10.1055/a-0982-2904Since the introduction of narrow-band imaging (NBI) in 2005, based on optic filters
that select wavelengths of light that correspond to the peak light absorption of hemoglobin,
and the further systems that digitally process endoscopic images to provide a series
of wavelengths of light, such as Fuji Intelligent Chromo Endoscopy (FICE) and i-Scan,
there has been a great improvement in colorectal lesions characterization, delineation
and classification, specially when combined to high-optical magnification. However,
when it comes to colorectal polyps or adenoma detection, virtual chromoendoscopy techniques
have not demonstrated significant improvement over white-light endoscopy (WLE) in
colonoscopy screening targeting an average-risk population. A meta-analysis of 42
studies concluded that only conventional dye-based chromoendoscopy was associated
with an increase in adenoma detection rate (ADR) in comparison to NBI, FICE, i-Scan,
autofluorescence imaging and cap-assisted colonoscopy [1]. Another single-center randomized controlled trial (RCT) enrolling 1650 patients
also showed no difference in ADR and adenoma miss rate between virtual chromoendoscopy
with NBI or FICE and WLE [2].
More recently, a second generation of equipment-based image-enhanced endoscopy (IEE)
technologies has been launched, which includes NBI in the latest Olympus scopes (2G-NBI);
i-Scan Optical Enhancement from Pentax; Blue Laser Imaging (BLI) and Linked Color
Imaging (LCI) from Fujifilm [3]. The new generation of IEE technologies has improved the brightness of the light
source, offering a better quality of image and overcoming the darkness associated
with far-field examination in earlier versions [3]. Conceivably, the new generation of IEE could perform much better in detection of
superficial neoplastic lesions and could be a powerful tool for colorectal cancer
screening. LCI particularly has been accepted with enthusiasm due to the lighter colors
and brighter image that resembles WLE, and the more reddish appearance of mucosal
abnormalities in comparison with NBI and BLI. A multicenter prospective RCT in China
demonstrated that LCI increased ADR and sessile serrated polyp detection compared
to WLE [4]. Another RCT, designed in Japan, reported an improvement in polyp detection rate
in the right colon by six non-expert endoscopists with use of LCI (92.3 ± 2.3 % with
LCI vs. 72.7 ± 11.5 % with WLE; P < 0.01) [5]. Moreover, a recently published European RCT revealed a reduction in the adenoma
miss rate with use of LCI in comparison to WLE (11.8 % vs 30.6 %; P < 0.001) [6].
In this issue of Endoscopy International Open, Sakamoto et al. [7] from National Cancer Center Hospital (NCCH) in Japan report an interesting study
aiming to evaluate the additional benefit of LCI in detection of colorectal polyps
compared to WLE in the insertion phase of colonoscopy. The authors enrolled 138 patients
equally distributed into two groups: colonoscope insertion with WLE or with LCI. All
patients underwent colonoscope withdrawal from cecum to rectum under LCI imaging.
A single, experienced endoscopist carried out all procedures. Interestingly no difference
was noted in detection of right colon polyps, because the majority of lesions were
detected during scope withdrawal. In addition, the efficacy of LCI for detection of
advanced adenomas (> 10 mm, villous component or presence of high-grade dysplasia)
was not superior to WLE. However, in the left colon, particularly in the sigmoid,
more polyps were detected during insertion with LCI imaging than with WLE, 16 % out
of 48 lesions for LCI against 0 % in the WLE group (P = 0.045). Of note, most of the lesions detected were small tubular adenomas smaller
than 5 mm. The authors concluded that colonoscopy with LCI improves colorectal polyp
detection during the insertion phase of the examination.
Many investigators have studied whether use of virtual chromoendoscopy is superior
to WLE in detection of colorectal polyps and flat lesions. The majority of studies
published to date have attempted to answer this question during the withdrawal phase
of colonoscopy, usually comparing two groups, the first under WLE (control group)
and the second with different types of virtual chromoendoscopy technologies (intervention
group), either in more sophisticated methodology such as in a back-to-back or tandem
fashion, with patients undergoing two consecutive examinations, whereas in other studies,
each individual undergoes a single exam (control or intervention), the design adopted
by Sakamoto et al [7]. What is unique about this study is that LCI impact was compared to WLE during the
insertion phase of colonoscopy. In this stage of the examination, the majority of
colonoscopists usually attempt to reach the cecum as fast as possible and do not particularly
carefully inspect the colonic mucosa. Occasionally when small polyps are identified
during insertion, immediate resection may be carried out, avoiding the risk of being
unable to find the lesion again during colonoscopy return. The group from NCCH proposes
an innovative approach, carrying out a meticulous examination during scope insertion
using LCI, to avoid missing polyps in the left colon. In their experience, approximately
10 % of polyps could only be detected during the insertion phase. It is an interesting
proposal, although their study has methodological limitations, as acknowledged by
the authors, particularly lacking a formal randomization design and enrolling a small
sample size, which could provide more robust evidence in favor of such an approach.
There is still a lot of space for improvement in colorectal cancer screening, and
virtual chromoendoscopy techniques may be the key to better performance in several
quality standards, such as ADR, adenoma miss rate and right colon examination. Although
IEE plays a definite role in lesion characterization and classification, its value
in routine inspection of the colon in screening programs is still being debated, especially
when taking into account studies that enrolled the first generation of virtual chromoendoscopy.
There is a growing body of evidence showing that newer techniques such as LCI improve
such parameters in small, prospective controlled trials; however, further large-scale
studies are warranted to add up to these findings.
In addition, the quality of colonoscopy screening involves several factors, ranging
from patient selection and pre-procedure counseling, to bowel preparation, sufficient
time dedicated to complete inspection, technology of the endoscopy system, and more
importantly, endoscopist expertise and training. Although the largest trials addressing
virtual chromoendoscopy’s impact on colorectal screening failed to demonstrate an
improvement in pathology detection, the first studies about LCI have presented promising
results. However, one must always take into account the importance of the knowledge
and experience of the endoscopist when analyzing either a scientific study or when
considering the addition of new technology to daily practice.
In our personal view, it will be difficult to prove that any type of virtual chromoendoscopy
technology per se will be able to overcome a highly trained endoscopist with a WLE
high-resolution scope in hands, performing a dedicated colonoscopy aimed at finding
and properly removing any colorectal polyp or flat lesion in his/her way. Therefore,
we endorse the recent statement of Matthew Rutter concerning this same issue: “pick
your endoscopist before you pick your technology.” [8] Perhaps, in the near future, the frontiers recently opened with artificial intelligence
will revolutionize gastrointestinal endoscopy practice and demonstrate that both of
us were wrong.
In conclusion, we would like to congratulate the authors from NCCH on developing a
study that prompts some new questions based on its conclusions: Should we change our
routine and proceed careful examination during scope insertion? What is the impact
of higher detection of diminutive left-sided polyps? Is there a definitive role for
new-generation virtual chromoendoscopy in colorectal cancer screening? And is LCI
the best technique for virtual chromoendoscopy?
