Editor: Joseph Sung
3. Barrett's esophagus and esophageal cancer
Norman E. Marcon, Umar Beejay & Shou-jiang Tang
The clinical significance of Barrett's esophagus (BE) has increased enormously since it was first described by the surgeon
Norman Barrett in the 1950s. This has occurred for three reasons (i) realization that BE develops as a direct consequence
of chronic gastroesophageal reflux disease (GERD) (ii) the appreciation that BE is a risk factor for the development of esophageal
adenocarcinoma (EAC), and (iii) the fact that the incidence of EAC, a disease with typically dismal outcomes because of its
advanced presentation, has increased over 400% over the past 2030 years . Accordingly, greater efforts have been directed towards detection of the premalignant BE in an attempt to reduce the incidence
of EAC. Despite, or perhaps because of, advances in our knowledge of BE, there has been a concomitant increase in the uncertainties
surrounding the condition. Controversies persist about its definition, nomenclature and indeed whether earlier endoscopic
detection and treatment truly affect outcome.
Esophageal cancer (EC) is an uncommon malignancy accounting for 1% of all cancers. Nevertheless, EC is the seventh leading
cause of cancer death worldwide. The two most common types of EC are esophageal squamous cell carcinoma (ESCC) and esophageal
adenocarcinoma (EAC). Because EC often presents at an advanced stage when first diagnosed, prognosis is poor, with a five-year
survival rate of less than 10%, and thus the emphasis of most management is on palliation rather than cure. Although screening
for EC has proven to be an effective strategy in parts of the world where the incidence is high, it has not been extensively
studied in North America. Endoscopy with biopsy remains the 'gold standard' diagnostic test and its role has been expanded both in the development of potentially curative treatments (e.g. endoscopic
mucosal resection (EMR) photodynamic therapy (PDT)) as well as in the palliative arena (e.g. esophageal endoprostheses (EE)
and ablative therapies). Although our major therapeutic paradigm is still curative/palliative rather than preventive, the latter approach (e.g. examination of dietary factors, tobacco cessation, and decreased
alcohol consumption) should not be ignored as a management approach. In addition, the recent focus on chemopreventive strategies
is an area of intense study. Hopefully, earlier detection (e.g. the 'optical biopsy' and biomarkers), more accurate staging, and advances in our understanding of tumor biology may eventually lead to an improvement
in outcome parameters.
This review of BE and EC cannot be comprehensive but will try to emphasize current technical and clinical aspects of endoscopy,
from an endoscopic perspective, using an evidence-based approach.
BE is a premalignant condition of the lower esophagus in which the normal squamous lining of the distal esophagus is replaced
by columnar epithelium. BE is best defined as the 'presence of columnar-appearing mucosa of any length coupled with the finding of incomplete (specialized) intestinal metaplasia
(SIM) on biopsy within the distal esophagus. Although three types of columnar epithelia ((i) junctional-type epithelium (a.k.a
cardiac epithelium), (ii) gastric fundic-type epithelium, and (iii) specialized intestinal metaplasia (SIM)), have been described
in BE, it is SIM that exists as the most common type, as well as the only one with a clear malignant potential. SIM can be
distinguished from the other two types of epithelia described above by the unique presence of goblet cells, that contain characteristic
acidic mucins which stain blue on Alcian Blue staining.
SIM occurring in BE is called incomplete (Types II and III) because it is less well differentiated than the more mature intestinal
metaplasia (IM) occurring at the squamo-columnar junction which is fully differentiated (Type I), and which is typified by
the presence of Paneth cells. In contrast, Paneth cells are only sporadically seen in incomplete intestinal metaplasia. Types
II and III SIM are characterized as incomplete because of their often villiform appearance and their tendency to manifest
significant immature and/or incomplete morphological and histochemical features.
This definition of BE highlights the fact that the critical defining feature of BE is the dependence on a histopathologic
verification of SIM rather than just the presence of columnar lined epithelium within the distal esophagus: this renders the
definition primarily pathologic than clinical.
Although most patients with BE do not progress to cancer, those who do seem to have markedly increased survival when the cancer
is detected at an early stage. At the same time it has been noted that the risk of developing EAC in patients with BE may
have been overrated in the past because of publication bias . Currently, the overall risk is estimated to be 0.5 per 100 patients per year (i.e. 1 cancer will occur out of 200 patients
followed for one year) . Current research efforts are focused in several areas, including the elucidation of better methods of detection and stratification
of the condition from both an endoscopic and molecular perspective, clarifying the natural history of BE and/or coexisting grades of dysplasia, investigating earlier preventive strategies (e.g. chemoprevention), as well as efforts
aimed at identifying optimal treatment modalities from survival, quality of life, and economic perspectives.
Definition of Barrett's esophagus
Long-segment Barrett's esophagus (LSBE)
Long-segment BE (LSBE) is defined as BE segments greater than 3 cm in length (Fig. 1). In fact, all BE was identified as LSBE up until the early 1990s when segments of SIM less than 3 cm long were discovered (these segments were correspondingly called short-segment BE). It remains unclear whether LSBE and
short-segment BE (SSBE) have the same pathogenesis and natural history, or indeed whether SSBE eventually progresses to LSBE.
Nevertheless, progression from SSBE to LSBE has rarely been reported in clinical practice. In addition, uncertainty exists
surrounding the issue of the varying risk of dysplasia associated with SSBE and LSBE: is it related purely to the respective
length/areas of the segments, or are other factors involved? In spite of these uncertainties, the distinction between the two forms
of BE persists in clinical practice, even though both subtypes continue to be treated in a similar fashion. The clinical gestalt
currently suggests that SSBE is less sinister than LSBE, though the data supporting this inference is unclear.
Although the definition of BE seems robust, difficulties may be encountered in the translation of this operative definition
to the real-life situation in the esophagus. Although it is easy to differentiate between squamous and glandular epithelium
on the initial endoscopic evaluation of the lower esophagus, greater difficulty is encountered in deciding on where the esophagus
ends and the stomach begins, and in defining the position of the lower end of the gastroesophageal junction (GEJ) . That the GEJ itself (i) includes part of the distal third of the esophagus as well as the gastric cardia (ii) can be defined
both anatomically and functionally, and (iii) can have its appearance made even more confusing by the presence of a hiatal
hernia or gastroesophageal reflux disease (making the sphincter more patulous), thus the distinction is all the more problematic.
One suggestion has been to use the top of the gastric rugae (folds) as a surrogate marker for the lower margin of the GEJ
when the proximal stomach is in a collapsed air state. Another has been to define the lower end of GEJ as the point at which
the diameter of the gastrointestinal lumen widens into the stomach. Neither recommendation may be the exact correlate of the
lower end of the GEJ but they offer more practical methods of defining the lower end of the GEJ endoscopically.
In addition, it should be appreciated that while the esophagus joins the stomach under the diaphragm after a short intra-abdominal
segment, the normal position of the squamocolumnar junction (SCJ) is 2 cm proximal to this junction in the tubular esophagus. In summary, one can surmise that the ability to diagnose BE (especially
SSBE), is confounded by the complex anatomy of the GEJ, and this has important implications with regards to standardizing
our approach to this condition.
Short-segment BE (SSBE)
Short-segment BE (SSBE) is defined as BE less than 3 cm long . The difficulties encountered in defining the lower margin of the GEJ lead to difficulties in delineating the point from
which a length of BE should be measured, which, in turn often leads to problems in deciding whether endoscopic biopsy specimens
from this area are indeed from BE or from the gastric cardia. Although normal cardiac mucosa is histologically identical to
antral mucosa and does not normally contain intestinal metaplasia, chronic injury can lead to intestinal metaplasia in this
area indistinguishable from BE.
The 3 cm cutoff is an arbitrary method of separating SSBE from LSBE. Some believe that the premalignant potential of BE is directly
related to the surface area of BE and that accordingly, the risk of development of EAC may be less in SSBE than in LSBE (a
notion yet to be confirmed). Others believe that the malignant potential of SSBE/LSBE may not be a function of length/area, but that the conditions may represent two separate disease entities. Interestingly, both the fact that adenocarcinoma
of the gastric cardia may arise within SSBE, and that IM seems to be a common precursor in both EAC and adenocarcinoma of
the gastric cardia, has led to suggestions that EAC and adenocarcinoma of the gastric cardia may ultimately be one and same
Ultra-short BE (SSBE)
Ultra-short BE (USBE)(a.k.a. microscopic BE) represents small amounts of BE found just at the GE junction. On endoscopic examination,
it is characterized by a small protuberant tongue/s of red tissue through which the superficial esophageal vessels may be seen. USBE can be more easily visualized by using
endoscopic adjunctive techniques (e.g. magnification endoscopy with either acetic acid or indigo carmine). Using mini-biopsy
forceps may help target these small areas more precisely. Although it can be classified as a subtype of SSBE, it has been
suggested that USBE should be distinguished from SSBE because it may behave differently from SSBE in terms of the natural
history and malignant potential of the lesion. This hypothesis remains unproven.
Risk factors for Barrett's esophagus
Gastroesophageal reflux disease (GERD)
The strongest risk factor associated with BE is chronic GERD . Up to 5% of people with long-standing GERD will develop LSBE, and between 10 and 15% will develop SSBE. Moreover, the extent
and duration of esophageal acid exposure appears to be an important contributing factor determining the length of Barrett's
mucosa. In a study of 701 patients, BE was discovered in 4% with GERD symptoms for less than a year, but seen in 21% of those
with a 10-year history of GERD. The mechanism by which GERD leads to BE is not entirely clear. Exposure of the esophageal
epithelium to acid is essential but this must ultimately occur primarily because of failure of the sphincteric antireflux
mechanism. The natural history of progression from a sphincter expressing healthy transient lower esophageal relaxations to
one which is incompetent structuro-functionally is unclear and deserves further investigation. Centripetal obesity and gastric
distension have been postulated as factors involved in distortion and stretching of the sphincteric antireflux mechanism.
Accordingly, prevention of BE begins with the effective control of GERD and should be attempted either with medical therapy
(antisecretory medication) or mechanical surgical therapy (fundoplication) . Although surgical antireflux procedures can result in long-term control of GERD symptoms, EC may still occur after surgery
mandating the need for continued endoscopic surveillance if BE persists.
H.pylori and GERD
The effect of Helicobacter pylori (HP) eradication on GERD is more contentious: some data suggests HP should not be treated because HP eradication may worsen
GERD, and that proton-pump inhibitor (PPI) treatment may be less effective in suppressing intragastric acidity in HP-negative
patients. The opposing viewpoint asserts that HP should be eradicated because of its association with an increased incidence
of atrophic gastritis (a premalignant condition) in patients on long-term acid-suppressive therapy. Interestingly, BE is rare,
and EAC uncommon, in many regions of sub-Saharan Africa-areas where the HP prevalence is high (61100%). In fact, some epidemiologic studies have suggested that childhood acquisition of HP may be a protective factor in the
development of GERD in these populations. These findings have not been corroborated in the North American population.
Age, sex, and race
BE occurs predominantly in middle-aged/elderly Caucasian males. While it is not clear why the racial and sex predilection exists, the 'elderly skewed' age distribution is felt to be due to the effects of the chronic interplay of environmental insults (e.g. acid and bile salts)
and genetic factors (e.g. molecular differences in adhesion molecules), which, over time, ultimately leads to metaplasia of
the multipotent mucosal stem cells within the distal esophagus.
Other risk factors
A variety of other independent risk factors for the development of BE have been suggested [6,7]. These include alcohol consumption, cigarette smoking, presence of a hiatal hernia, abnormal bile reflux, cholecystectomy,
a defective lower esophageal sphincter, defective distal esophageal contractions, abnormal number of reflux episodes lasting
longer than 5 min, obesity, and GERD lasting for more than 5 years .
That a hereditary BE has also been described (i.e. the occurrence of six cases of BE over three generations) would suggest
that hereditary factors also influence the development of the condition. Other postulated factors being examined are a diet
high in fat and low in fruits and vegetables, low serum vitamin C, vitamin E, carotenoids, selenium, and the use of medications
that promote reflux (e.g. calcium channel blockers).
While associations discovered by epidemiological studies can be helpful in generating causal hypotheses, it should be appreciated
that the associations may, or may not eventually lead to a link of causality between BE and the factor in question. The questionable
role of some risk factors is demonstrated by for example, various studies examining the purported relation between bile reflux
and BE. One study associated cholecystectomy, a procedure in itself associated with duodenogastric reflux, with an increased
incidence of adenocarcinoma of the esophagus. Another case-control study assessing the risk for BE after partial gastrectomy
found no association between BE and gastric surgery, and summarily stated that reflux of bile without acid was not sufficient
to damage the esophageal mucosa.
Epidemiology of Barrett's esophagus
BE occurs in middle-aged/elderly Caucasian males . The male to female sex ratio has been estimated to be 2.6 : 1. Prospective studies have reported incidence rates of BE from 0.2% to 1.9%. Currently the most reasonable estimate of incidence
is in the lower range at 0.5% per year. Prevalence in the US population is cited at 2%. BE is present in 12% of patients undergoing upper endoscopy for various reasons, 610% of patients complaining of heartburn and up to 15% of patients with long-standing GERD. A recent 34 years review of the
literature (over 750 studies), identified 1503 cases of EAC from 12 eligible studies, and found that the overall percentage
of patients undergoing resection who carried a prior diagnosis of Barrett's was only 4.7% ± 2.9%, further indicating how poor current screening/surveillance strategies were in affecting the outcome of EAC .
Natural history of Barrett's esophagus
The natural history of non-dysplastic and dysplastic BE is unclear . The overall incidence of any grade of dysplasia has been reported as low as 23% (median 4.8 years follow-up) and as high
as 80% (20 years of follow-up). Reported rates for the development of EC, once high-grade dysplasia (HGD) has developed have
varied from 16 to 59% over 58 years of follow-up.
One study involving 79 patients with BE/HGD, discovered EAC in 4/79 during one year of intensive surveillance. The remaining 75 were followed for a mean of 7.3 years during which 12/75 (16%) developed EAC. The low incidence of EAC in this study may be attributed to over-reading by the single pathologist
associated with the study: no second expert pathologist provided secondary confirmation of the HGD. Although 'weeding out' of HGD had occurred in the first year of intensive endoscopic surveillance, the extra 5% patients discovered in this fashion
did not fully account for the low incidence rate reported.
Another study followed 15 patients with 'unifocal' HGD for a mean follow-up of 3 years, and reported the development of EAC in 27% patients. In another study of 100 patients
with BE/HGD, the incidence of EC was reported as 32% during eight years of follow-up. Moreover 19/32 (59%) of the cancers were detected within the first 6 months of the 8 years follow-up for HGD.
A further study reported EAC in 59% of 76 BE/HGD patients after five years of follow-up. Finally, the highest incidence of EAC was reported in a small study of 8 patients:
5/8(62.5%) developed EAC after only 1 year of follow-up.
An interesting study of 100 consecutive patients (mentioned above) further suggested that the risk of development of EAC was
directly correlated with the extent of HGD found in biopsy specimens: if HGD was confined to a single focus of less than 5
crypts, it was termed 'focal'. If the HGD was seen in more than 5 crypts, or in more than one biopsy specimen, the dysplasia was deemed 'diffuse'. 33/100 patients had 'focal' HGD and 67/100 had 'diffuse' HGD. The three year cancer incidence was 14% in the focal HGD group and 56% in the 'diffuse' HGD group.
In summary, finding HGD within BE is ominous, with the lowest 5 years incidence extrapolated as 9%. Several studies have discovered
EAC in nearly a third of these patients within just one year of intensive follow-up. At the same time, it should be remembered
that the development of EAC in patients with BE is not inevitable, and the process may take several years. Moreover, data
about the very presence of EAC in patients with BE is confusing, with most current data not correlating endoscopic features
with histology and staging. The crucial factor remains whether the HGD/EAC has breached the muscularis mucosa, and is into the submucosa, or whether it remains above the muscularis mucosa. Clearer
information in this area will determine the newer modern treatment strategy of endoscopic vs. surgical therapy.
Pathogenesis of Barrett's esophagus
A vitally important prerequisite for the development of BE is an incompetent sphincteric antireflux mechanism. It is in this
context that a relative or absolute increase in acid becomes a prerequisite for the development of GERD, the most significant
risk factor for BE. When compared to patients with erosive/nonerosive esophagitis, patients with BE typically have greater acid exposure based on 24- h pH monitoring, perhaps more related
to the combination of the almost uniform presence of hiatal hernia, a lower basal lower esophageal sphincter (LES) pressure,
and defective esophageal acid clearance, than to an absolute increase in acid. In fact, many BE patients do not have increased
gastric acid and pepsin secretion when compared to patients without BE.
Moreover, recent evidence suggests that a positive symptomatic response to PPIs does not correlate with the results of pH
monitoringin one study up to 80% of study participants continued to have abnormal amounts of distal esophageal acid exposure (particularly
overnight), despite good symptomatic relief. These findings suggest that BE patients may be less sensitive to the presence
of acid in the esophagusa theory further supported by findings that demonstrate that BE patients are less likely to have symptoms with esophageal
acid-perfusion studies than those patients with reflux esophagitis. Moreover, at a cellular level, acid has been demonstrated
to directly increase BE cellular proliferation , as well as increase the expression of cyclo-oxygenase-2 (COX-2) and prostaglandin E2 (PGE2) levels . Accordingly, research has been directed at reducing cellular proliferation by reducing acid exposure in conjunction with
That BE may develop in a few patients with achlorhydria, and postgastrectomy, suggests a possible etiologic role for duodenal
contents (i.e. conjugated and unconjugated bile acids, lysolecithin and pancreatic trypsin) . While conjugated bile acids seem to need an acidic milieu to produce esophageal mucosal injury, unconjugated bile acids
and trypsin can cause esophageal injury at more neutral pH values.
Several studies have shown an incremental increases in both acid and duodenogastroesophageal reflux (DGER) across the entire
spectrum of nonerosive GERD, erosive GERD, and BE. Furthermore, the fact that DGER seems to occur more frequently in the settings
of abnormal acid reflux raises the possibility that a synergistic interaction between acid and bile may contribute to the
development of BE, and ultimately EAC. Moreover, some evidence suggests that aggressive acid suppression with PPIs can result
in a decrease in both acid levels and DGER. Just as with acid, bile exposure increases BE cellular proliferation and the expression
of cyclo-oxygenase-2 (COX-2) and prostaglandin E2 (PGE2) .
The role of HP in the development of BE is controversial with most studies finding no causal relationship between HP and BE.
Yet other evidence suggests a possible protective role for HP attributable to several factors : (i) HP induced corpus-predominant gastritis is associated with decreased acid secretion which in turn could lead to less
GERD; and (ii) the protective effect of HP may be related to colonization with cag A-positive strains of HP.
A recent study found that the prevalence of cag A-positive HP was inversely proportional to the manifestations of GERD severity when compared with a control population.
In addition, GERD patients with cag A-positive strains of HP had a threefold decreased risk of complications of GERD when compared to patients with cag A-negative strains of HP. Moreover, a recent study reported that the presence of cag A-positive strains of HP was associated with an increased risk of distal gastric cancers but a reduced risk of for EAC and
gastric cardiac adenocarcinomas.
Novel diagnostic techniques for Barrett's esophagus
Endoscopically, with the exception of SSBE, most BE is relatively easy to identify. Novel diagnostic techniques are directed
at diagnosing and mapping endoscopically inapparent dysplasia. The significant technological advances in the application of
light technologies, advances in our understanding and use of dyes and magnification, and improvements in the technical specifications
of endoscopes (e.g. magnification and high-resolution), have meant that numerous techniques are potentially available for
the purposes of screening, diagnosis and surveillance of BE .
The use of high-resolution optical techniques for tissue diagnosis without biopsy ('optical biopsy') offers significant potential advantages over standard techniques, such as tissue biopsy, both in terms of patient care and
medical costs. For example, optical techniques are potentially faster, and sedatives may not be needed. While many of the
new diagnostic modalities have shown promise, most remain at the investigational rather than clinical level of application.
Currently, chromoendoscopy and magnification endoscopy are the most clinically available techniques to the practicing endoscopist.
Conventional EGD is unable to delineate and capture the fine detail of the gastrointestinal mucosa. Tissue staining (dye spraying
or chromoscopy), used to enhance and improve the characterization of mucosal surfaces, was first described in 1966. Chromoendoscopy
describes the application and adaptation of chromoscopy to endoscopic practice . Although the absorptive stains have the greatest utility in evaluation of dysplasia and neoplasia in the esophagus, contrast
stains (e.g. indigo carmine) may also be helpful. Absorptive stains are stains that react with different cellular constituents
once absorbed into certain cells and include Lugol's iodine solution, methylene blue, and toluidine blue.
Lugol's iodine solution. Lugol's iodine solution is an absorptive stain that stains benign, glycogen-containing, non-keratinized squamous cells brown
but does not stain dysplastic, inflamed, or malignant squamous cells or columnar mucosa. By delineating benign squamous from
columnar mucosa, vital staining with Lugol's iodine solution improves the ability of EGD to diagnose BE with sensitivity,
specificity, and accuracies of 89%, 93%, and 91%, respectively. A 1.5% to 4% solution is sprayed on the esophagus and within
minutes the normal whitish squamous mucosa will change to dark brown/greenish brown as a result of binding of the iodine to glycogen within the non-keratinized squamous epithelium. Squamous cells
that are inflamed, dysplastic, or malignant will not stain, and columnar epithelium remains pink and unstained (Fig 2), parts (i) and (ii).
Methylene blue. Methylene blue (MB) is an absorptive stain that stains actively absorbing intestinal-type cells but does not stain squamous
and gastric cells or gastric-type metaplasia. It can help differentiate the SIM of BE from gastric-type epithelium. MB stains
SIM in BE with high accuracy, quickly, and with no side-effects . MB staining involves application of a mucolytic agent (acetyl cysteine or pronase), washing the mucolytic agent off after
2 min, applying dye and waiting for 2 min, finally followed by adequate washing with 100300 cc water to remove excess dye. The mucolytic agent helps to remove surface mucus, thereby increasing the uptake of dye
into epithelial cells. Although the technique would appear to take a short time in theory, the reality is that the technique
remains labor-intensive, subjective, and as yet, unstandardized.
Although more SIM at the SCJ has been detected with MB staining than without staining, the sensitivity and specificity of
staining was only 79%/66%, respectively. Its efficacy in improving endoscopic surveillance was shown in a randomized controlled trial (RCT) in which
MB-directed biopsy led to identification of a much larger proportion of specialized columnar epithelium in endoscopic biopsy
samples compared with those samples obtained by random biopsy (P = 0.0006). This was particularly evident in patients with SSBE (94% vs. 54%) than LSBE (92% vs. 72%).
Moreover, the estimated average cost of diagnosing a cancer using MB-directed biopsy was less than half the costs associated
with that of random biopsy. Nevertheless, other studies have not confirmed these findings, and currently the American Society
of Gastrointestinal Endoscopy (ASGE) classifies this technique as of unproven benefit.
Toluidine blue. Toluidine blue is an absorptive stain that stains the nuclei of columnar cells (gastric and intestinal-type), but does not
stain squamous cells. Accordingly, it is used in BE to selectively stain gastric and intestinal metaplasia. Staining is accomplished
by applying a 1% acetic acid rinse followed by a 1% aqueous solution of toluidine blue, followed by a final wash with 1% acetic
acid. The sensitivity and specificity rates for diagnosing BE are 98%/80%, respectively. While toluidine blue stains columnar-type mucosa deep blue, it cannot differentiate between the various
histologic subtypes of Barrett's esophagus.
Indigo carmine. Indigo carmine (IC) is a blue contrast stain not absorbed by cells. Most of the work with IC has been in the detection of
colonic adenomas, and only recently has attention moved to investigating its use in BE. It pools in the crevices between mucosal
projections and highlights subtle mucosal abnormalities. A 0.4% solution of IC is sprayed on columnar mucosa after delineation
of the SCJ with Lugol's solution. Areas of possible BE are then examined using a magnifying endoscope (Olympus JPG-200Z endoscope
(Olympus Corp, Tokyo, Japan)) that can provide adjustable image magnification from × 10 to × 35. BE is visualized as a slightly raised surface pattern with a villiform appearance (similar to small intestinal mucosa)
The use of magnification techniques offers another way of improving the inspection of the esophageal epithelium [16,19]. The advent of magnifying electronic video endoscopes, incorporating an optical zoom at the tip, permits magnification rates
of up to × 150. Magnification gastroscopes (e.g. GIF-200Z, Olympus America Inc., Melville, NY, USA) are similar in diameter (11.9 mm) to standard gastroscopes and thus do not normally create any additional difficulties in esophageal intubation. However,
the cost of high-magnification endoscopes is about $2000 to $3000 ($US, not including the video processor) more than standard
endoscopes. In addition, examinations can be hampered by the relative inability to keep the viewed field constant due to the
amplitude of cardiorespiratory movements. Adding a plastic hood cap will help to stabilize the motion of the mucosa. Moreover,
the procedure usually takes longer, and requires longer periods of conscious sedation (with all its attendant problems).
In addition, the development of high-resolution electronic video endoscopes with built-in structure enhancement has further
enhanced our visualization of BE. Resolution of an endoscopic image is defined as the ability to distinguish optically between
two closely approximated objects or points. The new high-resolution videoendoscopes offer a 60100% increase in resolution over the older models.
These two modalities were initially designed to be specifically used in combination with other adjunctive techniques such
as chromoendoscopy, or the use of acetic acid, in an effort to improve the specificity of identifying the more dysplastic
areas within BE .
The utility of magnification techniques in the evaluation of BE is good compared with standard endoscopy, and the reported
sensitivity/specificity rates are 82%/34%, respectively  One recent study of 48 patients utilized the combination of magnification endoscopy (x 35) with topical application of 1015 ml of 1.5% acetic acid to the distal esophagus. Acetic acid produced reversible intracellular cytoplasmic protein denaturation,
leading to a whitish discoloration of the squamous epithelium, as compared with the SIM and gastric epithelium which remained
Moreover, this study described 4 types of pit pattern within the red columnar lined epithelium, with subsequent histological
analysis of biopsies showing a close correlation between the type of pit pattern described and the extent of SIM. Incomplete
SIM occurred in low frequencies in types I and II pit patterns, but with high prevalence in types III and IV pit patterns
 (Fig 3), sections i-iv. Another study evaluated the use of magnification chromoendoscopy in suspected BE: in this study, finding
a ridged or villous pattern predicted SIM in 89% of cases, while only a third of cases with a circular pattern demonstrated
SIM on histology.
Magnification chromoendoscopy may also prove useful in the evaluation of BE/HGD patients. A recent abstract from France examined 23 patients referred for endoscopic treatment of a localized focus of
HGD/early esophageal carcinoma. While standard endoscopy was unable to localize the lesions of interest in 10 patients, magnification
endoscopy with indigo carmine detected lesions in 7/10 of these patients, and moreover, located 4 additional foci of HGD in the other 13 patients with visible lesions seen on
standard endoscopy. Although promising, these preliminary results need further validation before more widespread clinical
application of this technique can be recommended.
Despite the promise surrounding these preliminary studies, it must be stressed that relatively few studies in the esophagus
have been performed using magnification, and even fewer, evaluating the use of high-resolution techniques.
Other investigational techniques
Endoscopic ultrasonography (EUS)
EUS has traditionally been used in the realm of cancer staging (i.e. establishing tumor depth and evaluating loco-regional
lymph nodes) (Fig. 4). With the advent of high-resolution EUS, attention has focused on its ability to examine dysplasia within the mucosal layer.
High frequency miniprobes that may be passed through the biopsy channel of a regular gastroscope , or dedicated high frequency radial/linear-array echoendoscopes have been used in the preliminary evaluation of the mucosal layer of the esophagus. EUS plays
a role only in the staging, and not in the diagnosis of invasive carcinoma, but currently has no established role in screening
or surveillance of BE with or without dysplasia . As the limits of EUS resolution increase with the development of higher frequency devices, this technique may potentially
be of help in the future.
Light induced fluorescence (LIF)
The underlying principle involved in LIF is that biological tissues contain elements which, when stimulated by light energy,
can become excited, and then emit specific patterns of fluorescent light that can allow differentiation between various tissue
types . Such fluorescence may come from endogenous fluorophores (e.g. nicotinamide adenine dinucleotide (NAD), collagen, flavins,
porphyrins), or exogenous fluorophores (e.g. protoporpyhrin IX, porfimer sodium). Fluorescence studies using mucosa stimulated
at 437 nm blue light in dysplastic BE show reduced fluorescence in the 470480 nm range, and increased fluorescence at 630 nm. Although the differences in the autofluorescence spectra of squamous epithelium and BE with high-grade dysplasia are promising,
defining spectral differences between HGD and low-grade dysplasia (LGD) remains problematic.
There are three methods of applying the principles of LIF in BE. The first method, called Laser-induced fluorescence spectroscopy
(LIFS), involves the transmission of a monochromatic light source via a through-the-scope catheter probe to the tissue surface.
The tissue surface is excited and then releases light (with specific peaks), that can be collected via the collecting fibers
in the same probe from the BE, thereby obtaining point spectroscopic measurements of an area of 13 mm2. Most of the fluorescence emanates from within the first 500 µm depth of the mucosal surface.
LIFS has shown some promise in differentiating non-dysplastic from dysplastic BE, with preliminary studies demonstrating an
ability to detect the presence of HGD within BE with up to 90% accuracy, as compared with standard histopathologic examination.
In another study of 13 patients with EC, LIFS had a sensitivity of 97% and specificity of 95% for the diagnosis of esophageal
carcinoma . Nevertheless, its role in detecting dysplasia in BE has been hampered by false positive rates of up to 30%, and by the point
sampling technique which interrogate only a tiny area of the surface.
The second technique, fluorescence endoscopy (FE), applies a method called ratio fluorescence imaging in the production of
a real-time fluorescent endoscopic system (LIFE-GI system; Xillix Technologies, Vancouver, BC, Canada). The LIFE-GI system
utilizes a blue monochromatic excitatory light source, and two cameras attached to the fiber-optic endoscope to detect red
and green fluorescence, which can then be displayed as real-time endoscopic images (Fig. 5). With this system normal esophageal epithelium appears green, BE appears a brick red and high-grade dysplasia appears a
darker brick red (Fig. 6).
Several preliminary studies have shown that FE is more sensitive and specific for detecting HGD compared with random biopsies
in BE patients with sensitivity and specificity values of 87%/90%, respectively, in distinguishing between HGD and non-dysplastic BE. In a study of 111 BE patients, comparing autofluorescence
(AF) with fiber-optic white light endoscopy (WLE), AF was found to be superior to WLE, detecting 20/24 cases of HGD, compared with only 11/24 HGD cases for WLE. This technique allows screening of wide areas of many square centimetre.
The third technique, photodynamic diagnosis, involves examining the BE following the administration of exogenous fluorescent
agents such as porfimer sodium (Photofrin®, Axcan Pharma, Montreal, Canada), or with the pro-drug 5-aminolevulinic acid (5-ALA) that is converted to protoporphyrin
IX in vivo. By preferentially and selectively concentrating themselves (or their metabolites) in dysplastic/malignant epithelium relative to normal tissue, these substances enhance the intensity of the emitted fluorescence from these
areas and can improve the differentiation of HGD from non-dysplastic BE .
Preliminary studies show that the use of photodynamic diagnosis is able to distinguish between dysplastic BE and normal esophageal
mucosa. A study using 5-ALA in 47 patients with BE using an oral dosage of 15 mg/kg body weight and a light source in the 380440 nm range reported sensitivity/specificity rates for diagnosing dysplasia of 80%/56%, respectively. The ability of 5-ALA to be taken orally, in non-therapeutic doses, together with its shorter pharmacokinetic
half-life (and accordingly shorter photosensitive period) are factors that influence its preference as a possible detector
of dysplasia when an imaging system is used compared with Photofrin®. However, it should be appreciated that the depth of tissue injury with 5-ALA is more superficial than that induced by Photofrin®.
Despite promising preliminary results, all three techniques remain at the investigational stage, and are hampered by a number
of limitations (i.e. unclear dosimetry, timing, regulatory issues, expensive equipment, low specificity (because of an inability
to differentiate between inflammatory and neoplastic tissue), and in the case of LIFS, limited sampling areas).
Optical coherence tomography (OCT)
Optical coherence tomography (OCT) is an imaging modality that allows real-time, high-resolution, cross-sectional imaging
of depths of up to 23 mm of biologic tissue. The way in which OCT works has been compared to B-mode ultrasound: while ultrasound harnesses the differences
in reflected sound waves in the interrogation of subsurface structures, OCT uses tissue depth-dependent properties of backscattered
light to create a near histological 'optical biopsy'. OCT seems to work best in the esophagus yielding near histological quality (1020 µm) images of 12 mm depth, and permitting adequate visualization of the mucosa and submucosa.
Preliminary studies have demonstrated that OCT can differentiate between BE and normal squamous epithelium: squamous epithelium
appears brighter, layered and thin, while BE appears less bright, non-layered and thicker. The sensitivity and specificity
of OCT for detecting dysplasia or cancer in a recent study of 44 patients with BE was 75% and 91%, respectively.
Another study examining the use of OCT in evaluating BE, LGD and HGD demonstrated a 100% sensitivity and 85% specificity for
diagnosing HGD . Although it has been reported that OCT may be able to detect the development of cancer underneath neo-squamous epithelium
that develops after performing ablative endoscopic therapy for BE, or in the assessment of BE/HGD after the use of photodynamic therapy (PDT), more data is required to validate the utility of such approaches. Until then,
OCT remains strictly a research tool, given the significant expenses and technical demands associated with its use.
In summary, although the potential exists for an increase in cellular detail with OCT, true histological-grade quality images
are some time away. Probably the first benefit from OCT will likely be as a method of staging of dysplastic BE, necessarily
involving better definition of the mascularis mucosae/submucosal barrier.
Reflectance and elastic light scattering spectroscopy
When light enters biologic tissue it can interact in one of four ways: it can be scattered, absorbed, lead to fluorescence
or produce Raman scattering. Elastic light scattering spectroscopy (ELSS) is a technique dependent on the scattering interaction:
it measures the wavelength dependence of light that has entered the tissue, been scattered by numerous constituents (subcellular
structures and tissue layers) within the tissue and re-emitted . The detected scattered light provides information not only about the morphological properties of the examined epithelium
but also acts as an index of the tissue hemoglobin concentration. ELSS can distinguish between dysplastic and non-dysplastic
tissue based on differences in nuclear size, with malignant tissue displaying larger hyperchromatic nuclei, increased mitotic
rates, and crowding.
Although ELSS has shown some promise in distinguishing between cellular replication rates, further investigation is required
in order to more fully understand the fundamental interactions of light with tissue, and more specifically, the differences
between normal, reparative, inflammatory, infectious, dysplastic, and cancerous epithelia highlighted with this modality.
Confocal microscopy is an experimental method of studying the epithelium at the cellular and subcellular level . A thin probe is passed down the accessory channel of a fiber-optic gastroscope, and passes images up to a laser-scanning
confocal microscope. This technique can potentially compete with histological analysis of epithelial substructure, and offers
the potential for the development of a real-time optical biopsy system. Although the current resolution of confocal microscopy
is 1 micrometer, the technique remains at the experimental stage.
Raman spectroscopy (RS)
Raman spectroscopy (RS) measures changes in specific energy states (i.e. vibrational and rotational energies) of the molecular
bonds of atoms and molecules. Biomolecules possess a unique set of vibrational/rotational energies, a signature that in turn, reflects differences in the structuro-functional composition of various tissues
Because the vibrational signals are very weak, a very sensitive (a million times more sensitive than LIF), and expensive system
is required to detect these signals. Accordingly, RS can be used to distinguish between different tissues, or changes within
a single tissue. RS differs from LIFS because the Raman scattering is not contingent on the absorption of the incident light
photon by the molecule. Instead, incident light energy impinges on a biomolecule, exciting it to a higher rotational/vibrational state and, in doing so, emits light waves of typically longer wavelengths (lower frequency).
Moreover, the near infra-red wavelengths that are used in Raman spectroscopy can penetrate up to a depth of 1 mm in human tissues, providing the ability to potentially interrogate the entire esophageal epithelium. Preliminary data from
the use of RS in BE demonstrate a sensitivity of 75% and a specificity of 92% in the detection of dysplasia, but until further
data is available, the technique remains experimental.
Screening for Barrett's esophagus
Because approximately 95% of all people with BE remain undiagnosed, and EAC survival rates are so poor, it has been suggested
that one method of reducing the rate of EAC would be by instituting a screening program for BE. Although screening protocols
have been established, and in part, recommended by expert consensus (American College of Gastroenterology: http://www.acg.gi.org), decisions about who should be screened, and when they should be screened remain problematic. It is unclear whether limiting
a screening program to just GERD patients would significantly affect outcome.
The poor correlation of the presence of BE with just GERD is further supported by two recent studies that reported a prevalence
of BE of between 10 and 25% in patients undergoing colo-rectal cancer screening. Even if it were decided to limit the strategy
to GERD patients, it has proved difficult deciding which patients with GERD should be screened as BE represents only the tip
of the 'GERD iceberg'. Moreover, only limited evidence exists that endoscopic screening prevents deaths from EAC .
One attempt to clarify this area is a seven-point questionnaire developed by Gerson et al. the aim of which was to decide which patients should be screened endoscopically. Significant predictors of BE in the GERD
population that were identified included male sex, heartburn, nocturnal pain, and odynophagia. The sensitivity and specificity
of the questionnaire for predicting BE was 77% and 63%, respectively .
Interestingly, this study did not identify age, race, duration of symptoms, symptom severity, alcohol, or smoking history
as being associated with BE. Moreover, it has been calculated that if screening endoscopy were limited to Caucasian males
aged greater than 50 years with chronic GERD for greater than 5 years, only a limited impact on cancer death rates would result,
especially because up to 40% of patients with EAC report no previous GERD.
Another recent questionnaire attempting to identify screening predictors designed by the Cleveland Clinic group suggested
that BE screening be restricted even more severely to males over 70 years of age with more than 10 years of GERD. The limited
long-term benefits associated with this more restrictive strategy would seem intuitively obvious.
Esophagogastroduodenoscopy (EGD) with biopsy has retained its pivotal role in screening for BE. Although BE is ultimately
a pathologic diagnosis, information gained from an EGD (e.g. length of BE) can help decide on the frequency of future surveillance.
Unfortunately, once it is elected to perform endoscopic screening, no data is available to help guide how often an individual
patient should be screened, or inform as to what biopsy protocol should be employed. The generally accepted (by consensus)
biopsy protocol adopted is that used for endoscopic surveillance of pre-existing BE: four quadrant biopsies every 2 cm, commencing 2 cm above the proximal limit of the gastric folds (rugae), with an endoscope capable of obtaining large-particle biopsies,
and extending until one is clearly in squamous mucosa. In addition, any tongues of glandular epithelium should be biopsied.
Diagnosis of Barrett's esophagus
Although BE is often discovered incidentally in the course of the endoscopic investigation of GERD, with the condition typically
remaining otherwise asymptomatic unless symptoms occur from a malignancy arising from within the BE, it has been estimated
that physicians identify only approximately 5% of the BE population. BE typically appears as a salmon-colored mucosa extending
above the GEJ.
The salmon red appearance is in striking contrast to the whitish appearance of the squamous mucosa of the normal esophagus.
BE may appear as a circumferential sheet, as tongue-like projections creeping up from the GEJ, or as isolated islands of columnar
mucosa interspersed with squamous mucosa. Adjunctive techniques such as chromoendoscopy with Lugol's iodine solution (see
Novel diagnostic techniques/Chromoendoscopy above) may be helpful both in clarifying the presence of BE, and permitting a precise delineation of the field.
Endoscopic detection of BE may be complicated if endoscopic features suggestive of GERD are also present: this is because
even though BE may coexist with GERD, subsequent pathological diagnosis of BE associated with dysplasia may be obfuscated
by the histologic sequelae of acid reflux, because reparative changes can be indistinguishable from dysplasia on histological
Certainly, when a gross abnormality (e.g. ulcer, nodule) is seen within a suspected field of BE, biopsies should be concentrated
on that lesion as there is a much greater chance that a malignancy may be found. However if the mucosa is relatively flat,
with no focal abnormalities (ulcerations, strictures or nodules), the patient's presumed GERD should be treated with powerful
acid-suppressive therapy, typically a high-dose PPI for 812 weeks, before the patient is brought back for a repeat EGD with systematic biopsy.
It is difficult to predict which patient with erosive esophagitis will ultimately heal with complete squamous re-epithelialization,
and which patient will heal but manifest underlying BE. Certainly, the possible delay in subsequent diagnosis of BE associated
with optimizing healing should not put the patient at a higher risk for the development of dysplasia. Even though the primary
use of barium studies as a diagnostic modality is not recommended, any abnormalities on barium studies need to be investigated
Histologically, only the presence of SIM, characterized by the presence of goblet cell (specialized mucin-secreting cells
normally found in the small and large intestines) is diagnostic of BE. Goblet cells are columnar cells with a distended lateral
border (barrel shaped), a compressed basal nucleus, and a basophilic apical cytoplasm. Goblet cells can be identified either
on regular Hematoxylin and Eosin (H&E) stain, or better seen with the Periodic Acid-Schiff Alcian Blue stain .
Histologic interpretation of dysplasia is more problematic . Dysplasia is defined as neoplastic epithelium that remains confined with the basement membrane of the gland within which
it arose, and accordingly should thus be confined to the mucosal surface of the esophageal epithelium, with sparing of the
muscularis mucosa. The dysplastic cells should reach right up to the top of the surface epithelium for a diagnosis of dysplastic
BE to be made . Dysplasia is divided into four categories: negative for dysplasia, indefinite for dysplasia (ID), low-grade dysplasia (LGD)
and high-grade dysplasia (HGD). The term carcinoma-in-situ should be avoided.
Negative for dysplasia is self explanatory. But an interpretation of ID means that the pathologist cannot decide whether dysplasia
is present, and is most often made when esophageal inflammation confounds accurate interpretation. LGD is characterized by
glands lined by cells with crowded, stratified, hyperchromatic nuclei extending onto the mucosal surface. The diagnosis of
LGD is also often challenging because the histologic changes are not specific to neoplasia but may represent reparative changes
secondary to inflammatory injury. In this scenario, 812 weeks of powerful acid-suppressive therapy should be instituted before repeating the biopsies. Moreover, because interobserver
agreement for the diagnosis of LGD amongst experienced pathologists may be as low as 50% , it is important to have pathology reviewed by two independent, preferably gastrointestinal pathologists.
Histologically, HGD typically reveals marked distortion of the glandular architecture, branching and lateral budding of the
crypts, and intra glandular bridging of epithelium leading to a cribriform pattern of back to back glands. However, the most
important feature of HGD is the loss of nuclear polarity on the mucosal surface (absence of a consistent relationship of nuclei
to each other) . Close collaboration between the clinician and the pathologist is always critical for appropriate interpretation of biopsy
material (dysplastic vs. non-dysplastic BE), and often influences further clinical management substantially.
Management of Barrett's esophagus
The need and methods for surveillance remain in dispute.
A surveillance strategy demands the following prerequisites: (1) the method should be low risk; (2) it should make accurate
diagnoses of dysplasia; (3) it should incorporate solid proof that surgical resection for HGD would decrease the incidence
of EAC; (4) it should demand successful resection and a good outcome of any cancers thus detected. According to American Society
of Gastrointestinal Endoscopy (ASGE) guidelines , regular EGD satisfies the first criteria, is the optimal BE surveillance method, and is very useful, with or without adjunctive
techniques, in assessing the extent and progression of the BE into dysplasia.
Regular endoscopic surveillance in BE patients is based on the premise that BE advances through progressive degrees of cellular
dysplasia that typically predate the development of EAC. The efficacy of a surveillance program depends on the ability to
effectively detect premalignant mucosal changes and eliminate them before the development of EAC, which itself depends on
surveillance frequency, extent of sampling of the esophageal epithelium, and the accuracy with which the histological diagnosis
of dysplasia is made . That all patients diagnosed with BE are currently entered into a surveillance program, means that the diagnosis of BE should
be made very carefully, in order to minimize the risks, costs, and quality of life issues associated with a long-term surveillance
The surveillance interval in BE is determined by the presence and degree of dysplasia. For a patient with non-dysplastic BE,
endoscopy with biopsy should be repeated every 23 years. If LGD is noted, the surveillance interval should be shortened to every 6 months for one year and, if no further
progression of dysplasia is seen, the interval extended to every year. If HGD is seen on biopsy, the diagnosis should be confirmed
by a second experienced gastrointestinal pathologist and then the patient assessed for either endoscopic or surgical therapy.
When HGD is diagnosed, intensive surveillance every 23 months or the use of endoscopic ablative therapies may be presented as two alternatives to surgical esophagectomy. Use of
endoscopic ablative therapies can be recommended not only to patients either unwilling or unsuitable for surgical intervention,
but also well-staged patients who would ordinarily be operative candidates.
The American College of Gastroenterology (ACG) endorses only intensive endoscopic surveillance or surgical esophagectomy for
the management of HGD and recommends that endoscopic ablative therapies be limited to patients in clinical trials . More data on the strategy of intensive endoscopic surveillance is required before it can be fully incorporated into a treatment
algorithm: especially in the light of reports that suggest that a full 1/3 patients with BE and HGD with no apparent tumor mass may already have EAC that has been missed because of a biopsy sampling
error. The American Gastroenterological Association (AGA) and ASGE do not currently recommend either intensive endoscopic
surveillance or endoscopic ablative therapies for the treatment of HGD but instead advise that healthy patients with HGD be
referred for surgical esophagectomy. However, in most patients diagnosed with HGD/BE in the absence of nodules, the cancer is usually intramucosal, and should thus be curable by endoscopy.
The use of adjunctive techniques (e.g. chromoendoscopy, magnification) may offer the additional benefit of helping to target
endoscopic biopsies to more suspicious areas within the BE. The use of flow cytometry and immunocytochemistry has been harnessed
in attempts to identify parameters which might help identify a patient subset at increased risk for neoplastic progression.
Factors such as baseline 17p (p53) loss of heterozygosity, flow cytometric increased 4 N fractions, and aneuploidy have all
been suggested as being potential biomarkers for increased malignant potential . Although many of these biomarkers show some promise in identifying a higher risk subset within BE, it has been further suggested
that utilizing a panel of biomarkers may lead to increased sensitivity and specificity. Brush cytology has been advocated
by some aauthors, it has the advantage of sampling larger areas, being simple to use, and not expensive. However, there is
limited information about its efficacy in BE surveillance.
Utility of surveillance
Although surveillance endoscopy has been used to monitor the extent and progression of BE, it remains unclear whether surveillance
endoscopy leads to an improvement in outcomes such as life expectancy . Although some small studies have demonstrated that endoscopic surveillance can detect earlier and curable EACs compared
with patients who present with symptoms of EAC, and lead to an improvement in 5-year survival, it has been calculated that
surveillance endoscopy will only add to the quality of life of individuals in whom the EAC incidence in BE is greater than
Furthermore, evidence from a recent cohort study demonstrated that of 335 (< 20%) BE patients initially entered into an endoscopic surveillance program, only 52/335 would have ultimately benefitted from the program . Moreover, one recent study of 95 BE patients found that a 'transient diagnosis' of dysplasia (i.e. false positive dysplasia) was made in 20% of the cohort, and suggested that transient positive diagnoses
of dysplasia account for between 28 and 61% of endoscopies in BE surveillance programs.
Further support to the limitations of surveillance was offered by another study involving annual endoscopy in 143 BE patients
where only one patient with EAC was detected over a mean follow up of 4.4 years, and this patient died of the resultant esophageal
surgery. The authors concluded that current surveillance strategies were of limited value and that it might be more appropriate
to further restrict surveillance endoscopy to patients with additional risk factors (e.g. strictures, ulcers, or a > 8 cm long BE segment).
Cost considerations should also be considered in the surveillance equation: it was estimated that a US population-wide surveillance
program could potentially cost approximately $290 million dollars, and that outpatient management of BE patients would cost
$1241 dollars/year (of which half would be from medication use) .
Chemoprevention of Barrett's esopahgus
Increased epidemiological evidence suggests that non-steroidal anti-inflammatory drugs may have a protective role against
EC, and by extension BE. One epidemiological study of > 600 000 Americans with a 6-years follow-up demonstrated that regular aspirin use was associated with a decreased mortality risk
[36,37]. In addition, exposure of BE epithelia to acid and bile has been demonstrated to induce the expression of cyclo-oxygenase-2
(COX-2) , and COX-2 expression occurs at higher levels in dysplastic BE [38,39]. Prospective phase II chemoprevention studies are underway to examine the effect of COX-2 inhibitors in BE patients with
Medical reduction of acid load
Reducing acid load in patients with BE can be accomplished by either inhibiting acid secretion, increasing LESP, enhancing
acid clearance from the esophagus, or diverting acid away from the esophagus (see Management/Surgery).
Long-term acid suppression with PPI is more effective than with histamine-2 receptor antagonists in healing esophagitis and
preventing esophageal strictures, especially in SSBE . Symptom relief is not an effective surrogate marker for complete elimination of acid reflux, and thus is a poor gauge for
titrating acid-suppressive therapy. Moreover, some studies that have examined the effect of acid on cellular proliferation
indices and found that even small levels of acid exposure lead to increased cellular proliferation in BE compared with BE
exposed to no acid. One inference drawn from these studies advocates complete acid elimination, and supports the use of a
histamine-2 receptor antagonist at night to abolish nocturnal acid levels, in addition to a twice-daily PPI regimen.
What remains more controversial is whether acid suppression can delay the progression, or even result in the regression of
BE and/or dysplasia. Although some studies demonstrated only partial or complete regression in a few study subjects (in spite of
adequate documented acid control with a 24-h pH monitoring), most studies do not demonstrate that acid suppression can lead
to regression of BE. Nevertheless, data from these studies did suggest that acid suppression-related regression was more likely
to occur in SSBE than in LSBE. Furthermore, the absence of a hiatal hernia was noted to be the most important factor associated
with BE regression. In summary, complete regression of BE in response to acid suppressive therapy occurs in the minority of
BE patients, primarily those SSBE and no hiatal hernia.
Although promotility agents have been used in an effort to enhance esophageal acid clearance, their utility in the management
of BE has been disappointing. It has been suggested that because presence of HP has been postulated to decrease gastric acidity
and thereby minimize acid reflux, that it should not be treated (see Pathogenesis/Helicobacter pylori). Currently, routine screening for HP in patients with GERD or BE is not recommended.
In summary, despite some of the apparent benefits associated with reduction in esophageal acid load, no method has conclusively
been proven to decrease the risk of development of EAC.
Endoscopic therapy for Barrett's esophagus
Endoscopic techniques include thermal methods such as monopolar/bipolar cautery, heater probe, argon plasma coagulation, and laser therapy (Argon, Potassium Titanyl Phosphate (KTP), and
Neodymium-yttrium aluminum garnet (Nd:YAG)). Photodynamic therapy (PDT) and endoscopic mucosal resection (EMR) are two other
methods which may be used to treat dysplastic BE.
It is vitally important to contextualize the role of endoscopic therapy within the broader spectrum of therapeutic techniques
that are available, especially since no endoscopic therapy has been demonstrated to decrease the long-term risk of EAC. In
addition, no long-term data is available regarding the durability of dysplastic/non-dysplastic BE ablation/removal (whether the regenerated squamous epithelium is stable over longer follow-up periods). Moreover, even if successful
endoscopic therapy has been performed, the optimal method for preventing the return of BE has not yet been determined. In
addition, persistence of SIM underneath re-epithelialized squamous mucosa following endoscopic ablative therapies remains
a concern of unclear significance .
In summary, although endoscopic ablative therapies can be used to eliminate non-dysplastic BE, the most appropriate use of
endoscopic ablative therapies remains in those patients with BE/HGD who are deemed suboptimal candidates for esophageal surgery, and those who refuse surgery , but it should also be considered in those patients who are well-staged operative candidates. Because of the uncertainties
in long-term outcome associated with such endoscopic therapies, any use of such therapy for dysplastic BE necessitates both
careful and long-term follow-up.
Thermal ablation: Monopolar/bipolar/heater probe electrocoagulation
Monopolar, bipolar/multipolar (MPEC), and heater probes work by coagulating BE epithelium. Slight physical compression is essential to ensure
optimal effectiveness of this modality. For small lesions, a BICAP probe can be applied directly to the BE lesion typically
using between 30 and 50 W power (Fig. 7). Alternatively, for larger lesions, bipolar electro-coagulation can be accomplished using an olive-shaped BICAP tumor probe
(introduced under fluoroscopic guidance), that can deliver thermal energy to the esophageal mucosa. The probe comes in a variety
of diameters (615 mm) and delivers energy in a circumferential field. The procedure can be performed under endoscopic guidance by introducing
a small caliber endoscope proximally. The BICAP tumor probe offers the advantages of being able to treat circumferential lesions
quickly as well as being both cheap and portable. Nevertheless, the BICAP tumor probe is used primarily for debulking tumors
rather than for eradication of mucosal disease, and is used rarely these days. Complications include transient chest pain,
fever, leucocytosis, worsening of dysphagia from the resultant edema, and occasionally perforation.
Although multipolar electrocoagulation (in conjunction with acid suppression) has been used to treat non-dysplastic BE with
results varying from complete to partial regression, the technique cannot be recommended. Lack of a significant differences
between the various thermal methods suggests that for now, the specific technique chosen should be governed by local resource
availability and endoscopic proficiency with the technique, more than any other factors.
Thermal ablation: Argon Plasma coagulation (APC). Argon plasma coagulation (APC) is a relatively new non-contact non-coaptive thermal technique that transmits high-frequency
monopolar current via a probe emitting ionized, electrically conductive, argon plasma gas. Typical gas flow rates are 0.52.0 l/min with a power output of 4090 W. The technique allows tangential tissue ablation and limits the maximum coagulation depth achievable to approximately 23 mm, thereby minimizing the possibility of transmural injury. APC is not the optimal approach for BE lesions that are raised
or nodular because of the limited depth of tissue penetration. Although theoretically a non-contact method, the probe often
contacts the tissue because of patient movement.
The many advantages of APC over laser therapy include low equipment costs, portability, the relative absence of the adhesion
effect between the probe and coagulum, as well as the ability to deliver thermal energy (both en face and from the side).
APC (in conjunction with acid suppression) has also been used to treat non-dysplastic BE (Fig. 8). Ablation rates of BE using APC vary from 61 to 98.6% over follow-up periods of up to one year . A prospective study examined the efficacy of APC ablation (in combination with high-dose acid suppression) in 73 patients
with non-dysplastic BE over a median follow-up period of 12 months . The median length of BE was 4 cm (range 112 cm), and up to 4 cm segments were treated in a single session.
Of the 70 patients who completed treatment, complete squamous re-epithelialization was reported in 98.6% (69/70) after a median of 6 weeks (range 124 weeks), and the remaining one patient had his 8 cm BE segment reduced to 2 cm after 3 sessions (a residual segment that persisted despite 1 year of intense acid-suppressive therapy). Over a median
follow-up period of 12 months (range 251 months), none of the other 69 patients showed evidence of relapse of BE or evidence of the development of dysplasia. The
reported stricture rate was 4.3% (3/70), all of which responded well to a single bougie dilation.
In addition, cumulative long-term results (13 years) are available for 188 patients from 5 studies where APC was used to treat BE with/without dysplasia . Although the patient populations were not homogenous, ablation rates were reported as high, BE recurrence rates after ablation
varied from 3 to 62%, and one of the studies reported a complication rate of over 50%. The outcome encompassed a range from
complete reversal, through residual visible BE, to residual underlying SIM, with at least some of the variability in outcome
being attributed to the wattage used (90 W was deemed as optimal), and the degree of acid suppression.
Preliminary studies of APC in dysplastic BE patients with purely HGD have been reported: 10 BE patients with HGD or early
EAC were treated with high dose acid suppression and APC until the original lesion was eradicated or complete squamous re-epithelialization
had occurred, and reported complete eradication in 8/10 patients after a mean number of 3.3 APC sessions. In addition these 8 patients showed no evidence of local recurrence during
a median follow-up of 24 months. One patient with initial HGD had persistence of HGD 30 months after initial diagnosis, and
one patient progressed to invasive adenocarcinoma after failure of APC and PDT .
Of concern is the fact that EAC has been reported as arising from within squamous epithelium 18 months after APC ablation,
an observation that suggests SIM may have persisted underneath the re-epithelialized squamous mucosa. Certainly more RCTs
are needed to more fully assess the effect of APC as an ablative technique for removal of dysplastic and non-dysplastic BE:
until then its use should be limited to clinical trials.
Thermal ablation: Endoscopic laser therapy (ELT). Endoscopic laser therapy (ELT) is typically performed with a Potassium Titanyl Phosphate (KTP) or neodymium-yttrium-aluminum-garnet
(Nd:YAG) laser. The Nd:YAG laser can penetrate normal gastrointestinal mucosa up to a depth of 12 mm. ELT was most popular in the 1980s when it was used primarily for the palliation of EC, but more recently its use has become
ELT can be used in the obliteration of BE/HGD but results have been variable, and overall numbers treated have been small. In most cases, although the BE and tumor
were effectively ablated, subsequent follow-up often showed tumor recurrence [47,48].
One study of 17 BE patients, all of whom had undergone successful antireflux surgery, ablated the BE with laser in 11/17, and followed 6/17 as untreated controls. Complete squamous re-epithelialization of the esophagus was reported in all 11 of the laser treated
group over a 26-month follow-up, but no change in the length or quality of the SIM was seen in the 6 controls .
KTP laser was used in a study of 10 BE patients (4/10 LGD, 4/10 HGD, and 2/10 early EAC) for a mean of 2.4 sessions to ablate the dysplastic/cancerous mucosa. Complete ablation was seen in all 10 patients and squamous re-epithelialization was induced in combination
with high-dose acid suppression therapy over a 10.6-month mean follow-up period. No complications were reported .
In conclusion, ELT is beset by a number of disadvantages (e.g. high costs, lack of portability, and a higher rate of complications)
which make its use less common. More generally, all thermal techniques have the potential to destroy the BE. However, treatment
is over a small area at a time, and one cannot assess the uniformity or depth of damage at the time of the procedure. Moreover,
the techniques tend to be labor intensive. And again, persistence of SIM underneath re-epithelialized squamous mucosa remains
a significant problem: one study reported this occurrence in 2/10 patients treated with a KTP laser .
Photodynamic therapy (PDT)
Although Photodynamic therapy (PDT) has been around since 1960, regulatory approval in the mid 1990s permitted a major increase
in its use primarily as an ablative therapy for dysplastic BE . The two major photosensitizers used are porfimer sodium (Photofrin®, Axcan Pharma, Montreal, Canada), administered intravenously 48 h before endoscopy, and ä-aminolevulinic acid (ALA, marketed as Levulan®), administered orally 46 h before endoscopy. ALA is a pro-drug that is converted to the endogenous photosensitizer protoporphyrin IX in vivo at a dose of 60 mg/kilogram.
In each case, the lesion is illuminated using a red laser light transmitted through a fibreoptic probe that may be used either
free floating or with a balloon-centreing device. In cases where the lesion is not well defined, the use of adjunctive endoscopic
techniques (e.g. chromoendoscopy, LIF) may be helpful in defining the extent of the lesion. A portable diode (630 nm) laser, supplying 4002000 mW (using 1050 mm fibers), and running on conventional electrical power, is now available for performing PDT (Diomed Ltd, Cambridge, UK).
The light activates the drug (which has a degree of selectivity for premalignant and malignant tissue), to produce toxic free
radicals, an effect that ultimately leads to an obliterative endarteritis of the microvasculature, cytotoxicity and resultant
necrosis. Levulan® appears to be more selective for tumor tissue with a tumor:benign tissue ratio of 6:1 compared with 3:1 for Photofrin®.
Reassessment should be carried out 48 h later to assess the therapeutic effect, and if Photofrin®-PDT has been used, further therapy can be performed on areas that have not fully responded to the treatment. Typically patients
may complain of moderate chest pain, and sometimes anorexia and nausea, and oral liquid analgesia should be routinely prescribed.
The therapy is usually performed as an outpatient service. Patients should be limited to liquids for the first 2 days and
then the diet can be stepped up as tolerated, with most able to accommodate a full diet by 57 days post-treatment.
The main disadvantages associated with PDT include the high drug and equipment costs, as well as the photosensitivity that
occurs from the residual drug in the skin. Photofrin® light sensitivity can last from 4 to 6 weeks and the patient should strictly avoid direct sunlight for the first 2 weeks.
Indirect sunlight or room light is fine and should not cause any problems. Light sensitivity is much less of an issue with
Levulan®-PDT since the drug remains in the body for only approximately 4872 h.
Complications resulting from the use of PDT include esophageal strictures occurring in 1050% patients, transient fever and pleural effusion. Fistulas and perforations are rare. Typically 24 dilations are required to treat these strictures but occasionally they can be much more refractory to therapy. A recent
study compared 60 patients randomized to PDT or PDT/oral prednisone and found no significant difference in stricture formation between the two groups.
However, balloon length used was identified as a possible risk factor especially in long segments where there was overlap:
the stricture rate was 31% with a 7-cm balloon compared with 7% when a 5-cm balloon was used. Although esophageal strictures
are generally not a problem with Levulan®-PDT (likely due to less deep tissue injury), significant nausea and hypotension may occur following administration of the
One PDT trial randomized BE patients with only LGD to two groups, one receiving placebo and the other receiving ALA, and reported
an 89% response rate, and 30% decrease in the BE surface area in the ALA-treated patients, compared to a 11% partial response
rate, and 10% decrease in the BE surface area in 11% of the placebo group. More interestingly, 33% of the placebo group had
no evidence of LGD on follow-up biopsy .
The largest series employing PDT for BE/HGD, and/or superficial EC involved 100 BE patients using Photofrin®. ELT was subsequently used to treat any residual dysplastic or cancerous mucosa. Endoscopic follow-up 19 months after PDT
demonstrated eradication of 10/13 ECs, 88% of BE/HGD subjects and 92% of BE/LGD subjects, with a stricture rate of 34% .
A further study examined the use of PDT (60 mg/kg ALA) in the treatment of 10 patients with BE/HGD and 22 patients with superficial mucosal EC, and reported eradication of dysplasia in 100% (10/10) BE/HGD subjects and 77% (17/22) of the EC subjects after a median follow-up of 9.9 months (range 130 months) .
BE may still persist in a fair number of subjects after PDT, and may be treated either with repeat PDT at a lower energy level
(50100 J/cm), or indeed with another endoscopic ablative therapy (e.g. APC, MPEC, Nd:YAG laser) to completely ablate any residual BE
tissue. Employing adjunctive diagnostic techniques (e.g. chromoendoscopy, EUS) may aid in targeting suspicious areas postablation,
but this approach is unproven as yet. Of potential concern, if one considers using LIF or FE to detect residual dysplasia,
is the discovery that residual Photofrin®-related fluorescence (detected in persisting areas of BE) may interfere with the fluorescence pattern .
In summary, although PDT is a promising therapy, the lack of long-term follow-up, and high stricture rate, limits its current
application to those BE/HGD patients eligible, but unsuitable for surgical esophagectomy, as well as increasingly, in otherwise operative candidates.
Endoscopic mucosal resection (EMR)
Endoscopic mucosal resection (EMR) is a technique employed to obtain tissue, or remove focal lesions not previously amenable
to standard biopsy or excisional techniques. Although its use in dysplastic BE has not been fully defined, it is currently
used to remove dysplastic BE confined to the level of the esophageal mucosa [56,57]. While endoscopic ultrasound remains a critical staging tool that should be employed before EMR is contemplated, in theory
a well-orientd EMR specimen should theoretically offer more complete and definitive information as to the stage of the dysplastic
A variety of EMR techniques have been described . Most EMR techniques involve lifting the esophageal mucosa by a submucosal injection of saline or epinephrine/saline, sometimes in conjunction with a dye (e.g. methylene blue), and then resecting the mucosal layer down to the submucosa
[56,58](Fig. 9). The use of other solutions that may diffuse less rapidly has also been described (e.g. hypertonic saline, 50% dextrose,
and sodium hyaluronate). Between 2 and 30 ml may be initially injected using a standard injection needle inserted at an oblique angle. Failure to lift the lesion may
be due to incorrect positioning of the needle or may indicate deeper tissue involvement. The border should be marked with
electrocautery leaving at least a 2-mm border. Once the tissue is lifted, a specialized snare, needle knife excision or a
suction technique may be employed to resect the lesion.
The 'lift and cut' EMR technique uses a double channel endoscope through which the lesion is lifted with a forceps passed down one channel,
and then grasped, and resected by a snare passed down the second channel.
Cap-assisted EMR involves sucking the lesion into a straight/oblique transparent rimmed cap (Olympus America Inc., Melville, New York). Once the lesion is raised, a snare (SD-221 L-25, Olympus America Inc., Melville, New York) is passed down the biopsy channel and carefully positioned within the distal
rim of the cap. The lesion is then sucked into the cap and the snare tightened around the lesion. The snare is then advanced
out of the cap, the extent of the entrapped mucosa visually assessed, and then the piece resected using a blended current.
In addition, a technique employing a rubber band ligator (Steigman-Goff Endoscopic Ligator Kit, Bard Interventional Products,
Tewksbury, MA, USA; Sumitomo Bakelite Co., Akita, Japan) to create a pseudopolyp in the area of the lesion that can then be
snared off, has been described.
A recent ASGE Technology Status Evaluation Report suggested that EMR be limited to the following esophageal neoplasms: (i)
neoplasms with a diameter less than 2 cm (ii) involvement of < one-third of the esophageal wall circumference, and (iii) lesions limitated to the esophageal mucosa
on EUS. EMR offers the additional advantage over ablative therapies of permitting histological examination of the entire resected
specimen for staging.
Complications associated with EMR of esophageal lesions include hemorrhage (3.620%) and perforation (0.5%) . Because up to 15% patients treated with EMR may subsequently develop metachronous cancer, long-term endoscopic follow-up
is mandated in this group of patients. Some work, directed at examining the feasibility of circumferential EMR in the pig
model, has been performed with promising results.
Nevertheless given the relative paucity of data, the use of EMR should currently be limited to three indications: (i) SSBE
patients with a focal resectable area of HGD within a field of BE or BE/LGD, or (ii) removal of discrete nodules of LGD/HGD before application of a more uniform ablative therapy, and (iii) BE patients unsuitable for surgery, until prospective
comparative trial data (EMR vs. surgery) becomes available. Furthermore, patients undergoing EMR should undergo endoscopic
surveillance every 36 months for the first year, and then every 13 years for at least a period of five years.
Surgery for Barrett's esophagus
Anti-reflux surgery that diverts acid from the esophagus by creating an anatomic barrier, is highly effective at eliminating
the symptoms and signs of GERD. Furthermore, it has been suggested that surgical fundoplication might confer the additional
benefit associated with limiting pancreaticobiliary reflux into the esophagus, since bile has also been implicated in the
pathogenesis of BE. Nonetheless, no conclusive long-term data demonstrates that effective antireflux surgery leads to a reduction
in the development of dysplastic BE or its progression to EAC (partly because of the large number of patients needed to power
such a trial [60,61]).
Moreover, one study examining the efficacy of antireflux surgery in patients with BE demonstrated failure rates of 56% in
uncomplicated BE patients and 64% in complicated BE patients over 100 months follow-up [62,62]. In addition it is still unclear whether a surgical approach is capable of leading to significant cost savings . Although several small studies support the possibility that antireflux surgery might prevent the development of dysplasia
and EAC in patients with BE, surveillance should be continued after surgery because of the continuing risk of EAC (especially
in the first 34 years post surgery), and the lack of long-term outcome data.
A variety of antireflux operations are effective in the treatment of GERD. The classic technique is the Nissen fundoplication,
which involves the creation of a 360-degree wrap of the lower end of the esophagus with the gastric fundus. Postoperatively,
gas bloat and dysphagia, probably resulting from the hyper-competent lower esophageal sphincter produced by the Nissen are
Accordingly, various modifications of antireflux surgery (e.g. floppy Nissen, Toupet, Rosetti, Dor, the Hill operation, and
the Belsey Mark IV) of the Nissen procedure have had the aim of either minimizing these complications, making the surgery
easier, or addressing special situations .
The floppy Nissen involves both shortening and loosening of the fundoplication and markedly decreases the incidence of postoperative
dysphagia and gas bloat syndrome. The Dor procedure is an anterior partial fundoplication (the esophagus is wrapped from the
front and the edges of the stomach are attached to the posterior wall of the esophagus), and was initially designed as an
antireflux operation for patients who had previously undergone a Heller's myotomy for achalasia. In the Toupet procedure,
the esophagus is wrapped from behind and then the edges of the stomach are attached to the anterior wall of the esophagus,
rather than to each other, leaving a space in between: this approach is the procedure of choice for GERD patients with poor/absent peristalsis .
Since the late 1980s, laparoscopic or minimally invasive fundoplication has become the standard surgical method of treating
GERD and has yielded comparable, if not better results to the open cases. Laparoscopic surgery offers several advantages over
open surgery: surgical morbidity is lower, hospital stay is shorter than with open techniques, and re-operation can be performed
earlier and more easily . In addition, endoscopic manipulation of the GEJ to prevent acid reflux (the endoscopic full-thickness plication system (NDO
Surgical, Inc, MA, USA), Endocinch (Bard, USA), endoscopic injection of the polymer Enteryx (Enteric Medical Technologies,
USA), and the Stretta procedure (Curon Medical Inc, California, USA)), all currently a source of intensive preliminary investigation,
have no real long-term follow-up, and it will probably be some time before more definitive outcome data becomes available.
The Endocinch (Bard) and Stretta (Curon) were approved by FDA in April, 2000.
Surgical esophagectomy is the definitive treatment of choice for BE/HGD and offers the best chance of long-term cure. Nevertheless the operation carries an operative mortality of 312% and a morbidity of up to 50%. (See Esophageal cancer/Management/Surgery for a full description of the technique.)
Barrett's esophagus conclusion
Many uncertainties continue to permeate our knowledge and treatment of BE. For example, neither medical nor surgical techniques
of acid reduction in GERD have been convincingly demonstrated to reduce the rate of development of EAC. Controversy still
exists as to the degree of acid suppression required to ultimately influence the development of EAC. Until more data proves
otherwise, the most reasonable course would be to titrate the level of acid suppression to eliminate symptoms and endoscopic
signs of acid reflux. Moreover, no definite proof exists that confirms the efficacy or screening or surveillance for BE. Nevertheless,
screening should be considered especially in Caucasian males with chronic GERD symptoms > 5 years. BE surveillance should
be performed in BE patients every 23 years according to the ACG guidelines and more frequently if dysplasia is present. The role of intensive endoscopic surveillance
and endoscopic ablative therapies has yet to be fully defined. Either is a reasonable approach in patients unsuitable or unwilling
to undergo surgical esophagectomy. Although surgical esophagectomy is the current 'gold standard', endoscopic methods such as PDT in patients well staged by EUS or by EMR, will, in the future, likely replace surgical treatment
of mucosal disease.
Although EC accounts for 1% of all cancers, it surprisingly represents 12% of all cancer deaths. Marked variation in EC rates
exist by sex, ethnic group, and by geographic variation. The two most common types of EC are esophageal squamous cell carcinoma
(ESCC) and esophageal adenocarcinoma (EAC). Its often-late clinical presentation, combined with a lack of anatomic barriers
to early spread, means that both direct extension into the mediastinum, and spread to lymph nodes occurs early. Prognosis
is uniformly very poor once the tumor has extended beyond the esophageal wall, with 5 years survival rates of 10% for Caucasians,
and 6% for Blacks. Unfortunately, the early stages of the disease are only found by chance, by screening of precursor lesions
(e.g. BE), or through screening programs in endemic geographic areas [66,67].
Risk factors for esophageal cancer
The main risk factors for ESCC esophagus are alcohol (x 27 relative risk), and tobacco use (x 2.4 relative risk), with these two factors acting independently . Moderate smokers have up to a five-fold increase in the incidence of ESCC, while the risk appears only doubled for EAC . Alcohol and tobacco use appear to be less important risk factors in the development of EAC. Racial genetic differences in
the susceptibility to the effects of alcohol and tobacco in ESCC have been implicated as accounting for the differing incidence
between blacks and caucasians [66,67].
Another risk factor associated with an increased risk of EC is lower socio-economic status (independently of tobacco and alcohol
use): this may reflect inadequate diet. PlummerVinson syndrome (associated with iron and vitamin deficiency), chronic esophagitis, celiac disease, chronic lye strictures,
achalasia, esophageal diverticula, partial gastrectomy, familial tylosis, and hot beverages, are all associated with an increased
risk of EC . Factors specifically associated with ESCC include head and neck squamous cell carcinoma, malnutrition, and human papillomavirus
infection [66,67]. In addition, selective nutritional deficiencies may account for some of the regional geographic variations in the incidence
and prevalence of ESCC. Deficiencies in zinc, molybdenum, magnesium, iron, selenium, alpha and beta-carotene, retinol, alpha-tocopherol,
ascorbic acid, riboflavin, niacin, and thiamine have all been implicated [36,37,68].
As a corollary to the relationship between nutritional deficiencies and EC, diets that are high in vitamin C, carotenoids,
fruits, and vegetables are possibly associated with a decreased risk of EC. In addition, regular use of aspirin (16 aspirin/month >= 1 years) has been associated with a decreased risk of death from EC. A recent epidemiological study of > 600 000 Americans with a 6-years follow-up demonstrated that regular aspirin use was associated with a decreased mortality risk
[36,37]. Currently, prospective phase II chemoprevention studies examining the use of COX-2 inhibitors in BE patients with LGD/HGD are underway to examine more fully the implications of these findings.
Increasing obesity, and smoking, both perhaps acting by increasing the development of chronic GERD have been identified as
risk factors specifically associated with EAC. Obese adults (body mass index 30 kg/m2) have an increased odds ratio of 16.2 compared with the leanest adults (body mass index < 22 kg/m2) for developing EAC. Moreover the frequency, severity, and duration of GERD symptoms are all positively associated with an
increased risk of EAC (see various sections above: Barrett's Esophagus/Risk Factors, Epidemiology, Natural History, Pathogenesis). The declining incidence of HP infection, acting through decreased
acid production, has been suggested as being one factor accounting for the increased incidence of EAC recently (see Barrett's esophagus/Pathogenesis/Helicobacter pylori). In addition, past use of medications that lower the lower esophageal sphincter (LES) pressure (e.g. anticholinergics) has
also been associated with an increased risk of EAC.
Epidemiology of esophageal cancer
EC is 35 × more common in men than women in the USA, although the worldwide ratios vary from 2 to 20 fold, and is approximately 3 × more common in African-Americans than caucasians in the United States. Worldwide, the highest incidence rates in caucasian
men occurs in Calvados, France (26.5/100 000) while the lowest incidence occurs in Israeli Jews (0.6/100 000). For women, the highest incidence occurs in Bangalore, India (8.8/100 000), while the lowest incidence is amongst the Japanese living in Los Angeles, USA (0.1/100 000) .
ESCC incidence varies considerably around the world with the highest incidences occurring in Linxian County, Peoples Republic
of China, the Caspian region of Iran, and the Transkei region of the Republic of South Africa. Incidence rates in these areas
are between 50 and 100/100 000 populationwell above the incidence rate in North America of 5/100 000. Although the worldwide incidence of ESCC is declining, it is still 6 × more likely to occur in black males than white males .
In the 1960s, ESCC accounted for more than 90% of all ECs but over the past 20 years, the incidence of EAC has risen markedly
so that EAC now accounts for approximately half of the 13 000 occurring in the USA annually. However, the incident of EAC is much less than that of colon cancer. In fact EAC represents
one of the fastest growing cancers in the developed world. In contrast to ESCC, EAC occurs predominantly in middle-aged to
elderly male Caucasians (see Barrett's esophagus/Epidemiology) .
Natural history of esophageal cancer
The natural history of ESCC is divided into an initial phase (~20 years, characterized by basal cell hyperplasia and increasing degrees of dysplasia), a developing phase (~36 years, characterized by the development of ESCC from within the dysplastic tissue), an asymptomatic phase (~1 years), and an overt phase (~less than 1 year). The presence of dysplasia as opposed to hyperplasia, increases the risk of the development of ESCC nearly
15-fold. Unfortunately, most cases of ESCC are only discovered in the overt symptomatic phase, and the mean survival in this
set of patients is approximately 10 months.
The natural history of EAC is similar to ESCC. Most EACs present in the context of BE, a premalignant complication of chronic
GERD (see Barrett's esophagus/Natural history above). Because so few BE patients seek attention for GERD symptoms (often because of decreased pain sensitivity),
early cancers are often missed. EAC may present as a flat, ulcerated, or polypoid/fungating lesion, and lymph node metastases occur early and extensively.
Diagnosis of esophageal cancer
EC typically presents late, with symptoms of dysphagia, profound weight loss, and cachexia. Other symptoms that may occur
are slow food passage, retrosternal discomfort, odynophagia, hematemesis, or coughing from recurrent aspiration. In addition,
EC may present with symptomatic manifestations of extra-esophageal disease (e.g. hoarseness from recurrent laryngeal nerve
involvement, back pain, and skin metastases). Dysphagia is usually progressive and initially intermittent limited to liquids
but gradually progresses to solid and more continuous dysphagia as tumor bulk increases. Local tumor extension into the tracheo-bronchial
tree may give rise to esophago-pulmonary fistulas in up to 5% of patients with a median survival with this presentation of
between 1 and 4 months. Signs are typically non-specific and include profound weight loss, cachexia, local skin involvement
Although there is no evidence to suggest that screening regimens would result in decreased mortality rates in the US population,
screening may be more effective in areas of high incidence (e.g. China), or in high-risk conditions (see sections on Barrett's
Esophagus/Diagnosis, Barrett's esophagus/Management/Surveillance above) .
Although microcytic anaemia may result from bleeding, and malnutrition be more pronounced because of dysphagia in EC, most
other laboratory abnormalities typically reflect the non-specific systemic effects of cancer (e.g. hypoalbuminemia, low cholesterol,
and low leukocyte count). In addition, hypercalcemia may result from bony metastases or through the effects of humoral factors.
Liver metastases may result in abnormalities in liver transaminases, alkaline phosphatase or bilirubin.
A barium study remains the most common initial investigation ordered by primary care physicians for the evaluation of dysphagia,
and may demonstrate a variety of findings, ranging from subtle mucosal irregularities, through ulcerative lesions, to grossly
polypoid infiltrative lesions. Although any radiographic abnormality ultimately requires endoscopic evaluation, the barium
esophagram is essential to provide a 'lay of the land' before performing the endoscopy. Endoscopy should be performed even if the barium esophagram is negative, given the test's
low negative predictive value in this context.
A diagnostic chest X-ray should be performed to look for extra-esophageal spread (e.g. adenopathy, pulmonary and bony metastases,
aspiration pneumonia, fistulas, abscesses and perforation). In addition, the chest X-ray may demonstrate a widened mediastinum,
tracheal displacement, or air-fluid levels within the esophagus. Although computed tomography (C.T) of the neck and chest
can be used in the initial diagnostic work-up of dysphagia (especially in the context of a neck mass), its use is more appropriate
as a staging investigation. (See Esophageal cancer/Diagnosis/Staging for staging radiology below).
Endoscopy with biopsy remains the definitive diagnostic test in EC (see Barrett's esophagus/Diagnosis above), with the endoscopic picture aiding in determining the extent of local disease, and guiding further management.
The endoscopy can determine whether a lesion is partially, or totally obstructive, as well as delineating the location of
the proximal and distal margins of the lesion. The distance of the cancer from the cricopharyngeus muscle should be carefully
determined as this information is vital when surgical resection remains a possibility . If the lesion appears local, and limited to the submucosa, local endoscopic excision or ablation may be possible: if the
lesion is exophytic, extensive, or involves a fistula, palliative methods are more appropriate.
Multiple biopsies should be taken from non-necrotic areas to improve diagnostic yield: a recent study recommends that at least
seven biopsies should be taken. Occasionally, submucosal spread of the tumor results in non-diagnostic biopsies initially,
in which case repeat biopsies should be obtained .
In addition, chromoendoscopy with Lugol's iodine or Toluidine blue can be helpful in directing endoscopic biopsy of small
and subtle lesions, in delineating areas of interest for endoscopic ablation, endoscopic resection or definitive surgery,
and finally in the detection of multifocal areas of cancer (see Barrett's esophagus/Novel diagnostic techniques/Chromoendoscopy, magnification endoscopy above).
Brush cytology, incorporating a balloon encased with a mesh net has also proven to be a useful adjunctive technique, especially
in the diagnosis of ESCC in China. The sensitivity, specificity, and accuracy of brush cytology in China are 1436%, 9099%, ~ 90%, respectively. In other parts of the world, these results have been less spectacular. Although brush cytology is a useful
technique in high endemic areas, the efficacy of surveillance cytology or endoscopy for patients with other high-risk conditions
(e.g. tylosis, long-standing achalasia) remains unknown at present.
Once the diagnosis of EC is confirmed, accurate staging using the Tumor, Node, Metastases (TNM) classification remains a critical
part of optimizing future management strategy (Figs 11 and 12). Staging investigations initially include a C.T. of the neck, chest and liver to look for extra-esophageal spread . If there is no evidence of advanced local or metastatic disease on CT, endoscopic ultrasonography (EUS) should be performed
to more accurately delineate the extent of local mediastinal and lymph node involvement [71,72]. Moreover, in an attempt to improve the accuracy and precision associated with staging early EC, a subclassification of depth
of involvement of superficial EC has been developed (Fig. 13).
EUS is useful in both the pre- and post-treatment staging of EC. Moreover, EUS is the most accurate non-operative imaging
modality for T/N staging in EC with T accuracy varying from 59 to 92%, and N accuracy from 55 to 90% . In addition, EUS T/N staging can accurately predict survival prior to any surgical therapy. EUS can identify patients having early stage disease
(i.e. T1N0, T1N1) who would not likely benefit from the addition of neoadjuvant therapy to surgery alone, as well as identifying
those patients with advanced T4 disease (who would probably have equal survival whether treated aggressively or palliatively).
One recent prospective study of 55 subjects examined whether pretreatment staging with EUS could identify which subjects would
have a positive response to neoadjuvant chemoradiation. The study found that there was no significant difference among the
EUS-staged groups in the frequency of complete tumor response (CR) to neoadjuvant chemoradiation: T2 subjects had a 60% CR,
T3 subjects a 33% CR, and T4 subjects a 63% CR (P = 0.19). Interpretation of these findings is problematic: they might just reflect the efficacy of neoadjuvant chemoradiation
rather than the sensitivity of EUS-based criteria. Nevertheless, most clinical trials comparing neoadjuvant therapy to neoadjuvant
therapy and surgery demonstrate no significant survival differences.
The results of neoadjuvant protocols employing novel agents (e.g. paclitaxel, irinotecan, gemcitabine, flavopiridol, fludeoxyglucose
F 18, docetaxel, and carboplatin) are awaited to see if they demonstrate a significant improvement in the efficacy of neoadjuvant
therapy. Only then will the role of EUS in treatment selection or disease outcomes be more fully established.
The synchronicity of aerodigestive tumors may sometimes indicate the need for bronchoscopy in tumor staging to exclude tracheo-bronchial
involvement. In addition, in scenarios where curative surgery exists as a possibility, staging with conventional or minimally
invasive thoracoscopy and laparoscopy can help more precisely determine the N stage before resection is considered. Not only
can the whole esophagus be visually inspected by thoracoscopy/laparoscopy, but also periesophageal lymph nodes may be sampled, and occult pleural/pulmonary metastases be readily identified. Laparoscopy examines the liver, peritoneum, gastric wall and diaphragm and can
also allow biopsy of celiac and perigastric nodes.
In addition, evolving techniques (e.g. positron emission tomography (PET) scanning, and use of radio-labeled antibodies) may
prove to be potentially useful additional methods of staging, with PET scanning potentially being able to delineate lymph
node involvement more accurately and non-invasively, and perhaps altering management from a surgical approach to one involving
Management of esophageal cancer
Ultimately, the management of EC depends primarily on whether the disease is curative or palliative. It is surprising that
although surgery has formed the cornerstone of therapy for EC, there is very little data supporting a preferential survival
advantage in choosing surgery over radiation therapy/chemoradiation for those patients who have locally advanced disease.
For those few patients where the EC is detected early, curative surgery offers a reported five-year survival rate of between
5 and 15%. Unfortunately, most patients present with EC in an advanced state, and palliation of dysphagia/pain and nutritional support become the primary goals of therapy. The management of EC should be multidisciplinary in nature,
involving gastroenterologists, oncologists, surgeons, and a palliative care team.
Curative therapy with endoscopy for EC can be achieved either with ablative techniques, or by EMR [69,73]. Ablative techniques include monopolar/bipolar/heater probe electro-coagulation, endoscopic laser therapy (ELT), argon plasma coagulation (APC), and photodynamic therapy
(PDT). The step-wise incremental nature of bipolar electrocoagulation, APC, and ELT makes these techniques the preferred method
for removing tumor bulk, compared with PDT.
Accurate EUS is critical in the optimal selection of potentially curative cases. The depth and extent of invasion of the EC
is of great importance in deciding the appropriateness of curative endoscopic therapy. One small study of 10 subjects undergoing
surgery for a small EC within BE demonstrated that the thickness of BE mucosa varied between 0.8 and 3.3 mm, while the epithelium measured between 0.3 and 1.3 mm. Certainly, more data is needed to more adequately compare the depth of damage induced by ablative therapies with the depth
of seemingly superficial lesions.
Endoscopic palliative management of EC includes endoscopic dilation, thermal ablation of tumor tissue (i.e. bipolar electrocoagulation,
argon plasma coagulation (APC), endoscopic laser therapy (ELT), and photodynamic therapy (PDT)), injection techniques, the
insertion of esophageal endoprostheses, and the placement of feeding tubes [69,73].
Monopolar/bipolar/heater probe electrocoagulation.. See Barrett's esophagus/Management/Endoscopic Therapy/Thermal Ablation/Monopolar/Bipolar/Heater Probe Electrocoagulation for a description of the technique and complications. Monopolar, bipolar/multipolar, and heater probes work by coagulating malignant tissue. Although all techniques have been reported as being successful
in the elimination of tumor-associated dysphagia, limited data is available comparing the modalities. Until more comparative
data is available, current use is dictated more by resource availability and provider expertise. Complications include transient
chest pain, fever, leucocytosis, worsening of dysphagia from the resultant edema, and occasionally perforation .
Endoscopic laser therapy (ELT). See Barrett's esophagus/Management/Endoscopic Therapy/Thermal Ablation/Endoscopic Laser Therapy for a description of the technique. ELT is easiest to use for EC lesions in the mid to distal esophagus,
and although it can be used to palliatively treat both eccentric and circumferential lesions, it is best suited to treating
short, non-circumferential lesions. The laser probe is introduced through-the-scope and treatment begun in a caudal to rostral
direction. The procedure may be performed in several sessions for long or bulky tumors. In addition, ELT has been reported
as being a successful ablative modality in the treatment of tumor ingrowth associated with esophageal endoprostheses (EE).
Complications associated with ELT include all those associated with thermal therapies, including transient dysphagia, chest
pain, pleural effusions, fever and perforation (115%) .
Although ELT can restores luminal patency in up to 90% of patients, this effect translates into elimination of dysphagia in
only 7085% of cases. In one study employing the Nd:YAG laser to treat EC, dysphagia improved in 88% patients, and the mean dysphagia-free
period post ELT was 30 days. The KTP laser has also been used in the treatment of early EC: one study of 10 subjects demonstrated
complete initial eradication of all HGD/cancers without any complications. Unfortunately subsequent follow-up in this study found SIM underneath the re-epithelialized
squamous mucosa in 8/10 study subjects.
The major disadvantages associated with ELT include the high equipment costs, the lack of portability of the system, difficulty
in using the laser in angulated areas (cricopharyngeus and lower esophagus), and the not uncommon requirement for several
sessions before desired results are achieved.
Argon plasma coagulation (APC). See Barrett's esophagus/Management/Endoscopic therapy/Thermal ablation/Argon plasma coagulation (above) for a description of the technique and complications. The existence of both side and forward
viewing APC probes permits very precise targeting of the malignant tissue. Although APC has a limited role in the curative
therapy of all but the most superficial EC because of its limited depth of penetration it remains an effective therapy. One
study of APC in 83 subjects demonstrated a success rate of 84% in the palliation of malignant dysphagia (within two treatment
sessions)the perforation rate in this study was 8%. In addition, APC is particularly useful in managing malignant and hyperplastic
tissue overgrowth associated with the use of EE.
Photodynamic therapy (PDT). See Barrett's esophagus/Management/Endoscopic therapy/Photodynamic therapy (above) for a description of the technique and complications. Photodynamic therapy (PDT) can be employed
in both the cure of early EC, and the palliation of advanced EC .
A recent series of 77 subjects treated with PDT (1.52.0 mg Photofrin®/630 nm red dye laser) for advanced EC, demonstrated an improvement in dysphagia in 91% of subjects at 4 weeks, a stricture rate
of 5%, a skin photosensitivity rate of 10%, and the development of pleural effusions in 3% of subjects. Further PDT was needed
in 38% in this study, and 9% ultimately required expandable metal EE to treat the dysphagia . However, when compared to ELT, it was found that PDT alleviated dysphagia in an equivalent manner.
In contrast, an RCT compared 110 PDT subjects with 108 ELT subjects, and demonstrated a better tumor response in the PDT group
for certain tumor subtypes: tumors longer than 10 cm, tumors located within the upper third of the esophagus, and tumors located within the angulated portion of the lower esophagus.
Nevertheless, these slight differences in response have not manifested themselves as differences in survival outcomes between
the two modalities.
While early data examining the effect of using hyperbaric oxygen as a method of improving the efficacy of PDT has shown some
promising results in favor of hyperbaric oxygen, more data is required to validate these findings before such an approach
can be recommended. It should be mentioned that PDT has also been used to obliterate tumor ingrowth associated with the use
of EE but experience is limited. Finally, the use of ALA (Levulan®) is not likely to have any role in the dubulking of tumor in EC predominantly because of its limited tissue penetration.
Endoscopic mucosal resection (EMR)
See Barrett's esophagus/Management/Endoscopic therapy/Endoscopic mucosal resection (above) for a description of the technique and complications. EMR is used for early EC when the
tumor is limited to the mucosa . An additional advantage offered by EMR over ablative endoscopic therapies is that it permits histological examination of
the entire resected specimen for staging.
In one study of EMR in 64 subjects with BE/HGD (3/64) or early EC (61/64), those with low-risk intramucosal cancer (Types I, IIa, IIb, and IIc; lesions less than 20 mm; mucosal lesions; histological grade GI, G2 or HGD) had complete remission in 97% cases, while those with high-risk lesions
had complete remission in 59% cases. During the follow-up period of 12 ± 8 months (mean ± SD), recurrent or metachronous carcinomas were found in 14% .
Another study examined the efficacy of EMR in 25 subjects (15 of whom had early cancer): 22/25 subjects were free of recurrence at the end of a 42-month follow-up period.
An additional study of 25 subjects demonstrated the feasibility of EMR using the 'lift and cut' technique in 23/25 and rubber band ligation in 2/25. EMR was performed because of a nodule/polyp within BE in 11/25 subjects, and suspected superficial EC or BE/HGD in 14/25 subjects. EMR diagnosed superficial adenocarcinoma in 13/25 subjects (52%), HGD in 4/25, and confirmed lesions to be of a lower neoplastic risk in 8/25. No complications occurred due to the procedure itself .
A Japanese study used EMR to remove 25 superficial cancers from 21 subjects, either en bloc, or by piecemeal resection, and
demonstrated no recurrence after a 2-years follow-up. Bleeding necessitating endoscopic therapy occurred in 24% of cases in
A series of 56 subjects with superficial EC treated with cap-assisted EMR reported that 53 remained alive without evidence
of residual/recurrent disease after a median follow-up of 39 months. In addition, another study of 35 subjects using predominantly ligation-assisted
EMR reported a local remission rate of 97% after a mean follow-up period of 12 months .
In summary, while EMR seems to be a very effective technique in removing superficial EC, the availability of longer follow-up
data will help in further establishing its role. Currently, a recent ASGE Technology Status Evaluation Report suggests that
with the current data, EMR be limited to the following esophageal neoplasms: (i) neoplasms with a diameter less than 2 cm (ii) those neoplasms involving < one third of the esophageal wall circumference, and (iii) neoplasms limited to the esophageal
mucosa on EUS.
Perhaps the most readily available, and well-established technique for the treatment of EC is endoscopic dilation (ED). ED
can be used to effectively treat dysphagia secondary to malignant esophageal strictures in 90% of patients. In addition, ED
allows a more comprehensive evaluation of a malignant lesion where the distal margin cannot be assessed because the tumor
is impassable. ED can be performed either with SavaryGilliard dilators over a Savary wire, or with balloon dilators, introduced over a wire through-the-scope under direct endoscopic
vision. The use of fluoroscopy is recommended to ensure safe passage of the Savary wire in each technique. Complications associated
with ED include 0.2% perforation risk, chest pain, bleeding and bacteremia.
Injection therapy (EIT)
A more established, but now less commonly used technique in the palliation of malignant dysphagia is endoscopic injection
therapy (EIT). The procedure involves the direct injection of a cytotoxic agent into the tumor tissue employing an injection
catheter. Although 100% ethanol is the most commonly used agent, the successful use of other agents has been reported (e.g.
2.5% sodium morrhuate/5-fluorouracil mixture, cisplatin/epinephrine mixture). In one RCT comparing EIT with ELT, dysphagia improved in 78% of subjects treated with ethanol, and no
significant differences in efficacy or complications was found between the two groups. The extent of resultant tissue necrosis
with EIT is less predictable than the necrosis associated with the incremental thermal ablative therapies, and many patients
require further treatment sessions to sustain symptom relief. Although advantages of the technique include its universal availability
and low equipment costs, lack of standardization of the technique, and its relatively short efficacy has limited its utility
in recent years.
Esophageal endoprosthesis (EE)
The insertion of esophageal endoprostheses (EE) is another common approach to the palliative management of dysphagia secondary
to EC . Despite the apparent appeal of this approach, placement of an EE should only be undertaken after very careful scrutiny of
the clinical situation (e.g. tumor stage, length, position, and expected prognosis of the patient). Consideration of these
factors is especially important before an EE is placed for two reasons: (i) once in position, an EE should be considered permanent;
and (ii) because of the attendant risks associated with the procedure itself (e.g. stent migration, perforation, stent tumor
in/overgrowth). In order to minimize the risks associated with EE placement, tumors that can be adequately treated with thermal
ablative therapies should first be treated in that fashion for as long as possible before placement of an EE is contemplated.
Factors that may preclude successful placement of an EE include a tumor within 2 cm of the upper esophageal sphincter, and short and/or non-circumferential tumors.
EE may be plastic or metal, coated or uncoated, and the various designs offer different degrees of tensile strength, memory,
and expansive radial forces . The use of rigid plastic stents has fallen significantly since the introduction of metallic EE, partly because of their
high complication rates (associated with preplacement dilation), and partly because metallic EE are easier to use, deploy,
and tend to cause less pain. Nevertheless, the fact that plastic stents offer equally effective palliation rates means their
use should not be entirely discounted, especially when cost is a major consideration (Fig. 14).
Most available data on self-expandable metallic stents (SEMS) is retrospective, uncontrolled, and involves heterogenous groups
of patients and interventions, ultimately making comparison between studies difficult. However, most studies yield figures
for the immediate palliation of malignant dysphagia of 7896%, stent migration of 49%, tumor ingrowth of 337%, need for subsequent intervention of 378%, and for major complications (i.e. bleeding, perforation, aspiration, and fistula) of 317%.
Whether one SEMS is better than another was investigated in a RCT involving 100 subjects, randomized to an Ultraflex stent,
Flamingo Wallstent, or a Gianturco-Z stent: no difference in dysphagia improvement rates, subsequent recurrence of dysphagia
(2436%), or complication rates (1836%) was seen between the three groups .
Another RCT of 62 subjects compared covered with uncovered SEMS, and found more tumor ingrowth in the uncovered SEM arm (9/30), compared with the covered arm (1/32) This led to a significantly different endoscopic reintervention rates between uncovered/covered SEM groups (27% vs. 0%). Nevertheless, despite the large difference in subsequent endoscopic reintervention rates
between uncovered/covered SEMS, this difference did not translate into differences in survival or quality of life measures between the two groups
Data comparing the various therapeutic modalities available for treatment of EC is limited: one RCT compared ELT/External beam radiotherapy (EBR) in one arm, with SEMS in the other arm, in 39 subjects with unresectable EC. It found no
difference in improvement of dysphagia rates, or survival rates between the two groups, but did find a lower incidence of
re-stenosis, and major complications in the SEMS group compared with the ELT/EBR group. Stenting was also demonstrated to be more cost effective.
Another study examining dysphagia and quality of life outcomes, randomized 65 subjects with inoperable EC to endoscopic thermal
ablation or SEMS placement. Median survival was significantly longer in the thermal ablation group than in the SEMS group
(125 vs. 68 days), dysphagia was poorly treated in both groups, hospital stay and costs were higher in the thermal ablation
group than in the SEMS group, but SEMS subjects reported more pain and worse quality of life .
EE occlusion, and migration, are the most significant complications occurring in long-term EC survivors. EE occlusion is caused
by malignant tumor overgrowth in approximately 50% of cases and by reactive tissue hyperplasia/granulation tissue in the remainder. Long circumferential tumors offer the best chances of minimizing EE migration. EE migrate
more often distally than proximally, and the approach to migration depends on the type of stent, extent of migration, and
whether any complications have occurred as a consequence of migration. Although EE may migrate at any time, migration is more
likely to occur after chemoradiation because of tumor shrinkage. When EE have migrated into the stomach, controversy exists
in whether to adopt a 'watchful waiting approach' or whether to try to remove the stent immediately.
The availability of covered SEMS has significantly changed the management of tracheo-esophageal fistula (TEF). Untreated,
TEF carries a very poor prognosis (primarily from aspiration pneumonia), and using covered SEMS has become the treatment of
choice (Fig. 15). It is usually very easy to place the SEMS so that it covers the area of the fistula, and similar complication rates should
be expected to those of placing SEMS in non-fistulizing EC. It should be noted that placement of an SEMS does not prevent
further subsequent enlargement of the existing fistula or indeed development of further fistulas.
Ongoing advances in EE material and design aim to reduce tissue overgrowth, EE migration, and acid reflux in cases where the
EE traverses the GEJ. A modified covered self-expanding metallic stent (SEMS) is available which can be fixed by a silk thread
in an attempt to reduce migration, and a SEMS with a distal antireflux valve designed in an attempt to reduce acid reflux
(Fig. 16), are examples of novel developments currently being evaluated . In some instances, endoscopically placed clips may be used to prevent EE migration, especially in patients where the stricture
is extrinsic and/or not overly tight.
In summary, based on currently available data, no one SEMS can be recommended to be ideal for all causes of malignant dysphagia.
Although all SEMS used in the comparative trial provided adequate relief of dysphagia, a statistically-non-significant two-fold
higher complication rate was seen with the Gianturco Z-stent compared with the Flamingo Wallstent (36% vs. 18%) [77,80]. Although this finding has to be interpreted carefully given the inadequate power of this trial, it should not be ignored.
Migration rates may be affected by EE diameter, with larger-caliber EE being less likely to migrate. But this finding has
to be balanced by the possibility that larger caliber EE may increase the rate of complication from stent erosion through
the esophageal wall.
Finally partially covered stents with wider distal flanges may prove useful in reducing the rate of EE migration though this
is not proven by the available data. Local resource availability, technical expertise and familiarity, and price are other
factors that need to considered in the decision-making process .
Surgery for esophageal cancer
Although there have been no significant increases in the overall survival rates for EC regardless of treatment modality, surgical
removal remains the only definitive method for long-term cure . Sadly, it is the typically late presentation of EC that acts as the major limiting factor to improving the prognosis in
In fact, despite a striking increase in surgical resection rates and improvement in operative mortality; the curative potential
of surgery remains highest in the early stages of the disease. Indeed, several studies using early staging laparoscopy/operative exploration have confirmed the high incidence of metastatic EC (4095%), despite initially negative conventional preoperative staging, further supporting the use of laparoscopy as a critical
Another problem associated with more precisely defining the role of surgery in EC has been that of data collection. Collecting
data from prospective RCTs has been limited by the relatively low incidence of the disease, as well as by the many variables
that must be considered in planning therapy. Such variables can be divided into provider-related issues such as local expertise
and experience, and patient-related issues such as comorbidity and the characteristics of the tumor itself.
Most esophageal tumors are divided equally in incidence between ESCC on one hand, and EAC arising from BE or mucosal/submucosal glands on the other. Amongst ESCC, verrucous carcinoma, and squamous cell carcinoma with a spindle cell component,
are both subtypes of ESCC that offer a better prognosis than other forms of squamous cell carcinomas.
Precise localization of the carcinoma is essential in determining both the feasibility, and approach to surgery, that is ultimately
adopted. Tumors occurring in the cervical esophagus drain into paratracheal, deep cervical, and internal jugular nodes and
tend not to spread longitudinally throughout the lymphatics. In addition because the trachea, major vessels and spine are
often involved in such tumors, they are usually unresectable either for cure or palliation. In the few cases where the tumor
is deemed resectable, a total laryngectomy should be performed because of the high likelihood of microscopic tumor invasion
of the larynx. Operative mortality is 8% and the 5-year survival is 1520% .
In contrast, most authorities agree that within the thoracic esophagus, aggressive resection should only be performed for
tumors involving the lower esophagus/cardia (below the level of the tracheal bifurcation), since the proximity of tumors of the upper and middle esophagus to the
trachea and aorta often precludes an adequate en bloc resection . The procedures of choice for curative surgery of the lower esophagus are an en bloc dissection with 2/3-field lymphadenectomy, an en bloc esophagectomy with a 2-field lymph node dissection (the IvorLewis operation), or a transhiatal
esophagectomy. In all cases, the major influences ultimately affecting cure rates include the degree of resectabilty, the
stage of the disease, and the complete removal of all malignant tissue involvedlymph nodes included .
In addition, accurate staging with the TNM classification determines whether surgical resection is a possibility. 5-year survival
is directly related to stage: carcinoma in situ 100%, stage I 6080%, stage II 3550%, stage III 15%, and stage IV 0%.
The overall condition of the patient can also help determine optimal treatment strategies. Patient age, cardiopulmonary reserve
and impaired renal function should all be adequately assessed in the preoperative evaluation. Surgical risk is increased in
subjects over 75 years of age, those patients with a forced expiratory volume below 1 l, or those patients with a cardiac ejection fraction of less than 40%. Accordingly any of these parameter are contraindications
for performing a major en bloc resection .
The lack of large prospective data sets has perpetuated the variability in conceptions regarding the utility of the surgical
approach in the management of EC. Moreover, the low overall 5 years survival rates of between 4 and 15%, and the 33.3% hospital
death rate of the two-thirds of patients who are deemed resectable, have fueled this controversy.
In addition, interpretations surrounding the presence and significance of lymph nodes have varied . Some suggest that the presence of any lymph nodes reflects systemic disease with no possibility of long-term cure, to the
opposing view that believes that positive lymph nodes may be treated with aggressive en bloc dissection in conjunction with
Nevertheless, despite the disagreements in the optimal surgical approach, there is general consensus that the proximal esophageal
resection margin should be at least 10 cm and the distal clearance margin 5 cm because of the higher incidence of anastomotic recurrence when the margins are smaller .
Once it has been decided that a patient is a suitable candidate for curative surgery, the exact surgical operations used to
treat EC is governed by the philosophical approach to the disease as well as provider expertise and resources. The two standard
surgical approaches, a transhiatal abdominal approach, or approach via a thoracotomy, each may be supplemented by a neck incision,
or indeed used in combination.
The most radical forms of surgery for EC include the radical esophagectomy with 3-field lymphadectomy, an en bloc esophagogastrectomy
with colon replacement incorporating a 2-field lymphadectomy (used for carcinoma of the lower esophagus/cardia), and an en bloc esophagectomy with a 2-field lymph node dissection (the IvorLewis operation) .
Less aggressive approaches include a transhiatal esophagectomy without thoracotomy, in which the entire intrathoracic esophagus
and proximal cuff of stomach is mobilized through the abdomen. In this case, an anastomosis is subsequently created between
the cervical esophagus and the stomach, without opening of the chest cavity .
A minimally invasive laparoscopic version of this operation employing five trochar sites in the abdomen, and a small cervical
incision has also been developed. Provisional data suggests that by utilizing this approach, the number of days spent in the
intensive care unit can be reduced, as well leading to an earlier resumption of normal activities of daily living. There have
also been reports of using a purely thoracoscopic esophageal dissection in which the esophagus is mobilized and resected with
lymph node clearance under purely thoracoscopic control .
Despite our best efforts to elucidate whether a tumor is curative prior to surgery, there are many cases where the stage is
only truly discovered by operative findings. Discovery of positive nodes, a transmural non-resectable esophagus, or distant
metastasis may quickly change the aim of surgery from curative to palliative. In these cases, tumor debulking, or even continuation
of complete surgical removal, may still offer the best quality of life in cases of palliation, than surgical bypass procedures
or endoscopic therapy (e.g. ablative techniques, endoprosthesis placement).
Surprisingly, the natural history of EC can occasionally be quite variable. In some cases, early EC can remain asymptomatic
in the non-invasive stage for many years. That EC can occasionally be indolent, coupled with the not insubstantial morbidity/mortality associated with surgical esophagectomy should also not be forgotten. Ultimately, more data is required for us to
more fully be able to select the optimal strategy for each patient.
Radiation therapy for esphageal cancer
Radiation therapy (RT) has a limited role as a single modality in EC when the intent of management is curative. The median
survival amongst patients treated with RT alone is between 6 and 12 months, with a five-year survival of approximately 10%.
Although local tumors recur in between 60 and 80% patients after RT, the patients are usually free of dysphagia for 510 months, and if indicated, a further course can often be given .
When RT is combined with surgery with intent to cure, doses of between 50 and 80 Grays are used. Indeed, using peri-operative
RT has not led to unequivocally improved survival rates . Moreover, the relative efficacy of the RT and surgery has been difficult to assess, primarily because RT has historically
been reserved for advanced EC, and no large-scale RCTs have adequately compared surgery with radiation .
The patients who best respond to RT include those of female gender, and those patients with tumor characteristics including
a length of less than 5 cm, and a more proximal esophageal location. RT can be delivered as external beam radiation (EBR) and/or via intraluminal devices with an aim of reducing injury to the surrounding normal tissue.
The most common complications following RT are stricture formation, occurring in up to two thirds of all patients, and esophagitis.
Although most RT induced strictures respond well to esophageal dilation, some are more refractory, requiring either more frequent
dilation, or supplementation by inserting an EE. Extra-esophageal complication of RT include formation of tracheo-esophageal
fistula, transverse myelitis, pericardial effusion, constrictive pericarditis, pneumonitis, pulmonary fibrosis and the BrownSequard syndromeall complications which can develop up to 60 months following RT .
The role of RT in the presence of tumors with existing tracheoesophageal fistulas is problematic, given the possibility of
exacerbation of the fistula. If RT is contemplated in these circumstances, lower, more targeted doses, in conjunction with
insertion of an EE may reduce the likelihood of exacerbation of the fistula.
In addition, RT has been used as part of multimodality therapy in conjunction with both surgery and chemotherapy [87,88]. It was believed that when used preoperatively, RT (3040 Grays) could theoretically shrink the primary tumor, thereby improving the possibility of resectability, and perhaps reducing
future local tumor recurrence.
Several preliminary studies have suggested that preoperative high dose radiation combined with chemotherapy may reduce local
disease, improve the chance for resectability, and possibly prolong survival . Unfortunately, no prospective clinical trial has confirmed the theory. Accordingly, the use of preoperative RT is not recommended
as an effective clinical strategy . The use of RT in conjunction with chemotherapy, either as neoadjuvant, or adjuvant therapy, is discussed below in the chemotherapy
Chemotherapy for esophageal cancer
The use of single/multiagent chemotherapy as a primary modality in EC has been disappointing. Accordingly its role as a curative modality has
been limited to use as part of multimodal therapies where it can be used either preoperatively (neoadjuvant) or postoperatively
Preoperative chemotherapy has been used in an attempt to decrease tumor activity, increase resectability, and improve disease-free
and overall survival. Most studies have demonstrated that there is no benefit on long-term survival, or an improvement in
resectability . Nonetheless, a favorable preoperative response to chemotherapy is a good predictor in the subsequent planning of therapy
following postoperative recurrence. In addition, those patients who have shown a partial response to neoadjuvant chemotherapy
may show an increase, albeit limited, in short-term survival .
Pre-operative chemotherapy has also been combined with RT (preoperative CRT), and used to some effect in EC, especially radiosensitive
ESCC . A recent review of the literature found 46 non-RCTs and six RCTs of preoperative CRT: although treatment-related toxicity
was significant, 3/6 RCTs demonstrated a benefit in either overall survival or disease-free survival compared with surgery alone . Certainly larger RCTS are required to further examine the issue of whether preoperative CRT does indeed increase overall
survival and if so, to optimize the regimens to decrease toxicity, before the benefit of CRT can be adopted unequivocally
. The emphasis has been on administering CRT preoperatively rather than postoperatively in an attempt to limit the severe
toxicities associated with this approach .
Some studies have demonstrated that the use of preoperative external beam RT, in conjunction with cisplatin-based chemotherapy,
has led to tumor disappearance in up to 24% patients. Unfortunately this remarkable response does not translate into significantly
improved survival times, and a proportion of patients succumb no later than if they were just treated with surgery alone.
Nevertheless, patients who show a complete or partial response to this therapy survive longer than a patient whose disease
either did not respond or progressed during therapy. Certainly, the greater toxicity associated with combination chemoradiation
has limited its application as a standard therapy .
Chemotherapy has also been used postoperatively (adjuvant). When postoperative cisplatin-based chemotherapy was compared with
postoperative RT (50 Grays), both groups had equivalent survival. More data is required to fully establish the optimal role
of postoperative chemotherapy before any definitive recommendations can be made.
In summary, current data examining the role of radiation therapy, chemotherapy, and neoadjuvant/adjuvant chemoradiation is mixed. While many small RCTs have suggested a possible role for the some of the above strategies,
larger scale studies have failed to confirm such effects .Until more data is forthcoming, therapy should be individualized based on available evidence, local expertise, and resource
Esophageal cancer conclusion
Despite EC being an uncommon cancer, the bleak fact remains that it is responsible for 12% of all cancer deaths. A large proportion
of the reason for this markedly poor survival is to do with its often-late clinical presentation, often correlated with its
predisposition to spread early to extraesophageal structures both directly and indirectly. There has been a technological
expansion of resources available for dealing with both early and advanced esophageal cancer.
New endoscopic interventions such as EMR and the newer ablative therapies have heightened the need for more accurate staging
modalities. The distinction between many of these endoscopic techniques and the techniques of minimally invasive surgeries
are becoming so blurred that it is not inconceivable that one day the endoscopist and surgeon will be operating simultaneously
on the same patient. Certainly more data is essential to help ascertain the optimal management strategies for the different
stages of EC.
Although most of our current therapeutic paradigm has centred on improving interventions when the tumor is advanced, the recent
change in focus towards preventive strategies, screening programs (e.g. biomarkers), and increased concentration on intervention
on premalignant conditions such as BE, may ultimately lead to a greater success in reducing the mortality associated with
Outstanding issue and future trends
Prospective studies are needed to evaluate outcomes of the surveillance program, such as the optimal frequency of surveillance
examination. The increasing use of endoscopic screening and surveillance programs, advanced imaging techniques such as high-frequency
endoscopic ultrasound and optical coherence tomography, and new endoscopic interventions such as endoscopic mucosal resection
may lead to early detection and treatment of esophageal mucosal dysplasia and cancer. Early intervention could lead to better
prognosis or cure. There are now tremendous interests and efforts in finding potential biomarkers in Barrett's dysplasia that
can predict the development of dysplasia and cancer, hence resulting in a better and more cost-efficient surveillance program.
Likewise, future imaging techniques can potentially detect mucosal dysplasia in real time during endoscopy. There also has
been a technological expansion of resources available for dealing with both early and advanced esophageal cancer. Advances
in biotechnology offer hopes for patients with advanced cancer, such as gene-based therapy and molecule-targeted drugs delivered
either systemically or locally by endoscopic injection. Chemotherapeutics coated and radiation emitting metal stents may decrease
the stent occlusion rates. These endeavors may ultimately lead to a greater success in reducing the mortality associated with
1 Spechler, SJ. Barrett's esophagus. N Engl J Med 2002; 346: 83642. PubMed
2 Shaheen, NJ, Crosby, MA, Bozymski, EM & Sandler, RS. Is there publication bias in the reporting of cancer risk in Barrett's esophagus? [see comments]. Gastroenterology 2000; 119: 3338. PubMed
3 Axon, A, Lambert, R, Robaszkiewicz, M, Rosch, T & Sonnenberg, A. The Second European Endoscopy Forum (Sintra, Portugal, 1718 June 1999): twenty questions on the esophagogastric junction. Endoscopy 2000; 32: 41118. PubMed
4 Sharma, P, Morales, TG & Sampliner, RE. Short segment Barrett's esophagus the need for standardization of the definition and of endoscopic criteria. Am J Gastroenterol 1998; 93: 10336. PubMed
5 Lagergren, J, Bergstrom, R, Lindgren, A & Nyren, O. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med 1999; 340: 82531. PubMed
6 Cameron, AJ. Epidemiology of columnar-lined esophagus and adenocarcinoma. Gastroenterol Clinics North Am 1997; 26: 48794.
7 Campos, GM, DeMeester, SR, Peters, JH, Oberg, S, Crookes, PF, Hagen, JA, Bremner, CG, Sillin, LF, Mason, RJ & DeMeester, TR. Predictive factors of Barrett esophagus: multivariate analysis of 502 patients with gastroesophageal reflux disease. Arch Surg 2001; 136: 126773. PubMed
8 Conio, M, Filiberti, R, Blanchi, S, Ferraris, R, Marchi, S, Ravelli, P, Lapertosa, G, Iaquinto, G, Sablich, R, Gusmaroli, R, Aste, H & Giacosa, A. Gruppo Operativo per lo Studio delle Precancerosi E. Risk factors for Barrett's esophagus: a case-control study. Int J Cancer 2002; 97: 2259. PubMed
9 Dulai, GS, Guha, S, Kahn, KL, Gornbein, J & Weinstein, WM. Preoperative prevalence of Barrett's esophagus in esophageal adenocarcinoma: a systematic review. Gastroenterology 2002; 122: 2633. PubMed
10 Souza, RF, Shewmake, K, Terada, LS & Spechler, SJ. Acid exposure activates the mitogen-activated protein kinase pathways in Barrett's esophagus. Gastroenterology 2002; 122: 299307. PubMed
11 Kaur, BS, Ouatu-Lascar, R, Omary, MB & Triadafilopoulos, G. Bile salts induce or blunt cell proliferation in Barrett's esophagus in an acid-dependent fashion. Am J Physiol Gastrointest Liver Physiol 2000; 278: G10009. PubMed
12 Shirvani, VN, Ouatu-Lascar, R, Kaur, BS, Omary, MB & Triadafilopoulos, G. Cyclooxygenase 2 expression in Barrett's esophagus and adenocarcinoma: Ex vivo induction by bile salts and acid exposure. Gastroenterology 2000; 118: 48796. PubMed
13 Fitzgerald, RC, Lascar, R & Triadafilopoulos, G. Review article: Barrett's esophagus, dysplasia and pharmacologic acid suppression. Aliment Pharmacol Ther 2001; 15: 26976. PubMed
14 Katz, PO. Review article. the role of non-acid reflux in gastro-oesophageal reflux disease. Aliment Pharmacol Ther 2000; 14: 153951. PubMed
15 Chow, WH, Blaser, MJ, Blot, WJ, Gammon, MD, Vaughan, TL, Risch, HA, Perez-Perez, GI, Schoenberg, JB, Stanford, JL, Rotterdam, H, West, AB & Fraumeni, JF Jr. An inverse relation between cagA+ strains of Helicobacter pylori infection and risk of esophageal and gastric cardia adenocarcinoma. Cancer Res 1998; 58: 58890. PubMed
16 Pfau, PR & Sivak, MV Jr Endoscopic diagnostics. Gastroenterology 2001; 120: 76381. PubMed
17 Canto, MI. Methylene blue chromoendoscopy for Barrett's esophagus: coming soon to your GI unit? [Letter; Comment] Gastrointest Endosc 2001; 54: 4039. PubMed
18 Stevens, PD, Lightdale, CJ, Green, PH, Siegel, LM, Garcia-Carrasquillo, RJ & Rotterdam, H. Combined magnification endoscopy with chromoendoscopy for the evaluation of Barrett's esophagus. Gastrointest Endosc 1994; 40: 7479. PubMed
19 Guelrud, M, Herrera, I, Essenfeld, H & Castro, J. Enhanced magnification endoscopy: a new technique to identify specialized intestinal metaplasia in Barrett's esophagus. Gastrointest Endosc 2001; 53: 55965. PubMed
20 Guelrud, M & Herrera, I. Acetic acid improves identification of remnant islands of Barrett's epithelium after endoscopic therapy. Gastrointest Endosc 1998; 47: 51215. PubMed
21 Pfau, PR & Chak, A. Endoscopic ultrasonography. Endoscopy 2002; 34: 218. PubMed
22 Vazquez-Sequeiros, E & Wiersema, MJ. High-frequency US catheter-based staging of early esophageal tumors. Gastrointest Endosc 2002; 55: 959. PubMed
23 DaCosta, RS, Wilson, BC & Marcon, NE. Light-induced fluorescence endoscopy of the gastrointestinal tract. Gastrointest Endosc Clinics North Am 2000; 10: 3769, vi.
24 Mayinger, B, Horner, P, Jordan, M, Gerlach, C, Horbach, T, Hohenberger, W & Hahn, EG. Light-induced autofluorescence spectroscopy for the endoscopic detection of esophageal cancer. Gastrointest Endosc 2001; 54: 195201. PubMed
25 Bown, SG & Lovat, LB. The biology of photodynamic therapy in the gastrointestinal tract. Gastrointest Endosc Clinics North Am 2000; 10: 53350.
26 Li, XD, Boppart, SA, Van Dam, J, Mashimo, H, Mutinga, M, Drexler, W, Klein, M, Pitris, C, Krinsky, ML, Brezinski, ME & Fujimoto, JG. Optical coherence tomography: advanced technology for the endoscopic imaging of Barrett's esophagus. Endoscopy 2000; 32: 92130. PubMed
27 Gerson, LB, Edson, R, Lavori, PW & Triadafilopoulos, G. Use of a simple symptom questionnaire to predict Barrett's esophagus in patients with symptoms of gastroesophageal reflux. Am J Gastroenterol 2001; 96: 200512. PubMed
28 Haggitt, RC. Barrett's esophagus, dysplasia, and adenocarcinoma. Human Pathol 1994; 25: 98293.
29 Spechler, SJ. Disputing dysplasia. Gastroenterology 2001; 120: 18648. PubMed
30 Anonymous. The role of endoscopy in the surveillance of premalignant conditions of the upper gastrointestinal tract. Guidelines for clinical
application. Gastrointest Endosc 1988; 34: 18S20S. PubMed
31 Shaheen, NJ. Does surveillance endoscopy improve life expectancy in those with Barrett's esophagus? Gastroenterology 2001; 121: 151618. PubMed
32 DeVault, KR & Castell, DO. Updated guidelines for the diagnosis and treatment of gastroesophageal reflux disease. The Practice Parameters Committee of
the American College of Gastroenterology. Am J Gastroenterol 1999; 94: 143442. PubMed
33 Krishnadath, KK, Reid, BJ & Wang, KK. Biomarkers in Barrett esophagus. Mayo Clinic Proc 2001; 76: 43846.
34 Gudlaugsdottir, S, van Blankenstein, M, Dees, J & Wilson, JH. A majority of patients with Barrett's esophagus are unlikely to benefit from endoscopic cancer surveillance. Eur J Gastroenterol Hepatol 2001; 13: 63945. PubMed
35 Arguedas, MR & Eloubeidi, MA. Barrett's esophagus: a review of costs of the illness. Pharmacoeconomics 2001; 19: 100311. PubMed
36 Brown, LM, Silverman, DT, Pottern, LM, Schoenberg, JB, Greenberg, RS, Swanson, GM, Liff, JM, Schwartz, AG, Hayes, RB & Blot, WJ. Adenocarcinoma of the esophagus and esophagogastric junction in white men in the United States: alcohol, tobacco, and socioeconomic
factors. Cancer Causes Control 1994; 5: 33340. PubMed
37 Brown, LM, Swanson, CA, Gridley, G, Swanson, GM, Schoenberg, JB, Greenberg, RS, Silverman, DT, Pottern, LM, Hayes, RB & Schwartz, AG. Adenocarcinoma of the esophagus: role of obesity and diet. J Natl Cancer Inst 1995; 87: 1049. PubMed
38 Kandil, HM, Tanner, G, Smalley, W, Halter, S, Radhika, A & Dubois, RN. Cyclooxygenase-2 expression in Barrett's esophagus. Dig Dis Sci 2001; 46: 7859. PubMed
39 Morris, CD, Armstrong, GR, Bigley, G, Green, H & Attwood, SE. Cyclooxygenase-2 expression in the Barrett's metaplasia-dysplasia-adenocarcinoma sequence. Am J Gastroenterol 2001; 96: 9906. PubMed
40 Van Laethem, JL, Peny, MO, Salmon, I, Cremer, M & Deviere, J. Intramucosal adenocarcinoma arising under squamous re-epithelialization of Barrett's esophagus. Gut 2000; 46: 5747. PubMed
41 Byrne, JP, Armstrong, GR & Attwood, SE. Restoration of the normal squamous lining in Barrett's esophagus by argon beam plasma coagulation. Am J Gastroenterol 1998; 93: 181015. PubMed
42 Van Laethem, JL, Cremer, M, Peny, MO, Delhaye, M & Deviere, J. Eradication of Barrett's mucosa with argon plasma coagulation and acid suppression: immediate and mid term results. Gut 1998; 43: 74751. PubMed
43 Mork, H, Barth, T, Kreipe, HH, Kraus, M, Al-Taie, O, Jakob, F & Scheurlen, M. Reconstitution of squamous epithelium in Barrett's esophagus with endoscopic argon plasma coagulation: a prospective study. Scand J Gastroenterol 1998; 33: 11304. PubMed
44 Schulz, H, Miehlke, S, Antos, D, Schentke, KU, Vieth, M, Stolte, M & Bayerdorffer, E. Ablation of Barrett's epithelium by endoscopic argon plasma coagulation in combination with high-dose omeprazole. Gastrointest Endosc 2000; 51: 65963. PubMed
45 Rosch, T. Gastroesophageal reflux disease and Barrett's esophagus. Endoscopy 2001; 33: 90916. PubMed
46 Van Laethem, JL, Jagodzinski, R, Peny, MO, Cremer, M & Deviere, J. Argon plasma coagulation in the treatment of Barrett's high-grade dysplasia and in situ adenocarcinoma. Endoscopy 2001; 33: 25761. PubMed
47 Ertan, A, Zimmerman, M & Younes, M. Esophageal adenocarcinoma associated with Barrett's esophagus: long-term management with laser ablation. Am J Gastroenterol 1995; 90: 22013. PubMed
48 Fremond, L, Bouche, O, Diebold, MD, Demange, L, Zeitoun, P & Thiefin, G. [Partial regression of Barrett esophagus with high grade dysplasia and adenocarcinoma after photocoagulation and endocurietherapy
under antisecretory treatment]. Gastroenterol Clin Biol 1995; 19: 11216. PubMed
49 Salo, JA, Salminen, JT, Kiviluoto, TA, Nemlander, AT, Ramo, OJ, Farkkila, MA, Kivilaakso, EO & Mattila, SP. Treatment of Barrett's esophagus by endoscopic laser ablation and antireflux surgery. Ann Surg 1998; 227: 404. PubMed
50 Gossner, L, May, A, Stolte, M, Seitz, G, Hahn, EG & Ell, C. KTP laser destruction of dysplasia and early cancer in columnar-lined Barrett's esophagus. Gastrointest Endosc 1999; 49: 812. PubMed
51 Wang, KK. Photodynamic therapy of Barrett's esophagus. Gastrointest Endosc Clinics North Am 2000; 10: 40919.
52 Ackroyd, R, Brown, NJ, Davis, MF, Stephenson, TJ, Marcus, SL, Stoddard, CJ, Johnson, AG & Reed, MW. Photodynamic therapy for dysplastic Barrett's esophagus: a prospective, double blind, randomised, placebo controlled trial. Gut 2000; 47: 61217. PubMed
53 Overholt, BF, Panjehpour, M & Haydek, JM. Photodynamic therapy for Barrett's esophagus: follow-up in 100 patients. Gastrointest Endosc 1999; 49: 17. PubMed
54 Gossner, L, Stolte, M, Sroka, R, Rick, K, May, A, Hahn, EG & Ell, C. Photodynamic ablation of high-grade dysplasia and early cancer in Barrett's esophagus by means of 5-aminolevulinic acid. Gastroenterology 1998; 114: 44855. PubMed
55 Pfefer, TJ, Schomacker, KT & Nishioka, NS. Long-term effects of photodynamic therapy on fluorescence spectroscopy in the human esophagus. Photochem Photobiol 2001; 73: 6648. PubMed
56 Shim, CS. Endoscopic mucosal resection: an overview of the value of different techniques. Endoscopy 2001; 33: 2715. PubMed
57 Leiper, K & Morris, AI. Treatment of oesophago-gastric tumors. Endoscopy 2002; 34: 13945. PubMed
58 Ponchon, T. Endoscopic mucosal resection. J Clin Gastroenterol 2001; 32: 610. PubMed
59 Ell, C, May, A, Gossner, L, Pech, O, Gunter, E, Mayer, G, Henrich, R, Vieth, M, Muller, H, Seitz, G & Stolte, M. Endoscopic mucosal resection of early cancer and high-grade dysplasia in Barrett's esophagus. Gastroenterology 2000; 118: 6707. PubMed
60 Fennerty, MB & Triadafilopoulos, G. Barrett's-related esophageal adenocarcinoma: is chemoprevention a potential option? Am J Gastroenterol 2001; 96: 23025. PubMed
61 DeMeester, TR. Antireflux surgery in the management of Barrett's esophagus. J Gastrointest Surg 2000; 4: 1248. PubMed
62 Csendes, A, Braghetto, I, Burdiles, P, Puente, G, Korn, O, Diaz, JC & Maluenda, F. Long-term results of classic antireflux surgery in 152 patients with Barrett's esophagus. clinical, radiologic, endoscopic,
manometric, and acid reflux test analysis before and late after operation. Surgery 1998; 123: 64557. PubMed
63 Spechler, SJ, Lee, E, Ahnen, D, Goyal, RK, Hirano, I, Ramirez, F, Raufman, JP, Sampliner, R, Schnell, T, Sontag, S, Vlahcevic, ZR, Young, R & Williford, W. Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease: follow-up of a randomized controlled
trial. Jama 2001; 285: 23318. PubMed
64 Demeester, SR, Peters, JH & Demeester, TR. Barrett's esophagus. Curr Prob Surg 2001; 38: 558640.
65 Mamazza, J, Schlachta, CM & Poulin, EC. Surgery for peptic strictures. Gastrointest Endosc Clinics North Am 1998; 8: 399413.
66 Wong, R & Malthaner, R. Esophageal cancer: a systematic review. Curr Prob Cancer 2000; 24: 297373.
67 Heitmiller, RF. Epidemiology, diagnosis, and staging of esophageal cancer. Cancer Treat Res 2001; 105: 37586. PubMed
68 Brown, LM. The role of race/ethnicity in the epidemiology of esophageal cancer. J Assoc Acad Minority Phys 2000; 11: 327.
69 Lambert, R. An overview of the management of cancer of the esophagus. Gastrointest Endosc Clinics North Am 1998; 8: 41534.
70 Koch, J & Halvorsen, RA Jr Staging of esophageal cancer: computed tomography, magnetic resonance imaging, and endoscopic ultrasound. Seminars Roentgenol 1994; 29: 36472.
71 Kelly, S, Harris, KM, Berry, E, Hutton, J, Roderick, P, Cullingworth, J, Gathercole, L & Smith, MA. A systematic review of the staging performance of endoscopic ultrasound in gastro-oesophageal carcinoma. Gut 2001; 49: 5349. PubMed
72 Krasna, MJ. Surgical staging and surgical treatment in esophageal cancer. Semin Oncol 1999; 26: 911. PubMed
73 Ponec, RJ & Kimmey, MB. Endoscopic therapy of esophageal cancer. Surg Clinics North Am 1997; 77: 1197217.
74 Lightdale, CJ. Role of photodynamic therapy in the management of advanced esophageal cancer. Gastrointest Endosc Clinics North Am 2000; 10: 397408.
75 Nijhawan, PK & Wang, KK. Endoscopic mucosal resection for lesions with endoscopic features suggestive of malignancy and high-grade dysplasia within
Barrett's esophagus. Gastrointest Endosc 2000; 52: 32832. PubMed
76 Shimi, SM. Self-expanding metallic stents in the management of advanced esophageal cancer: a review. Sem Laparoscopic Surg 2000; 7: 921.
77 Siersema, PD, Hop, WC, van Blankenstein, M, van Tilburg, AJ, Bac, DJ, Homs, MY & Kuipers, EJ. A comparison of 3 types of covered metal stents for the palliation of patients with dysphagia caused by esophagogastric carcinoma:
a prospective, randomized study. Gastrointest Endosc 2001; 54: 14553. PubMed
78 Dallal, HJ, Smith, GD, Grieve, DC, Ghosh, S, Penman, ID & Palmer, KR. A randomized trial of thermal ablative therapy versus expandable metal stents in the palliative treatment of patients with
esophageal carcinoma. Gastrointest Endosc 2001; 54: 54957. PubMed
79 Dua, KS, Kozarek, R, Kim, J, Evans, J, Medda, BK, Lang, I, Hogan, WJ & Shaker, R. Self-expanding metal esophageal stent with anti-reflux mechanism. Gastrointest Endosc 2001; 53: 60313. PubMed
80 Baron, TH. A practical guide for choosing an expandable metal stent for GI malignancies: is a stent by any other name still a stent?
[Letter; Comment ] Gastrointest Endosc 2001; 54: 26972. PubMed
81 Poulin, EC, Schlachta, CM & Mamazza, J. Operative treatment of malignancy. Gastrointest Endosc Clinics North Am 1998; 8: 43550.
82 Law, S & Wong, J. What is appropriate treatment for carcinoma of the thoracic esophagus? World J Surg 2001; 25: 18995. PubMed
83 Law, S & Wong, J. Does lymphadenectomy add anything to the treatment of esophageal cancer? Adv Surg 1999; 33: 31127. PubMed
84 Wajed, SA & Peters, JH. Laparoscopic and endoscopic surgery in esophageal malignancy. Surg Oncol Clinics North Am 2001; 10: 493510. vii.
85 Suntharalingam, M. Radiation therapy for esophageal cancer. Chest Surg Clinics North Am 2000; 10: 56981.
86 Arnott, SJ, Duncan, W, Gignoux, M, Girling, DJ, Hansen, HS, Launois, B, Nygaard, K, Parmar, MK, Rousell, A, Spiliopoulos, G, Stewart, LA, Tierney, JF, Wang, M & Rhugang, Z. Oeosphageal Cancer Collaborative G (2000) Preoperative radiotherapy for esophageal carcinoma. Cochrane Database of Systematic Reviews[Computer File: CD001799.
87 Kelsen, DP. Multimodality therapy of esophageal cancer: an update. Cancer J Sci Am 2000; 6: S17781.
88 Sugarbaker, DJ, Jaklitsch, MT & Liptay, MJ. Thoracoscopic staging and surgical therapy for esophageal cancer. Chest 1995; 107: 218S223S. PubMed
89 Walsh, TN, Noonan, N, Hollywood, D, Kelly, A, Keeling, N & Hennessy, TP. A comparison of multimodal therapy and surgery for esophageal adenocarcinoma. N Engl J Med 1996; 335: 4627. PubMed
90 Urba, S. Combined-modality treatment of esophageal cancer. Oncology (Huntington) 1997; 11: 637.
91 Kok, TC. Chemotherapy in oesophageal cancer. Cancer Treat Rev 1997; 23: 6585. PubMed
92 Daly, JM, Fry, WA, Little, AG, Winchester, DP, McKee, RF, Stewart, AK & Fremgen, AM. Esophageal cancer: results of an American College of Surgeons Patient Care Evaluation Study. J Am Coll Surg 2000; 190: 56272; discussion 5723. PubMed
93 Kelsen, DP. Neoadjuvant therapy of squamous cell carcinoma of the esophagus. Recent Results Cancer Res 2000; 155: 10512. PubMed
94 Geh, JI, Crellin, AM & Glynne-Jones, R. Preoperative (neoadjuvant) chemoradiotherapy in oesophageal cancer. Br J Surg 2001; 88: 33856. PubMed
95 Kelsen, DP, Ginsberg, R, Pajak, TF, Sheahan, DG, Gunderson, L, Mortimer, J, Estes, N, Haller, DG, Ajani, J, Kocha, W, Minsky, BD & Roth, JA. Chemotherapy followed by surgery compared with surgery alone for localized esophageal cancer. N Engl J Med 1998; 339: 197984. PubMed
Copyright © Blackwell Publishing, 2004