and esophageal adenocarcinoma (EAC; n = 28) with immunohistochemical analysis. The results revealed significantly more overexpression of DcR3 in high-.
Anatomic Pathology / DCR3 OVEREXPRESSION IN ESOPHAGEAL CANCER
Overexpression of Decoy Receptor 3 in Precancerous Lesions and Adenocarcinoma of the Esophagus Huixiang Li, MD, PhD,1 Lurong Zhang, MD, PhD,2 Hong Lou, MD,3 Ivan Ding, MD,2 Sunghee Kim, PhD,2 Luping Wang, MD,2 Jiaoti Huang, MD, PhD,2 P. Anthony Di Sant’Agnese, MD,2 and Jun-Yi Lei, MD, PhD2 Key Words: Decoy receptor 3; Barrett esophagus; Esophageal cancer; Immunohistochemistry DOI: 10.1309/XK594E4B5WU82QR6
Abstract Overexpression of decoy receptor (DcR) 3 protein, a recently discovered member of the tumor necrosis factor receptor superfamily, was examined in 40 esophagogastrectomy specimens containing areas of Barrett esophagus (n = 27), low-grade dysplasia (n = 27), high-grade dysplasia or carcinoma in situ (n = 22), and esophageal adenocarcinoma (EAC; n = 28) with immunohistochemical analysis. The results revealed significantly more overexpression of DcR3 in highgrade dysplasia or carcinoma in situ and EAC than in benign esophageal mucosa (both P < .0001), Barrett esophagus (both P < .001), and low-grade dysplasia (P < .01 and P = .033, respectively). Low-grade dysplasia also showed significant overexpression of DcR3 compared with benign esophagus (P < .05) but not with Barrett esophagus (P > .05). DcR3 overexpression seems to negatively correlate with the grade of EAC. Our results suggest that overexpression of DcR3 protein might aid in the diagnosis of high-grade dysplasia or carcinoma in situ and EAC and also might serve as a potential therapeutic target.
282 282
Am J Clin Pathol 2005;124:282-287 DOI: 10.1309/XK594E4B5WU82QR6
Esophageal adenocarcinoma (EAC) is becoming more common worldwide, with its incidence tripling in the United States from 1976 to 1990 and an annual increase of approximately 10%.1 More than 13,900 new cases with 13,000 deaths were anticipated in the United States in 2003.2 Barrett esophagus is considered a key precancerous lesion with a strong association with the development of dysplasia and subsequent EAC,3 but the pathogenic mechanisms of this process remain unclear. Decoy receptor (DcR) 3 is a recently discovered member of the tumor necrosis factor receptor superfamily, which had 23 groupings and 30 members as of May 2005.4 The DcR group consists of 4 members, DcR1, DcR2, and DcR3 and osteoprotegerin.5,6 Although osteoprotegerin has very low physiologic affinity for TRAIL (tumor necrosis factor–related apoptosis-inducing ligand), DcR1 and DcR2 are capable of binding TRAIL but incapable of transducing the death signal because of lack of a death domain in their structures.6,7 In fact, DcR1 and DcR2 compete with the TRAIL receptor for the ligands and subsequently inhibit apoptosis. In addition, DcR can inhibit T-cell activation via the repression of actin polymerization and pseudopodium formation by T cells.8 DcR3 is a unique member in the DcR group because it is secreted rather than membrane-bound like DcR1 and DcR2. In addition, DcR3 binds to the Fas ligand, not the TRAIL, as other DcRs do.9 DcR3 also can modulate immune cell interaction by down-regulating the functions of macrophages and T cells.8,10 DcR’s functions of blocking apoptosis and escaping immune surveillance are important in carcinogenesis. The gene amplification and overexpression of DcR3 messenger RNA (mRNA) and protein have been demonstrated in lymphoma, glioblastoma, and cancer of the lung, esophagus, colon, and pancreas.9,11-14 The serum level of DcR3 also was significantly elevated in © American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
56.2% of a variety of tumors, including cancers of the digestive system, thyroid, lung, and breast.15 In the present study, we examined the expression of DcR3 in 40 cases of Barrett esophagus–associated lesions by immunohistochemical analysis. Association of DcR3 overexpression with the pathogenesis of EAC and the potential application of DcR3 overexpression in the diagnosis and treatment of EAC also are discussed.
Materials and Methods From the cases of Barrett esophagus–associated lesions in our surgical pathology files, we selected 40 esophagogastrectomy specimens that contained areas of normal esophageal mucosa (n = 40), Barrett esophagus (n = 27), low-grade dysplasia (n = 27), high-grade dysplasia or carcinoma in situ (CIS; n = 22), and EAC (n = 28). The diagnoses of low- and highgrade dysplasia were based on the World Health Organization criteria.16 Of 28 EACs, 7, 13, and 8 were well-, moderately, and poorly differentiated, respectively. Two measures were taken to ensure that all lesions were Barrett esophagus but not gastric cardiac–associated: (1) All cases were selected from patients who had a history of Barrett esophagus and/or Barrett esophagus–associated problems. (2) Strict topographic criteria were used to select only the cases in which the main tumors and/or lesions were clearly in the esophagus. The patients were 33 men and 7 women whose ages ranged from 42 to 84 years (mean, 66.2 years). All patients had a history of Barrett esophagus ranging from 2 to 12 years and had been followed up for 2 to 98 months (mean, 27 months). We selected 1 or 2 of the most representative sections (slides) from each case and stained them with a monoclonal antibody against DcR3 (provided by L.Z.). The procedure was done as described previously.17 Briefly, 5-µm-thick sections were deparaffinized, quenched with 3% hydrogen peroxide for 6 minutes, and heated for 40 minutes at 95°C to 99°C in DAKO antigen retrieval solution (citrate buffer, pH 6.1, DAKO, Carpinteria, CA) in a Black and Decker steamer (Shelton, CT). The slides were stained for 60 minutes with antihuman DcR3 monoclonal antibody (clone MD3B1, 1:1,000 dilution), followed by 30-minute incubations each in goat antimouse IgG (Vector Laboratories, Burlingame, CA) and streptavidin–horseradish peroxidase (Jackson Laboratories, West Grove, PA). The positive reaction was visualized with 3-amino-9-ethylcarbazole (DAKO) and showed cytoplasmic staining. We counted 100 cells from 5 representative areas in each lesion and normal esophageal mucosa. An average of 10% or more of tumor cells staining was considered positive. The degree of staining was subdivided as follows: 1+, focal or fine granular, weak staining; 2+, linear or cluster, strong staining; and 3+, diffuse, intense staining. Overexpression of DcR3 was
defined as staining of 2+ or more. Immunohistochemical staining was interpreted by 2 pathologists (J.H. and J-Y.L.). The reproducibility was 95.8% (138/144) between the 2 observers. The cases with disagreement in diagnoses were discussed by the 2 pathologists and categorized according to consensus. The Fisher exact test was used for comparing overexpression of DcR3 among different groups. A P value of less than .05 was considered statistically significant. Multivariate analysis was used for identifying correlation of DcR3 overexpression with patients’ age and survival, tumor size, and clinical and pathologic stages. A P value of less than .05 (for 1 side) or less than .025 (for 2 sides) was considered statistically significant.
Results DcR3 Expression in Normal Esophagus and Stomach Focal DcR3 overexpression was observed in 2 of 38 cases of normal esophageal (squamous) or gastric (glandular) mucosa, although it was seen more often in the esophageal or gastric mucosa adjacent to Barrett esophagus (3/27), dysplasia (low-grade, 8/27; high-grade, 8/22), or EAC (14/28). DcR3 also stained plasma cells. DcR3 Overexpression in Barrett-Associated Lesions DcR3 antibody showed similar granular stains in Barrett esophagus and Barrett-associated lesions, but the staining patterns were different. In Barrett esophagus, DcR3 staining was predominantly in the apical surface or basal portion of the columnar epithelium ❚Image 1A❚ and ❚Image 1B❚. It did not stain the mucin of goblet cells. In dysplastic and malignant cells, more cytoplasmic staining of DcR3 was seen ❚Image 1C❚, ❚Image 1D❚, ❚Image 1E❚, ❚Image 1F❚, ❚Image 1G❚, and ❚Image 1H❚. The percentage of positive cells and staining intensity of DcR3 were proportional to the severity of the esophageal lesions. Overexpression of DcR3 was observed significantly more often in EAC and high-grade dysplasia or CIS than in benign esophageal mucosa (both P < .0001), Barrett esophagus (both P < .001), and low-grade dysplasia (P < .01 and P < .05, respectively) ❚Table 1❚. Low-grade dysplasia showed significant overexpression of DcR3 compared with benign esophagus (P < .05) but not with Barrett esophagus (P > .05; Table 1). There was no significant difference in DcR3 overexpression between high-grade dysplasia or CIS and EAC (P = .23). Correlation of DcR3 Overexpression With BarrettAssociated Lesions In addition to its correlation with the grade of dysplasia, overexpression of DcR3 also tended to correlate with the degree of EAC differentiation. DcR3 overexpression was seen more often in well-differentiated EAC than in poorly Am J Clin Pathol 2005;124:282-287
© American Society for Clinical Pathology 283
DOI: 10.1309/XK594E4B5WU82QR6
283 283
Li et al / DCR3 OVEREXPRESSION IN ESOPHAGEAL CANCER
A
B
C
D
❚Image 1❚ Overexpression of decoy receptor (DcR) 3 in Barrett esophagus and its associated lesions. A and B, Barrett esophagus (A, H&E, ×100; B, DcR3, ×100). C and D, Low-grade dysplasia (C, H&E, ×100; D, DcR3, ×100).
differentiated EAC, although the difference was not significant (P = .051) ❚Table 2❚. Multivariate analysis revealed no association of DcR3 overexpression in EAC with tumor size, clinical and pathologic stage, patient age, and survival time (data not shown).
Discussion EAC remains one of the least studied and deadliest cancers in the world, despite its rapid increase in the United States between 1976 and 1990.1,2 The precursor lesion of EAC is thought to be Barrett esophagus, which can lead to dysplasia and eventually EAC. The pathogenesis of EAC remains 284 284
Am J Clin Pathol 2005;124:282-287 DOI: 10.1309/XK594E4B5WU82QR6
unclear. The diagnoses of dysplasia and early EAC also remain a challenge. Overexpression of DcR3 has been demonstrated in several human malignant neoplasms, including esophageal adenocarcinoma.9,11-14 However, the number of esophageal adenocarcinomas studied in literature is small (8 cases), and, to our knowledge, no study of overexpression of DcR3 in Barrett esophagus and Barrett-associated lesions has been reported. The present study was designed for such an investigation with more cases, and the results might provide better understanding of the pathogenesis and permit more accurate diagnosis of esophageal dysplasia and EAC. As is true for most human cancers, complete removal has been the most effective treatment for EAC. Thus, early and accurate diagnosis is essential. The finding of significant © American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
E
F
G
H
❚Image 1❚ E and F, High-grade dysplasia (E , H&E, ×100; F, DcR3, ×100). G and H, Esophageal adenocarcinoma (G, H&E, ×200; H, DcR3, ×200).
overexpression of DcR3 in high-grade dysplasia and EAC in the present study suggests its possible role as a biomarker for the diagnoses of high-grade esophageal dysplasia and/or EAC. The percentage of DcR3 protein overexpression in esophageal adenocarcinoma in the present study was higher than that previously reported14 (79% vs 63%). Such variation might result from a difference in the number of samples (28 vs 8). It is critical to differentiate high-grade dysplasia from low-grade dysplasia in Barrett esophagus because the former might result in an esophagectomy. Many biomarkers have been tested to help characterize Barrett-associated dysplasia, and some progress has been achieved.18-21 Expression of the bcl-xl protein and human telomerase reverse transcriptase mRNA is significantly higher in Barrett-associated dysplasia than in Barrett esophagus.19,21 However, correlation with grade
of dysplasia is not known. The significant overexpression of DcR3 in high-grade dysplasia or CIS vs low-grade dysplasia arising in Barrett esophagus in the study might provide some clue but not a specific marker to help diagnose high-grade dysplasia. Moreover, the positive correlation of DcR3 overexpression with degree of dysplasia might aid in monitoring the progression of esophageal dysplasia. Many genes, including oncogenes, tumor suppressor genes, growth factors and their receptors, apoptosis genes, DNA mismatch repair genes, and the p53 gene show aberrant transcription and translation in EAC.3,22 It is known that EAC can result from the metaplasia-dysplasia-carcinoma sequence, ie, Barrett esophagus → low-grade dysplasia → high-grade dysplasia → EAC. However, the molecular mechanisms in such a process remain unclear. The present study demonstrates a correlation between Am J Clin Pathol 2005;124:282-287
© American Society for Clinical Pathology 285
DOI: 10.1309/XK594E4B5WU82QR6
285 285
Li et al / DCR3 OVEREXPRESSION IN ESOPHAGEAL CANCER
❚Table 1❚ Overexpression of Decoy Receptor 3 in Different Types of Esophageal Tissue* P Compared With Decoy Receptor 3 Benign mucosa Barrett esophagus Low-grade dysplasia High-grade dysplasia Esophageal adenocarcinoma *
Esophageal Adenocarcinoma
2/40 (5) 6/27 (22) 12/27 (44) 16/22 (73) 22/28 (79)
High-Grade Dysplasia
