This will be a feasibility study to evaluate the role of endocytoscopy in classifying
colorectal polyps in vivo. The primary outcomes will be to determine the key endocytoscopy
image features of neoplastic and non-neoplastic colorectal polyps. The target population will
include adult subjects undergoing screening and surveillance colonoscopies.
Colorectal cancer has been recognized as the second most common cause of cancer related death
in the US (1). Most colorectal cancers arise from adenomatous polyps that progress. Colon
polyps are usually classified as neoplastic (adenoma and carcinoma) and non-neoplastic (most
commonly hyperplastic). Standard endoscopic inspection cannot reliably distinguish between
polyp types. Thus, all visualized polyps during standard colonoscopies are typically removed.
Since almost half of all polyps are hyperplastic, a large proportion of unnecessary
polypectomies increase the time, risk, and cost of colonoscopies.
Various studies have showed improved diagnostic accuracy of different types of polyps when
broad field techniques such as chromoendoscopy with the use of topical stains were used to
detect and characterize lesions (2-4). Chromoendoscopy, though an approved method of lesion
characterization during endoscopic evaluation of colorectal lesions, is time consuming and
non-standardized. Recently small field techniques such as confocal microscopy and
endocytoscopy have been introduced enabling visualization of the gastrointestinal tract at a
cellular level, thus allowing diagnosis and classification of colorectal lesions in vivo.
Confocal endomicroscopy used along with chromoendoscopy to detect and characterize lesions
has been studied extensively and reported to have a high accuracy in diagnosing neoplastic
and nonneoplastic lesions of the gastrointestinal tract (5-12). One of the major limitations
of the confocal system is the mandatory use of dyes such as topical Acriflavine or
intravenous Fluorescein. However, there are issues with the application of these dyes; the
risk of DNA damage with Acriflavine reduced its use. In addition, although intravenous
fluorescein is FDA approved for diagnostic fluorescein angiography, its gastrointestinal
application is an off-label indication and considered class IIb by the FDA. The other major
limitation of the current confocal system is that it requires a dedicated confocal endoscope
marketed by a single manufacturer. Thus, the use of confocal imaging requires purchase of
specific confocal endoscopes.
Endocytoscopy is based on the technology of light-contact microscopy. The current
endocytoscopy system (ECS) consists of two prototypes; Prototype one gives a low
magnification (XEC300) with a maximal 450X magnification and field of view of 300X300 um,
Prototype two provide a high magnification (XEC120) with a maximal 1100X magnification and a
field of view of 120X120 um. In contrast to confocal endomicroscopy, the endocytoscopes can
be easily passed through an accessory channel of the conventional therapeutic endoscope. This
method does not require the use of intravenous contrast agents. Instead, it uses topical
staining such as methylene blue or crestyl violet, which is applied routinely to facilitate
visualization and removal of advanced colorectal polyps through endoscopic mucosal resection
The endocytoscopy system has only recently been introduced and hence there are very few
studies reporting its use.
A prospective study from Japan used endocytoscopy on 113 patients to obtain real time
histological images in vivo during colonoscopy (13). With the ECS system, it was possible to
observe lesions at the cellular level, evaluate cellular atypia and distinguish between
neoplastic and nonneoplastic lesions when compared to histology which was used as the gold
standard. The correlation between the endocytoscopic and histological diagnosis was
statistically significant. Eleber et al (14) also reported a sensitivity and specificity of
79% and 90 %, respectively of the ECS system in diagnosing neoplastic lesions in 28 patients
with colonic lesions. Furthermore the study by Cipoletta et al (15) demonstrated that the ECS
system provides real time imaging in vivo with clear visualization of cellular details and
features of dysplasia of aberrant crypt foci, considered to be the earliest precursor of
colorectal cancer. In addition Rotondano et al (15) also confirmed high positive predictive
values for diagnosing hyperplastic polyps as well as dysplastic polyps including low, high
grade dysplasia as well as invasive cancer.
A recent systematic review of all published studies also confirmed that endocytoscopy is a
safe and effective new endoscopic imaging technique to obtain in vivo histology and guided
biopsies with high diagnostic accuracy (17).
No associated risks related to the endocystoscopy procedures have been reported in all
published studies (12-17).
This specific diagnostic tool of in vivo histology with the use of high resolution
endocytoscopy system would allow endoscopists to perform targeted biopsies and in some cases
(if warranted based on the in vivo images) to proceed directly to endoscopic resection of
lesions. It may also guide assessment of the completeness of the endoscopic intervention with
detection of any residual neoplastic tissue at the index exam as well as on follow up
colonoscopy exams. Other potential benefits include elimination of random biopsies for
surveillance of mucosal disease, elimination of sampling error, limitation of unnecessary
polypectomies, hence ultimately resulting in cost effectiveness and improved patient
In summary, endocytoscopy has the potential to fundamentally change the way endoscopy and
pathology interact by allowing near histological quality imaging in vivo, without the need,
risk, and cost of tissue removal.
• Age above 18, any patient undergoing screening or surveillance colonoscopy.
- Unwillingness to consent
- Lack of any pathological state that would require biopsy at the time of endoscopy
(will be considered "screen failure" since this will not be known until after consent
is obtained and sedated endoscopy performed).