Improving practicality of radiofrequency ablation for eradication of Barrett*s mucosa: a randomized trial comparing two different treatment regimens for focal ablation using the HALO90 System

Radiofrequency ablation (RFA) is a new endoscopic ablation technique that has
been shown to be an easy, safe and effective treatment modality for complete
eradication of Barrett*s esophagus (BE) containing early neoplasia. Compared to
PDT and APC, RFA seems to be more easy to use, better tolerated by patients,
and is not associated with esophageal stricturing or the occurrence of buried
Barrett*s.
In RFA, the Barrett*s segment is ablated by radiofrequency energy through two
specially designed devices for circumferential and focal ablation respectively.
The HALO360 System consists of a balloon which contains a spindle-shaped
electrode on its outer surface. Balloons with different diameters and lengths
of electrodes are available. For focal ablation a cap-based electrode, the
HALO90 System is used. The instruments have been developed by BÂRRx Medical
Inc, Sunnyvale CA, USA and are FDA and EC approved for ablation of Barrett*s
mucosa.

Currently, most patients undergo primary circumferential ablation with a
balloon based electrode, the HALO360 System, followed by multiple focal
ablation sessions using a cap-based electrode, the HALO90 System. During these
HALO90 procedures, residual islands undergo 2 separate ablation passes, each
pass with two sequential *hits* at 15 J/cm2. In between these 2 ablation passes
the ablation zone is cleaned of its coagulum and the HALO90 catheter is removed
in order to clean the surface of the electrode. The actual HALO90 ablation is
relatively easy to perform. Subsequent cleaning of the ablation zone, removal,
cleaning and reintroduction of the catheter are, however, impractical and
uncomfortable to the patient. We propose a simplified HALO90 ablation procedure
in which cleaning of the ablation zone is left out, thereby reducing the number
of introductions with the endoscope and the HALO ablation device.
We hypothesize that this simplified HALO90 ablation procedure results in an
easier and faster ablation procedure, while maintaining efficacy and safety.

To develop a simplified HALO90 ablation protocol at the same energy level and
with compared efficacy and safety as the current protocol.

Patients scheduled for HALO90 ablation for BE (with or without neoplasia) after
prior circumferential ablation using the HALO360 System for BE with flat
low-grade dysplasia (LGD) or high-grade dysplasia (HGD) or for BE (with or
without neoplasia) after prior endoscopic resection (ER) for lesions containing
HGD or and early cancer are being asked for inclusion in the study.

Treatment protocol: HALO90 procedure
The BE is evaluated for the presence of strictures and the presence of residual
BE mucosa using high-resolution white light (WL) endoscopy and narrow band
imaging (NBI). The number, size (maximum and minimum diameter) and localization
(insertion depth of the endoscope and orientation in the endoscopic field) of
all islands and BE tongues are documented on video recordings and still images
(WL + NBI). Islands are then numbered sequentially from proximal to distal.
A minimum of one and a maximum of four islands per HALO90-procedure will be
randomized. Tongues and *large islands* (larger than the total surface of four
adjacent HALO90 treatment zones) are excluded from randomization. The treatment
regimen to which the first island is randomized will be the first treatment
regimen performed.
Eligible islands will be randomized to either the standard *double-double*
regimen at 15 J/cm2 (two ablation passes each with two *hits* at 15 J/cm2 and
with cleaning of the ablation zone in between the two passes), or a *single-
triple* ablation at 15 J/cm2 (single ablation with three *hits* at 15 J/cm2 and
no subsequent cleaning).
Islands will be randomized in couples (islands 1-2 and 3-4, or 1-2 and 3, or
1-2), always randomizing one island out of a couple and automatically
redirecting the other island of the couple to the other treatment regimen, to
guarantee an even spread of both treatment regimens within a HALO90-procedure.
This means that all patients with two or more eligible BE islands will undergo
both the standard HALO90 ablation regimen and the simplified HALO90 regimen.

Standard HALO90 ablation
regimen:

After mapping and randomization the Barrett*s segment is flushed with the
mucolytic agent acetylcysteine (1%) followed by flushing with tap water. The
endoscope is removed and the proximal esophagus is thoroughly inspected to
exclude the presence of a Zenker*s diverticulum that may make subsequent
introduction of the HALO90 cap difficult or dangerous.
The HALO90 cap is attached to the tip of the endoscope at the twelve o*clock
position and introduced into the distal esophagus. Visible islands are then
treated with 2x15 J/cm2 (40 Watt)): the cap is brought into close contact with
the target area and is then activated. The endoscope and the cap are kept in
position and immediately a second ablation of the same area is performed. After
ablation of an island, the endoscope is gently removed from the mucosa and
rotated to remove the coagulum by suctioning through the endoscope.
If the squamocolumnar junction (SQJ) has an irregular appearance the SQJ is
treated circumferentially, allowing an overlap of 5-10 mm between the adjacent
ablations. After all islands and Z-line have been ablated in this manner, the
necrotic debris is cleaned off by a combination of suctioning and irrigating
tap water. In addition, the HALO90 cap can be used to gently push off the
coagulum from the ablation zone. Subsequently, the ablated areas are cleaned by
vigorous flushing of water through a spraying catheter. After emptying the
stomach, the endoscope is removed, the HALO90 electrode is cleaned and then
reintroduced to ablate all treated areas again with 2x15 J/cm2 (40 Watt).
Simple HALO90 ablation regimen:
In the simplified HALO90 ablation regimen the target area is treated with 3x15
J/cm2 (40 Watt) *single-triple* regimen without cleaning in between the
ablations. Using this regimen the HALO90 ablation device mounted on the
endoscope needs to be introduced only once.

Consequences for eligible patients
In this study design individual islands instead of individual patients are
randomized. This ensures that the two ablation regimens are compared under
equal circumstances (it reduces variability due to interpatient differences)
and increases the power of the study since generally multiple islands are
eligible per patient.
This does imply, however, that patients who have more than one eligible island
will undergo both the standard 2x2x15 J/cm regimen (for treatment of islands
randomized to this regimen as well as non-eligible islands and Z-line) AND the
simplified 3x15 J/cm regimen (for treatment to this regimen). This requires an
additional introduction of the endoscope with the HALO90 catheter and a maximum
of 2 ablations. It is estimated that this will prolong the procedure by less
than 5 minutes. The total duration of a HALO90 ablation using the standard
regimen is approximately 35 minutes.

Follow-up
At two months, the first post-HALO90-treatment endoscopy will be performed with
narrow band imaging. Endpoints will be scored at this time. Still images are
made of any residual columnar epithelium (WL + NBI) at the sites of the
previously randomized and treated isles. Percentage of endoscopically visible
surface regression will be scored by two endoscopists, blinded for the
administered treatment regimens, using images taken directly before HALO90
treatment and at the first post-HALO90-treatment endoscopy.
When complete eradication of BE has not been achieved, further treatment will
be according to current RFA guidelines.

Outcome Parameters
Outcome parameters will be assessed after a single HALO90 treatment session:
- Rate of complete removal of BE islets
- Percentage of endoscopically visual surface regression of BE epithelium after
2 months as scored by two endoscopists blinded to the treatment regimen.

Eradication of Barrett's mucosa.

Study patients will have an extra introduction of the endoscope with cap.
During the study treatment, patients will be treated with 2 different HALO90
catheters, one will be used for the current protocol and one will be used for
the single triple regime.
The ablation procedure will be prolongated by less than 5 minutes.