Published: September 2020
Barrett esophagus (BE) is a premalignant condition associated with chronic gastroesophageal reflux disease (GERD) in which normal squamous epithelium in the distal tubular esophagus is replaced by columnar intestinal epithelium.1 BE is a known risk factor for esophageal adenocarcinoma (EAC), an aggressive malignancy in which 40 % of patients have metastatic disease at the time of diagnosis.2 BE has become especially relevant in recent years as a rise in the obesity epidemic in the United States has led to a parallel rise in the incidence of EAC. Mathematical models predict near doubling of mortality from EAC between the years of 2011 and 2030.3 This clinical review summarizes clinically relevant pathophysiology, epidemiology, and risk factors for BE and EAC and elucidates the latest recommendations in the screening and management of BE and early EAC.
As is widely known, GERD is the most significant risk factor for BE. Chronic acid and bile exposure in the esophagus facilitates an inflammatory environment and induces several metabolic and genetic changes at the cellular level including
As acid exposure in the distal esophagus continues, replacement of squamous epithelium by intestinal columnar epithelium ensues. The progenitor cells for origin of BE epithelium is controversial: it is thought to be either squamous cells in esophagus or from the stem cells in the esophagus, gastric cardia or bone marrow.4 In the setting of continued acid exposure over time, metaplastic cells proliferate and acquire several genetic abnormalities. This stage, pathologically termed as dysplasia, is potentially reversible if a healthy environment is restored. If left untreated, dysplastic BE cells have the potential to undergo malignant transformation, leading to esophageal cancer.
In the United States, it is estimated that 2% to 5% of the general population has BE.5 For patients with chronic or frequent GERD, the prevalence is 5% to 15%.6-7 Men are at 2 to 3 times higher risk of BE than women and tend to develop disease earlier in life.8 Overall, the median age at diagnosis is 50 to 60, although BE may be diagnosed in younger patients as well.9 The prevalence of Barrett esophagus is highest in whites (6.1%) and lower in blacks (1.6%) and Hispanics (1.7%) with a similar higher risk for men than women across races.10
As previously mentioned, there is a strong association between GERD and BE. Nearly half the population in the United States reports heartburn, however most never develop BE.11 It has been suggested that reflux symptoms experienced weekly or more frequently over the course of 5 years or more can induce metaplastic changes.12 Patients diagnosed with reflux at a young age are at risk for increased acid exposure over their lifetime.13 A case series of Veterans Affairs patients found that those with weekly GERD symptoms before age 30 were 15 times more likely to develop BE.13 Of note, symptom severity in GERD does not correlate with degree of acid exposure and subsequent development of BE.14-15 Half of patients with BE do not report heartburn symptoms prior to diagnosis, known as silent acid reflux. Additionally, the presence of hiatal hernia has been identified as an independent risk factor for BE. As the stomach herniates through the diaphragmatic hiatus, it pushes the gastroesophageal junction upwards and creates a straight path for entry of gastric refluxate into the distal esophagus. Patients with hiatal hernia are 4 times more likely to have long segments of BE.16
Obesity has been consistently identified as a risk factor for BE.17 The mechanism is thought to involve increased pressure on the stomach and also hormones secreted in adipose tissue that cause the lower esophageal sphincter to relax. Moreover, there is an elevated risk for hiatal hernia in obese patients.18 In a 2009 meta-analysis, patients with body mass index (BMI) greater than 30 were 1.4 times more likely to develop BE compared with those with a BMI less than 30.19 A pooled analysis from 2013 found that first quartile for waist circumference yielded a 2.24 times risk of BE in males and a 3.75 times risk of BE in males.20 A case control analysis noted that a high waist-to-hip ratio of 0.9 or greater for males and 0.85 or greater for females increased the risk of BE fourfold.21
Patients with a family history of BE or EAC are not only at increased risk of BE but also increased risk of malignant progression.22. BE was found in 24 % of those with positive family history compared with 5% in a control group.22 Family history was described as an independent risk factor for BE when adjusted for age, gender, and BMI.22 This association may be due to environmental exposures or genetic predispositions. Although many genes have been identified in familial cases of EAC in small studies, larger prospective studies are needed to determine their clinic-pathological relevance.23
Smoking has been associated with BE in the majority of the clinical studies conducted so far.24 This relationship is likely due to relaxation of the lower esophageal sphincter induced by nicotine. In a pooled analysis of 5 case-control studies, patients with BE were found to be 70% more likely to have ever smoked.24 BE was twice as common in smokers, with the risk accruing up to 20 pack years, after which there was a plateau.24
Alcohol consumption has not been found to be associated with BE development. Rather, moderate alcohol consumption may have a protective effect. High-quality evidence furnished by the International Barrett’s and Esophageal Adenocarcinoma Consortium pooled 5 studies to assess the risk of alcohol use on BE showed no consistent association between alcohol consumption and the risk of BE and there was no evidence for a dose-response relationship. Wine appears to be associated with a 29% reduced risk of BE.25
The transition point between squamous esophageal mucosa and columnar gastric mucosa is called the Z-line. Squamous mucosa is typically whitish-pink in color and columnar mucosa is salmon-colored or red (Figure 1a). According to the American College of Gastroenterology, BE is defined as salmon-colored mucosa that extends 1 cm or greater proximal to gastroesophageal junction (Figure 1b). BE typically starts circumferentially and ascends into the esophagus with irregular extensions called tongues. Gross description of lesions has been standardized according to the Prague classification, which measures the extent of circumferential and maximal involvement.26 Long-segments are 3 cm or greater and short-segments are less than 3 cm.26
Histologic diagnosis of intestinal metaplasia is needed to confirm the diagnosis of BE. If suspected on endoscopic evaluation, 8 random biopsies should be acquired to maximize yield. If segments are shorter (less than 2 cm) 4 biopsies per centimeter and one biopsy per tongue is adequate.27 It is worth noting that the British Society of Gastroenterology requires only columnar mucosa, with or without intestinal metaplasia, for diagnosis of BE as cardiac or fundic mucosa, may precede development of intestinal metaplasia.28 Intervention is not currently recommended in the United States, however repeat esophagogastroduodenoscopy (EGD) in 1 to 2 years may be reasonable.29
Dysplasia may be classified as indefinite for dysplasia (IND), low-grade dysplasia (LGD) or high-grade dysplasia (HGD) based on degree of cytological and architectural changes. LGD is particularly difficult to diagnosis as there is significant interobserver variability.30 It has been observed that 75% to 85 % of specimens labeled LGD by a community pathologists are downgraded to nondysplastic after review by an expert pathologist.31 Therefore, if any degree of dysplasia is suspected, it should be confirmed by a pathologist with expertise in BE.32 Occasionally, a specimen will be labeled IND in the setting of mucosal inflammation when underlying dysplasia cannot be confidently ruled out. Once IND is confirmed, acid suppression should be optimized for 3 to 6 months with a repeat EGD.1
Up to 40% of patients with EAC deny any history of reflux symptoms.33 This indicates that symptomatology alone is not predictive of BE or EAC. Factors such as demographics, lifestyle, and family history further stratify patients into high- and low-risk populations to aid in decision making (Table 1).1,34
Although EGD is currently the gold standard for screening, it is an expensive and invasive procedure that carries small risk of complications related to sedation and the procedure itself. Physicians may be reluctant to refer patients for invasive endoscopy in the setting of controlled or absent reflux. Despite screening efforts, BE remains underdiagnosed, as fewer than 10 % of patients with EAC carry a prior diagnosis of BE.35 Current research is focused on creating risk prediction models and efficient and cost-effective screening methods via alternative modalities.
|American College of Gastroenterology||American Gastroenterological Association||American College of
Men with chronic (> 5 yrs) or frequent (> once a week) GERD symptoms and 2 or more of the following:
• Age ≥ 50
Patients with the risk factors:
• Age ≥ 50
Men ≥ 50 with chronic (> 5 yrs) GERD symptoms and risk factors:
• Nocturnal reflux symptoms
BE = Barrett esophagus; BMI = body mass index; EAC = esophageal adenocarcinoma; GERD = gastroesophageal reflux disease.
Data from references 1 and 35.
Newer non-endoscopic screening methods have been developed and are showing promise as viable alternatives to EGD.
Unsedated transnasal endoscopy (TNE) utilizes a thin scope that is passed through the nose to examine the esophagus with minimal gagging and without requiring sedation. In a randomized trial, patients were assigned to TNE and EGD screening groups, then subsequently underwent an alternative procedure 3 weeks later.36 There was no difference in detection of BE using either endoscopy. At the end of the study, more than half of patients preferred TNE to EGD.36 Although identification of BE is comparable between TNE and EGD, diagnostic yield on biopsy has varied between studies.37 TNE forceps are small and consequently yield small biopsy specimens. A randomized trial from 2015 found that biopsy acquisitions was 83% for TNE and 100%.38 It is worth noting that patient willingness to undergo TNE is low at approximately 46%.38 Additionally, the narrow scope channel results in poor suction and air function and its flexibility makes the scope difficult to manipulate.33
Cytosponge is an ingestible capsule attached to a string that quickly degrades in the stomach and releases a sponge that may be removed through the mouth, brushing the esophagus along the way. Specimens are assessed for BE via immunohistochemical staining for biomarker trefoil factor 3. The Barrett's Esophagus Screening in a Case-Control Study BEST2 trial in 2014 was a case-control trial that described the sensitivity of Cytosponge was 80 % for 1 cm segments, which increased to 87% for segments 3 cm or greater and the overall specificity was 92%.39 The Cytosponge procedure is safe and well tolerated by patients and may be appropriate in the primary care setting.40
Similar devices are developed to sample the esophageal mucosa (EsophaCap, Esochek) and diagnose BE using DNA biomarkers such as methylated CCNA1 DNA and VIM DNA.41 Esophageal capsule endoscopy is another noninvasive, unsedated imaging technique for visualization of the esophagus using a camera. However, biopsies cannot be obtained and it proved to be cost ineffective. Other screening techniques in development are liquid biopsies where dysregulated microRNA which are cancer specific are tested in the blood and “electronic nose” devices that detect exhaled volatile organic compounds that are altered in different disease states. While these devices have very good patient acceptance, validation in large studies is warranted.
The risk of EAC in BE depends on the degree of dysplasia. On average, the yearly risk of progression to EAC for BE based on dysplasia are as follows: nondysplastic BE (NDBE) 0.2% to 0.5 %, LGD 0.7 %, and HGD 7 %.1 Increasing age (odds ratio [OR] 1.03; 95% confidence interval [CI] 1.01–1.05) and male gender (OR 2.16, 95% CI 1.84–2.53) are known to increase risk of progression in BE patients.42 Other modifiable risk factors include smoking and visceral obesity. Proton pump inhibitor (PPI) and statin use are protective against progression (OR 0.55 and 0.48, respectively).43 Risk of progression increases by 25 % for every 1cm increase in BE length.44 Nodularity is associated with a six fold increased risk of progression in patients with LGD BE.42 Biomarkers for progression include P53 overexpression, loss of heterozygosity of TP53and CDKN2A, and presence of aneuploidy.45
Several risk predictions models have been designed based on these factors to identify patients at higher risk of progression that may benefit from early intervention. In one model, patients are assigned a sum total of a score based on 4 variables (male gender 9 points, smoking 5 points, length of BE segment 1 point for every1 cm increase and LGD 11 points). Patients are then stratified into low risk (0-10 points, 0.13% per year,), intermediate risk (11-19 points, 0.73% per year) and high risk (>20 points, 2.1% per year). However, this model need to be externally validated before they are used in clinical practice.46
Interestingly, nonsteroidal anti-inflammatory drugs and aspirin seem to be protective against development of BE progression in some epidemiological studies. This association was noted in a large randomized controlled trial of patients with BE (AspECT trial) in which time to event analysis (death, cancer, and high-grade dysplasia) was performed using a combination of PPI and aspirin therapy.47 Combining high-dose PPI with aspirin had the strongest effect compared with low-dose PPI without aspirin (time to event ratio of 1·59, 1·14–2·23, P = 0·0068).47 Weighing the moderate risk of bleeding associated with anti-inflammatories and the minimal risk of disease progression in NDBE patients, chemoprevention is not recommended. Statins have also been shown to reduce the risk of BE progression, especially among obese patients. A 2014 case-control study showed 43% reduction in risk with statin use, which increased to 74% in obese patients (defined as BMI ≥ 30). This observation is not seen with other cholesterol-lowering medications.48
Once BE or dysplasia have been established or treated, enrollment in a surveillance program should be discussed. Recommendations regarding surveillance intervals are currently dictated by presence and grade of dysplasia (Table 2).1 If endoscopic eradication therapy (EET) is subsequently performed and NDBE is eradicated, exam intervals are drawn out to assess intervention durability (Table 3).1
|American College of Gastroenterology||American Gastroenterological Association||American College of
Every 3-5 years
Every 3-5 years
None or every 3-5 years
|Indefinite for dysplasia|
|Low-grade dysplasia||Every 12 months if untreated||Every 6-12 months if untreated||None or every 3-5 years|
|High-grade dysplasia||Every 3 months if untreated||Every 3 months if untreated||Every 3 months if untreated|
Data from reference 1.
No current recommendations
|Low-grade dysplasia||Every 6 months for the first year, annually thereafter|
|High-grade dysplasia/Intramucosal cancer||Every 3 months for the first year, every 6 months for the second year, annually thereafter|
Data from reference 1.
Surveillance biopsies are acquired using the Seattle protocol, which suggests 4-quadrant biopsies along the length of BE.49-50 For NDBE, 4-quadrant biopsy is recommended every 2 cm. LGD is bit more intensive requiring 4-quadrant biopsies every 1 to 2 cm. Patients with HGD should have 4-quadrant biopsies every 1 cm and additional biopsies of gross mucosal abnormalities. Any degree of dysplasia should be confirmed by an expert pathologist.
Before a patient commits to a surveillance program, they should be appropriately counseled regarding the limitations of these programs. Although most studies support the role of routine surveillance,51 others refute its usefulness. This is due to observation that patients with known BE who progress to EAC are often adherent to a surveillance program.52 Scoring systems and biomarkers are being studied to separate high- from low-risk patients and will likely influence future recommendations regarding surveillance intervals.53-54
Effective management of GERD and risk-modification are important for prevention of neoplastic progression in BE patients. Acid control is typically achieved by use of PPI therapy. PPI regimen should only be intensified to relieve bothersome symptoms of reflux or in setting of esophagitis. Similarly, anti-reflux surgery should be reserved for patients with symptoms uncontrolled by medical therapy. Fundoplication has not been shown to protect BE patients from progression of disease.55 Other modifiable risk factors for progression to EAC such as smoking and obesity need to be addressed.
Advances in endoscopic technology over the past two decades have expanded treatment options to a wide variety of BE patients who may otherwise have required invasive surgery. EET combines mucosal resection and ablation techniques to eliminate aberrant mucosa from the esophagus and prevent progression to EAC. EET begins with careful inspection of the mucosa for abnormalities such as nodules or ulcers, which often contain neoplastic tissue. If found, the next step is resection of the lesions for histologic evaluation, staging and therapy. This is typically accomplished using endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD) techniques. Subsequently, patients undergo ablation of flat BE mucosa to prevent metachronous neoplasia.56 Radiofrequency ablation (RFA) has the highest-quality evidence and is the most utilized modality.57,58
EET is not routinely recommended for non-dysplastic BE due to low risk of progression, however there may be some exceptions. EET may be considered in NDBE patients at high risk of progression such as young patients, familial cases of BE and those with long segment BE.59 An individualized approach with discussion regarding risks and benefits should take place in these situations. There is debate surrounding the utility EET for LGD given the low risk of progression and unclear long-term benefit.60 In general, once LGD is confirmed on biopsy, anti-secretory therapy should be optimized followed by repeat EGD in 3-6 months. This may result in downgrading of LGD.1 If LGD persists on repeat exam, EET is recommended.56 If patient does not wish to pursue EET, annual surveillance is an acceptable alternative.61 Management of HGD and early cancer is well-defined given the high risk for progression to invasive EAC.
In patients with early EAC, ie, disease confined to mucosa or submucosa, EMR is more sensitive than EUS for accurate staging. In patients with IMC on EMR, EUS findings were T1a in 23.6%, upstaged in 18.4% and down-staged in 57.8%.62 In patients with IMC, risk of lymph node metastases is negligible (0 to 2%) and hence can be considered for EET in the absence of lymphovascular invasion and poor differentiation. EET is also considered in selected cases of submucosal tumors if they are well differentiated, confined to upper one thirds of submucosa and do not have any clinical evidence of lymph node metastases.1
There a several modalities of EET available. Resection techniques include EMR and ESD. Ablation techniques include RFA, photodynamic therapy (PDT), cryotherapy, and argon plasma coagulation (APC). Similar recurrence rates have been found among the different strategies.
EMR is accomplished by raising of lesions followed by snare resection for therapeutic and diagnostic purposes. ESD utilizes resection of lesions en bloc, regardless of size. Lesions are raised by injection and resected circumferentially with an electrosurgical knife with subsequent dissection of the submucosa. Removal of large segments in one piece allows for accurate histologic assessment and staging. Although ESD tends to achieve greater disease-free margins than EMR, there is no difference in complete remission from neoplasia.63 ESD is associated with a greater risk for perforation and strictures. It is typically reserved for patients with lesions larger than 15 mm, poor ability to raise lesion, or those with concern for submucosal invasion.64
RFA is the most widely used ablation technique for the management of dysplasia in BE. The fundamental principle for the use of RFA is the generation of an electrical current by radiofrequency waves that, when applied to mucosa, induces thermal injury and necrosis. This method is highly effective and safe. RFA achieves eradication of dysplasia in 92% to 98% and metaplasia in 88% to 91%.32,60 Complications are infrequent. Esophageal strictures may occur in 5% to 6% of patients and may require subsequent dilation.65 few patients have post-procedural chest pain, which may be severe, but is self-limited to 4 to 5 days. Bleeding and perforation are extraordinarily rare.56
Cryotherapy is a newer technique that has demonstrated high eradication rates in recent cohorts. In this technique, cryogens freeze tissues to cause rapid cellular necrosis followed by self-induced cellular apoptosis over several days owed to membrane disruption, protein denaturation, and impedance of blood flow. Cryogens are delivered via spray or expandable balloon. Cryoballoon has achieved eradication of dysplasia and intestinal metaplasia in 95% and 88% of patients, respectively.66,67 Cryospray eliminates dysplasia in 75% of cases and intestinal metaplasia in 88%.68 Although Cryospray demonstrates lower efficacy than Cryoballoon, it carries the advantage of being non-contact, so it may ablate uneven surfaces. Pain is significantly less with cryotherapy than with RFA, however strictures occur in 35 to 13% of cases and minor bleeding in 2 %.67-69 Currently, indication for cryotherapy includes disease refractory to RFA, including those with HGD and early esophageal cancer.
PDT was one of the first endoscopic interventions developed for BE. A photosensitizer is injected intravenously and is preferentially taken up by neoplastic tissue. It is activated by application of laser and leads to tissue destruction. Efficacy is inferior to RFA, with elimination of dysplasia in 80%.70,71 Additionally, PDT has a significant side effect profile, including strictures and photosensitivity. PDT may still be used for palliative purposes in patients with advanced disease who are poor surgical candidates.1
APC is a non-contact method that uses argon gas to deliver a high-frequency current to cause thermal injury to target tissues and was used for ablation of NDBE. This technique was historically associated with strictures, perforation, and buried BE. Its safety and efficacy were improved by addition of a saline injection step to raise lesions prior to ablation, which is now known as hybrid APC. In a recent trial of 50 patients, there was 96% complete eradication of intestinal metaplasia after a median of 3.5 hybrid APC sessions and 2% stricture rate.57