JPM Vol27. No. 4 July 1992:245-250
Gastric Mucosal Damage Due to Aspirin and Copper Aspirinate Assessed by Gastric Mucosal Potential Difference Changes A. A. Alich,
L. E. Wittmers,
Jr., Lee A. Anderson,
Department of Chemistry (A.A.A., Physiology (L.E. W., L.A.A.),
Elise M. Rieschl,
and Patricia
L. Peterson
E.M.R., P.L.P.), The College of St. Scholastica, Duluth; and Department University of Minnesota-Duluth School of Medicine, Duluth, Minnesota
of
When the gastric mucosa is damaged by antiinflammatory agents, such as aspirin, gastric mucosal potential difference (GPD) decreases and may or may not return to predamage values after the agent is removed. The magnitude and time course of the gastric potential difference changes have been suggested as a measure of mucosal damage. Male Sprague-Dawley rats were fasted, anesthetized, and surgically prepared for measurement of GPD by placement of electrodes in the gastric lumen and spleen. Test mixtures of aspirin, copper aspirinate, copper sulfate, or mixtures of aspirin and copper sulfate were administered by gavage, and subsequent changes in GPD were recorded. The area between the extrapolated control (baseline) GPD and the damage GPD was determined. The product of this area and the maximum change in GPD, the Reizindex (RI), was calculated. Values for all copper (II)-containing systems, including those with no aspirin, were significantly greater than aspirin alone. It was concluded that factors other than mucosal damage may contribute to a reduction in GPD and, therefore, an increase in Reizindex. Keywords:
Nonsteroidal
antiinflammatory
Introduction Many
antiinflammatory
drugs
damage
the gastric
mucosa on contact and/or during absorption. This damage is manifested as erosions or ulcers that appear primarily in the body and antrum of the stomach. A variety of techniques have been used to assess gastric mucosal damage, including visual lesion assessment, with (Brodie et al., 1970) or without (Davenport, 1964; Guth, 1973; Robins, 1980) the aid of an intravenous administration of protein-binding dye, blood loss into the gastric lumen (Lussier et al., 1978), radiography of the mucosal surface (Garner, 1978), changes in gastric potential difference Caspary, 1975; Ganote and Lucker, 1982; Geall et al., 1970; McGreevy, 1984; Sernka and Hogben, 1969; Tarnawski and Ivey, 1978), and loss of large molecules from the stomach contents (Flemstrom and Marsden, 1974; Malling, 1991; Wittmers et al., 1990). The ongoing efforts to develop new
Address reprint requests to Dr. A.A. Alich, The College of St. Scholastica, 1200 Kenwood Avenue, Duluth, MN 55811, U.S.A. Received January 15, 1992, revised and accepted April 28, 1992.
agents; Gastric damage assessment
methods are an indication of the difficulty in the interpretation and/or lack of precision that continue to plague present methods of evaluating gastric mucosal damage. Measurement of changes in gastric potential difference (GPD) is among the most recently developed methods of mucosal damage evaluation. Vance and coworkers (Vance et al., 1982) have proposed that changes in gastric potential difference provide the most quantitative, rapid, and reproducible means available for assessing the integrity of the gastric mucosal barrier. While the drop in gastric mucosal potential difference has been attributed to diffusion of ions down their concentration gradients after damage to the mucosa, (Sernka and Hogben, 1969), Caspary (1975) attributed it to the back-diffusion of hydrogen ion, secretion of sodium ion, and/or ultrastructural changes. It has also been observed that potential difference recovered to normal values even though some structural damage was still apparent (Baskin et al., 1976). Baskin’s group concluded that factors other than mucosal damage must contribute to changes in GPD. The correlation between gastric potential difference and damage/repair
Journal of Pharmacological and Toxicological Methods 27, 245-250 (1992) 0 1992Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas,
New York, NY 10010
1056-8719/92/$5.00
246
JPM Vol 27. No. 4 July 1!392:245-250
of gastric mucosal tissue appears to leave unanswered questions (Cummins and Vaughn, 1965; Tarnawski et al., 1978; Takeuchi et al., 1988). Recently, a new approach to interpretation of potential difference data was proposed in which Reizindex (RI), an irritation index, was calculated (Ganote and Lucker, 1982; Laule et al., 1982; Bruhn et al., 1983). The change in gastric potential difference was recorded as a function of time, and the area under the baseline was computed [AUB (the area between the extrapolated control baseline potential difference and the instantaneous potential as shown in Equation l)].
troesophageal junction, care being taken not to disturb the major nerve or vascular supplies. A small catheter was introduced into the stomach through an incision in the duodenum and was secured 0.5 cm below the gastroduodenal junction by a ligature around the duodenum. This catheter was used for washing the stomach and introduction of test mixtures. The stomach was initially irrigated with repeated aliquots of 0.9% saline (0.5 ml per 100 g body weight) until all food was removed.
Gastric Potential Difference AUB =
J 0
APDdt
(mV x min).
The Reizindex was then calculated RI
=
(APDnaxbUJB) 1000
(I)
by Equation 2:
(mV2
X
min).
(2)
A PDmaxwas the difference
between the baseline potential and the lowest potential (maximum change) achieved during the experiment. The A PD,,, term was considered to be proportional to the extent of loss of barrier integrity, and AUB to be proportional to the extent of disruption of barrier function. RI was thus suggested to be a more reliable and sensitive estimate of barrier dysfunction than either APD,,, or AUB alone (Laule et al., 1982). In the last several years copper aspirinate, a compound consisting of two copper(I1) and four aspirinate ions (Alich and Wittmers, 1980), has been proposed as a replacement drug for aspirin, based on the report that copper aspirinate was much more effective than aspirin as an antiinflammatory agent as well as being less damaging to the gastric mucosa (Sorenson, 1976). However, subsequent publications do not support the latter observation (Alich et al., 1983; Rainsford and Whitehouse, 1976; Kishore, 1988; McGahan, 1990). In the work reported here, GPD was measured after administration of aspirin, copper aspirinate, and various mixtures containing copper and/or aspirin and the RI was computed in order to evaluate damage to the gastric mucosa.
Methods Rat Preparation Male Sprague-Dawley rats (130-200 g weight, fasted for 24 hr with water allowed ad libitum) were anesthetized by an intraperitoneal injection of sodium pentobarbitol (Nembutal, Abbott Laboratories, North Chicago, IL). A tracheal cannula was inserted through a midline neck incision. The abdomen was opened with a midline incision, and a ligature was placed at the gas-
Measurement
The electrode system for measurement of gastric potential difference consisted of balanced calomel electrodes attached to a recording millivoltmeter (Beckman Expandomatic IV output to a strip chart recorder). Connection of the electrodes was accomplished by two salt bridges consisting of 36-cm lengths of 1.4-mm i.d. polyethylene tubing filled with 3% agar in saturated KC1 (Tarnawski et al., 1978). One end of one of these bridges was placed in the stomach through a second duodenal incision and secured with a ligature, taking care to minimize interruption of blood flow. One end of the other salt bridge was cut to give a sharp point and was inserted into the spleen through a small slit. The bridges each terminated in a beaker of saturated KC1 containing one of the calomel electrodes. A 35”-38”C, 0.9% saline solution, 0.5 mL/lOO g body weight, was instilled into the stomach, and a steady (baseline) potential was allowed to develop (for 15 min or until GPD stabilized at at least - 35 mV; lumen negative with respect to the vascular compartment). The baseline was recorded, the saline was withdrawn and replaced with one of the test mixtures (Table 1). The test mixture was allowed to remain in the stomach for 15 min and then replaced with an equal volume of saline. The measurement of GPD was continued for 45 min or until the GPD had recovered to the baseline value. All test compounds were suspended in 0.1% Tween 80 (ICN Biochemicals, Cleveland, OH) in glass-dis-
Table 1. Test Agent Dose Drug
Suspension
ASA @ASA.,
cuso4 ASA
+ CuSO4
Low Dose (FmoYkg rat)
High Dose (pmol/kg rat)
5.55 (ASA) or 10 mgikg 5.55 (ASA) 2.78 (Cu) 2.78 (01) 5.55 (ASA) 2.78 (Cu)
55.5 (ASA) or 100 mgikg 55.5 (ASA) 27.8 (Cu) 27.8 (Cu) 55.5 (ASA) 27.8 (Cu)
ALICH ET AL. GASTRIC MUCOSAL
247 DAMAGE
DUE TO ASPIRIN AND COPPER ASPIRINATE
dropped from the baseline to intersect the measured gastric potential difference at the 60-min point [Figure l(b)]. The enclosed area was the AUB used in computing the Reizindex (Equations 1 and 2). The enclosed area was tilled with a specific color and pattern and saved as a PICT file. Image 1.29, software designed for digital image processing and analysis on the Macintosh II (public domain program by Wayne Rasband) was used to calculate the area of the PICT images. Reizindex values were then obtained by Equation 2. Student’s t test was applied to both Reizindex and APD,, in order to evaluate differences between test suspension treatments. All statistical analyses employed the Statistical Package for the Social Sciences (SPSS) 4.0 for the Apple Macintosh.
AUB
Results -40 -20 0 !
,y, 0
30
60
Time (minutes) Figure 1.Illustration of the steps in computation of the Reizindex from a strip chart recording of gastric potential as a function of time before, during, and after treatment (in this case, the treatment was aspirin, high dose). Panel a is the data digitized from the strip chart recording and replotted from the graphics file. Panel b shows the definition of the area (AUB) used in computing the Reizindex. AUB is defined by extrapolation of the control potential and placement of the perpendicular line at 60 min posttreatment.
tilled water and sonicated prior to administration. Test suspensions were warmed to 35”-38°C and administered at 0.5 mWlO0 g body weight. Dose sizes of aspirin were 10 and 100 mg/kg rat as indicated in Table 1. Dose sizes of copper aspirinate were based on equivalent aspirin doses, while the other copper-containing systems were based on the copper aspirinate doses. The low dose of aspirin approximates two standard 325mg aspirin tablets administered to a 70-kg person. Test suspensions are described in Table 1.
Computer Reduction of Data Chart recordings of the gastric potential difference were digitized with a Summagraphics Bit Pad [Figure l(a)] into Canvas 2.1 graphics software (Deneba). The baseline was extrapolated to 60 min posttreatment, and in those cases in which the gastric potential difference did not return to baseline, a perpendicular line was
Gastric potential difference changes induced by aspirin treatment were not large in this experiment. Even with the high dose, the maximum decrease in GPD usually did not exceed 20 mV, and recovery to control potential after removal of the aspirin suspension was observed in most experiments. This was not the case for any of the copper(II)-containing mixtures. Even the low doses of the copper(II)-containing mixtures caused the GPD to decrease by at least 20 mV (high doses as much as 30 mV). With the low doses of copper(D), there was a partial recovery of GPD after the test suspension was removed and replaced with saline. However, with the high doses, almost no recovery was observed. Figure 1 contains a representative example of the response of a treatment (intermediate system) that did not recover fully during the experimental period. The control potential was recorded with the vehicle control solution (0.9% saline) in the stomach prior to the beginning of each experiment. Thus the control values of RI and APD,,, would, of necessity, always be zero. In order to evaluate the relationship between changes in GPD and dose of the agent administered, RI was plotted against the three treatment doses of aspirin (Figure 2). Linear regression analysis of this data yielded a regression coefficient (R) of 0.999, indicating an excellent correlation between treatment dose of aspirin and potential difference response. The remainder of the treatment groups consisted of two sets of animals each: a low-dose set corresponding to the lo-mg ASA/kg rat group, and a high-dose set corresponding to the 100 mg ASA/kg group. From Figure 3, which summarizes the Reizindex values for all treatment groups, it is immediately clear that all copper(B)-containing systems exhibit higher Reizindices than does aspirin alone. The group of animals tested
248
JPM Vol 27. No. 4 July 1%2:245-250
T
14-
12-
=lSEM
10 -
8-
Figure 2. Changes in Reizindex with dose of aspirin. Error bars represent standard error of the mean.
6-
40
= -1.18+0.12x
R=0.999
80
100
60
120
Dose of Aspirin (mg/Kg)
70
60
with Cu(I1) (as copper sulfate) have an RI of approximately the same magnitude as those of copper aspirinate-treated animals. The group of animals tested with a mixture of copper(I1) and aspirin exhibit larger (borderline) significance values than does copper aspirinate . The use of the Reizindex was expected to yield results that would be more amenable to interpretation. Lucker and coworkers (Ganote and Lucker, 1982; Laule et al., 1982; Bruhn et al., 1983) proposed that the RI represents a more reliable estimate of mucosal damage than the simple change in GPD. In the experiments presented here, if the A PD,,, comparisons were significantly different, then RI was also. Of 30 comparisons made, RI actually produced four more test comparisons that were significantly different than did A PD,,, (Table 2).
/-J ASA
Cu,ASA,
fl
n cuso, q
ASA + CuSO,
Discussion Low
dose
High dose
Dose of Test compound (mg/Kg)
Figure 3. Summary of Reizindex values for the four test mixtures. Aspirin-containing doses have either 5.55 (low) or 55.5 (high) wmol of aspirin/kg body weight. Copper-containing systems have either 2.78 or 27.8 p,mol of Cu(II)/kg body weight. Error bars represent standard error of the mean. Nthe number of animals in the group and is shown on each bar.
A dose response with more than adequate sensitivity was observed for increasing doses of aspirin as shown in Figure 2. The best fit for these data was obtained with a linear plot of dose versus response (RI). The mean values of RI were tested for significance (Student’s t test; significance level p < 0.05). RI for both low- and high-dose ASA were significantly less than all copper(containing systems, but the copper(II)-containing systems were not significantly dif-
ALICH ET AL. GASTRIC MUCOSAL
249 DAMAGE
DUE TO ASPIRIN AND COPPER ASPIRINATE
Table 2. Test Comparisons
in which the Reizindex Produced a Significant Difference But APD,,, Did Not Treatment Comparison” 1 compared 1 compared 6 compared 5 compared
with with with with
2 3 4 6
Reizindex (mV2 X min)
p
0.014 0.048 0.034 0.006
0.01 0.05 0.04 0.01
A PD,,, @VI
0.168 0.614 0.808 0.214
P
0.15 0.66 0.81 0.27
This research was supported in part by grants from Miller-Dwan Medical Center Foundation, Duluth Clinic Education and Research Foundation, the College of St. Scholastica Faculty Development Fund, and Minnesota Medical Foundation, and a 3M Fellowship to Elise M. Rieschl. The authors wish to acknowledge the assistance of Bradley Ingersoll for statistical analysis, and David Nyback for technical assistance.
a 1: high dose of aspirin; 2: low dose of copper aspirinate; 3: low dose of copper sulfate; 4: high dose of copper sulfate; 5: low dose of copper sulfate plus aspirin; and 6: high dose of copper sulfate plus aspirin.
References Alich AA, Wittmers LE, Jr. (1980) Hydrolysis of tetrakis-p,-acetylsalicylato-dicopper(I1). J Pharm Sci 69~125-127.
ferent from each other. The mixture of copper(U) and aspirin yielded the largest RI values, approximating an additive response for the two agents. The similarity between values for copper aspirinate on the mixtures of aspirin and copper sulfate led to the conclusion that hydrolysis of the complex was not a significant factor under the conditions of this experiment. RI of copper(I1) alone was as large as that of copper aspirinate. If the change in gastric mucosal potential is interpreted as indicating mucosal damage, then this was a surprising result, as it has been previously shown (Alich et al., 1983) by both lesion assessment using Guth’s method of visual lesion assessment (Guth, 1973) and microscopic histological examination that copper(I1) alone does not damage gastric mucosa. Because the system containing copper(H) with no aspirin at all had an RI as large as the other copper-containing systems with aspirin, we concluded that factors other than, or in addition to, mucosal damage are involved in the change in gastric potential difference in systems containing copper(I1). The use of the RI produced slightly improved results (over the use of APD,,, alone). It appears that for animals not showing recovery of GPD during the experiment, that is, those with more serious damage, the inclusion of the time factor yields a more complete picture of damage. Thus, for experiments yielding severe mucosa damage, the RI calculations may be preferred. Where damage is not serious, the APD,,, would take less time and yield just as acceptable results. In summary, whereas Reizindex is a reliable mucosal damage indicator, with an adequate dose response for aspirin, the same is not true for copper(II)containing systems. Factors other than mucosal damage cause changes in gastric potential difference in systems containing copper(I1). Furthermore, the use of the RI in data analysis results in a slight improvement over the use of APD,,, alone. Finally, hydrolysis of copper aspirinate does not occur to a significant extent during these experiments.
Alich AA, Welsh VJ, Wittmers LE, Jr. (1983) Comparison of aspirin and copper aspirinate with respect to gastric mucosal damage in the rat. J Pharm Sci 72: 1457-1461. Baskin WN, Ivey KJ, Krause WJ, Jeffrey GE, Gemmell RT (1976) Aspirin-induced ultrastructural changes in human gastric mucosa. Ann Intern Med 85~299-303. Brodie DA, Tate CL, Hooke KF (1970) Aspirin: Intestinal damage in rats. Science 170:183-185. Bruhn R, Ganote DP, Lucker PW, Vance J, Procaccini RL (1983) Comparative gastric irritation by two buffered acetylsalicylic acid formulations and acetysalicylic acid. ArzneimittelForschung 33: 158-162.
Caspary WF (1975) Einfluss von Aspirin, Antacida, Alkohol, und Gallensluren auf die transmurale elektrische Potentialdifferenz des menschlichen Magens. Dtsch Med Wochenschr 100: 1263-1268. Cummins JT, Vaughn BE (1965) Ionic relationships of the bioelectrogenie mechanism in isolated rat stomach. Biochem Biophys Acta 94~280-292.
Davenport HW (1964) Functional significance of gastric mucosal barrier to sodium. Gastroenterology 47: 142-152. Flemstrom G, Marsden NVB (1974) Increased inulin absorption from the cat stomach exposed to acetylsalicylic acid. Acta Physiol Stand
92:517-525.
Ganote DP, Lucker PW (1982) A new mathematical approach to describe gastric potential difference changes. Arzneim Forsch Drug Res 321552-556.
Gamer A, (1978) Mechanisms of action of aspirin on the gastric mucosa of the guinea pig. Acta Physiol Stand (Special Suppl.): 101-110. Geall MG, Phillips SF, Summer&ill WHJ (1970) Profile of gastric potential difference in man. Gastroenterology 58:437-443. Guth PH (1973) Experimental production of peptic ulcer. Gastroenterology 64:1187-1189. Kishore V (1988) Antiinflammatory activities of copper-aspirinate and aspirin in adjuvant arthritic rats. Res Comm Chem Pathol Pharmacol60:257-260.
Laule H, Lucker PW, Altmayer P, Eldon MA (1982) Gastric potential difference as a model in clinical pharmacology. Eur J Cfin Pharmaco/22:147-151. Lussier A, Arsenault A, Varady J (1978) Gastrointestinal microbleeding after aspirin and naproxen. C/in Pharmacol Ther 23: 402-407.
Malling T (1991) Use of radioactive inulin and sucrose in the assessment of gastric mucosal damage. Abstracts: 32nd Annual National Student Research Forum, Galveston, TX, April 11-13, pp. 14.
250
JPM Vol27. No. 4 July 19!32:245-250
McGahan M (1990) Copper and aspirin treatment increase the antioxidant activity of plasma. Agents Actions 31:59-64.
Sorenson JRJ (1976) Copper chelates as possible active forms of the antiarthritic agents. J Med Chem 19:135-148.
McGreevy JM (1984) Gastric surface cell function: Potential difference and mucosal barrier. Am J Physiol247:G79-G87.
Takeuchi K, Nishiwaki H, Dsano H, Ebara S, Okabe S (1988) Repair of mucosal damage induced by ethanol in the rat stomach. Digestion 40: l-10.
Rainsford KD, Whitehouse MW (1976) Concerning the merits of copper aspirin as a potential antiinflammatory drug. J Pharm Pharmaco/ 28:83-86. Rainsford KD, Whitehouse MW (1976) Gastric mucus effusipn elicited by oral copper compounds: Potential antiulcer activity. Experientia 32:1172-1173.
Tamawski A, Ivey KJ (1978) Transmucosal potential difference protile in rat upper gastrointestinal tract. Can J Physiol Pharmacol 56:471-473.
Robins PG (1980) Ultrastructural observations on the pathogenesis of aspirin induced gastric erosions. Br .I Exp Pathol61:497-504.
Vance J, Lucker PW, Tilling W, Procaccini R, Wetzelsbergen N (1982) The transmural gastric potential difference in combination with pharmacokinetic profiles. Methods Find Exp Clin Pharmaco1 4533-538.
Sernka TJ, Hogben CAM (1969) Active ion transport by isolated gastric mucosa of rat and guinea pig. Am J Physiol 215: 1419-1423.
Wittmers JR LE, Anderson LA, Fall MM, Alich AA (1990) Intragastric inulin as a measure of mucosal damage caused by aspirin. .I Pharmacol Methods 241229-239.
Gastric Mucosal Damage Due to Aspirin and Copper Aspirinate Assessed by Gastric Mucosal Potential Difference Changes A. A. Alich,
L. E. Wittmers,
Jr., Lee A. Anderson,
Department of Chemistry (A.A.A., Physiology (L.E. W., L.A.A.),
Elise M. Rieschl,
and Patricia
L. Peterson
E.M.R., P.L.P.), The College of St. Scholastica, Duluth; and Department University of Minnesota-Duluth School of Medicine, Duluth, Minnesota
of
When the gastric mucosa is damaged by antiinflammatory agents, such as aspirin, gastric mucosal potential difference (GPD) decreases and may or may not return to predamage values after the agent is removed. The magnitude and time course of the gastric potential difference changes have been suggested as a measure of mucosal damage. Male Sprague-Dawley rats were fasted, anesthetized, and surgically prepared for measurement of GPD by placement of electrodes in the gastric lumen and spleen. Test mixtures of aspirin, copper aspirinate, copper sulfate, or mixtures of aspirin and copper sulfate were administered by gavage, and subsequent changes in GPD were recorded. The area between the extrapolated control (baseline) GPD and the damage GPD was determined. The product of this area and the maximum change in GPD, the Reizindex (RI), was calculated. Values for all copper (II)-containing systems, including those with no aspirin, were significantly greater than aspirin alone. It was concluded that factors other than mucosal damage may contribute to a reduction in GPD and, therefore, an increase in Reizindex. Keywords:
Nonsteroidal
antiinflammatory
Introduction Many
antiinflammatory
drugs
damage
the gastric
mucosa on contact and/or during absorption. This damage is manifested as erosions or ulcers that appear primarily in the body and antrum of the stomach. A variety of techniques have been used to assess gastric mucosal damage, including visual lesion assessment, with (Brodie et al., 1970) or without (Davenport, 1964; Guth, 1973; Robins, 1980) the aid of an intravenous administration of protein-binding dye, blood loss into the gastric lumen (Lussier et al., 1978), radiography of the mucosal surface (Garner, 1978), changes in gastric potential difference Caspary, 1975; Ganote and Lucker, 1982; Geall et al., 1970; McGreevy, 1984; Sernka and Hogben, 1969; Tarnawski and Ivey, 1978), and loss of large molecules from the stomach contents (Flemstrom and Marsden, 1974; Malling, 1991; Wittmers et al., 1990). The ongoing efforts to develop new
Address reprint requests to Dr. A.A. Alich, The College of St. Scholastica, 1200 Kenwood Avenue, Duluth, MN 55811, U.S.A. Received January 15, 1992, revised and accepted April 28, 1992.
agents; Gastric damage assessment
methods are an indication of the difficulty in the interpretation and/or lack of precision that continue to plague present methods of evaluating gastric mucosal damage. Measurement of changes in gastric potential difference (GPD) is among the most recently developed methods of mucosal damage evaluation. Vance and coworkers (Vance et al., 1982) have proposed that changes in gastric potential difference provide the most quantitative, rapid, and reproducible means available for assessing the integrity of the gastric mucosal barrier. While the drop in gastric mucosal potential difference has been attributed to diffusion of ions down their concentration gradients after damage to the mucosa, (Sernka and Hogben, 1969), Caspary (1975) attributed it to the back-diffusion of hydrogen ion, secretion of sodium ion, and/or ultrastructural changes. It has also been observed that potential difference recovered to normal values even though some structural damage was still apparent (Baskin et al., 1976). Baskin’s group concluded that factors other than mucosal damage must contribute to changes in GPD. The correlation between gastric potential difference and damage/repair
Journal of Pharmacological and Toxicological Methods 27, 245-250 (1992) 0 1992Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas,
New York, NY 10010
1056-8719/92/$5.00
246
JPM Vol 27. No. 4 July 1!392:245-250
of gastric mucosal tissue appears to leave unanswered questions (Cummins and Vaughn, 1965; Tarnawski et al., 1978; Takeuchi et al., 1988). Recently, a new approach to interpretation of potential difference data was proposed in which Reizindex (RI), an irritation index, was calculated (Ganote and Lucker, 1982; Laule et al., 1982; Bruhn et al., 1983). The change in gastric potential difference was recorded as a function of time, and the area under the baseline was computed [AUB (the area between the extrapolated control baseline potential difference and the instantaneous potential as shown in Equation l)].
troesophageal junction, care being taken not to disturb the major nerve or vascular supplies. A small catheter was introduced into the stomach through an incision in the duodenum and was secured 0.5 cm below the gastroduodenal junction by a ligature around the duodenum. This catheter was used for washing the stomach and introduction of test mixtures. The stomach was initially irrigated with repeated aliquots of 0.9% saline (0.5 ml per 100 g body weight) until all food was removed.
Gastric Potential Difference AUB =
J 0
APDdt
(mV x min).
The Reizindex was then calculated RI
=
(APDnaxbUJB) 1000
(I)
by Equation 2:
(mV2
X
min).
(2)
A PDmaxwas the difference
between the baseline potential and the lowest potential (maximum change) achieved during the experiment. The A PD,,, term was considered to be proportional to the extent of loss of barrier integrity, and AUB to be proportional to the extent of disruption of barrier function. RI was thus suggested to be a more reliable and sensitive estimate of barrier dysfunction than either APD,,, or AUB alone (Laule et al., 1982). In the last several years copper aspirinate, a compound consisting of two copper(I1) and four aspirinate ions (Alich and Wittmers, 1980), has been proposed as a replacement drug for aspirin, based on the report that copper aspirinate was much more effective than aspirin as an antiinflammatory agent as well as being less damaging to the gastric mucosa (Sorenson, 1976). However, subsequent publications do not support the latter observation (Alich et al., 1983; Rainsford and Whitehouse, 1976; Kishore, 1988; McGahan, 1990). In the work reported here, GPD was measured after administration of aspirin, copper aspirinate, and various mixtures containing copper and/or aspirin and the RI was computed in order to evaluate damage to the gastric mucosa.
Methods Rat Preparation Male Sprague-Dawley rats (130-200 g weight, fasted for 24 hr with water allowed ad libitum) were anesthetized by an intraperitoneal injection of sodium pentobarbitol (Nembutal, Abbott Laboratories, North Chicago, IL). A tracheal cannula was inserted through a midline neck incision. The abdomen was opened with a midline incision, and a ligature was placed at the gas-
Measurement
The electrode system for measurement of gastric potential difference consisted of balanced calomel electrodes attached to a recording millivoltmeter (Beckman Expandomatic IV output to a strip chart recorder). Connection of the electrodes was accomplished by two salt bridges consisting of 36-cm lengths of 1.4-mm i.d. polyethylene tubing filled with 3% agar in saturated KC1 (Tarnawski et al., 1978). One end of one of these bridges was placed in the stomach through a second duodenal incision and secured with a ligature, taking care to minimize interruption of blood flow. One end of the other salt bridge was cut to give a sharp point and was inserted into the spleen through a small slit. The bridges each terminated in a beaker of saturated KC1 containing one of the calomel electrodes. A 35”-38”C, 0.9% saline solution, 0.5 mL/lOO g body weight, was instilled into the stomach, and a steady (baseline) potential was allowed to develop (for 15 min or until GPD stabilized at at least - 35 mV; lumen negative with respect to the vascular compartment). The baseline was recorded, the saline was withdrawn and replaced with one of the test mixtures (Table 1). The test mixture was allowed to remain in the stomach for 15 min and then replaced with an equal volume of saline. The measurement of GPD was continued for 45 min or until the GPD had recovered to the baseline value. All test compounds were suspended in 0.1% Tween 80 (ICN Biochemicals, Cleveland, OH) in glass-dis-
Table 1. Test Agent Dose Drug
Suspension
ASA @ASA.,
cuso4 ASA
+ CuSO4
Low Dose (FmoYkg rat)
High Dose (pmol/kg rat)
5.55 (ASA) or 10 mgikg 5.55 (ASA) 2.78 (Cu) 2.78 (01) 5.55 (ASA) 2.78 (Cu)
55.5 (ASA) or 100 mgikg 55.5 (ASA) 27.8 (Cu) 27.8 (Cu) 55.5 (ASA) 27.8 (Cu)
ALICH ET AL. GASTRIC MUCOSAL
247 DAMAGE
DUE TO ASPIRIN AND COPPER ASPIRINATE
dropped from the baseline to intersect the measured gastric potential difference at the 60-min point [Figure l(b)]. The enclosed area was the AUB used in computing the Reizindex (Equations 1 and 2). The enclosed area was tilled with a specific color and pattern and saved as a PICT file. Image 1.29, software designed for digital image processing and analysis on the Macintosh II (public domain program by Wayne Rasband) was used to calculate the area of the PICT images. Reizindex values were then obtained by Equation 2. Student’s t test was applied to both Reizindex and APD,, in order to evaluate differences between test suspension treatments. All statistical analyses employed the Statistical Package for the Social Sciences (SPSS) 4.0 for the Apple Macintosh.
AUB
Results -40 -20 0 !
,y, 0
30
60
Time (minutes) Figure 1.Illustration of the steps in computation of the Reizindex from a strip chart recording of gastric potential as a function of time before, during, and after treatment (in this case, the treatment was aspirin, high dose). Panel a is the data digitized from the strip chart recording and replotted from the graphics file. Panel b shows the definition of the area (AUB) used in computing the Reizindex. AUB is defined by extrapolation of the control potential and placement of the perpendicular line at 60 min posttreatment.
tilled water and sonicated prior to administration. Test suspensions were warmed to 35”-38°C and administered at 0.5 mWlO0 g body weight. Dose sizes of aspirin were 10 and 100 mg/kg rat as indicated in Table 1. Dose sizes of copper aspirinate were based on equivalent aspirin doses, while the other copper-containing systems were based on the copper aspirinate doses. The low dose of aspirin approximates two standard 325mg aspirin tablets administered to a 70-kg person. Test suspensions are described in Table 1.
Computer Reduction of Data Chart recordings of the gastric potential difference were digitized with a Summagraphics Bit Pad [Figure l(a)] into Canvas 2.1 graphics software (Deneba). The baseline was extrapolated to 60 min posttreatment, and in those cases in which the gastric potential difference did not return to baseline, a perpendicular line was
Gastric potential difference changes induced by aspirin treatment were not large in this experiment. Even with the high dose, the maximum decrease in GPD usually did not exceed 20 mV, and recovery to control potential after removal of the aspirin suspension was observed in most experiments. This was not the case for any of the copper(II)-containing mixtures. Even the low doses of the copper(II)-containing mixtures caused the GPD to decrease by at least 20 mV (high doses as much as 30 mV). With the low doses of copper(D), there was a partial recovery of GPD after the test suspension was removed and replaced with saline. However, with the high doses, almost no recovery was observed. Figure 1 contains a representative example of the response of a treatment (intermediate system) that did not recover fully during the experimental period. The control potential was recorded with the vehicle control solution (0.9% saline) in the stomach prior to the beginning of each experiment. Thus the control values of RI and APD,,, would, of necessity, always be zero. In order to evaluate the relationship between changes in GPD and dose of the agent administered, RI was plotted against the three treatment doses of aspirin (Figure 2). Linear regression analysis of this data yielded a regression coefficient (R) of 0.999, indicating an excellent correlation between treatment dose of aspirin and potential difference response. The remainder of the treatment groups consisted of two sets of animals each: a low-dose set corresponding to the lo-mg ASA/kg rat group, and a high-dose set corresponding to the 100 mg ASA/kg group. From Figure 3, which summarizes the Reizindex values for all treatment groups, it is immediately clear that all copper(B)-containing systems exhibit higher Reizindices than does aspirin alone. The group of animals tested
248
JPM Vol 27. No. 4 July 1%2:245-250
T
14-
12-
=lSEM
10 -
8-
Figure 2. Changes in Reizindex with dose of aspirin. Error bars represent standard error of the mean.
6-
40
= -1.18+0.12x
R=0.999
80
100
60
120
Dose of Aspirin (mg/Kg)
70
60
with Cu(I1) (as copper sulfate) have an RI of approximately the same magnitude as those of copper aspirinate-treated animals. The group of animals tested with a mixture of copper(I1) and aspirin exhibit larger (borderline) significance values than does copper aspirinate . The use of the Reizindex was expected to yield results that would be more amenable to interpretation. Lucker and coworkers (Ganote and Lucker, 1982; Laule et al., 1982; Bruhn et al., 1983) proposed that the RI represents a more reliable estimate of mucosal damage than the simple change in GPD. In the experiments presented here, if the A PD,,, comparisons were significantly different, then RI was also. Of 30 comparisons made, RI actually produced four more test comparisons that were significantly different than did A PD,,, (Table 2).
/-J ASA
Cu,ASA,
fl
n cuso, q
ASA + CuSO,
Discussion Low
dose
High dose
Dose of Test compound (mg/Kg)
Figure 3. Summary of Reizindex values for the four test mixtures. Aspirin-containing doses have either 5.55 (low) or 55.5 (high) wmol of aspirin/kg body weight. Copper-containing systems have either 2.78 or 27.8 p,mol of Cu(II)/kg body weight. Error bars represent standard error of the mean. Nthe number of animals in the group and is shown on each bar.
A dose response with more than adequate sensitivity was observed for increasing doses of aspirin as shown in Figure 2. The best fit for these data was obtained with a linear plot of dose versus response (RI). The mean values of RI were tested for significance (Student’s t test; significance level p < 0.05). RI for both low- and high-dose ASA were significantly less than all copper(containing systems, but the copper(II)-containing systems were not significantly dif-
ALICH ET AL. GASTRIC MUCOSAL
249 DAMAGE
DUE TO ASPIRIN AND COPPER ASPIRINATE
Table 2. Test Comparisons
in which the Reizindex Produced a Significant Difference But APD,,, Did Not Treatment Comparison” 1 compared 1 compared 6 compared 5 compared
with with with with
2 3 4 6
Reizindex (mV2 X min)
p
0.014 0.048 0.034 0.006
0.01 0.05 0.04 0.01
A PD,,, @VI
0.168 0.614 0.808 0.214
P
0.15 0.66 0.81 0.27
This research was supported in part by grants from Miller-Dwan Medical Center Foundation, Duluth Clinic Education and Research Foundation, the College of St. Scholastica Faculty Development Fund, and Minnesota Medical Foundation, and a 3M Fellowship to Elise M. Rieschl. The authors wish to acknowledge the assistance of Bradley Ingersoll for statistical analysis, and David Nyback for technical assistance.
a 1: high dose of aspirin; 2: low dose of copper aspirinate; 3: low dose of copper sulfate; 4: high dose of copper sulfate; 5: low dose of copper sulfate plus aspirin; and 6: high dose of copper sulfate plus aspirin.
References Alich AA, Wittmers LE, Jr. (1980) Hydrolysis of tetrakis-p,-acetylsalicylato-dicopper(I1). J Pharm Sci 69~125-127.
ferent from each other. The mixture of copper(U) and aspirin yielded the largest RI values, approximating an additive response for the two agents. The similarity between values for copper aspirinate on the mixtures of aspirin and copper sulfate led to the conclusion that hydrolysis of the complex was not a significant factor under the conditions of this experiment. RI of copper(I1) alone was as large as that of copper aspirinate. If the change in gastric mucosal potential is interpreted as indicating mucosal damage, then this was a surprising result, as it has been previously shown (Alich et al., 1983) by both lesion assessment using Guth’s method of visual lesion assessment (Guth, 1973) and microscopic histological examination that copper(I1) alone does not damage gastric mucosa. Because the system containing copper(H) with no aspirin at all had an RI as large as the other copper-containing systems with aspirin, we concluded that factors other than, or in addition to, mucosal damage are involved in the change in gastric potential difference in systems containing copper(I1). The use of the RI produced slightly improved results (over the use of APD,,, alone). It appears that for animals not showing recovery of GPD during the experiment, that is, those with more serious damage, the inclusion of the time factor yields a more complete picture of damage. Thus, for experiments yielding severe mucosa damage, the RI calculations may be preferred. Where damage is not serious, the APD,,, would take less time and yield just as acceptable results. In summary, whereas Reizindex is a reliable mucosal damage indicator, with an adequate dose response for aspirin, the same is not true for copper(II)containing systems. Factors other than mucosal damage cause changes in gastric potential difference in systems containing copper(I1). Furthermore, the use of the RI in data analysis results in a slight improvement over the use of APD,,, alone. Finally, hydrolysis of copper aspirinate does not occur to a significant extent during these experiments.
Alich AA, Welsh VJ, Wittmers LE, Jr. (1983) Comparison of aspirin and copper aspirinate with respect to gastric mucosal damage in the rat. J Pharm Sci 72: 1457-1461. Baskin WN, Ivey KJ, Krause WJ, Jeffrey GE, Gemmell RT (1976) Aspirin-induced ultrastructural changes in human gastric mucosa. Ann Intern Med 85~299-303. Brodie DA, Tate CL, Hooke KF (1970) Aspirin: Intestinal damage in rats. Science 170:183-185. Bruhn R, Ganote DP, Lucker PW, Vance J, Procaccini RL (1983) Comparative gastric irritation by two buffered acetylsalicylic acid formulations and acetysalicylic acid. ArzneimittelForschung 33: 158-162.
Caspary WF (1975) Einfluss von Aspirin, Antacida, Alkohol, und Gallensluren auf die transmurale elektrische Potentialdifferenz des menschlichen Magens. Dtsch Med Wochenschr 100: 1263-1268. Cummins JT, Vaughn BE (1965) Ionic relationships of the bioelectrogenie mechanism in isolated rat stomach. Biochem Biophys Acta 94~280-292.
Davenport HW (1964) Functional significance of gastric mucosal barrier to sodium. Gastroenterology 47: 142-152. Flemstrom G, Marsden NVB (1974) Increased inulin absorption from the cat stomach exposed to acetylsalicylic acid. Acta Physiol Stand
92:517-525.
Ganote DP, Lucker PW (1982) A new mathematical approach to describe gastric potential difference changes. Arzneim Forsch Drug Res 321552-556.
Gamer A, (1978) Mechanisms of action of aspirin on the gastric mucosa of the guinea pig. Acta Physiol Stand (Special Suppl.): 101-110. Geall MG, Phillips SF, Summer&ill WHJ (1970) Profile of gastric potential difference in man. Gastroenterology 58:437-443. Guth PH (1973) Experimental production of peptic ulcer. Gastroenterology 64:1187-1189. Kishore V (1988) Antiinflammatory activities of copper-aspirinate and aspirin in adjuvant arthritic rats. Res Comm Chem Pathol Pharmacol60:257-260.
Laule H, Lucker PW, Altmayer P, Eldon MA (1982) Gastric potential difference as a model in clinical pharmacology. Eur J Cfin Pharmaco/22:147-151. Lussier A, Arsenault A, Varady J (1978) Gastrointestinal microbleeding after aspirin and naproxen. C/in Pharmacol Ther 23: 402-407.
Malling T (1991) Use of radioactive inulin and sucrose in the assessment of gastric mucosal damage. Abstracts: 32nd Annual National Student Research Forum, Galveston, TX, April 11-13, pp. 14.
250
JPM Vol27. No. 4 July 19!32:245-250
McGahan M (1990) Copper and aspirin treatment increase the antioxidant activity of plasma. Agents Actions 31:59-64.
Sorenson JRJ (1976) Copper chelates as possible active forms of the antiarthritic agents. J Med Chem 19:135-148.
McGreevy JM (1984) Gastric surface cell function: Potential difference and mucosal barrier. Am J Physiol247:G79-G87.
Takeuchi K, Nishiwaki H, Dsano H, Ebara S, Okabe S (1988) Repair of mucosal damage induced by ethanol in the rat stomach. Digestion 40: l-10.
Rainsford KD, Whitehouse MW (1976) Concerning the merits of copper aspirin as a potential antiinflammatory drug. J Pharm Pharmaco/ 28:83-86. Rainsford KD, Whitehouse MW (1976) Gastric mucus effusipn elicited by oral copper compounds: Potential antiulcer activity. Experientia 32:1172-1173.
Tamawski A, Ivey KJ (1978) Transmucosal potential difference protile in rat upper gastrointestinal tract. Can J Physiol Pharmacol 56:471-473.
Robins PG (1980) Ultrastructural observations on the pathogenesis of aspirin induced gastric erosions. Br .I Exp Pathol61:497-504.
Vance J, Lucker PW, Tilling W, Procaccini R, Wetzelsbergen N (1982) The transmural gastric potential difference in combination with pharmacokinetic profiles. Methods Find Exp Clin Pharmaco1 4533-538.
Sernka TJ, Hogben CAM (1969) Active ion transport by isolated gastric mucosa of rat and guinea pig. Am J Physiol 215: 1419-1423.
Wittmers JR LE, Anderson LA, Fall MM, Alich AA (1990) Intragastric inulin as a measure of mucosal damage caused by aspirin. .I Pharmacol Methods 241229-239.
