The primary endpoint of the study will be to compare the accuracy of two procedures (FICE with target biopsies only, versus conventional white light colonoscopy with recommended targeted and random biopsies) in the endoscopic surveillance of patient with long-standing UC. Accuracy will be measured based on the number of patients with confirmed neoplasia using each technique. The combined histological outcome following the two procedures will represent the gold-standard diagnosis for each patient. Secondary outcomes will be the number of patients with false-positive findings, the number of neoplastic lesions detected, the number of false-positive lesions per patient for each technique and the total time required for each procedure.

Patients will undergo two colonoscopies each, with an interval of three months between procedures. This minimum time interval is chosen in order to allow for the healing of the mucosa on sampled areas and thus prevent recognition of biopsy sites. The first procedure will be randomly allocated to be either conventional white light endoscopy coupled with targeted and random biopsies or high-resolution endoscopy with FICE system and magnification. Randomisation will be achieved prior to the first endoscopy by means of sealed envelopes. In each recruitment centre, one of the two endoscopists with experience in endoscopic surveillance and treatment of ulcerative colitis will be assigned to carry out the first procedure. The second procedure will automatically be scheduled with the second endoscopist, who will be blinded as to the clinical and histological findings of the first investigation.

The two participating centres are already endowed with identical endoscopic equipment. All examinations will be performed using the same high resolution endoscope (EC-590 ZW, Fujinon Inc., Daitama, Japan). The zoom function on the device will only be used during the FICE procedures. The system is equipped with the EPX 4400 processor (Fujinon Inc., Japan) that enables the CVC technology. This digital processing system can switch between conventional imaging and CVC imaging at any time during the procedure by means of a simple pushbutton on the endoscope. The system has up to ten (# 10) settings designed to select the most suitable wavelengths. In this study the CVC procedure will be performed using setting number three (# 3).

The colonoscopy protocol will be the same in both participating centres. All patients will undergo a bowel preparation consisting in the intake of four litres of hypertonic polyethylene glycol solution. The procedures will be performed under conscious sedation using propofol. The caecum will be reached in white light endoscopy in all cases. Cecal intubation will be confirmed by identification of the ileocecal valve and appendiceal orifice. Upon extubation, 20 mg of butyl scopolamine will be given intravenously, barring any contraindication, to reduce colonic motility and facilitate the examination of the colon. When performing the FICE procedure, the imaging mode will be switched to CVC at the caecum and will then be used throughout withdrawal. The endoscopist will classify the degree of inflammation in each segment of the colon on a scale and give the Mayo Clinic score (proctosigmoiditis - left-sided colitis - Pan Colitis). The quality of the bowel preparation will be noted. During the extubation phase, washing of the colon and aspiration of waste will be accomplished in an optimal way to maximise the detection capabilities of each procedure.

The biopsy protocol is meant to reflect observed mucosal abnormalities and, in the case of conventional colonoscopy, it will be supplemented by random samples taken every 10 cm of the colon. A standard biopsy forceps will be used (Radial Jaw 4, Boston Scientific Inc., USA). To reduce the risk of sampling error, a minimum of two biopsies for each suspicious lesion will be performed. The number of lesions suspect of neoplasia will be noted and targeted by each procedure. In the case of high-resolution FICE colonoscopy, an analysis of the surface pattern will be performed for each targeted lesion according to the pit pattern classification. Suspicious FICE lesions will be defined as having a polyploidy, flat or irregular mucosal structure with Kudo pit pattern III - V, unusual ulcers, strictures or areas with increased and disrupted vascular intensity revealed by dark coloration/discoloration and confirmed with magnification (annexe 9). In conventional endoscopy without FICE, suspicious lesions will be defined as polypoid or irregular mucosa, and unusual ulcers or strictures. During the conventional white light endoscopy (but not during FICE) additional four-quadrant random biopsies will be taken every 10 cm of colon and placed in a specimen container of formalin. Targeted biopsy samples will be sent separately for analysis.

The histopathological evaluation will be performed twice, by two different pathologists, at each participating centre. The pathologists were recruited according to their expertise in digestive histology. For the purposes of this study, they will be blinded to the assessment of the endoscopist when analysing biopsy samples. The inflammation activity level of each specimen will be ranked into the following categories: no inflammation, mild to moderate inflammation or severe inflammation. Dysplasia will be classified according to the new Vienna classification 21. Lesions classified as "indefinite for neoplasia" with no differentiation between adenoma and colitis-associated dysplasia in biopsy material will be not considered as neoplastic. The final histopathology findings will then be compared with the endoscopic assessment with regards to the presence of intraepithelial neoplasia and colorectal cancer.

Inclusion Criteria:

• Clinically and histologically verified UC
• Disease duration ≥ 8 years
• A Mayo score ≤ 8 with an endoscopic sub score ≤ 2
• CPAM affiliation
• Able to give written informed consent to participate in the study

Exclusion Criteria:

• Known intraepithelial neoplasia or colorectal cancer or any other active malignancy
• Previous colo-rectal surgery
• Non-treatable coagulopathy or hemostatic dysfunction (prothrombin index < 50% of control or/and partial thromboplastin time > 50 seconds and/or thrombopenia < 60000 / mm3)
• Pregnancy
• Inability to give informed consent