Screening and Risk Stratification for Barrett’s Esophagus




Barrett’s esophagus (BE) and gastroesophageal reflux disease are the strongest risk factors for esophageal adenocarcinoma. To reduce the clinical impact of this disease, endoscopic screening to detect BE has been proposed and nonendoscopic diagnostic techniques are under investigation. Because screening would result in new diagnoses of BE and additional costs related to endoscopic surveillance, novel tools for risk stratification are also warranted. Dysplasia is the gold standard for risk stratification. Molecular biomarkers may provide a more objective and reproducible estimation of the individual risk, and further prospective studies are required as a prelude to introducing biomarkers into routine clinical practice.


Key points








  • Esophageal adenocarcinoma (EAC) is the solid malignancy with the fastest increasing incidence in the Western world over the last 3 decades.



  • Barrett’s esophagus (BE) is the only known precursor to EAC and the strongest risk factor for this type of esophageal cancer.



  • Other risk factors for EAC include gastroesophageal reflux disease, obesity, smoking, and male sex.



  • Endoscopic screening for BE has the potential to reduce the clinical impact of the changing epidemiology of EAC, but it is not cost-effective.



  • Less invasive and more inexpensive modalities (eg, Cytosponge, office-based transnasal endoscopy, and capsule endoscopy) are under investigation, with the aim of making screening feasible in primary care.



  • Histologic dysplasia in BE is the only biomarker available in clinical practice; however, it is subject to high interobserver variability among pathologists and sampling error from random biopsies.



  • Molecular biomarkers can provide a more objective estimate of the individual cancer risk and are under investigation.






Epidemiology of esophageal adenocarcinoma


The epidemiology of esophageal adenocarcinoma (EAC) has changed dramatically over the past 50 years in the Western world. Surgical series reported before the 1980s showed that esophageal squamous cell carcinoma (ESCC), which arises from the native multilayered squamous epithelium, was the most common malignancy in the esophagus. However, subsequent Western case series published 10 to 15 years later indicated that the EAC had become the most common esophageal malignancy, exceeding the number of cases of ESCC. A study looking at the relative distribution of these 2 esophageal tumor types over 25 years spanning this transition period showed that this change was statistically significant. Despite a slight decline in the incidence of ESCC, the trend inversion seems to be mostly caused by a dramatic increase in the incidence of EAC during the last 3 to 4 decades, such that in the early years of the twenty-first century EAC has been the fastest rising solid malignancy in the United States. This epidemiologic trend for EAC involves mostly Western countries, although there is some geographic variation. Between the early 1980s and late 1990s, among all European countries, Ireland and the United Kingdom showed the largest increase in the age-standardized incidence rates (∼7% per annum), a trend comparable with the United States and Australia. Data have established that this is a true increase in incidence, rather than the effect of histologic reclassification or overdiagnosis associated with technological advancements. Another potential confounding factor is the anatomic classification of adenocarcinomas around the gastroesophageal junction. EAC and gastric cardia adenocarcinoma (GCA) are often grouped together in clinical studies. However, studies looking separately at cancer incidence at these 2 locations between the early 1980s and late 2000s showed that, although incidence of EAC has risen by 6-fold, the incidence of GCA increased initially by 2-fold and declined latterly, remaining overall stable across this temporal period. Data from more recent periods show that the overall incidence of EAC continues to increase, although at a slower rate, which is approximately 2% per year.




Epidemiology of esophageal adenocarcinoma


The epidemiology of esophageal adenocarcinoma (EAC) has changed dramatically over the past 50 years in the Western world. Surgical series reported before the 1980s showed that esophageal squamous cell carcinoma (ESCC), which arises from the native multilayered squamous epithelium, was the most common malignancy in the esophagus. However, subsequent Western case series published 10 to 15 years later indicated that the EAC had become the most common esophageal malignancy, exceeding the number of cases of ESCC. A study looking at the relative distribution of these 2 esophageal tumor types over 25 years spanning this transition period showed that this change was statistically significant. Despite a slight decline in the incidence of ESCC, the trend inversion seems to be mostly caused by a dramatic increase in the incidence of EAC during the last 3 to 4 decades, such that in the early years of the twenty-first century EAC has been the fastest rising solid malignancy in the United States. This epidemiologic trend for EAC involves mostly Western countries, although there is some geographic variation. Between the early 1980s and late 1990s, among all European countries, Ireland and the United Kingdom showed the largest increase in the age-standardized incidence rates (∼7% per annum), a trend comparable with the United States and Australia. Data have established that this is a true increase in incidence, rather than the effect of histologic reclassification or overdiagnosis associated with technological advancements. Another potential confounding factor is the anatomic classification of adenocarcinomas around the gastroesophageal junction. EAC and gastric cardia adenocarcinoma (GCA) are often grouped together in clinical studies. However, studies looking separately at cancer incidence at these 2 locations between the early 1980s and late 2000s showed that, although incidence of EAC has risen by 6-fold, the incidence of GCA increased initially by 2-fold and declined latterly, remaining overall stable across this temporal period. Data from more recent periods show that the overall incidence of EAC continues to increase, although at a slower rate, which is approximately 2% per year.




Risk factors for EAC


Table 1 lists the clinical and epidemiologic factors that have been studied in relation to the incidence of EAC.



Table 1

Factors associated with EAC








































Evidence References
GERD Weekly symptoms: OR for EAC 4.9
Daily symptoms: OR for EAC 7.4
BE Annual risk 0.12%–0.38% for nondysplastic BE
Annual risk 0.51%–14.0% for BE with LGD
Obesity BMI 25 to <30 kg/m 2 : OR for EAC 1.7
BMI >30 kg/m 2 : OR for EAC 2.3
Gender EAC incidence 6 times higher in men than women
EAC risk in male patients with BE 2–3 times higher than female
Smoking Current smokers: OR for EAC 2.3
Ex-smokers: OR for EAC 1.6
Alcohol Evidence of lack of association
Helicobacter pylori Inverse correlation to EAC (particularly for cytotoxin-associated antigen A-positive strains: OR for EAC 0.4)
Diet Weak evidence of positive association with red meat and processed food and inverse association with consumption of fruit and vegetables

Abbreviations: BE, Barrett’s esophagus; BMI, body mass index; GERD, gastroesophageal reflux disease; LGD, low-grade dysplasia; OR, odds ratio.


Gastroesophageal Reflux Disease


Gastroesophageal reflux disease (GERD) and its associated pathologic condition hiatus hernia are well-documented risk factors for the development of EAC. A recent meta-analysis that included 5 studies showed that weekly and daily GERD symptoms are associated with an odds ratio (OR) for EAC of 4.92 (95% confidence interval [CI], 3.90–6.22) and 7.4 (95% CI, 4.94–11.1), respectively. Although the pathophysiologic mechanisms of this association are not fully understood, it is believed that oxidative and genotoxic damage provoked by exposure of esophageal epithelium to acid and bile induces genetic and epigenetic changes that support the carcinogenic process.


Barrett’s Esophagus


GERD is believed to induce esophageal carcinogenesis through the premalignant condition Barrett’s esophagus (BE), which is a columnar metaplasia of the distal esophagus generally containing intestinal differentiation. Endoscopic studies have shown that BE can be found in approximately 10% of individuals with reflux disease sufficient to warrant referral for endoscopy. Most cases of EACs are believed to develop in this context, so that BE is the single strongest risk factor for the development of EAC with a relative risk (RR) of 11.3 (95% CI, 8.8–14.4). The metaplastic conversion to columnar lined epithelium explains how adenocarcinoma can develop in an organ normally lined by a squamous epithelial type. Historically, the annual incidence of EAC in patients with BE was believed to be between 0.5% and 1% per year ; however, more recent population studies and a meta-analysis have set this risk at around 0.12% to 0.38% per year.


Obesity


Obesity is a risk factor for many types of cancer. Most studies examining the correlation between obesity and EAC have found a positive association. Two meta-analyses published in 2006 and 2012 reached similar conclusions, namely that high body mass index (BMI, calculated as weight in kilograms divided by the square of height in meters) is associated with both EAC and GCA, with the strongest effect for EAC. The more recent of these meta-analyses found ORs for EAC of 1.7 (95% CI, 1.5–1.96) and 2.34 (95% CI, 1.95–2.81), for a BMI between 25 and 30 kg/m 2 and greater than 30 kg/m 2 , respectively. The mechanisms whereby obesity increases EAC risk are likely to be multiple. Obesity leads to increased intra-abdominal pressure, is associated with more frequent transient relaxations of the lower esophageal sphincter, and also correlates with a higher incidence of hiatus hernia. This finding may explain the positive association between obesity and GERD found in many studies and confirmed by a meta-analysis, in which the ORs for GERD symptoms were 1.43 (95% CI, 1.16–1.776) and 1.94 (95% CI, 1.47–2.57), for BMIs between 25 and 30 kg/m 2 and greater than 30 kg/m 2 , respectively. However, this correlation also seems to go beyond the simple predisposition to acid reflux. It is now clear that the distribution of fat, and visceral adiposity in particular, may play a role in the promotion of BE. Furthermore, obesity acts not only on the acquisition of the metaplastic precursor but also on the development of genetic abnormalities that drive the transformation to cancer. This situation is likely to be related to the metabolic activity of visceral fat and its ability to produce cytokines, growth factors, and hormones, which can alter insulin resistance and lipid metabolism, promoting the condition known as metabolic syndrome. This syndrome has been independently associated with a predisposition to cancer and BE. Overall, it is believed that the increasing prevalence of obesity in Western countries is likely to be a significant contributor to the epidemic increase in the incidence of EAC.


Smoking


Another risk factor for EAC is tobacco smoking, although to a lesser degree compared with ESCC. A recent meta-analysis found a pooled RR for ever-smokers of 1.76 (95% CI, 1.54–2.01), with a direct correlation between the risk of cancer and the dose and the duration of the exposure. The recent decline in smoking in many of the Western countries may account for the slower rate of increase in the incidence of EAC since the late 1990s.


Gender


Male sex is associated with an increased risk of EAC, and 2 recent population studies confirmed this association. Bhat and colleagues showed that the annual cancer incidence in patients with BE with intestinal metaplasia was 0.45% (95% CI, 0.36–0.56) in men and 0.26% (95% CI, 0.18–0.38) in women. Similarly, Hvid-Jensen and colleagues found a significant difference in the cancer incidence between male and female patients. The meta-analysis from Sikkema and colleagues before these studies found an RR associated to male gender of 1.7 (95% CI, 0.6–4.5), whereas a more recent meta-analysis did not perform subgroup analysis related to gender. Independently of a previous diagnosis of BE, EAC incidence is 6-fold to 8-fold higher in men compared with women. The reasons for this gender bias are not entirely clear, although hormonal factors have been proposed.


Alcohol Consumption


Although alcohol is a well-known carcinogen and is a major risk factor for the development of ESCC, alcohol consumption is not associated with an increased risk of EAC and there has been some evidence that a moderate intake may even lead to a slight decreased risk of EAC. Recently a meta-analysis provided definitive evidence of the absence of any association between alcohol consumption and EAC and GCA, even at high doses.


Helicobacter pylori


Helicobacter pylori is a World Health Organization class I carcinogen and is associated with an increased risk of gastric cancer and other gastrointestinal (GI) malignancies. However, studies have highlighted an inverse association between H pylori gastric colonization and the risk of EAC, such that eradication strategies have been linked to the increased incidence of EAC. A meta-analysis of 19 studies on the subject confirmed that H pylori infection is inversely associated with EAC risk, with an OR of 0.56 (95% CI, 0.46–0.68), which appeared to relate to cytotoxin-associated antigen A (CagA)-positive strains (OR, 0.41) and not to CagA-negative strains (OR, 1.08).


Diet


There is some evidence that dietary factors can modulate the risk of EAC. Epidemiologic studies suggested a protective effect of high fruit and vegetable intake, whereas red meat and processed food can confer an increased risk for EAC. However, there are substantial obstacles in accurately assessing an individual dietary intake over time, and further research is needed to better estimate the impact of dietary factors on EAC risk.




Intervention to reduce clinical impact of esophageal cancer


Modification of Risk Factors


The epidemiologic trends of EAC in the Western world, discussed earlier, have raised concern about this disease amongst public health officials and as a result, there has been a call for clinical and research strategies to limit the impact of this disease. Because the modifiable risk factors implicated in the pathogenesis are still not completely understood and because compliance with healthy-living campaigns is poor, it is difficult to promote strategies to change the risk profile at a population level. For example, obesity, which has been related to the increase of incidence of EAC, has been implicated in a variety of diseases for many years and social campaigns have been instigated to sensitize the population. However, despite these efforts, obesity is an increasing problem of the modern Western world. Chemoprevention is an alternative strategy; however, there is lack of robust evidence to support this approach using currently available drugs to control reflux disease or to interfere with inflammatory pathways. However, the results of randomized controlled trials such as AspECT are awaited to shed further light on this. An alternative approach is therefore to focus on early detection.


Screening


The poor survival rates of EAC are mostly related to the fact that symptomatic disease correlates with advanced stage (≥T3N1), whereby only 1 in 7 patients is predicted to be alive at 5 years, despite all the therapeutic efforts to achieve a cure. Diagnosis of disease at an earlier stage is therefore paramount and is a distinct possibility, because EAC has a well-established pathologic sequence, whereby it is preceded in most cases by BE. Furthermore, the progression of BE to cancer is gradual and occurs through dysplastic stages, namely low-grade dysplasia (LGD) and high-grade dysplasia (HGD). Several retrospective series have shown that a diagnosis of EAC in patients with who were previously in endoscopic surveillance correlates with improved pathologic staging and better survival. However, randomized studies confirming this finding are lacking, and surveillance strategies for BE do not seem to have improved survival from a population perspective. One prime reason for this situation is likely to be that most persons with BE remain undiagnosed in the community, and therefore EAC is more likely to occur de novo in individuals without the benefit of a diagnosis of BE. To overcome this situation, endoscopic screening for BE has been considered for individuals with GERD. Studies on the cost-effectiveness of endoscopic screening have shown a wide range of incremental cost-effectiveness ratios (ICER), leaving uncertainty about its feasibility in clinical practice. Moreover, the recent evidence that the cancer risk in BE is significantly lower than previously believed makes this strategy even less cost-effective. The implications are that an expensive and invasive test (endoscopy) would be used to diagnose a prevalent condition such as BE (about 10% of the GERD population), which carries a relatively low cancer risk (0.12%–0.38% per year ). This large cohort of patients would then need to be monitored with even more expensive and invasive tests (endoscopy + multiple biopsies). This regime is clearly difficult to propose in the current financial climate, and it is also doubtful whether the invasiveness of this approach and the consequent impact on the quality of life could be justified. A solution would therefore require a less costly and invasive alternative. Furthermore, to make this screening strategy more cost-effective, once BE is diagnosed, there needs to be a method for identifying high-risk patients so that surveillance (and treatment) programs can be restricted to those at significant risk for EAC. The recent advent of less invasive endoscopic techniques for the ablation of dysplastic BE fits well into this clinical algorithm, in that patients with precancerous conditions and at high risk of progression can be offered ablation therapy for the prevention of cancer.




Screening modalities


Table 2 provides a summary of the diagnostic modalities proposed for BE and EAC screening.



Table 2

Summary of interventions proposed for BE and EAC screening




















































Conventional Endoscopy Transnasal Endoscopy Conventional Capsule Endoscopy Cytosponge Balloon Cytology Occult Blood Bead
Patient preference + ++ ++ + ++
Accuracy +++ ++ + +
Sampling +++ ++ ++ + +
Primary care + ++ +++ + +++
Cost-effective ? ++ ? ?


Conventional Endoscopy


Conventional endoscopy (CE) is the gold standard for the surveillance of BE, because it allows complete and high-quality visualization of the distal esophagus as well as tissue sampling for histologic diagnosis. CE has been studied as a screening modality for BE and early EAC. Two studies used CE to examine the prevalence of BE in patients undergoing colonoscopy. In the study by Rex and colleagues of 961 individuals, the overall prevalence of BE was 6.8% (5.6% in patients without a previous history of heartburn and 8.3% in patients with a previous history of heartburn, although this difference did not reach significance in the multivariate analysis). Gerson and colleagues detected BE in 27 of 110 asymptomatic veterans (25%). In a further study by this group, BE was detected in 6% of women undergoing either colonoscopy or bariatric surgery. Although these studies investigated a small and selected population of individuals, they suggest that BE, defined as the presence of intestinal metaplasia above the gastroesophageal junction, is a common finding even in the absence of reflux symptoms. Gupta and colleagues analyzed the cost-effectiveness of CE screening in the general population at the time of the screening colonoscopy. Considering that this strategy can lead to diagnosis of any esophageal malignancy (EAC and ESCC) as well as gastric cancer, the scenario of screening endoscopy plus BE surveillance was associated with an ICER of $95,559 per quality-adjusted life-year (QALY), which is comparable with that of other screening interventions performed in the United States, such as mammography for breast cancer or endoscopic surveillance for ulcerative colitis. When considering endoscopic screening as a separate intervention distinct from colonoscopy, most of the studies published so far have assumed a cancer risk that was higher than that indicated by more recent studies, and furthermore, they did not model endoscopic therapies, making the conclusions difficult to extrapolate to the current clinical practice. Further studies are therefore needed, and endoscopy cannot be recommended as a screening modality for the general population. However, American societies do suggest discussing the risks and benefits of screening with high-risk patients (white men older than 50 years, with a high BMI and long-standing GERD).


Transnasal Endoscopy


Transnasal endoscopy (TNE) has been studied and shown to be a valid alternative to CE for a diagnosis of BE. TNE has the advantage that it does not involve contact with the root of the tongue and does not trigger the gag reflex, and as a result, it does not require sedation and is better tolerated than CE. The study from Shariff and colleagues compared CE and TNE using a randomized crossover design and found a similar diagnostic accuracy for both techniques. TNE is now available with office-based technologies that are compatible with a primary care setting, with the advantage of significantly reducing the costs of screening. One of them (EndoSheath, Vision-Sciences, Orangeburg, NY) has been compared with CE in a single-center randomized study in a cohort of 121 patients enriched for BE. This study found that office-based TNE can be used as a screening modality in a tertiary referral center and had a moderate agreement (κ = 0.59) and a significantly smaller biopsy size when compared with CE. Importantly, 71% of patients preferred TNE over CE. A prospective multicenter study, which enrolled more than 400 individuals, assessed the feasibility of this technique as a screening modality in a primary care setting and found that office-based TNE is feasible and can significantly affect the clinical management of screened individuals, in whom the prevalence of erosive esophagitis and BE was 34% and 4%, respectively. It remains to be established whether a screening approach with this technology in patients with GERD is cost-effective.


Esophageal Capsule Endoscopy


Esophageal capsule endoscopy (ECE) has been extensively investigated as a screening modality in patients with GERD. ECE has the advantage of avoiding intubation and therefore has the potential to be well tolerated. Limitations include the absence of histologic sampling and the smaller number of total frames per centimeter compared with the small bowel capsule endoscopy, which negatively affects the diagnostic accuracy. Initial experience with the first-generation ECE, which had an image quality lower than 10 frames/s, showed a suboptimal sensitivity for BE of between 60% and 67%. A small study evaluating a second-generation ECE (18 frames/s) reported 100% sensitivity and 74% specificity for BE, but included only 28 patients. A third-generation ECE that images the esophagus at a rate of 35 frames/s is now available and studies are awaited to assess its diagnostic accuracy in esophageal diseases. The previous limitation of low image frequency has also been addressed with the use of string capsule endoscopy (SCE), to slow the transit through the esophagus. Ramirez and colleagues evaluated SCE in 100 veterans with reflux disease and assessed its diagnostic accuracy. Using the endoscopic diagnosis alone as the gold standard, the sensitivity and specificity of SCE for BE were 78.3% and 82.8%, whereas when SCE was compared with endoscopy with histologic confirmation, the sensitivity and specificity were 93.5% and 78.7%, respectively. Although we look forward to studies evaluating the new-generation ECE, the question of the cost-effectiveness of the ECE remains to be addressed, because a cost-usefulness study performed in 2007 showed that ECE and standard esophagogastroduodenoscopy performed similarly, suggesting that in the current context, the ECE may not be cost-effective.


Balloon Cytology


Nonendoscopic devices have been proposed as screening tools for premalignant and early malignant esophageal conditions. Balloon cytology allows collection of superficial cells for cytologic analysis and is cost-effective and better tolerated when compared with endoscopy. However, although cytology from balloon cytology sampling can identify abnormal cells in patients with HGD and EAC, the yield of goblet cells for a reliable diagnosis of BE has been shown to be poor and not adequate for it to be proposed as a screening tool for detecting premalignant lesions.


Cytosponge


The Cytosponge coupled to a diagnostic biomarker for BE has recently been proposed as a screening modality. The Cytosponge is a nonendoscopic cell collection device in which a capsule is swallowed and a sponge expands in the proximal stomach over a period of 5 minutes. This sponge samples the gastric cardia, gastroesophageal junction, and distal and proximal esophagus in turn before being pulled out through the mouth. The cytologic specimen is then examined for expression of the biomarker Trefoil factor 3, which was ascertained to be a Barrett’s-specific marker from a gene expression profiling experiment that compared gene expression between samples of gastric cardia, BE, and normal squamous esophagus. In a primary care screening study, this test was applied to more than 500 individuals with a history of reflux disease and the diagnosis compared with that obtained from standard endoscopy and biopsies. The primary aims of the study were to determine patient acceptability as well as the feasibility of administering the test in primary care. The prevalence of BE in this population was 3%, and hence there was limited power to determine sensitivity and specificity. Nevertheless, the data were encouraging, with a sensitivity of 73% and 90% for BE segments of at least 1 and 2 cm, respectively, with specificity figures more than 90%. These preliminary data compare well with other screening tests for other conditions such as the fecal occult blood test for colorectal cancer and the prostate specific antigen test for prostate cancer. A cost-usefulness study comparing Cytosponge and endoscopy as screening interventions in patients with GERD found that the 2 did not differ much in terms of the number of QALYs gained; however, the Cytosponge was more cost-effective. Further studies in larger cohorts are needed to confirm this diagnostic accuracy and whether this test can be proposed as a screening tool at a population level.


Occult Blood Bead


Another noninvasive test studied for screening for upper GI malignancy, including EAC, is the occult blood bead, which consists of a blood detector connected to a string that is swallowed by the patient. This test has been studied in large Chinese cohorts at high risk for upper GI malignancy and yielded a positive result in 12% to 24%. Subsequent upper GI endoscopy showed a malignancy in the stomach or esophagus only in a few individuals, leading to a specificity of about 3%. Because only people with positive test underwent endoscopy, the sensitivity of the test remains unknown, and in addition, this test is unlikely to detect BE.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Sep 6, 2017 | Posted by in GASTROENTEROLOGY | Comments Off on Screening and Risk Stratification for Barrett’s Esophagus

Full access? Get Clinical Tree

Get Clinical Tree app for offline access