal bicarbonate secretion (Figure 2, bottom). The integrated net 30-minute responses to PGE₂ during sham smoking and smoking were comparable: 159 mol/30 min (CI, 31 to 288 mol/30 min) and 122 [5 to 249] mole/30 min, respectively. During smoking, however, the alkaline response was delayed and did not reach statistical significance until cigarette smoking was discontinued (Figure 2, bottom).

Discussion

Our results indicate that smoking markedly inhibits acid-stimulated proximal duodenal mucosal bicarbonate secretion in humans. Further, smoking delays but does not appear to suppress the bicarbonate response to PGE₂. Smoking does not, however, alter basal bicarbonate secretory rates nor does it appear to have any long-term effect on acid- or PGE₂-stimulated duodenal mucosal bicarbonate secretion, because the smokers who had abstained from smoking for at least 12 hours had bicarbonate responses during sham smoking comparable to those of nonsmokers.

Cigarette smoking inhibited acid-stimulated, but not PGE₂-stimulated, duodenal mucosal bicarbonate secretion. Cryer and colleagues [30] reported recently that duodenal as well as gastric PGE₂ and PGF_2a concentrations measured in mucosal biopsies were significantly diminished in cigarette smokers compared with nonsmokers. Because the stimulatory effect of duodenal acidification is probably mediated in part through the release of PGE₂ [38], taken together, these results suggest that smoking may inhibit the response to luminal acidification through inhibition of acid-induced release of PGE₂. The inhibitory effect of smoking on human duodenal mucosal bicarbonate secretion is undoubtedly related to the constituents of tobacco smoke itself because sham smoking had no effect. Our study does not, however, allow us to formulate any firm conclusions as to which of the many constituents of tobacco smoke produces this effect. Nicotine itself is not likely to be the mediator, because intravenous infusion of nicotine, in doses corresponding to those occurring during cigarette smoking, had no effect on the duodenal mucosal bicarbonate secretion in anesthetized rabbits [39]. Other, more likely, mechanisms include inhibition of endogenous prostaglandin production [28-30], the stimulation of sympathetic neurons, or both [39].

Limited information is available regarding the effect of tobacco smoke on duodenal mucosal bicarbonate secretion; we are aware of no studies that have examined this phenomenon in humans. Granstam and colleagues [39] observed that tobacco smoke inhibited basal duodenal mucosal bicarbonate secretion in rodents. (The effect of smoking on stimulated duodenal bicarbonate output was not reported.) Our findings differed somewhat from theirs because we saw no effect of cigarette smoking on human basal bicarbonate output. This discrepancy may have been caused by differences in the quantity of tobacco smoke administered to animals, by differences between the species, or both.

The ulcerogenic effect of cigarette smoking is probably related to either stimulation of aggressive, inhibition of defensive factors, or both. Stimulation of aggressive factors is less likely because the effect of smoking on a major aggressive factor, gastric acid secretion, is inconsistent (ranging from no change to either inhibition or stimulation [16-22]). Information on the effect of smoking on duodenal defensive factors is scarce. Cigarette smoking has been reported to inhibit human pancreatic bicarbonate secretion [20, 25-27]. However, these studies examined mixed duodenal secretions (that is, pancreatic as well as duodenal plus hepatobiliary). Therefore, the decreases in luminal bicarbonate content attributed to a pancreatic effect may have been largely caused by smoking-induced decreases in mucosal or hepatobiliary bicarbonate secretion or both rather than by a specific effect on the pancreas. We selectively measured the proximal duodenal mucosal bicarbonate secretion.

Bicarbonate secreted by the gastroduodenal mucosa creates a pH gradient across the mucus layer immediately adjacent to the surface epithelium; pH at the cell membrane remains nearly neutral despite acidic luminal pH [31, 32, 40]. The juxtamucosal neutralization of acid by surface epithelium is an important factor in the protection of the duodenal mucosa from acid-peptic injury [31, 32, 41]. Further, in humans the major and minor pancreatic ducts enter the duodenum at approximately 6 to 10 cm distal to the pylorus [42]. Therefore, acid neutralization and inactivation of pepsins in the duodenal bulb are achieved, at least in part, by intrinsic mucosal bicarbonate production. Smoking-induced inhibition of mucosal bicarbonate secretion, therefore, would impair acid neutralization and probably pepsin inactivation by the duodenal bulb. This finding may contribute to the clinical observation that cigarette smokers are predisposed to develop duodenal ulcer disease and have delayed ulcer healing and a greater risk for ulcer-related complications than nonsmokers.

These data have been published in abstract form: Gastroenterology. 1992; 102:A32.