PDF icon Download PDF
Endoscopy 2015; 47(07): 579-581
DOI: 10.1055/s-0034-1392105
Editorial
© Georg Thieme Verlag KG Stuttgart · New York

What is the ideal ablation therapy for Barrett’s esophagus?

Raf Bisschops
Department of Gastroenterology, University Hospitals Leuven, Leuven, Belgium
› Author Affiliations
Further Information

Publication History

Publication Date:
26 June 2015 (online)

Also available ateRef
   Permissions and Reprints
Preview

Over the past decade, endoscopic resection of early mucosal neoplasia with subsequent ablation of the remainder of the Barrett’s epithelium has replaced subtotal esophagectomy as the first-line treatment for early mucosal Barrett’s-associated neoplasia. Endoscopic resection has been shown to be oncologically safe and comparable to surgery [1]. However, metachronous lesions develop in up to 30 % of patients, and therefore a two-step approach involving the removal of all visible neoplasia followed by ablation of the remaining Barrett’s mucosa is adopted in order to reduce this risk significantly [2].

In this issue of Endoscopy, Canto et al. [3] report, for the first time, long-term efficacy and safety data on carbon dioxide (CO2) cryotherapy for the ablation of Barrett’s esophagus, through a retrospective analysis of prospectively collected data. Consecutive patients with high grade dysplasia or early adenocarcinoma after removal of visible lesions were included in the study. Patients were stratified into two groups: those receiving cryotherapy as primary treatment for residual Barrett’s esophagus after endoscopic resection of visible lesions (n = 21), and those receiving cryotherapy as a rescue therapy after failed endoluminal treatment including endoscopic resection, photodynamic therapy (PDT), and radiofrequency ablation (RFA) (n = 47). For the entire group, complete remission for dysplasia at 1 year was 89 % (95 % for primary treatment and 86 % for rescue treatment). At 3-year follow-up, complete remission for dysplasia and early cancer was 100 % for primary treatment and 82 % for rescue treatment. Complete remission of intestinal metaplasia after 1 year was 55 % (70 % for primary treatment and 48 % for rescue treatment). This improved at the 3-year follow-up, where 67 % of the whole group were disease free (85 % for primary treatment and 59 % for rescue treatment). The study also showed a favorable safety profile for CO2 cryotherapy, with adverse events being limited to postprocedural pain in four patients (6 %) and stricture in one patient (2 %). Two serious adverse events were reported and required hospitalization: one suspected perforation (minute sub-diaphragmatic air) and one postprocedural bradycardia. Buried glands during follow-up were noted in 7 % of patients.

What does this study add to our knowledge of treatment for dysplastic Barrett’s esophagus? First, the safety data appear to be very favorable, especially given that 48 % of the patients had a history of upper gastrointestinal bleeding or were at risk of bleeding, including one patient with esophageal varices. No bleeding occurred after CO2 cryotherapy. Therefore, there might be a place for this technique in patients who are at a higher risk of bleeding or when RFA is contraindicated in patients with varices. Second, the majority of patients underwent a rescue treatment for failed endoluminal therapy. These are the first results to show that CO2 cryotherapy may be useful in patients who are refractory to other treatments. However, it has to be emphasized that 1) it is not clear from the study how proton pump inhibitor intake had been controlled during previous RFA treatment and 2) three treatments of RFA would not have been considered a failure in the EURO-II, AIM dysplasia, or SURF trials, where a maximum of five RFA sessions was permitted [4] [5] [6]. Third, the study confirms that meticulous follow-up after treatment of patients with degenerated Barrett’s esophagus is mandatory to ensure a good oncological outcome. In the Canto study, the risk of dysplasia in recurrent intestinal metaplasia was as high as 40 %. This is much higher than what is reported for APC and RFA, but may very well be due to the selection of patients with refractory disease in 68 % of the cohort.

A variety of methods for ablation of Barrett’s esophagus has been assessed over past decades. The ideal ablation tool should be safe, easy to apply, and have excellent long term outcomes for sustained endoscopic and histological remission of all intestinal metaplasia. A perfect tool with a 100 % remission rate and 0 % complication rate does not exist. In the era of evidence-based medicine, a literature search reveals clearly that the best scientific data and systematic assessment of an ablation technique are only available for RFA [4] [5] [6] [7] [8] [9] [10] [11]. Indeed, the technique has been assessed from the bench [12] [13] and feasibility studies [7] to well-conducted randomized controlled trials and prospectively controlled multicenter trials, showing high efficacy and safety for ablation of flat high grade dysplasia [6], low grade dysplasia [5], and, recently, for ablation of residual Barrett’s mucosa after endoscopic resection of visible lesions [4] [6].

The company behind the RFA technique has set a new benchmark against which the efficacy and safety of future promising techniques must now be assessed. However, the study by Canto et al. exemplifies the shortcomings of the manufacturers in terms of device testing. For a long time, companies have been bombarding endoscopists with new CE-labelled or FDA-approved devices, exposing patients to techniques that have not been comprehensively assessed in a clinical study. The fact that long term data are the result of a retrospective analysis from a single centre, and that the study combined dosimetry assessment with clinical outcome, merely illustrate Dr Canto’s merit in gathering and analyzing the data in order to overcome the shortcomings of companies that failed to properly investigate this new, possibly promising, device.

To answer the question of the best ablation tool for Barrett’s esophagus, only one randomized controlled trial has compared two different techniques [14]. Stepwise radical endoscopic resection has been compared with staging endoscopic resection of visible lesions followed by RFA. Both techniques had a comparable outcome in terms of complete removal of dysplasia and Barrett’s esophagus; however, stepwise endoscopic resection was technically more challenging and resulted in significantly more stenoses and, as a consequence, a higher number of interventions compared with RFA.

For other ablation methods, there are currently no comparative data available to make a final conclusion on whether one ablation device is better than another. However, PDT has shown insufficient efficacy for complete remission of intestinal metaplasia and was associated with an unacceptable rate of complications, mainly stenosis and phototoxicity. In addition, there is a significant cost associated with PDT and it is no longer widely used [15].

Recently, a randomized controlled trial showed a significant reduction of recurrent metachronous disease after argon plasma coagulation (APC) of residual Barrett’s mucosa in patients after endoscopic resection for high grade dysplasia and early mucosal cancer [2]. After a mean of 4 APC sessions, complete removal of Barrett’s esophagus was achieved in 26 /33 patients (79 %). The study by Canto et al. [3] suggests a similar efficacy for CO2 cryotherapy but provides a longer follow-up period. Although the device cost of APC and CO2 cryotherapy may be lower compared with RFA, other factors need to be taken into account, such as operator dependency of the technique and number of treatment sessions. In the study by Canto et al., the goal of each treatment session was to freeze all visible columnar mucosa and perform focal or circumferential ablation in patients with a Barrett’s segment length of up to 8 cm. Although 68 % of patients had previously undergone other ablation treatments, the median number of sessions for Barrett’s segments of 3 – 8 cm was 4, increasing to 7 sessions for segments > 8 cm. This is in contrast to RFA data where a median of 2 – 3 treatments were required for complete remission [6] [8] [10]. Even for long Barrett’s segments, complete remission was achieved in 83 % of patients after a median of three RFA ablation sessions [11]. Similarly, it seems that APC as a spot-by-spot ablation technique would be too time-consuming in long-segment Barrett’s esophagus. In addition, CO2 cryotherapy, and in particular APC, are very operator dependent both for energy settings and delivery. This may explain the variable results for APC in the literature [16]. To date, good efficacy data have been provided mainly by single-center or single-operator studies, and results need to be confirmed in multicenter prospective trials.

Considering the simplicity of the RFA technique, the fact that energy delivery is operator independent, and the excellent long term clinical results, it appears that, at this time, RFA will be hard to beat as the preferred ablation tool for long Barrett’s segments. Even with the publication of studies assessing the effect of new ablation devices, the jury is still out as to which device is preferable for focal ablation of residual Barrett’s esophagus or short Barrett’s segments. Only studies with a head-to-head comparison, assessing hard clinical end points and cost effectiveness, will provide a definitive answer to this question.