Detection of a Shared Colon Epithelial Epitope on Barrett Epithelium by a Novel Monoclonal Antibody

Methods

Tissue Samples

Retrospective Tissue Materials from the Esophagus and Gastroesophageal Junction

Fifty-three biopsy specimens taken during endoscopy at different levels of esophagus from 44 persons, and 12 surgical resection specimens for carcinoma of the esophagus arising in Barrett epithelium (7 patients) and squamous cell carcinoma of the esophagus (5 patients) were obtained from the archival material in the Department of Pathology.

Prospective Tissue Specimens from the Esophagus, Stomach, and Upper Small Intestine

Fifty-one biopsy specimens were obtained from 12 consecutive persons who were evaluated by upper gastrointestinal endoscopy for persistent symptoms of esophageal reflux disorders and acid peptic syndromes. All patients were evaluated by a single gastroenterologist, and biopsy specimens were taken systematically from various sites of the esophagus, stomach, duodenum, and jejunum (Table 1). Among 12 specimens with variable degrees of esophagitis, 6 specimens were taken from the gastroesophageal junction, and the remaining 6 were taken from distal esophagus just proximal to the squamocolumnar junction. Tissue specimens were fixed in formalin, and paraffin blocks were prepared for routine histologic study. Serial sections from each block were processed to study the immunoreactivity against the 7E₁₂H₁₂ monoclonal antibody by the immunoperoxidase method. The pathologist and the investigators who did the immunocytochemical studies were not informed about the history and clinical diagnosis of the patients, and they evaluated their findings independently.

Immunoperoxidase Method

The method of production and characterization of the 7E₁₂H₁₂ monoclonal antibody has been previously reported [8]. An unrelated mouse monoclonal antibody of IgM isotype (MOPC-104E) was used as a control. Normal colonic mucosal biopsy specimens were included in each experiment as a positive control against the 7E₁₂H₁₂ monoclonal antibody. The immunohistochemical analysis was done as previously described [8] with some modifications as described below [9]. The tissues were sectioned (5 microns), mounted on poly-L-lysine-coated slides, deparaffinized by heating at 56 °C for 1 hour, immersed in xylene, rehydrated in 100%, 95%, and 70% alcohol, and finally in phosphate-buffered saline (pH, 7.2). Free aldehydes were reduced with 0.05% sodium borohydride in phosphate-buffered saline (pH, 7.2) for 30 minutes at 4 °C. Sections were then sequentially incubated with normal swine serum, 7E₁₂H₁₂ monoclonal antibody, or control murine IgM monoclonal antibody, biotinylated swine antimouse IgM (Dakopatts; Carpinteria, California), hydrogen peroxide solution (3%), and streptavidin-peroxidase (Dakopatts), respectively. Tissue sections were washed in phosphate-buffered saline, treated with 3-3′ diaminobenzidene hydrochloride (50 µg/150 mL of 0.5 mol/L TRIS-buffer; pH, 7.2) for 30 minutes. The sections were washed, counter stained in hematoxylin or toluidine blue for 1 minute, dehydrated in graded ethanol solutions and then in xylene, and mounted for microscopic examination. The presence of clear brown staining of the tissue was graded as positive and its absence as negative.

Results

The location and histologic diagnosis of the 116 specimens collected retrospectively and prospectively from 53 patients are shown in Table 1. The mean age of these patients was 57 years for the retrospective group and 51 years for the prospective group (range, 14 to 85 years). Of 22 biopsy specimens, 21 (95%) with established diagnoses of specialized Barrett epithelium reacted with the 7E₁₂H₁₂ monoclonal antibody (Figure 1 C, Table 1). These included 19 of 20 retrospective specimens and 2 of 2 prospective specimens. Among these 21 positive patients, three biopsy specimens were taken from the esophagus at 20 to 25 cm, four from 25 to 30 cm, and 14 from 30 cm or below from the incisor teeth, as defined by the endoscopist. The one biopsy specimen that did not react with 7E₁₂H₁₂ monoclonal antibody was from the distal esophagus lower than 30 cm.

The specialized columnar epithelial cells of Barrett epithelium reacted strongly with 7E₁₂H₁₂ monoclonal antibody (Figure 1 C). The reactivity was more intense in the periphery of the cells (probably the membrane area compared with the cytoplasm). The 7E₁₂H₁₂ monoclonal antibody also reacted with some of the goblet cells, including their contents and basolateral regions of the cells. Each of the 11 specimens from normal gastroesophageal junction (squamocolumnar junction) was negative when reacted with the 7E₁₂H₁₂ monoclonal antibody (Figure 1 D and E). The availability of six operative specimens (from patients with cancer) from this area allowed us to obtain multiple tissue samples to evaluate both structure and immunoperoxidase staining. All of the 16 esophageal tissue specimens with normal squamous epithelium were negative when reacted with 7E₁₂H₁₂ monoclonal antibody (Figure 1 C, D, and E). Twelve specimens (six from distal esophagus and six from squamocolumnar junction) obtained from patients with endoscopic and histologic diagnosis of active esophagitis did not react with 7E₁₂H₁₂ monoclonal antibody.

Each of the 12 cases of adenocarcinoma arising in Barrett epithelium reacted with 7E₁₂H₁₂ monoclonal antibody and the staining was intense, mostly cytoplasmic (Figure 1 G and H). However, 12 of the 13 esophageal squamous cell carcinomas did not react with 7E₁₂H₁₂ monoclonal antibody (Figure 1)I. Only one specimen showed some focal and patchy staining in the tumor. Normal colonic biopsy specimens that were examined in parallel during each experiment as positive controls consistently reacted with the monoclonal antibody (Figure 1 F). None of the 21 specimens from sites of the stomach (cardia, fundus, body, and antrum), duodenum, and jejunum reacted with the 7E₁₂H₁₂ monoclonal antibody.

Discussion

Our study is the first report of a unique epitope shared between normal colonic mucosa and distinct specialized type of Barrett epithelium. Most cases of Barrett epithelium have a histologic similarity with incomplete intestinal metaplasia or small-bowel-like histologic findings and occasional gastric or fundic type of epithelium. However, 7E₁₂H₁₂ failed to react with small-bowel or gastric mucosa, confirming previous reports by us [8] and others [10], using both immunoperoxidase and immunofluorescence assays. These data suggest a histogenetic relation between specialized Barrett epithelium and colonic-type epithelium.

The origin of specialized Barrett epithelial cells in distal esophagus has been debated because no specific marker distinguishes Barrett epithelium. Metaplastic columnar epithelial cells of Barrett epithelium develop at higher frequency in the distal esophagus as a complication of reflux esophagitis [1, 2]. The cases that show the “intestinal type” histologic pattern tend to progress to adenocarcinoma at relatively higher frequency than other types of Barrett epithelia [11]. This type of Barrett epithelium has a colon epithelial phenotype. Several independent groups of investigators have reported a higher frequency of colonic neoplasia in patients with Barrett epithelium than in controls [12-14]. The reason for this clinical association is unknown. However, it is intriguing that a common epitope detected by the 7E₁₂H₁₂ monoclonal antibody is shared in these two tissues. Data from both the retrospective and prospective specimens show that the monoclonal antibody 7E (₁₂H)₁₂ reacts frequently (21 of 22) with the specialized type of benign Barrett epithelium but not with any other tissue systematically sampled from different sites of normal esophagus (including gastroesophageal junction); cardia, fundus, body, and antrum of the stomach; duodenum and jejunum. Further, the 7E₁₂H₁₂ monoclonal antibody reacted with all 12 cases of adenocarcinoma derived from Barrett epithelium; however, it did not react with the esophageal squamous cell carcinomas.

In addition to its potential as a new tool to study the histogenesis of Barrett epithelium, the 7E₁₂H₁₂ monoclonal antibody may be valuable in the diagnosis of Barrett epithelium. Endoscopic identification of the transition zone can be difficult, especially when the gastric mucosa extends into the distal esophagus and the structure of the gastric mucosa blends with the esophageal mucosa, which occurs in active esophagitis. Although some aids for endoscopic identification of gastroesophageal junction have been described [15, 16], the endoscopist may find that the area of diaphragmatic compression, gastroesophageal junction, and the gastroesophageal muscular junction are not identical. Because the immunoreactive epitope recognized by the 7E₁₂H₁₂ monoclonal antibody is not destroyed by formalin fixation, an additional tissue section can be stained by the immunoperoxidase method easily and at little added cost. Positive reactivity with the 7E₁₂H₁₂ monoclonal antibody strengthens the clinical suspicion of Barrett epithelium and in cases of any clinical and histologic uncertainty may alert the endoscopist to consider a second biopsy. The intensity of the immunoreactivity by 7E₁₂H₁₂ monoclonal antibody in the adenocarcinoma associated with Barrett epithelium appears to be higher compared with benign Barrett epithelium. Exploring the potential diagnostic role of the monoclonal antibody in defining patients with dysplasia involves answering such questions as whether the cellular distribution, pattern, or intensity of the reactivity changes with dysplasia in Barrett epithelium and establishing the timing of appearance of the immunoreactive epitope in patients with chronic reflux esophagitis.