Histopathologic evaluation of an animal model for Barrett's esophagus and adenocarcinoma of the distal esophagus.

INTRODUCTION: Barrett's esophagus and adenocarcinoma of the esophagus are related to long-standing duodeno-gastroesophageal reflux. The development of an animal model in which Barrett's esophagus and/or carcinoma is induced by duodeno-(gastro-)esophageal reflux could provide better understanding of the pathogenesis of the metaplasia-dysplasia-carcinoma sequence and would create the possibility of investigating new treatment strategies for this aggressive disease.
MATERIALS AND METHODS: Two rat models were analyzed. In the first experiment, 44 male Sprague Dawley rats underwent end-to-side esophagojejunostomy with gastric resection, to ensure duodenoesophageal reflux without gastric acid. In the second experiment a side-to-side esophago-gastrojejunostomy was performed in 30 rats, ensuring duodeno-gastroesophageal reflux. In both experiments animals were not exposed to any exogenous carcinogens during the experiment. Sequential morphological changes (i.e., esophagitis, intestinal metaplasia, dysplasia, and carcinoma) were studied after 4, 6, and 12 months. To analyze histopathologic characteristics, evaluation of the hematoxylin and eosin specimens was combined with immunohistochemical stainings for high-iron diamine-alcian blue, alcian blue/periodic acid-Schiff, the proliferation marker PCNA, and mutations in the tumor suppressor gene p53.
RESULTS: In the first experiment, only 11 animals survived the postoperative period. These animals had to be sacrificed at a median of 11 weeks due to persistent weight loss and failure to thrive. Severe ulcerative esophagitis was seen in all animals, with a 2-mm segment of metaplastic epithelium found at the anastomosis. In four animals a large, well-differentiated, mucinous tumor without malignant characteristics was observed. In the second experiment, eight animals died postoperatively. Twelve animals were sacrificed according to protocol at 4 or 6 months. In these animals, extensive esophagitis with squamous cell hyperplasia was found. In addition, a short (2 mm) segment of metaplastic epithelium was observed, without dysplasia. The remaining animals survived 1 year. After 1 year, 9 of the 10 animals had developed a glandular metaplastic segment (median length, 10 mm), which was histologically and immunohistologically characteristic for the specialized columnar epithelium of Barrett's esophagus without signs of dysplasia. Finally, in seven animals a mucinous tumor with cytologic characteristics of a well-differentiated mucinous adenocarcinoma was found without infiltrative growth. These tumors were always found at the site of the anastomosis, originated in the submucosa, and did not reach either the luminal surface or the muscular layer. The mucinous lesions were not positive for p53, and PCNA was only slightly increased. Although they showed cytological characteristics of malignancy, histopathologic evaluation was more suggestive of a reactive mucous producing lesion fitting the diagnosis "esophagitis cystica profunda."
CONCLUSION: This study demonstrates the development of a long Barrett's segment in an animal duodeno-gastroesophageal reflux model. Although mucinous tumors resembling adenocarcinomas develop around the anastomosis, these are probably not reflux induced and are more likely to be reactive lesions. However, the true nature of these tumors remains to be elucidated.