Case provided by:
Prof. dr. med. Stefan Seewald
Center of Gastroenterology, Hirslanden Clinic Zürich, Zürich, Switzerland

A 60-yr old patient presented with a long segment Barrett’s esophagus (C7M8) and a reflux stenosis at 29cm. The patient was prescribed high-dose proton-pump inhibitors and was rescheduled for endoscopy 6 weeks later. Biopsies from the stenosed area, at the 12 o’clock position, showed high-grade intraepithelial neoplasia.

To get an impression of the lesion and its accessibility for endoscopic resection, it was tested if the mucosa could be suctioned into an ER-cap. This proved to be no problem.

However, the fibrotic ring may make impede with the endoscopic resection procedure. In addition, scarring after an endoscopic resection in this area may result in a more severe stricture, which may make subsequent circumferential ablation with the HALO360 system very difficult.

The video nicely illustrates this interesting case.

The question is:
What would be the best treatment strategy for this patient? Endoscopic resection or radiofrequency ablation?

Get the Flash Player to see this player.

Suggestion provided on behalf of the RFA Academia training committee by:
Dr. Jacques Bergman
Dept. of Gastroenterology and Hepatology, Academic Medical Center, Amsterdam, the Netherlands

It may be advisable to perform stepwise dilatation to 18 mm and wait for any laceration to heal. Then, RFA and endoscopic resection can be performed in the same treatment session. For this, the lesion first needs to be delineated with nicely visible electrocoagulation marks. After marking the lesion, RFA can be performed as usual, using the 18-mm wide HALO³⁶⁰ catheter. Since the patient was dilatated to 18 mm, no sizing procedure is necessary, since this could only increase the risk of lacerations during the procedure.

Immediately after the RFA procedure, the electrocoagulation marks use still visible, and they can guide an endoscopic resection of the suspicious area, using either multiband mucosectomy or the ER-cap technique with submucosal lifting.
The superficial layers of the mucosa may be damaged by the RFA procedure, yet the true staging information of the endoscopic resection: invasive cancer, differentiation grade, vertical resection margins, sm-invasion and vascular invasion, can still be assessed.

By performing RFA and endoscopic resection in the same session, after dilatation of the stricture, you prevent that the stenosis becomes worse after the endoscopic resection and that subsequent RFA is more difficult.

Case provided by:
Dr. Adolfo Parra Blanco
Hospital Universitario Central de Asturias, Oviedo, Spain

A 76-yr old lady without any relevant past medical history, was recently diagnosed with multifocal inconspicuous high-grade dysplasia in a C10M11 Barrett´s esophagus (Figure 1).  Three sessions of RFA were applied so far (1x HALO³⁶⁰ 12J/cm²), and 2x HALO⁹⁰ sessions 15J/cm2 (Figure 2). In all three sessions it was difficult to pass the ablation catheters to the oesophagus, probably because of osteophyte compression. In the first RF session (with the HALO³⁶⁰) ablation was attempted from 21 cm, but the ablation balloon slipped inwards when inflated, probably because of reduced diameter of the esophagus at that location. Then at the second RFA procedure, 2 cm circumferential Barrett´s mucosa was found at 21 cm, and ablation was performed with HALO⁹⁰ (15J/cm2). At the time of the third procedure, three Barrett´s islands were found at 21 cm, the largest diameter of the whole lesions together being 2 cm. It was very difficult to pass to the esophagus the HALO⁹⁰ and a guidewire was required. A double ablation was performed on those remaining islands, and some others including the cardia. All visible islands were successfully targeted, and well burnt. Then the endoscopist balanced the risk of leaving some Barrett´s mucosa left in this session if the second double application of RFA was not performed, against the risk of cervical oesophagus perforation if the HALO⁹⁰ was withdrawn to clean the electrode and reinserted. The endoscopist decided not to perform a second “double” ablation, and the procedure was finished.

The questions are:

- Is one double ablation with HALO⁹⁰ enough in cases with just remaining islands, if the endoscopist is confident that the desired areas have been well targeted?

- Would it be possible to perform only one double ablation regimen in patients with marked difficulty to pass the ablation HALO⁹⁰ catheter, in order to reduce the risk of perforation?

Suggestion provided on behalf of the RFA Academia training committee by:
Dr. Jacques Bergman
Dept. of Gastroenterology and Hepatology, Academic Medical Center, Amsterdam, the Netherlands

In some cases it can indeed be difficult to pass the HALO⁹⁰ catheter and in the case of narrowing it is sometimes even necessary to perform dilatation to allow passage of the HALO⁹⁰ catheter to reach the residual Barrett mucosa. The optimal use of the HALO⁹⁰ has not yet been established, but until now we have found the 2x2x15J/cm² regimen more effective than 2x 15J/cm², and we regression percentage is higher when using the 2x2x regimen. But in some cases, 2x 15J/cm² may also be enough.

At this moment, the Amsterdam group is comparing the 2x2x15 J/cm² with cleaning in between the ablation passes, with 3x 15J/cm² (3 burns immediately after each other). If the latter turns out to be equally effective as the 2x2x15J/cm² that will increase the practical use of the HALO⁹⁰ a lot, but we have to await the results of this study.

In this case, however, you can really only wait and see what the result is. It sounds like this patient with her long Barrett segment responded really good to the prior HALO³⁶⁰ and HALO⁹⁰ treatment, so your last treatment may have been effective as well.

I fully agree with you that in this case the risk of a second introduction is not balanced with a higher efficacy of the ablation by a second 2×12 ablation. You may also consider to use APC to patch up small islands the remain at further follow-up visits in this case.

And maybe in the future you can use the 3x15J/cm² regimen if you are having these kind of difficulties introducing the HALO⁹⁰ catheter again.

Case provided by:
Dr. Adolfo Parra Blanco
Hospital Universitario Central de Asturias, Oviedo, Spain

An elderly lady with a C10M11 Barrett´s esophagus with high-grade dysplasia in flat mucosa, without any definite lesions. The patient was treated initial HALO³⁶⁰. At follow-up 3 months after the initial RFA session, 70% of the Barrett’s segment had been eradicated. There were still a number of circular areas larger than 2 cm that needed to be treated with additional HALO³⁶⁰ ablation. However, somewhat irregular mucosa was detected at 28-29 cm (Figure 1). Biopsies from this area were obtained and treatment was postponed to await the histological diagnosis of the biopsies.

The question is:

How can the irregular mucosa be explained, and could the RFA treatment have been performed in this case?

Figure 1, images obtained with iScan, A: Cardia, B & C: Irregular mucosa at 29 and  28 cm.

Suggestion provided on behalf of the RFA Academia training committee by:
Dr. Jacques Bergman
Dept. of Gastroenterology and Hepatology, Academic Medical Center, Amsterdam, the Netherlands

If the patient has undergone thorough endoscopic work-up and had inconspicuous high-grade dysplasia in the absence of any visible lesions, it is unlikely that new lesions will develop so quickly after the initial RFA treatment.

Most likely the mucosa irregularities reflect inflammatory changes after the prior RFA. Nevertheless, one should keep an eye out for any lesions that pop up during the RFA protocol.

In this case taking 1-2 biopsies and then proceeding with the HALO³⁶⁰ procedure would have been fine. On the other hand, postponing treatment until the remaining Barrett’s mucosa has become less inflamed (and thus more flat) may also have its advantages since it (theoretically) increases the success rate of the 2nd HALO³⁶⁰ treatment.

Case provided by:
Dr. Adolfo Parra Blanco
Hospital Universitario Central de Asturias, Oviedo, Spain

In our center, we are doing all RFA cases under deep sedation with Propofol by the anesthesiologist. However, I know that other centers also perform RFA under conscious sedation. An advantage of endoscopist-administered sedation (e.g. midazolam, fentanyl) is of course that it is more efficient.

The question is:
I would be interested to read about the different types of sedation that are used for RFA treatment, and what the experiences are with these different sedation approaches?

Case provided by:
Dr. Adolfo Parra Blanco (who was consulted about this case by a colleague)
Hospital Universitario Central de Asturias, Oviedo, Spain

A type 0-IIa lesion in a Barrett’s esophagus was removed with multiband mucosectomy. After the first resection, the adjacent mucosa was suctioned into the cap and a rubber band was released. However, the second band was partly applied on the resection wound from the first resection, and the endoscopist could see the tissue from the first resection wound in the pseudopolyp that was formed (Figure a & b). Given his lack of experience with the multiband mucosectomy technique, the endoscopist did not resect the pseudopolyp, because he was afraid to perforate the esophagus.

The questions are:

- Could the resection have been finished without perforating the esophagus?

- Is there a risk on development of a delayed perforation, since the rubber band is still in situ?

Suggestion provided on behalf of the RFA Academia training committee by:
Dr. Jacques Bergman
Dept. of Gastroenterology and Hepatology, Academic Medical Center, Amsterdam, the Netherlands

As long as the band stays on and doesn’t slip off immediately, the risk that there is muscular tissue captured in the band is really small if the patient has no history of anything that may have caused esophageal scarring.

Overlap is acceptable, and even advisable to prevent residual tissue bridges, as long as it is less than 25%. In case of doubts:

1) push gently with the edge of the cap onto the band so that any muscle tissue can slip out;

2) wait a bit longer and inflate the esophagus (has the same effect as pushing onto the band);

3) place and close the snare above the band.

A combination of the above measures always works.

In this case were the psuedopolyp with the band was left in place, part of the tissue encaptured may partially slip out, the other part will necrose and fall of. This may be a good approach for treating patients with lesions on the top of esophageal varices. In this case I think I would have used the measures 1-3 to resect the lesion.