http://orcid.org/0000-0002-9103-9757
Figures
Abstract
Introduction
The low incidence of oesophageal adenocarcinoma (EAC) in Barrett's oesophagus (BE) patients reinforces the need for risk stratification tools to make BE surveillance more effective. Therefore, we have undertaken a systematic review and meta-analysis of published studies on immunohistochemical (IHC) biomarkers in BE to determine the value of IHC biomarkers as neoplastic predictors in BE surveillance.
Materials and methods
We searched MEDLINE, EMBASE, Web of Science, CENTRAL, Pubmed publisher, and Google scholar. All studies on IHC biomarkers in BE surveillance were included. ORs were extracted and meta-analyses performed with a random effects model.
Results
16 different IHC biomarkers were studied in 36 studies. These studies included 425 cases and 1835 controls. A meta- analysis was performed for p53, aspergillus oryzae lectin (AOL), Cyclin A, Cyclin D and alpha-methylacyl-CoA racemase. Aberrant p53 expression was significantly associated with an increased risk of neoplastic progression with an OR of 3.18 (95% CI 1.68 to 6.03). This association was confirmed for both non-dysplastic BE and BE with low-grade dysplasia (LGD). Another promising biomarker to predict neoplastic progression was AOL, with an OR of 3.04 (95% CI 2.05 to 4.49).
Citation: Janmaat VT, van Olphen SH, Biermann KE, Looijenga LHJ, Bruno MB, Spaander MCW (2017) Use of immunohistochemical biomarkers as independent predictor of neoplastic progression in Barrett's oesophagus surveillance: A systematic review and meta-analysis. PLoS ONE 12(10): e0186305. https://doi.org/10.1371/journal.pone.0186305
Editor: John Green, University Hospital Llandough, UNITED KINGDOM
Received: July 18, 2017; Accepted: September 28, 2017; Published: October 23, 2017
Copyright: © 2017 Janmaat et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Development of oesophageal adenocarcinoma (EAC) is related to Barrett's oesophagus (BE), a premalignant condition of the distal oesophagus. In BE, the pre-existent squamous epithelium is replaced by columnar epithelium which develops under the influence of chronic acid and bile reflux and frequently contains goblet cells [1–3]. The progression from BE to EAC is a gradual process, in which intestinal metaplasia (IM) evolves to low-grade dysplasia (LGD), high-grade dysplasia (HGD) and eventually EAC [4]. Therefore, current guidelines recommend endoscopic surveillance in BE patients to detect HGD or EAC at an early stage, with the aim to improve survival rates [5, 6]. Several studies have shown that patients diagnosed with EAC during BE surveillance have earlier staged tumors and probably better survival compared to those diagnosed after the onset of symptoms [7–10].
The estimated incidence of EAC in patients with BE was reported to be between 0.5 and 1% per year [11–14]. However, more recent population-based studies and two meta-analyses have set this risk around 0.12% to 0.38% per year [15–18]. This relatively low annual risk reinforces the need for risk stratification tools to make BE surveillance more effective. BE length, male gender, smoking, and LGD are known risk factors for progression to HGD and EAC [13, 15, 18–20]. Two large population studies confirmed that patients with LGD have an approximately five times higher risk of progression compared to patients with non-dysplastic BE [15, 18]. Thus, more intensive surveillance is recommended in BE patients with LGD [5, 6]. However, the histological diagnosis of LGD is subject to a considerable inter- and intra-observer variation, because of sample error and overlap with features of non-neoplastic regenerative changes [21–24].
Because none of the current clinical and histologic criteria are able to accurately predict which patients are likely to progress to HGD or EAC, there is an increasing interest in (molecular) biomarkers. Many immunohistochemical (IHC) biomarkers have been studied in BE progression, mainly because they can be applied to standard histological samples. In clinical practice, IHC biomarkers are relatively easily applicable compared to other techniques. Currently, the addition of p53 IHC to the histological assessment is recommended in the guideline of the British Society of Gastroenterology as it may improve the diagnostic reproducibility of a histological diagnosis of LGD [5]. The use of IHC biomarkers as independent predictor of neoplastic progression is not yet performed in routine clinical care, neither for p53, nor for other IHC biomarkers. Therefore, this study aims to provide a systematic review and meta-analyses of all retrospective case control or cohort studies and prospective cohort studies investigating IHC biomarkers as predictor of neoplastic progression in patients with BE.
Materials and methods
This review was conducted according to the PRISMA and MOOSE guidelines (S1 MOOSE checklist, S1 PRISMA checklist) [25, 26].
Definitions
BE was defined as columnar lined oesophagus (CLE). Neoplastic progression was defined as the development of HGD or EAC during follow up. Patients with neoplastic progression were classified as cases and patients without neoplastic progression as controls.
Data sources and searches
Records were identified by searching the following electronic databases: 1. EMBASE, 2. MEDLINE, 3. Web of Science, 4. CENTRAL, 5. PubMed Publisher, 6. Google scholar until 09-12-2016 (S1 Search). The search strategy was constructed by applying a sensitivity maximizing approach. A combination of MESH subject headings and text words were used related to IHC markers for progression in patients with BE. The search was confined to English language publications. Conference abstracts indexed in Embase from the years 2014–2016 were included in order to be able to include new and unpublished papers.
Study selection
Search results were combined and duplicates removed. Every article was screened on title and abstract level for relevance by a single author (SvO or VJ). Articles were reviewed full text by the same two independent authors and included if they met the following criteria: (1) association between IHC biomarker expression on formalin fixed paraffin embedded material and risk of neoplastic progression was assessed; (2) a cohort or case-control study design; (3) patients with known or newly diagnosed BE with or without LGD at baseline; (4) patients defined as cases had to have progressed to either HGD or EAC during follow-up; (5) mean follow-up of at least one year from the time of initial BE diagnosis; (6) the possibility to extract an OR. Studies were excluded if: (1) BE cohorts included patients with HGD at baseline; (2) endoscopic therapies affecting neoplastic progression were performed during follow-up (Fig 1). Some manuscripts studied different biomarkers within the same population, these were considered as individual studies on the level of the individual biomarker.
Data extraction
For each included study two independent authors extracted data according to a standardized data extraction form and assessed the quality of the eligible studies (S1 Standardized data extraction form). In case of disagreement consensus was reached by consulting a third author (MS). Odds ratio's (OR)s and 95% confidence intervals (CI)s of individual IHC biomarkers were extracted or estimated from the data. If ORs could not be extracted directly from the text or the tables, ORs were calculated indirectly by using the numbers of cases and controls with an aberrant versus a normal IHC biomarker expression from text, tables, or figures.
Quality assessment
The quality aspects defined were: a difference at baseline between cases and controls of at least 10% (concerning baseline histology, age, sex, length of BE segment, and follow–up time), adjustments in the form of regression for differences of known predictors of progression (such as baseline histology, age, sex, and length of BE segment), exclusion of prevalent cases, control stainings, number of pathologists, pathologist agreement and pathologist blinding. These aspects were assessed and reported on but not used as exclusion criteria.
Data synthesis and analyses
Meta-analysis were performed if at least two studies were available [27]. If multiple studies in a single analysis included the same patients, the oldest study was excluded. An inverse variance random-effect model was used. If data on multiple definitions of aberrant staining were available, definitions were chosen to resemble those from other included studies for that IHC biomarker. If only one study was available, definitions of the authors of that study were used. Pooled estimates of effect, in the form of ORs, were calculated and investigated for statistical heterogeneity by visual inspection and the I-squared test (I2) = [(Q-df)/Q]*100%, where Q was the chi squared statistic and df was its degree of freedom. Where possible, ORs adjusted for most factors were used in the analysis and unadjusted and adjusted ORs were pooled if necessary. Small study effects such as publication bias were assessed using a funnel plot.
Sensitivity and subgroup analyses
Sensitivity analyses were performed in case of a large standard error (if small study effects were likely present as observed in the funnel plot), if no adjustments were made for known predictors of progression (such as sex, age, histology, i.e. non-dysplastic or LGD, and BE-length), and if only an abstract was available. We excluded individual studies from the most reliable analysis to evaluate the impact of single studies on pooled risk estimates and heterogeneity. Additional analyses were performed to assess if an IHC biomarker can be used as a predictor of neoplastic progression, independent of the presence of dysplasia. Therefore, all studies were summarized which included only non-dysplastic BE patients, only BE with LGD patients, or in which adjustments were made for histology type. Additionally, two subgroup analyses were performed including either non-dysplasic or LGD BE patients.
Stringency of the definition for aberrant staining used and its interpretation
The stringency level of the definition for aberrant staining and its interpretation could lead to variation in the predictive ability of the IHC biomarker investigated. To investigate whether this effect might be present, the proportion of controls deemed positive was plotted against the OR of each study.
Results
Included studies
2081 records were retrieved, after removal of duplicates. After excluding 2050 records based on title and abstract, a total of 27 full text articles and four abstracts were assessed in detail (Fig 1). Of these, 19 full text articles and two abstracts were included in this review [28–49]. These articles contained a total of 36 studies.
Characteristics.
A total of 36 studies were included, containing 2260 patients of which 425 cases, selected from a populations of more than 7.000 BE patients. The proportion of male patients ranged from 66% to 100%. Mean duration of follow-up varied from 11.3 months to 120 months. Most studies were retrospective case-control studies (n = 33), and three prospective cohort studies. One study defined BE as CLE without IM, other studies defined BE as CLE with IM (n = 23), or gave no definition (n = 12). Endpoint was EAC in six studies and either HGD or EAC in 30 studies. (Table 1).
Dilutions and definitions for aberrant staining used.
For p53, the antibody DO-7 (Dako, Glostrup, Denmark) was frequently used with a dilution ranging from 1:20 to 1:1000. The definition for aberrant IHC staining was heterogeneous. Very intense staining was considered aberrant by all studies (being independent of the concentration used). However, intensity of staining was not a prerequisite for considering a staining pattern aberrant in seven studies[29–33, 37, 47]. Three more recent studies also considered a total absence of staining as aberrant [35, 36, 39]. For aspergillus oryzae lectin (AOL), one study calculated the OR for aberrant AOL IHC staining in 2 or 3 epithelial compartments versus 0 or 1 compartment [34]. Another study reported multiple ORs for aberrant AOL in 1, 2, or 3 versus 0 epithelial compartments [39]. The OR of aberrant AOL IHC staining in 2 or 3 versus 0 or 1 compartments was extracted from this second study and analyzed together with the data from the first study for the meta-analysis.
Quality of studies
In 14 of the 36 studies there was at least a 10% difference in baseline histology between cases and controls. In these studies, around 32% of the cases had IND or LGD at baseline, versus 9% in the controls. In five studies an age difference at baseline of at least 5 years was found between cases and controls, in four of these studies the case group was older. In six studies at least 10% more males were included in the case groups, 96% males on average in the cases, versus 73% in controls. Information on length of the BE segment for both cases and controls was provided in 13 studies. In cases a longer BE segment was present; on average 5.9 cm versus 4.8 cm in the controls. In 17 studies the total follow-up time differed by at least 10% between case and control groups. On average, follow up time was 51 months versus 59 months for cases versus controls, respectively. Seven studies excluded possible prevalent cases [29, 32, 34, 47, 49]. Fourteen studies adjusted for known predictors of progression (S1 Table). 16 studies did not describe technical validation of the staining. IHC staining was scored by one observer in 15 studies, by two observers in 13 studies, and by three observers in three studies. Kappa values were mentioned in only eight studies. Whether slides were assessed in a blinded manner was not mentioned in six studies, all other studies reported the use of blinding (S1 Table).
Meta-analyses
These were possible for p53, AOL, Cyclin A, Cyclin D, and alpha-methylacyl-CoA racemase (AMACR), which were studied 13, 2, 4, 3, and 2 times respectively. Of the 13 studies, two included patients from the same population, which resulted in exclusion of the older study in analyses for which both would have been eligible [32, 34]. The most frequent reasons for excluding articles were the absence of follow-up data and LGD being defined as neoplastic progression and end-point of the study (Fig 1). Biomarkers studied only once were MCM2, CD1a, β-catenin, COX2, Ki67, HER2, Sialyl Lewis, Wheat germ, Lewis, and SOX2. The same group published two studies on hERG1, both including patients from the same population[46, 48]. Therefore, both studies were individually included without summary in a meta-analysis.
p53
A total of 12 studies were included in the meta-analysis. These contained 1905 patients, of which 342 cases. One study gave multiple ORs for various expression levels of p53 [34]. For this study, only the OR for intense overexpression of p53 staining was considered positive. Individual patient data of one study were converted in order to extract an adjusted OR [35]. The overall OR for neoplastic progression was 7.04 (95% CI 3.68 to 13.46) for patients with aberrant p53 expression (Table 2 and Fig 2). Aberrant p53 expression, detected in both non-dysplastic BE and LGD patients, was significantly associated with the development of HGD or EAC. Significant heterogeneity (I2 = 56%, P<0.010) was observed between the included studies, which can be considered a moderate amount of heterogeneity [50]. The 12 studies were plotted in a funnel plot which shows that small study effects can be present (S1 Fig). In order to reduce the influence of such effects a sensitivity analysis was performed, which excluded all studies with a standard error above one. Based on this criterion, five studies remained, containing 1413 patients and 289 cases. The overall OR for neoplastic progression was 4.15 (95% CI 1.96 to 8.81) in patients with an aberrant p53 expression. (Table 2). The use of a more stringent definition of aberrant staining may lead to loss of aberrant expression in cases, in controls, or in both. In order to investigate this, the proportion of controls deemed aberrant was plotted against the OR of each study (S2 Fig). Studies with a higher point estimate of the OR appeared to have had less positive non-progressors. The same is seen if this is analyzed in individual studies where multiple cut-offs for positivity are described[32, 34]. Using a more stringent cut off resulted in a higher OR. Further sub sensitivity analyses were performed excluding studies for which no adjusted ORs were available. Four studies remained, containing 1322 patients and 278 cases. The overall OR for neoplastic progression was 3.18 (95% CI 1.68 to 6.03) in patients with an aberrant p53 expression. (Table 2). Subsequently, individual studies were excluded from this analysis, and finally also studies presented as abstracts. These sensitivity analyses showed similar results with slightly lower point estimates compared to the main analysis. (Table 2). For three studies both unadjusted and adjusted ORs were available, and all three adjusted ORs had a lower point estimate compared to the unadjusted ones, in line with the outcome of our meta-analyses.
Twelve studies were included.
p53 as independent predictor of neoplastic progression.
For this analysis studies that did not adjust for histology at baseline were excluded. This led to the inclusion of six studies. These studies contained a total of 1340 BE patients, of which 282 cases. The overall OR, for aberrant p53 IHC on neoplastic progression, after stratification for histology, was 3.86 (95% CI 2.03 to 7.33). (Table 2).
p53 in non-dysplastic Barrett's oesophagus.
Two studies were included for this analysis. These contained a total of 720 BE patients, of which 61 cases. Individual patient data of one study was re-analyzed to provide an OR for non-dysplastic BE patients only [35]. The overall OR for neoplastic progression to HGD or EAC in non-dysplastic BE patients was 6.12 (95% CI 2.99 to 12.52). (Table 2).
p53 in low-grade dysplasia Barrett.
For this analysis four studies were included. These contained a total of 182 BE patients, of which 37 cases. One study was re-analyzed to provide an OR for the LGD subgroup only [35]. The overall OR for neoplastic progression to HGD or EAC was 8.64 (95% CI 3.62 to 20.62). (Table 2).
AOL
Two studies were included in this meta-analysis. These contained 573 BE patients, of which 204 cases. The overall OR for neoplastic progression in BE patients with an aberrant AOL staining in 2 or 3 compartments, versus 0 or 1 compartments of the tissue was 3.04 (95% CI 2.05 to 4.49) (Table 3 and S3 Fig). Results of the two studies were consistent in their findings (I2 = 0%, P = 0.85).
CYCLIN A
Four studies were included in this meta-analysis. These contained 1275 patients, of which 285 cases. The overall OR for neoplastic progression in BE patients with cyclin A positivity was 1.90 (95% CI 0.85 to 4.22) (Table 3 and S3 Fig). Results of the three studies were inconsistent in their findings (I2 = 76%, P = 0.005).
CYCLIN D
Three studies were included in this meta-analysis. These contained 337 patients, of which 50 cases. The overall OR for neoplastic progression in BE patients with cyclin D positivity was 1.01 (95% CI 0.14 to 7.03) (Table 3 and S3 Fig). Results of the two studies were inconsistent in their findings (I2 = 80%, P = 0.007).
Alpha-methylacyl-CoA racemase
Two studies were included in this meta-analysis. These contained 712 patients, of which 53 cases. The overall OR for neoplastic progression in BE patients with alpha-methylacyl-CoA racemase positivity was 4.07 (95% CI 0.66 to 25.12) (Table 3 and S3 Fig). Results of the two studies were moderately consistent in their findings (I2 = 53%, P = 0.14).
Studies on other IHC biomarkers
The following IHC biomarkers were investigated only once: β-catenin, CD1a, COX2, HER2, Ki67, Lewis, Mcm2, Sialyl Lewis, SOX2, and WGA. The same group published two studies on hERG1, both including patients from the same population[46, 48]. Therefore, both studies were individually included without summary in a meta-analysis. In the CD1a study CLE without IM was used as baseline histology [45]. When considering study size and point estimate, CD1a, SOX2, and hERG1 appeared most promising. (S4 Fig).
Discussion
This is the first systematic review and meta-analysis to assess if IHC biomarkers can be used as an independent predictor for neoplastic progression in BE surveillance. Sixteen biomarkers have been investigated in this setting, of which five biomarkers have been investigated more than once. The meta-analysis showed that aberrant p53 expression was associated with a significantly increased risk of neoplastic progression. Moreover, aberrant p53 expression predicted neoplastic progression in both non-dysplastic BE patients and BE patients with LGD. Of the other four IHC biomarkers, AOL appeared to be most promising in predicting neoplastic progression, whereas Cyclin A, Cyclin D, and alpha-methylacyl-CoA racemase are still of limited value.
Current use of p53 IHC in BE patients differs in international guidelines. The guideline of the British Society of Gastroenterology recommends the addition of p53 IHC staining for the pathological assessment of BE to improve the diagnostic reproducibility of dysplasia [5]. While the American Gastroenterological Association guideline states that: “data supporting the use of biomarkers to confirm the histologic diagnosis of dysplasia must be considered preliminary [51]. No guideline has yet adopted the use of IHC biomarkers to predict neoplastic progression. Two large population based studies confirmed that patients with LGD have an approximately 5 times higher risk of neoplastic progression compared to patients without LGD [15, 18]. Our meta-analysis is the first to show that BE patients, independent of the presence of LGD, with aberrant p53 IHC have a similar increased risk to develop cancer compared to patients with LGD. A recent publication claims to have investigated the predictive ability of immunohistochemical biomarkers [52]. However, they reported on samples either obtained from a resection specimen or from cases and controls without follow-up. Therefore, based on their current dataset, their current conclusion, i.e. that p53 overexpression predicts malignant progression, is not justified [53].
Although routine p53 IHC will incur higher cost than histological assessment alone, application of this marker has the potential to reduce the overall costs related to BE surveillance by improved risk stratification using expression of p53 IHC in combination with other predictors of progression, such as histology, sex, age, and length of the BE segment. Better risk stratification could result in both earlier detection of lesions in patients at risk, and a reduction in endoscopic and pathology recources for patients that will never develop progression. The disparity in ORs of neoplastic progression found in the various studies may be explained by differences in staining methods, including antibodies used, antigen retrieval methods, definitions, and interpretations of aberrant staining used. Therefore, special consideration should be given to the protocol of staining and the definition and interpretation used for aberrant expression. Some studies did not consider loss of p53 staining aberrant, which might have contributed to the protocol being less predictive compared to other studies. By using a more stringent definition of aberrant expression, cases appeared to remain p53 aberrant, while controls were not considered aberrant (S2 Fig). Therefore, the use of more stringent definitions and interpretations for aberrant staining appears to lead to a higher predictive ability of p53 IHC.
The strength of this paper is the focus on IHC biomarkers as a relatively easy applicable tool to improve risk stratification in BE surveillance. Additionally, we performed a broad search, and the extraction of ORs from text, tables and figures resulted in the inclusion of quite a large number of studies. The inclusion of abstracts results in an up to date overview of this field. Because meta-analysis is the synthesis technique that is most transparent and most likely valid also with small amount of studies included, some of the meta-analysis were performed with only two studies, as no more studies were available. [27] This study also has its limitations, such as the confinement to English language publications, the apparent presence of publication bias, differences in baseline comparability within studies, and the various adjustments made for these baseline differences. Therefore, we performed sensitivity analyses of the p53 meta-analyses, these show that the point estimate of the OR decreased from 7.04 to 3.18 when we accounted for these limitations. Because aberrant p53 IHC co-occurs with LGD, separate analyses were performed in which we stratified for dysplastic and non-dysplastic patients. These analyses show that aberrant p53 expression is an independent prognostic factor for neoplastic progression.
In conclusion, we show that sixteen IHC biomarkers in BE surveillance have been studied. Aberrant p53 expression is the most studied IHC biomarker and associated with a significantly increased risk to develop HGD or EAC, this association was independent of the presence of LGD. Consensus amongst pathologists concerning the appropriate staining method, definition, and interpretation of aberrant p53 expression is currently low, and more consensus is required. Other promising biomarkers such as AOL need further investigation.
Supporting information
S1 Fig. Funnel plot of all studies investigating p53 IHC as a predictor of progression.
The exact patient numbers and the SE of these studies can be found in Fig 2.
https://doi.org/10.1371/journal.pone.0186305.s001
(EPS)
S2 Fig. Stringency of the definition and interpretation of aberrant p53 IHC.
A more stringent definition of aberrant staining, and interpretation of that definition, may lead to loss of aberrant expression in cases, in controls, or in both. In order to investigate this, the proportion of controls deemed aberrant was plotted against the OR of each study with a standard error below 1. The use of a more stringent definition and interpretation for aberrant p53 staining appeared to result in a bigger reduction in the number of controls considered to have aberrant staining, compared to cases. Thus, by applying a more stringent definition and interpretation, the predictive value of p53 for neoplastic progression appears to increase. Formal statistical tests were not performed due to the limited number of data points and the post hoc nature of this analysis.
https://doi.org/10.1371/journal.pone.0186305.s002
(EPS)
S3 Fig. Forest plot of all studies investigating AOL, Cyclin A, Cyclin D, and AMACR as a predictor of progression.
(A) Two studies were included in the forest plot for AOL. (B) Four studies were included in the forest plot for Cyclin A. (C) Three studies were included in the forest plot for Cyclin D. (D) Two studies were included in the forest plot for AMACR.
https://doi.org/10.1371/journal.pone.0186305.s003
(TIF)
S4 Fig. Forest plot, without meta-analysis, of all studies investigating IHC biomarkers which have been studied only once.
12 studies were included in this forest plot.
https://doi.org/10.1371/journal.pone.0186305.s004
(EPS)
S1 Table. Additional characteristics of included studies.
https://doi.org/10.1371/journal.pone.0186305.s005
(DOCX)
Acknowledgments
B.E. Hansen, Department of Gastroenterology and Hepatology, Erasmus University Medical Center Rotterdam, The Netherlands. The biomedical information specialists of the medical library of the Erasmus MC, University Medical Center Rotterdam, The Netherlands for their support with building and running the search.
References
- 1. American Gastroenterological A, Spechler SJ, Sharma P, Souza RF, Inadomi JM, Shaheen NJ. American Gastroenterological Association medical position statement on the management of Barrett's esophagus. Gastroenterology. 2011;140(3):1084–91. pmid:21376940.
- 2. Kastelein F, Spaander MC, Biermann K, Vucelic B, Kuipers EJ, Bruno MJ. Role of acid suppression in the development and progression of dysplasia in patients with Barrett's esophagus. Digestive diseases. 2011;29(5):499–506. pmid:22095018.
- 3. Winters C Jr., Spurling TJ, Chobanian SJ, Curtis DJ, Esposito RL, Hacker JF 3rd, et al. Barrett's esophagus. A prevalent, occult complication of gastroesophageal reflux disease. Gastroenterology. 1987;92(1):118–24. pmid:3781178.
- 4. Buttar NS, Wang KK. Mechanisms of disease: Carcinogenesis in Barrett's esophagus. Nature clinical practice Gastroenterology & hepatology. 2004;1(2):106–12. pmid:16265072.
- 5. Fitzgerald RC, di Pietro M, Ragunath K, Ang Y, Kang JY, Watson P, et al. British Society of Gastroenterology guidelines on the diagnosis and management of Barrett's oesophagus. Gut. 2014;63(1):7–42. pmid:24165758.
- 6. Wang KK, Sampliner RE, Practice Parameters Committee of the American College of G. Updated guidelines 2008 for the diagnosis, surveillance and therapy of Barrett's esophagus. The American journal of gastroenterology. 2008;103(3):788–97.
- 7. Cooper GS, Kou TD, Chak A. Receipt of previous diagnoses and endoscopy and outcome from esophageal adenocarcinoma: a population-based study with temporal trends. The American journal of gastroenterology. 2009;104(6):1356–62. pmid:19491849.
- 8. Fountoulakis A, Zafirellis KD, Dolan K, Dexter SP, Martin IG, Sue-Ling HM. Effect of surveillance of Barrett's oesophagus on the clinical outcome of oesophageal cancer. The British journal of surgery. 2004;91(8):997–1003. pmid:15286961.
- 9. Kastelein F, van Olphen SH, Steyerberg EW, Spaander MC, Bruno MJ, ProBar-study g. Impact of surveillance for Barrett's oesophagus on tumour stage and survival of patients with neoplastic progression. Gut. 2015. pmid:25903690.
- 10. Rubenstein JH, Sonnenberg A, Davis J, McMahon L, Inadomi JM. Effect of a prior endoscopy on outcomes of esophageal adenocarcinoma among United States veterans. Gastrointestinal endoscopy. 2008;68(5):849–55. pmid:18547567
- 11. Reid BJ, Levine DS, Longton G, Blount PL, Rabinovitch PS. Predictors of progression to cancer in Barrett's esophagus: baseline histology and flow cytometry identify low- and high-risk patient subsets. The American journal of gastroenterology. 2000;95(7):1669–76. pmid:10925966
- 12. Sikkema M, de Jonge PJ, Steyerberg EW, Kuipers EJ. Risk of esophageal adenocarcinoma and mortality in patients with Barrett's esophagus: a systematic review and meta-analysis. Clinical gastroenterology and hepatology: the official clinical practice journal of the American Gastroenterological Association. 2010;8(3):235–44; quiz e32. pmid:19850156.
- 13. Yousef F, Cardwell C, Cantwell MM, Galway K, Johnston BT, Murray L. The incidence of esophageal cancer and high-grade dysplasia in Barrett's esophagus: a systematic review and meta-analysis. American journal of epidemiology. 2008;168(3):237–49. pmid:18550563.
- 14. Hage M, Siersema PD, van Dekken H, Steyerberg EW, Dees J, Kuipers EJ. Oesophageal cancer incidence and mortality in patients with long-segment Barrett's oesophagus after a mean follow-up of 12.7 years. Scandinavian journal of gastroenterology. 2004;39(12):1175–9. pmid:15742992.
- 15. Bhat S, Coleman HG, Yousef F, Johnston BT, McManus DT, Gavin AT, et al. Risk of malignant progression in Barrett's esophagus patients: results from a large population-based study. Journal of the National Cancer Institute. 2011;103(13):1049–57. pmid:21680910
- 16. Desai TK, Krishnan K, Samala N, Singh J, Cluley J, Perla S, et al. The incidence of oesophageal adenocarcinoma in non-dysplastic Barrett's oesophagus: a meta-analysis. Gut. 2012;61(7):970–6. pmid:21997553.
- 17. Desai TK, Singh J, Samala N, Subbiah P. The incidence of esophageal adenocarcinoma in Barrett's esophagus has been overestimated. The American journal of gastroenterology. 2011;106(7):1364–5; author reply 5–6. pmid:21731021.
- 18. Hvid-Jensen F, Pedersen L, Drewes AM, Sorensen HT, Funch-Jensen P. Incidence of adenocarcinoma among patients with Barrett's esophagus. The New England journal of medicine. 2011;365(15):1375–83. pmid:21995385.
- 19. Wani S, Falk G, Hall M, Gaddam S, Wang A, Gupta N, et al. Patients with nondysplastic Barrett's esophagus have low risks for developing dysplasia or esophageal adenocarcinoma. Clin Gastroenterol Hepatol. 2011;9(3):220–7; quiz e6. pmid:21115133.
- 20. Coleman HG, Bhat S, Johnston BT, McManus D, Gavin AT, Murray LJ. Tobacco smoking increases the risk of high-grade dysplasia and cancer among patients with Barrett's esophagus. Gastroenterology. 2012;142(2):233–40. pmid:22062359.
- 21. Curvers WL, ten Kate FJ, Krishnadath KK, Visser M, Elzer B, Baak LC, et al. Low-grade dysplasia in Barrett's esophagus: overdiagnosed and underestimated. The American journal of gastroenterology. 2010;105(7):1523–30. pmid:20461069.
- 22. Kerkhof M, van Dekken H, Steyerberg EW, Meijer GA, Mulder AH, de Bruine A, et al. Grading of dysplasia in Barrett's oesophagus: substantial interobserver variation between general and gastrointestinal pathologists. Histopathology. 2007;50(7):920–7. pmid:17543082.
- 23. Cameron AJ, Carpenter HA. Barrett's esophagus, high-grade dysplasia, and early adenocarcinoma: a pathological study. The American journal of gastroenterology. 1997;92(4):586–91. pmid:9128304.
- 24. 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):40–50. pmid:8514061.
- 25. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. J Clin Epidemiol. 2009;62(10):1006–12. pmid:19631508.
- 26. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. Jama. 2000;283(15):2008–12. pmid:10789670.
- 27. Valentine JC, Pigott TD, Rothstein HR. How Many Studies Do You Need? A Primer on Statistical Power for Meta-Analysis. Journal of Educational and Behavioral Statistics. 2010;Vol. 35(No. 2):pp. 215–47. Epub 247.
- 28. Younes M, Ertan A, Lechago LV, Somoano JR, Lechago J. p53 Protein accumulation is a specific marker of malignant potential in Barrett's metaplasia. Digestive diseases and sciences. 1997;42(4):697–701. pmid:9125634.
- 29. Bani-Hani K, Martin IG, Hardie LJ, Mapstone N, Briggs JA, Forman D, et al. Prospective study of cyclin D1 overexpression in Barrett's esophagus: association with increased risk of adenocarcinoma. Journal of the National Cancer Institute. 2000;92(16):1316–21. pmid:10944553.
- 30. Weston AP, Banerjee SK, Sharma P, Tran TM, Richards R, Cherian R. p53 protein overexpression in low grade dysplasia (LGD) in Barrett's esophagus: immunohistochemical marker predictive of progression. The American journal of gastroenterology. 2001;96(5):1355–62. pmid:11374668.
- 31. Skacel M, Petras RE, Rybicki LA, Gramlich TL, Richter JE, Falk GW, et al. p53 expression in low grade dysplasia in Barrett's esophagus: correlation with interobserver agreement and disease progression. The American journal of gastroenterology. 2002;97(10):2508–13. pmid:12385431.
- 32. Murray L, Sedo A, Scott M, McManus D, Sloan JM, Hardie LJ, et al. TP53 and progression from Barrett's metaplasia to oesophageal adenocarcinoma in a UK population cohort. Gut. 2006;55(10):1390–7. pmid:16682429
- 33. Sikkema M, Kerkhof M, Steyerberg EW, Kusters JG, van Strien PM, Looman CW, et al. Aneuploidy and overexpression of Ki67 and p53 as markers for neoplastic progression in Barrett's esophagus: a case-control study. The American journal of gastroenterology. 2009;104(11):2673–80. pmid:19638963.
- 34. Bird-Lieberman EL, Dunn JM, Coleman HG, Lao-Sirieix P, Oukrif D, Moore CE, et al. Population-based study reveals new risk-stratification biomarker panel for Barrett's esophagus. Gastroenterology. 2012;143(4):927–35 e3. pmid:22771507.
- 35. Kastelein F, Biermann K, Steyerberg EW, Verheij J, Kalisvaart M, Looijenga LH, et al. Aberrant p53 protein expression is associated with an increased risk of neoplastic progression in patients with Barrett's oesophagus. Gut. 2013;62(12):1676–83. pmid:23256952.
- 36.
Davelaar AL, Calpe S Fau—Lau L, Lau L Fau—Timmer MR, Timmer Mr Fau—Visser M, Visser M Fau—Ten Kate FJ, Ten Kate Fj Fau—Parikh KB, et al. Aberrant TP53 detected by combining immunohistochemistry and DNA-FISH improves Barrett's esophagus progression prediction: a prospective follow-up study. (1098–2264 (Electronic)).
- 37. Gimenez A, de Haro LM, Parrilla P, Bermejo J, Perez-Guillermo M, Ortiz MA. Immunohistochemical detection of p53 protein could improve the management of some patients with Barrett esophagus and mild histologic alterations. Arch Pathol Lab Med. 1999;123(12):1260–3. pmid:10583932.
- 38.
Brown K, Younes M, Ertan A, Verm R, Meriano FV. M1928 P53 Immunostaining Predicts Malignant Progression in Barrett's Metaplasia: WB Saunders; 2008.
- 39. Wolf WA, Lao-Sirieix P, Duits LC, Chak A, Shaheen NJ, Fitzgerald R, et al. Utility of a biomarker panel to predict progression to esophageal adenocarcinoma in barrett's esophagus. Gastroenterology. 2014;146(5):S–330.
- 40. Lao-Sirieix P, Lovat L, Fitzgerald RC. Cyclin A immunocytology as a risk stratification tool for Barrett's esophagus surveillance. Clinical cancer research: an official journal of the American Association for Cancer Research. 2007;13(2 Pt 1):659–65. pmid:17255290.
- 41. Sirieix PS, O'Donovan M, Brown J, Save V, Coleman N, Fitzgerald RC. Surface expression of minichromosome maintenance proteins provides a novel method for detecting patients at risk for developing adenocarcinoma in Barrett's esophagus. Clinical cancer research: an official journal of the American Association for Cancer Research. 2003;9(7):2560–6. pmid:12855631.
- 42. Kastelein F, Biermann K, Steyerberg EW, Verheij J, Kalisvaart M, Looijenga LH, et al. Value of alpha-methylacyl-CoA racemase immunochemistry for predicting neoplastic progression in Barrett's oesophagus. Histopathology. 2013;63(5):630–9. pmid:24004067.
- 43. van Olphen S, Biermann K, Spaander MC, Kastelein F, Steyerberg EW, Stoop HA, et al. SOX2 as a Novel Marker to Predict Neoplastic Progression in Barrett's Esophagus. The American journal of gastroenterology. 2015. pmid:26323187.
- 44.
Rossi E, Grisanti S Fau—Villanacci V, Villanacci V Fau—Della Casa D, Della Casa D Fau—Cengia P, Cengia P Fau—Missale G, Missale G Fau—Minelli L, et al. HER-2 overexpression/amplification in Barrett's oesophagus predicts early transition from dysplasia to adenocarcinoma: a clinico-pathologic study. (1582–4934 (Electronic)).
- 45.
Cappello F, Rappa F Fau—Anzalone R, Anzalone R Fau—La Rocca G, La Rocca G Fau—Zummo G, Zummo G. CD1a expression by Barrett's metaplasia of gastric type may help to predict its evolution towards cancer. (0007–0920 (Print)).
- 46. Lastraioli E, Taddei A, Messerini L, Comin CE, Festini M, Giannelli M, et al. hERG1 channels in human esophagus: evidence for their aberrant expression in the malignant progression of Barrett's esophagus. J Cell Physiol. 2006;209(2):398–404. pmid:16883575.
- 47. Horvath B, Singh P, Xie H, Thota PN, Sun X, Liu X. Expression of p53 predicts risk of prevalent and incident advanced neoplasia in patients with Barrett's esophagus and epithelial changes indefinite for dysplasia. Gastroenterol Rep (Oxf). 2016;4(4):304–9. pmid:26486567.
- 48. Lastraioli E, Lottini T, Iorio J, Freschi G, Fazi M, Duranti C, et al. hERG1 behaves as biomarker of progression to adenocarcinoma in Barrett's esophagus and can be exploited for a novel endoscopic surveillance. Oncotarget. 2016;7(37):59535–47.
- 49. van Olphen SH, Ten Kate FJ, Doukas M, Kastelein F, Steyerberg EW, Stoop HA, et al. Value of cyclin A immunohistochemistry for cancer risk stratification in Barrett esophagus surveillance: A multicenter case-control study. Medicine (Baltimore). 2016;95(47):e5402. pmid:27893678.
- 50. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Bmj. 2003;327(7414):557–60. pmid:12958120.
- 51. Spechler SJ, Sharma P, Souza RF, Inadomi JM, Shaheen NJ. American gastroenterological association technical review on the management of Barrett's esophagus. Gastroenterology. 2011;140(3):e18–e52. pmid:21376939
- 52. Altaf K, Xiong JJ, la Iglesia D, Hickey L, Kaul A. Meta-analysis of biomarkers predicting risk of malignant progression in Barrett's oesophagus. The British journal of surgery. 2017;104(5):493–502. pmid:28295252.
- 53. Janmaat VT, Peppelenbosch MP, Bruno MJ, Spaander MCW. Comment on: 'Meta-analysis of biomarkers predicting risk of malignant progression in Barrett’s oesophagus’ (Br J Surg 2017; 104: 493–502). British Journal of Surgery; 2017 [updated 14-06-2017; cited 14-06-2017 14-06-2017]. Available from: https://www.bjs.co.uk/article/meta-analysis-of-biomarkers-predicting-risk-of-malignant-progression-in-barretts-oesophagus/.