Seriate Histomorphometry of Whole Rat Stomach: An Accurate and Reliable Method for Quantitative Analysis of Mucosal Damage

Toxicology and Applied Pharmacology 174, 17–26 (2001) doi:10.1006/taap.2001.9193, available online at http://www.idealibrary.com on

Seriate Histomorphometry of Whole Rat Stomach: An Accurate and Reliable Method for Quantitative Analysis of Mucosal Damage Gianfranco Natale,* Gloria Lazzeri,† Corrado Blandizzi,† Giorgio Gherardi,‡ Paola Lenzi,* Antonio Pellegrini,* and Mario Del Tacca† *Department of Human Morphology and Applied Biology and †Department of Oncology, Transplants and Advanced Technologies in Medicine, University of Pisa, Via Roma 55, I-56126 Pisa, Italy; and ‡Department of Pathology, Fatebenefratelli Hospital, Corso di Portanuova 23, I-20121 Milan, Italy Received January 20, 2001; accepted April 12, 2001

observation and measurement of gastric lesions is often performed with the naked eye or with a magnifying glass (Yamamoto et al., 1986; Konturek et al., 1993; Hamajima et al., 1994; Kim and Hong, 1995), under a dissecting stereomicroscope with or without square grids or an ocular micrometer (Okabe and Kunimi, 1981; Matsuo et al., 1986; Esplugues et al., 1988; Matsumoto et al., 1992), by means of computerassisted planimetric analysis (Romano et al., 1992; Brzozowski et al., 1993), or by transmission densitometry (Witt et al., 1985). Once detected, gastric injury is quantified by scoring the lesions according to a scale of increasing severity, from 0 (no lesion) to the highest grade of damage. Different criteria are adopted to establish the various grades of severity. Chromatic variations (presence or not of hemorrhagic extravasation, petechiae, or hyperemia) (Schiantarelli et al., 1984; Wallace and Whittle, 1985; Kim and Hong, 1995), shape (punctiform, linear, or circular lesions) (Schiantarelli et al., 1984; Esplugues et al., 1988), and size (length of linear injuries, diameter of roundish lesions, and area of damaged mucosa) (Schiantarelli et al., 1984; Yamamoto et al., 1986; Matsuo et al., 1986; Romano et al., 1992; Brzozowski et al., 1993) are the most common evaluation elements. In some studies, different criteria are often combined to characterize the grades of severity, and various scales are usually adopted in the same study to quantify the lesions induced by ulcerogenic agents (Schiantarelli et al., 1984; Wallace and Whittle, 1985; Konturek et al., 1993). The damage is then expressed as the “ulcer index”, which may be represented by the absolute value of the score (for instance, the sum of linear length of all lesions) or by the percentage ratio between damaged and total mucosa (for instance, the area of damaged mucosa divided by the total area of glandular mucosa). However, when using such approaches, the estimation of gastric damage may be affected by the subjectivity of the examiner and quantitative analysis is possible only in terms of surface extension. For instance, chromatic variations due to hemorrhagic extravasation or marked hyperemia may be erroneously interpreted as necrotic lesions and apparently uninjured areas of mucosa may contain microscopic but deep and severe lesions that remain undetec-

Seriate Histomorphometry of Whole Rat Stomach: An Accurate and Reliable Method for Quantitative Analysis of Mucosal Damage. Natale, G., Lazzeri, G., Blandizzi, C., Gherardi, G., Lenzi, P., Pellegrini, A., and Del Tacca, M. (2001). Toxicol. Appl. Pharmacol. 174, 17–26. The evaluation of mucosal damage in experimental models of gastric injury is commonly based on macroscopic detection of gross lesions and/or histological examination of tissue samples and is limited by the subjectivity of the examiner and by the paucity of nonrepresentative samples. This study proposes a novel method for the histomorphometric analysis of gastric damage, based on the examination of seriate parallel strips taken from whole rat stomachs. Strips were cut perpendicular to the lesser curvature, placed on a glass slide, with the side of each strip distal to the pylorus upward, and processed for routine histology. Sections were then observed by light microscopy: the length of damaged mucosa divided by the total length of mucosa, measured on a micrometric scale and expressed in percentage values, was indicated as the lesion index. Furthermore, to evaluate the severity of the damage, three types of lesions were discriminated depending on their depth: type I, lysis of luminal cells; type II, damage involving the cells lying on both surface mucosa and gastric pits; and type III, damage involving the lower part of the lamina propria with injury of glands associated with detachment of whole mucosal layers. Three models of acute gastric damage (ethanol, hemorrhagic shock, and indomethacin) were examined and treatment was also carried out with the antiulcer drugs omeprazole, ranitidine, and misoprostol, to show the advantages of this histomorphometric approach. The results indicate that this method allows an accurate quantitative analysis of gastric damage, and the effects of different antiulcer drugs can be better discriminated. © 2001 Academic Press Key Words: histomorphometry; gastric mucosa; gastric damage; ethanol; hemorrhagic shock; indomethacin; rat.

In several studies reporting different models of experimentally induced gastric injury, the evaluation of mucosal damage is mainly based on the macroscopic detection of gross lesions by a direct inspection of the gastric mucosal surface. The 17

0041-008X/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

18

NATALE ET AL.

ted. Furthermore, the different grading scales adopted do not allow an appropriate comparison between the studies. An accurate histological evaluation of the gastric mucosa represents the first step for an objective examination of gastric injury. In this case, gastric mucosal strips or samples are processed for histology (Ito and Lacy, 1985; Wallace and Whittle, 1985; Morini et al., 1995, 1997) and gastric lesions can be appropriately described. However, the damage does not necessarily occur to the same extent and with the same severity within the gastric mucosa, and ulcerogenic agents can promote different patterns of mucosal injury. Therefore, the number and sites of gastric mucosal samples may not be representative of overall gastric damage. In the present study, we propose a histomorphometric approach to the analysis of whole stomachs through the examination of seriate histological sections. This method takes into account both the extent (Willems et al., 1976) and the depth of gastric lesions, allowing an accurate quantitative estimation of mucosal injury, as well as a detailed description of morphological changes occurring within the gastric wall after the induction of mucosal damage. Three representative models of acute gastric damage were used (absolute ethanol, hemorrhagic shock, and indomethacin) (Glavin and Szabo, 1992) and treatment was also carried out with three antiulcer drugs (misoprostol, ranitidine, and omeprazole) in order to validate and analyze the advantages of this histomorphometric approach. MATERIALS AND METHODS Animals. Albino male Wistar rats, 200 –220 g body wt, were used throughout the study. They were fed standard laboratory chow and tap water ad libitum and were not used for at least 1 week after their delivery to the laboratory. The animals were housed, six in a cage, in temperature-controlled rooms on a 12-h light cycle at 22–24°C and 50 – 60% humidity. The study was performed in accordance with the Guiding Principles in the Use of Animals in Toxicology. Twenty-four hours before the experiments, animals were deprived of food and maintained in single cages provided with wire net bottoms to prevent coprophagy. Free access to water ad libitum was allowed until 1 h before the beginning of experiments. Induction of gastric mucosal damage by absolute ethanol. Mucosal damage was induced with 1 mL/200 g body wt of absolute ethanol administered by intragastric gavage using a polyethylene orogastric catheter. In this model of gastric injury, the mucosa was also examined at different times after damage induction, in order to evaluate the time course of the lesions: at intervals of 1, 2, 4, 8, 12, 24, or 48 h after the injection of the necrotizing agent, animals were euthanized by cervical dislocation and their stomachs were rapidly removed and processed for histomorphometric evaluation of mucosal damage. Control animals received 1 mL of saline. Induction of gastric mucosal damage by hemorrhagic shock. At the time of the experiment, animals were anesthetized with urethane 1.2 g/kg ip. The trachea was exposed and cannulated to ensure a patent airway. A polyethylene catheter was introduced into the esophagus and advanced as far as 5 mm beyond the gastroesophageal junction. Following a midline incision of the abdomen, a second catheter was introduced into the duodenum and pushed forward until its tip was about 5 mm beyond the pylorus. The stomach lumen was perfused continuously with 0.1 N HCl solution (pH 1.5) at 37°C, at a rate of 1 mL/min. Rectal temperature was monitored and maintained between 37 and 39°C with an infrared lamp. The right carotid artery was carefully exposed

at the cervical level and isolated from the vagus nerve. The cranial portion of the carotid was occluded by a ligature, while the caudal one was cannulated with a thin polyethylene tube, both for monitoring of blood pressure and for removal of blood. Blood pressure was recorded using a pressure transducer (Bentley Trantec, model 800, Basile, Italy) connected to a microdynamometer (Unirecords, Basile, Italy). After 30 min, to allow stabilization of blood pressure, hemorrhagic shock was induced by subtracting 3 mL blood/200 g body wt in 1 min. The blood was collected in a plastic test tube containing 30 IU/mL heparin and was maintained at 37°C. After 40 min, the blood was reinfused through the femoral vein in 10 min, and 10 min after blood reinfusion the animal was euthanized. The stomach was quickly removed and used for histomorphometric evaluation of mucosal damage. Control animals underwent the same procedure apart from blood subtraction, and their stomachs were perfused with saline. Induction of gastric mucosal damage by indomethacin. Gastric mucosal damage was induced with 20 mg/kg indomethacin administered by intragastric gavage using a polyethylene orogastric catheter. Four hours after the administration of indomethacin, animals were euthanized by cervical dislocation, and their stomachs were rapidly removed and processed for histomorphometric evaluation of mucosal damage. Control animals received 1 mL of indomethacin vehicle. Drug treatment. The effects of drug treatment were tested on absolute ethanol-induced gastric damage. Omeprazole (30 mg/kg), misoprostol (300 ␮g/kg), ranitidine (30 mg/kg), or their vehicles, were administered by the intraduodenal (id) route 30 min before the induction of gastric mucosal damage. At the time of the experiment, during a short anesthesia with diethyl ether, the abdomen was opened by a midline laparatomy and the duodenum was exposed. Omeprazole, misoprostol, ranitidine, or their vehicles then were directly injected into the distal portion of the duodenum by means of a 25-gauge needle, the abdominal incision was closed with clips, and animals were allowed 10 min to recover from anesthesia. Thirty minutes later, gastric mucosal damage was induced with absolute ethanol administered by intragastric gavage using a polyethylene orogastric catheter. One hour after the administration of ethanol, rats were euthanized by cervical dislocation, and their stomachs were rapidly removed and processed for quantitative evaluation of mucosal damage. Doses of omeprazole, misoprostol, and ranitidine were selected on the basis of previous studies dealing with gastroprotection against necrotizing agents in rats (Bauer, 1985; Okabe et al., 1988; Shiratsuchi et al., 1988). Histomorphometric evaluation of gastric mucosal damage. Each stomach was opened along the greater curvature, gently washed with saline (154 mM NaCl), pinned upon a cork plate with the mucosal surface turned upward (Figs. 1A and 1B), and fixed in 10% formalin buffered with phosphate for 24 h at 4°C. The glandular portion of the stomach was then dissected into 2-mmwide parallel strips, perpendicular to the lesser curvature. The strips obtained from each stomach were sequentially placed on a glass slide and oriented with the side of each strip distal to the pylorus upward. In order to maintain this arrangement, a solution of melted 3% agar was gently poured over the strips and quickly cooled at 4°C to induce solidification. The agar block was then removed from the glass slide, introduced into a tissue-embedding cassette for histology, and finally dehydrated and embedded in paraffin wax (Vogel HistoComp, Giessen, Germany). Three-micrometer-thick paraffin sections were cut using a microtome HM 330 Microm (Heidelberg, Germany) and stained with H & E (Fig. 1C). Sections were observed by light microscopy and the length of the luminal border of both total and damaged mucosa was evaluated by means of an ocular micrometric scale (Periplan GF 12.5⫻/18 M equipped with a grid subdivided into 200 units; Leitz, Wetzlar, Germany), which could be rotated to be adapted to the undulating mucosa. From each block, three sections were taken at different depths, in order to detect possible lesions lying within the 2 mm of thickness of the strips. Three sections of gastric strips were then examined for each rat, leading to a more accurate histological reading. The histomorphometric analysis was performed by two independent observers (G.N. and G.G.) who were unaware of the model of gastric damage and of the treatment received by animals. The lesion index (total damage) was estimated

SERIATE HISTOMORPHOMETRY OF WHOLE RAT STOMACH

19

as the length of damaged mucosa divided by the total length of mucosa (fractional damage) and expressed as percentage values. The effects of drugs were expressed as percentages of mucosal damage measured in rats treated with absolute ethanol. Furthermore, taking into account the depth of the mucosal damage, three types of lesions were also discriminated. Type I lesion consists of lysis of mucosal cells on the luminal free surface, gastric pit cells being undamaged. Type II lesion consists of damage confined to the upper part of the lamina propria involving the cells lying on both surface mucosa and gastric pits, gastric gland cells being undamaged. Type III lesion consists of damage that involves the lower part of the lamina propria as well, with injury of gastric glands associated with detachment of whole layers of necrotic superficial mucosa. In absolute ethanol-treated animals the distribution of both number and length of total damage and type I, II, and III lesions within the gastric corpus and the antrum was also evaluated. Furthermore, when examining the distribution of gastric damage, four types of lesions were discriminated, depending on their length: less than 1 mm, 1 to 2 mm, 2 to 4 mm, and more than 4 mm. Data analysis. Results are given as means ⫾ SEM. The statistical significance of data was evaluated by Student’s t test for unpaired data and p values lower than 0.05 were considered significant: n indicates the number or animals. All statistical procedures were performed by means of personal computer programs (InStat for MacIntosh, Version 2.01). Drugs. The following drugs and reagents were used: omeprazole (Astra, Milan, Italy); ranitidine (Guidotti, Pisa, Italy); misoprostol (Schiapparelli– Searle, Turin, Italy); indomethacin; urethane ethyl carbamate; heparin (Sigma Chemicals Co., St. Louis, MO); diethyl ether (Mallinckrodt Baker BV, Deventer, The Netherlands). Other reagents were of analytical grade. Omeprazole, misoprostol, and ranitidine were suspended in 1% methocel and administered in a volume of 0.5 mL. Indomethacin was dissolved in 5% sodium bicarbonate and administered in a volume of 1 mL.

RESULTS

Histomorphometric Evaluation of Gastric Mucosal Damage on Rats Treated with Absolute Ethanol

FIG. 1. (A) Gross appearance of the gastric mucosa from an untreated rat. The stomach was opened along the greater curvature and pinned upon a cork plate with the mucosal surface turned upward. The mucosa does not present evident lesions. The upper part of the stomach, in white, is called the forestomach and does not contain gastric glands. The lower glandular part of the stomach can be divided into the antrum (within dotted lines) and the corpus. (B) Gross appearance of gastric mucosa from a rat 1 h after the intragastric injection of absolute ethanol. Macroscopic mucosal lesions of different sizes can be recognized clearly as dark areas. (C) Histological slide with gastric strips taken from a rat 1 h after the intragastric injection of absolute ethanol.

In untreated animals injected with saline, gastric mucosa appeared undamaged (Fig. 1A). One, 2, 4, 8, 12, 24, or 48 h after the intragastric injection of saline, the histomorphometric analysis showed a negligible amount of lesions or no damage. Injection of absolute ethanol into the gastric lumen induced gross lesions in the glandular part of the stomach (Fig. 1B). Histological examination of gastric mucosa showed lesions that varied in extent and depth (Figs. 2A–C). Epithelial cells appeared vacuolated with pyknotic nuclei and had a lightly stained cytoplasm. Cellular lysis induced exfoliation of either a few glandular cells or entire portions of mucosa. A marked edema was evident in the underlying submucosal layer, as well as in the exfoliated portions of mucosa, which appeared to be swollen and detached. Histomorphometric analysis showed that, 1 h after intragastric injection of absolute ethanol, total damage included 12.84 ⫾ 1.12% of gastric mucosa (n ⫽ 8)

Gastric strips are serially disposed from cardias (top) to pylorus (bottom). The thinner antral mucosa, just beside the lesser curvature, lies within the two dotted lines. Some type III lesions are evident to the naked eye in the corpus, close to the greater curvature (arrows). Some diffuse exfoliated material is present in the antrum, around the lesser curvature (arrowheads).

20

NATALE ET AL.

FIG. 3. Total damage (A) and type I, type II, and type III lesions (B) of gastric mucosa induced by intragastric injection of absolute ethanol in conscious rats. Gastric mucosal damage was assessed 1, 2, 4, 8, 12, 24, or 48 h after absolute ethanol administration. Data are expressed as percentage damage length out of the total mucosal length. Each column indicates the mean value obtained from eight animals ⫾ SEM (vertical lines).

(Figs. 3A and 4A), with type I, type II, and type III lesions accounting for 4.66 ⫾ 0.57, 4.23 ⫾ 0.46, and 3.95 ⫾ 0.44% of gastric mucosa, respectively (Figs. 3B and 4B). Subsequently, total damage gradually declined, and 48 h after intragastric injection of absolute ethanol it accounted for 2.86 ⫾ 0.72% (n ⫽ 8) (Figs. 3A and 3B). When the distribution of the lesion number was evaluated at the level of the corpus and the antrum, lesions were found more frequently in the corpus (70.6%) and consisted mainly of type I lesions (about 50%). The incidence of type III lesions was higher in the antrum (about 40%). Considering the whole stomach, the number of type I lesions appeared to be slightly higher than types II and III. Extensive lesions (more than 4 mm) did not seem to occur very frequently (6.1%), whereas small lesions (⬍1 mm) accounted for 48.4% of the total

FIG. 2. Histological appearance of gastric damage from a rat 1 h after the intragastric injection of absolute ethanol. (A) In type I lesions, the damage occurs in the superficial epithelium with detachment of a few cells. (B) In type II lesions, the damage is limited to gastric pits with detachment of the surface

epithelium. (C) In type III lesions, the injury involves the gastric glands with swelling and whole-thickness necrosis of gastric mucosa. All histological pictures refer to H & E-stained sections of rat gastric mucosa. Original magnification 80⫻.

SERIATE HISTOMORPHOMETRY OF WHOLE RAT STOMACH

21

tomorphometric analysis revealed that, 1 h after the induction of hemorrhagic shock, total necrotic damage included 3.86 ⫾ 0.22% of gastric mucosa (n ⫽ 8) (Fig. 4A), with type I, type II, and type III lesions accounting for 0.82 ⫾ 0.09, 2.20 ⫾ 0.19, and 0.84 ⫾ 0.05% of gastric mucosa, respectively (Fig. 4B). Histomorphometric Evaluation of Gastric Mucosal Damage in Rats Treated with Indomethacin In untreated animals injected with indomethacin vehicle, histomorphometric analysis showed a negligible amount of lesions or no damage. Injection of 20 mg/kg indomethacin into the gastric lumen induced gross lesions in the glandular part of the stomach (not shown). Histological examination of gastric mucosa showed lesions that varied in extent and depth (Figs. 6A–C). Necrotic damage appeared as a homogeneous area where the glandular architecture was completely lost. In the most severe lesions, the submucosal layer showed signs of intense inflammation with hyperemia, edema, and cell infiltration. Histomorphometric analysis showed that, 4 h after intragastric injection of indomethacin, total damage included 10.23 ⫾ 1.27% of gastric mucosa (n ⫽ 8) (Fig. 4A), with type I, type II, and type III lesions accounting for 4.12 ⫾ 0.81, 3.18 ⫾ 0.62, and 2.93 ⫾ 0.79% of gastric mucosa, respectively (Fig. 4B). Effects of Drug Treatment on Gastric Damage Induced by Absolute Ethanol FIG. 4. Total damage (A) and type I, type II, and type III lesions (B) of gastric mucosa induced by intragastric injection of absolute ethanol or indomethacin in conscious rats and hemorrhagic shock in anesthetized rats. Data are expressed as percentage damage length out of the total mucosal length. Each column indicates the mean value obtained from eight animals ⫾ SEM (vertical lines). *Significant difference from absolute ethanol and indomethacin, p ⬍ 0.05.

number (Table 1). Similarly, when examining the distribution of the lesion length, the corpus exhibited about two thirds of the total damage (68.54%), with type I and type II lesions mostly represented. On the contrary, type III lesions were found mainly in the antrum (19.23%). Furthermore, in this part of the stomach the longest type III lesions could be detected (Table 2). Histomorphometric Evaluation of Gastric Mucosal Damage in Rats Subjected to Hemorrhagic Shock In control animals, histomorphometric analysis showed a negligible amount of lesions or no damage. Hemorrhagic shock induced gross lesions in the glandular part of the stomach (not shown). Histological examination of gastric mucosa showed lesions that varied in extent and depth (Figs. 5A–C). The main histological finding was the presence of necrosis associated with hemorrhage and colliquation. His-

Pretreatment with omeprazole (30 mg/kg id; n ⫽ 8) caused a significant inhibition of ethanol-induced mucosal damage (Fig. 7A). The degree of protection provided by omeprazole was more pronounced toward type III lesions than type I and type II lesions (Fig. 7B). Similarly, misoprostol (300 ␮g/kg id; n ⫽ 8) markedly reduced the ethanol-induced total damage (Fig. 7A). The degree of protection provided by misoprostol was more pronounced toward type II and type III lesions than type I lesions (Fig. 7B). By contrast, ranitidine (30 mg/kg id; n ⫽ 8) failed to significantly affect the damage evoked by absolute ethanol (Figs. 7A and 7B). DISCUSSION

The present results show that morphometric analysis of seriate histological sections allows a quantitative estimation of both the extent and the depth of mucosal damage in the whole stomach. With this method, the histological discrimination between necrotic and nonnecrotic lesions, the microscopic examination of lesions involving either surface epithelium or gastric pits, as well as the detection of other qualitative features of the damage, do not simply result from the evaluation of limited samples of gastric mucosa, but they rather reflect the overall injury occurring in the stomach. This histomorphomet-

22

NATALE ET AL.

TABLE 1 Distribution of the Lesion Number in Rat Gastric Mucosa Lesion length Corpus ⬍1 mm 1 to 2 mm 2 to 4 mm ⬎4 mm Antrum ⬍1 mm 1 to 2 mm 2 to 4 mm ⬎4 mm Whole stomach ⬍1 mm 1 to 2 mm 2 to 4 mm ⬎4 mm

Type I lesion

Type II lesion

Type III lesion

Total damage

4.4 ⫾ 1.2 (11.8 ⫾ 3.2) 4.4 ⫾ 0.9 (11.8 ⫾ 3.2) 3.0 ⫾ 0.7 (8.0 ⫾ 1.9) 1.4 ⫾ 0.7 (3.7 ⫾ 1.9) 13.1 ⫾ 1.8 (35.0 ⫾ 4.8)

6.4 ⫾ 1.6 (17.11 ⫾ 4.3) 1.9 ⫾ 0.5 (5.1 ⫾ 1.3) 0.6 ⫾ 0.3 (1.6 ⫾ 0.8) 0 (0) 8.9 ⫾ 1.9 (23.8 ⫾ 5.1)

1.4 ⫾ 0.5 (3.7 ⫾ 1.3) 2.1 ⫾ 0.6 (5.6 ⫾ 1.6) 0.8 ⫾ 0.4 (2.1 ⫾ 1.1) 0.1 ⫾ 0.1 (0.3 ⫾ 0.3) 4.4 ⫾ 1.2 (11.8 ⫾ 3.2)

12.1 ⫾ 2.9 (32.4 ⫾ 7.8) 8.4 ⫾ 1.4 (22.5 ⫾ 3.7) 4.4 ⫾ 0.9 (11.8 ⫾ 3.2) 1.5 ⫾ 0.7 (4.0 ⫾ 1.9) 26.4 ⫾ 4.5 (70.6 ⫾ 12.0)

0.9 ⫾ 0.5 (2.4 ⫾ 1.3) 0.6 ⫾ 0.3 (1.6 ⫾ 0.8) 0.3 ⫾ 0.3 (0.8 ⫾ 0.8) 0.3 ⫾ 0.3 (0.8 ⫾ 0.8) 2.0 ⫾ 0.7 (5.4 ⫾ 1.9)

1.7 ⫾ 0.5 (4.5 ⫾ 1.3) 0.8 ⫾ 0.3 (2.1 ⫾ 0.8) 0 (0) 0 (0) 2.5 ⫾ 0.6 (6.7 ⫾ 1.6)

3.4 ⫾ 0.8 (9.1 ⫾ 2.1) 1.5 ⫾ 0.5 (4.0 ⫾ 1.3) 1.1 ⫾ 0.4 (2.9 ⫾ 1.1) 0.5 ⫾ 0.2 (1.3 ⫾ 0.5) 6.5 ⫾ 1.3 (17.3 ⫾ 3.5)

6.0 ⫾ 1.6 (16.0 ⫾ 4.3) 2.9 ⫾ 0.6 (7.8 ⫾ 1.6) 1.4 ⫾ 0.4 (3.7 ⫾ 1.1) 0.8 ⫾ 0.3 (2.1 ⫾ 0.8) 11.0 ⫾ 1.9 (29.4 ⫾ 5.1)

5.3 ⫾ 1.3 (14.2 ⫾ 3.2) 5.0 ⫾ 1.2 (13.4 ⫾ 3.2) 3.3 ⫾ 0.8 (8.8 ⫾ 2.1) 1.7 ⫾ 0.7 (4.5 ⫾ 1.9) 15.1 ⫾ 2.1 (40.4 ⫾ 5.6)

8.1 ⫾ 1.9 (21.7 ⫾ 5.1) 2.7 ⫾ 0.8 (7.2 ⫾ 2.1) 0.6 ⫾ 0.2 (1.6 ⫾ 0.5) 0 (0) 11.4 ⫾ 2.0 (30.5 ⫾ 5.3)

4.8 ⫾ 1.1 (12.8 ⫾ 2.9) 3.6 ⫾ 0.9 (9.6 ⫾ 3.2) 1.9 ⫾ 0.7 (5.1 ⫾ 1.9) 0.6 ⫾ 0.1 (1.6 ⫾ 0.3) 10.9 ⫾ 2.0 (29.1 ⫾ 5.3)

18.1 ⫾ 1.7 (48.4 ⫾ 4.7) 11.3 ⫾ 1.4 (30.2 ⫾ 3.7) 5.8 ⫾ 1.1 (15.5 ⫾ 2.9) 2.3 ⫾ 0.9 (6.1 ⫾ 3.2) 37.4 ⫾ 5.1 (100)

Note. Values represent the mean number of lesions (⫾SEM) calculated in eight rats 1 h after injection of absolute ethanol. Values in parentheses represent the percentage of lesions out of the total number of lesions assumed as 100%.

ric approach also allows the detection of fine changes among different models of experimental damage. Among the models of experimental mucosal injury used in the present study, absolute ethanol, hemorrhagic shock, and indomethacin, histomorphometric analysis showed that, in terms of total damage, the highest incidence of mucosal lesions occurred after absolute ethanol administration, whereas the total damage

induced by hemorrhagic shock was one third of that induced by ethanol. Furthermore, the analysis of lesion depth provided interesting information: absolute ethanol induced similar percentages of type I, type II, and type III lesions; type I lesions were slightly prevalent in indomethacin-treated rats; and type II lesions occurred mainly in animals subjected to hemorrhagic shock. In the present study, the extent of ethanol-induced micro-

TABLE 2 Distribution of the Lesion Length in Rat Gastric Mucosa Lesion length Corpus ⬍1 mm 1 to 2 mm 2 to 4 mm ⬎4 mm Antrum ⬍1 mm 1 to 2 mm 2 to 4 mm ⬎4 mm Whole stomach ⬍1 mm 1 to 2 mm 2 to 4 mm ⬎4 mm

Type I lesion

Type II lesion

Type III lesion

Total damage

0.42 ⫾ 0.08 (3.27 ⫾ 0.72) 1.18 ⫾ 0.31 (9.19 ⫾ 1.18) 1.08 ⫾ 0.24 (8.41 ⫾ 1.23) 1.14 ⫾ 0.27 (8.88 ⫾ 1.30) 3.82 ⫾ 0.80 (29.75 ⫾ 3.91)

1.39 ⫾ 0.11 (10.83 ⫾ 1.50) 1.18 ⫾ 0.41 (9.19 ⫾ 1.42) 0.93 ⫾ 0.08 (7.24 ⫾ 1.72) 0 (0) 3.50 ⫾ 0.60 (27.26 ⫾ 3.10)

0.20 ⫾ 0.07 (1.56 ⫾ 0.60) 0.67 ⫾ 0.06 (5.22 ⫾ 0.87) 0.47 ⫾ 0.07 (3.66 ⫾ 0.61) 0.14 ⫾ 0.03 (1.09 ⫾ 0.05) 1.48 ⫾ 0.45 (11.53 ⫾ 2.52)

2.01 ⫾ 0.13 (15.66 ⫾ 1.89) 3.03 ⫾ 0.76 (23.60 ⫾ 2.03) 2.48 ⫾ 0.88 (19.31 ⫾ 1.36) 1.28 ⫾ 0.71 (9.97 ⫾ 1.17) 8.80 ⫾ 1.03 (68.54 ⫾ 4.57)

0.13 ⫾ 0.04 (1.01 ⫾ 0.09) 0.34 ⫾ 0.05 (2.65 ⫾ 0.15) 0.13 ⫾ 0.03 (1.01 ⫾ 0.22) 0.24 ⫾ 0.06 (1.87 ⫾ 0.11) 0.84 ⫾ 0.06 (6.54 ⫾ 1.93)

0.39 ⫾ 0.07 (3.04 ⫾ 0.28) 0.34 ⫾ 0.06 (2.65 ⫾ 0.41) 0 (0) 0 (0) 0.73 ⫾ 0.12 (5.69 ⫾ 0.97)

0.55 ⫾ 0.04 (4.28 ⫾ 0.73) 0.59 ⫾ 0.05 (4.59 ⫾ 0.92) 0.57 ⫾ 0.07 (4.44 ⫾ 0.30) 0.76 ⫾ 0.08 (5.92 ⫾ 0.71) 2.47 ⫾ 0.67 (19.23 ⫾ 2.31)

1.07 ⫾ 0.12 (8.33 ⫾ 1.16) 1.27 ⫾ 0.29 (9.89 ⫾ 1.27) 0.70 ⫾ 0.08 (5.45 ⫾ 0.86) 1.00 ⫾ 0.28 (7.79 ⫾ 1.07) 4.04 ⫾ 0.78 (31.46 ⫾ 3.18)

0.55 ⫾ 0.07 (4.28 ⫾ 0.45) 1.52 ⫾ 0.09 (11.84 ⫾ 1.03) 1.21 ⫾ 0.50 (9.42 ⫾ 1.10) 1.38 ⫾ 0.09 (10.75 ⫾ 1.71) 4.66 ⫾ 0.57 (36.29 ⫾ 4.53)

1.78 ⫾ 0.08 (13.87 ⫾ 1.45) 1.52 ⫾ 0.11 (11.84 ⫾ 1.22) 0.93 ⫾ 0.07 (7.24 ⫾ 0.5) 0 (0) 4.23 ⫾ 0.46 (32.95 ⫾ 3.83)

0.75 ⫾ 0.04 (5.84 ⫾ 0.29) 1.26 ⫾ 0.12 (9.81 ⫾ 1.16) 1.04 ⫾ 0.14 (8.10 ⫾ 1.56) 0.90 ⫾ 0.07 (7.01 ⫾ 1.47) 3.95 ⫾ 0.44 (30.76 ⫾ 3.23)

3.08 ⫾ 0.31 (23.99 ⫾ 2.15) 4.30 ⫾ 0.67 (33.49 ⫾ 3.77) 3.18 ⫾ 0.86 (24.76 ⫾ 2.78) 2.28 ⫾ 0.08 (17.76 ⫾ 1.76) 12.84 ⫾ 1.12 (100)

Note. Values represent the percentage length of damaged mucosa out of the total length of mucosa (⫾SEM) calculated in eight rats 1 h after injection of absolute ethanol. Values in parentheses represent the percentage length of damaged mucosa out of the fractional total damage assumed as 100%.

SERIATE HISTOMORPHOMETRY OF WHOLE RAT STOMACH

FIG. 5. Histological appearance of gastric damage from a rat subjected to hemorrhagic shock. (A) In type I lesions, the superficial epithelium is damaged with detachment of a few cells and denudation of gastric pits (arrows). (B) In type II lesions, gastric pits are markedly damaged with detachment of the surface epithelium. (C) In type III lesions, the injury involves the gastric glands with necrosis of gastric mucosa. All histological pictures refer to H & E-stained sections of rat gastric mucosa. Original magnification 80⫻.

23

scopic damage (ranging from 2.86 to 12.84 at the various times examined) was somewhat lower than that estimated by Lacy and Ito (1982) under similar experimental conditions. However, it must be noted that, after gross examination of mucosal surface, only two types of gastric samples were examined by these authors: samples from nonnecrotic lesion areas (with 60 – 87% of total damage 30 min after ethanol administration) and samples from necrotic lesions areas (with 100% of total damage 15 min after ethanol administration). Overall, the discrepancy between our results can be explained by taking into account that only a few specimens of damaged mucosa cut from each stomach were examined in the study of Lacy and Ito (1982). Similarly, Guth et al. (1984) found that mucosal damage accounted for 86% after 1 h of absolute ethanol. In this case, the histological evaluation was performed on a single gastric strip taken from the most damaged area of rat gastric mucosa, as assessed by direct examination, a sample that may not be representative of the damage within the whole gastric mucosa. In the study by Tarnawski et al. (1985), the length of damaged mucosa was calculated only on some coded gastric strips on the basis of macroscopic examination. In this report the length of mucosal damage accounted for 87% of total strip length, which is very high in comparison with our data, as well. Furthermore, these authors examined the lesion depth, but only lesions deeper than 0.2 mm were discriminated, accounting for 59%. The extent of hemorrhagic shock-induced microscopic damage was found to be greater in the study by Itoh et al. (1986) (about 80%) in comparison with the present data (3.86%). These authors examined accurately both the total damage and the depth of gastric lesions according to criteria practically identical to the present ones. However, they examined only three gastric strips. Other authors, who induced gastric damage by clamping the left gastric artery, limited their analysis to qualitative descriptions of histological pictures, because their findings were considered “minimum and induced inconsistently” (Kawai et al., 1994). As far as indomethacin-induced gastric damage is concerned, in previous studies the gross injury was mainly assessed by calculating a lesion area (Trevethick et al., 1993; Scheiman et al., 1997) or summing the lengths of all lesions (Santucci et al., 1994). When histological assessment was performed, only a part of the rat stomach was examined: in the study by Morini et al. (1995), a single gastric strip was evaluated and about 100% of damage was found 3 h after indomethacin administration. Despite the high percentage of damage found by the above authors, the actual extent of the damage in the whole stomach is underestimated. Indeed, the apparently undamaged areas were not evaluated. Wallace and Whittle (1985) compared the macroscopically visible damage, which was evaluated planimetrically, with the histological assessment of gastric mucosa and they found that the superficial damage was histologically evident in gastric regions that appeared undamaged by macro-

24

NATALE ET AL.

scopic examination. According to these authors, at least the most severe lesions, histologically detected, can correlate with macroscopic examination. Itoh et al. (1986) also compared gross and histological findings, concluding that the evaluation of gross damage only is misleading. In the present study, the whole stomach was histologically examined and type I lesions, which represent superficial damage and are macroscopically undetectable, accounted for 36.29% of total damage. Furthermore, cutting the stomach into several strips also makes it possible to transform a planimetric measure (area of damaged gastric mucosa) into a linear measure (length and depth of damaged gastric mucosa). Besides quantification of gastric damage, the histological pattern of the lesion must be taken into account. The attribution of a lesion to a particular type of damage is merely due to its size and extent within the gastric mucosa. Nevertheless, the histological pattern of a type III lesion induced by absolute ethanol is different from that of a type III lesion induced by hemorrhagic shock or indomethacin. In the most severe lesions

FIG. 6. Histological appearance of gastric damage from a rat 4 h after the intragastric injection of 20 mg/kg indomethacin. (A) In type I lesions, the surface epithelium is damaged with detachment of a few cells. (B) In type II lesions, gastric pits are markedly damaged with necrosis. (C) In type III lesions, whole-thickness necrosis of gastric mucosa: in this case the glandular architecture is completely lost. All histological pictures refer to H & E-stained sections of rat gastric mucosa. Original magnification 80⫻.

FIG. 7. Effects of omeprazole (30 mg/kg id), misoprostol (300 ␮g/kg id), or ranitidine (30 mg/kg id) on total damage (A) and on type I, type II, and type III lesions (B) of gastric mucosa induced by absolute ethanol in conscious rats. Omeprazole, misoprostol, ranitidine, or their vehicle (control) were given 30 min before the induction of gastric mucosal injury. Gastric mucosal damage was assessed 1 h after absolute ethanol administration. Data are expressed as percentages of damage measured in control rats. Each column indicates the mean value obtained from eight animals ⫾ SEM (vertical lines). *Significant difference from control values, p ⬍ 0.05.

SERIATE HISTOMORPHOMETRY OF WHOLE RAT STOMACH

induced by ethanol, there is necrosis with exfoliated portions of mucosa, which appears to be swollen and detached because of the marked edema. In fact, when the superficial gastric epithelium is exposed to ethanol, both venoconstriction and dilation of arterioles occur with consequent hyperemia, edema, and hemorrhage (Oates and Hakkinen, 1988). In animals subjected to hemorrhagic shock, necrosis with colliquation is the main characteristic. In this case, impairment of gastric blood flow, excess of free oxygen radical generation, and reduction of mucus and bicarbonate secretion seem to be responsible for the damage (Itoh and Guth, 1985). Nonsteroidal antiinflammatory drugs (NSAIDs) possess severe side effects, especially at the gastrointestinal level, with bleeding, ulceration, and perforation, and they are widely used to promote experimental gastric injury. Necrosis with complete destruction of the glandular architecture and submucosal edema was observed under indomethacin treatment. Inhibition of endogenous prostaglandin secretion and ion trapping seem to be the main causes of NSAID-induced gastric damage (Schoen and Vender, 1989). Independently of the histological pattern of gastric damage, the present histomorphometric criteria provide a unique evaluation scale that makes it possible to describe the extent and the depth of the lesions. In this respect, apart from the complete detection of damage in whole stomachs, this approach has other advantages: the gastric injury, quantified according to the same criteria, no longer requires different grading scales (Schiantarelli et al., 1984; Wallace and Whittle, 1985; Konturek et al., 1993) and can be appropriately compared when examining experimental models with different damaging agents or pathophysiological conditions (Glavin and Szabo, 1992). As far as the distribution of damage length after ethanol administration is concerned, the present data show that the three types of lesions are equally represented in the whole stomach. The corpus, which is the largest part of the glandular mucosa, exhibits 68.54% of the total damage, mainly due to type I and type II lesions. On the other hand, the antrum, which is characterized by a thinner mucosa and contains the lesser curvature, shows about two thirds of the most severe type III lesions. In most studies, the lesion lengths are simply summed (Matsumoto et al., 1992; Hamajima et al., 1994) or the grade of damage severity is discriminated (Wallace and Whittle, 1985; Chan et al., 1995), and the final value (ulcer index) does not provide the incidence of these grades of severity. In the present study, the contribution of each lesion to the total damage, in terms of length and depth, can be discriminated. In other studies, only the number of lesions was taken into account: Robert et al. (1985) and Konturek et al. (1993) found a mean number of lesions of 12 and 15, respectively, after intragastric administration of absolute ethanol. The most macroscopically evident type III lesions accounted for a similar total number (10.9) in the present study. Nevertheless, our data show that the distribution of both lesion number and lesion length, although histologically assessed, is not always super-

25

imposable. This means that the simple counting of lesions is not enough to characterize the gastric damage sufficiently. In order to assess the suitability of our morphometric method for pharmacological investigations, in the present study, the effect of omeprazole, misoprostol, and ranitidine were evaluated. The effects of drug treatment on absolute ethanol-induced gastric injury revealed that the benzimidazole derivative omeprazole reduced the total damage significantly. Previous studies also reported a gastroprotective action of this drug as well as of other benzimidazole derivatives, such as timoprazole, pantoprazole, and lansoprazole (Mattsson et al., 1983; Ruwart et al., 1982; Blandizzi et al., 2000). The degree of gastroprotection provided by omeprazole was mainly evident toward type III lesions. It is conceivable that this drug, by strengthening defensive mechanisms, does not allow the occurrence of severe lesions: some type III lesions may be converted into type II lesions and some type II lesions may be converted into type I lesions; then, type I and type II lesions may be more frequent than type III lesions as a sign of gastroprotection. Furthermore, an acceleration of restitution processes by omeprazole may also contribute to this pattern of lesion. The prostaglandin analogue misoprostol also exerted a similar profile of gastroprotection from ethanol-induced damage, with type III lesions markedly reduced. On the other hand, the antisecretive drug ranitidine failed to prevent gastric damage. This finding is in agreement with previous studies indicating that this drug is not effective when the damage is induced by exogenously administered necrotizing agents (Del Soldato et al., 1984). In conclusion, the present findings indicate that histomorphometry of whole gastric mucosa (1) allows an accurate quantification of experimentally induced gastric injury by which the severity of the damage can be appropriately defined and the distribution of lesions within gastric mucosa is easily studied; (2) represents an innovative combination of classic histological techniques that aims to obtain qualitative results superior to the simple lesion counting or sampling; (3) can be applied to other hollow viscera and to all small-sized animals (rats, mice, rabbits, and guinea pigs); and (4) might be suitable for the evaluation of other histological features in whole stomachs: stained and immunostained cells, glands, mucosal thickness, mucus secretion, blood distribution, and immunoreactivity to different antigens. REFERENCES Bauer, R. F. (1985). Misoprostol: Preclinical pharmacology. Dig. Dis. Sci. 30, 118S–125S. Blandizzi, C., Natale, G., Gherardi, G., Lazzeri, G., Marveggio, C., Colucci, R., Carignani, D., and Del Tacca, M. (2000). Gastroprotective effects of pantoprazole against experimental mucosal damage. Fundam. Clin. Pharmacol. 14, 89 –99. Brzozowski, T., Konturek, S. J., Majka, J., Dembinski, A., and Drozdowicz, D. (1993). Epidermal growth factor, polyamines, and prostaglandins in healing of stress-induced gastric lesions in rats. Dig. Dis. Sci. 38, 276 –283.

26

NATALE ET AL.

Chan, C-C., Boyce, S., Brideau, C., Ford-Hutchinson, A. W., Gordon, R., Guay, D., Hill, R. G., Li, C-S., Mancini, J., Penneton, M., Prasit, P., Rasori, R., Riendeau, D., Roy, P., Tagari, P., Vickers, P., Wong, E., and Rodger, I. W. (1995). Pharmacology of a selective cyclooxygenase-2 inhibitor, L-745,337: A novel nonsteroidal anti-inflammatory agent with an ulcerogenic sparing effect in rat and nonhuman primate stomach. J. Pharmacol. Exp. Ther. 274, 1531–1537. Del Soldato, P., Foschi, D., Varin, L., and Daniotti, S. (1984). Comparison of the gastric cytoprotective properties of atropine, ranitidine and PGE2 in rats. Eur. J. Pharmacol. 106, 53–58. Esplugues, J. V., Brage, R., Cortijo, J., Marti-Cabrera, M., Hernandez, G., Sarria, B., and Esplugues, J. (1988). Differential effects of verapamil on various gastric lesions in rats. Pharmacology 36, 69 –72. Glavin, G. B., and Szabo, S. (1992). Experimental gastric mucosal injury: Laboratory models reveal mechanisms of pathogenesis and new therapeutic strategies. FASEB J. 6, 825– 831. Guth, P. H., Paulsen, G., and Nagata, H. (1984). Histologic and microcirculatory changes in alcohol-induced gastric lesions in the rat: Effect of prostaglandin cytoprotection. Gastroenterology 87, 1083–1090. Hamajima, E., Sugiyama, S., Hoshino, H., Goto, H., Tsukamoto, Y., and Ozawa, T. (1994). Effects of FK506, an immunosuppressive agent, on genesis of water-immersion stress-induced gastric lesions in rats. Dig. Dis. Sci. 39, 713–720. Ito, S., and Lacy, E. R. (1985). Morphology of rat gastric mucosal damage, defense, and restitution in the presence of luminal ethanol. Gastroenterology 88, 250 –260. Itoh, M., and Guth, P. H. (1985). Role of oxygen-derived free radicals in hemorrhagic shock-induced gastric lesions in the rat. Gastroenterology 88, 1162–1167. Itoh, M., Paulsen, G., and Guth, P. H. (1986). Hemorrhagic shock and acid gastric injury in the rat. Comparison of gross and histologic findings. Gastroenterology 90, 1103–1110. Kawai, T., Joh, T., Iwata, F., and Itoh, M. (1994). Gastric epithelial damage induced by local ischemia-reperfusion with or without exogenous acid. Am. J. Physiol. 266, G263–G270. Kim, C. D., and Hong, K. I. W. (1995). Preventive effect of rebamipide on gastric lesions induced by ischemia-reperfusion in the rat. J. Pharmacol. Exp. Ther. 275, 340 –344. Konturek, S. J., Brzozowski, T., Majka, J., Szlachcic, A., Bielanski, W., Stachura, J., and Otto, W. (1993). Fibroblast growth factor in gastroprotection and ulcer healing: Interaction with sucralfate. Gut 34, 881– 887. Lacy, E. R., and Ito, S. (1982). Microscopic analysis of ethanol damage to rat gastric mucosa after treatment with a prostaglandin. Gastroenterology 83, 619 – 625. Matsumoto, J., Ueshima, K., Ohuchi, T., Takeuchi, K., and Okabe, S. (1992). Induction of gastric lesions by 2-deoxy-D-glucose in rats following chemical ablation of capsaicin-sensitive sensory neurons. Jpn. J. Pharmacol. 60, 43– 49. Matsuo, Y., Kuwayama, H., Itoh, H., Seki, A., Fukuda, S., Kawasaki, T., Hotta, T., Ohsaka, Y., and Komatsu, K. (1986). Gastric anti-ulcer and cytoprotective properties of 2-(3,4-dimethoxyphenyl)-5-methyl-thiazolidine-4-one in rat experimental ulcers. Arzneim-Forsch. 36, 1236 –1240. Mattsson, H., Andersson, K., and Larsson, H. (1983). Omeprazole provides protection against experimentally induced gastric mucosal lesions. Eur. J. Pharmacol. 91, 111–114. Morini, G., Grandi, D., Arcari, M. L., and Bertaccini, G. (1995). Indomethacininduced morphological changes in the rat gastric mucosa, with or without prior treatment with two proton pump inhibitors. Aliment. Pharmacol. Ther. 9, 615– 623.

Morini, G., Grandi, D., Arcari, M. L., Galanti, G., and Bertaccini, G. (1997). Histological effect of (R)-(⫺)methylhistamine on ethanol damage in rat gastric mucosa. Influence on mucus production. Dig. Dis. Sci. 42, 1020 – 1028. Oates, P. J., and Hakkinen, J. P. (1988). Studies on the mechanism of ethanol-induced gastric damage in rats. Gastroenterology 94, 10 –21. Okabe, S., Akimoto, Y., Yamasaki, S., and Nagai, H. (1988). Effects of NC-1300-B, a new benzimidazole derivative, on hog gastric H ⫹,K ⫹-ATPase, gastric acid secretion and HCl-ethanol-induced gastric lesions in rats. Dig. Dis. Sci. 33, 1425–1434. Okabe, S., and Kunimi, H. (1981). Effects of N-acetyl-L-carnosine alumin (CL-1700) on various acute gastric lesions and gastric secretion in rats. Jpn. J. Pharmacol. 31, 941–950. Robert, A., Lancaster, C., Davis, J. P., Field, S. O., Wickrema Sinha, A. J. W., and Thornburgh, B. A. (1985). Cytoprotection by prostaglandin occurs in spite of penetration of absolute ethanol into the gastric mucosa. Gastroenterology 88, 328 –333. Romano, M., Polk, W. H., Awad, J. A., Arteaga, C. L., Nanney, L. B., Wargovich, M. J., Kraus, E. R., Boland, R., and Coffey, R. J. (1992). Transforming growth factor protection against drug-induced injury to the rat gastric mucosa in vivo. J. Clin. Invest. 90, 2409 –2421. Ruwart, M. J., Lancaster, C., Nezamis, J. E., Davis, J. P., Rush, B. D., Friedle, N. M., and Pugh, C. (1982). Timoprazole: A unique antisecretory and cytoprotective agent. Gastroenterology 82, 1166. Santucci, L., Fiorucci, S., Giansanti, M., Brunori, P. M., Di Matteo, F. M., and Morelli, A. (1994). Pentoxifylline prevents indomethacin induced acute gastric mucosal damage in rats: Role of tumor necrosis factor alpha. Gut 35, 909 –915. Scheiman, J. M., Tillner, A., Pohl, T., Oldenburg, A., Angermu¨ller, S., Go¨rlach, E., Engel, G., Usadel, K. H., and Kusterer, K. (1997). Reduction of non-steroidal anti-inflammatory drug induced injury and leucocyte endothelial adhesion by octreotide. Gut 40, 720 –725. Schiantarelli, P., Cadel, A., and Folco, G. C. (1984). Gastroprotective effects of morniflumate, an esterified anti-inflammatory drug. Arzneim-Forsch. 34, 885– 890. Schoen, R., and Vender, R. (1989). Mechanisms of non-steroidal anti-inflammatory drug-induced gastric damage. Am. J. Med. 86, 449 – 458. Shiratsuchi, K., Fuse, H., Hagiwara, M., Mikami, T., Miyasaka, K., and Sakuma, H. (1988). Cytoprotective action of roxatidine acetate HCl. Arch. Int. Pharmacodyn. Ther. 294, 295–304. Tarnawski, A., Hollander, D., Stachura, J., Krause, W. J., and Gergely, H. (1985). Prostaglandin protection of the gastric mucosa against alcohol injury—A dynamic time-related process. Gastroenterology 88, 334 –352. Trevethick, M. A., Clayton, N. M., Strong, P., and Harman, I. W. (1993). Do infiltrating neutrophils contribute to the pathogenesis of indomethacin induced ulceration of the rat gastric antrum? Gut 34, 156 –160. Wallace, J. L., and Whittle, B., Jr. (1985). Role of prostanoids in the protective actions of BW755C on the gastric mucosa. Eur. J. Pharmacol. 115, 45–52. Willems, G., Nyst, M., Delince´, P., and De Graef, J. (1976). Direct method for estimating the total gastrin cell number in the stomach of rats. Gastroenterology 71, 533–536. Witt, C. G., Will, P. C., and Gaginella, T. S. (1985). Quantification of ethanol-induced gastric mucosal injury by transmission densitometry. J. Pharmacol. Methods 13, 109 –116. Yamamoto, H., Nakamura, Y., Kunoh, Y., Ichihara, K., Nagasaka, M., and Asai, H. (1986). Antiulcer effect of (⫺)-cis-2,3-dihydro-3-(4-methylpiperazinyl-methyl)-2-phenyl-1,5-benzothiazepin-4-(5H)-one-hydrochloride (BTM-1086) in experimental animals. Jpn. J. Pharmacol. 41, 283–292.