BARRETT’S ESOPHAGUS 1052–3359/02 $15.00
.

00

APPEARANCE AND
PROGNOSIS OF DYSPLASIA IN
BARRETT’S ESOPHAGUS
Richard E. Sampliner, MD

Barrett’s esophagus (BE) is a change in the lining of the distal

esophagus from normal squamous epithelium to intestinal metaplasia.
The inclusion of intestinal metaplasia as part of the definition of BE over
the past 2 decades is the result of the recognition of its malignant
potential.3 The risk for malignancy has been highlighted by the continu-
ing rise in the incidence of adenocarcinoma of the esophagus in North
America and Western Europe. 2, 6
Dysplasia is the first step in the neoplastic process. As such, its
appearance and progression assume a central role in the clinical manage-
ment of BE. To date, dysplasia remains the most widely used and best-
validated marker of cancer risk in patients who have BE. Documentation
of dysplasia is the key to surveillance endoscopy and biopsy. The grad-
ing of dysplasia determines the intervals of surveillance endoscopy and
the threshold for therapeutic intervention in clinical practice. This article
updates the risk for dysplasia progressing to adenocarcinoma of the
esophagus and the recommendations for surveillance and therapy of
differing grades of dysplasia in patients who have BE.

DATA

Dysplasia is a cytologic and architectural change in the intestinal

metaplasia as explained in Riddell’s review. The database of the natural

From the Southern Arizona VA Health Care System, University of Arizona, Tucson, Ari-
zona

CHEST SURGERY CLINICS OF NORTH AMERICA

VOLUME 12 • NUMBER 1 • FEBRUARY 2002 69

70 SAMPLINER

Table 1. DYSPLASIA GRADE AND PROGRESSION TO ESOPHAGEAL

ADENOCARCINOMA
Dysplasia n Cancer Percent
Negative 382 9 2
Low grade 115 7 6
High grade 95 40 42

history of dysplasia is limited in the number of patients and the general-

izability of the information. A total of 592 patients can be gleaned from
five published series of patients who have BE with defined absence
of dysplasia or grade of dysplasia followed for the development of
adenocarcinoma over a mean of 2.9 to 4.2 years (Table 1).15, 18, 25, 28, 42 The
risk for progression to cancer increases with the grade of dysplasia.
As estimates of the incidence of cancer may be overstated because of
publication bias,34 so may be the estimate of progression of dysplasia.
These estimates remain the best basis for surveillance endoscopy in
patients who have BE until the future performance of randomized trials
of differing surveillance intervals (Table 2).
The limitations of the information on the natural history of dysplasia
include differing definitions of BE, referral bias, small numbers of pa-
tients, short duration of follow-up, differing definitions of prevalent
cancers, and differing intensity of the biopsy protocol and the interob-
server variability of the reading of dysplasia.
The earlier series defined BE as esophageal columnar epithelium of
3 to 5 cm in length.15, 28 In the remainder of the series, all patients who
had esophageal columnar mucosa had intestinal metaplasia. These series
also include patients who had short-segment BE ranging from 22% (⬍3
cm)25 to 47% (⬍2cm)42 of the patients. Patients who have short-segment
BE may well have a lower risk for developing cancer, but this has not
been well defined.29 Clearly, patients who have short-segment BE can
develop dysplasia10, 40 and adenocarcinoma.33
Potential referral bias is demonstrated in comparing two of the
series. The largest series has 327 patients.25 At baseline, 23% had high-
grade dysplasia (HGD), and a total of 42 patients developed cancer over
a mean follow-up of 3.9 years. The relative risk of patients who had
HGD developing cancer was 28 times that of patients who had negative,
indefinite, or low-grade dysplasia (LGD). In contrast, in a registry of 136
patients, only 0.7% had HGD at baseline and only 2 patients developed

Table 2. SURVEILLANCE INTERVAL BY GRADE OF DYSPLASIA

Dysplasia Interval
Negative at two endoscopies 3 years
Low grade—confirmed by second endoscopy 1 year
High grade Intervention
 APPEARANCE AND PROGNOSIS OF DYSPLASIA IN BARRETT’S ESOPHAGUS 71

cancer over a mean follow-up of 4.2 years.18 A series of patients enriched

with HGD will yield many more cancers, some of them prevalent cancers
missed at initial endoscopy.
In one series, all cancers in the BE cohort are identified in the
Kaplan–Meier plot. Of the cancers identified, 22% occurred within 1
year of follow-up.25 Such cancers usually are defined as prevalent and
would have been excluded by an analysis of cancer incidence. This
inclusion inflates the risk for cancer progression in this series.
The most intensive biopsy protocol described is that of the Univer-
sity of Washington.13 A large-channel endoscope and a large-capacity
biopsy forceps are used. For HGD, four-quadrant biopsies are taken at
1-cm intervals. In addition, visible mucosal abnormalities are biopsied.
This testing represents self-admitted research endoscopy that takes an
extended time commitment (as long as 90 minutes) to obtain the system-
atic biopsies (a maximum of 120 per procedure) and routinely is not
performed in the clinical setting. Two surveys of endoscopists in the
United States documented that only 17% used large-capacity biopsy
forceps.7, 9 Additionally, with this optimized biopsy protocol, only 39%
of patients who had an endoscopic biopsy diagnosis of cancer had
cancer in the esophagectomy specimen.26 This finding may be explained
by biopsy removal of small cancers, sampling limitations of the surgical
specimen, or observer variation. The latter cause, however, is unlikely
because 38% of cancers diagnosed by a single endoscopic biopsy were
detected in the surgical specimen even when the same pathologist inter-
preted the endoscopic and the surgical samples. In cancers detected in
one endoscopic biopsy, an average of 17.6 biopsies were taken per 1-cm
interval. In the esophagectomy specimens, an average of 25 blocks of
the entire specimen were evaluated, which is not as intensive as the
research endoscopy.
The interobserver variability of dysplasia is well documented.
Agreement in separating HGD and intramucosal cancer from all other
categories is good at 85% to 87%.27 This separation is critical in clinical
decision making. In a three-tiered comparison of negative versus indif-
ferent and LGD versus HGD and intramucosal cancer, however, the
agreement is only 58% to 61%. The variability of the reading of LGD is
emphasized by a series documenting that when two expert gastrointesti-
nal (GI) pathologists agree on LGD, there is a significant association with
progression to HGD or cancer that was not present with an individual
pathologist’s diagnosis of dysplasia.36 Additionally, the transient positive
diagnosis of dysplasia—a patient with at least one reading of dysplasia
but negative at the last study endoscopy over a 2-year follow-up—is a
major contributor to the endoscopy workload and costs of surveillance
endoscopy.19

SURVEILLANCE RECOMMENDATIONS
The first issue to address is whether an individual patient is a
candidate for surveillance. Is the patient able and willing to undergo
72 SAMPLINER

repeat endoscopy? Does the patient have major comorbidity that would
preclude serial endoscopy or an intervention necessitated by the devel-
opment of HGD or cancer?
Ideally, patients should have their reflux symptoms controlled be-
fore surveillance. Control of symptoms with proton pump inhibitor
therapy maximizes the likelihood of a healed mucosa.4 A healed mucosa
prevents the presence of erosive esophagitis from obscuring the recogni-
tion of BE by erosive esophagitis41 and permits the clinician to measure
the extent of BE more accurately and to obtain a reading of dysplasia
that is less likely to be affected by inflammation. The methodology of
surveillance is a topic informed by habit and few data. The pragmatic
practice limitations of time may take priority over the protocol recom-
mendation based on research endoscopy optimizing the detection of
early cancer. The usual intensity of biopsy is four quadrant every 2 cm.
Although large-capacity forceps are recommended, in practice surveys
they are used only by 17% of endoscopists. By use of large-capacity
biopsy forceps and a more intensive biopsy protocol, one study, in which
a historical control group was used, failed to demonstrate an increased
yield of cancer in patients who had HGD.8
Systematic biopsy of new squamous epithelium in patients treated
with proton pump inhibitor therapy may be necessary. Many patients
who have squamous islands have underlying intestinal metaplasia35 and,
rarely, cancer.32 Further evaluation of the possibility of biopsying new
squamous islands and squamous extensions is necessary.
For a patient negative for dysplasia in systematic biopsies from two
endoscopies, a 3-year surveillance interval should be sufficient. If LGD
is found, a repeat endoscopy should be performed, increasing the inten-
sity of biopsy at the level with dysplasia because LGD can be adjacent to
frank cancer. Once LGD is confirmed, annual surveillance is appropriate.
As with patients who have LGD, rebiopsy is necessary in patients
who have HGD. Perhaps in this setting the intensity of the University of
Washington research biopsy protocol is appropriate. Once intramucosal
cancer is present, there is the potential for lymph-node metastases, and
surgical resection clearly is indicated. If HGD is found again, then
confirmation of this reading by an expert GI pathologist is essential
before any therapeutic intervention is initiated.
The management of a patient who has HGD is perhaps the greatest
dilemma in BE. Because the average age of recognition of BE is 60
years, many patients are elderly and have comorbid cardiopulmonary
conditions that increase the risk of surgery. Although expert centers
have an operative mortality less than 5% in the fit patient, low-volume
centers, at which most procedures are performed, have an unacceptably
high mortality.1
Patient morbidity and high operative mortality have led to endo-
scopic approaches to the treatment of HGD. Preliminary 6-month results
of a large randomized trial of photodynamic therapy are promising in
documenting downgrading of HGD and in decreasing the progression
to cancer.20 Five-year survival data are necessary to demonstrate the
 APPEARANCE AND PROGNOSIS OF DYSPLASIA IN BARRETT’S ESOPHAGUS 73

successful outcome of endoscopic therapy and to match the outcome of

esophagectomy in patients who have early stage cancer.11, 16, 30, 37, 38
The contemporary 5-year survival of patients who have resected
esophageal adenocarcinoma limited to the mucosa and submucosa ex-
ceeds 80%. Endoscopy therapy must be judged against this benchmark.
Downgrading of HGD is not a sufficient endpoint. As is well docu-
mented in three cases, patients who have residual intestinal metaplasia
can continue to progress to HGD. Because genetic mutations persist in
the residual intestinal metaplasia, the subsequent progression is to be
expected.

ENHANCEMENT OF SURVEILLANCE

Several current and developing techniques offer the opportunity to

improve the accuracy of surveillance strategies. Endoscopic ultrasonog-
raphy provides better local staging of adenocarcinoma of the esophagus
than does computerized tomography. To date, however, even with the
use of 20-mHz probes, the detection of cancer in the absence of visible
mucosal defects has not been improved.
Endoscopic mucosal resection (EMR) offers the opportunity to im-
prove the histologic accuracy of areas of ulcerated, polypoid, or nodular
mucosa by providing larger tissue samples safely. Of 25 patients, 44%
had a significant change in diagnosis in the spectrum of metaplasia to
cancer with EMR.17 This offers the possibility of better timing of the
intervention strategy.
One endoscopist has been able to document an increased yield of
dysplasia or cancer by using methylene blue-directed biopsies compared
with a random four-quadrant biopsy technique (44% versus 28%).3 Al-
though fewer biopsies are required with the methylene blue-detected
techniques, the duration of the procedure is prolonged, and similar
results have not been reproduced by other investigators.
Various optical techniques that can detect dysplasia in real time are
in varying stages of development. These techniques include photody-
namic diagnosis,14 optical coherence tomography,22 laser-induced fluo-
rescence,21 Raman spectroscopy, and fluorescent spectroscopy.39 They
offer the potential to focus biopsies on areas of dysplasia, to improve
the accuracy and efficiency of surveillance biopsy, and even the potential
to bypass histology. Further, technical developments and validation are
necessary before clinical application.
The search for biologic markers that can enhance risk stratification
of patients who have BE has been an ever-expanding field. The under-
standing of the molecular pathogenesis of adenocarcinoma in patients
who have BE continues to progress and offers the possibility of identi-
fying genetic mutations that can be used in combination with the grad-
ing of dysplasia to improve the identification of high-risk patients for
cancer. Surveillance and potentially endoscopic interventions then could
be focused on these patients.
74 SAMPLINER

Mutations in the tumor suppressor gene p53 are found with increas-
ing frequency in the metaplasia, dysplasia, and adenocarcinoma se-
quence.24, 44 In one retrospective study,44 only patients who had LGD and
mutated p53 progressed to HGD or cancer.43 This observation needs to
be documented in prospective series. In a subgroup of patients who had
p53 mutation, a field effect enables easier detection of the mutation.23
Antibody to p53 mutation has even been detected in patients who had
BE before they developed cancer.5 In the future, enhanced predictive
value may be achieved with the detection of antibody to genetic muta-
tions. Of patients who have adenocarcinoma in BE, 25% had hyper-
methylated adenomatosis polyposis coli DNA in the plasma.12 These
preliminary results offer the exciting potential of serum molecular prog-
nostication.
In a large series, the finding of aneuploidy and increased 4N frac-
tions by flow cytometry has been documented to increase the predictive
value when added to histology. Patients who have negative, indefinite,
or LGD and who lacked flow-cytometric abnormality had a 5-year
cumulative cancer incidence of 0 compared with 20% in those who had
a cytometric abnormality.25
The surveillance recommendations herein proposed are based on
limited data, will be refined, and certainly will be at longer time intervals
when better risk stratification of patients who have BE is accomplished.
Cohort studies of patients who have BE have provided preliminary
evidence supporting potential risk factors, including age, gender, grade
of dysplasia, length of BE, size of hiatal hernia, flow-cytometry abnor-
malities, and p53 mutations. Prospective multicenter studies will be
needed to validate these factors and to define better the specific criteria
to risk stratify patients who have BE.

References

1. Begg C, Cramer L, Hoskins W, et al: Impact of hospital volume on operative mortality

for major cancer surgery. JAMA 280:1747–1751, 1998
2. Bytzer P, Christensen P, Damkier P, et al: Adenocarcinoma of the esophagus and
Barrett’s esophagus: A population based study. Am J Gastroenterol 94:86–91,1999
3. Canto M, Setrakian S, Willis J, et al: Methylene blue directed biopsies improve detection
of intestinal metaplasia and dysplasia in Barrett’s esophagus. Gastrointest Endosc
51:560–568, 2000
4. Carlsson R, Galmiche J, Dent J, et al: Prognostic factors influencing relapse of oesopha-
gitis during maintenance therapy with antisecretory drugs: A meta-analysis of long
term omeprazole trials. Aliment Pharmacol Ther 11:473–482, 1997
5. Cawley H, Meltzer S, DeBenedetti V, et al: Anti-p53 antibodies in patients with Barrett’s
esophagus or esophageal carcinoma can predate cancer diagnosis. Gastroenterology
115:19–27, 1998
6. Devesa S, Blot W, Fraumeni J: Changing patterns in the incidence of esophageal and
gastric carcinoma in the United States. Cancer 83:2049–2053, 1998
7. Falk G, Ours T, Richter J: Practice patterns for surveillance of Barrett’s esophagus in
the United States. Gastrointest Endosc 52:197–203, 2000
8. Falk G, Rice T, Goldblum J, et al: Jumbo biopsy forceps protocol still misses unsus-
 APPEARANCE AND PROGNOSIS OF DYSPLASIA IN BARRETT’S ESOPHAGUS 75

pected cancer in Barrett’s esophagus with high grade dysplasia. Gastrointest Endosc
49:170–176, 1999
9. Gross G, Canto M, Hixson J, et al: Management of Barrett’s esophagus: A national
study of practice patterns and their cost implications. Am J Gastroenterol 94:3440–
3447, 1999
10. Hirota W, Loughney T, Lazas D, et al: Is H. pylori associated with specialized intestinal
metaplasia of the esophagus or stomach? A prospective study of 889 patients. Gastro-
enterology 112:A149, 1997
11. Holscher A, Bollschweiler E, Schneider P, et al: Early adenocarcinoma in Barrett’s
oesophagus. Br J Surg 84:1470–1473, 1997
12. Kawakami K, Brabender J, Lord R, et al: Hypermethylated APC DNA in plasma and
prognosis of patients with esophageal adenocarcinoma. J Natl Cancer Inst 92:1805–
1811, 2000
13. Levine D, Blount P, Rudolph R, et al: Safety of a systematic endoscopic biopsy protocol
in patients with Barrett’s esophagus. Am J Gastroenterol 95:1152–1157, 2000
14. Messmann H, Knuchel R, Baumler W, et al: Endoscopic fluorescence detection of
dysplasia in patients with Barrett’s esophagus, ulcerative colitis, or adenomatous
polyps after 5-aminolevulinic acid-induced protoporphyrin IX sensitization. Gastroint-
est Endosc 49:97–101, 1999
15. Miros M, Kerlin M, Walker N: Only patients with dysplasia progress to adenocarci-
noma in Barrett’s oesophagus. Gut 32:1441–1446, 1991
16. Nigro J, Hagen J, DeMeester T, et al: Occult esophageal adenocarcinoma. Extent of
disease and implications for effective therapy. Ann Surg 230:433–440, 1999
17. Nijhawan P, Wang K: Endoscopic mucosal resection for lesions with endoscopic fea-
tures suggestive of malignancy and high-grade dysplasia within Barrett’s esophagus.
Gastroenterology 52:328–332, 2000
18. O’Connor J, Falk G, Richter J: The incidence of adenocarcinoma and dysplasia in
Barrett’s esophagus. Am J Gastroenterol 94:2037–2042, 1999
19. Ofman J, Lewin K, Ramers C, et al: The economic impact of the diagnosis of dysplasia
in Barrett’s esophagus. Am J Gastroenterol 95:2946–2952, 2000
20. Overholt B: A multicenter, partially blinded, randomized study of the efficacy of
photodynamic therapy using porfimer sodium for the ablation of high grade dysplasia
in Barrett’s esophagus: Results of 6-month follow up. Gastroenterology 120:A-79, 2001
21. Panjehpour M, Overholt B, Vo-Dihn T, et al: Endoscopic fluorescence detection of high
grade dysplasia in Barrett’s esophagus. Gastroenterology 111:93–101, 1996
22. Poneros J, Brand S, Bouma B, et al: Diagnosis of specialized intestinal metaplasia by
optical coherence tomography. Gastroenterology 120:7–12, 2001
23. Prevo L, Sanchez C, Galipeau P, et al: p53-mutant clones and field effects in Barrett’s
esophagus. Cancer Res 59:4784–4787, 1999
24. Ramel S, Reid B, Sanchez C, et al: Evaluation of p53 protein expression in Barrett’s
esophagus by two-parameter flow cytometry. Gastroenterology 102:1220–1228, 1992
25. Reid B, Levine D, Longton G, et al: Predictors of progression to cancer in Barrett’s
esophagus: Baseline histology and flow cytometry identify low and high risk patient
subsets. Am J Gastroenterol 95:1669–1676, 2000
26. Reid B, Blount P, Feng Z, et al: Optimizing endoscopic biopsy detection of early
cancers in Barrett’s high-grade dysplasia. Am J Gastroenterol 95:3089-3096, 2000
27. Reid B, Haggitt R, Rubin C, et al: Observer variation in the diagnosis of dysplasia in
Barrett’s esophagus. Hum Pathol 19:166–178, 1988
28. Robertson C, Mayberry J, Nicholson D, et al: Value of endoscopic surveillance in the
detection of neoplastic change in Barrett’s oesophagus. Br J Surg 75:760–763, 1988
29. Rogers E, Goldkind S, Iseri O, et al: Adenocarcinoma of the lower esophagus: A disease
primarily of white men with Barrett’s esophagus. J Clin Gastroenterol 8:613–618, 1986
30. Ruol A, Merigliano S, Baldan N, et al: Prevalence, management and outcome of early
adenocarcinoma (pT1) of the esophagogastric junction. Comparison between early
cancer in Barrett’s esophagus (type I) and early cancer of the cardia (type II). Dis
Esophagus 10:190–195, 1997
31. Sampliner R: The Practice Parameters Committee of the American College of Gastroen-
76 SAMPLINER

terology. Practice guidelines on the diagnosis, surveillance, and therapy of Barrett’s

esophagus. Am J Gastroenterol 93:1028–1032, 1998
32. Sampliner R, Fass R: Partial regression of Barrett’s esophagus: An inadequate endpoint.
Am J Gastroenterol 88:2092–2094, 1993
33. Schulz H, Miehlke S, Antos D, et al: Ablation of Barrett’s epithelium by endoscopic
argon plasma coagulation in combination with high-dose omeprazole. Gastrointest
Endosc 51:659–663, 2000
34. Shaheen N, Crosby M, Bozymski E: Is there publication bias in the reporting of cancer
risk of Barrett’s esophagus? (Abstract). Gastroenterology 116:G0404, 1999
35. Sharma P, Morales T, Bhattacharyya A, et al: Squamous islands in Barrett’s esophagus:
What lies underneath? Am J Gastroenterol 93:332–335, 1998
36. Skacel M, Petras R, Gramlich T, et al: The diagnosis of low-grade dysplasia in Barrett’s
esophagus and its implications for disease progression. Am J Gastroenterol 95:3383–
3387, 2000
37. Stein H, Feith M, Mueller J, et al: Limited resection of early adenocarcinoma in Barrett’s
esophagus. Ann Surg 6:733–742, 2000
38. vanSandick J, vanLanschot J, tenKate F, et al: Pathology of early invasive adenocarci-
noma of the esophagus or esophagogastric junction. Cancer 88:2429–2437, 2000
39. Wallace M, Perelman L, Backman V, et al: Endoscopic detection of dysplasia in
patients with Barrett’s esophagus using light-scattering spectroscopy. Gastroenterology
119:677–682, 2000
40. Weinstein W, Ippoliti A: The diagnosis of Barrett’s esophagus: Goblets, goblets, goblets.
Gastrointest Endosc 44:91–96, 1996
41. Weinstein W, Lieberman D, Lewin K, et al: Omeprazole induced regression of Barrett’s
esophagus: A 2 year randomized, controlled double blind trial. Gastroenterology
110:A294, 1996
42. Weston A, Badr A, Hassanein R: Prospective multivariate analysis of clinical, endo-
scopic, and histological factors predictive of the development of Barrett’s multifocal
high grade dysplasia or adenocarcinoma. Am J Gastroenterol 94:3413–3419, 1999
43. Younes M, Ertan A, Lechago L, et al: p53 protein accumulation is a specific marker of
malignant potential in Barrett’s metaplasia. Dig Dis Sci 42:697–701, 1997
44. Younes M, Lebovitz R, Lechago L, et al: p53 protein accumulation in Barrett’s metapla-
sia, dysplasia and carcinoma. Gastroenterology 105:1637–1642, 1993

Address reprint requests to

Richard E. Sampliner, MD
Southern Arizona VA Health Care System
University of Arizona College of Medicine
Gastroenterology 111 G1
3601 South 6th Avenue
Tucson, Arizona 85723

email: [email protected]