SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

Background : The incidence of Barrett's oesophageal adenocarcinoma is increasing more rapidly than any other malignancy in industrialized countries. Cyclo-oxygenase-2 appears to play an important role in gastrointestinal carcinogenesis. Previous studies on cyclo-oxygenase-2 expression in Barrett's oesophageal carcinogenesis have utilized tissue samples obtained from different patients. We sought a definitive comparison of cyclo-oxygenase-2 expression in the sequence of Barrett's metaplasia–dysplasia–adenocarcinoma within the same patients.

Methods : Paraffin-embedded oesophago-gastrectomy specimens from 20 patients, containing successive stages of Barrett's metaplasia, high-grade dysplasia and adenocarcinoma, were analysed for cyclo-oxygenase-2 expression by immunohistochemistry.

Results : Cyclo-oxygenase-2 was constitutively expressed in the basal layers of cells in the adjacent normal squamous oesophageal epithelium, but a higher cyclo-oxygenase-2 expression was observed in Barrett's metaplasia. A further increase in cyclo-oxygenase-2 expression was detected in high-grade dysplasia, but cyclo-oxygenase-2 was decreased in adenocarcinoma tissue, regardless of its stage or level of differentiation.

Conclusions : Cyclo-oxygenase-2 expression is progressively increased when squamous oesophageal epithelium develops into Barrett's metaplastic epithelium and then into high-grade dysplasia, but appears to decrease when adenocarcinoma develops. These findings may be significant for an effective chemo-prevention strategy with selective cyclo-oxygenase-2 inhibitors.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

The rapidly increasing incidence of oesophageal adenocarcinoma in the Western world is a matter of great concern.1 Oesophageal adenocarcinoma occurs with an annual incidence of 12–16 cases per 100 000 population in the UK,2 and is currently increasing more rapidly than any other malignancy.3, 4 Over the last three decades, the overall 5-year survival rate for oesophageal adenocarcinoma has remained at less than 20%. In view of the extremely poor prognosis of this disease, and a lack of unequivocal data documenting the efficacy of anti-reflux and/or endoscopic therapies, it is extremely important to understand the pathogenesis of Barrett's oesophageal adenocarcinoma and to develop strategies to prevent it. Like other cancers of the alimentary tract, oesophageal adenocarcinoma develops through a multi-step process in which the metaplastic epithelium sequentially develops low- and high-grade dysplasia, early cancer and eventually advanced cancer. This sequence is associated with a series of molecular changes that include p53 and p16 mutations, aneuploidy and microsatellite instability.5 The earliest recognized change is Barrett's oesophagus, in which the stratified squamous epithelium in the lower oesophagus is replaced by metaplastic columnar intestinal-type mucosa.6 It has been estimated that patients with Barrett's oesophagus have a 30–125-fold increased risk of developing oesophageal adenocarcinoma compared with the general population.7

There is a substantial body of epidemiological data documenting a reduction in risk of oesophageal cancer of up to 90% in individuals taking aspirin or other non-steroidal anti-inflammatory drugs (NSAIDs).8–11 This has focused the interest of researchers on the role of cyclo-oxygenase (COX) in the pathophysiology of oesophageal neoplasia. In contrast with COX-1, which is a constitutive enzyme involved in the maintenance of tissue integrity and homeostasis, COX-2 is an inducible enzyme predominantly expressed at sites of inflammation in response to cytokines, growth factors, interleukins and tumour promoters.12–15 COX-2 catalyses the production of prostaglandins that stimulate cancer cell proliferation, inhibit apoptosis and enhance cancer-induced angiogenesis16 and invasiveness.17 Wilson et al. reported that COX-2 was over-expressed in dysplastic oesophageal mucosa and in adenocarcinoma, but not in Barrett's intestinal metaplasia, nor in the adjacent normal squamous epithelium.18 In contrast, Zimmermannet al. reported COX-2 immunoreactivity in 21 of 27 patients with oesophageal adenocarcinoma, but could not detect expression in the adjacent Barrett's metaplastic epithelium.19 Shirvani et al. described a progressive increase in COX-2 expression from squamous epithelium to Barrett's metaplasia, dysplasia and adenocarcinoma.20 Recently, Morris et al. also reported a significantly higher expression of COX-2 activity in high-grade dysplasia or oesophageal adenocarcinoma when compared with low-grade dysplasia and Barrett's oesophagus.21 There was no difference in COX-2 expression between Barrett's oesophagus and low-grade dysplasia, or between high-grade dysplasia and oesophageal adenocarcinoma.

One important limitation in previous studies describing COX-2 expression at each stage of the metaplasia–dysplasia–adenocarcinoma sequence is the fact that comparisons have been made using tissue samples obtained from different patients. Such data inevitably reflect variations in COX-2 expression between individuals.19, 22 The inherent inter-patient variability reduces even further the statistical power of already small studies. In the present study, we performed an immunohistochemical investigation of COX-2 in the sequence of Barrett's metaplasia–dysplasia–adenocarcinoma from oesophago-gastrectomy specimens demonstrating all three stages of the carcinogenesis sequence within each patient.

Patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

Twenty paraffin-embedded tissue specimens obtained from patients who had undergone oesophago-gastrectomy for oesophageal adenocarcinoma between 1995 and 2000 were collected retrospectively from the pathology archive at the Norfolk and Norwich University Hospital. Patients were included if the specimens contained Barrett's metaplasia and high-grade dysplasia adjacent to the adenocarcinoma. Based on retrospective examination of the patients' clinical notes, those who had received pre-operative chemo- or radiotherapy, or who had regularly consumed aspirin or NSAIDs before the operation, were excluded. The study was carried out with the approval of the Norwich District Ethics and Research Governance Committees.

Histopathological analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

All tissues were reviewed by at least two senior histopathologists. TNM staging of the oesophagus (ICD-O C15) followed the Union Internationale Contre le Cancer criteria. Barrett's epithelium was characterized by specialized intestinal-type epithelium with a villiform surface, mucous glands and the presence of goblet cells. Dysplasia was defined as the development of neoplastic tissue confined within the superficial layer of the epithelium by an intact basement membrane, and graded according to criteria similar to those used by the Inflammatory Dysplasia Morphology Study Group.23 High-grade dysplasia was characterized by marked distortion of the crypt architecture with a cribriform pattern, stratified nuclei involving the cell apex, loss of nuclear polarity, nuclear anisocytosis and pleomorphism, prominent nucleoli and numerous mitotic figures. Neoplastic lesions with infiltration through the basement membrane were identified as adenocarcinoma.

Immunohistochemistry for COX-2

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

COX-2 protein expression in the tissues was examined by immunohistochemistry with a COX-2-specific mouse antihuman monoclonal antibody (160112, Cayman Chemical, Ann Arbor, MI, USA). Formalin-fixed, paraffin-embedded samples of oesophageal tissue, together with positive control samples containing human colonic adenocarcinoma, normal colonic epithelium and colonic muscularis propria, were cut into 4-µm sections. All oesophageal and positive control slides were processed in one batch in an Optimax Plus 2.0 automated cell stainer (Biogenex, San Ramon, CA, USA). The biotin–avidin system (PK-6200, Vectastain Universal Elite ABC kit, Vector Laboratories Inc, Burlingame, CA, USA) was used for staining. Tissue sections were incubated with the COX-2 antibody at a dilution of 1 : 200, followed by biotinylated secondary antibody solution and avidin–biotylated enzyme complex (ABC) reagents, for 30 min each at room temperature. Colour was developed for 10 min with diaminobenzidine tetrachloride. The sections were then counter-stained with haematoxylin. The positive control slides containing colonic adenocarcinoma, normal colonic epithelium and colonic muscularis propria displayed strong, moderate and weak COX-2 staining, respectively. In addition, the oesophageal stromal cells (fibroblasts, lymphocytes and endothelial cells) and smooth muscle of the muscularis propria were used as internal positive controls. Negative controls were performed after pre-absorption of the COX-2 antibody with human COX-2 peptide at 10 µg/mL (360107, Cayman Chemical, Ann Arbor, MI, USA), and no staining was seen (not shown).

Image analysis for COX-2

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

Digital pictures of the four tissue areas (squamous epithelium, Barrett's oesophagus, high-grade dysplasia and oesophageal adenocarcinoma) of each tissue section were photographed by a histopathologist (LI) with a Nikon Coolpix 950 digital camera under uniform optical conditions at × 400 magnification. For each of the four tissue areas, five representative fields were photographed and assessed. Only cells in the deep or basal layers of the four tissue areas were assessed because they had higher levels of COX-2 expression compared with cells located more superficially.20 Only cells from the epithelial component of the four tissue areas were assessed, and no stromal component was included in the analysis. COX-2 expression was determined by assessing 100 cells from each selected field. Semi-quantitative computer-assisted image analysis for COX-2 expression was performed using Image-Pro Plus version 4.0. All digital pictures were acquired for image analysis under uniform conditions. The immunostained images were separated into hue (colour), saturation (colour density) and intensity (contrast of an image). The hue and the intensity were fixed at 0–130 and 0–255, respectively, throughout the analyses. The values of the colour density (saturation) and total area stained were acquired for each image. The mean saturation of each image per unit area stained was used as the value of the density of COX-2 immunostaining, and it was expressed in arbitrary units (AU). The whole analysis was repeated on two separate occasions, 1 month apart, and the same outcome was found.

Statistics

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

COX-2 expression in the four tissue areas was analysed using two-way analysis of variance (anova). The Scheffé test for multiple comparisons was used to examine the associations between COX-2 expression and each stage of Barrett's oesophageal carcinogenesis. The association between the clinicopathological parameters and COX-2 expression was analysed by one-way anova. P values of 0.05 were accepted as statistically significant.

Clinical data

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

The 20 tumours included in the study were from 18 male and two female patients, with a mean age of 66 years (range, 50–86 years). Eight tumours were stage I, five stage II, one stage III and six stage IV. Six of the tumours were well differentiated, nine were moderately differentiated and five were poorly differentiated.

COX-2 expression in normal squamous epithelium

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

The normal squamous epithelium lying adjacent to Barrett's epithelium was studied. All 20 samples displayed relatively weak COX-2 immunostaining, predominantly in the basal and spinous layers of the stratified squamous epithelium (Figure 1a). COX-2 was located within the cytoplasm of these cells. The mean COX-2 immunostaining score was 25.8 AU, being significantly lower than that for Barrett's oesophagus (P = 0.044), high-grade dysplasia (P < 0.0001) and oesophageal adenocarcinoma (P = 0.0004) (Figure 2).

image

Figure 1. Immunostaining for cyclo-oxygenase-2 (COX-2) in human oesophageal squamous epithelium, Barrett's metaplastic epithelium and Barrett's metaplasia–dysplasia–adenocarcinoma sequence in a typical patient. (a) COX-2 protein expression occurs predominantly in the basal and spinous layers of the squamous epithelium (original magnification, × 100). (b) In Barrett's metaplastic epithelium, COX-2 expression is more prominent towards the base of the glandular crypt (original magnification, × 100). COX-2 expression in a typical patient showing Barrett's metaplasia (c), high-grade dysplasia (d) and adenocarcinoma (e) (c–e: original magnification, × 400).

Download figure to PowerPoint

image

Figure 2. Cyclo-oxygenase-2 (COX-2) immunostaining scores for squamous epithelium, Barrett's metaplasia, high-grade dysplasia (HGD) and oesophageal adenocarcinoma (OAC) from 20 patients. COX-2 expression progressively increases from squamous epithelium to Barrett's metaplasia and then high-grade dysplasia, but decreases when oesophageal adenocarcinoma develops. High-grade dysplasia has a significantly higher COX-2 expression than squamous epithelium ( P  < 0.0001), Barrett's metaplasia ( P  = 0.0002) and oesophageal adenocarcinoma ( P  = 0.02). Squamous epithelium has a significantly lower COX-2 expression than Barrett's metaplasia ( P  = 0.044) and oesophageal adenocarcinoma ( P  = 0.0004). There is no statistically significant difference in COX-2 expression between Barrett's epithelium and oesophageal adenocarcinoma ( P  = 0.48). Data shown are means ± S.E.M.

Download figure to PowerPoint

COX-2 expression in Barrett's epithelium

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

All 20 samples of Barrett's epithelium without dysplasia displayed moderate cytoplasmic immunoreactivity for COX-2, but there was also evidence of microheterogeneity for COX-2. The epithelial cells located deeper in the glandular crypts displayed stronger COX-2 immunostaining compared with those located more superficially (Figure 1b). The mean COX-2 immunostaining score was 38.6 AU in Barrett's oesophagus. This analysis was taken from the epithelium located in the deep part of the glandular crypts. It was significantly lower than that for high-grade dysplasia (P = 0.0002) and significantly higher than that for squamous epithelium (P = 0.044), but not oesophageal adenocarcinoma (P = 0.483) (Figure 2).

COX-2 expression in high-grade dysplasia and adenocarcinoma

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

All 20 high-grade dysplasia samples had strong immunoreactivity for COX-2 (Figure 1d). However, there was also inter-patient heterogeneity in COX-2 immunostaining. In high-grade dysplasia, the mean COX-2 immunostaining score was 59.7 AU, which was significantly higher than that for Barrett's oesophagus (P = 0.0002) and oesophageal adenocarcinoma (P = 0.021) (Figure 2). Eighteen of the 20 (90%) adenocarcinomas had moderate immunoreactivity for COX-2 (Figure 1e). A marked inter-tumoral heterogeneity in COX-2 immunostaining was also noted, with some tumours having low COX-2 immunoreactivity. Within the tumour, there was also evidence of microheterogeneity in COX-2 immunostaining. The mean COX-2 immunostaining score from the 20 oesophageal adenocarcinoma samples was 45.6 AU, which was not significantly higher than that for Barrett's oesophagus (P = 0.483), but was significantly lower than that for high-grade dysplasia (P = 0.021) (Figure 2). Of the two oesophageal adenocarcinomas that did not express COX-2, one was a well-differentiated stage I tumour and the other was a moderately differentiated stage II tumour with lymph node metastasis.

Relationship between COX-2 expression and clinicopathological parameters

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

The relationship between COX-2 immunostaining and clinicopathological parameters was assessed for the 20 tumours (Table 1). There was no statistically significant correlation between the COX-2 immunostaining score and the TNM classification, overall stage or grade of the tumour.

Table 1.  Clinicopathological parameters and cyclo-oxygenase-2 (COX-2) expression in 20 tumours
 Mean COX-2 (AU)P
TNM
 T148.2 (n = 9)0.574
 T235.8 (n = 6) 
 T347.5 (n = 3) 
 T460.2 (n = 2) 
 N048.2 (n = 13)0.484
 N140.7 (n = 7) 
 M044.8 (n = 14)0.814
 M147.5 (n = 6) 
Tumour stage
 I51.1 (n = 8)0.625
 II33.9 (n = 5) 
 III48.8 (n = 1) 
 IV47.5 (n = 6) 
Tumour grade
 Well43.5 (n = 6)0.956
 Moderate45.9 (n = 9) 
 Poor47.7 (n = 5) 

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References

In the present study, we evaluated the expression of the COX-2 isoenzyme in the Barrett's metaplasia–dysplasia–adenocarcinoma sequence. Immunohistochemistry was preferred to immunoblotting in the analysis of COX-2 protein expression to minimize errors due to the presence of focal areas of heterogeneous COX-2 expression.20 Because the tumour stroma contains cells that frequently over-express COX-2, especially at sites of ulcers or necrosis, the COX-2 level in tumours may be over-estimated in studies where immunoblotting or reverse transcriptase polymerase chain reaction methods are used.19, 22 We chose to compare high-grade rather than low-grade dysplasia, so as to minimize the ambiguity due to the very high inter-observer variation which exists in the diagnosis of low-grade dysplasia.24, 25 We compared each level of the metaplasia–dysplasia–adenocarcinoma sequence within the same patients, thereby reducing the effects of inter-patient variability in COX-2 expression.19, 21, 22

COX-2 is constitutively expressed in the basal layers of the normal squamous oesophageal epithelium adjacent to the neoplastic tissue. However, a significantly higher COX-2 expression was observed in all Barrett's metaplastic tissues without dysplasia. This was seen especially in the cells at the base of the crypts, compared with those on the surface where the level of COX-2 expression was consistently lower. Negative18, 19 and positive20–22 COX-2 expression in Barrett's oesophagus without dysplasia has been reported. It may relate to the different sensitivity of the COX-2 antibody used. Recently, several COX-2 antibody preparations have been evaluated,26 and it was concluded that the monoclonal antibody used in our study provided the most specific and reproducible immunoreactivity.

A further increase in COX-2 expression occurred when Barrett's oesophagus progressed to high-grade dysplasia, but the level of COX-2 expression appeared to decline in oesophageal adenocarcinoma. There was no significant difference in COX-2 expression between Barrett's metaplastic and oesophageal adenocarcinoma tissues. These observations contrast with the findings of Shirvani et al., who reported that COX-2 expression progressively increased throughout the metaplasia–dysplasia–adenocarcinoma sequence.20 One possible explanation is a lack of differentiation between high- and low-grade dysplasia stages in their study, leading in effect to the lower COX-2 expression values in their dysplastic tissue group. Recently, it has been reported that the level of COX-2 expression in low-grade dysplasia is indistinguishable from that of Barrett's metaplasia, whereas there is a significantly higher COX-2 expression in high-grade dysplasia when compared with low-grade dysplasia or Barrett's oesophagus.21 Another possibility is that, as the tissues analysed by Shirvani et al. were unpaired, the high level of inter-patient variation in COX-2 expression may have led to an over-estimation in the adenocarcinoma group, which was small (n = 5).20 It should also be noted that our analysis for COX-2 in Barrett's oesophagus was focused on cells in the deep regions of the glandular crypts, where a higher level of COX-2 was seen when compared with those on the surface. Our analysis of adenocarcinoma tissues was also focused on the deeper layers of the tumour, but the gradient of COX-2 is less steep in tumour tissue,20 and this may explain the lack of a significant difference between Barrett's oesophagus and oesophageal adenocarcinoma.

When we focused our analysis on the 18 patients with COX-2-positive tumours, there was a similar progressive significant increase in COX-2 expression as squamous epithelium developed into Barrett's oesophagus (P = 0.022), and as Barrett's oesophagus developed into high-grade dysplasia (P < 0.0001), but no significant difference in COX-2 expression between high-grade dysplasia (60.1 AU) and oesophageal adenocarcinoma (50.7 AU) (P = 0.158). In addition, there was a statistically significant difference between Barrett's oesophagus and oesophageal adenocarcinoma (P = 0.049). This result is similar to that reported by Morris et al.21 Interestingly, all their oesophageal adenocarcinoma samples (n = 12) demonstrated positive COX-2 expression. Zimmermannet al. also reported a significant increase in COX-2 expression in oesophageal adenocarcinoma, but only in 21 of 27 oesophageal adenocarcinomas when compared with COX-2-negative Barrett's oesophagus epithelium.19 Hence, we concur with the findings of Morris et al. that the significant change in COX-2 expression occurs ‘down-stream’ during the progression from benign Barrett's metaplasia to high-grade dysplasia.21 In addition, we found that COX-2 expression peaked at high-grade dysplasia and then appeared to be reduced during the progression from high-grade dysplasia to oesophageal adenocarcinoma.

We did not find a significant correlation between COX-2 expression and the TNM classification, overall stage and grade of the tumour. Although our study size was probably too small to address this issue rigorously, this finding is consistent with that of previous studies.19, 22 However, Buskens et al. showed that patients with high COX-2 expression had a poorer survival rate after surgery for oesophageal adenocarcinoma, and were more likely to have locoregional recurrence and distant metastasis.22

COX-2 expression in high-grade dysplasia was significantly higher than that in the other three stages of Barrett's oesophageal carcinogenesis studied, suggesting that COX-2 plays a significant role in the transformation from benign to malignant cells. A similar finding was noted in the carcinogenesis of oesophageal squamous cell carcinoma, where COX-2 progressively increased from normal oesophageal epithelium to low-grade dysplasia, and peaked at high-grade dysplasia, before subsequently decreasing during progression to early and advanced squamous cell carcinoma.27 In the oesophagus, high-grade dysplasia lesions may already have irreversibly progressed. At least 50% of high-grade dysplasia lesions in Barrett's oesophagus have immediately adjacent oesophageal adenocarcinoma,28, 29 and high-grade dysplasia lesions have a four-fold to eight-fold greater risk of developing oesophageal adenocarcinoma,28–30 compared with low-grade dysplasia.29, 31, 32 In addition, long-term follow-up studies have revealed that squamous epithelial dysplasia in the oesophagus is associated with a high risk of developing squamous cell carcinoma, and 70% of patients with squamous dysplasia will subsequently develop squamous cell carcinoma.33–36

The over-expression of COX-2 in the early stages of carcinogenesis has significant implications on the malignant transformation of benign cells. Over-expression of COX-2 in permanently transfected rat intestinal epithelial cells led to phenotypical alterations typical of malignant cells.33–37 In addition, treatment with Sulindac inhibited the growth of rat intestinal epithelial cells by markedly increasing apoptosis. Hence, the up-regulation of COX-2 prolongs the survival of abnormal cells, favouring the accumulation of sequential genetic changes and increasing the risk of malignancy.

The over-expression of COX-2 in the early stages of Barrett's oesophageal carcinogenesis may carry important implications in relation to chemo-prevention. Selective COX-2 inhibition can normalize both proliferation and apoptosis in Barrett's epithelial cells in an in vitro model.38 COX-2 inhibition also suppresses growth and induces apoptosis in human oesophageal adenocarcinoma cells.39 Recently, a randomized, placebo-controlled, chemo-prevention study using a selective COX-2 inhibitor in a rat surgical model for oesophageal adenocarcinoma was reported.40 Selective COX-2 inhibition reduced the relative risk of development of oesophageal cancer by 55%. Selective COX-2 inhibition in Barrett's oesophagus patients was also associated with significant suppression of COX-2 expression, prostaglandin E2 release and cell proliferation in Barrett's tissues.41 These are compelling data suggesting that chemo-prevention may be a plausible option against dysplasia and oesophageal adenocarcinoma.

We conclude that increased expression of COX-2 occurs in Barrett's metaplasia, reaches a maximum in high-grade dysplasia, and diminishes significantly with the transition to adenocarcinoma. The potential use of selective COX-2 inhibitors to arrest the transition from Barrett's metaplasia to high-grade dysplasia may be an effective chemo-preventive strategy that seems worthy of further investigation.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Patients
  6. Histopathological analysis
  7. Immunohistochemistry for COX-2
  8. Image analysis for COX-2
  9. Statistics
  10. Results
  11. Clinical data
  12. COX-2 expression in normal squamous epithelium
  13. COX-2 expression in Barrett's epithelium
  14. COX-2 expression in high-grade dysplasia and adenocarcinoma
  15. Relationship between COX-2 expression and clinicopathological parameters
  16. Discussion
  17. Acknowledgements
  18. References
  • 1
    Watson A. Barrett's oesophagus — 50 years on. Br J Surg 2000; 87: 52931.
  • 2
    Jankowski J. CRC Cancerstats: Oesophageal Cancer. London: Cancer Research Campaign Press, 2001.
  • 3
    Blot WJ, Devesa SS, Kneller RW, Fraumeni JF Jr. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA 1991; 265(10): 12879.
  • 4
    Lagergren J, Bergstrom R, Lindgren A, Nyren O. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med 1999; 340(11): 82531.
  • 5
    Jankowski JA, Wright NA, Meltzer SJ, et al. Molecular evolution of the metaplasia–dysplasia–adenocarcinoma sequence in the esophagus. Am J Pathol 1999; 154(4): 96573.
  • 6
    Spechler SJ, Goyal RK. Barrett's esophagus. N Engl J Med 1986; 315(6): 36271.
  • 7
    Dent J, Bremner CG, Collen MJ, Haggitt RC, Spechler SJ. Barrett's oesophagus. J Gastroenterol Hepatol 1991; 6(1): 122.
  • 8
    Thun MJ, Namboodiri MM, Calle EE, Flanders WD, Heath CW Jr. Aspirin use and risk of fatal cancer. Cancer Res 1993; 53(6): 13227.
  • 9
    Funkhouser EM, Sharp GB. Aspirin and reduced risk of esophageal carcinoma. Cancer 1995; 76(7): 11169.
  • 10
    Farrow DC, Vaughan TL, Hansten PD, et al. Use of aspirin and other nonsteroidal anti-inflammatory drugs and risk of esophageal and gastric cancer. Cancer Epidemiol Biomarkers Prev 1998; 7(2): 97102.
  • 11
    Langman MJ, Cheng KK, Gilman EA, Lancashire RJ. Effect of anti-inflammatory drugs on overall risk of common cancer: case–control study in general practice research database. Br Med J 2000; 320(7250): 16426.
  • 12
    Jones DA, Carlton DP, McIntyre TM, Zimmerman GA, Prescott SM. Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem 1993; 268(12): 904954.
  • 13
    DuBois RN, Awad J, Morrow J, Roberts LJ2nd, Bishop PR. Regulation of eicosanoid production and mitogenesis in rat intestinal epithelial cells by transforming growth factor-alpha and phorbol ester. J Clin Invest 1994; 93(2): 4938.
  • 14
    Inoue H, Yokoyama C, Hara S, Tone Y, Tanabe T. Transcriptional regulation of human prostaglandin-endoperoxide synthase-2 gene by lipopolysaccharide and phorbol ester in vascular endothelial cells. Involvement of both nuclear factor for interleukin-6 expression site and cAMP response element. J Biol Chem 1995; 270(42): 24 965–71.
  • 15
    Subbaramaiah K, Telang N, Ramonetti JT, et al. Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Res 1996; 56(19): 44249.
  • 16
    Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 1998; 93(5): 70516.
  • 17
    Tsujii M, Kawano S, DuBois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 1997; 94(7): 333640.
  • 18
    Wilson KT, Fu S, Ramanujam KS, Meltzer SJ. Increased expression of inducible nitric oxide synthase and cyclooxygenase-2 in Barrett's esophagus and associated adenocarcinomas. Cancer Res 1998; 58(14): 292934.
  • 19
    Zimmermann KC, Sarbia M, Weber AA, Borchard F, Gabbert HE, Schror K. Cyclooxygenase-2 expression in human esophageal carcinoma. Cancer Res 1999; 59(1): 198204.
  • 20
    Shirvani VN, Ouatu-Lascar R, Kaur BS, Omary MB, Triadafilopoulos G. Cyclooxygenase 2 expression in Barrett's esophagus and adenocarcinoma: ex vivo induction by bile salts and acid exposure. Gastroenterology 2000; 118(3): 48796.
  • 21
    Morris CD, Armstrong GR, Bigley G, Green H, Attwood SE. Cyclooxygenase-2 expression in the Barrett's metaplasia–dysplasia–adenocarcinoma sequence. Am J Gastroenterol 2001; 96(4): 9906.
  • 22
    Buskens CJ, Van Rees BP, Sivula A, et al. Prognostic significance of elevated cyclooxygenase 2 expression in patients with adenocarcinoma of the esophagus. Gastroenterology 2002; 122(7): 18007.
  • 23
    Riddell RH, Goldman H, Ransohoff DF, et al. Dysplasia in inflammatory bowel disease: standardized classification with provisional clinical applications. Hum Pathol 1983; 14(11): 93168.
  • 24
    Reid BJ, Haggitt RC, Rubin CE, et al. Observer variation in the diagnosis of dysplasia in Barrett's esophagus. Hum Pathol 1988; 19(2): 16678.
  • 25
    Skacel M, Petras RE, Gramlich TL, Sigel JE, Richter JE, Goldblum JR. The diagnosis of low-grade dysplasia in Barrett's esophagus and its implications for disease progression. Am J Gastroenterol 2000; 95(12): 33837.
    Direct Link:
  • 26
    Saukkonen K, Nieminen O, Van Rees B, et al. Expression of cyclooxygenase-2 in dysplasia of the stomach and in intestinal-type gastric adenocarcinoma. Clin Cancer Res 2001; 7(7): 192331.
  • 27
    Shamma A, Yamamoto H, Doki Y, et al. Up-regulation of cyclooxygenase-2 in squamous carcinogenesis of the esophagus. Clin Cancer Res 2000; 6(4): 122938.
  • 28
    Reid BJ, Blount PL, Rubin CE, Levine DS, Haggitt RC, Rabinovitch PS. Flow-cytometric and histological progression to malignancy in Barrett's esophagus: prospective endoscopic surveillance of a cohort. Gastroenterology 1992; 102(4 Part 1): 12129.
  • 29
    Levine DS, Haggitt RC, Blount PL, Rabinovitch PS, Rusch VW, Reid BJ. An endoscopic biopsy protocol can differentiate high-grade dysplasia from early adenocarcinoma in Barrett's esophagus. Gastroenterology 1993; 105(1): 4050.
  • 30
    Morales TG, Bhattacharyya A, Johnson C, Sampliner RE. Is Barrett's esophagus associated with intestinal metaplasia of the gastric cardia? Am J Gastroenterol 1997; 92(10): 181822.
  • 31
    Altorki NK, Sunagawa M, Little AG, Skinner DB. High-grade dysplasia in the columnar-lined esophagus. Am J Surg 1991; 161(1): 97100.
  • 32
    Wright TA. High-grade dysplasia in Barrett's oesophagus. Br J Surg 1997; 84(6): 7606.
  • 33
    Qiu SL, Yang GR. Precursor lesions of esophageal cancer in high-risk populations in Henan Province, China. Cancer 1988; 62(3): 5517.
  • 34
    Nagamatsu M, Mori M, Kuwano H, Sugimachi K, Akiyoshi T. Serial histologic investigation of squamous epithelial dysplasia associated with carcinoma of the esophagus. Cancer 1992; 69(5): 10948.
  • 35
    Dawsey SM, Lewin KJ, Wang GQ, et al. Squamous esophageal histology and subsequent risk of squamous cell carcinoma of the esophagus. A prospective follow-up study from Linxian, China. Cancer 1994; 74(6): 168692.
  • 36
    Morita M, Kuwano H, Yasuda M, et al. The multicentric occurrence of squamous epithelial dysplasia and squamous cell carcinoma in the esophagus. Cancer 1994; 74(11): 288995.
  • 37
    Tsujii M, DuBois RN. Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 1995; 83(3): 493501.
  • 38
    Buttar NS, Wang KK, Anderson MA, et al. The effect of selective cyclooxygenase-2 inhibition in Barrett's esophagus epithelium: an in vitro study. J Natl Cancer Inst 2002; 94(6): 4229.
  • 39
    Souza RF, Shewmake K, Beer DG, Cryer B, Spechler SJ. Selective inhibition of cyclooxygenase-2 suppresses growth and induces apoptosis in human esophageal adenocarcinoma cells. Cancer Res 2000; 60(20): 576772.
  • 40
    Buttar NS, Wang KK, Leontovich O, et al. Chemoprevention of esophageal adenocarcinoma by COX-2 inhibitors in an animal model of Barrett's esophagus. Gastroenterology 2002; 122(4): 110112.
  • 41
    Kaur BS, Khamnehei N, Iravani M, Namburu SS, Lin O, Triadafilopoulos G. Rofecoxib inhibits cyclooxygenase 2 expression and activity and reduces cell proliferation in Barrett's esophagus. Gastroenterology 2002; 123(1): 607.