Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 7.
Published in final edited form as: Am J Gastroenterol. 2025 Mar 7;120(11):2520–2528. doi: 10.14309/ajg.0000000000003383

High Risk of Persistence and Risk of Dysplasia after Ultrashort Barrett’s Diagnosis

Wasseem Skef a,b,*, Jasmine Haydel c,*, Ashwin Rao a,b, Ronan Allencherril d, Rollin George a, Gyanprakash A Ketwaroo e, Aaron P Thrift f,g, Hashem B El-Serag a,b, Theresa Nguyen Wenker a,b
PMCID: PMC12266607  NIHMSID: NIHMS2060994  PMID: 40052679

Abstract

Background

The management of ultrashort (< 1cm) Barrett’s esophagus (BE) remains unclear. We aimed to determine the prevalence of ultrashort BE (USBE) at index diagnosis, identify factors associated with persistent BE after USBE diagnosis, and identify risk of dysplasia after initial USBE in a population of United States veterans.

Methods

This was a retrospective cohort study at the Veterans Affairs hospital in Houston, TX of consecutive patients with new BE diagnosis from 11/1990 to 6/2022 with follow-up through 4/2023. Using a pathology database, we identified patients with a new USBE diagnosis and any subsequent follow up EGD. We examined the association of sociodemographic and clinical risk factors of persistent USBE cohort compared to longer length segment BE and those with a negative follow up EGD after index USBE with chi-square tests and logistic regression models. Lastly, we calculated the prevalence and incidence of any dysplasia in persistent BE after USBE at index diagnosis compared to BE≥1cm. We excluded patients without at least 1 follow up endoscopy.

Results

Of 739 patients with BE, 167 (22.6%) had USBE on index EGD. Of those with index USBE, 86 (51.5%) had persistent BE and 67 (40.1%) had negative IM on follow up EGD. There was a greater proportion of non-Hispanic white and Hispanic than non-Hispanic black patients with persistent BE after index USBE and the negative follow up EGD cohorts (p=0.012), but no significant difference in age, sex, smoking status, alcohol status, and body mass index (BMI) between the two groups. White race (adjusted aOR 3.80; 95% CI 1.35–10.7) and Hispanic ethnicity (aOR 4.85; 95% CI 1.19–19.7; ref: non-Hispanic black) were associated with an increased likelihood of persistent BE. During 3,880.7 person-years of follow-up, 112 patients (10 persistent BE after index USBE) developed definite dysplasia/neoplasia. The incidence rate of definite dysplasia/neoplasia was 19.5 per 1,000 person-years (95% CI, 10.5–36.3 per 1,000 person-years) in those with persistent BE after USBE and 33.8 per 1,000 person-years (95% CI, 27.9–41.1 per 1,000 person-years) in those with longer segment BE (p-value by log-rank test=0.23; hazard ratio 0.67; 95% CI, 0.35–1.29). We did not identify any significant predictors of dysplasia in persistent BE after index endoscopy with USBE.

Conclusion

The prevalence of persistent BE after index USBE is high, and there is a risk of developing dysplasia/neoplasia in persistent BE after USBE. We were unable to detect a difference in the risk of dysplasia/neoplasia between persistent BE after index USBE and BE ≥1cm on index endoscopy in a cohort of United States veterans. White race and Hispanic ethnicity are associated with persistent BE after index USBE and may be target demographics for surveillance.

Keywords: Barrett’s esophagus, esophageal adenocarcinoma, dysplasia, incidence, risk factors

INTRODUCTION

The incidence of esophageal adenocarcinoma (EAC) rose from the 1970s to 2000s and has since remained elevated in the United States1,2. The 5-year survival remains poor at only 20%, and curative treatment is only possible in <30% of cases due to late stage at EAC diagnosis3,4. Barrett’s esophagus (BE), the replacement of normal squamous epithelium in the lower esophagus with columnar intestinal epithelium, is the only known precursor to EAC5. Detection and surveillance of BE have been shown to lead to earlier stage at EAC diagnosis and decreased EAC-related mortality6.

BE is diagnosed by endoscopic visualization of salmon-colored columnar mucosa in the tubular esophagus and histologic confirmation of the presence of intestinal metaplasia on targeted biopsies. Currently, some societal guidelines (American College of Gastroenterology [ACG], British Society of Gastroenterology [BSG], and European Society for Gastrointestinal Endoscopy [ESGE]) recommend biopsies of endoscopic salmon-colored columnar mucosa limited to those with length ≥1cm in the absence of any other visible abnormalities7,8,9,10 This recommendation was made due to the low interrater reliability for describing lengths <1cm (0.22) in prior studies of the Prague classification11,12.

However, other societal guidelines (American Gastroenterological Association [AGA] and the American Society of Gastrointestinal Endoscopy [ASGE] do not make recommendations for a minimum length threshold for the diagnosis of BE13,14. Data suggest that BE segments <1cm, also termed ultrashort BE (USBE), may harbor, or predispose to dysplasia and neoplasia. In a retrospective analysis of the GI Quality Improvement Consortium (GIQuIC)15 in the US, the authors found endoscopists routinely biopsy irregular z-lines and 3.8% of 10,986 endoscopies of patients with irregular z-line had concomitant indefinite or low-grade dysplasia, while 0.7% had high-grade dysplasia. In another German cohort study of 1017 patients with newly diagnosed T1 esophageal adenocarcinoma, 20% had had intestinal metaplasia segments <1 cm16. This information, coupled with the rising incidence of esophageal adenocarcinoma (EAC) in the United States17, has raised questions about our present screening and surveillance strategies and whether we are mistakenly ignoring USBE.18 In contrast, only 4 previous studies (with methodological limitations) suggest that IM segments < 1 cm have low risk of neoplastic progression.19,20,21,22

In this study, we aimed to (1) identify the prevalence of persistent BE after initial diagnosis of USBE and factors associated with persistent BE after USBE, and (2) estimate the incidence and prevalence, and identify risk factors for dysplasia after initial USBE in a population of United States veterans.

METHODS

Study Design and Population

We conducted a retrospective cohort study at the Michael E. DeBakey VA Medical Center (MEDVAMC) in Houston, TX using a pathology database. We included patients diagnosed with BE during esophagogastroduodenoscopy (EGD) from 11/1990 to 06/2022 at the MEDVAMC. Patients were followed through April 2023. We excluded patients without at least 1 follow up endoscopy.

BE was defined as the presence of intestinal metaplasia on histopathological exam of at least one biopsy specimen obtained from endoscopically suspected areas in the tubular esophagus. The index EGD was defined as the first endoscopy demonstrating BE with at least one esophageal biopsy. BE segment length was prospectively recorded using the established Prague classification system23. USBE was defined as columnar lined epithelium (CLE) <1 cm at the gastroesophageal junction (reported as irregular z-line or <1cm tongue on endoscopy findings) with biopsy proven goblet cell metaplasia. BE was defined as ≥ 1 cm CLE with biopsy proven intestinal metaplasia. Records of patients with ultrashort BE on index endoscopy were then reviewed and categorized into the 7 surveillance sequence groups based on findings and pathology of subsequent EGDs outlined in Table 1. We then organized the surveillance sequences into 2 cohorts (persistent BE after USBE on index endoscopy [groups 1, 2, 3] vs. negative BE on follow up EGD of USBE on index endoscopy [groups 4,5, 6]) for analysis. Patients with BE segment ≥ 1 cm on index endoscopy were categorized into group 8. Dysplasia status in all cases of low-grade dysplasia (LGD) or high-grade dysplasia (HGD) was assessed by 2 board-certified pathologists as part of institutional protocol. We considered the final endoscopy before any endoscopic eradication therapy as their progression.

Table 1.

Surveillance sequence groups after ultrashort Barrett’s esophagus on index endoscopy. Groups 1, 2 and 3 represent persistent Barrett’s esophagus cohort. Groups 4, 5 and 6 represent negative Barrett’s esophagus on surveillance.

Surveillance Sequence Interpretation of Sequence N (%)
Persistent BE after USBE on Index Endoscopy Group 1 +, +, + Strict persistent BE (consistently detected) 44 (5.9%)
Group 2 +, -, + ‘Persistent’ BE or potential sampling error vs recurrence on subsequent endoscopy 17 (2.3%)
Group 3 +, +, x Less than 3 EGDs, Strict persistent BE (consistently detected) 25 (3.4%)
Negative BE after USBE on Index Endoscopy Group 4 +, -, - Strict negative IM on surveillance 30 (4.1%)
Group 5 +, +, - ‘Negative IM’ on surveillance (or potential sampling error) 22 (3.0%)
Group 6 +, -, x Less than 3 EGDs, ‘Negative IM’ on surveillance (or potential sampling error) 14 (1.9%)
No surveillance after USBE on Index Endoscopy Group 7 +, x Less than 2 EGDs 15 (2.0%)
Longer BE on Index Endoscopy Group 8 BE segment≥1cm on IE BE on index endoscopy 572 (77.4%)
Open in a new tab

Abbreviations: BE= Barrett’s esophagus. EGD= esophagogastroduodenoscopy. IM= intestinal metaplasia. “+” indicates the presence of BE on endoscopy and histology. “-” indicates no endoscopic BE or BE suspected endoscopically but biopsies negative for BE. “x” indicates incomplete surveillance or no biopsy.

We performed a structured manual review of the VA Computerized Patient Record System (CPRS) for information on baseline variables including age at index EGD and follow-up EGD(s), sex, height, weight at each endoscopy, body mass index (BMI), ethnicity, smoking status (never, former, current), alcohol use status (never, former, current), hiatal hernia, PPI or H2 antagonist (H2RA) use, and biopsy volume (cm3). Smoking and alcohol use status were obtained from the endoscopist’s history and physical note at the time of the index EGD. This research was approved by the Institutional Review Boards for Human Subjects Research for Baylor College of Medicine and the VA Research and Development Committee of the Michael E. DeBakey Veteran Affairs Medical Center.

Statistical Analysis

For descriptive statistics, the mean and standard deviation (SD) were calculated. Continuous data were compared using Student’s t-test. Categorical variables were compared using Pearson’s chi-square test. The Fisher’s exact test was used when expected values were too small (≤5). Multivariable logistic regression analyses containing identified significant variables on univariate analysis (p<0.05) and known BE’s risk factors selected a priori were performed to identify predictors for persistent BE after USBE on index EGD.

We calculated the overall cumulative incidence and 95% confidence interval (CI) as well as incidence rate per 1,000 person-years for the whole study cohort and stratified by the two main groups: 1) persistent BE after USBE on index endoscopy and 2) segment length BE ≥1cm on index EGD. Follow-up time was calculated from index EGD to the EGD date of progression for those who developed dysplasia over the study period. For those who did not develop dysplasia, follow-up was from index EGD to the last EGD or date of death in those who passed during the study period. We calculated the incidence of any dysplasia/neoplasia (indefinite for dysplasia, LGD, HGD, and EAC) and definite dysplasia/neoplasia (LGD, HGD, EAC). Differences in incidence rates between groups were evaluated by log-rank test. We also estimated hazard ratios (HR) and 95% CIs with Cox Proportional Hazards models predicting any dysplasia and definite dysplasia for those with persistent BE after index USBE compared to those with BE length ≥1cm on index EGD.

Statistical analysis was performed using Stata version 15.1 (StataCorp, College Station, TX). A 2-tailed p-value of <0.05 was considered statistically significant.

RESULTS

Ultrashort BE Prevalence and Risk Factors

We identified 739 patients with BE, of which USBE on index endoscopy was present in 167 (22.6%) patients (Figure 1). Fifteen patients with USBE on index endoscopy were excluded from the analysis due to lack of follow-up EGD, while 86 (11.6% of the total cohort with BE) had persistent BE after USBE on index EGD. The overall average follow-up was 5.36 years (SD 5.8 years), while the median follow-up was 4.15 years. We compared patients with persistent BE after USBE on index endoscopy (n=86) to those with longer segment BE on index EGD (n = 572) and found no significant differences in age, sex, ethnicity, BMI, alcohol or smoking status, hiatal hernia, use of PPI or H2RA, biopsy acquisition volume or mean follow-up time (Table 2).

Figure 1.

Figure 1.

Open in a new tab

Flow diagram demonstrating the study population. Of 739 patients with Barrett’s esophagus (BE), 167 had ultrashort BE on index endoscopy and 572 had normal segment BE.

Table 2.

Socio-demographic and clinical characteristics of those with persistent Barrett’s esophagus (BE) after ultrashort BE on index endoscopy compared to those with a longer BE on index endoscopy.

Persistent BE after USBE on Index Endoscopy (n=86) Longer BE on Index Endoscopy (n=572) p-value
Age at diagnosis (years) 61.5 [SD±8.9] 61.5 [SD±9.1] 0.951
Sex 0.147
 Male 81 (94.2%) 555 (97.0%)
 Female 5 (5.8%) 17 (3.0%)
Mean Follow Up Time (years) 5.95 [SD±9.9] 5.27 [SD±4.7] 0.302
BMI 0.913
 < 25 17 (19.8%) 100 (17.4%)
 25–29.9 34 (39.5%) 228 (39.9%)
 ≥30 35 (40.7%) 242 (42.3%)
 Missing 0 (0.0%) 2 (0.4%)
Race/Ethnicity 0.731
 Non-Hispanic White 69 (80.2%) 463 (80.9%)
 Non-Hispanic Black 6 (7.0%) 51 (8.9%)
 Hispanic 11 (12.8%) 55 (9.6%)
 Missing 0 (0.0%) 3 (0.6%)
Smoking Status 0.193
 Never smoker 17 (19.8%) 156 (27.3%)
 Former smoker 48 (55.8%) 263 (46.0%)
 Current smoker 21 (24.4%) 153 (26.7%)
Alcohol Status 0.202
 Never drinker 29 (33.7%) 230 (40.2%)
 Former drinker 16 (18.6%) 127 (22.2%)
 Current drinker 41 (47.7%) 215 (37.6%)
Hiatal Hernia 0.072
 No 34 (39.5%) 171 (29.9%)
 Yes 52 (60.5%) 401 (70.1%)
PPI or H2RA use 0.312
 No 20 (23.3%) 163 (28.5%)
 Yes 66 (76.7%) 409 (71.5%)
Open in a new tab

Abbreviations: BE=Barrett’s esophagus. SD=standard deviation. BMI=body mass index. PPI=proton pump inhibitor. H2RA=H2 receptor antagonist.

Persistent BE after USBE on Index Endoscopy

Eighty-six patients had persistent BE on at least 1 follow up endoscopy after initial diagnosis with USBE on index endoscopy versus 66 patients with no BE on follow up EGD after USBE on index endoscopy (Table 3). There was no difference in follow-up time between the two surveillance cohorts (mean 5.85 vs 4.45 years, p = 0.277) or biopsy volume on index endoscopy (0.03 (SD±0.03 vs 0.04 (SD±0.03 cm3, p =0.156). Those with persistent BE after USBE on index endoscopy group were more likely to be non-Hispanic white (80.2% v 66.7%) or Hispanic (12.8% v 9.1%; p =0.011) than those with no BE on follow-up EGD after index USBE. On multivariable logistic regression controlling for known BE risk factors (Table 4), non-Hispanic white patients had 3.85 increased odds of persistent BE after USBE on index endoscopy compared to non-Hispanic Black patients (95% CI 1.37–10.8), while Hispanic patients had 4.78 increased odds of persistent BE after USBE on index endoscopy compared to non-Hispanic Black patients (95% CI 1.18–19.4).

Table 3.

Socio-demographic and clinical characteristics of those with persistent Barrett’s esophagus (BE) after ultrashort BE on index endoscopy compared to those with negative surveillance.

Persistent BE after USBE on Index Endoscopy 1
(n=86)		 Negative BE after USBE on Index Endoscopy 2
(n=66)		 p-value
Age at diagnosis (years) 61.5 [SD±8.9] 60.5 [SD±8.3] 0.480
Sex 1.000
 Male 81 (94.2%) 62 (93.9%)
 Female 5 (5.8%) 4 (6.1%)
Mean Follow Up Time (years) 5.85 [SD±10.0] 4.45 [SD±3.4] 0.277
BMI 0.368
 < 25 17 (19.8%) 14 (21.2%)
 25–29.9 34 (39.5%) 19 (28.8%)
 ≥30 35 (40.7%) 33 (50.0%)
Race/Ethnicity 0.011
 Non-Hispanic White 69 (80.2%) 44 (66.7%)
 Non-Hispanic Black 6 (7.0%) 16 (24.2%)
 Hispanic 11 (12.8%) 6 (9.1%)
Smoking Status 0.152
 Never smoker 17 (19.8%) 22 (33.3%)
 Former smoker 48 (55.8%) 29 (43.9%)
 Current smoker 21 (24.4%) 15 (22.7%)
Alcohol Status 0.280
 Never drinker 29 (33.7%) 27 (40.9%)
 Former drinker 16 (18.60%) 16 (24.2%)
 Current drinker 41 (47.7%) 23 (34.9%)
Hiatal Hernia 0.864
 No 34 (39.5%) 27 (40.9%)
 Yes 52 (60.5%) 39 (59.1%)
PPI or H2RA use 0.722
 No 20 (23.3%) 17 (25.8%)
 Yes 66 (76.7%) 49 (74.2%)
Biopsy Volume on index endoscopy (cm3) 0.03 [SD±0.03] 0.04 [SD±0.03] 0.156
Open in a new tab

Footnote:

1

Denotes surveillance sequence groups 1, 2 and 3.

2

Denotes surveillance sequence groups 4, 5 and 6.

Abbreviations: BE=Barrett’s esophagus.SD=standard deviation. BMI=body mass index. PPI=proton pump inhibitor. H2RA=H2 receptor antagonist.

Table 4.

Multivariate logistic regression predicting the odds of persistent any Barrett’s esophagus (BE) after diagnosing ultra-short BE on index endoscopy.

Adjusted Odds Ratio 95% Confidence Interval
Age ≥60 years (ref: <60 years) 1.26 0.60–2.63
Male Sex (ref: Female) 0.74 0.17–3.27
Race/Ethnicity (ref: non-Hispanic Black)
 Non-Hispanic white 3.85 1.37–10.85
 Hispanic 4.78 1.18–19.4
Body Mass Index (ref: <25)
 25–29.9 1.71 0.65–4.49
 ≥30 1.13 0.46–2.81
Smoking (ref: Never smoker)
 Former smoker 2.06 0.90–4.73
 Current smoker 1.63 0.60–4.42
Open in a new tab

Of the 86 with persistent BE after ultrashort BE on index EGD, 36 (41.9%) continued to have ultrashort BE (length <1cm) on follow-up regardless of endoscopist. Of the remaining 50, the great majority (n=39) had short segment BE (length ≥1 and <3cm) during follow-up, with most (n=30) having 1cm of BE as their longest length recorded during follow-up.

Dysplasia in Persistent BE after USBE on Index Endoscopy

Of those with persistent BE after USBE on index endocopy, 48 (55.8%) had nondysplastic BE, compared with 277 (48.4%) of those with longer segment BE (Table 5, Supplemental Figure 1). Definite dysplasia or neoplasia (defined as LGD, HGD or EAC) was detected in 5 patients (5.8%) on index EGD of those with persistent BE after USBE on index endoscopy, while 40 (7.0%) with BE segment ≥1cm, 1 with no surveillance after USBE, and none with no BE after index USBE had prevalent definite dysplasia/neoplasia (p-value 0.053 by Fisher’s exact test).

Table 5.

Frequency of dysplasia in those with persistent Barrett’s esophagus (BE) after ultrashort BE on index endoscopy, negative BE after USBE-IE, no follow-up after USBE on index endoscopy, and normal segment BE on index endoscopy.

Highest Detected Dysplasia Persistent BE after USBE on Index Endoscopy (n=86) Negative BE after USBE on Index Endoscopy
(n=66)		 No Follow-Up Endoscopy after USBE on Index Endoscopy
(n=15)		 Longer Segment BE on Index Endoscopy
(n=572)
Nondysplastic BE 48 (55.8%) 57 (85.1%) 12 (85.7%) 277 (48.4%)
BE Indefinite for Dysplasia 24 (27.9%) 9 (13.4%) 2 (14.3%)
prevalent		 164 (28.6%)
19 at index EGD 5 on follow-up 6 at index EGD 3 on follow-up 87 prevalent 77 incident
BE with Low-Grade Dysplasia 8 (9.3%) 0 1 (1.5%)
at index EGD		 98 (17.1%)
2 at index EGD 6 on follow-up 32 at index EGD 66 on follow-up
BE with High-Grade Dysplasia 5 (5.8%) 0 0 22 (3.6%)
3 at index EGD 2 on-follow-up 7 at index EGD 15 on follow-up
Esophageal Adenocarcinoma 1 (1.2%)
on-follow-up		 0 0 11 (1.9%)
1 at index EGD 10 on follow-up
Open in a new tab

Abbreviations: BE=Barrett’s esophagus. NDBE=nondysplastic BE.

Of note, in Table 5, under “Negative BE after index USBE,” there are 3 patients who developed BE indefinite for dysplasia on follow-up. These 3 patients were in surveillance group 5: their first EGD had nondysplastic ultrashort BE, the second EGD had incident BE indefinite for dysplasia, and the third EGD had no endoscopic or histologic BE.

The cumulative incidence of all dysplasia (defined as indefinite for dysplasia, LGD, HGD, or EAC) was 17.4% (95% CI 10.8–27.0%) in those with persistent BE after USBE on index endoscopy and 31.2% (95% CI 27.4–35.0%) in those with BE segment ≥1cm on index EGD. The cumulative incidence of definite dysplasia/neoplasia in those with persistent BE after USBE on index endoscopy was 11.6% (95% CI, 6.34–20.4%), and 17.8% (95% CI, 14.9–21.2%) in those with BE segment ≥1cm on index EGD.

During 3,527.1 person-years of follow-up, the incidence rate of all dysplasia/neoplasia in those with persistent BE after USBE on index endoscopy was 30.2 per 1,000 person-years (95% CI, 18.2–50.1 per 1,000 person-years), while the incidence rate among those with normal segment BE on index EGD was 66.2 per 1,000 person-years (95% CI, 57.2–76.7 per 1,000 person-years; p-value by log-rank test=0.02). Those with persistent BE after ultrashort index BE had a decreased risk of developing any dysplasia/neoplasia (including indefinite for dysplasia) compared to those with BE length ≥1cm on index EGD (HR, 0.54; 95% CI, 0.32–0.91).

During 3,880.7 person-years of follow-up, the incidence rate of definite dysplasia/neoplasia (LGD, HGD, or EAC) was 19.5 per 1,000 person-years (95% CI, 10.5–36.3 per 1,000 person-years) in the group with persistent BE after USBE-on index endoscopy and 33.8 per 1,000 person-years (95% CI, 27.9–41.1 per 1,000 person-years) in those with BE segment ≥1cm. There was no difference in incidence rates of definite dysplasia/neoplasia between persistent BE after USBE on index endoscopy and BE segment ≥1cm on index endoscopy (p-value by log-rank test = 0.23). We did not detect a difference in the risk of definite dysplasia/neoplasia between those with persistent BE after index USBE and those with index BE ≥1cm (HR, 0.67; 95% CI, 0.35–1.29).

Predictors of Dysplasia in those with Persistent BE after USBE- on Index Endoscopy

We performed univariate analysis among those with persistent BE after USBE on index endoscopy comparing patients with definite dysplasia or neoplasia (LGD, HGD, EAC) to patients with indefinite and nondysplastic BE. We found no significant differences in age, sex, race/ethnicity, BMI, alcohol or smoking status, hiatal hernia, and use of antisecretory medication despite statistically similar biopsy volume and mean follow-up time.

DISCUSSION

In a cohort of United States veterans, USBE on index endoscopy and persistent USBE constituted 22.6% and 11.6% of all patients with BE, respectively. We found no significant differences in risk factors between persistent USBE, and longer segment BE on index endoscopy. We identified that patients of white race and Hispanic ethnicity were at increased risk of persistent BE after initial diagnosis of USBE on index endoscopy. Prevalent definite dysplasia or neoplasia was present in 5.8% of patients with USBE on index endoscopy, while the cumulative incidence was 11.6% in those with persistent BE after USBE on index EGD. Notably, definite dysplasia prevalence and incidence rates were not different for USBE and longer segment BE, and there were no distinguishing risk factors to predict dysplasia in patients with USBE.

A high proportion of BE patients in our study had persistent BE after index USBE, however this may underrepresent the true prevalence of persistent BE after index USBE given that there is heterogeneity in endoscopist sampling of irregular z-lines at our center. Indeed, some data suggest that up to 44% of irregular z-line biopsies may harbor intestinal metaplasia.24 Curiously, the sociodemographic profile for persistent BE after index USBE and longer BE segment length was similar in our study. This is interesting because our current understanding is that Barrett’s rapidly progresses proximally over time25 to a maximum length with little change afterward26. Other unmeasured variables such as esophageal pH data, which has been shown to correlate with extent to intestinal metaplasia, may differentiate between USBE and BE ≥1cm.27Our findings of Hispanic ethnicity association with persistent BE after USBE on index endoscopy compared to those with no BE after USBE on index endoscopy warrants consideration. It is well known that White race is a risk factor BE28, but Hispanic ethnicity is typically considered lower risk for BE29. However, two prior studies have found Hispanic populations may have a similar BE prevalence to non-Hispanic white populations. In a retrospective cohort study of 1,541 patients, Bersentes et al30 found the prevalence of BE in White and Hispanic veterans was similar (5.3% and 3.8%, respectively, p = 0.563). Similarly, Keyashian and colleagues31 found that the prevalence of BE in Hispanic patients was also similar to non-Hispanic whites in a retrospective cohort study of 663 urban county hospital patients (10% vs 12%, p-value 0.75). Our findings are consistent with these observations yet are unique in that our results are among those with USBE diagnosed on index endoscopy. It is plausible that the higher prevalence of known risk factors for BE (i.e., GERD, central obesity, smoking) among those of Hispanic ethnicity may account for the increased prevalence of BE in these populations.

The four previous studies found a low risk of dysplasia or neoplasia progression in USBE but had some limitations. Jung et al19 conducted a population-based cohort study of 86 subjects with intestinal metaplasia at the GEJ at baseline, in which there was no progression to HGD of EAC in a median follow up time of 8 years. However, it should be noted that only 59 (65%) had endoscopic follow up. Similarly, Itskoviz et al20 evaluated the long-term development of dysplasia in 50 patients with USBE on index endoscopy and found that only 2 (4%) progressed to LGD and none progressed to HGD or EAC in more than 5 years of follow up. Similarly, 14 (21.9%) patients were lost to follow up, the study was conducted in Israel and may not be generalizable to a US population and there was also concern that in subjects with hiatal hernias biopsies maybe have been taken from the cardia instead of the GEJ. In another long term follow up study by Horwat el al21 of 47 patients with EGJ intestinal metaplasia on index endoscopy, none developed dysplasia over mean follow-up 3.5 years. However, it should be noted two had prevalent dysplasia at diagnosis that regressed on follow up EGD, and one had prevalent cancer and underwent esophagectomy. Furthermore, again this study was limited by poor follow-up with only 34 (72.3%) having follow-up endoscopy. Finally, in Thota et al’s study22, none of the 167 patients with non-dysplastic USBE on index developed dysplasia or neoplasia during median follow-up of 4.8 years. It should be noted that those who developed any dysplasia in the 1st year were excluded and many were excluded due to lack of follow-up. Our study is unique from the aforementioned studies because of its large cohort size (167 patients with USBE on index endoscopy) and is notable for follow up data on 154/167 (91.6%).

Perhaps the most alarming finding in our study was that the prevalence and incidence rates of dysplasia were high at 5.8% and 11.6% respectively in the persistent BE after index USBE cohort. However, our data is consistent with a previous study of 20% detected dysplasia in patients with BE segment <1cm32. This may be related to the high prevalence of known risk factors for dysplasia, including increased age, male sex, obesity, and smoking in our cohort33.This challenges the literature that mainly longer BE segment is associated with increased dysplasia risk34. Indeed, neoplasia risk in USBE may be higher than expected based on length because the esophagogastric junction is a hotspot for neoplasia given higher exposure to known carcinogens such as nitric oxide and gastric acid5. This is further supported by the fact that the anatomic location of BE neoplasia recurrence after endoscopic eradication therapy is centered around the esophagogastric junction35.

While our observations may be used to support a practice of initial biopsies of suspected endoscopic BE <1cm as well as surveillance of patients with these lesions, the cost and burden of both practices may be prohibitive. A previous study suggests that 12,000 endoscopies would be required to diagnose one EAC in USBE36 and coupled with the fact that 70% of EAC cases occur in the setting of visible BE length >1cm, the most cost-effective prevention program remains to focus on longer segment disease. However, given practice patterns of providers routinely diagnosing USBE15, it is prudent to continue to collect and analyze this information to better identify determinants of persistent BE and progression to dysplasia in patients diagnosed with USBE. Tissue systems pathology assay testing, which has been shown to predict progression in patients with BE, could potentially be applied in USBE.37

Strengths of our study include the large cohort and longitudinal design with a long follow-up period of 5.4 years coupled with a lack of a significant difference in follow-up length among the study groups. Additionally, we attempted to control for adequate sampling by comparing quantitative biopsy volume and found no difference among study groups. We believe that this novel method is reproducible and can be used for future studies.

There are a few limitations in our study. Our study was retrospective and both misclassification and residual confounding are possible. Furthermore, we considered the final endoscopy before any eradication therapy as their progression, and this could conceivably impact BE progression and dysplasia rates. Additionally, our study population consisted exclusively of United States veterans which are predominately white and male and thus our observations may not be generalizable to a wider population. Misclassification of USBE was also possible; BE segment length variability among endoscopists can be as high as 2 cm38. This misclassification may partly explain the lack of difference in risk factors and dysplasia rates between USBE and longer segment BE. Most follow-up endoscopies in this study were performed by different endoscopists than the initial EGD, and there may have been measurement variations due to inter-observer variation. However, the variation appears to be mostly minimal between 1–3cm. Lastly, and despite the large sample size and follow up, the frequency of patients who progressed to dysplasia is small and therefore a type 2 error may still explain the lack of significant differences in progression between USBE and longer segment BE. Additionally, our sample size may not be large enough in the context of the small risks being evaluated.

In conclusion, we found that patients of white race and Hispanic ethnicity were at an increased risk of persistent BE after USBE on index endoscopy. We found no significant difference in dysplasia prevalence between patients with USBE and longer segment BE. Consideration to biopsy endoscopic BE <1cm in length and surveillance of established USBE in certain populations may be warranted, but further risk factor stratification is needed to identify which patients with USBE are at risk of dysplastic progression.

Supplementary Material

Supplementary File

Supplementary Figure: Progression and Regression in USBE on index endoscopy cohort

Acknowledgements:

We would like to acknowledge and thank our two gastrointestinal pathologists, Daniel Rosen and Linda K. Green, for their assistance.

Financial Support:

This work is funded in part by National Institutes of Health grant NCI R01 116845, and the Texas Digestive Disease Center NIH DK58338. Dr. El-Serag is also supported by NIDDK K24–04-107. Dr. Nguyen Wenker is supported by the American College of Gastroenterology Junior Faculty Development Award. This research was supported in part with resources at the VA HSR&D Center for Innovations in Quality, Effectiveness and Safety (#CIN 13–413), at the Michael E. DeBakey VA Medical Center, Houston, TX. The opinions expressed reflect those of the authors and not necessarily those of the Department of Veterans Affairs, the US government, or Baylor College of Medicine.

Footnotes

Potential Competing Interests: The authors report no competing interests for this publication.

REFERENCES

  • 1.Thrift AP, Whiteman DC. The incidence of esophageal adenocarcinoma continues to rise: analysis of period and birth cohort effects on recent trends. Annals of Oncology. 2012;23(12):3155–3162. doi: 10.1093/annonc/mds181 [DOI] [PubMed] [Google Scholar]
  • 2.Liu KS, Raza Syed Ahsan, El-Serag HB, Thrift AP. Trends in Esophageal Adenocarcinoma and Esophageal Squamous Cell Carcinoma Incidence in the United States from 1992 to 2019. Cancers. 2022;14(24):6049–6049. doi: 10.3390/cancers14246049 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Haiyu Zhang, Xiaofeng P, Xiangqiong Mo, et al. Incidence and Survival Changes in Patients with Esophageal Adenocarcinoma during 1984–2013. BioMed Research International. 2019;2019:1–11. doi: 10.1155/2019/7431850 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chang DT, Chapman C, Shen J, Su Z, Koong AC. Treatment of Esophageal Cancer Based on Histology. American Journal of Clinical Oncology. 2009;32(4):405–410. doi: 10.1097/coc.0b013e3181917158 [DOI] [PubMed] [Google Scholar]
  • 5.Spechler SJ. Barrett’s esophagus and esophageal adenocarcinoma: pathogenesis, diagnosis, and therapy. Medical Clinics of North America. 2002;86(6):1423–1445. doi: 10.1016/s0025-7125(02)00082-2 [DOI] [PubMed] [Google Scholar]
  • 6.El-Serag HB, Naik AD, Duan Z, et al. Surveillance endoscopy is associated with improved outcomes of oesophageal adenocarcinoma detected in patients with Barrett’s oesophagus. Gut. 2016;65(8):1252–1260. doi: 10.1136/gutjnl-2014-308865 [DOI] [PubMed] [Google Scholar]
  • 7.Shaheen NJ, Falk GW, Iyer PG, et al. Diagnosis and Management of Barrett’s Esophagus: An Updated ACG Guideline. Official journal of the American College of Gastroenterology | ACG. 2022;117(4):559–587. doi: 10.14309/ajg.0000000000001680 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Fitzgerald RC, di Pietro M, Ragunath K, et al. British Society of Gastroenterology guidelines on the diagnosis and management of Barrett’s oesophagus. Gut. 2013;63(1):7–42. doi: 10.1136/gutjnl-2013-305372 [DOI] [PubMed] [Google Scholar]
  • 9.Weusten B, Bisschops R, Coron E, et al. Endoscopic management of Barrett’s esophagus: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy. 2017;49(02):191–198. doi: 10.1055/s-0042-122140 [DOI] [PubMed] [Google Scholar]
  • 10.Clermont M, Falk GW. Clinical Guidelines Update on the Diagnosis and Management of Barrett’s Esophagus. Digestive Diseases and Sciences. 2018;63(8):2122–2128. doi: 10.1007/s10620-018-5070-z [DOI] [PubMed] [Google Scholar]
  • 11.Herrero Lorenza Alvarez, Curvers WL, van Vilsteren, et al. Validation of the Prague C&M classification of Barrett’s esophagus in clinical practice. Endoscopy. 2013;45(11):876–882. doi: 10.1055/s-0033-1344952 [DOI] [PubMed] [Google Scholar]
  • 12.Vahabzadeh B, Seetharam AB, Cook MB, et al. Validation of the Prague C & M criteria for the endoscopic grading of Barrett’s esophagus by gastroenterology trainees: a multicenter study. Gastrointestinal Endoscopy. 2012;75(2):236–241. doi: 10.1016/j.gie.2011.09.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Qumseya B, Sultan S, Bain P, et al. ASGE guideline on screening and surveillance of Barrett’s esophagus. Gastrointestinal Endoscopy. 2019;90(3):335–359.e2. doi: 10.1016/j.gie.2019.05.012 [DOI] [PubMed] [Google Scholar]
  • 14.Muthusamy VR, Wani S, Gyawali CP, et al. AGA Clinical Practice Update on New Technology and Innovation for Surveillance and Screening in Barrett’s Esophagus: Expert Review. Clinical Gastroenterology and Hepatology. 2022;20(12):2696–2706.e1. doi: 10.1016/j.cgh.2022.06.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wani S, Williams JL, Falk GW, Komanduri S, Muthusamy VR, Shaheen NJ. An Analysis of the GIQuIC Nationwide Quality Registry Reveals Unnecessary Surveillance Endoscopies in Patients With Normal and Irregular Z-Lines. American Journal of Gastroenterology. 2020;115(11):1869–1878. doi: 10.14309/ajg.0000000000000960 [DOI] [PubMed] [Google Scholar]
  • 16.Pohl H, Pech O, Arash H, et al. Length of Barrett’s oesophagus and cancer risk: implications from a large sample of patients with early oesophageal adenocarcinoma. Gut. 2015;65(2):196–201. doi: 10.1136/gutjnl-2015-309220 [DOI] [PubMed] [Google Scholar]
  • 17.Pohl H, Welch HG. The Role of Overdiagnosis and Reclassification in the Marked Increase of Esophageal Adenocarcinoma Incidence. JNCI Journal of the National Cancer Institute. 2005;97(2):142–146. doi: 10.1093/jnci/dji024 [DOI] [PubMed] [Google Scholar]
  • 18.Spechler SJ. Screening for Barrett’s oesophagus: are we looking for the right thing? Gut. 2020;70(8):1426–1427. doi: 10.1136/gutjnl-2020-323378 [DOI] [PubMed] [Google Scholar]
  • 19.Jung Kee Wook, Talley NJ, Romero Y, et al. Epidemiology and Natural History of Intestinal Metaplasia of the Gastroesophageal Junction and Barrett’s Esophagus: A Population-Based Study. American Journal of Gastroenterology. 2011;106(8):1447–1455. doi: 10.1038/ajg.2011.130 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Itskoviz D, Levi Z, Boltin D, et al. Risk of Neoplastic Progression Among Patients with an Irregular Z Line on Long-Term Follow-Up. Digestive Diseases and Sciences. 2018;63(6):1513–1517. doi: 10.1007/s10620-018-4910-1 [DOI] [PubMed] [Google Scholar]
  • 21.Horwhat J David, Baroni D, Maydonovitch C, et al. Normalization of Intestinal Metaplasia in the Esophagus and Esophagogastric Junction: Incidence and Clinical Data. The American Journal of Gastroenterology. 2007;102(3):497–506. doi: 10.1111/j.1572-0241.2006.00994.x [DOI] [PubMed] [Google Scholar]
  • 22.Thota PN, Vennalaganti Prashanth, Vennelaganti Sreekar, et al. Low Risk of High-Grade Dysplasia or Esophageal Adenocarcinoma Among Patients With Barrett’s Esophagus Less Than 1 cm (Irregular Z Line) Within 5 Years of Index Endoscopy. Gastroenterology. 2017;152(5):987–992. doi: 10.1053/j.gastro.2016.12.005 [DOI] [PubMed] [Google Scholar]
  • 23.Sharma P, Dent J, Armstrong D, et al. The Development and Validation of an Endoscopic Grading System for Barrett’s Esophagus: The Prague C & M Criteria. Gastroenterology. 2006;131(5):1392–1399. doi: 10.1053/j.gastro.2006.08.032 [DOI] [PubMed] [Google Scholar]
  • 24.Dickman R, Levi Z, Vilkin A, Zvidi Ibrahim, Niv Y. Predictors of specialized intestinal metaplasia in patients with an incidental irregular Z line. European Journal of Gastroenterology & Hepatology. 2010;22(2):135–138. doi: 10.1097/meg.0b013e3283318f69 [DOI] [PubMed] [Google Scholar]
  • 25.Souza RF, Spechler SJ. Mechanisms and pathophysiology of Barrett oesophagus. Nature Reviews Gastroenterology & Hepatology. 2022;19(9):605–620. doi: 10.1038/s41575-022-00622-w [DOI] [PubMed] [Google Scholar]
  • 26.Cameron AJ, Lomboy CT. Barrett’s esophagus: Age, prevalence, and extent of columnar epithelium. Gastroenterology. 1992;103(4):1241–1245. doi: 10.1016/0016-5085(92)91510-b [DOI] [PubMed] [Google Scholar]
  • 27.Öberg S, Johansson J, Wenner J, Walther B. Metaplastic Columnar Mucosa in the Cervical Esophagus After Esophagectomy. Annals of Surgery. 2002;235(3):338–345. doi: 10.1097/00000658-200203000-00005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Abrams JA, Fields S, Lightdale CJ, Neugut AI. Racial and ethnic disparities in the prevalence of Barrett’s esophagus among patients who undergo upper endoscopy. Clinical Gastroenterology and Hepatology: The Official Clinical Practice Journal of the American Gastroenterological Association. 2008;6(1):30–34. doi: 10.1016/j.cgh.2007.10.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Jacobson BC, Barry CT, Liu Z. The Prevalence of Barrett’s Esophagus Among Hispanics Is Lower Than That of Caucasians. Gastrointestinal Endoscopy. 2006;63(5):AB137. doi: 10.1016/j.gie.2006.03.236 [DOI] [Google Scholar]
  • 30.Bersentes Korina, Fass R, Padda S, Johnson C, Sampliner RE. Prevalence of Barrett’s esophagus in Hispanics is similar to Caucasians. Digestive diseases and sciences. 1998;43(5):1038–1041. doi: 10.1023/a:1018834902694 [DOI] [PubMed] [Google Scholar]
  • 31.Keyashian K, Hua V, Narsinh K, Kline M, Chandrasoma PT, Kim JJ. Barrett’s esophagus in Latinos undergoing endoscopy for gastroesophageal reflux disease symptoms. Diseases of the Esophagus. 2012;26(1):44–49. doi: 10.1111/j.1442-2050.2011.01316.x [DOI] [PubMed] [Google Scholar]
  • 32.Barrie J, Yanni F, Sherif M, Dube AK, Tamhankar AP. Length of Barrett’s esophagus in the presence of low-grade dysplasia, high-grade dysplasia, and adenocarcinoma. Surgical Endoscopy. 2020;35(8):4756–4762. doi: 10.1007/s00464-020-07950-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Falk GW. Barrett’s oesophagus: Frequency and prediction of dysplasia and cancer. Best Practice & Research Clinical Gastroenterology. 2015;29(1):125–138. doi: 10.1016/j.bpg.2015.01.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Prasad GA, Bansal A, Sharma P, Wang KK. Predictors of Progression in Barrett’s Esophagus: Current Knowledge and Future Directions. American Journal of Gastroenterology. 2010;105(7):1490–1502. doi: 10.1038/ajg.2010.2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Omar M, Thaker AM, Wani S, et al. Anatomic location of Barrett’s esophagus recurrence after endoscopic eradication therapy: development of a simplified surveillance biopsy strategy. Gastrointestinal endoscopy. 2019;90(3):395–403. doi: 10.1016/j.gie.2019.04.216 [DOI] [PubMed] [Google Scholar]
  • 36.Pohl H, Pech O, Arash H, et al. Length of Barrett’s oesophagus and cancer risk: implications from a large sample of patients with early oesophageal adenocarcinoma. Gut. 2015;65(2):196–201. doi: 10.1136/gutjnl-2015-309220 [DOI] [PubMed] [Google Scholar]
  • 37.Davison JM, Goldblum J, Grewal Udhayvir Singh, et al. Independent Blinded Validation of a Tissue Systems Pathology Test to Predict Progression in Patients With Barrett’s Esophagus. The American journal of gastroenterology. 2020;115(6):843–852. doi: 10.14309/ajg.0000000000000556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Tschanz ER. Do 40% of Patients Resected for Barrett Esophagus With High-Grade Dysplasia Have Unsuspected Adenocarcinoma? Archives of Pathology & Laboratory Medicine. 2005;129(2):177–180. doi: 10.5858/2005-129-177-doprfb [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary File

Supplementary Figure: Progression and Regression in USBE on index endoscopy cohort

NIHMS2060994-supplement-Supplementary_File.pdf (36.9KB, pdf)

RESOURCES