A Comparison of Volumetric Laser Endomicroscopy (VLE) and Endoscopic Mucosal Resection (EMR) in Patients with Barrett*s Dysplasia or Intramucosal Adenocarcinoma

In patients with Barrett oesophagus (BO) malignant degeneration may occur
through a series of phenotypic cellular changes detected and graded on
microscopy; beginning with non-dysplastic intestinal metaplasia (IM), then low-
(LGIN) and high-grade intraepithelial neoplasia (HGIN), and eventually early
cancer (EC) may arise1,2. Endoscopic surveillance of patients with BO is,
therefore, recommended to detect early neoplasia at a curable stage3. When
using standard endoscopy, however, it may be difficult to distinguish areas
with early neoplasia (i.e. HGIN a/o EC) within the normal Barrett mucosa4.
Thus, in the absence of visible abnormalities random four-quadrant biopsies are
obtained every 1-2 cm of the BO, to allow for histological evaluation for the
presence of neoplasia (Seattle protocol)4,5. However, random biopsies are
associated with a high rate of sampling error and may miss malignant lesions in
the BO6. Moreover, the extensive biopsy protocol poses significant burden on
the patient, the endoscopist and the health care system, due to prolonged
endoscopy time and high costs. To increase the detection rate of early
neoplasia during endoscopic surveillance of BO patients, different imaging
techniques have been developed. In this respect, roughly two imaging goals have
to be distinguished: first and foremost, suspicious lesions will have to be
identified in the BO, which requires a *red flag* imaging modality with the
ability to draw attention to a certain area of interest. Second, a
differentiating tool will have to be able to distinguish between truly
suspicious areas (i.e. HGIN/EC) or false positive areas. The N-Vision pVLE
system is a newly developed diagnostic tool that will allow high resolution
imaging of the oesophageal mucosa through Optical Frequency Domain Imaging
(OFDI), a second generation Optical Coherence Tomography (OCT) technology. OFDI
compares backscattered light from tissue to a reference signal, which allows
high resolution depth resolved imaging of the investigated tissue. In essence,
OFDI is a kind of optical ultrasound imaging. The N-Vision probe based
Volumetric Laser Endomicroscopy (pVLE) system incorporates OFDI in a rotating
endoscopic probe, that allows for real-time, 3D high resolution imaging of the
oesophageal mucosa. The N-Vision system can be used as an additional tool
during standard surveillance endoscopy for Barrett's oesophagus or work-up of
early neoplasia. The 3D mucosal map that is projected on the screen of the
n-Vision system may identify suspicious areas that would otherwise have been
overlooked by standard white light endoscopy. VLE is a new technique. Before it
can be applied in the clinic, VLE imaging needs to be validated. Therefore, the
VLE images have to be correlated to the histopathological features of the
imaged tissue. A standardized, ex-vivo set-up will ensure spot-on correlation
between the VLE images and the imaged tissue.

In patients undergoing surveillance endoscopy for Barrett's oesophagus or
work-up and treatment for early neoplasia in Barrett's oesophagus we will
evaluate the N-Vision pVLE system for the following items: 1) Correlating the
VLE images with the corresponding histology of the biopsy specimen of Barrett's
neoplasia in the oesophagus and in the endoscopic biopsy specimen. 2)
Correlating the VLE images with the corresponding histology of the biopsy
specimen of non-dysplastic Barrett tissue in the oesophagus and in the
endoscopic biopsy specimen. 3) To define VLE image characteristics and develope
a VLE classification system for the imaging of Barrett's oesophagus. 4) To
creating a VLE imaging atlas with corresponding histology. 5) To optimize and
validate the VLE classification by independent observers.

Endoscopic procedure: During standard endoscopy for surveillance or work-up,
the oesophagus will first be examined with white light endoscopy, recording all
marks, distances and possible suspicious areas. Subsequently, the N-Vision
probe will be deployed through the working channel of the endoscope, the
balloon inflated and the inner lining of the oesophagus imaged. Areas
suspicious for early neoplasia identified on the N-Vision 3D image will be
recorded. Mapping with the N-Vision is followed by standard biopsies: all
suspicious areas and random four-quadrant biopsies, as required by the official
guidelines. The endoscopist will remove the system after performing this. If
there is an abnormality, this will be marked by electrocoagulation according to
the guidelines, followed by endoscopic resection of the abnormality by the
Cap-technique according to the guidelines. When indicated, biopsies will follow
of other (suspected) abnormalities. All abnormalities will be visualised with
the N-Vision VLE system. After the procedure the biopsy specimen will be placed
in an especially constructed mold and the second pVLE imaging will take place.
Afterwards diagnostic work up will be done on the specimen by the pathology
department. All histological evaluation is done by both a junior and a senior
pathologist. All histology will be reviewed by a GI-expert pathologist. The
histological data will be correlated to the N-Vision data. These images and
histology results will be used to create an image atlas and to develop and
validate a VLE classification system for imaging and reviewing Barrett's mucosa
and neoplasia in the oesophagus. The study will be done in 2 phases: 1) First
phase is a single-centre pilot study, in which set up and logistics of the
system will be optimalised, the VLE image characteristics will be defined and
the VLE classification determined. In this phase,10 patients will be included:
5 patients with Barrett's oesophagus without dysplasia and 5 patients with an
early neoplastic abnormality of Barrett's oesophagus. 2) In the second phase,
up to 30 patients will be included in 1 centre (the AMC); the first 15
inclusions will be patients with an early neoplastic abnormality of Barrett's
oesophagus. After the first 15 an evaluation will be performed to determine the
number of additional patients (maximum of 15), these can be dysplastic and
non-dysplastic. In this phase, the image-atlas will be created and the VLE
classification optimized and validated by indepent observers.

Endoscopic treatment is standard policy in patients with an early neoplastic
lesion in Barrett's oesophagus. An endoscopic resection is prefered because of
the diagnostic and therapeutic value. During this study patients will undergo a
nVision pVLE procedure before and after endoscopic resection. The resection
specimen will be imaged in a standardized ex-vivo set-up with pVLE. The regular
diagnostic and therapeutic proces is not influenced by these extra procedures.
The endoscopy will take 15 minutes longer compared to the standard endoscopy.
The performance of an ER in patients in the dysplastic group does not pose
additional risks, given the fact that these patients undergo an endoscopic
resection as a regular therapy for carcinoma or high-grade dysplasia. The
additional risc is associated with the application of nVision VLE (mucosal
laceration due to the extension of the balloon). In the non-dysplastic group, a
regular follow-up endoscopy will be performed, with standard biopsy sampling.
Additionally, an endoscopic resection will be performed by using the
cap-technique. Minor bleeding may occur in 6% of the cases, usually easily
managed with endoscopic hemostatic techniques. During this study non-dysplastic
patients will undergo a nVision pVLE procedure before and after endoscopic
resection. The resection specimen will be imaged in a standardized ex-vivo
set-up with pVLE. The regular diagnostic and therapeutic proces is not
influenced by these extra procedures. The endoscopy will take 30 minutes longer
compared to the standard endoscopy. The additional risc is associated with the
application of nVision VLE (mucosal laceration due to the extension of the
balloon). Moreover, patients in the non-dysplastic group undergo an extra
intervention (ER), which may mean an extra burden and risk for these patients,
compared to normal surveillance endoscopy.