GASTROENTEROLOGY
1991;100:1626-1629
Age Does Not Influence Acute Aspirin-Induced Gastric Mucosal Damage JOHN G. MOORE,
DAVID J. BJORKMAN,
MURRAY
D. MITCHELL,
and ANDRE AVOTS-AVOTINS Salt Lake City Veterans Affairs Medical Center and Departments and Gynecology, University of Utah, Salt Lake City, Utah
Gastroduodenal ulceration and complications occur more commonly in elderly patients consuming aspirin and other nonsteroidal antiinflammatory drugs than in younger cohorts. To test the hypothesis that the gastroduodenal mucosa of the elderly is more sensitive to nonsteroidal antiinflammatory drug damage, acetylsalicyclic acid in low (325 mg) and high (1300 mg) doses was administered on separate days to 10 elderly (median age, 67.5; range, 60-74 years) and 10 young (median age, 24.5; range, 19-30 years) healthy, male volunteer subjects. Gastroduodenal mucosal damage was assessed 2 hours after aspirin dosage by videoendoscopic techniques. At the conclusion of the study, the videotapes of all procedures were randomly evaluated by two endoscopists blinded to subject identification or study sequence. Endoscopically observed lesions were counted in the duodenum, gastric antrum, and gastric body. Aspirin damage was dose-related and more severe in the stomach than in the duodenum. There were no significant differences in total gastrodenodenal lesion counts between age groups. In addition, no differences in median gastric pH or tissue 6-ketoprostaglandin F,, content were observed between age groups. These data suggest that age is not a significant factor in determining the extent of acutely administered aspirin-induced damage.
cetylsalicyclic acid (ASA) administration produces predictable, dose-related damage to the gastric mucosa (l-3). With acute dosing, the damage is limited to the superficial mucosal layers and consists of mucosal hemorrhages or shallow erosions that develop within minutes or hours after exposure (4). Chronic ASA use, however, leads to gastroduodenal ulceration with complications in susceptible recipients (4,s). Other nonsteroidal antiinflammatory drugs (NSAIDs) typically produce less acute damage but, when consumed chronically, are as &erogenic as
of Medicine
and Obstetrics
ASA (1,2,4,6-g). Older people are major NSAID users and especially susceptible to gastrointestinal complications (9-11). The relative risk of hospitalization from ulcer complications in NSAID users over 60 years of age is 2.56 of that of NSAID users under age 60 (11). This suggests that the gastroduodenal mucosa in the elderly may be more susceptible to acute and chronic NSAID-induced damage. In this report, we compare acute ASA damage in a group of young and old healthy male subjects. Materials
and Methods
Subjects Twenty healthy male volunteer subjects particiin the study. Ten subjects were young (median age, 24.5; range, 19-30 years) and 10 elderly (median age, 67.5; range 60-74 years). All denied significant past or recent history of gastroduodenal illness. None of the younger subjects took any medication on a long-term basis. The older subjects were allowed to continue taking the following medications during study: imipramine (Z patients), enalapril (1 patient), and bumetanide (1 patient). All other medications were prohibited for 3 weeks before and during the study. Alcohol use was prohibited for 3 days before each endoscopic study. Tobacco use was not allowed during the day of the study. The study was approved by the Salt Lake City Veterans Administration Medical Center Institutional Review Board in 1988. pated
A
Study Design Each subject
underwent
three endoscopic
examina-
tions. Study after placebo
I. The first endoscopic study was performed administration. Subjects arrived in the Salt
used in thispaper: ASA, acetylsalicyclic acid. o 1991 by the American Gastroenterological Association 00s5oa5/91/$3.00
Abbreviation
AGE AND ASPIRIN GASTRIC DAMAGE
June 1991
Lake City Veterans Affairs Medical Center Gastrointestinal (GI) laboratory at 10 AM after an overnight fast. All subjects were given four placebo tablets in a cup of water (240 mL) and instructed to lie on their left side for 1 hour and then supine for a second hour. At noon an upper GI endoscopic examination was performed with a Pentax EG-2900 videoendoscope (Pantax Precision Instruments, Inc., Orangebury, NY). Anesthesia consisted of topical throat spray (10% lidocaine), IV Valium for sedation, and 1-2 mg IV glucagon to promote gastric atony. The entire procedure was videotaped and stored for later review. Three biopsy specimens from the gastric body were obtained for 6-keto-prostaglandin F,, (6-keto-PGF,,) determinations. Indicator strips (Baxter, McGraw, IL] were used to measure gastric pH on aspirated samples. Study 2. The second endoscopic study was p,erformed after an interval of at least 1 week. Subjects arrived as in study 1 at 10 AM. In a random design, subjects received either one ASA tablet (325 mg) with three placebo tablets or four ASA tablets (1300 mg) in a cup of water and were instructed to lie down as above. The ASA tablets provided for all studies came from a single bottle (Bayer Aspirin, lot no. 7K8G, expiration date, 10/92; Sterling Drug, Inc., New York, NY] stored in the dark at room temperature. At noon, endoscopic examination was performed as in study 1. Gastric pH was measured. Gastric biopsy specimens were not obtained. Study 3. The third endoscopic study was performed at least 2 weeks after study 2. Subjects arrived at the GI laboratory at 10 AM and received either one ASA tablet with three placebo tablets or four ASA tablets in a cup of water, depending on the randomization schedule established in study 2. Anesthesia and endoscopic examination were performed as in study 1.Gastric pH was measured. Gastric biopsy specimens were not obtained.
1627
Statistics Standard analysis of variance statistics were applied to the lesion counts, gastric pH values, and mucosal 6-ketoPGF,, determinations as shown in Figures l-3. Wilcoxon signed rank tests were applied to all comparisons between young and old subjects and to the lesion counts produced by placebo and low and high doses of ASA. Spearman rank correlations were calculated between total lesion count, at both ASA doses, and gastric pH and mucosal 6-keto-PGF,, concentrations.
Results Placebo (Study 1)
As expected, placebo administration produced the least damage (Figure 1). Fourteen of the 20 subjects had no lesions. The largest number of lesions (150) was seen in a subject who admitted taking two ASA tablets for a “headache” 2 days before the endoscopic examination, despite instructions prohibiting use of this drug. Aspirin (Studies 2 and 3)
Aspirin produced dose-related damage to the gastroduodenal mucosa (Figure 1). The fewest lesions were observed in the duodenum. Only 4 of 20 subjects on low-dose ASA demonstrated duodenal lesions,
@$ M
YOUNG MEN (N:~lO) OLD MEN (N--l@
__
_A_
t Gastroduodenal
Mucosa Lesion Counting
The videotapes of all procedures (n = 60) were randomly sorted and read at the same sitting by two endoscopists (J.G.M. and D.J.B.) at the conclusion of study. The endoscopists were blinded to subject identification and study sequence. The original taping and videorecorder replay were set at standard speed; slow motion, stop, and zoom modes were used when necessary for lesion identification. The number of separately identified punctate hemorrhagic lesions were recorded in the duodenum, gastric antrum, and gastric body. Gastric erosions, observed in only one study, were counted as single lesions. Linear “red streaks” were more frequently observed; most of them represented confluent punctate lesions that could be counted separately. When not, red streaks were counted as single lesions.
Gastric Biopsy 6-Keto-Prostaglandin Determinations Mucosal biopsy samples, obtained measured for 6-keto-PGF,, concentrations sue) by radioimmunoassay (12).
F,U
in triplicate, were (pg/mg wet tis-
‘I
1
PL.ACEBO Figure 1. Box-whisker
‘-x-1 325 mg ASPIRIN
: 1300 mg ASPIRIN
plot of total number of endoscopically counted gastroduodenal lesions during placebo, 325-mg ASA, and 1300-mg ASA studies. Solid line represents median value; upper and lower margins of box represent 75% and 25% quartiles of median value, respectively. Whiskers show the greatest or smallest values within one interquartile distance from the quartiles (box margins). Dashed line represents mean value; in 325-mg ASA plot with old men, the median and mean values coincide. Dots represent observed extreme values. No significant differences between age groups were detected.
1628 MOORE
ET AL.
GASTROENTEROLOGYVol.100,No.6
Figure 1 reflects the total lesion count (duodenum, antrum, and body) for placebo and each ASA dose. Both ASA doses produced significantly (P < 0.0001) more damage than placebo (median number of lesions for placebo, 0; for 325 mg ASA dose, 133; and for 1300-mg ASA dose, 271). The high ASA dose produced significantly (P < 0.0025) more lesions than the low ASA dose. There was no significant difference in total lesion count, at either ASA dose, between age groups, but the elderly group had significantly (P = 0.028) more damage to the duodenum at the high ASA dose. There was no significant correlation between total lesion count and gastric pH (325 mg, r = -0.092; 1300 mg, r = 0.172) or total lesion count and 6-keto-PGF,, concentrations (325 mg, r = -0.133; 1300 mg, r = -0.06).
9
8
m
YOUNG
MEN (N=lO)
M
OLD MEN (N=lO)
7 6
x
5
a4 3 2 1 0
PLACEBO
1300 rng ASPIRIN
325 mg ASPIRIN
Figure 2. Box-whisker plot of gastric pH values obtained during the three studies. The median value for young placebo, young 32-Smg ASA, and old 325-mg ASA studies is the lower margin of each box (corresponds to pH 1.0); the upper margin of each box corresponds to the 75% quartile. Other designations are as in Figure 1. No significant differences between age groups were detected.
whereas 8 of 20 on high dose had duodenal lesions; 7 of the 8 were elderly. The high and low dose of ASA caused an equivalent amount of damage in the gastric antrum and body. Only 3 of 40 ASA studies (20 subjects) failed to demonstrate antral lesions and only 1 study failed to demonstrate gastric body lesions.
c
m
YOUNG MEN (N=9)
M
OLD MEN (N=lO)
.
I..!
--T. Figure 3. Box-whisker plot of tissue 6-keto-PGF,, concentrations. Legend as in Figure 1. No significant differences between age groups were noted.
Gastric pH Median gastric pH values for the older subjects, compared with the younger subjects, were higher in all three endoscopic studies. However, these differences did not achieve statistical significance (Figure 2). Gastric Mucosal6-Keto Tissue Concentrations
Prostaglandin
G,,
6-Keto-PGF,, levels did not significantly differ between age groups (Figure 3). There was no significant correlation between tissue 6-keto-PGF,a concentrations and placebo study gastric pH (r = -0.384). Discussion This study demonstrates that age is not a significant factor in determining the severity of acute ASA-induced gastroduodenal mucosal damage. However, it should be noted that the healthy population of aged subjects examined in this study may not be comparable with elderly patients receiving NSAIDs who, in general, have arthritis, are more often female, and use NSAIDs chronically (10,ll). It has been suggested that arthritic patients are more susceptible to ulcer disease than age-matched controls, independent of the effect of therapy (13). However, others argue that this association is entirely the result of the medications used to treat arthritis (14,15). Endoscopic studies comparing genders in their susceptibility with acute NSAID gastrointestinal toxicity are not available. The current study addresses only the acute toxic effects on the gastric mucosa of a single ASA dose. This effect may or may not relate to gastroduodenal ulcerations or their complications in patients consuming NSAIDs on a chronic basis (1,4). The pathogenesis of acute NSAID-induced gastrointestinal toxicity dif-
AGE AND ASPIRIN GASTRIC
June 1991
fers from that of chronic NSAID-induced ulcerations. Acute damage is dependent on topical application of the NSAID, which, as an organic acid, is capable of penetrating the gastric mucosal barrier to cause disrupThe damage is tion of the epithelial surface (1,16,17). superficial and typically produces punctate hemorrhagic lesions and, less frequently, shallow erosions. This type of damage is universally observed after acute ASA administration, is considered of little clinical consequence, and may even resolve with continued therapy (4,18). Chronic NSAID GI toxicity is dependent on systemic inhibition of prostaglandin synthesis, a property shared by all NSAIDs and believed to be the central mechanism underlying the (1,17,19). production of gastroduodenal ulcerations In this regard, gastric mucosal6-keto-PGF,, concentrations did not differ between age groups in this study, a finding observed in other human studies (20,21). However, tissue 6-keto-PGF,, levels were not measured in response to ASA challenge, and it is possible that qualitative or quantitative differences in this response exist in the elderly, compared with younger subjects. The elderly also have diminished renal clearance of these drugs and, theoretically, enhanced susceptibility to gastrointestinal toxicity. Gastric pH levels did not significantly differ between age groups, contrary to the observation that gastric acidity declines with advancing age (21). However, 24hour intragastric pH profile comparisons between elderly and younger subjects are almost identical (22,231. It is possible that total acid production (a function of gastric volume and acid concentration) does decline with age but that pH (a measure of hydrogen ion concentration alone) is relatively unaffected. This observation is pertinent to the present study because the extent of ASA-induced acute gastric mucosal damage is directly related to luminal hydrogen ion concentration (16). In summary, the current study shows that acute ASA-induced gastric mucosal injury is not age dependent. Additionally, there was no significant difference in gastric pH or prostaglandin (6-keto-PGF,,) levels between age groups. These data suggest that there is no age-related physiological predisposition to acute NSAID injury in otherwise healthy subjects. References Graham DY, Smith JL. Aspirin and the stomach. Ann Intern Med 1986;104:390-398. Katza KA, Sunshine AG, Cohen S. The effect of non-steroidal anti-inflammatory drugs on upper gastrointestinal tract symptoms and mucosal integrity. J Clin Gastroenterology 1987;9:142148. Lanza FL. Endoscopic studies of gastric and duodenal injury after the use of ibuprofen, aspirin and other non-steroidal antiinflammatory agents. Am J Med 1984;77(Suppl):19-24. Graham DY. The relationship between non-steroidal antiinflam-
DAMAGE
matory drug use and peptic ulcer disease. Gastroenterol 5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16. 17. 18.
19.
20.
21.
22.
23.
1629
Clin
North Am 1990;19:171-182. Levy M. Aspirin use in patients with major upper gastrointestinal bleeding and peptic ulcer disease. New Engl J Med 1974;290: 1158-1162. Somerville K, Faulkner G, Langman M. Non-steroidal antiinflammatory drugs and bleeding peptic ulceration. Lancet 1986; 1:462-464. Armstrong CP, Blower AL. Non-steroidal anti-inflammatory drugs and life threatening complications of peptic ulceration. Gut 1987;28:527-532. Griffin MR, Roy WA, Schaffer W. Non-steroidal antiinflammatory drug use and death from peptic ulcer in elderly persons. Ann Int Med 1988;109:359-363. Langman MJS. Epidemiologic evidence on the association between peptic ulceration and antiinflammatory drug use. Gastroenterology 1989;96:640-646. Baum C, Kennedy D, Forbes M. Utilization of non-steroidal anti-inflammatory drugs. Arthritis Rheum 1985;28:666-692. Fries JF, Miller SR, Spitz PW. Toward an epidemiology of gastropathy associated with non-steroidal antiinflammatory drug use. Gastroenterology 1989;96:647-655. Magness RR, Osei-Boatren K, Mitchell MD, Rosenfeld CR. In vitro prostacyclin production by ovine uterine and systemic arteries: effects of angiotensin II. J Clin Invest 1985;76:22062212. Atwater EC, Morgan ES, Weicche DR, Jacox RF. Peptic ulcer and rheumatoid arthritis: a prospective study. Arch Int Med 1965;115:184-189. Kimberly R, Plotz P. Salicylates including aspirin and sulfasalozine. In: Kelley WN, Harris ED, Ruddy S, Sledge CB, eds. Textbook of rheumatology. 3rd ed. Philadelphia: Saunders, 1989:739-764. Malone DE, McCormick PA, Daly L, Jones B, Long A, Bresnihan B, Maloney J, O’Donoghue DP. Peptic ulcer in rheumatoid arthritis: intrinsic or related to drug therapy? Br J Rheum 1986;25:342-344. Davenport HW. Salicylate damage to the gastric mucosal barrier. New Eng J Med 1967;276:1307-1312. Ivey JK. Mechanisms of non-steroidal anti-inflammatory druginduced gastric damage. Am J Med 1988;84(Suppl 2A]:41-48. Graham DY, Smith JL, Spjut HJ, Torres E. Gastric adaptation studies in humans during continuous aspirin administration. Gastroenterology 1988;95:327-333. Sol1 AH, Kurata J, McGuiganJE. Ulcers, nonsteroidal antiinflammatory drugs and related matters. Gastroenterology 1989;96: 561-568. Dobo I, Tihanzi K, Banai J, Szanto I, Rosza I. Mucosal prostaglandin levels of the gastric stump. Gastroenterol Jpn 1988;23:514-520. Hawkey CJ. Synthesis of prostaglandin E,, thromboxane B, and prostaglandin catabolism in gastritis and gastric ulcer. Gut 1986;27:1484-1492. Grossman MI, Kirsner JB, Gillespie IE. Basal and histologystimulated gastric secretion in control subjects and in peptic ulcer or gastric ulcer. Gastroenterology 1963;45:14-26. Frank WO, Brauerman A, Palmer R, Young MD. Comparison of the pharmacodynamics of cimetidine in elderly population (abstr). Gastroenterology 1987;92:1395.
Received September 25,199O. Accepted November 15,199O. Address requests for reprints to: David J. Bjorkman, M.D., Veterans Affairs Medical Center, 500 Foothill Boulevard, Salt Lake City, Utah 84148. The authors thank the Department of Research, Salt Lake Veterans Affairs Medical Center, for support of this work.