Endoscopy 2020; 52(01): 5-6
DOI: 10.1055/a-1026-4514
© Georg Thieme Verlag KG Stuttgart · New York

Barrett’s surveillance: take the full package or leave it altogether

Referring to Roumans CAM et al. p. 17–28
Massimiliano di Pietro
MRC Cancer Unit at the University of Cambridge, Cambridge, United Kingdom
› Author Affiliations
Further Information

Publication History

Publication Date:
18 December 2019 (online)

Barrett’s esophagus (BE) is the only known precursor to esophageal adenocarcinoma (EAC), the solid malignancy with the steepest increase in incidence rate over the past 4 – 5 decades and one of the most aggressive cancers, with an overall 5-year survival of only 15 % [1]. It is therefore intuitive that most gastroenterology guidelines recommend endoscopic surveillance of BE with the aim of detecting neoplasia at the pre-invasive (dysplasia) or early invasive (T1) stage [2 – 4]. This opens a window to offer patients timely treatment with minimally invasive modalities, such as endoscopic resection and/or ablation, which are highly effective for the eradication of BE and related neoplasia. Endoscopic treatment also offers the possibility of preserving the anatomy and avoiding the mortality and morbidity related to esophagectomy. It is not surprising that evidence indicates that endoscopic surveillance is associated with detection of cancer at earlier stage; however, studies looking at the impact of endoscopic monitoring on patient outcome are conflicting, with some studies suggesting that mortality from EAC detected within surveillance programs does not differ from that of patients diagnosed de novo with symptomatic cancer [5,6]. An interesting Dutch population study recently investigated the relationship between EAC mortality in patients who had a prior BE diagnosis and the adequacy of endoscopic surveillance, defined as adherence to recommended monitoring intervals [7]. Using patients with known BE who did not receive any surveillance as reference, it was found that only patients who received adequate surveillance had a significantly lower adjusted hazard ratio for EAC mortality (0.79, 95 % confidence intervals 0.64 – 0.92), whereas those who received inadequate monitoring did not benefit from it. These results help to explain why some studies have failed to find an association between improved cancer and endoscopic surveillance, if the latter is not performed adequately. However, surveillance interval is not the only key element of adequacy of clinical management, which also relies on strict adherence to biopsy and standardized endoscopic protocols.

“The clinical effectiveness of surveillance programs, in terms of reduction of cancer-related mortality, can only be guaranteed by strict adherence to guidelines.”

In an effort to shed light on the adequacy of clinical practice in BE management and to identify the factors that may predict poor performance, Roumans et al. conducted a meta-analysis that investigated adherence to guidelines in studies reporting on patients receiving endoscopic surveillance [8]. The study, which is published in this issue of Endoscopy, analyzed 19 different recommendations, divided into 5 domains, which included intended adherence to one guideline, surveillance interval, biopsy protocol, use of endoscopic landmarks for reporting, and histopathologic information. Surveillance intervals, Seattle protocol, and Prague classification were regarded as primary recommendations. Any study reporting on adherence to at least one of the 19 recommendations was included, as long as the patient cohort under investigation had not reached the end point of high grade dysplasia or cancer. The primary outcome of the study was variability in guideline adherence between studies explained by specific subgroup, for example type of practice, year of publication, type of data collection, and country. The secondary outcome was the association between adherence rates and explanatory parameters, related to patient, endoscopist or facility-related factors. A total of 475 publications were screened and 56 studies, published between 1997 and 2007, were included in the analysis. These consisted of 19 questionnaire-based cross-sectional studies, 26 retrospective studies, 6 prospective studies, and 5 with mixed methodology. The findings of the study are very interesting. Approximately a half of studies did not adhere to recommendations for both surveillance intervals and biopsy protocol. With regard to surveillance intervals, most of the nonadherence was caused by intervals that were too short, whereas most of the lack of adherence to the recommended number of biopsies was due to insufficient number. The heterogeneity in surveillance intervals mostly related to the type of practice, with university hospitals adhering more than community hospitals. In terms of biopsy protocol, it appears that the country of origin of the study had the biggest impact, with US centers being the top performers. The following factors were found to be associated with better adherence to guidelines: shorter BE segments, younger endoscopists, salaried employment, university hospital setting, and the availability of dedicated surveillance lists.

It can be argued therefore that in general practice there is a tendency to compensate for suboptimal biopsy sampling by scheduling shorter surveillance intervals. This has important implications for the cost-effectiveness of surveillance, as additional endoscopic procedures create an unnecessary burden on the health care system. Moreover, endoscopists are more likely to take insufficient numbers of biopsies when dealing with long Barrett’s segments in a community hospital setting. This finding strengthens the significance of the European Society of Gastrointestinal Endoscopy recommendation that patients with BE segments longer than 10 cm should be referred for surveillance in BE expert centers, to ensure adequacy especially when the risk of neoplastic degeneration is higher. These are more likely to be academic hospitals with dedicated surveillance programs, both of which are associated with adherence to biopsy protocol.

In summary this study suggests that the quality of BE surveillance is often suboptimal in general practice. The clinical effectiveness of surveillance programs, in terms of reduction of cancer-related mortality, can only be guaranteed by strict adherence to guidelines. In addition, the uncertainty about the quality of the endoscopic diagnosis when the Seattle protocol is not applied, leads to unnecessary procedures, undermining the cost-effectiveness of the intervention.