Barrett's oesophagus : towards improving surveillance

Abstract

Introduction In Barrett's Oesophagus the squamous epithelial lining of the distal oesophagus transforms into a columnar mucosa by a metaplastic process which is strongly associated with the reflux of acid and bile. Population studies have shown that this condition is a strong risk factor in the development of oesophageal adenocarcinoma, which has a notoriously bad prognosis when discovered late. The incidence of this cancer has increased dramatically over the past few decades and this has caused alarm in the ranks of health-care providers. Barrett's Oesophagus is thought to lead to carcinoma through a dysplastic phase in which the "unstable" specialised intestinal mucosa becomes progressively atypical and in time frankly neoplastic. The rationale for surveillance is the detection of early neoplasia which is amenable to surgery, hence offering a chance of cure. Surveillance practices vary because the reported dysplasia and cancer incidences themselves vary within, as well as between populations. Moreover, the time-line for this metaplasia-dysplasia-carcinoma sequence seems erratic. There is no consensus over the utility of surveillance, with some authors voicing doubt over its cost-effectiveness. This uncertain scenario has prompted considerable research into the development of reliable methods of screening and effective and noninvasive treatment modalities.

Aim In this dissertation surveillance for Barrett's Oesophagus is tackled from the screening and diagnostic perspective as well as the therapeutic one. It is postulated that improvement of the quality and accuracy of the diagnostic test will result in a higher yield of early neoplastic lesions which are amenable to curative therapy. Likewise, improvements in the therapeutic modalities are expected to enhance the scope for further surveillance.

Methodology The present-day screening standard is white-light endoscopy with biopsy which is offered at regular endoscopic intervals guided by the histological history. In this project, a systematic review of published Barrett's Oesophagus series was undertaken in order to compare follow-up practices and their resultant outcomes. Using a set of stringent inclusion criteria 30 English-language articles were chosen and the reported series were divided into three groups. In total, 3176 patients were considered. In the group of series employing regular endoscopic follow-up with a rigorous multi-level biopsy protocol, 75% of the neoplastic lesions identified, were high-grade dysplasia or T1 cancers. In the other two groups of series (random biopsies and no surveillance) early neoplasia was much less common (45% and 15% respectively).

In this project, this observation was further tested in a prospective cohort study performed at the Glasgow Royal Infirmary wherein two demographically similar groups of patients (n=362) were followed up with annual endoscopies. The cohorts differed in their biopsy technique i.e. the first group was sampled with a protocol of biopsies from the four quadrants of the oesophagus at 1 or 2 cm intervals, while the second group was biopsied at random. Whereas low-grade dysplasia was diagnosed reliably and early on in the first group (35 prevalent and 4 incident cases), its detection was haphazard and not as frequent when using random biopsies (6 prevalent and 12 incident cases) (P=0.0003). In the protocol biopsy cohort 11 cases of high-grade dysplasia and T1 cancer were diagnosed with confidence and treated with curative intent. No invasive cases emerged in this group. In contrast, there were no early lesions in the random biopsy group and 3 invasive cancers received treatment with palliative intent. In this project, a novel screening tool, in the form of exhaled ethane concentration, was assessed. Breath testing is an eminently simple and non-invasive screening modality. Clearly, a breath test for dysplasia and early cancer would be a considerable achievement. Recently, exhaled ethane gas concentration has been used as a surrogate marker for oxidative stress in a number of benign and malignant conditions. In this dissertation a proof of concept clinical experiment, exploring the relationship of exhaled ethane concentrations and upper gastrointestinal cancers, will be described. The author compared breath samples from 20 patients with advanced oesophago-gastric cancers, with those from 10

age- and sex-matched voluntary healthy controls. Ethane concentrations were measured using a highly accurate (to 0.2 parts per billion) tunable diode laser spectroscope. No statistically or clinically useful relationship was identified. Oesophagectomy is the standard treatment for high-grade and early carcinomatous lesions. This is a major and costly procedure which can only be offered to patients who prove to be physically fit for major surgery. Unfortunately, a significant proportion of patients with dysplastic and neoplastic Barrett's Oesophagus will have co-morbidities which render them unsuitable for this type of surgery. For these patients less invasive (and cheaper) treatment options, such as endoscopic mucosal resection (EMR), laser, photodynamic and radiofrequency ablative (RFA) modalities, may offer a suitable alternative to surgery. Being much less invasive than oesophagectomy and tolerated well under conscious sedation, endoscopic ablative procedures may be offered to the vast majority of patients. In this project, a series of 30 patients with oesophageal high-grade dysplasia and early cancer treated with a combination of EMR and Nd:YAG laser is presented. In a retrospective study the outcome of these patients was compared with that of 8 stage-matched patients treated with oesophagectomy. There was no statistically significant difference in the three year disease-specific (and overall) survival between the 2 groups.

Radiofrequency ablation is at the cutting edge of endoscopic ablation and in this project, the author describes a series of 58 patients with a diagnosis of Barrett's high-grade dyspalsia or intramucosal cancer treated with a combination of EMR to raised neoplastic areas followed by intensive RFA to the residual flat metaplastic segments. Using these complimentary techniques, resolution of dysplasia was achieved in almost 90% of patients receiving the full ablative programme, whereas metaplasia was eradicated in nearly 70%.

Conclusion The aim of this research project was to improve the surveillance of patients with Barrett's Oesophagus by improving the diagnostic and therapeutic arms of surveillance programmes. The systematic review and cohort study assessing biopsy technique suggest that a rigorous biopsy protocol yields more curable neoplastic lesions than a random biopsy approach. EMR, laser and RFA are novel and relatively non-invasive ablative modalities which in the treatment of high-grade dysplasia and intramucosal cancer have promising and encouraging results. These techniques may emerge as suitable alternatives to oesophagectomy.

Keywords: Barrett's oesophagus, dysplasia, cancer, surveillance, breath testing, endoscopic mucosal resection, laser, radiofrequency ablation