Epidemiology of Barrett’s Esophagus and Esophageal Adenocarcinoma

Background: Barrett’s esophagus (BE) can progress to esophageal adenocarcinoma (EAC), a tumor characterized by rapidly increasing incidence and poor survival. We describe patterns of occurrence and risk factors for BE and EAC and how they may influence healthcare. Methods: We searched PubMed and Cochrane databases for English-language publications using the search terms Barrett’s esophagus and esophageal adenocarcinoma. We mainly considered systematic reviews with meta-analysis, randomized clinical trials, population-based observational studies, and international clinical guidelines. The results were synthesized into a narrative review. Results: The prevalence of BE (currently approximately 1% among adults) and incidence of EAC (current global age-standardized incidence rate of 0.9/100 000 person-years) have increased in many regions during the last 5 decades, particularly in North America, Northwestern Europe, and Australia. The increasing incidence is likely attributed to the increasing prevalence of the main risk factors, that is, gastroesophageal reflux disease (GERD) and obesity, combined with a decreasing prevalence of the protective exposure Helicobacter pylori-infection. GERD and obesity are now increasingly prevalent also in Asia, indicating that the incidence of EAC will continue to increase globally. Dysplastic BE and early EAC are readily endoscopically treated, and represent ideal conditions for screening because of their strong association with more advanced EAC. But such screening requires targeting of high-risk individuals, who remain to be better delineated. Conclusions: The changing prevalence rates of GERD, obesity, and Helicobacter pylori-infection might drive continued increasing EAC incidence rates worldwide. Targeted screening programs incorporating combinations of risk factors for BE and EAC may reduce mortality in EAC.


Introduction
Barrett's esophagus (BE) is a specialized columnar metaplastic epithelium of the distal esophagus caused by longstanding gastroesophageal reflux disease (GERD). BE is a premalignant condition that in some cases develop into low-grade and high-grade dysplasia which might further progress into invasive cancer, that is, esophageal adenocarcinoma (EAC). 1 This tumor is highly lethal and biologically aggressive and often presents first at advanced stages when distant metastases have developed. This explains the poor overall prognosis (<20% 5 year survival in Western populations). The incidence of EAC has increased manifold in North America, Northwestern Europe, and Australia during the last 5 decades. This increase parallels the increasing prevalence rates of GERD and the other main risk factor, obesity, and the decreasing prevalence of the preventive factor, Helicobacter pylori (H. pylori)-infection. 1 Tailored detection of dysplastic BE or early cancer, where recent clinical research has revolutionized the treatment, may help decrease the overall mortality and morbidity in EAC. In this review, we provide an update of the occurrence and etiology of BE and EAC, and how this may have implications for healthcare.

Methods
Using the search terms Barrett's esophagus and esophageal adenocarcinoma, we searched PubMed and Cochrane databases for relevant publications written in English language. We particularly assessed systematic reviews with meta-analyses, randomized clinical trials, populationbased observational studies, and international clinical guidelines. The results were synthesized into a narrative review.

Barrett's Esophagus
Definitions BE develops as a consequence of long-standing gastroesophageal reflux, which induces a replacement of damaged esophageal squamous epithelium with acid-resistant columnar epithelium. For a diagnosis of BE, United States', Australian, and European medical societies require histopathological evidence from biopsies of intestinal-type epithelium containing prominent goblet cells, 2-6 supported by studies showing a low risk of EAC in metaplastic epithelium without these particular characteristics. But this definition is debated, and guidelines from the United Kingdom and the Asian-Pacific region are broader and include all types of esophageal metaplastic columnar epithelium in the definition of BE. [7][8][9] Occurrence The onset of BE, and thus its incidence, is almost impossible to ascertain. The condition develops indolently in patients with chronic gastroesophageal reflux without adding symptoms but may rather reduce existing symptoms of reflux. The diagnosis of BE requires endoscopy and biopsy and is often identified when performing endoscopy in patients with gastroesophageal reflux symptoms or by coincidence for any indication. 10 Instead of assessing the incidence, the prevalence is therefore the measure of choice for assessing its occurrence. The prevalence rates are also unpredictable because they heavily depend on clinical recommendations and access to endoscopy, which vary considerably geographically and over time. Yet, studies have indicated an increasing prevalence of BE outpacing the increase in endoscopy volume. 11,12 Because the prevalence of BE has mainly been examined among patients referred for endoscopy, it is not possible to make inference from these selected populations to prevalence rates in the general background populations. Comparisons of prevalence rates between studies, study populations, and geographical areas are also uncertain. The regional prevalence rates of BE seem to roughly follow the global variability of the prevalence rates of gastroesophageal reflux disease (GERD), which are highest in North America, Northern Europe, and Southern Asia. 13,14 Two European studies have attempted to assess the prevalence of BE in the general population. In the Kalixanda study, 3000 inhabitants in 2 neighboring communities in Northern Sweden were invited for a survey on abdominal symptoms. 15 Among these, 1000 out of 1568 (64%) agreed to undergo upper endoscopy, of whom 1.6% had a histologically verified BE. 15 A slightly lower prevalence of BE (1.3%) was found in the Loiano-Monghidoro study, where 1033 out of 1533 asked adults (67%) from either of 2 Italian villages underwent upper endoscopy. 16 Although selection bias is likely to have been present also in these studies and the results had low statistical precision, the findings should be closer to the true population prevalence than those originating from hospital-based studies.
A study from the United States of 961 patients who underwent routine screening colonoscopy in tertiary settings showed a high (6.8%) prevalence of BE. 17 The pooled prevalence of BE in Asia has been estimated at around 1.3%, but again without population-based design, and with large geographical variations, ranging from 4.1% in South Central Asia to 0.9% in Eastern Asia. 18 These variations might mirror the regional prevalence rates of GERD in Asia, which are substantially higher in South Asia than in the Asian-Pacific regions. 14 Data from South America, the Middle East, and Africa are scarce. A Brazilian study of 104 patients who underwent upper endoscopy for any indication estimated the prevalence of BE at 3.8%. 19 The prevalence of BE in the Middle East seems lower than in Western countries, with recent studies from Lebanon and Saudi Arabia showing 0.2% to 0.3% prevalence. 20,21

Risk Factors
Etiological research aiming to identify risk factors for BE is hampered by the lack of data on incidence of the metaplasia, and unreliable prevalence rates are used as surrogate measures. Nevertheless, it is well established that BE is caused by chronic and long-standing gastroesophageal reflux. A systematic review and meta-analysis of 6 studies analyzed the association between GERD and BE by length of the BE mucosa, finding a strong association in long-segment BE (≥3 cm in length) (odds ratio [OR] 4.9, 95% CI 2.0-12.0), but not in short-segment BE (<3 cm) (OR 1.2, 95% CI 0.8-1.7). 10 BE seems to have a polygenic hereditary component and familial aggregation can occur, although its role in non-Western countries is uncertain. 22 Men are at an approximately 2-fold increased risk of BE compared to women, and the risk increases gradually with older age. 23 Studies from the United States and Malaysia indicate that the prevalence of BE varies greatly between ethnic subgroups. In the United States, Caucasians are at highest risk of BE, followed by Hispanic, Asian, and Afro-American minorities. 24 In Malaysia, which has large subgroups of Chinese and Indian residents, Indians have a higher prevalence of BE than those of Chinese and Malayan ethnicities. 25 Obesity, defined by a body mass index ≥30, and in particular a central or intra-abdominal distribution of adipose tissue, increases the risk of BE. A systematic review and meta-analysis found that patients with central adiposity were at two-fold increased odds of BE (OR 2.0, 95% CI 1.5-2.6). 26 Abdominal obesity mechanically reduces the efficacy of the lower esophageal sphincter to prevent reflux, but obesity in general may also increase the risk through hormonal and inflammatory mechanisms. 26 A meta-analysis of 5 case-control studies indicated that tobacco smoking increases the risk of BE in a dosedependent manner (OR 1.5, 95% CI 1.0-2.2, for those with <15 pack-years; OR 1.5, 95% CI 1.0-2.1, for 15 < 30 pack-years; and OR 2.2, 95% CI 1.1-4.4, for 30-<45 years, compared to never smokers), in part due to relaxation of the lower esophageal sphincter and increased gastroesophageal reflux. 27 The above mentioned risk factors have been examined and validated as risk factors also in Asian countries. 7 Infection with H. pylori may lead to atrophy of the gastric corpus, thus decreasing the production of gastric acid and the risk of BE. A meta-analysis of 72 studies demonstrated a decreased risk of BE among H. pyloriinfected individuals (OR 0.7, 95% CI 0.6-0.8), and this negative association is a seemingly global pattern. 28

Implications for Healthcare
Most Western clinical guidelines recommend endoscopy screening for BE in patients with long-standing GERD combined with additional risk factors for BE, that is, age ≥50 years, Caucasian ethnicity, male sex, obesity, and tobacco smoking. [2][3][4][5]8 The American Gastroenterological Association's guidelines even suggest that patients with ≥2 of these risk factors may be considered for endoscopy screening irrespective of GERD symptoms. 6 Patients with heredity for BE may be considered for screening also in the absence of any risk factors. A systematic review with meta-analysis showed that the prevalence of BE was 12.2% among patients with GERD combined with at least one additional risk factor, and the prevalence increased for each incremental risk factor. 29 A recent Nordic cohort study indicated that a one-time screening endoscopy in patients with GERD was followed by a substantially decreased incidence of EAC for 5 years and a decreased mortality from EAC for over 10 years. 30 Yet, any routine use of screening is probably more effective in persons with additional risk factors for BE. It should also be noted that about half of all BE patients do not seem to have a history of reflux symptoms, meaning that a screening strategy requiring reflux symptoms will have low sensitivity. On the other hand, a screening strategy not requiring reflux symptoms will have low specificity. How to improve both sensitivity and specificity of screening for BE remains to be thoroughly investigated. 31 Asian-Pacific clinical guidelines do not recommend screening for BE because of the low prevalence reported in the region. 7 Another argument for not using screening in these regions is that most patients with upper abdominal discomfort undergo upper endoscopy to rule out gastric cancer, which is common in these regions. 7 Only 10% of EACs are identified in patients with known BE, mirroring the difficulties in screening for BE. 32

Esophageal Adenocarcinoma
Definitions EAC typically arises from BE mucosa and is thus usually located just above the gastroesophageal junction (where the longitudinal gastric folds begin). 33 Early symptoms of EAC are scarce or vague, and >75% of patients present with advanced disease, rendering an overall 5 year survival <20%. 1 The main symptoms are progressive dysphagia, involuntary weight loss, and fatigue, which should prompt urgent upper endoscopy. Endoscopy with biopsy for histopathological confirmation is needed to establish a diagnosis of EAC. EAC includes tumors with an epicenter up to 2 cm (Siewert II) below the gastroesophageal junction and all adenocarcinomas located above this level in the esophagus. 34 Occurrence EAC was once a rarity, but the incidence has from the early 1970s increased rapidly and continuously in many countries. 35 In 2020, 85 700 new cases of EAC occurred worldwide, translating to an average global age-standardized incidence rate of 0.9 per 100 000 person-years. 36 There are large geographical variations, however, with the highest incidence reported in North America, Northern Europe (particularly in the United Kingdom and the Netherlands), and Australia. 36 In the United Kingdom, recent incidence rates are as high as 7.2/100 000 personyears in men and 2.5/100 000 person-years in women, corresponding to an overall average annual increase in incidence of 5%. 35 As a consequence, EAC had surpassed squamous cell carcinoma as the most common histological subtype of esophageal cancer in most Western countries. Some reports indicate that the pace of the increasing incidence may have slowed down in more recent birth cohorts, 37,38 but worldwide data suggest a continuous increase. 39 If the rates continue to increase in the current pace, 141 300 new EAC cases are expected to occur in year 2040. 36 Compared to Northern Europe, the age-standardized incidence of EAC in Southern Europe is clearly lower, that is, 0.5/100 000 person-years, which is surprising given that the prevalence of BE seems to be similar in these two European regions. 15,16 Contrasting the development in the Western world, the incidence of EAC has remained lower in Asian populations, although the incidence may be on the increase also in some Asian countries, that is, in Israel, Japan, and Singapore. 40

Risk Factors
The risk factors for EAC are similar to those of BE, and include GERD, older age, male sex, Caucasian ethnicity, obesity, and tobacco smoking, whereas infection with H. pylori and high intake of fruits and vegetables prevent EAC development. 41 A meta-analysis found that at least weekly GERD symptoms increases the odds nearly fivefold (OR 4.9, 95% CI 3.9-6.2), but are not reported by 40% of patients with EAC. 42,43 GERD and obesity are associated risk factors, and combined result in seemingly multiplicative relative risk estimates of EAC, 44 and they also increase the risk independent of each other. 45 Both GERD and obesity demonstrate a dose-response association with the risk of EAC. 42,46 The association between GERD and obesity in relation to EAC varies with tumor subsite, with 2 population-based case-control studies showing weaker association with EAC situated at the gastric cardia (Siewert II) for both GERD symptoms (OR 2.0, 95% CI 1.4-2.9) and obesity (OR 4.3, 95% CI 2.1-8.7) compared to EAC located higher up in the esophagus (OR 7.7, 95% CI 5.3-11.4 for GERD symptoms; OR 16.2, 95% CI 6.3-41.4 for obesity). 42,46 EAC has an even stronger male predominance than BE, although the sex ratio varies considerably across geographical regions. 39 The male-to-female ratio is 8:1 in the United States, which is stronger than in any other non-sex-specific cancer. 47 This ratio is less pronounced in Asia (4:1) and Africa (1:1). 47 The cause of the male predominance remains largely unclear, but the increased severity of GERD with higher rates of esophagitis and a higher prevalence of abdominal obesity and tobacco use among men might contribute. However, these factors do not explain the geographical variations in the sex ratio.
The increasing incidence of EAC has largely been attributed to changes in the prevalence of its main risk factors. North American and European studies conducted in the 2000 to 2004 report an approximate 50% higher prevalence of GERD compared to studies conducted before 1995. 48 GERD is also increasing in East Asia, but this trend became apparent first in 2005 to 2009 (52% increase compared to before 1995). 48 The prevalence of obesity is now 35% to 40% of adults in the United States compared to 13% in the 1970s. 49,50 The prevalence of obesity in Asia is also rapidly increasing, but in similarity to GERD remains substantially lower than in North America and Europe, with 16% of adults classified as obese in China. 51 The decreasing prevalence rates of H. pylori infection is a third potential explanation for the increasing incidence of EAC, because H. pylori counteracts EAC development. 52 The risk of EAC development specifically in patients with BE is difficult to ascertain because of methodological issues. As described above, the onset of BE is impossible to determine, and alterations to the genome heralding cancer development may be well underway once the diagnosis is made, leading to an overestimation of the EAC risk. In fact, BE is often first noted just before or at the time of EAC diagnosis, with a meta-analysis indicating that up to 25% of all EAC arising from BE are diagnosed within the first year of the index upper endoscopy. 53 Particular attention and vigilance should therefore be put in the index endoscopic investigation determining the presence of BE, because small lesions in the mucosa representing early EAC might otherwise be overlooked. Endoscopic therapy of dysplastic BE may also artificially decrease the natural incidence of EAC, while the superior access to upper endoscopy in patients with BE may lead to biased earlier detection in comparison to the general population. The presence and degree of dysplasia in BE, which serves as a reflection of the mutational load of the metaplastic epithelium, is the most important risk factor for tumor progression. 54 Most patients with BE are negative for dysplasia, and these have a very low absolute risk of EAC (<0.5% per year). 9 The annual risk of EAC in patients with low-and high-grade dysplasia is approximately 0.6% and 7%, respectively, but the reported risks of tumor progression differ substantially between studies and whether or not the dysplasia is confirmed by an expert pathologist in this field. [55][56][57] Apart from visible lesions (nodularity or mucosal irregularities) which indicate the presence of early EAC, a most important endoscopic risk factor is the length of the BE. This length is associated with increased risk of tumor progression. A systematic review and meta-analysis found that each additional centimeter confers a 25% increased odds (OR 1.3, 95% CI 1.2-1.4) of developing EAC. 58 Apart from the main risk factors BE segment length and dysplasia, a systematic review and meta-analysis has highlighted other risk factors for tumor progression, that is, older age (OR 1.03, 95% CI 1.01-1.05 per year), male sex (OR 2.2, 95% CI 1.8-2.5), and tobacco smoking (OR 1.5, 95% CI 1.1-2.0). 58 Prediction models using various combinations of risk factors have shown relatively good discriminative ability and may become clinically useful to identify patients with BE at particularly high or low risk of EAC, although the routine use of such models are not currently endorsed in clinical guidelines. [59][60][61] Observational studies suggest that use of medication with proton pump inhibitors, statins, and non-steroidal anti-inflammatory drugs decrease the risk of tumor progression in BE, but these findings might be explained methodological errors and remain to be verified by randomized clinical trials. 58 In a multicenter randomized clinical trial (AspECT trial), where patients with BE were randomly assigned to high-or low-dose PPI combined with or without aspirin, aspirin use did not decrease the risk of developing EAC (time ratio 1.02, 95% CI 0. 64-1.64). 62 Similarly, high-dose PPI (40 mg esomeprazole twice daily) was not superior to low-dose PPI (20 mg once daily) in preventing EAC (time ratio 1.04, 95% CI 0.67-1.61) in this trial. 62 Antireflux surgery with laparoscopic fundoplication is at least as effective as PPI in controlling heartburn, and superior in preventing regurgitation and normalizing esophageal pH-levels. 63 Researchers have therefore hypothesized that fundoplication may prevent tumor progression in BE. Some indirect evidence of a protective effect of fundoplication exists, with reports showing downstaging of dysplasia and regression of short-segment BE following fundoplication, although the effect on long-segment BE seems limited. 64 Whether this translates into prevention of EAC remains largely uncertain and has not been found in large-scale observational studies. 65 In the light of current knowledge, the treatment of patients with BE should focus on relieving reflux symptoms rather than preventing EAC.

Implications for Healthcare
The last decades have witnessed a rapid development of the management and treatment of BE and EAC. Patients with dysplastic BE or early EAC (T1a) often undergo endoscopic treatment, that is, endoscopic resection of visible lesions and radiofrequency ablation of any remaining metaplastic epithelium. Endoscopic eradication therapy is associated with less perioperative morbidity compared to esophagectomy while retaining comparable survival rates. 66 The results of endoscopic eradication therapy are promising, but studies have thus far had a relatively short follow-up. The esophagus remains exposed to gastroesophageal reflux, and recent cohort studies have indicated that the cumulative incidence of recurrent BE after eradication may be over 50% at 8 years. 67 Thus, patients with complete eradication of BE need to undergo endoscopic surveillance, although the optimal frequency and duration of such practice remains to be established. Patients with non-dysplastic BE are typically enrolled in endoscopy surveillance programs to detect dysplasia and early EAC, but their efficacy is uncertain. Results from meta-analysis of observational studies indicate that patients enrolled in surveillance programs are diagnosed with EAC at earlier tumor stage, translating into improved survival, but at least some of this survival benefit is likely attributed to lead and length time bias. 68 Because of the low incidence of EAC in Asian-Pacific regions, surveillance of BE is not recommended in the absence of dysplasia. 7 Tumors invading deeper into the submucosa (≥T1b sm2) or into the muscularis propria (T2) should be treated with esophagectomy because of the higher risk of lymph node metastasis. These tumors are currently often resected using a minimally invasive approach, because well-designed studies have found fewer intraoperative and postoperative pulmonary complications, better quality of life, and improved long-term survival after minimally invasive compared to open surgery. [69][70][71][72] In locally advanced tumors (invading the adventitia or adjacent structures; T3-T4) or cN1-N3 (lymph node metastasis according to clinical evaluation), clinical trials have demonstrated improved survival with neoadjuvant chemoradiotherapy or perioperative chemotherapy in addition to esophagectomy. 73,74 Patients with residual disease following neoadjuvant chemoradiotherapy and esophagectomy may benefit from adjuvant treatment with the monoclonal antibody nivolumab, which in a recent randomized trial showed improved median disease-free survival compared to placebo (22.4 vs 11.0 months). 75 The overall survival benefits of prevention of EAC in BE is limited because of the low (10%) proportion of patient diagnosed with BE before EAC diagnosis, but advances in the treatment of EAC have contributed to improved survival trends, at least in North America and Northern Europe. 76,77 Tumor stage remains the dominating prognostic factor, while the recognition, risk-stratification, and treatment of dysplastic BE may be paramount for the prevention of mortality in EAC in the future.

Conclusions
GERD and obesity are the main risk factors for BE and EAC and, together with a decreasing prevalence of H. pylori, the increasing prevalence rates of these exposures contribute to the increasing incidence of EAC in Western countries. The prevalence of BE and incidence of EAC remain low in East Asia and many other parts of the world, but changes in the global prevalence of obesity and H. pylori infection might lead to increases of EAC also in these regions in the coming decades. Riskstratification of patients with GERD and BE combined with novel therapeutic options may improve the efficacy of the prevention of EAC in patients with dysplastic BE, but the role of screening remains uncertain.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Swedish Research Council (2019-00209), Swedish Cancer Society (180684)