ANTICANCER RESEARCH 27: 175-182 (2007)

Association Between Duodenal Contents Reflux and Squamous Cell – Establishment of an Esophageal Cell Line Derived from the Metastatic Tumor in a Rat Reflux Model

KUAN-HAO CHEN, KEN-ICHI MUKAISHO, ZHI-QIANG LING, AKIHIKO SHIMOMURA, HIROYUKI SUGIHARA and TAKANORI HATTORI

Department of , Shiga University of Medical Science, Seta-tsukinowa-cho, Otsu, Shiga, 520-2192, Japan

Abstract. (SCC) of the human world-wide are ESCCs (1). It is widely accepted that ESCC is has a multifactorial etiology involving several associated with smoking and (3-5). On the other hand, environmental and/or genetic factors. Recently, gastroesophageal a risk of ESCC is reported not to be associated with reflux has been implicated as a causative factor in upper gastroesophageal reflux. In addition, individuals with long- aerodigestive tract carcinogenesis. The development of standing and severe symptoms of reflux display odds ratios of esophageal squamous cell carcinoma (ESCC) in a duodenal 43.5 for EAC and 1.1 for ESCC (6). However, recent studies contents reflux model without any known carcinogen present has have demonstrated that refluxed duodenal contents cause been reported previously. In this study, the duodenal contents esophageal carcinoma in rats without exposure to carcinogens reflux model without gastrectomy was used. At 60 weeks post- (7, 8). The histological spectrum of these includes operatively, all surviving animals had malignant as ESCC, EAC and (7, 8). N-nitroso- follows: ESCC (40%), esophageal (EAC) compounds from multiple sources, including reflux of gastric (20%) and adenosquamous carcinoma (40%). In one subject, juice, have been postulated as one of the causative factors of a well-differentiated ESCC was detected with thoracic esophageal cancer (9, 10). Recently gastroesophageal reflux dissemination and metastases in lymph nodes. A novel cell line, has also been implicated as a causative factor in upper designated ESCC-DR, was established from the thoracic aerodigestive tract carcinogenesis, with SCC as the most metastatic tumor at the 60th post-operative week. These cells frequent histological type (11-15). To elucidate the causative were transplanted into nude mice, and the developed nodules factors underlying esophageal carcinomas including SCC, the represented a well differentiated ESCC, resembling that of the histological features of a rat duodenal contents reflux model parent site. Duodenal contents reflux has a great potential for were investigated with slight modification, i.e., esophago- malignant initiation and plays a role in developing not only EAC jejunostomy without gastrectomy. From a metastatic tumor in but also ESCC. this model, a novel cell line was established, ESCC-DR, and its characteristics were determined. The high potential of Esophageal cancer is the eighth most common cancer and the duodenal contents reflux for malignant initiation and the sixth most common cause of cancer-related mortality in the association between duodenal contents reflux and ESCC are world (1). The overall 5-year survival rate for patients discussed. diagnosed with esophageal cancer is poor (e.g. 3%-10%) (2). This exists in two main forms with distinct Materials and Methods etiological and pathological characteristics, esophageal squamous cell carcinoma (ESCC) and esophageal All procedures complied with the ethical guidelines for animal experimentation on the care and use of laboratory animals at Shiga adenocarcinoma (EAC). The majority of esophageal cancer University of Medical Science, Japan.

Duodenal contents reflux model and histological examination of esophageal tumors. Male Wistar rats weighing approximately 250 g Correspondence to: T. Hattori, Department of Pathology, Shiga were used in this experiment. After fasting for 24 hours, a midline University of Medical Science, Seta-tsukinowa-cho, Ohtsu, Shiga, laparotomy incision was made under inhalation anesthesia using 520-2192, Japan. Tel: +81 77 548 2166, Fax: +81 77 543 9880, e- diethyl ether. Reflux of duodenal contents was induced according mail: [email protected] to previously reported procedures with a slight modification (16) in that the esophago-gastric junction was transected, and the distal Key Words: Esophagus, duodenal reflux, squamous cell carcinoma, cut end closed, the proximal cut end was anastomosed end-to-side cell line, rat reflux model. to the upper jejunum from about 2 cm anal to the origin.

0250-7005/2007 $2.00+.40 175 ANTICANCER RESEARCH 27: 175-182 (2007)

Table I. Histological findings at 10, 20, and 30 weeks post-operatively.

Histological findings 10 weeks 20 weeks 30 weeks (number of cases/total cases)

Esophageal ulcer 15/16 16/16 16/16 Squamous hyperplasia 12/16 16/16 16/16 Squamous 2/16 9/16 12/16 Squamous cell carcinoma 0/16 0/16 3/16 Barrett esophagus 0/16 5/16 8/16 Barrett dysplasia 0/16 1/16 2/16 Adenocarcinoma 0/16 0/16 2/16

USA) and these were left at an angle of 90Æ for 30 min before adding culture medium. The growth medium used for primary cultures was Dulbecco’s modified Eagle’s medium (DMEM; Nakalai Tesque, Kyoto, Japan) with 1% Antibiotic-Antimycotic solution (GIBCO, NY, USA), and 10% fetal bovine serum (FBS). Culture flasks were left undisturbed for 3 days in a humidified incubator containing 5% CO2 at 37ÆC, then the culture medium (4 ml/flask) was changed twice a week. After 2 weeks, 3 flasks containing 2-3 colonies was kept. The colonies with small cell numbers by sterile cotton swab were excluded and only the largest colony in the same flask was kept. Then one of these flasks containing one colony was used. Confluent cultures were Figure 1. Animal model. We established a duodenal contents reflux subcultured after 30 days from the start of culturing by replacing model, using esophago-jejunostomy without gastrectomy. To reduce the growth medium with PBS. After draining off the solution, 0.05% volume of duodenal contents reflux compared with the previously reported Trypsin/0.53 mM-EDTA (Nakalai Tesque) was added to transfer models, a serosal suture between the esophagus and the jejunum was culture cells into a monolayer culture system. Cells were transferred added after esophago-jejunostomy as indicated. at dilutions of 1:4 using 0.05% Trypsin/0.53 mM-EDTA solution to detach cells.

Backtransplantation. Five nude mice (BALB/cA Jcl-nu) were Subsequently, to reduce the volume of duodenal contents reflux subcutaneously inoculated with 1x106 ESCC-DR cells into the back compared with the previously reported model, a serosal suture skin. Developed tumors were removed 4 weeks after injection, between the esophagus and the jejunum was added after esophago- fixed in 10% formalin in PBS for 4 hours and embedded in jejunostomy. As a result, duodenal contents flowed back into the paraffin. Serial 4-Ìm sections were stained with HE. esophagus through the esophago-jejunal stoma (Figure 1). All sutures were performed using interrupted 7-0 nylon sutures. Results Animals were allowed access to water 12 hours post-operatively and food 36 hours later, and were not treated using any known carcinogens. Animals were sacrificed using an overdose of diethyl Histological findings in this model. All histological findings are ether at post-operative weeks 10 (n=16), 20 (n=16), 30 (n=16), summarized in Table I. In the esophagus of the present reflux and 60 (n=5). Resected esophagus were fixed in 10% formalin in model, we detected esophageal ulcers (93.8%), hyperkeratosis phosphate-buffered saline (PBS) for 4 hours and embedded in (80.9%) and dysplasia with mild atypia (12.5%) at 10 weeks paraffin. Serial 4-Ìm sections were made and stained using hematoxylin and eosin (HE). The histological features were post-operatively. At 20 weeks post-operatively, esophageal classified into esophageal ulcer, squamous cell hyperplasia, ulcers (100%) and dysplasia with moderate to severe atypia squamous cell dysplasia, Barrett epithelium (so-called specialized (56.3%) were detected. Barrett esophagus (31.3%) and columnar epithelium), Barrett dysplasia (Barrett epithelium with Barrett dysplasia (0.82%) were also detected. At 30 weeks atypical cells), ESCC, adenosquamous carcinoma and EAC. post-operatively, we detected Barrett esophagus (50.0%), squamous cell dysplasia (75.0%), ESCC (18.8%), and EAC Establishment of a cell line. A cell line from a thoracic metastatic (12.5%). No adenosquamous carcinoma was detected within tumor at the 60th week post-operative was established according to a previously reported procedure (17). In briefly, the tissue sample 30 weeks post-operatively. At 60 weeks post-operatively, all was rinsed using PBS (Nakalai Tesque, Kyoto, Japan) and then animals had malignant lesions as follows; ESCC (40%), EAC culture medium, and then minced using a sharp pair of scissors. (20%) and adenosquamous carcinoma (40%). When we Fragments were placed into 25 cm 2 plastic flasks (Corning, NY, opened the thoracic cavity of 1 of 5 surviving rats at 60 weeks

176 Chen et al: Association between Duodenal Reflux and SCC

Figure 2. Macroscopic findings of a case of thoracic disseminated case. a) Serosal side in the thoracic cavity: Swelling of lymph nodes, and multiple small nodules in the surface of , heart and peritoneum were detected. b) Mucosal side: Marked thickness of esophageal wall, deep ulceration and an unevenly eroded surface were detected.

post-operatively, we detected an increase in the wall thickness Macroscopically, all mice exhibited nodules, measuring on the serosal side of esophagus, swelling of lymph nodes, and about 1 cm in diameter, during the 4 weeks after injection multiple small nodules on the surface of the lung, heart and (Figure 5). Histologically, nodules represented well peritoneum (Figure 2). There were deep ulcerations in the differentiated ESCC, resembling those of the primary and lower portion of the esophagus and an unevenly eroded metastatic sites (Figure 6). surface in the middle portion apart from the anastomosis. We tried to establish the cell line using the metastatic site. Discussion Microscopic examination revealed the tumor was a well- differentiated SCC with thoracic dissemination and In a duodenal contents reflux model of rats, we detected in mediastinal lymph nodes and the lung (Figure 3). ESCCs with thoracic dissemination and metastases in lymph nodes developed at 60 weeks post-operatively. Although Establishment of a cell line. A cell line, ESCC-DR, was many previous studies have reported that refluxed duodenal established from the metastatic site at 60 weeks post- contents caused esophageal carcinomas in rats without operatively. This cell line maintained a doubling time of 24- exposure to carcinogens, this is, to our knowledge, the first hour through more than 30 passages from the initial culture report that describes the development of esophageal (Figure 4). carcinoma with thoracic dissemination and metastases in the reflux model without using any known carcinogens. This Backtransplantation. Injection of 106 cultured cells on the indicates that the duodenal reflux can induce the tumors of 15th passage produced tumors in 5 out of 5 mice. highly malignant behavior as well as of malignant .

177 ANTICANCER RESEARCH 27: 175-182 (2007)

Figure 3. Histological findings of main and metastatic tumors at 60 weeks postoperatively. a) Main tumor. b) Invasion to the . c) Metastatic tumor in the lung. d) Metastatic tumor in the regional (arrow).

A cell line from the metastatic tumor was also established. presence of bacterial flora in duodenal juice that are capable The cultured cells were transplanted into nude mice, and of catalyzing endogenous reactions to produce nitroso produced tumors similar to that of the parent site. These compounds (21, 22). Recently, we evaluated the inhibitory results have confirmed a high potential of duodenal effect of thiazolidine-4-carboxylic acid (thioproline) as a contents reflux for malignant initiation. nitrite scavenger and as a most sensitive probe to detect N- However, little is known about the mutagenic potential of nitroso compounds. Thioproline was found to prevent refluxed material on esophageal mucosa. Thisen et al. esophageal and remnant stomach carcinogenesis, thereby recently reported preliminary evidence of the mutagenic indicating the possible role of N-nitroso compounds in potential of bile reflux on esophageal epithelium, using carcinogenesis (10, 23). In addition, extensive research in Sprague Dawley/Big Blue F1 lacI transgenic rats which China has also suggested that N-nitroso compounds and underwent esophagoduodenostomy to surgically create their precursors are probable etiological factors for duodeno-gastric-esophageal reflux (18). On the other hand, esophageal cancer in the areas of high incidence (24). many researchers have already addressed the role of bile Although it is widely accepted that duodenogastric- acids and pancreatic juice in elucidating carcinogenic effects esophageal reflux is directly linked to Barrett’s esophagus of duodenal contents (19, 20), and others referred to the and to the development of EAC, a risk of ESCC is reported

178 Chen et al: Association between Duodenal Reflux and SCC

Figure 4. Morphology of an established novel cell line designated ESCC-DR on the 15th passage. The cultured cells appear epithelial in shape and grow in multi-layers without contact inhibition.

Figure 5. Macroscopic findings of the tumor developed in the nude mouse. The multi-nodular tumor, measuring about 1 cm in diameter, was detected at the 4th week after subcutaneous inoculation into the back of the nude mouse.

not to be associated with gastroesophageal reflux (6). In compared to EAC (8). This means that histological features contrast, results of several studies using rat duodenal may depend on the volume of reflux contents; small contents reflux models have shown that development of amounts of reflux causes ESCC and a large volume of reflux esophageal carcinomas includes squamous-cell carcinoma causes EAC. In our modified model, we added a serosal (8, 10, 25). In this study, the incidence of pure suture between the esophagus and the jejunum after adenocarcinoma is lower than that of squamous cell esophago-jejunostomy. This addition of a serosal suture may carcinoma. It is unclear what factors lead to the formation decrease the reflux of duodenal contents compared with of carcinomas of specified histology. Miwa et al. suggested other models, so that the incidence of ESCC was higher ESCC developed in places distant from the anastomosis than that of EAC in this study.

179 ANTICANCER RESEARCH 27: 175-182 (2007)

Figure 6. Histological findings of the tumor developed at the site of inoculation. The tumor was shown to be a well-differentiated squamous cell carcinoma, similar in morphology to the metastatic tumor from which it was derived.

Great interest has been focused on gastroesophageal could represent a suitable material for molecular analyses of reflux as an independent carcinogenic factor and a co- ESCC and tests of chemotherapeutic agents against ESCC. carcinogen of laryngeal SCC in association with smoking and alcohol assumption (11-15). Commarota et al. reported Acknowledgements that the risk of laryngeal SCC increased more than 10-fold This work was supported in part by Grant-in-aid for Cancer in 20 years or more after gastric resection in a retrospective Research from the Ministry of Health, Labour and Welfare. case-control study of subjects admitted in the same hospital (14). This study included 828 consecutive patients with References laryngeal cancer and 825 controls with acute myocardial infarction. The term of 20 years is quite long. Although they 1 Parkin DM, Bray F, Ferlay J and Pisani P: Estimating the world mentioned the carcinogenetic effect of deudenogastro- cancer burden: Globocan 2000. Int J Cancer 94: 153-156, 2001. esophageal reflux in laryngeal cancer, they also suggested 2 Cilley RE, Strodel WE and Peterson RO: Cause of death in that the distance of the from the stomach and the carcinoma of the esophagus. Am J Gastroenterol 84: 147-149, 1989. presence of a low-grade, chronic stimulation could explain 3 Nozoe T, Korenaga D, Kabashima A and Sugimachi K: the long time needed for duodenogastroesophageal reflux Smoking-related increase of O(6)-methylguanine-DNA to become dangerous. This hypothesis also supports the methyltransferase expression in squamous cell carcinoma of the possibility that continuous small amounts of reflux contents esophagus. Cancer Lett 8: 49-55, 2002. might be associated with the development of SCC. 4 Devesa S, Blot WJ and Fraumeni JF Jr: Changing patterns in In conclusion, duodenal contents reflux has a great the incidence of esophageal and gastric carcinoma in the United potential for malignant initiation, inducing not only EAC but States. Cancer 83: 2049-2053, 1998. 5 Yokoyama A, Ohmori T, Muramatsu T, Higuchi S, Yokoyama also ESCC, and even developing highly malignant tumors T, Matsushita S, Matsumoto M, Maruyama K, Hayashida M with thoracic dissemination and metastases in lymph nodes and Ishii H: Cancer screening of upper aerodigestive tract in and the lung. From the metastatic tumor, we have Japanese alcoholics with reference to drinking and smoking established the first culture model that was derived from the habits and aldehyde dehydrogenase-2 genotype. Int J Cancer reflux-induced tumors. This cell line, designated ESCC-DR, 68: 313-316, 1996.

180 Chen et al: Association between Duodenal Reflux and SCC

6 Lagergren J, Bergstrom R, Lindgren A and Nyren O: 16 Hanby AM, Pera M, Filipe I, Duranceau A, Wright NA, Pera Symptomatic gastroesophageal reflux as a risk factor for M, Grande L and Poulsom R: Duodenal content reflux esophageal adenocarcinoma. N Engl J Med 340: 825-831, 1999. esophagitis in the rat: an animal model for the ulcer-associated 7 Kumagai H, Mukaisho K, Sugihara H, Bamba M, Miyashita T, cell lineage (UACL)? Am J Pathol 151: 1819-1824, 1997. Miwa K and Hattori T: Cell kinetic study on histogenesis of 17 Mukaisho K, Miwa K, Totsuka Y, Shimomura A, Sugihara H, Barrett's esophagus using rat reflux model. Scand J Wakabayashi K and Hattori T: Induction of gastric GIST in rat and Gastroenterol 38: 687-692, 2003. establishment of GIST cell line. Cancer Lett 231: 295-303, 2006. 8 Miwa K, Sahara H, Segawa M, Kinami S, Sato T, Miyazaki I and 18 Theisen J, Peters JH, Fein M, Hughes M, Hagen JA, Demeester Hattori T: Reflux of duodenal or gastro-duodenal contents induces SR, Demeester TR and Laird PW: The mutagenic potential of esophageal carcinoma in rats. Int J Cancer 17: 269-274, 1996. duodenoesophageal reflux. Ann Surg 241: 63-68, 2005. 9 Rogers MA, Vaughan TL, Davis S and Thomas DB: 19 Busby WF Jr, Shuker DE, Charnley G, Newberne PM, Consumption of nitrate, nitrite, and nitrosodimethylamine and Tannenbaum SR and Wogan GN: Carcinogenicity in rats of the the risk of upper aerodigestive tract cancer. Cancer Epidemiol nitrosated bile acid conjugates N-nitrosoglycocholic acid and N- Biomarkers Prev 4: 29-36, 1995. nitrosotaurocholic acid. Cancer Res 45: 1367-1371, 1985. 10 Kumagai H, Mukaisho K, Sugihara H, Miwa K, Yamamoto G 20 Kaibara K, Takebayashi M and Koga H: Deoxycholic acids as and Hattori T: Thioproline inhibits development of esophageal a promoter in the development of gastric carcinoma induced by adenocarcinoma induced by gastroduodenal reflux in rats. N-methyl-N'-nitrosoguanidine in rats. Cancer J 2: 14-16, 1988. Carcinogenesis 25: 723-727, 2004. 21 Domellof L, Reddy BS and Weisburger JH: Microflora and 11 El-Serag HB, Hepworth EJ, Lee P and Sonnenberg A: deconjugation of bile acids in alkaline reflux after partial Gastroesophageal reflux disease is a risk factor for laryngeal and gastrectomy. Am J Surg 140: 291-295, 1980. pharyngeal cancer. Am J Gastroenterol 96: 2013-2018, 2001. 22 Muscroft TJ, Deane SA, Youngs D, Burdon DW and Keighley 12 Galli J, Cammarota G, Calo L, Agostino S, D'Ugo D, Cianci R MR: The microflora of the post-operative stomach. Br J Surg and Almadori G: The role of acid and alkaline reflux in 68: 560-564, 1981. laryngeal squamous cell carcinoma. Laryngoscope 112: 1861- 23 Suo M, Mukaisho K, Shimomura A, Sugihara H and Hattori T: 1865, 2002. Thioproline prevents carcinogenesis in the remnant stomach 13 Galli J, Calo L, Agostino S, Cadoni G, Sergi B, Cianci R and induced by duodenal reflux. Cancer Lett 237: 256-262, 2006. Cammarota G: Bile reflux as possible risk factor in 24 Lin K, Shen W, Wu Y and Lu S: Exposure level of total N- laryngopharyngeal inflammatory and neoplastic lesions. Acta nitroso compounds in residents of high-and low-risk areas for Otorhinolaryngol Ital 23: 377-382, 2003. esophageal cancer in southern Zhonghua. Yu Fang Yi Xue Za 14 Cammarota G, Galli J, Cianci R, De Corso E, Pasceri V, Palli Zhi 36: 386-389, 2002. D, Masala G, Buffon A, Gasbarrini A, Almadori G, Paludetti G, 25 Miwa K, Hattori T and Miyazaki I: Duodenogastric reflux and Gasbarrini G and Maurizi M: Association of laryngeal cancer foregut carcinogenesis. Cancer 75: 1426-1432, 1995. with previous gastric resection. Ann Surg 240: 817-824, 2004. 15 Cianci R, Galli J, Agostino S, Bartolozzi F, Gasbarrini A, Almadori G, D'Ugo D, Gasbarrini G and Cammarota G: Received May 29, 2006 Gastric as a long-term risk factor for malignant lesions Revised December 14, 2006 of the larynx. Arch Surg 138: 751-754, 2003. Accepted December 19, 2006

181