Polyp Surface Pattern Recognition in the Identification of Neoplastic Polyps
Polyps are a common finding during colonoscopy. It is current practice to remove these
lesions, as some have the potential to turn into cancer. However, not all polyps are the
same. Polyps can be of three different types:
1. Hyperplastic, which have negligible potential to turn into cancer and if left would
cause no harm. These account for around one third of all small polyps encountered
2. Adenomas, which can turn into cancer and should be removed
3. Polyp cancers, which should be either biopsied or removed completely
The management pathway for polyps >10mm is very simple as they are either likely to be
adenomas which need removal or cancers which need a biopsy and tattoo. There is a rare
possibility of these big polyps being hyperplastic / serrated adenomas and given the risk of
malignant transformation, they need removal as well.
It has been traditionally felt that hyperplastic polyps cannot be separated clinically from
adenomas or polyp cancers by the endoscopist. It is for this reason that all polyps are
removed. However, polypectomy increases the risk of perforation and results in a significant
cost in processing the samples. Recently it has been suggested that it is possible to
differentiate neoplastic from non neoplastic lesions in the colon using skills in polyp
surface pattern recognition.
Kudo's pit pattern is an effective way of in-vivo prediction of histology and
differentiating neoplastic from non neoplasic polyps. However, it was originally described
using vial staining which is cumbersome, time consuming and is not possible to perform
outside Japan due to lack of availability and fears of toxicity related to gentian violet.
It does however produce excellent results when used for pit pattern recognition. (Hurlstone
DP C. S., 2004) It works by irreversibly binding microbial DNA and directly inhibiting cell
replication. (Wakelin LPG, 1981).
Gentian violet is not the only way to examine the surface patterns of polyps. Many of the
features can be seen with white light endoscopy. (Fu KI, 2004). Furthermore, there is
extensive experience with other dies which are commonly used in Western Europe. Indigo
carmine dye spray has been performed in studies in Japan, and has proven to be very
effective in the assesment of polyp characteristics. It has been used for over 14 years in
colonoscopy (Axelrad, 1996). Overall diagnostic accuracy by conventional view,
chromoendoscopy and chromoendoscopy with magnification ranged from 68% to 83%, 82% to 92%,
and 80% to 96%, respectively. (Axelrad AM, 1996) (Tung SY, 2001) (Eisen GM, 2002) (Su MY,
2004) (Apel D, 2006). It has the advantage over gentian violet that it does not bind to
tissues and is therefore very safe. It has been questioned whether widespread applications
of the techniques could influence the indications for biopsy sampling during colonoscopy
and the indication for mucosectomy (Yasushi Sano, 2005).
When compared to standard colonoscopy, indigo carmine chromoendoscopy, with magnification,
has been shown to increase the accuracy for polyp differentiation from 84% to 96% (Fu KI,
2004). High-resolution indigo carmine chromoendoscopy demonstrated a small increase in
accuracy over chromoendoscopy without magnification from 81% to 83% . (Hurlstone DP K. M.,
Because of its numerous benefits in lesion detection and assessment of polyps indigo carmine
has become a standard part of colonoscopy and is recognised by the British Society of
Gastroenterology as an important skill for the practicing endoscopist. It has become
standard practice within Portsmouth Hospitals NHS trust when performing colonoscopy. This is
a position which is reflected nationally in most large units which participate in the bowel
cancer screening programme.
White light endoscopy and chromoendoscopy are not the only methods of examining colonic
polyps. Computed virtual colonoscopy, where a video computer processor within the endoscopy
equipment enhances the endoscopic picture have become commonly available. There are systems
available from all of the main endoscope manufacturers, including narrow band imaging from
Olympus and i-scan from Pentax. Fujinon has developed a very effective post processor
technology called FICE which helps define the surface pattern of polyps in great detail.
This is now a standard feature on all of the endoscopes produced by the company. It has been
shown in a limited number of studies to offer a potentially useful alternative to dye spray
in predicting in-vivo histology. (J. Pohl, 2008). A prospective randomised multicentre study
looked at 764 patients with FICE compared to white light with targeted indigocarmine spray.
It found that FICE and indigocarmine were both able to differentiate adenomas from
neoplasia. There was a sensitivity in differentiating adenomas from non-neoplastic polyps of
92.7%,with FICE, comparable but not superior to that of indigocarmine (90.4%), with no
statistically significant difference between the two techniques observed. (p=0.44.) FICE did
not shorten the procedure time. A prospective series looking at FICE in the evaluation of
colononic polyps up to 2cm in size by the same team suggested a sensitivity of 89% could be
achieved. Unfortunately these studies looked at larger lesions and therefore it is difficult
to say whether this could be achieved with polyps <1cm, where accurate assessment is clearly
more difficult. However, it is in the smaller lesions where the greatest gain in terms of a
change in management exists. (Pohl J N.-T. M., 2008).
It is generally accepted that further research is needed in this field. If it is possible
for endoscopists to differentiate neoplastic from non neoplastic polyps the potential
benefits both in terms of safety and histopathology costs would be significant. The current
policy of polypectomy or biopsy of all polyps detected during colonoscopy is a very
expensive option (£ 80-120) and carries a significant risk of perforation.
Within Portsmouth Hospitals NHS trust it has become standard practice to assess all polyps
with FICE and dye spray with indigo carmine prior to removal. This is consistent with
current guidelines from the British society of Gastroenterologists and is important both in
the detection and assessment of polyps. These techniques are established and their
importance in colonoscopy is not under investigation. What is unclear is whether
endoscopists are able to differentiate neoplastic from non-neoplastic lesions on a basis of
surface pattern recognition accurately enough to be able to change clinical practice.
We aim to assess the impact of surface pattern assessment on the potential management of
patients, and whether it has the potential to safely replace the policy of biopsy /
polypectomy for pateints with polyps <10mm in size.
Observational Model: Cohort, Time Perspective: Prospective
The accuracy of predicted in-vivo polyp histology
Pradeep Bhandari, MBBS,MD,MRCP
United Kingdom: Medicines and Healthcare Products Regulatory Agency