Skip to main content
NCBI home page
The BMJ logoLink to The BMJ
. 2001 May 5;322(7294):1124.

Surveillance for Barrett's oesophagus

The conundrum of Barrett's oesophagus is changing

J A Eksteen 1,2, Janusz A Jankowski 1,2
PMCID: PMC1120248  PMID: 11360913

Editor—Macdonald et al in their paper and McGarrity in his accompanying editorial reviewed the value of endoscopic surveillance of Barrett's oesophagus.1,2 Both articles highlighted the major problems with detection of oesophageal adenocarcinoma in an unselected group of individuals with Barrett's oesophagus.

Much attention has been devoted to risk stratification of individuals who are at high risk of malignant change in Barrett's oesophagus. Men over 45 years, those with at least 3 cm of Barrett's metaplasia, those with severe and frequent reflux symptoms (>3 times week), those with chronic heartburn for 10 years or more, obese patients, those taking drugs which relax the lower oesophageal sphincter (such as nitrates), and perhaps those with eradicated Helicobacter pylori infection are most at risk of Barrett's associated adenocarcinoma.3

Pathology has also made a major contribution to understanding the pathogenesis as intestinal type metaplasia gives rise to dysplastic clones from which the adenocarcinoma arises.3 Molecular genetics has been rigorously applied to samples along the sequence encompassing Barrett's metaplasia, dysplasia, and adenocarcinoma, and it has yielded important information about key genetic alterations.4 Furthermore, information of those with a family history of gastro-oesophageal cancer has also yielded rare, but none the less important, genetic defects, which can and should be considered for application to familial clusters of disease including germ line mutations of the cell-cell adhesion molecule E-cadherin.5

We believe therefore that even in those patients with Barrett's oesophagus who are fit for surgery further selection for repeated endoscopic surveillance should be undertaken. In particular, a combination of clinical criteria, and, perhaps in the near future, genetics, can be used to stratify for surveillance those at high risk of Barrett's adenocarcinoma.

References

  • 1.Macdonald CE, Wicks AC, Playford RJ. Final results from 10 year cohort of patients undergoing surveillance for Barrett's oesophagus: observational study. BMJ. 2000;321:1252–1255. doi: 10.1136/bmj.321.7271.1252. . (18 November.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.McGarrity TJ. Barrett's oesophagus: the continuing conundrum. BMJ. 2000;321:1238–1239. doi: 10.1136/bmj.321.7271.1238. . (18 November.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Jankowski J, Wright NA, Meltzer S, Triadafilopoulos G, Geboes K, Casson A, et al. Molecular evolution of the metaplasia dysplasia adenocarcinoma sequence in the esophagus (MCS) Am J Pathol. 1999;154:965–974. doi: 10.1016/S0002-9440(10)65346-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Jankowski J, Harrison RF, Perry I, Balkwill F, Tselepis C. Barrett's metaplasia. Lancet. 2000;356:2079–2085. doi: 10.1016/S0140-6736(00)03411-5. [DOI] [PubMed] [Google Scholar]
  • 5.Jankowski J, Perry I, Harrison RF. Gastro-oesophageal cancer: death at the junction. Understanding changes at the molecular level could lead to screening opportunities. BMJ. 2000;321:463–464. doi: 10.1136/bmj.321.7259.463. [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2001 May 5;322(7294):1124.

It is too early to dismiss surveillance programmes

Christopher Paul Wild 1,2, David Forman 1,2

Editor—Macdonald et al conclude that current surveillance strategies for patients with Barrett's oesophagus provide no benefit in terms of reduced risk of mortality from oesophageal adenocarcinoma.1-1 In their annual surveillance programme of 143 patients with an average follow up period of 4.4 years, cancer was diagnosed in only one patient, as a result of a biopsy taken within the programme. Although Barrett's oesophagus confers a 30-fold to 125-fold increased risk of oesophageal adenocarcinoma compared with people without Barret's oesophagus, the comparatively low absolute incidence of cancer (around 1 per 100 patient years) results in few cancers being detected by routine surveillance.1-2

Macdonald et al also say that molecular markers may identify patients at greatest risk of developing cancer. This view is supported by our data from a similar surveillance cohort in Leeds, United Kingdom, in which we used cyclin D1 overexpression as a marker of risk.1-3 In this prospective surveillance cohort, 307 patients with specialised columnar (intestinal) epithelium were included and had an endoscopic examination annually. A total of 12 incident cases of oesophageal adenocarcinoma were detected between 1984 and 1995 with a mean follow up of 4.3 years. These patients with cancer were matched by age, sex, length of columnar epithelium, and length of follow up with up to six patients from the cohort who did not develop adenocarcinoma. The biopsies obtained at recruitment were stained for cyclin D1 overexpression, and cases were found to be approximately seven times more likely to be positive for the marker than controls (odds ratio 6.85, 95% confidence interval 1.57 to 29.91; P=0.0106).

These results are promising, but there was a significant prevalence of positive cyclin D1 staining in biopsies from patients with Barrett's disease who had not yet developed cancer, 14 of 49 (29%), and biopsies from four of eight of those who did develop adenocarcinoma did not stain positive at recruitment. Therefore the sensitivity and specificity of the cyclin D1 marker were 67% and 71%, respectively. Nevertheless, if this prevalence of cyclin D1 positive staining applied to the whole cohort and had been used as a criterion for entry into the surveillance programme then about 90 people would have been subject to follow up with the detection of eight cancers.

Given that multiple genetic alterations are implicated in the natural history of Barrett's oesophagus and adenocarcinoma,1-4 a combination of carefully validated biomarkers might improve still further the predictive value of the molecular approach. The rising incidence of the disease and the advances in the understanding of its molecular pathology suggest that it is premature to dismiss refined surveillance programmes for early detection and more effective treatment of this cancer.

References

  • 1-1.Macdonald CE, Wicks AC, Playford RJ. Final results from 10 year cohort of patients undergoing surveillance for Barrett's oesophagus: observational study. BMJ. 2000;321:1252–1255. doi: 10.1136/bmj.321.7271.1252. . (18 November.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 1-2.Kim R, Weissfeld JL, Reynolds JC, Kuller LH. Etiology of Barrett's metaplasia and esophageal adenocarcinoma. Cancer Epidemiology Biomarkers Prev. 1997;6:369–377. [PubMed] [Google Scholar]
  • 1-3.Bani-Hani K, Martin IG, Hardie LJ, Mapstone N, Briggs JA, Forman D, et al. Prospective study of cyclin D1 overexpression in Barrett's esophagus: association with increased risk of adenocarcinoma. J Natl Cancer Inst. 2000;92:1316–1321. doi: 10.1093/jnci/92.16.1316. [DOI] [PubMed] [Google Scholar]
  • 1-4.Jankowski JA, Harrison RF, Perry I, Balkwill F, Tselepis C. Barrett's metaplasia. Lancet. 2000;356:2079–2085. doi: 10.1016/S0140-6736(00)03411-5. [DOI] [PubMed] [Google Scholar]
BMJ. 2001 May 5;322(7294):1124.

Patients need to be appropriately selected for follow up

Richard A Carr 1

Editor—Macdonald et al present results of a 10 year cohort of patients diagnosed with Barrett's oesophagus.2-1 Seven (5.2%) of 134 deaths were related to carcinomas of stomach or oesophagus in a median follow up period less than 10 years, and only five of 409 patients had evidence of mild dysplasia. Macdonald et al concluded that annual surveillance (143 patients for an average of 4.4 years) was not of direct benefit to any individual patient.2-1

The 5.2% death rate from oesophageal or stomach carcinoma in less than 10 years of follow up is fairly high, and risk factors may have been modified by treatment. The time interval between development of Barrett's metaplasia and carcinoma may be far in excess of the 10 year follow up in their series. Barrett's oesophagus is only one of many predisposing factors in the development of malignancy, and the natural history of the condition is poorly understood.

Centres that have recruited large numbers of patients into surveillance programmes should audit the value of their programmes and publish the results so that meta-analyses can be performed. But it would require a huge number of deaths and analyses of subgroups—for example, young patients with Barrett's oesophagus and those with dysplasia—before results could be used to advise individual patients regarding the benefit of surveillance. Although there are many consistent variables in the pathways to developing malignancy, there are also many individual factors—both genetic and environmental—that suggest that follow up should be tailored to individual patients' circumstances.

If I were 34 and had Barrett's oesophagus would I want regular surveillance? Maybe not on the basis of current evidence, but if I had some adverse genetic, environmental, or pathological features I might think differently. The knowledge might also encourage modifications in lifestyle or treatment of the risk factors and early presentation if I developed symptoms. The ratio of cost to benefit for routine surveillance of uncomplicated Barrett's metaplasia in medically fit, symptom-free patients may not be good for the NHS (underresourced in terms of cash and staff) when a second class service prevails. In a well resourced service the conclusion might be different particularly for appropriately selected patients.

References

  • 2-1.Macdonald CE, Wicks AC, Playford RJ. Final results from 10 year cohort of patients undergoing surveillance for Barrett's oesophagus: observational study. BMJ. 2000;321:1252–1255. doi: 10.1136/bmj.321.7271.1252. . (18 November.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2001 May 5;322(7294):1124.

Appropriate practice must be studied

Ian L P Beales 1

Editor—The paper by Macdonald et al will be quoted by those who believe that screening for oesophageal adenocarcinoma in columnar-lined oesophagus is a waste of effort and resources.3-1 The data presented do not add anything useful to the debate on the effectiveness of screening if modern surveillance programmes are being discussed.

The protocol used by Macdonald et al is inappropriate given current understanding of columnar-lined oesophagus and does not match modern recommendations. During the course of their study it was reported elsewhere that it was the presence of specialised intestinal metaplasia in the oesophagus which was the risk factor for malignant transformation.3-2 By performing surveillance on all patients with an endoscopically visible glandular oesophageal lining (both with and without intestinal metaplasia) Macdonald et al enrolled patients not at risk of malignant transformation. Thus it is not surprising that the intensive annual endoscopy protocol had such a low yield overall. Streitz et al reported that screening endoscopies in patients with columnar-lined oesophagus, at a mean interval of 17 months, detected one cancer per 73 patient years of follow up and overall was more cost effective than mammographic screening for breast cancer.3-3 The current recommendations of the American College of Gastroenterology for screening suggest endoscopy at intervals of two to three years in those with intestinal metaplasia.3-4 The findings of Macdonald et al, although of historical interest, do therefore not provide useful data on which to discuss current practice.

An additional point that deserves comment is the prominent and incorrect use of the term Barrett's oesophagus. By placing this in the title, Macdonald et al have helped perpetuate the confusion surrounding the diagnosis. Their use of the historical definition is at odds with the modern definition (a change in the appearance of the oesophageal mucosa recognisable at endoscopy with intestinal metaplasia on histology).3-3,3-5 Many of the patients in the study did therefore not really ever have Barrett's oesophagus. The term is so well established that it may be impossible to remove it from the lexicon, but it would be preferable if the more precise descriptive terms—columnar-lined oesophagus with or without specialised intestinal metaplasia and specialised intestinal metaplasia at the gastro-oesophageal junction—were used in future.3-5

Although I agree with Macdonald et al that screening policies for oesophageal adenocarcinoma require careful scrutiny for their effectiveness, appropriate practice must be studied.

References

  • 3-1.Macdonald CE, Wicks AC, Playford RJ. Final results from 10 year cohort of patients undergoing surveillance for Barrett's oesophagus: observational study. BMJ. 2000;321:1252–1255. doi: 10.1136/bmj.321.7271.1252. . (18 November.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3-2.Reid BJ. Barrett's esophagus and esophageal adenocarcinoma. Gastroenterol Clin North Am. 1991;20:817–834. [PubMed] [Google Scholar]
  • 3-3.Streitz JM, Jr, Ellis FH, Jr, Tilden RL, Erickson RV. Endoscopic surveillance of Barrett's esophagus: a cost-effectiveness comparison with mammographic surveillance for breast cancer. Am J Gastroenterol. 1998;93:911–915. doi: 10.1111/j.1572-0241.1998.00275.x. [DOI] [PubMed] [Google Scholar]
  • 3-4.Sampliner RE. Practice guidelines on the diagnosis, surveillance, and therapy of Barrett's esophagus. The Practice Parameters Committee of the American College of Gastroenterology. Am J Gastroenterol. 1998;93:1028–1032. doi: 10.1111/j.1572-0241.1998.00362.x. [DOI] [PubMed] [Google Scholar]
  • 3-5.Spechler SJ, Goyal RK. The columnar-lined esophagus, intestinal metaplasia, and Norman Barrett. Gastroenterology. 1996;110:614–621. doi: 10.1053/gast.1996.v110.agast960614. [DOI] [PubMed] [Google Scholar]
BMJ. 2001 May 5;322(7294):1124.

Authors' response

R J Playford 1,2, A C B Wicks 1,2, C E Macdonald 1,2

Editor—We agree about the potential usefulness of molecular markers, but which markers will prove to be of clinical value is debatable. Further research may provide a way to target patients at the highest risk of malignant transformation, who might gain from some form of surveillance programme.

Carr makes some interesting points about selecting suitable patients for surveillance. In our cohort, however, future analyses will provide information only about cause of death and will not influence the ratio of cost to benefit of our programme because nearly all the patients have left for the reasons specified.

Screening and surveillance are different. We examined the benefit of surveillance of a population at known increased risk as opposed to screening a general population to detect an unidentified problem.

The original case reports described by Barrett did not mention intestinal metaplasia. Its presence may be an absolute requirement for subsequent development of adenocarcinoma and its detection is therefore considered by some to be essential in diagnosing Barrett's oesophagus. However, nearly all cases of oesophageal mucosal metaplasia containing columnar epithelium will also have intestinal metaplasia if enough samples are taken (N Shepherd, eighth united European gastroenterology week, Brussels, November 2000). Beales's argument is therefore circular as nearly all of our patients will have had intestinal metaplasia even if it was not visible at biopsy.

We consider that the guidelines issued by the American Association of Gastroenterology4-1 are not sufficiently evidence based to prove their value. Blindly to follow them without auditing the ratio of cost to benefit is unwise in the current evidence based culture. The guidelines are similar to those of the World Congress, and following them would have been unlikely to have influenced patient outcome in our cohort for the reasons we gave. The survey we quoted showed that few centres in the United Kingdom follow either of the guidelines. Practice is also far from perfect (defined as following the American guidelines4-1)—for example, an important factor determining how often an endoscopy was performed and how many biopsy specimens were taken was whether the doctor was being paid as a “fee for service.”4-2

Our protocol (annual surveillance with four-quadrant biopsies at midpoint plus additional biopsy at strictures or ulcers) did not seem to be beneficial. Many centres in the United Kingdom currently use similar approaches, so Beales's suggestion that our work has only historical interest is therefore erroneous. Other centres have recently come to a similar, disappointing, conclusion about the value of their programmes.4-3

The British Society of Gastroenterology is currently developing guidelines on surveillance of Barrett's oesophagus and has found little strong evidence to support recommendations (personal communication). Guidelines may reduce litigation, but they do not necessarily help patients. Large formal trials, with differences in survival as the end point, are needed to collect evidence.

References

  • 4-1.Sampliner RE. Practice guidelines on the diagnosis, surveillance, and therapy of Barrett's esophagus. The Practice Parameters Committee of the American College of Gastroenterology. Am J Gastroenterol. 1998;93:1028–1032. doi: 10.1111/j.1572-0241.1998.00362.x. [DOI] [PubMed] [Google Scholar]
  • 4-2.Gross CP, Canto MI, Hixson J, Powe NR. Management of Barrett's esophagus: a national study of practice patterns and their cost implications. Am J Gastroenterol. 1999;94:3440–3447. doi: 10.1111/j.1572-0241.1999.01606.x. [DOI] [PubMed] [Google Scholar]
  • 4-3.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 Gastroenterol. 2000;35:10–16. [PubMed] [Google Scholar]

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Publishing Group

RESOURCES