role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

6
J Cancer Res Clin Oncol (1989) 115:253-258 J~om~lof Cancer esearch Clinical 9 9 Springer-Verlag 1989 Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat * A. Materia a, G. Silecchia a, E. Spaziani 1, p. Mariani 1, L. Scucchi 2, p. Mingazzini 2, C. Favalli a, E. Garaci a and N. Basso a 1 II Clinica Chirurgica University "La Sapienza", 2 Department of Biopathology University "La Sapienza", 3 Department of Experimental Medicine, University "Tor Vergata", Rome, Italy Summary. The study was initiated to evaluate the ef- fect of N-methyl-N-nitro-N-nitrosoguanidine (NG) on gastric intraluminal prostaglandin release during a 30-day treatment period and to investigate the effect of a stable prostaglandin E a analogue (misoprostol) on NG-induced gastric mucosal damage during the same time period. Samples of gastric juice (1 h) were obtained from 40 male Sprague-Dawley rats with chronic gastric fistulas, in basal conditions and after 5, 15 and 30 days of continuous oral administration of NG (120 rag/l) or tap water. Aliquots of gastric juice were titrated with 0.1 M NaOH.Other aliquots were extracted with ethyl acetate and subjected to specific radioimmunoassay for prostaglandin E 2. The severity of gastric mucosal lesions was evaluated in 60 rats after 5 days and 30 days of continuous oral adminis- tration of NG (120 rag/l) or NG plus misoprostol (200 gg/kg-a/day-1) or tap water, and a histological study was carried out. Administration of NG induced a significant decrease of gastric intraluminal prosta- glandin E 2 concentration at 15 and 30 days. Oral ad- ministration of misoprostol, at non-antisecretory doses, protected the rats against NG-induced gastric mucosal damage. Prostaglandins may be involved in the early phases of experimental gastric carcinogen- esis. Key words: Gastric experimental carcinogenesis - N- Methyl-N-nitro-N-nitrosoguanidine - Prostaglandins - Prostaglandin analogues - Gastric juice. * This work was supported by grants from the Ministry of Education (MPI) Offprint requests to." Alberto Materia, II Clinica Chirurgica, Poli- clinico Umberto I, Universita' degli Studi "La Sapienza', 1-00161 Roma, Italy Introduction Several studies have confirmed that N-methyl-N-ni- tro-N-nitrosoguanidine (NG) is a valuable tool for production of adenocarcinomas in the rat stomach since it is easily dissolved in tap water and adminis- tered orally, and produces lesions in over 70% of rats treated for 10-12 months (Deschner et al. 1979; Sugi- mura and Fujimura 1967; Kunze et al. 1979; Nagai et al. 1980; Ochara et al. 1980; Saito et al. 1970; Sugimura et al. 1970; Sugimura and Kawachi 1973). The sequential morphological changes of NG-induced gastric mu- cosal damage in the rat are characterized by four types of lesions: erosions, regenerative hyperplasia, adeno- matous hyperplasia and adenocarcinoma (Saito et al. 1970). Previous reports showed that repeated NG ac- tion on regenerating cells favors carcinogenesis in the eroded area, suggesting a close relationship between cytotoxic and carcinogenic effects for this compound (Saito et al. 1970; Tabuchi et al. 1974). However, the mechanism of formation of early NG-induced gastric mucosal lesions is not completely understood, nor it is known whether the prevention of mucosal erosions, and consequently of regenerative hyperplasia, could interfere with gastric carcinogenesis. Several studies have emphasized the prominent role played by prostaglandins in the protection of the gas- tric mucosa against a number of ulcerogenic agents and procedures (Jacobson et al. 1976; Robert 1979; Robert et al. 1979). However, the role of endogenous prostaglandins in experimental gastric carcinogenesis is not known. This study was initiated to investigate the changes of gastric intraluminal prostaglandins during the early phases (30 days) of NG-induced gastric carcinogenesis in the rat and to evaluate the effects of exogenous prostaglandin administration on NG-induced gastric mucosal lesions during the same period.

Upload: a-materia

Post on 10-Aug-2016

214 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

J Cancer Res Clin Oncol (1989) 115:253-258

J~om~lof

Cancer esearch Clinical �9 �9 Springer-Verlag 1989

Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat * A. Materia a, G. Silecchia a, E. Spaziani 1, p. Mariani 1, L. Scucchi 2, p. Mingazzini 2, C. Favalli a, E. Garaci a and N. Basso a 1 II Clinica Chirurgica University "La Sapienza", 2 Department of Biopathology University "La Sapienza", 3 Department of Experimental Medicine, University "Tor Vergata", Rome, Italy

Summary. The study was initiated to evaluate the ef- fect of N-methyl-N-nitro-N-nitrosoguanidine (NG) on gastric intraluminal prostaglandin release during a 30-day treatment period and to investigate the effect of a stable prostaglandin E a analogue (misoprostol) on NG-induced gastric mucosal damage during the same time period. Samples of gastric juice (1 h) were obtained from 40 male Sprague-Dawley rats with chronic gastric fistulas, in basal conditions and after 5, 15 and 30 days of continuous oral administration of NG (120 rag/l) or tap water. Aliquots of gastric juice were titrated with 0.1 M NaOH.Other aliquots were extracted with ethyl acetate and subjected to specific radioimmunoassay for prostaglandin E 2. The severity of gastric mucosal lesions was evaluated in 60 rats after 5 days and 30 days of continuous oral adminis- tration of NG (120 rag/l) or NG plus misoprostol (200 gg/kg-a/day-1) or tap water, and a histological study was carried out. Administration of NG induced a significant decrease of gastric intraluminal prosta- glandin E 2 concentration at 15 and 30 days. Oral ad- ministration of misoprostol, at non-antisecretory doses, protected the rats against NG-induced gastric mucosal damage. Prostaglandins may be involved in the early phases of experimental gastric carcinogen- esis.

Key words: Gastric experimental carcinogenesis - N- Methyl-N-nitro-N-nitrosoguanidine - Prostaglandins - Prostaglandin analogues - Gastric juice.

* This work was supported by grants from the Ministry of Education (MPI)

Offprint requests to." Alberto Materia, II Clinica Chirurgica, Poli- clinico Umberto I, Universita' degli Studi "La Sapienza', 1-00161 Roma, Italy

Introduction

Several studies have confirmed that N-methyl-N-ni- tro-N-nitrosoguanidine (NG) is a valuable tool for production of adenocarcinomas in the rat stomach since it is easily dissolved in tap water and adminis- tered orally, and produces lesions in over 70% of rats treated for 10-12 months (Deschner et al. 1979; Sugi- mura and Fujimura 1967; Kunze et al. 1979; Nagai et al. 1980; Ochara et al. 1980; Saito et al. 1970; Sugimura et al. 1970; Sugimura and Kawachi 1973). The sequential morphological changes of NG-induced gastric mu- cosal damage in the rat are characterized by four types of lesions: erosions, regenerative hyperplasia, adeno- matous hyperplasia and adenocarcinoma (Saito et al. 1970). Previous reports showed that repeated NG ac- tion on regenerating cells favors carcinogenesis in the eroded area, suggesting a close relationship between cytotoxic and carcinogenic effects for this compound (Saito et al. 1970; Tabuchi et al. 1974). However, the mechanism of formation of early NG-induced gastric mucosal lesions is not completely understood, nor it is known whether the prevention of mucosal erosions, and consequently of regenerative hyperplasia, could interfere with gastric carcinogenesis. Several studies have emphasized the prominent role played by prostaglandins in the protection of the gas- tric mucosa against a number of ulcerogenic agents and procedures (Jacobson et al. 1976; Robert 1979; Robert et al. 1979). However, the role of endogenous prostaglandins in experimental gastric carcinogenesis is not known.

This study was initiated to investigate the changes of gastric intraluminal prostaglandins during the early phases (30 days) of NG-induced gastric carcinogenesis in the rat and to evaluate the effects of exogenous prostaglandin administration on NG-induced gastric mucosal lesions during the same period.

Page 2: Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

254 A. Materia et al.: Prostaglandins and gastric carcinogenesis

Materials and Methods

Animals and chemicals. One hundred Sprague-Dawley male rats, weighing 220 250 g (Charles River, Como, Italy) were utilized in the study. The rats were conditioned to the room environment for at least 15 days prior to the beginning of each experiment. All rats were deprived of food, but not of water, for 24 h prior to surgery or col- lection of gastric juice and were housed in individual cages with wide wire bottoms to prevent coprophagia. Experiments were performed between 8 a.m. and 2 p.m. 1-Methyl-3-nitro-l-nitrosoguanidine (N N' N) 97% (Aldrich-Chemic, Steinheim, FRG) was used as the car- cinogen, diluted with tap water at a concentration of 120 rag/l. The solution was kept in dark bottles to avoid degradation by light. The solution was prepared and changed every other day. The animals were allowed to drink NG solution and to eat commercial rat chow ad libitum.

Experimental design

Study 1. In 40 rats a chronic gastric fistula was created, with a steel cannula (external diameter 12 ram) positioned in the anterior wall of the glandular stomach through a laparotomy under light ether anes- thesia. The cannula was secured to the abdominal wall and dosed with a screw steel plug. One week after surgery, the rats were trained to moderate restraint in Bollman cages for 20-min-periods each day for 2 weeks. The rats were then divided into two groups: group A (20 animals) received NG solution for 30 days; group B (20 animals) re- ceived tap water only for 30 days. In all rats gastric juice was col- lected through the gastric fistula for i h on days 0, 5, 15, and 30. After collection, aliquots of gastric juice were titrated with 0.1 M NaOH. Other aliquots were centrifuged at 4 ~ (2500 rpm x 5 min) and 2 ml supernatant was immediately extracted with 10 ml ethyl acetate. After centrifugation, the organic phase was evaporated in a stream of nitrogen and then dissolved in 2 ml phosphate buffer. Prostaglandin E2 concentration was measured by radioimmunoas- say using specific antibodies with minimal cross-reactivity with other types of prostaglandins (< 4%). l~5I-labeled prostaglandin E 2 was used as tracer. Antibodies, iodinated prostaglandin E2 and standards were obtained from New England Nuclear, Boston, Mass.

Study 2. Three groups of 20 rats each were studied: group A received NG solution for 30 days; group B received NG plus a stable analogue of prostaglandin E 1 (Dajani et al. 1976) (misoprostol, Searle Chemi- cals, Skokie, Ill, USA) (200 lag/kg- 1 day- 1) dissolved in tap water; group C received tap water only. Ten rats from each group were sac- rificed on day 6 and the remaining ten on day 30. The stomach of each rat was removed, opened along the greater curvature, pinned flat on a cork board and rinsed with saline solution for gross examination. The severity of gastric erosions was evaluated in a blind fashion us- ing a dissecting microscope. Only gastric mucosal lesions occurring in the glandular part of the mucosa were considered. The length and width of each lesion were measured to the nearest 0.2 mm and the surface area of erosion was calculated by assuming each erosion to be an ellipse (length x width/4) and the total area was summed and expressed in mm z (Kaufman and Grossman 1978). The average le- sion area for a group of animals was calculated as the mean of the lesion area for each rat +_ the standard error of the mean (SEM). After fixation in 10% neutral formalin, all the glandular stomach and duodenum was cut into 4-ram-wide strips and embebbed in paraffin. Several sections 4 lam thick were obtained from each spec- imen and stained with hematoxylin/eosin, periodic and / Schiff and phosphotungstic and / hematoxylin methods.

Study 3. In a separate experiment, the effect 0fmisoprostol on gastric secretion, at the doses cited above, was investigated in five rats that had chronic gastric fistula. In these animals gastric acid was measured in basal conditions and after 24 h of continuous oral

administration of the prostaglandin analogue at the dose 200 lag/ kg-1 day-1. Statistical evaluation of the data was performed by means of the two-way analysis of variance and by Student's t-test for paired values. P values less than 0.05 were considered as significant.

Results

Water intake was comparable in all rats with an aver- age of 45 ___ 12 ml/day and was consistent throughout the entire experimental period. Food intake was also comparable and consistent in all rats.

Study 1

The mean intake of NG was 6.0 +_ 1.5 rag/day for each rat. The gastric secretion volume and HC1 concentra- tion on days 0, 5, 15, and 30 are summarized in Figs. 1 and 2. No significant changes in either secretion vol- ume or HC1 concentration were found. Prostaglandin E 2 concentration in the gastric juice was shown to be 1105_+98.5 at day 0, 865_+125.8 at day 5, 645_+66.3 at day 15, and 596+_48.3 at day 30 with statistically significant differences (P<0.05) between day 0 and day 15 and day 0 and day 30 (Fig. 3). Consequently,

ml/h 5

al

E -~4 >

= :3

cn 2 ._o

0 0 5 days 15 days 30 days

[ ] control [ ] NG n.s,

Fig. 1. Gastric secretion volume in N-methyl-N-nitro-N-nitrosogua- nidine (NG)-treated animals and controls

J~Ea/ml 20 T

15

10 =o

8 U 5 -r

0 [ ] control

5 days 15 days I~NG

30 days

n.s.

Fig. 2. Gastric HC1 concentration in NG-treated animals and con- trols

Page 3: Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

A. Materia et al.: Prostaglandins and gastric carcinogenesis

Pg/ml

1200

900

o ~ 600 c o

300

O.

0 0 5 days 1,5 days 30 days

[] Control [] NG = # =p<O,05

Fig. 3. Gastric intraIuminal prostaglandin E 2 concentration in rats treated with NG and in control rats

Z mm

20

255

15 113

m 10 t -

9_ t~ _w 5

6 days

[ ] control [ ] NG

30 days

[ ] NG+PG * # = p<O,05

Fig. 4. Average lesion area (mm 2) on days 6 and 30 in rats treated with NG, NG plus misoprostol and tap water (control)

since the secretion volume did not change over the ex- perimental period, the total output of prostaglandin E 2 paralleled the changes in concentration. No varia- tions in gastric instraluminal prostaglandin Ez were observed in control animals.

Study 2

The mean intake of NG was 6,7 + 2.3 rag/day for each rat; the mean intake of misoprostol was 180 + 25 I-tg/ kg-1 day-1 for each rat. The average lesion area at day 6 was 15.6+1.4mm 2 in NG-treated rats, 1.2___ 0.2 mm 2 in rats treated with NG plus misopros- tol (P < 0.05) and 0 in the controls. At day 30, the aver- age lesion area was 9.8 +0.7 mm 2 in NG-treated ani- mals and 0.7 + 0.2 mm 2 in animals treated with NG plus misoprostol (P< 0.05). No mucosal lesions were observed in controls (Fig.4). No animal in either group experienced diarrhea or showed discomfort.

Study 3

Gastric secretion volume, measured in rats with gas- tric fistulas, was 6,1 _+ 0.4 ml/h in basal conditions and 5.2_+ 0.4 ml/h (P > 0.05) after 24 h of continuous oral administration of misoprostol (200 lag/kg). The HC1 concentration was 18.0 ___ 0.6 gequiv/ml in basal condi- tions and 16.6+0.04 gequiv/ml after misoprostol ad- ministration (P > 0.05).

Histological findings

At day 6, 24 deep erosions were observed in the glan- dular stomach of rats treated with NG (Fig. 5). Out of the eroded areas, antral mucosa showed degenerative and regenerative features (Fig. 6).

In the rats treated with NG plus misoprostol, no deep erosions were seen, and only 4 superficial ero- sions were observed. Out of these areas, the mucosa

Fig. 5. Mucosa of the glandular stomach of a rat treated for 5 days with NG. A deep erosion filled with necrotic debris is shown (H&E x 100)

appeared normal (Fig. 7). No significant gastric mu- cosal lesions were found in control animals.

At the end of the experiment (day 30), 21 deep ero- sions were found in NG-treated animals, 15 of them showing a pattern of coagulative necrosis. Regenerat- ive hyperplasia, without atypical cells, was present (Fig. 8).

In the animals treated with NG plus misoprostol, only 1 deep erosion of the antrum-duodenal junction and 2 small superficial erosions of the glandular stom- ach were found (Fig. 9). In these animals, slight areas of epithelial degeneration and mild antral inflamma-

Page 4: Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

256 A. Materia et al.: Prostaglandins and gastric carcinogenesis

Fig. 6. Antral mucosa of a rat treated with NG for 5 days. Nuclear pyknosis and karyorrhexis are evident. The nuclear enlargement, hy- perchromasia and the high mitotic rate are expressions of regenera- tion (H&E x 400)

Fig. 8, Mucosa of the glandular stomach of a rat, treated for 30 days with NG, showing regenerative hyperplasia of the foveolar epithe- lium (PAS x 100)

Fig.7. Mucosa of the glandular stomach of a rat treated with NG plus misoprostol for 5 days. No abnormalities of the mucosa are evi- dent except for a slight edema of the lamina prepria (H&E x 400)

Fig. 9. Mucosa of the glandular stomach of a rat, treated for 30 days with NG plus misoprostol. A shallow erosion of foveolar epithelium sparing the a, trat glands is shown (H&E x 400)

Page 5: Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

A. Materia et al.: Prostaglandins and gastric carcinogenesis 257

tory reaction were observed. Otherwise, gastric mu- cosa appeared normal and regenerative hyperplasia was a sporadic feature.

Discussion

On the basis of current data there is strong evidence that N-nitroso compounds can be causative factors in the aetiology of human gastric cancer (Mirvish 1983; Read 1986) and NG has been shown to be one of the most valuable agents employed in experimental gas- tric carcinogenesis (Deschner et al. 1979).

It is generally accepted that the formation of ero- sions, which represent the first NG-induced gastric mucosal lesions, is followed by regenerative hyper- plasia. The possibility that this lesion is an initial step in cancer development has been postulated (Jacobson et al. 1976; Tabuchi et al. 1974).

In the present study we observed, from the 5th day of treatment with NG, erosions mostly located in the body of the stomach. After 30 days, deep erosions with a pattern of coagulative necrosis were present, as- sociated with areas of regenerative hyperplasia. These findings can be related to a direct effect of NG on the surface epithelium. In this study NG did not affect gastric secretion volume or HC1 concentration over the entire experimental period confirming that NG acts directly on gastric epithelial cells. On the other hand, NG significantly decreased the concentration of intraluminal prostaglandin E2, expressing a reduced availability of prostaglandins at the gastric mucosal surface. Prostaglandins have been shown to be ac- tively synthesized by the gastric mucosa and a de- creased prostaglandin biosynthesis has been observed in several acute and chronic gastric ulcer models (Basso et al. 1983, 1985; Konturek 1984; Schlegel et al. 1977; Wright et al. 1982) suggesting that endogenous prostaglandins may play a role in the protection of the gastric mucosa against cytotoxic agents. On these bases, it is possible to hypothesize that endogenous prostaglandins are involved in the formation of NG- induced musosal lesions. Whether the prostaglandin inhibition represents a consequence of the extensive gastric mucosal damage or a preliminary effect of NG action contributing to the damage itself it is not known. Furthermore, the administration of cytopro- tective doses of a stable prostaglandin analog signifi- cantly prevented the cytotoxic effects of NG, with a marked reduction of mucosal erosions. At day 30, in rats treated with misoprostol the regenerative hyper- plasia was a sporadic feature and deep erosions were absent. These results indicate that exogenously admin- istered non-antisecretory doses of prostaglandins are able to antagonize the early changes of the gastric ep- ithelium induced by NG. Since, repeated sequences of

tissue damage and regenerative processes tend to pro- mote cancer de,~elopment, and since exogenous pros- taglandins seem to decrease the NG-induced early cy- totoxic lesion, an influence of these compounds on the outcome of dysplasia or carcinoma may be hypothe- sized. Long-term studies are underway to investigate the effects of prostaglandins on the development of neoplastic lesions. As for the mechanism of prosta- glandin-induced protection on the early gastric mu- cosal damage by NG, it can be found in the growing list of hypotheses already compiled to explain the pro- tection exerted by these compounds against several ul- cerogenic agents and methods (Robert 1979; Robert et al. 1979). In conclusion, the results of this study sug- gest that endogenous prostaglandins may play a role in the early phases of gastric experimental carcinogen- esis and that exogenous prostaglandins, at non-antise- cretory doses, prevent the formation of early NG-in- duced gastric mucosal lesions.

References

Basso N, Materia A, Forlini A, Jaffe BM (1983) Prostaglandin gen- eration in the gastric mucosa of rats with stress ulcer. Surgery 94:104-109

Basso N, Materia A, Bagarani M, Debes HT, Yamada T, Todisco A (1985) Selective inhibition of gastric somatostatin and prosta- glandins in stressed rats. Dig Dis Sci 30:368

Dajani EZ, Driskill DR, Bianchi RG, Collins PW, Pappo R (1976) SC-29333, a potent inhibitor of canine gastric secretion. Am J Dig Dis 21:1049-1057

Deschner EE, Tamura K, Bralow SP (J 979) Sequential histopathol- ogy and cell kinetic changes in rat pyloric mucosa during gastric carcinogenesis induced by N-methyl-N-nitro-N-nitrosogua- nidine. JNCI 63:1,171-177

Hattori T, Helpap B, Gedigk P (1984) Cell proliferation and growth of gastric carcinoma induced in inbred wistar rats by N-methyl- N-nitro-N-nitrosoguanidine. Cancer Res 44:5266-5272

Jacobsen ED0 Chaudhury TK, Thompson WJ (1976) Mechanism of gastric mucosal cytoprotection by prostaglandins. Gastroenter- ology 70:A897

Kauffman GL, Grossman MI (1978) Prostaglandins and cimetidine inhibit the formation of antral ulcers produced by parenteraI salycilates. Gastroenterology 75:1099-1103

Konturek SJ (1984) Actions of nonsteroid anti-inflammatory com- pounds on gastric mucosal integrity and prostaglandin forma- tion in healthy subjects and peptic ulcer patients. Adv Inflam- mation Res 6:29-37

Kunze E, Schauer A, Eder M, Seefledt C (1979) Early sequential le- sions during developement of experimental gastric cancer with special reference to dysplasias (1979) J Cancer Res Clin Oncol 95:247-264

Mirvish SS (1983) The etiology of gastric cancer: intragastric nitros- amides formation and other theories. JNCI 71:631-647

Nagai T, Pfeiffer CJ, Fuiimura M, Hattori T, Tabe T (1980) The re- lationship between gastric cancer and gastric carcinogenesis in- duced by N-methyl-N-nitro-nitrosoguanidine (MNNG). In: Ad- vances in Experimental Ulcer, Proc. 4th Int Conf Exp Ulcer, To- kyo. S. Uruahara, Tokyo, pp 650-655

Oohara T, Miyakuni T, Sadatsuki E, Mitarai Y, Kaminishi M (1980) Development of experimental gastric ulcer-cancer. In: Advances in Experimental Ulcer, Proc. 4th Int Conf Exp Ulcer, Tokyo. S. Uruahara, Tokyo, pp 656~63

Page 6: Role of prostaglandins in the early phases of experimental gastric carcinogenesis in the rat

258 A. Materia et al.: Prostaglandins and gastric carcinogenesis

Reed PI (1986) The role of nitrosamines in cancer formation. BiN Nutr Dicta 30:130-138

Robert A (1979) Cytoprotection by prostaglandin. Gastroenterol- ogy 77:761-7

Robert A, Nezamis JE, Lancaster C (1979) Cytoprotection by pros- taglandins in rats: prevention of gastric necrosis produced by al- cohol, HC1, NaOH, hypertonic NaC1 and thermal injury. Gas- troenterology 77:433-43

Saito T, Inokuchi K, Takayama S, Sugimura T (1970) Sequential morphological changes in N-methyl-N-nitro-N-nitrosogua- nidine carcinogenesis in the glandular stomach of rats. JNCI 44:769-783

Schlegel W, Wenk K, Dollinger HC (1977) Concentrations of pros- taglandin A-, E-, and F-like substances in gastric mucosa of nor- mal subjects and of patients with various gastric diseases. Clin Sci Mol Med 52:255-8

Sngimura T, Fujimura S (1967) Tumor production in glandular stomach of rat by N-methyl-N-nitro-N-nitrosoguanidine. Na- ture 216:943-4

Sugimura T, Kawachi T (1973) Experimental stomach cancer. Methods Cancer Res 7:245:308

Sngimura T, Fujimura S, Baba T (1970) Tumor production in the glandular stomach and alimentary tract of the rat by N-methyl- N-nitro-N-nitrosoguanidine. Cancer Res 30:455-65

Tabuchi Y, Ogino T, Mitsuno T, Sugiyama T (1974) Possible role of mucosal damage in stomach carcinogenesis with N-methyl-N- nitro-N-nitrosoguanidine in the rat. JNCI 52:1589-1594

Wright JP, Young GO, Klaff LF (1982) Gastric mucosal prostaglan- din E levels in patients with gastric ulcer disease and carcinoma. Gastroenterology 82:63-7

Received 5 December 1988/Accepted 31 January 1989