June issue.indd

A.C. Belo R.J. Playford
Gastroenterology Section, Imperial College, Faculty of Medicine, Hammersmith Hospital Campus, DuCane Road, London, W12 0NN U.K.

Correspondence to: R.J. Playford, Gastroenterology Section, Imperial College, Faculty of Medicine, Hammersmith Hospital Campus, DuCane Road, London, W12 0NN U.K. Email: r.playford@imperial.ac.uk

Introduction

Keywords: Oesophageal carcinoma, surveillance, screening Surg J R Coll Edinb Irel., 1 June 2003, 152-156

Metaplastic change of the oesophageal epithelium from normal stratified squamous to columnar-lined with intestinal metaplasia results in an increased risk of development of adenocarcinoma. As a result, endoscopic surveillance has been recommended for the surgically-fit patient. The evidence that these programmes are altering clinical outcome to any major degree, however, is weak. This review highlights some of the areas of controversy and outstanding points that need to be clarified to allow establishment of evidence-based medicine for this condition

INTRODUCTION
Barrett’s oesophagus was initially thought to be due to a congenital short oesophagus with a resultant section of intra-thoracic stomach.¹ This concept was superseded by the idea that this is an acquired condition resulting from gastro-oesophageal reflux.² It is now generally accepted that Barrett’s oesophagus is a metaplastic change, with the major factor stimulating this being reflux of gastric contents. This article highlights the major areas of controversy, drawing conclusion where possible but generally showing that we are currently at the stage of ‘many questions but few answers.’

WHAT IS BARRETT’S OESOPHAGUS?
The definition of what is required to make a diagnosis of Barrett’s oesophagus has varied over the years. The major factors that are relevant are the length of the metaplastic segment and the form that this metaplasia takes.

Length
Some of the earlier descriptions required the presence of a >3cm segment of altered mucosa to make the diagnosis. More recently, the term ‘short segment’ Barrett’s oesophagus has been used to describe a length of macroscopically visible Barrett’s type mucosa of <3cm. In addition, the somewhat confusing term of ‘ultrashort segment’ Barrett’s oesophagus has been coined to describe the histological identification of gastric tissue with intestinal metaplasia at the gastro-oesophageal junction.

Type
The replacement of the normal stratified squamous epithelium with gastric type columnar lined oesophagus (CLO) was initially considered sufficient to establish a diagnosis of Barrett’s type mucosa. More recently, it has also been required to identify intestinal metaplasia (IM) in addition to the CLO to diagnose a patient as having Barrett’s oesophagus (Figure 1). This decision was based on the idea that the development of adenocarcinoma probably requires the presence of IM to allow the sequential change from IM to IM and low-grade dysplasia to IM and high-grade dysplasia and finally carcinoma. However, it has recently been reported that virtually all cases of columnar lined oesophagus will also have some areas of IM if enough biopsies are taken (N. Shepherd, 8th European Gastroenterology Week, Brussels, November 2000). Based on this report, if a patient’s biopsies do not show IM, is it logical to state that the patient does not have Barrett’s? Many groups believe that this is not the case.

SIZE OF THE PROBLEM
Barrett’s mucosa is present in a large proportion of the general population. If diagnosis is restricted to a length >3cm, about 0.25% of the general population have Barrett’s oesophagus.³ If any length of columnar lined epithelium is considered sufficient for diagnosis, post-mortem findings suggest it affects about 2% of all subjects.⁴ The chances of finding areas of Barrett’s mucosa are much higher in patients with symptomatic reflux; 10% of subjects undergoing endoscopy with symptoms of longstanding reflux are found to have Barrett’s oesophagus and this rises to 15-20% if erosive oesophagitis is found. ^5,6

WHAT ARE THE RISKS OF HAVING BARRETT’S OESOPHAGUS?
By far the most important concern regarding patients with CLO +/- IM is the risk of developing adenocarcinoma. A wide range of figures has been quoted for the increased risk from 30-125 times that found in the general population. ^7,8 However, an important article by Shaheen et al (2000) with its associated editorial has shed considerable light on this area. ^9,10 The authors found that publication bias towards positive results (i.e., studies with small numbers of subjects only get published if a high risk is found) appears to have occurred for the risks of development of carcinoma in these subjects. A figure that seems closer to the truth is around 0.5% cancer per year (i.e.1 per 200 patient years), although regional variations probably exist with a recent report on the risk in the UK population being about 1% of cancers per year. ^9,11 As the vast majority of patients with Barrett’s oesophagus will die from unrelated diseases, enthusiasm for intervention has to be balanced by cost-benefit considerations. It would be extremely useful, therefore, to be able to identify subgroups with a particularly high risk of cancer. On a community basis, the most relevant factor in identifying subjects who have a higher risk of developing adenocarcinoma is the presence and severity of symptoms of reflux. Lagergren et al (1999) showed that patients with longstanding severe symptoms of reflux have an eight to ten-fold higher risk of cancer development than controls. ¹² It is important to note, however, that some patients with ‘long segment’ Barrett’s oesophagus have significant reflux but an insensitive gullet, making them virtually asymptomatic. Additional aspects, which are thought to be relevant risk factors, are presence of dysplasia (obviously an extremely important factor), ‘long segment’ Barrett’s of >8cm, smoking, co-presence of ulcer or stricture and male gender. ^13-16

Figure 1: Haematoxylin and eosin stain of oesophageal biopsy showing intestinal metaplasia and dysplasia. Original magnification 340. Image kindly provided by Dr. J.J. Boyle

SHOULD WE SURVEY?
In view of the concerns regarding development of cancer, many centres have established surveillance programmes for patients who have been identified as having Barrett’s mucosa present. Recommendations from the World Congress of Gastroenterology, American College of Gastroenterology and other specialist organisations usually state that patients with Barrett’s oesophagus should be recruited into a surveillance programme that entails regular (usually 1-3 yearly) endoscopic examination with multiple biopsies (usually 4-quadrant every 2cm along the affected length). ^8,17 Patients with evidence of mild dysplasia should undergo more frequent endoscopic review (every six months x 2, then yearly) and high-grade dysplasia should result in expert confirmation of the diagnosis with subsequent endoscopy every three months, selective resection or more radical surgical intervention.⁸

When considering the benefit of such procedures, however, it is important to appreciate that enthusiasm to help patients is not enough. Many other specialties undertake screening or surveillance (e.g. cervical screening, breast cancer screening) and the World Health Organisation has provided a useful list of criteria that should be met to ensure that screening/surveillance is of benefit (Table 1). ¹⁸

The evidence that we are fulfilling these criteria for Barrett’s oesophagus is lacking, with even the natural history of Barrett’s oesophagus being poorly understood; figures quoted for progression of high-grade dysplasia to cancer vary between a cumulative cancer incidence at three years of 56% to a five-year cumulative cancer incidence of only 9%. ^19,20

Because of the uncertainty surrounding the benefit of surveillance programmes, we recently reviewed the benefits of a surveillance programme that was ongoing in Leicester for many years. ²¹ The clinical progress and outcome of all subjects who were identified as having Barrett’s oesophagus in a 10-year cohort (1984-94) was followed until the end of 1999.

• Two-thirds of all patients with Barrett’s oesophagus were unfit to enter the programme, usually due to concomitant illness or general frailty, which would have precluded surgery

• Average length of time in the programme was short (4.4 years), again usually due to development of co-morbidity or other factors making potential surgery inappropriate

• Five people who were entered into the programme developed adenocarcinoma of the oesophagus, a figure in keeping with the risks reported by Spechler (2000) of about 0.5% per year. ¹⁰ Only one patient was identified as a result of a surveillance procedure. The remainder would have undergone endoscopy for change in symptoms necessitating endoscopy anyway or returning with symptoms having defaulted from the programme

• As the one subject who was detected by surveillance died post-operatively, no individual gained from participation

• Even if all five cancers had been detected by surveillance and the patients cured, this would have had virtually no impact on the amount of disease in the community. This is because most patients with adenocarcinoma of the oesophagus are not known about until symptoms of dysphagia develop

Review of the literature shows that many of these conclusions are not unique to our experience; for example, van der Burgh et al (1996) followed 155 patients for five years (not entered into a surveillance programme). Of the 79 that died, eight had cancer of the oesophagus but in only two was it the cause of death with one of them being an early post-operative related death. ²² Other publications reporting comparable studies come to the same conclusion; i.e., that the vast majority of patients with Barrett’s oesophagus will die from unrelated causes. Articles such as these provide important information on the risks of having Barrett’s oesophagus, but do not address the issue regarding whether a surveillance programme is beneficial.

The numbers of published works that allow a critical review of surveillance programmes are relatively few. Notable exceptions include the work of Nilsson et al (2000) who identified five cancers from 199 patients with mean surveillance duration of 3.9 years, quoting a figure of $38,000 per cancer detected. ²³ However, they found that of these five, one was unfit for surgery and two died early post-operatively so received little benefit from such intervention. In addition, they do not state whether their patients were symptomatic at the time of diagnosis (and, thus, may have received an endoscopy anyway, as found in some of our patients).

Wright et al (1996) identified 301 patients with Barrett’s oesophagus of which 166 entered the annual endoscopy programme. As found by other groups, there was a relatively short mean surveillance period (2.8 years), and the vast majority died from causes unrelated to Barrett’s oesophagus or became unfit for further surveillance. ²⁴

This study did detect six cancers, five of which were apparently asymptomatic. The subsequent outcome of these patients is unclear, although five were reported as having apparent node negative disease. Interestingly, their reporting of the patients’ progression show that entry into the programme of three of these six patients was based on identification of CLO without IM and, therefore, would have been excluded from surveillance by some groups. The rate of cancer development in this cohort (6 out of 166 over 2.8 years) gave a risk of about 1.25 cases per 100 patient years, about three times higher than that reported in the recent overview by Shaheen et al (2000) (0.5 per100 patient years).⁹

REDUCING THE RISKS IN PATIENTS WITH BARRETT’S
Many genetic abnormalities have been detected in segments of Barrett’s oesophagus, and the idea that this is a simple fixed metaplastic change is probably an oversimplification. ²⁵ The most obvious intervention, which might influence subsequent development of cancer, is to reduce the amount of acid reflux. This can be achieved medically (usually by proton pump inhibitors) or surgical intervention such as fundoplication. There is very limited evidence, however, that such intervention alters malignant progression. Islands of regeneration of stratified squamous epithelium may be induced by vigorous acid suppression. ²⁶ However, IM continues to be present in the deep layers of the mucosa, hidden from endoscopic surveillance but continuing to harbour malignant potential.

Surgical intervention has a definite risk of morbidity and mortality and the evidence that anti-reflux procedures alter outcome is minimal. ²⁷ In view of the fact that a huge number of patients would undergo operation for no cancer-related benefit, we consider that the current thought processes around who should be referred to surgery should relate to symptom control of reflux symptoms rather than altering cancer risks.

An additional intervention, which is currently receiving interest, is the use of laser ablation to facilitate reepithelialisation with stratified squamous mucosa.²⁸ This is an interesting approach but is currently limited by the requirement for multiple procedures per patient and the realisation that the vast majority of patients are being treated for a condition that will have no effect on their lifespan. The cost/benefit of such intervention will have to be monitored closely.

CONCLUSIONS AND OUTSTANDING QUESTIONS
As stated in the introduction, this article highlights many queries but can provide few definite answers. There is a need for randomised control trials regarding the benefits of surveillance programmes although debate exists around the ethics of having a control arm involving no endoscopic surveillance.

The sheer scale of patients with reflux symptoms means that we will probably never be in the position of screening for the presence of Barrett’s oesophagus in the community. At the other extreme, it is generally accepted that patients who are known to have Barrett’s oesophagus with evidence of dysplasia should be closely monitored by surveillance or have some form of intervention. This advice was previously limited to those subjects who were potentially suitable for radical surgery, but the advent of less aggressive therapy such as laser therapy or local resection means that we probably need to consider this on an individual basis. At present, most groups consider it unwise to perform random biopsies of the gastro-oesophageal junction in an apparently normal looking oesophagus.⁸ This is because we currently have no idea of what to do if histology identified CLO and IM (ultrashort segment Barrett’s oesophagus).

For the vast majority of patients who have a segment of Barrett’s oesophagus detected as a coincidental finding, we currently have a clinical conundrum. Separation of patients according to the length of Barrett’s oesophagus (> or <3cm) provides little clinical benefit, as recent reports suggest a similar risk in both groups. ²⁹ Following recommended guidelines of surveillance every two years with 4-quadrant biopsies every 2cm is the ‘easiest’ decision to make but, as this review shows, is based on little evidence and goes against the opinions of many clinicians. ³⁰ Our own opinion is that when an area of Barrett’s is first detected, it should be rigorously biopsied to detect dysplasia. If none is found there is a case for not entering subjects in a surveillance programme outside of a clinical trial. Until further data are forthcoming, specific advice needs to be tailored to individual patient circumstances (age, medical fitness, etc.), and we must accept that these decisions are unfortunately being based on medicine as an ‘art’ rather than a ‘science’.

In order to improve the cost-benefit outcome of surveillance programmes, the relatively high workload to low incidence of cancer detected means that we should be attempting to find ways of identifying those subjects without dysplasia who are at higher risk. Research into molecular markers of genetic instability may prove useful although the identification of the optimum one(s) to use is still unclear. ³¹ Acid suppression is of undoubted benefit for symptom control and laboratorybased studies suggest theoretical reasons why it might reduce progression. ³² Clinical trials proving this point are, however, lacking. Probably one of the more disturbing aspects of this area is the promotion of guidelines from various specialist bodies on how to deal with these patients. Although guidelines are excellent when data are strong, in situations such as this they lead to clinicians following the recommendations to reduce the risk of litigation rather than because they are sure they are helping patients.

REFERENCES
1. Barrett, NR. Chronic peptic ulcer of the oesophagus and ‘oesophagitis’. Br J Surg 1950; 38: 175-82
2. Allison, JR, Johnston, AS. The oesophagus linked with gastric mucous membrane. Thorax 1953; 8: 87-101
3. Stein, HJ, Siewert, JR. Barrett’s esophagus: pathogenesis, epidemiology, functional abnormalities, malignant degener-ation and surgical management. Dysphagia 1993; 8: 276-88
4. Cameron, AJ, Zinnmeister, AR, Ballard, DJ, Carey, JA. Prevalence of columnar lined Barrett’s esophagus: comparison of population based clinical and autopsy findings. Gastroenterology 1990; 99: 918-22
5. Johnston, MH, Hammond, AS, Laskin, W, Jones, DM. The prevalence and clinical characteristics of short segments of specialised intestinal metaplasia in the distal esophagus on routine endoscopy. Am J Gastroenterol 1966; 91: 1507-11
6. Bartlesman, JF, Hameeteman, W, Tytgat, GN. Barrett’s oesophagus. Eur J Cancer Prev 1992; 1: 323-25
7. Spechler, SJ, Goyal, RK. The columnar lined esophagus, intestinal metaplasia and Norman Barrett. Gastroenterology 1996; 110: 614-21
8. Sampliner, RE and The Practice Parameters Committee of the American College of Gastroenterology. Practice Guidelines on the diagnosis, surveillance and therapy of Barrett’s esophagus. Am J Gastroenterol 1998; 93: 1028-32
9. Shaheen, NJ, Crosby, MA, Bozymski, EM, Sandler, RS. Is there publication bias in the reporting of cancer risk in Barrett’s esophagus? Gastroenterology 2000; 119: 333-38
10. Spechler, SJ. Barrett’s esophagus: An overrated cancer risk factor: (Editorials) Gastroenterology 2000; 119: 587-89
11. Jankowski, JA, Provencale, D, Moayyedi, P. Oesophageal adenocarcinoma arising from Barrett’s metaplasia has regional variations worldwide. Gastroenterology 2002
12. Lagergren, J, Bergstrom, R, Lindgren, A et al. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. New Engl J Med 1999; 40: 825-31
13. van Sandick, JW, van Lauschott, JJ, Kuiken, BW, Tytgat, GN, Offerhaus, GJ, Obertop, H. Impact of endoscopic biopsy surveillance of Barrett’s oesophagus on pathological stage and clinical outcome of Barrett’s carcinoma. Gut 1998; 43: 216-22
14. Iftikhar, SY, James, PD, Steele, RJC, Hardcastle, JD, Atkinson, M. Length of Barrett’s oesophagus: an important factor in the development of dysplasia and adenocarcinoma. Gut 1992; 33: 1155-58
15. Gray, MR, Donnelly, RJ, Kingsnorth, AN. The role of smoking and alcohol in metaplasia and cancer risk in Barrett’s columnar lined oesophagus. Gut 1993; 34: 727-31
16. Menke-Pluymers, MB, Hop, WC, Dees, J, van Blankenstein, M, Tilanus, HW. Risk factors for the devel-opment of an adenocarcinoma in columnar-lined (Barrett’s) esophagus. The Rotterdam Esophageal Tumour Study Group. Cancer 1993; 72: 1155-58
17. Dent, J, Bremner, CG, Collen, MJ et al. Working Party report to the World Congress of Gastroenterology, Sydney 1990. Barrett’s oesophagus J Gastroenterol Hepatol 1991; 6:1-22
18. Wilson, JMG, Junger, G. Principles and practice of screening for disease. Geneva: World Health Organisation, 1968
19. Buttar, NS, Wang, KK, Sebo, TJ et al. Extent of high-grade dysplasia in Barrett’s esophagus correlates with risk of adenocarcinoma. Gastroenterology 2000; 120: 1630-39
20. Schnell, TG, Sontag, SJ, Chejfec, G et al. Long-term nonsurgical management of Barrett’s esophagus with highgrade dysplasia. Gastroenterology 2001; 120: 1607-1619
21. Macdonald, CE, Wicks, AC, Playford, RJ. Final results from 10-year cohort of patients undergoing surveillance for Barrett’s oesophagus: observational study. BMJ 2000; 321: 1252-55
22. van der Burgh, A, Dees, J, Hop, WCJ, van Blankstein, M. Oesophageal cancer is an uncommon cause of death in patients with Barrett’s oesophagus. Gut 1996; 39: 5-8
23. Nilsson, J, Skobe, V, Johansson, J, Willen, R, Johnsson, F. Screening for oesophageal adenocarcinoma: an evaluation of a surveillance programme for columnar metaplasia of the oesophagus. Scand J Gastroenterology 2000; 35: 10-16
24. Wright, TA, Gray, MR, Morris, AI, Gilmore, IT, Ellis, A, Smart, HL et al. Cost effectiveness of detecting Barrett’s cancer. Gut 1996; 39: 574-79.
25. Jankowski, JA, Harrison, RF, Perry, I, Balkwill, F, Tselepis, C. Barrett’s metaplasia. Lancet 2000; 356: 2079-85
26. Sharma, P, Morales, TG, Bhattacharyga, A et al. Squamous islands in Barrett’s esophagus: what lies underneath? Am J Gastroenterol 1998; 93: 332-35
27. Lambert, R. Review article: current practice and future perspectives in the management of gastro-oesophageal reflux disease. Aliment Pharmacol Ther 1997; 11: 651-62
28. Salo, JA, Salminen, JT, Kiviluoto, TA et al. Treatment of Barrett’s esophagus by endoscopic laser ablation and antire-flux surgery. Ann Surg 1998; 227: 40-44
29. Rudolph, RE, Vaughan, TL, Storer, BE et al. Effect of segment length on risk for neoplastic progression in patients with Barrett’s esophagus. Ann Intern Med 2000; 132: 612-20
30. Heading, RC. Barrett’s oesophagus - Time for a reality check? Digest Liver Dis 2001; 33: 313-15
31. van Lieshout, EM, Jansen, JB, Peters, WH. Biomarkers in Barrett’s esophagus (review). Int J Oncol 1998; 13: 855-64
32. Fitzgerald, RC, Omary, MB, Triadafilopoulos, G. Acid modulation of HT-29 cell growth and differentiation. An in vitro model for Barrett’s esophagus. J Cell Sci 1997; 110: 663-71

Education Section

Surveillance for Barrett’s oesophagus: is there light at the end of the metaplastic tunnel?