The adenoma detection rate by colonoscopy for asymptomatic individuals aged 50 years and older is estimated to be at least 25%. It is known that during colonoscopy, lesions may be missed. Image enhanced endoscopy techniques have been evaluated for the detection and differentiation of colonic polyps. Narrow band imaging (NBI), is one such technique. The common classification systems used predict histology is the NICE and the Sano and JNET classification. The NICE classification can be used without optical magnification; it evaluates the color of the lesion, regularity of the overlying vessels and regularity of the surface pattern. The Sano and JNET classification requires optical magnification in order to assess the capillary patterns such as whether there is dilation, irregularity or loss of irregular capillaries over the lesion. In the context of adenoma detection, the results are more contentious. A meta-analysis of randomized studies examining the utility of the first generation NBI system when compared to high definition WLI showed no difference in detection rates. A criticism of the NBI system had been the dark endoscopic view; this is a result of the optical filter, and can limit the far view. A second generation NBI system has since been developed. It is characterized by much brighter illumination despite the optical filter, and thus the far view is improved. A recent randomized controlled study compared the second-generation NBI system with high definition WLI. NBI was shown to improve polyp and adenoma detection rates compared to WLI. Blue laser imaging (BLI) is another form of narrow bandwidth imaging developed by Fujifilm Corporation (Tokyo, Japan). Instead of using an optical filter for white light to produce narrow bandwidths, the BLI system has a unique feature of illumination using two lasers and a white light phosphor to accomplish the visual enhancement of surface vessels and structures. This study aims to determine whether BLI can increase the detection rate of colonic polyps and adenomas when compared to white light endoscopy, with the null hypothesis being no difference in detection rates. This study will also examine the use of NICE and Sano/ JNET classification systems to predict histology with the BLI system.

BACKGROUND The adenoma detection rate by colonoscopy for asymptomatic individuals aged 50 years and older is estimated to be at least 25%. It is known that during colonoscopy, lesions may be missed. The miss rates for lesions depend on its size. In a meta-analysis, the overall miss rates for colonic polyps of any size was 22% (95% confidence interval [CI]: 19 - 26%) but based on size of adenomas, the miss rates ranged from 26% (95% CI: 21 - 30%) for lesions 1 - 5mm, to 13% (95% CI: 8 - 20%) for lesions 5 - 9mm and 2 (95% CI: 1 - 8%) for lesions larger than 10mm. Factors implicated in missed lesions include poor bowel preparation, short withdrawal time, lack of meticulous examination, and possibly subtle mucosal changes, especially for small or flat adenomas, that may be easily missed on white light imaging (WLI).

Image enhanced endoscopy techniques have been evaluated for the detection and differentiation of colonic polyps. Narrow band imaging (NBI), developed by Olympus Corporation (Tokyo, Japan) is one such technique and is a function available in all colonoscopies as a press button. When NBI is activated, an optical filter is applied to the white light source, such that only the narrow bandwidths of blue (440 - 460 nm) and green (540 - 560 nm) wavelengths are transmitted. These narrow bandwidths enhance the visualization of blood vessels and mucosal pit patterns. This technique has been shown to be useful in predicting colonic polyp histology and depth of mucosal invasion in the context of intra-mucosal cancer. The two common classification systems used predict histology is the NICE and the Sano classification. The NICE classification can be used without optical magnification; it evaluates the color of the lesion, regularity of the overlying vessels and regularity of the surface pattern. The Sano classification requires optical magnification in order to assess the capillary patterns such as whether there is dilation, irregularity or loss of irregular capillaries over the lesion. In the context of adenoma detection, the results are more contentious. A meta-analysis of randomized studies examining the utility of the first generation NBI system when compared to high definition WLI showed no difference in detection rates; it was only superior when compared to non high definition WLI, which may not be so relevant now since most new systems use high definition WLI. A criticism of the NBI system had been the dark endoscopic view; this is a result of the optical filter, and can limit the far view. A second generation NBI system has since been developed. It is characterized by much brighter illumination despite the optical filter, and thus the far view is improved. A recent randomized controlled study compared the second-generation NBI system with high definition WLI. NBI was shown to improve polyp and adenoma detection rates compared to WLI (adenoma: 48.3% vs. 34.4%, p = 0.01; polyps: 61.1% vs. 48.3%, p = 0.02).

Blue laser imaging (BLI) is another form of narrow bandwidth imaging developed by Fujifilm Corporation (Tokyo, Japan). It is a function integrated into the colonoscopy systems and can be activated and alternated with white light endoscopy by a push button. There is also optical magnification capability, thereby allowing detailed examination of micro-surface and micro-capillary structures. Instead of using an optical filter for white light to produce narrow bandwidths, the BLI system has a unique feature of illumination using two lasers and a white light phosphor to accomplish the visual enhancement of surface vessels and structures. A laser with a wavelength of 450nm stimulates the phosphor to irradiate a white-color illumination. The other laser, with a wavelength of 410nm, is used to enhance the blood vessels at shallow depth in the mucosa. Early data has shown its usefulness in predicting the histology of mucosal lesions. A comparative study showed that BLI had a greater far view compared to NBI due to its much brighter illumination. Thus far there has been no study to determine whether the use of BLI will increase the detection rate of colonic polyps and adenomas when compared to WLE. In addition, there has been no prior study of applying the NICE and Sano classifications developed using NBI to BLI. Recently in Japan a workgroup has proposed the JNET classification system to characterize polyps. This has not been applied using the BLI system either, being developed under NBI. Similar to Sano classification, this uses optical magnification. The main difference between the JNET and San classification systems is that in JNET classification system, Sano 2 is named JNET 2A, and Sano 3A is subsumed under JNET 2B, and Sano 3A becomes JNET 3.

AIM This study aims to determine whether BLI can increase the detection rate of colonic polyps and adenomas when compared to white light endoscopy, with the null hypothesis being no difference in detection rates. This study will also examine the use of NICE and Sano and JNET classification systems to predict histology with the BLI system.

METHODS This is a prospective randomized controlled study. Patients will be randomized in a 1:1 ratio in blocks of 10 to undergo either BLI or WLI colonoscopy. Randomization will be carried out by computer-generated random sequences. Once informed consent is obtained, the research assistant will disclose the assigned imaging technique (BLI or WLI) to the responsible endoscopist before the procedure.

Technique of colonoscopy and imaging All patients will be given dietary instructions before colonoscopy and will be prescribed four litres of polyethylene glycol in a split dose for bowel cleansing 1 day before colonoscopy. The Fujifilm colonoscope with WLI and BLI functions, and optical magnification capability (EC-L590ZW) and Fujifilm LASERO video endoscopy system, which supports WLI and BLI functions of the colonoscope, will be used. Colonoscopy will be performed under conscious sedation with intravenous midazolam and/or fentanyl. In the BLI group, insertion to cecum will be performed under WLI and once the cecum is reached, the BLI mode is switched on during withdrawal of endoscope for complete colonic examination. In the WLI group, WLI will be used during both insertion and withdrawal. Withdrawal time is defined as the time of the initiation of cecal inspection to the time when the colonoscope is completely removed from anus. The time for polypectomy will not be included. A dedicated research assistant measured the withdrawal time by using a stopwatch. The withdrawal time is set to a minimum of 6 min even in patients in whom no polyps are found. Bowel preparation of the whole colon will be graded according the Boston Bowel Preparation Scale which is routinely used in clinical practice and captured in the electronic reporting system.

The sizes and locations of all colonic polyps will be recorded contemporaneously. The locations of colonic lesions will be identified by anatomical landmarks or by transillumination. The size of colonic lesions will be measured against the span of an opened biopsy forceps. Regardless of the assigned group, once a polyp is detected during withdrawal, prior to removal, the surface structure of each polyp detected will be first assessed without optical magnification under BLI using the NICE classification. Thereafter optical magnification will be applied and the polyp will be classified using the Sano and JNET classification. Images will be captured electronically and stored in the electronic medical reporting system and can be retrieved for audit and verification and study of inter-observer agreement. All lesions will be resected or biopsied and labelled for histological examination. If more than 5 polyps are detected, the largest 5 polyps will be used for endoscopic-histological correlation. All procedures will be performed by either experienced endoscopists who have performed at least 1,000 colonoscopies or by endoscopy fellows under direct supervision of the experienced specialists. All endoscopists will have prior experience with image enhanced endoscopy using the NBI system. Prior to the start of the study, the NICE and Sano and JNET classifications will be formally reviewed with all participating endoscopists to ensure familiarity with these classifications for polyp assessment. To familiarize endoscopists with the BLI system, all endoscopists will be asked to carry out at least five examinations with the BLI system before performing study cases.

Definitions Complete colonoscopy is defined as successful cecal intubation. All colonic polyps removed during each examination will be sent for histological examination with clear labelling of location. Histological interpretation of all polyps will follow the World Health Organization system. Advanced adenoma will be defined as adenoma ≥ 10 mm in diameter, with any villous histology, high-grade dysplasia, or invasive carcinoma. Adenoma detection rate and polyp detection rate will be defined as the proportion of patients with at least one adenoma and one polyp respectively.

Outcomes The primary outcomes of this study will be the differences in adenoma (overall and flat) and polyp detection rates of BLI with WLI during diagnostic colonoscopy. Sub-analysis will be performed for the utility of NICE and Sano and JNET classification systems to predict histology with the BLI system.

Statistics We estimate the overall prevalence of colorectal adenoma in the WLI colonoscopy group to be 25%. In order to show a clinically important improvement of adenoma detection by BLI, the new system should increase the adenoma detection rate by 15%. With a statistical power of 80% and a two-sided significance level of 0.05, 152 patients will be needed in each study arm (total 304).

Adenoma detection rate or polyp detection rate will be calculated on the actual number of patients randomized to each group. Adenoma miss rate or polyp miss rate will be based on the percentage of patients who had completed the second colonoscopy. Bowel preparation will be regrouped into satisfactory (excellent to good) and unsatisfactory (fair to poor) for statistical analysis. The differences in detection rates between the BLI and WLI groups will be compared by the student's t-test. Statistical significance is taken as a two sided p value < 0.05. Factors associated with adenoma detection on first colonoscopy will be first identified by univariate analysis. Factors with a P value < 0.1 on univariate analysis will be further entered into forward stepwise logistic regression analysis. The adjusted odds ratio with 95% CI will be used to describe the influence of various factors on adenoma detection rate. All the above statistical analysis will be performed by SPSS statistics software (version 19.0, SPSS, Chicago, IL).

Using histology as gold standard, we will calculate the accuracy, sensitivity, specificity, and negative and positive predictive values for each component of the NICE and Sano and JNET classifications and for the overall prediction by using these classification systems. The differences in performance characteristics between the NICE and Sano and JNET classifications will be compared. To evaluate inter-observer variability in the diagnosis of colonic polyps using the NICE and Sano and JNET classifications a set of endoscopic digital images of histologically confirmed colonic polyps will be reviewed independently by the participating endoscopists. The k statistic, a measure of interobserver agreement over and above chance, will be calculated using the statistical software StatsDirect version 2.6.2 (Stats-Direct, Cheshire, UK). The strength of agreement is defined as follows: very good agreement: k more than 0.8 but less than 1; good agreement: k more than 0.6 but less than 0.8; moderate agreement: k more than 0.4 but less than 0.6; fair agreement: k more than 0.2 but less than 0.4; poor agreement: k less than 0.2.

Inclusion Criteria:

• An individual undergoing diagnostic colonoscopy
• An individual undergoing screening colonoscopy

Exclusion criteria:

• Acute lower gastrointestinal bleeding
• Familial colorectal cancer syndrome
• Inflammatory bowel disease
• Bloody diarrhoea
• Colonic resection
• Extensive abdominal or pelvic surgery where colonoscopy may be considered difficult
• Patients considered to be unsafe for biopsies or polypectomy due to bleeding tendency
• Situations where complete colonoscopy cannot be completed or performed
• Severe comorbid illnesses (ASA 3 and above)