Background and Aims
Surveillance endoscopy is recommended after endoscopic eradication therapy (EET) for Barrett’s esophagus (BE) because of the risk of recurrence. Currently recommended biopsy protocols are based on expert opinion and consist of sampling visible lesions followed by random 4-quadrant biopsy sampling throughout the length of the original BE segment. Despite this protocol, some recurrences are not visibly identified. We aimed to identify the anatomic location and histology of recurrences after successful EET with the goal of developing a more efficient and evidence-based surveillance biopsy protocol.
Methods
We performed an analysis of a large multicenter database of 443 patients who underwent EET and achieved complete eradication of intestinal metaplasia (CE-IM) from 2005 to 2015. The endoscopic location of recurrence relative to the squamocolumnar junction (SCJ), visible recurrence identified during surveillance endoscopy, and time to recurrence after CE-IM were assessed.
Results
Fifty patients with BE recurrence were studied in the final analysis. Seventeen patients (34%) had nonvisible recurrences. In this group, biopsy specimens demonstrating recurrence were taken from within 2 cm of the SCJ in 16 of these 17 patients (94%). Overall, 49 of 50 recurrences (98%) occurred either within 2 cm of the SCJ or at the site of a visible lesion. Late recurrences (>1 year) were more likely to be visible than early (<1 year) recurrences (P = .006).
Conclusions
Recurrence after EET detected by random biopsy sampling is identified predominately in the distal esophagus and occurs earlier than visible recurrences. As such, we suggest a modified biopsy protocol with targeted sampling of visible lesions followed by random biopsy sampling within 2 cm of the SCJ to optimize detection of recurrence after EET. (Clinical trial registration number: NCT02634645.)
Graphical abstract
Graphical Abstract
Abbreviations:
BE (Barrett’s esophagus), CE-IM (complete eradication of intestinal metaplasia), EAC (esophageal adenocarcinoma), EET (endoscopic eradication therapy), GEJ (gastroesophageal junction), HGD (high-grade dysplasia), IM (intestinal metaplasia), RFA (radiofrequency ablation), SCJ (squamocolumnar junction)Endoscopic eradication therapy (EET) is a safe and effective treatment for Barrett’s esophagus (BE)-related dysplasia and intramucosal esophageal adenocarcinoma (EAC).
1
, 2
, 3
, 4
EET typically involves a combination of therapies including EMR to remove any areas of nodularity or visible focal lesions followed by radiofrequency ablation (RFA) to treat the remaining flat BE with the goal of achieving complete eradication of intestinal metaplasia (CE-IM). However, recurrence of IM, dysplasia, and/or EAC remains a challenge. A 2016 meta-analysis by Krishnamoorthi et al5
reported a pooled incidence rate for any recurrence of 9.5% per patient year (95% confidence interval, 6.7-12.3). A 2017 meta-analysis by Fujii-Lau et al6
reported a pooled incidence rate for any recurrence of 7.5 per 100 person-years (95% confidence interval, 6.1-9.0). Among the most notable included studies was a multicenter randomized controlled trial for RFA, which discovered that 35 of 110 patients (32%) who previously achieved CE-IM experienced recurrence during a 5-year follow-up.7
Based on this well-established risk of recurrence after achieving CE-IM, current guidelines recommend that these patients undergo surveillance endoscopies at regular intervals.
1
, 2
, 3
During endoscopy, sampling of any visible lesions followed by systematic, 4-quadrant random biopsy sampling every 1 to 2 cm through the length of the original BE segment is suggested.2
, 3
These random biopsy specimens serve to identify nonvisible recurrence, which may include microscopic foci of BE or “buried” or subsquamous BE. However, the current tissue sampling protocols are based on expert opinion only.2
A rigorous biopsy protocol, of uncertain benefit, also results in additional pathology costs and can be time consuming, potentially leading to decreased compliance.8
The aims of this study were to identify the anatomic location of recurrences in patients who underwent successful EET, to assess for any differences between visible and nonvisible recurrences with respect to anatomic location and timing of recurrence, and to propose a simpler, more efficient random biopsy protocol to optimize identification of recurrent disease.Methods
This was an analysis of a large multicenter cohort, the TREAT-BE Consortium (ie, the Treatment with Resection and Endoscopic Ablation Techniques for Barrett’s Esophagus Consortium), that included 4 tertiary referral centers in the United States aimed at assessing outcomes of patients undergoing EET for BE. Patients were enrolled prospectively from 2013 onward. Patients treated before this time were also identified and contacted for enrollment into the registry and ongoing prospective data collection. Study approval was obtained from each participating center’s institutional review board. The study was registered at clinical trials.gov (NCT02634645).
The analysis for this study included patients who received EET and achieved CE-IM between 2005 and 2015. Inclusion criteria were patients who were ≥18 years of age who underwent EET for BE with monotherapy or combination therapy with EMR and/or ablation, achieved CE-IM, received surveillance at the primary site, and had documented recurrent IM and/or dysplasia after achieving CE-IM. When performed, the treatment zone of ablation typically included the gastroesophageal junction (GEJ) and a short distance into the cardia. Patients without precise documentation of the GEJ and location of recurrence of nonvisible lesions were excluded from the analysis.
Definitions
CE-IM was defined as the absence of endoscopically visible BE plus the absence of IM on all surveillance biopsy samples in the tubular esophagus or at the GEJ on a single endoscopy after EET, which is consistent with the definition of several prior studies.
9
, 10
, 11
, 12
, 13
, 14
Recurrence of IM or neoplasia was defined by the detection of IM or neoplasia on surveillance biopsy specimens in tubular esophagus or at the GEJ after achieving CE-IM. Consistent with prior reports, the discovery of IM or neoplasia from the cardia was considered GEJ recurrence for the purposes of this study.15
Recurrences were further stratified into visible and nonvisible recurrences depending on whether IM or neoplasia was discovered at the site of visible BE or only via random sampling, respectively.16
Study protocol
This prospective database includes the following data that were specifically used for this study: patient demographics, objective data from endoscopy and pathology reports, pretreatment histology, length of pretreatment BE (based on the Prague criteria), number of EET sessions until CE-IM achievement, and EET modalities.
17
These data were used to determine the time between CE-IM and recurrence, histology of recurrence, number of sites of the recurrent BE, location of the recurrence based on distance to the squamocolumnar junction (SCJ), and association of recurrence with a visible esophageal lesion. The surveillance biopsy methodology (every 1 cm, every 2 cm, targeted, or random biopsy sampling) was also documented. Tissue acquisition at the GEJ included sampling of the cardia, whether by sampling across the SCJ or by separately sampling just below the top of the gastric folds. Biopsy sampling across the SCJ or distinctly from the cardia was all considered together as biopsy sampling at the GEJ. Other biopsy samples were labeled and considered separately by their anatomic location in the tubular esophagus by distance proximal to the GEJ.Measured outcomes
The primary outcome of the study was the endoscopic location of recurrence of IM and/or dysplasia relative to the SCJ. The mean recurrence location ratio was calculated by dividing the distance between the recurrence and GEJ by the maximum length of the original BE segment. Secondary outcomes were whether recurrences were visible or nonvisible in nature, the timing of recurrence after CE-IM, and histology of recurrence.
Statistical analysis
Data were analyzed using SPSS statistical software (version 22; IBM, Armonk, NY). Numerical variables were expressed as mean (standard deviation) or as a median (range) when they followed a normal or skewed distribution, respectively. Categorical variables were expressed as numbers and percentages. The independent t test was used to compare means for numerical variables between different groups. The Mann-Whitney U test was used to compare medians between groups for nonparametric numerical variables. The χ2 test was used to compare categorical variables (with the Fisher exact test used when appropriate). Among the secondary analyses, timing of recurrence was assessed as both a continuous variable and a binary variable, using the median time to recurrence to define early versus late recurrence. Kaplan-Meyer curves were used to demonstrate time to recurrence and the log-rank test was used to compare the differences in time to recurrence among different groups.
Results
Study population
The cohort of the 549 patients who underwent EET during the study period was evaluated. Overall, 80.7% successfully achieved CE-IM (85.6% for low-grade dysplasia, 79.1% for high-grade dysplasia [HGD], and 76% for EAC). Of 443 patients who achieved CE-IM after EET between 2005 and 2015, recurrence of BE with or without neoplasia was noted in 61 patients (13.8%) at a median time of 12 months (range, 1-54) after achieving CE-IM. Of these, 13 (21.3%) had recurrence of dysplasia and 7 (11.5%) had recurrence of HGD or EAC. Ten patients (including 2 patients who underwent endoscopic submucosal dissection) did not have explicit documentation of the levels of recurrence in their endoscopy and pathology reports and were excluded. An additional patient who underwent endoscopic submucosal dissection was excluded because of the absence of documentation of the level of the GEJ. Of the 3 patients who underwent endoscopic submucosal dissection, 2 had HGD and 1 had EAC. All 3 recurrences were IM (without dysplasia or cancer) at the site of the previous resection, and all 11 of the excluded recurrences were IM. This resulted in a total of 50 patients who were included in the final analysis to determine an appropriate surveillance biopsy protocol (Fig. 1).
Figure 1Flow diagram to demonstrate patients included in the final study cohort. CE-IM, Complete eradication of intestinal metaplasia; NDBE, nondysplastic Barrett’s esophagus; HGD, high-grade dysplasia; EAC, esophageal adenocarcinoma .
The mean original BE length was 4 cm circumferentially and 6 cm in maximal extent by Prague criteria. Pretreatment histology included nondysplastic BE in 8% and HGD/EAC in 62%. A mean of 3.3 ± 1.8 EET procedures was needed to achieve initial CE-IM. The median time to recurrence was 12 months. Additional pretreatment characteristics of patients who developed recurrence are summarized in Table 1.
Table 1Pretreatment characteristics of patients who developed recurrent Barrett’s esophagus (n = 50)
| Characteristic | Value |
|---|---|
| Mean age, y (D) | 67.7 (11.3) |
| Gender | |
| Male | 40 (80) |
| Female | 10 (20) |
| Ethnicity | |
| White | 45 (90) |
| Hispanic | 1 (2) |
| African American | 1 (2) |
| Asian and Pacific Islander | 1 (2) |
| Unknown | 2 (4) |
| Mean Barrett’s length by Prague criteria, cm (SD) | |
| Circumferential extent | 4 (4) |
| Maximal extent | 6 (4) |
| Pre-EET histology | |
| NDBE | 4 (8) |
| LGD/indeterminate | 15 (30) |
| HGD | 24 (48) |
| EAC | 7 (14) |
| Mean no. of EET procedures to achieve CE-IM (SD) | 3.3 (1.8) |
| Visible lesions before RFA | 27 (54) |
| Type of EET treatment used | |
| RFA only | 20 (40) |
| EMR only | 6 (12) |
| EMR + RFA | 15 (30) |
| Cryotherapy only | 2 (4) |
| EMR + cryotherapy | 2 (4) |
| Other | 4 (8) |
Values are n (%) unless otherwise defined.
SD, Standard deviation; NDBE, nondysplastic Barrett’s esophagus; LGD, low-grade dysplasia; HGD, high-grade dysplasia; EAC, esophageal adenocarcinoma; EET, endoscopic eradication therapy; CE-IM, complete eradication of intestinal metaplasia; RFA, radiofrequency ablation.
∗ Argon plasma coagulation in combination with EMR, RFA, and/or cryotherapy.
Recurrence location and histology
Of the patients with recurrence, 33 of 50 (66%) developed visible recurrence, whereas 17 of 50 (34%) had recurrence on pathology but no visible lesion on biopsy samples (“nonvisible recurrence”). The characteristics and locations of recurrences are presented in Tables 2 and 3. Most recurrences (37/50, 74%) occurred within 2 cm proximal to the SCJ, including 21 of 33 visible recurrences (63.6%) and 16 of 17 nonvisible recurrences (98%). The only nonvisible recurrence >2 cm from the SCJ was IM without dysplasia. Overall, 49 of 50 BE recurrences (98%) were detected on random biopsy samples at the GEJ or ≤2 cm proximal to the GEJ or found at the site of a visible lesion.
Table 2Characteristics of recurrent Barrett’s esophagus
| Characteristic | Value |
|---|---|
| Median time between CE-IM and recurrence, mo (range) | 12 (1-54) |
| Recurrence histology | |
| NDBE | 37 (74) |
| Indeterminate for dysplasia | 1 (2) |
| LGD | 5 (10) |
| HGD | 5 (10) |
| EAC | 2 (4) |
| No. of sites of Recurrence | |
| One | 44 (88) |
| Two | 3 (6) |
| Three | 2 (4) |
| Not documented | 1 (2) |
| Visible recurrences | 33 (66) |
| Nonvisible recurrences | 17 (34) |
| Mean recurrence location divided by the pretreatment BE length, ratio (SD) | .17 (.29) |
| Biopsy protocol used to verify CE-IM | |
| Q1 cm | 8 (16) |
| Q2 cm | 30 (60) |
| Not specifically documented | 12 (24) |
| Biopsy protocol for surveillance | |
| Q1 cm | 16 (32) |
| Q2 cm | 24 (48) |
| Targeting visible lesion only | 8 (16) |
| Not specifically documented | 2 (4) |
| Forceps used in surveillance | |
| Large capacity | 2 (4) |
| Jumbo | 11 (22) |
| Not documented | 37 (74) |
Values are n (%) unless otherwise defined.
CE-IM, Complete eradication of intestinal metaplasia; NDBE, nondysplastic Barrett’s esophagus; LGD, low-grade dysplasia; HGD, high-grade dysplasia; EAC, esophageal adenocarcinoma; BE, Barrett’s esophagus; SD, standard deviation.
Table 3Location of Barrett’s esophagus recurrences within a relative distance to the SCJ, stratified by pretreatment and recurrence histology and visibility
| Distance to GEJ | Any IM, dysplasia, or EAC (n = 50) | Any dysplasia (n = 44) | HGD/EAC (n = 30) | |||
|---|---|---|---|---|---|---|
| Nonvisible recurrences (n = 17) | Visible recurrences (n =33) | Nonvisible recurrences (n = 16) | Visible recurrences (n = 28) | Nonvisible recurrences (n = 13) | Visible recurrences (n = 17) | |
| Pretreatment histology | ||||||
| At GEJ | 12 (70.6) | 16 (48.5) | 13 (81.3) | 17 (60.7) | 10 (76.9) | 10 (58.8) |
| ≤1 cm proximal to GEJ | 13 (76.5) | 19 (57.6) | 14 (87.5) | 21 (75) | 11 (84.6) | 13 (76.5) |
| ≤2 cm proximal to GEJ | 16 (94.1) | 21 (63.6) | 16 (100) | 23 (82.1) | 13 (100) | 15 (88.2) |
| ≤3 cm proximal to GEJ | 17 (100) | 23 (69.7) | 16 (100) | 26 (92.9) | 13 (100) | 15 (88.2) |
| >3 cm from GEJ | — | 4 (12.1) | — | 2 (7.1) | — | 2 (11.8) |
| Not documented | — | 6 (18.2) | — | 2 (7.1) | — | 2 (11.8) |
| Distance to GEJ | Any IM, dysplasia, or EAC (n = 50) | Any dysplasia (n = 13) | HGD/EAC (n = 7) | |||
| Nonvisible recurrences (n = 17) | Visible recurrences (n = 33) | Nonvisible recurrences (n = 6) | Visible recurrences (n = 7) | Nonvisible recurrences (n = 4) | Visible recurrences (n = 3) | |
| Recurrence histology | ||||||
| At GEJ | 12 (70.6) | 16 (48.5) | 4 (66.7) | 4 (57.1) | 2 (50) | 1 (33.3) |
| ≤1 cm proximal to GEJ | 13 (76.5) | 19 (57.6) | 5 (83.3) | 6 (85.7) | 3 (75) | 3 (100) |
| ≤2 cm proximal to GEJ | 16 (94.1) | 21 (63.6) | 6 (100) | 6 (85.7) | 4 (100) | 3 (100) |
| ≤3 cm proximal to GEJ | 17 (100) | 23 (69.7) | 6 (100) | 6 (85.7) | 4 (100) | 3 (100) |
| >3 cm from GEJ | — | 4 (12.1) | — | — | — | — |
| Not documented | — | 6 (18.2) | — | 1 (14.3) | — | — |
Values are n (%).
SCJ, Squamocolumnar junction; IM, intestinal metaplasia; HGD, high-grade dysplasia; EAC, esophageal adenocarcinoma; GEJ, gastroesophageal junction; —, none.
∗ Including cardia.
† Patients without documented nonvisible recurrence levels or GEJ levels were excluded from analysis.
Of the 21 patients with visible recurrences within 2 cm of the GEJ, short-segment pretreatment BE (≤3 cm) was present in 6 (28.6%). Of the 16 patients with nonvisible recurrences within 2 cm of the GEJ, it was present in 4 (25%). The mean ratio of BE recurrence location (distance proximal to GEJ) divided by the maximum pretreatment BE length was .17 ± .29 (95% confidence interval, .08-.26; P > .001), suggesting that recurrences typically occurred in the lower ∼20% of the treated segment of BE.
The histology of the recurrences was most commonly nondysplastic BE (74%, n = 37) followed by low-grade dysplasia (10%, n = 5), HGD (10%, n = 5), EAC (4%, n = 2), and indeterminate for dysplasia (2%, n = 1). Subsquamous buried BE was not reported in any of the pathology reports. Recurrent BE occurred at only a single location in 44 of 50 patients (88%). Consistent with prior reports, the histology of recurrent BE was the same or less advanced than the pretreatment histology in all patients (Fig. 2).
Figure 2Association between pretreatment and recurrent histology. BE, Barrett’s esophagus; NDBE, nondysplastic Barrett’s esophagus, LGD, low-grade dysplasia; HGD, high-grade dysplasia; EAC, esophageal adenocarcinoma.
Of the 13 patients with dysplasia or EAC as recurrence histology, 7 had precise documentation of the anatomic landmarks and the location of the pre-EET lesion. Three of the 7 patients had recurrence at the same location as the prior dysplasia/EAC, including 2 lesions that were at the GEJ both before and after EET and 1 lesion that was at 1 cm proximal to the GEJ both before and after EET. The pre-EET lesion of the remaining 4 lesions were ≥2 cm proximal to the GEJ, whereas the post-EET recurrence all occurred within 1 cm proximal to the GEJ (including 2 at the GEJ).
Timing of recurrence
Patients who developed recurrence in the first year were compared with patients who developed recurrence after the first year (Table 4). Recurrent BE tended to occur closer to the SCJ in the first year (P = .024). Recurrence was more likely to be at the site of a visible lesion in recurrences >1 year when compared with within the first year (P = .006). Conversely, visible recurrences tended to occur later than nonvisible (P < .001) on Kaplan-Meyer analysis (Fig. 3). All nonvisible recurrences were detected within 2 years after achieving CE-IM.
Table 4Comparison of early and late recurrences
| Recurrence ≤1 year (n = 30) | Recurrence >1 year (n = 20) | P value | |
|---|---|---|---|
| No. of EET sessions to CE-IM | 3.16 ± 1.75 | 3.42 ± 1.92 | .626 |
| Pretreatment BE length, cm | 5.98 ± 3.63 | 7.23 ± 3.88 | .325 |
| Location of the most proximal nonvisible recurrences relative to the SCJ, cm | .37 ± .9 | 2 ± 0 | .024 |
| Recurrence location divided by the pretreatment BE length | .16 ± .29 | .19 ± .28 | .701 |
| Pre-EET histology, % | .023 | ||
| NDBE | 6.5 (2) | 10.5 (2) | |
| LGD/indeterminate | 16.1 (5) | 52.6 (10) | |
| HGD | 61.3 (19) | 26.3 (5) | |
| EAC | 16.1 (5) | 10.5 (2) | |
| Recurrence histology, % | .285 | ||
| NDBE | 71 (22) | 78.9 (15) | |
| LGD/indeterminate | 9.7 (3) | 15.8 (3) | |
| HGD | 16.1 (5) | 0 (0) | |
| EAC | 3.2 (1) | 5.3 (1) | |
| Recurrence at the site of a visible lesion, % | .006 | ||
| Yes | 51.6 (16) | 89.5 (17) | |
| No | 48.4 (15) | 10.5 (2) |
Values in parentheses are number of cases.
EET, Endoscopic eradication therapy; CE-IM, complete eradication of intestinal metaplasia; BE, Barrett’s esophagus; SCJ, squamocolumnar junction; NDBE, nondysplastic Barrett’s esophagus; LGD, low-grade dysplasia; HGD, high-grade dysplasia; EAC, esophageal adenocarcinoma.
Figure 3Timing of the recurrences in subjects with visible versus nonvisible recurrences.
We also aimed to determine if changing the definition of CE-IM from 1 negative to 2 negative endoscopies impacted our results. In total, 4 patients were found to have recurrent IM or dysplasia on the first endoscopy after CE-IM, all within about 6 months after last EET and within about 3 months of first clearance endoscopy. All 4 patients had HGD before EET. Two of these 4 had nonvisible IM of the cardia and 1 had HGD .5 cm above the SCJ at a site distinct from the original location of HGD. The fourth patient had visible IM detected at the top of the original BE segment, at 5 cm above the SCJ. Thus, most of our recurrences would not have been affected by this change in definition, and the primary outcome and recommendations regarding anatomic location were unchanged.
Discussion
This multicenter study of 443 patients who achieved CE-IM after EET discovered a 13.8% overall recurrence rate, which is consistent with values reported by prior studies.
2
, 3
, 5
, 6
As also previously noted, most recurrences occurred within the first year of surveillance and at an earlier histologic stage than in the original BE segment (eg, patients with BE and HGD who underwent EET usually had recurrence with IM or low-grade dysplasia).6
These findings support the need for post-EET surveillance.However, current guidelines with regards to surveillance intervals and biopsy protocols are based on expert opinion.
2
, 18
The current practice of random 4-quadrant biopsy sampling every 1 to 2 cm through the original BE segment can be time consuming and can result in a large number of pathology specimens. In our cohort, nearly all random biopsy specimens that contained nonvisible recurrence were within 2 cm of the SCJ. There have also been reports of recurrence occurring at the GEJ and/or proximal cardia after EET.19
These findings suggest that current practices may be unnecessarily excessive, particularly in more proximal segments of the treated regions, and that more focused sampling around the SCJ is warranted.The rationale behind random sampling is theoretically to evaluate for nonvisible mucosal foci of recurrences or to evaluate for subsquamous IM or neoplasia (“buried Barrett’s”). All recurrences in our cohort were within the mucosal layer with no reported instances of buried BE. Furthermore, we discovered that nonvisible recurrences were more likely to occur early (<1 year), whereas late recurrences were more likely to be endoscopically visible. Although we acknowledge that the visible lesions may have previously been missed despite meticulous inspection by experienced endoscopists, we postulate that nonvisible recurrences represent minute foci of early IM or neoplasia that will eventually become endoscopically visible over time for targeted biopsy sampling.
7
, 9
Using this information, we applied a hypothetical simplified random biopsy protocol to our data (Fig. 4). This protocol would begin with a meticulous visual examination using high-definition white light and advanced imaging (eg, narrow-band imaging) to exclude or biopsy any visible lesions before random biopsies. Random sampling would then be performed in 4 quadrants at the GEJ plus at 2 additional levels within 2 cm of the SCJ. Such a protocol would have resulted in a substantial 57% reduction in the minimum number of required biopsy samples while detecting 98% of recurrences when applied to our cohort. Reducing the extent of random biopsy specimens may also potentially reduce procedure times and costs associated with pathology. A recent publication by Cotton et al
18
that aimed to develop evidence-based surveillance intervals after RFA for BE proposed a reduced frequency of endoscopies while still detecting most premalignant and malignant lesions. A prior single-center study by this group also reported the anatomic concentration of recurrences within the distal esophagus, data that are now corroborated by our larger multicenter cohort.10
Our study further characterizes the temporal association between nonvisible and visible findings and suggests that a simplified biopsy protocol can potentially be applied in conjunction with these newly proposed decreased intensity surveillance intervals. Visible recurrences may represent microscopic foci of BE that were missed by random sampling and become visible over time to allow for subsequent targeted rather than random biopsy sampling when discovered.
Figure 4Simplified biopsy protocol applied to our recurrence cohort, which originally consisted of patients who averaged C4M6 BE and achieved complete eradication of intestinal metaplasia. GEJ, Gastroesophageal junction; BE, Barrett’s esophagus; SCJ, squamocolumnar junction.
Our data also found that patients with early recurrence (<1 year) had more advanced pretreatment histology than those with recurrence >1 year from CE-IM (77% HGD/EAC vs 37% HGD/EAC, respectively). This finding supports a similar observation from a study of the US RFA database.
16
One potential explanation for this finding is that advanced pretreatment histology represents more-aggressive disease that confers a higher risk of recurrence. Alternatively, what is defined as early recurrence may actually represent obscure residual disease missed by random biopsy sampling because of sampling error. The latter hypothesis would suggest that improved methods to truly ensure complete eradication of BE-related neoplasia are required.Our study is subject to several limitations. There was variation in EET techniques and surveillance protocols. In particular, there was variation in biopsy protocols (including biopsy specimen spacing, types of forceps) to assess CE-IM and recurrence. However, these data reflect real-world practice by endoscopists with significant experience in BE treatment. This issue also highlights the fact that surveillance recommendations for post-EET BE management have evolved, and variations exist even among expert endoscopists.
9
, 11
, 14
, 15
, 16
, 20
, 21
There was also variation in documentation of biopsy sampling locations, which resulted in the exclusion of a substantial proportion of the total recurrences because of missing data and a smaller sample size of 50 recurrences. Future studies will require consistent sampling protocols, specific reporting, and a larger sample size to verify the findings of this study. Last, random biopsy protocols likely have limited sensitivity to truly exclude all cellular recurrence because of sampling error, regardless of the intensity of the protocol.22
The most effective surveillance biopsy protocol is therefore a topic of substantial uncertainty.Future work should focus on additional factors that may influence the rates of identified recurrence during post-EET surveillance. For example, the definition of CE-IM (1 negative endoscopy vs 2) does not have universal consensus and may impact the total number of recurrences found. However, as also seen in prior reports, in the present study our choice of definition did not impact the study’s overall conclusions.
5
, 6
, 16
, 23
It may therefore be time to develop consensus in the field and discard the use of multiple endoscopies to define CE-IM, acknowledging the possibility of sampling error with fewer endoscopies. In addition, the ideal random sampling protocol (every 1 or 2 cm) is uncertain and has not been validated specifically for the purposes of post-EET surveillance.24
In particular, there is a need to determine if additional focused sampling of prior areas of neoplasia is required (particularly if previously >2 cm above the SCJ). Next, there are mixed data regarding the importance of forceps size.25
, 26
, 27
Finally, it is unclear whether IM of the cardia should be considered recurrence and whether biopsy specimens should be obtained from the proximal cardia separately (to differentiate this from IM of the tubular esophagus) or whether biopsy specimens straddling the SCJ are suitable. Each of these practices has been previously reported and may account for some of the variation in practice.15
, 28
, 29
, 30
Clarifying each of these issues in conjunction with potentially incorporating adjunct imaging may allow more accurate identification of true CE-IM and recurrence and reduce substantial variability currently seen in studies of post-EET recurrence.In conclusion, nonvisible recurrences after EET appear to be concentrated within the distal 2 cm of the esophagus, tend to occur earlier than visible recurrences, and appear to be mucosal in depth. These findings suggest that random biopsy sampling of the entire neosquamous zone may be unnecessary. A simplified surveillance protocol limited to obtaining biopsy specimens of any visible lesions along with random biopsy sampling of the GEJ and in 1-cm intervals within the distal 2 cm of the esophagus would appear to capture all recurrences. Such an approach can potentially reduce procedure times and associated pathology costs. These data could be used in conjunction with newly proposed evidence-based surveillance intervals to significantly reduce both the frequency of endoscopies and the number of random biopsy specimens obtained from patients who have achieved CE-IM.
References
- Diagnosis and management of low-grade dysplasia in Barrett’s esophagus: expert review from the Clinical Practice Updates Committee of the American Gastroenterological Association.Gastroenterology. 2016; 151: 822-835
- Endoscopic eradication therapy for patients with Barrett's esophagus-associated dysplasia and intramucosal cancer.Gastrointest Endosc. 2018; 87: 907-931
- ACG clinical guideline: diagnosis and management of Barrett's esophagus.Am J Gastroenterol. 2016; 111: 30-50
- Efficacy and safety outcomes of multimodal endoscopic eradication therapy in Barrett's esophagus-related neoplasia: a systematic review and pooled analysis.Gastrointest Endosc. 2017; 85: 482-495
- Risk of recurrence of Barrett's esophagus after successful endoscopic therapy.Gastrointest Endosc. 2016; 83: 1090-1106.e3
- Recurrence of intestinal metaplasia and early neoplasia after endoscopic eradication therapy for Barrett’s esophagus: a systematic review and meta-analysis.Endosc Int Open. 2017; 5: 430-449
- Late recurrence of Barrett’s esophagus after complete eradication of intestinal metaplasia is rare: final report from Ablation in Intestinal Metaplasia Containing Dysplasia Trial.Gastroenterology. 2017; 153: 681-688
- Adherence to biopsy guidelines for Barrett’s esophagus surveillance in the community setting in the United States.Clin Gastroenterol Hepatol. 2009; 7 (736-10)
- Durability of radiofrequency ablation in Barrett's esophagus with dysplasia.Gastroenterology. 2011; 141: 460-468
- Recurrent intestinal metaplasia after radiofrequency ablation for Barrett's esophagus: endoscopic findings and anatomic location.Gastrointest Endosc. 2015; 81: 1362-1369
- Intestinal metaplasia recurs infrequently in patients successfully treated for Barrett's esophagus with radiofrequency ablation.Am J Gastroenterol. 2013; 108: 187-196
- Comparative risk of recurrence of dysplasia and carcinoma after endoluminal eradication therapy of high-grade dysplasia versus intramucosal carcinoma in Barrett’s esophagus.Gastrointest Endosc. 2015; 81: 1158-1166
- Detection of intestinal metaplasia after successful eradication of Barrett's Esophagus with radiofrequency ablation.Dig Dis Sci. 2011; 56: 1996-2000
- Stepwise radical endoscopic resection versus radiofrequency ablation for Barrett's oesophagus with high-grade dysplasia or early cancer: a multicentre randomised trial.Gut. 2011; 60: 765-773
- Recurrence of esophageal intestinal metaplasia after endoscopic mucosal resection and radiofrequency ablation of Barrett's esophagus: results from a US Multicenter Consortium.Gastroenterology. 2013; 145: 79-86
- Durability and predictors of successful radiofrequency ablation for Barrett's esophagus.Clin Gastroenterol Hepatol. 2014; 12: 1840-1847
- The development and validation of an endoscopic grading system for Barrett's esophagus: the Prague C & M criteria.Gastroenterology. 2006; 131: 1392-1399
- Development of evidence based surveillance intervals following radiofrequency ablation of Barrett’s esophagus.Gastroenterology. 2018; 155: 316-326.e6
- Clinical outcomes following recurrence of intestinal metaplasia after successful treatment of Barrett's esophagus with radiofrequency ablation.Am J Gastroenterol. 2017; 112: 87-94
- Endoscopic radiofrequency ablation for Barrett's esophagus: 5-year outcomes from a prospective multicenter trial.Endoscopy. 2010; 42: 781-789
- Multimodality endoscopic eradication for neoplastic Barrett oesophagus: results of a European multicentre study (EURO-II).Gut. 2016; 65: 555-562
- The Seattle protocol does not more reliably predict the detection of cancer at the time of esophagectomy than a less intensive surveillance protocol.Clin Gastroenterol Hepatol. 2009; 7: 653-658
- Higher rate of Barrett's detection in the first year after successful endoscopic therapy: meta-analysis.Am J Gastroenterol. 2018; 113: 959-971
- Optimizing endoscopic biopsy detection of early cancers in Barrett's high-grade dysplasia.Am J Gastroenterol. 2000; 95: 3089-3096
- The role of endoscopy in Barrett's esophagus and other premalignant conditions of the esophagus.Gastrointest Endosc. 2012; 76: 1087-1094
- Randomized comparison of 3 different-sized biopsy forceps for quality of sampling in Barrett's esophagus.Gastrointest Endosc. 2010; 72: 935-940
- Use of a new jumbo forceps improves tissue acquisition of Barrett's esophagus surveillance biopsies.Gastrointest Endosc. 2009; 70: 1072-1078
- Can you stop surveillance after radiofrequency ablation of Barrett's esophagus? A glass half full.Gastroenterology. 2013; 145: 39-42
- Radiofrequency ablation and endoscopic mucosal resection for dysplastic Barrett’s esophagus and early esophageal adenocarcinoma: outcomes of the UK National Halo RFA Registry.Gastroenterology. 2013; 145: 87-95
- Remission of Barrett's esophagus with early neoplasia 5 years after radiofrequency ablation with endoscopic resection: a Netherlands cohort study.Gastroenterology. 2013; 145: 96-104
Article Info
Publication History
Published online: April 17, 2019
Accepted:
April 2,
2019
Received:
October 4,
2018
Footnotes
DISCLOSURE: The following author received research support for this study from AGA-Takeda Research Scholar Award in GERD and Barrett’s esophagus and University of Colorado Department of Medicine Outstanding Early Scholars Program: S. Wani. In addition, the following authors disclosed financial relationships relevant to this publication: A.M. Thaker, D. Mullady: Consultants for Boston Scientific. S. Wani, B.C. Brauer, V. Kushnir, S. Komanduri, and V. Raman Muthusamy: Consultants for Boston Scientific and Medtronic. S. Edmundowicz: Endostim stock options, Medtronic research support and consultant, and Olympus advisory board and consultant. B.C. Brauer: Research grant support from ERBE. All other authors disclosed no financial relationships relevant to this publication.
See CME section, p. 506.
Identification
Copyright
© 2019 by the American Society for Gastrointestinal Endoscopy
ScienceDirect
Access this article on ScienceDirectLinked Article
- Random biopsy of the neosquamous epithelium is a waste of time and moneyGastrointestinal EndoscopyVol. 90Issue 3
31599-8&id=fx1.jpg){kind=link}
31599-8&id=gr1.jpg){kind=link}
31599-8&id=gr2.jpg){kind=link}
31599-8&id=gr3.jpg){kind=link}
31599-8&id=gr4.jpg){kind=link}