Article Text

Original article
Aberrant p53 protein expression is associated with an increased risk of neoplastic progression in patients with Barrett's oesophagus
  1. Florine Kastelein1,
  2. Katharina Biermann2,
  3. Ewout W Steyerberg3,
  4. Joanne Verheij4,
  5. Marit Kalisvaart1,
  6. Leendert H J Looijenga2,
  7. Hans A Stoop2,
  8. Laurens Walter2,
  9. Ernst J Kuipers1,5,
  10. Manon C W Spaander1,
  11. Marco J Bruno1,
  12. on behalf of the ProBar-study group
  1. 1Department of Gastroenterology and Hepatology, Erasmus University Medical Center, Rotterdam, The Netherlands
  2. 2Department of Pathology, Erasmus University Medical Center, Rotterdam, The Netherlands
  3. 3Department of Public Health, Erasmus University Medical Center, Rotterdam, The Netherlands
  4. 4Department of Pathology, Academic Medical Center, Amsterdam, The Netherlands
  5. 5Department if Internal Medicine, Erasmus University Medical Center, Rotterdam, The Netherlands
  1. Correspondence to Florine Kastelein, Department of Gastroenterology and Hepatology, Erasmus Medical Center, PO Box 2040, Rotterdam 3000 CA, The Netherlands; f.kastelein{at}erasmusmc.nl

Abstract

Objective The value of surveillance for patients with Barrett's oesophagus (BO) is under discussion given the overall low incidence of neoplastic progression and lack of discriminative tests for risk stratification. Histological diagnosis of low-grade dysplasia (LGD) is the only accepted predictor for progression to date, but has a low predictive value. The aim of this study was therefore to evaluate the value of p53 immunohistochemistry for predicting neoplastic progression in patients with BO.

Design We conducted a case–control study within a prospective cohort of 720 patients with BO. Patients who developed high-grade dysplasia (HGD) or oesophageal adenocarcinoma (OAC) were classified as cases and patients without neoplastic progression were classified as controls. P53 protein expression was determined by immunohistochemistry in more than 12 000 biopsies from 635 patients and was scored independently by two expert pathologists who were blinded to long-term outcome.

Results During follow-up, 49 (8%) patients developed HGD or OAC. P53 overexpression was associated with an increased risk of neoplastic progression in patients with BO after adjusting for age, gender, Barrett length and oesophagitis (adjusted relative risks (RRa) 5.6; 95% CI 3.1 to 10.3), but the risk was even higher with loss of p53 expression (RRa 14.0; 95% CI 5.3 to 37.2). The positive predictive value for neoplastic progression increased from 15% with histological diagnosis of LGD to 33% with LGD and concurrent aberrant p53 expression.

Conclusions Aberrant p53 protein expression is associated with an increased risk of neoplastic progression in patients with BO and appears to be a more powerful predictor of neoplastic progression than histological diagnosis of LGD.

  • Barrett's Oesophagus
  • Barrett's Carcinoma
  • Barrett's Metaplasia

Statistics from Altmetric.com

Significance of this study

What is already known about this subject?

  • Low-grade dysplasia (LGD) is the only accepted predictor for neoplastic progression in Barrett's oesophagus (BO) but has a low predictive value.

  • Use of biomarkers such as p53 may improve risk stratification.

  • P53 overexpression is related to neoplastic progression.

What are the new findings?

  • P53 overexpression is associated with an increased risk of neoplastic progression in BO, but the risk is even higher with loss of p53 expression.

  • P53 immunohistochemistry appears to be a more powerful predictor for neoplastic progression in BO than histological diagnosis of LGD.

  • Interobserver agreement for p53 expression is very good, indicating that p53 is a clinically suitable biomarker for predicting neoplastic progression.

How might it impact on clinical practice in the foreseeable future?

  • P53 immunohistochemistry may be useful as a discriminative test to improve risk stratification and hence the cost effectiveness of Barrett surveillance programmes.

Introduction

Over the past decades, the incidence of oesophageal adenocarcinoma (OAC) has been rising at a rate exceeding that of any other cancer.1 In many cases the development of OAC is related to Barrett's oesophagus (BO), a premalignant condition in which the normal squamous epithelium of the distal oesophagus is replaced by columnar epithelium-containing goblet cells.2 Patients with BO have an increased risk of developing OAC with an estimated incidence of 0.5% per year.3 ,4 The development of OAC is thought to be a gradual process, in which metaplastic epithelium evolves to low-grade dysplasia (LGD), high-grade dysplasia (HGD) and eventually OAC under the influence of chronic oesophageal acid exposure.5 ,6 Early identification of a premalignant stage provides the opportunity to prevent progression to BO-related adenocarcinoma and endoscopic surveillance is therefore recommended for patients with BO.7 However, the value of BO surveillance is under discussion given the overall low incidence of neoplastic progression, the large screening base which is estimated at 1–2% of the general population, and lack of discriminative tests for risk stratification.8 ,9

Histological diagnosis of LGD is currently the only accepted predictor for neoplastic progression and international guidelines recommend more intensive surveillance in patients with BO with LGD (yearly instead of every 3 years).7 ,10 However, histological diagnosis of LGD is subject to sample error and considerable interobserver variation, mainly because features of dysplasia may overlap with features of non-neoplastic regenerative changes.11 Although the predictive value of LGD increases with consensus of multiple pathologists, still one-third of patients with BO are diagnosed with LGD during surveillance, while the 10-year cumulative incidence of neoplastic progression is only around 15% in this subgroup.12 ,13 Use of biomarkers in addition to histological assessment may improve risk stratification for these patients, whereby more stringent surveillance is applied to individuals at high risk for neoplastic progression, while surveillance intervals are prolonged in those at low risk. P53 appears to be one of the most promising biomarkers and previous studies have shown that p53 overexpression is associated with an increased risk of neoplastic progression.12–18 P53 overexpression can be caused by TP53 mutations which stabilise the inactivated protein. However, truncating TP53 mutations or epigenetic silencing may result in protein inactivation and subsequently loss of p53 expression. Although little is known about loss of p53 expression, the first results indicate that it is also associated with an increased risk of neoplastic progression.19 The aim of the present study was therefore to evaluate the value of p53 immunohistochemistry for predicting neoplastic progression in patients with BO.

Methods

Study design

We conducted a case–control study within a large prospective cohort of patients with BO. In this cohort 720 patients were included with known or newly diagnosed BO of at least 2 cm, confirmed by the presence of intestinal metaplasia and without a history of HGD or OAC. Patients were included between November 2003 and December 2004 from three university medical centres and 12 regional hospitals throughout the Netherlands and had endoscopic surveillance according to the guidelines of the American College of Gastroenterology.7 Patients without dysplasia underwent gastroscopy with biopsy sampling every 3 years and patients with LGD underwent testing every year. All endoscopic procedures were performed by experienced gastroenterologists, according to a standardised protocol. Endoscopic landmarks such as the diaphragm impression, gastroesophageal junction and squamocolumnar junction were noted, the presence of oesophagitis was graded according to the Los Angeles Classification, and abnormalities were reported, including nodules, ulcers and erosions.20 At each endoscopy targeted biopsies were taken from mucosal abnormalities and quadrant biopsies were taken every 2 cm from the most distal to the most proximal part of the Barrett's epithelium, according to the Seattle protocol.21

Histology

Biopsy specimens were fixed with buffered formalin and embedded in paraffin, according to standard procedures. From each biopsy set 4 µm thick sections were cut and stained with haematoxylin eosin to assess the presence of BO and to define the grade of dysplasia. After examining all biopsies, the highest degree of abnormality was reported for each endoscopy. Slides were first graded by a local pathologist and then by an expert pathologist for a second opinion. When the local and expert pathologists disagreed on the grade of dysplasia, the slides were reviewed by a second expert pathologist. Pathologists were blinded to the diagnosis of each other and a final diagnosis was made only if at least two pathologists agreed on the grade of dysplasia. If there was still disagreement, a panel of expert pathologists reviewed the slides as well and a final diagnosis was made based on consensus agreement.

Patient selection

We collected paraffin material suitable for immunohistochemistry from all patients in our BO cohort. Paraffin material was not available in 85 patients, leaving 635 patients to be included in this analysis. Patients who developed HGD or OAC during follow-up were identified as cases and patients without neoplastic progression were identified as controls (figure 1). Immunohistochemistry was performed on paraffin material of all surveillance endoscopies of patients who developed any form of dysplasia. In patients without any dysplasia, immunohistochemistry was performed on biopsies of a random surveillance endoscopy.

Figure 1

Flowchart of patients included in the study. Patients with neoplastic progression were classified as cases and patients without progression were classified as controls.

Immunohistochemistry

Immunohistochemistry was performed as a single batch at the pathology department of the Erasmus University Medical Center (Rotterdam, the Netherlands) using an automatic immunohistochemical staining machine (Ventana Medical Systems, Tucson, Arizona, USA). A sample of tumour tissue was used as positive control for each section. Sections were deparaffinised prior to the staining procedure and heat-induced epitope retrieval was performed at 97°C for 15 min. Endogenous peroxidase activity was blocked by incubating the slides for 15 min in a solution of 0.3% hydrogen peroxide in phosphate-buffered saline. Monoclonal mouse anti-human p53 protein was used as the primary antibody for immunohistochemistry with a dilution of 1:25 (Clone DO-7, Dako, Glostrup, Denmark). The slides were incubated for 30 min with the primary antibody. Then amplification and visualisation was performed by using the Dako REAL EnVision system (peroxidase/DAB, Rabbit/Mouse, Dako, Glostrup, Denmark). Finally, the slides were counterstained with haematoxylin.

Immunohistochemically stained slides were examined in tandem with haematoxylin eosin stained slides to determine p53 expression in areas with dysplasia. P53 expression was scored independently by two expert pathologists who were blinded to long-term outcome on a three-point scale (normal expression, overexpression or complete loss of expression). Only intense nuclear staining for p53 was considered as overexpression (figure 2). Aberrant expression was defined as either p53 overexpression or complete loss of p53 expression. P53 protein expression was considered as aberrant when at least one gland showed overexpression or complete loss of expression. After examining all biopsies, the highest degree of abnormality was reported for each endoscopy. When there was disagreement between the pathologists, the slides were evaluated by both pathologists simultaneously to reach a consensus diagnosis.

Figure 2

Haematoxylin-eosin staining and p53 immunohistochemistry of (A) Barrett's oesophagus (BO) with low-grade dysplasia (LGD) and normal p53 expression, (B) BO with LGD and p53 overexpression and (C) oesophageal adenocarcinoma with loss of p53 expression. Access the article online to view this figure in colour.

Ethics

The study protocol was approved by the institutional review board of Erasmus University Medical Center, and those of all participating hospitals. Before the first endoscopy, written informed consent was obtained from all patients.

Statistical analysis

Characteristics of cases and controls were compared using Mann–Whitney U tests for continuous variables and χ2 tests for categorical variables. To compare p53 expression in biopsies with different grades of dysplasia, Mann–Whitney U tests and Kruskal–Wallis tests were used, thereby ignoring that multiple biopsies could be from the same patient. The value of p53 immunohistochemistry for predicting neoplastic progression was estimated in loglinear regression models. Since immunohistochemistry was not performed on all biopsy series, data were split up by endoscopy. Neoplastic progression was defined as the development of HGD or OAC after inclusion in the study and follow-up time was defined as the time between each endoscopy and the next surveillance endoscopy. Loglinear regression models were used to calculate relative risks (RRs) and 95% CIs with the logarithm of follow-up time as offset variable. In multivariable models, RRs were calculated adjusted for age, gender, BO length and oesophagitis. With 49 cases and 586 controls, 80% power was provided to detect a RR of at least 2.5 at a significance level of 5%. Interobserver agreement for p53 expression was determined using Cohen κ statistics. Two-sided p values <0.05 were considered to be statistically significant. Data were analysed using SPSS Statistics (V.20.0).

Results

Patient characteristics

Six hundred and thirty-five patients with BO (73% men, median age 60 years (IQR 53–69)) were included in this study and followed during surveillance for a median duration of 6.6 years (IQR 5.1–7.3) and with a median of four follow-up endoscopies (IQR 4–5). Thirty-five (6%) patients developed HGD and 14 (2%) patients developed OAC during surveillance, resulting in 49 (8%) patients with neoplastic progression which were identified as cases. The remaining 586 (92%) patients without neoplastic progression were identified as controls (figure 1). The incidence rate of HGD and OAC together was 1.4 per 100 patient-years (95% CI 1.0 to 1.8) and the incidence rate of OAC alone was 0.4 per 100 patient-years (95% CI 0.2 to 0.6). Histology and p53 expression were assessed in biopsy series of 1481 endoscopies. The highest degree of abnormality was reported for each endoscopy after examining all biopsies. In total, more than 12 000 biopsies were reviewed. Cases had a smaller number of follow-up endoscopies and longer Barrett length than controls, but otherwise there were no significant differences (table 1).

Table 1

Characteristics of cases and controls

Histology

Normal BO without dysplasia was seen in 1085 (73%) biopsy series, LGD in 347 (23%), HGD in 35 (3%) and OAC in 14 (1%). Presence of LGD was more common in biopsy series of cases (44%) than in biopsy series of controls (22%) and was associated with an increased risk of neoplastic progression (RR 4.2; 95% CI 2.4 to 7.3). This association remained after adjusting for age, gender, BO length and oesophagitis (RRa 4.0; 95% CI 2.3 to 7.0) (table 2).

Table 2

Histology and p53 immunohistochemistry in biopsy series of cases and controls

In total 223 (35%) patients were diagnosed with LGD during follow-up with an incidence rate of 3.6 per 100 patient-years (95% CI 3.0 to 4.4). Of these patients, 34 (15%) eventually developed HGD or OAC with an incidence rate of 4.2 per 100 patient-years (95% CI 2.8 to 5.9). The percentage of patients with LGD gradually increased from 5 years before neoplastic progression (figure 4). The sensitivity of LGD for predicting neoplastic progression was 44% with a specificity of 78%.

P53 immunohistochemistry

Normal p53 expression was seen in 1188 (80%) biopsy series, p53 overexpression in 262 (18%) and loss of p53 expression in 31 (2%). Aberrant p53 expression was more common with higher grades of dysplasia and was seen in 11% of biopsy series without dysplasia, 38% of biopsy series with LGD, 83% of biopsy series with HGD and all biopsy series with OAC (p<0.001) (figure 3). Loss of p53 expression was especially seen in biopsy series with HGD (6%) and OAC (43%). Aberrant p53 expression was more common in biopsy series of cases (49%) than in biopsy series of controls (14%) and was associated with an increased risk of neoplastic progression (RR 6.2; 95% CI 3.6 to 10.9). This association remained after adjusting for age, gender, BO length and oesophagitis (RRa 6.4; 95% CI 3.6 to 11.3) and was seen in biopsy series without dysplasia and biopsy series with LGD. P53 overexpression and loss of p53 expression were associated with an increased risk of neoplastic progression (RRa 5.6; 95% CI 3.1 to 10.3 and RRa 14.0; 95% CI 5.3 to 37.2) (table 2).

Figure 3

Percentage of biopsy series with aberrant p53 expression, stratified by grade of dysplasia. Embedded Image p53 overexpression Embedded Image loss of p53 expression. The highest degree of abnormality was reported for each endoscopy after examining all biopsies.

In total 118 (19%) patients were diagnosed with aberrant p53 expression during follow-up. Of these patients, 31 (26%) eventually developed HGD or OAC with an incidence rate of 7.4 per 100 patient-years (95% CI 5.0 to 10.5). During follow-up, aberrant p53 expression was confirmed in 37% of biopsy series without dysplasia, 78% of biopsy series with LGD and all biopsy series with HGD or OAC. In approximately 45% of patients, aberrant p53 expression was already seen up to 5 years before neoplastic progression and this percentage remained stable over time (figure 4). The sensitivity of aberrant p53 expression for predicting neoplastic progression was 49% with a specificity of 86%. Interobserver agreement for p53 expression was good (κ=0.79; 95% CI 0.75 to 0.83). Both expert pathologists agreed on p53 expression in 1379 (93%) biopsy series (table 3). When p53 expression was scored on a two-point scale (normal or aberrant expression) interobserver agreement was similar.

Table 3

Interobserver agreement for p53 expression between two expert pathologists

Figure 4

Percentage of patients with (A) low-grade dysplasia (LGD) or (B) aberrant p53 expression before neoplastic progression. Embedded Image cases Embedded Image controls.

Histology and p53 immunohistochemistry

Normal BO with normal p53 expression was seen in 968 (68%) biopsy series, LGD with normal p53 expression in 214 (15%), normal BO with aberrant p53 expression in 117 (8%) and LGD with aberrant p53 expression in 133 (9%). Aberrant p53 expression was more common in biopsy series of cases than in biopsy series of controls, in normal BO (18% vs 7%) and in BO with LGD (31% vs 7%). Aberrant p53 expression in normal BO was associated with an increased risk of neoplastic progression (RRa 4.3; 95% CI 1.9 to 9.8), but the risk was even higher with concurrent LGD (RRa 12.2; 95% CI 6.1 to 24.5).

During follow-up, 73 (11%) patients were diagnosed with LGD and concurrent aberrant p53 expression. Of these patients 24 (33%) eventually developed HGD or OAC with an incidence rate of 11.2 per 100 patient-years (95% CI 7.1 to 16.8).

Discussion

In this case–control study, we evaluated the value of p53 immunohistochemistry for predicting neoplastic progression in patients with BO. P53 overexpression was associated with an increased risk of neoplastic progression, but the risk was even higher with complete loss of p53 expression. Although aberrant p53 expression appeared to be a more powerful predictor than histological diagnosis of LGD, the risk of neoplastic progression was the highest in patients with LGD and concurrent aberrant p53 expression.

During surveillance, up to 35% of patients were diagnosed with LGD, while only 15% of these patients eventually developed HGD or OAC. The predictive value of LGD was thus low, despite using a consensus diagnosis for dysplasia. The incidence rate of LGD was 3.6% per year in our study, which is similar to rates observed previously.22 ,23 Although patients with LGD were at increased risk of neoplastic progression, the absolute risk of developing HGD or OAC was low with an incidence rate of 4.2% per year. The results of previous studies are highly variable, but show an average incidence rate of 1–2% per year (range 0.6–13%), which is only slightly lower than the incidence rate observed in our study.24–29 Aberrant p53 expression was observed more frequently with higher grades of dysplasia. The percentage of biopsy series with aberrant p53 expression increased from 11% in samples without dysplasia to even 100% in samples with OAC. These findings correspond to results reported in previous studies.30–33 Aberrant p53 expression was identified more frequently in cases than in controls and was associated with an increased risk of neoplastic progression. P53 overexpression was associated with an increased risk of developing HGD or OAC and the risk was even higher with complete loss of p53 expression. The positive predictive value for neoplastic progression increased from 15% with histological diagnosis of LGD to 33% with LGD and concurrent aberrant p53 expression. To our knowledge this is the first large case–control study evaluating the value of p53 overexpression and loss of p53 expression for predicting neoplastic progression in BO. Previous studies have shown that p53 overexpression is associated with an increased risk of neoplastic progression in non-dysplastic BO and BO with LGD. The results of these studies are in line with the results of our study.12 ,13 ,16 ,17 ,18 ,34

In this study we have also shown good interobserver agreement for the assessment of p53 expression, which indicates that p53 is a theoretical and a clinically suitable marker for predicting progression in BO. Although routine p53 immunohistochemistry is associated with higher costs than histological assessment alone, application of this marker may lead to the identification of a much smaller high-risk group needing intensive surveillance. At this moment all patients with LGD receive intensive surveillance, which can be up to one-third of all patients with BO. In this study we have shown that only 11% of patients are diagnosed with LGD and aberrant p53 expression and that the risk of neoplastic progression is much higher in this subgroup. Surveillance of such a small high-risk group may eventually result in lower costs of surveillance, less burden on endoscopy units and higher quality of life for patients with BO.

Previous studies have shown that p53 is an early molecular marker of genetic instability and may precede the development of dysplasia. In addition, studies have shown that aberrant p53 expression can be detected in non-dysplastic epithelium of patients with dysplasia.33 This may explain why aberrant p53 expression was also associated with an increased risk of neoplastic progression in biopsy series without dysplasia. Although aberrant p53 expression was more common in cases it was also seen in controls. The development of OAC in BO is a gradual process and it is unknown how much time this process of neoplastic progression takes.5 It is therefore possible that patients with aberrant p53 expression but without neoplastic progression will develop HGD or OAC in the future. However, not all patients with neoplastic progression showed aberrant p53 expression. It may be that these patients actually have mutations in p53, but that these mutations may not lead to accumulation or complete loss of p53 in the cell nucleus.

This study has several strengths including the large cohort of patients with BO and long follow-up time. Patients were prospectively followed according to a stringent follow-up scheme and during follow-up clinical and pathological data were collected. In addition, a standardised endoscopy and biopsy protocol were used. All slides were reviewed by at least two pathologists to obtain a final diagnosis based on consensus. In contrast to previous studies, the slides were not only evaluated for p53 overexpression, but also for loss of p53 expression.35

Our study also has some limitations. Although patients were only classified as controls when they did not develop HGD or OAC during follow-up, we cannot exclude that these patients will develop HGD or OAC in the future. If so, this may have lead to an underestimation of the value of p53 for predicting progression. In addition, a disadvantage of p53 immunohistochemistry is that the antibody directed to p53 stains protein derived from the mutant TP53 and wild-type TP53. Nevertheless, p53 expression is considered to be indicative for the presence of mutant TP53, because the latter has a longer half life than wild-type p53 and is not degraded in the normal way. This results in accumulation of p53 in the cell nucleus, which is detectable by immunohistochemistry.36

In conclusion, this case–control study shows that aberrant p53 expression is associated with an increased risk of neoplastic progression in patients with BO. Aberrant p53 expression appears to be a more powerful predictor for neoplastic progression than histological diagnosis of LGD. P53 immunohistochemistry may be useful as a discriminative test to improve risk stratification and hence the cost-effectiveness of Barrett surveillance programmes.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors All authors listed have contributed substantially to the design, data collection and analysis, and editing of the manuscript.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Medical Ethics Committee Erasmus University Medical Center.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.