AG

AA

bst

ract

sfar after an average follow up period of 3 years. 12 of those cases were high grade dysplasiaor intramucosal carcinoma and 6 were invasive cancers in stage 1A or 1B. The intervalbetween the most recent endoscopy with no abnormal findings and the endoscopy wherecancer was diagnosed is 4-25 months. Conclusions: Endoscopic surveillance is feasible, andhas already detected high grade dysplasia or early gastric cancer in a high risk SingaporeanChinese population.

532

Serrated Polyposis Syndrome: Monitored Colonoscopic Surveillance in aTertiary Setting Is SafeSusan Parry, Sonja Woodall, Julie Arnold, George Willdridge, Michael D. Walsh, DanielD. Buchanan, Christophe Rosty, Joanne P. Young

Aim To assess colorectal cancer (CRC) risk for a cohort of patients with the Serrated PolyposisSyndrome(SPS) during prospectively monitored colonoscopy surveillance in a tertiary setting.Method Colonoscopy and histology records from January 2000 to time of interview for 101New Zealand patients, meeting WHO criteria for a diagnosis of SPS and enrolled in theGenetics of Serrated Neoplasia study, were reviewed. Polyp demographics and family historyof cancer were recorded. Results Of 101 SPS patients enrolled in the study, 27 had CRCat initial presentation (median age 58 yrs, 18 female, range 22 - 80 yrs), 14 located in theproximal colon and 13 in the distal colon. Only two of the 27 reported a first degree relative(FDR) with CRC. Over a median follow up of nine years (range 1-11 yrs) from 2000with an average interval of 15 months between colonoscopies, two patients developedmetachronous CRC: one identified at prophylactic completion colectomy and one in associa-tion with first diagnosis of SPS, 19 years following initial CRC and 30 months after a previouscolonoscopy. A further patient proceeded to completion colectomy and ileostomy due topolyp burden, 4 years following previous surgery for CRC. Of 69 patients with polyps alone(median age 52 yrs, 41 female, range 19 - 82 yrs) followed for a median of 6 years (range1-12 yrs) with an average interval of 17 months between colonoscopies, none developedCRC despite 63 patients having multiple pan colonic polyps. Of these 63, 36 had an adenomaand 10 had a sessile serrated polyp. In addition, 37 of the 69 patients had serrated polypsgreater than 10mm in size dispersed throughout the proximal and distal colon. A FDR withCRC was reported in 17 of 69 patients. Five patients with multiple pan colonic polyps andno FDR with CRC underwent prophylactic subtotal colectomy (median age 58 yrs, 4 females,range 28 - 63 yrs) after being followed for a median of 5 years with an average intervalbetween colonoscopies of 8 months. All had an associated adenoma and two a sessile serratedpolyp. Conclusion If early endoscopic control is feasible SPS patients with pancolonicpolyposis, and those with associated conventional adenomas and large serrated polyps,can be judiciously managed by one to two yearly surveillance colonoscopy in a tertiaryclinical setting.

533

Attendance and Diagnostic Yield of Repeated Two-Sample FecalImmunochemical Test (FIT) Screening in a Randomized Population-BasedColorectal Cancer TrialAtija Kapidzic, Aafke H. van Roon, Miriam P. van der Meulen, Anneke van Vuuren,Marjolein van Ballegooijen, Wolfert Spijker, Monique E. van Leerdam, Ernst J. Kuipers

Introduction: Colorectal cancer (CRC) screening by means of fecal immunochemical tests(FITs) requires successive rounds for an optimal preventive effect. The diagnostic yield ofadvanced neoplasias may increase with use of 2 FITs per round. We therefore conducteda population-based CRC screening trial focusing on attendance and diagnostic yield ofrepeated 2-FIT screening. Methods: A representative sample of the Dutch population (n =3197) aged 50-75 years was randomly selected and invited by mail for two rounds (R1resp. R2) of FIT screening with a 2-year interval. Per round, invitees received 2 FITs (OC-Sensor Micro, Eiken Chemical, Japan) to sample from two consecutive bowel movements.FIT was per round considered positive if at least one of both tests was positive at a cut-offof 50 ng Hb / mL. Results: After excluding individuals who at either round met exclusioncriteria, had died, had moved away, or were positive at R1, attendance was 61.3% at R1(1875/3061; CI 59.6-63.0%, and 61.2% at R2 (1644/2688; CI: 59.3-63.0%). At R1, thepositivity rate was 12.8% (5.0% 2 FITs pos, 9.0% 1 FIT pos (i.e. 1-sample FIT)). In R2,the positivity rate was 8.6% (2.9% 2FITs, 5.8% 1 FIT pos (i.e. 1-sample FIT)) (p ,0.001for R2 vs. R1). At R2, R1 participants had lower positivity rates than R1 non-participants(7.6 vs. 14.5%; p,0.001). The detection rate (DR) of advanced neoplasia was 4.1% (CI3.3-5.1) at R1 and 1.5% (CI 1.2-2.5) at R2 (p ,0.001), with a positive predictive value ofat least one positive FIT being 38% (CI 31-45) at R1, vs. 18% (CI 12-25) at R2 (p ,0.001).Compared to a historical control series of 2-round 1-FIT screening (van Roon AH, Gut2012), repeated 2-FIT screening led to persistently higher positivity rates (R1 and R2 with1-FIT 8.4+6.0% vs. 2-FIT 12.8+8.6%; p,0.001) with a detection rate for advanced neoplasiaof 5.1% (R1 and R2;3.3+1.8%) for 1-FIT screening and a detection rate of 5.6% (R1 andR2;4.1+1.5%) for 2-FIT screening. Conclusion: 2-sample FIT screening is associated withstable, high attendance similar to screening with 1-sample FIT, but with higher positivityrates. Over 2 rounds, the detection rate of advanced neoplasia drops with 2-FIT screening,but still remains higher compared to 2-round 1-FIT screening. This implies that 2-FITscreening has a benefit in both first and second round screening to detect a maximumnumber of individuals with advanced neoplasia.Overview of participation and FIT performance characteristics for 2-sample FIT screeningin the first and second screening round

S-96AGA Abstracts

FIT: fecal immunochemical test; advanced advanced neoplasia: advanced adenoma or colo-rectal cancer; advanced adenoma: ≥10mm, villous component (≥ 25% villous) or high-grade dysplasia

534

Systematic Colonoscopic Screening of Adults With Cystic FibrosisDemonstrates High Incidence of Advanced Adenomatous PolypsAleh Bobr, Joanne Billings, Jordan Dunitz, Sandy McAllister, Tyler Herzog, AlexanderKhoruts

Cystic fibrosis (CF) is associated with increased incidence of gastrointestinal malignancies,including colon cancer. Survival of CF patients has increased dramatically over the past twodecades due to advances in medical care. Therefore, colon cancer can be anticipated tobecome significant emerging problem in this patient population. However, no specificguidelines exist for their colon screening. Starting January 2010 all adult patients with CFreceiving care at the Minnesota Cystic Fibrosis Center, aged 40 or above and having FEV1≥ 50% predicted, were recommended to have screening colonoscopies. 21 patients had theirexaminations done following a rigorous cleansing protocol introduced in our institutionspecifically for CF patients because regular colonoscopy preps are typically suboptimal inthis population due of viscous intestinal secretions. In addition, results of 17 examinationsdone with standard preparations in patients with the same inclusion criteria were availablefor analysis. The mean age of the studied population was 47 years old. Significantly moremales than females were found to have adenomatous polyps (60% versus 11%, p = 0.0039).However, one female patient was found to have rectal adenocarcinoma, and ultimately diedfrom this disease. The majority of patients with polyps had multiple polyps (mean 4.60 ±0.98 in men, 1.5 ± 0.50 in women, p = 0.2044). Furthermore, the majority of positiveexaminations found advanced lesions (size ≥ 1 cm, villous features, high grade dysplasia).The rigorous prep described here performed significantly better than the routine split preps(median 4 versus 2 on the Aronchik 0-4 Scale, p = 0.0002). Finally, in all patients wherefollow-up surveillance examinations were available, multiple additional adenomatous polypswere found within 1-3 years. Conclusions. CF is associated with early development ofadenomatous colon polyps. This finding fits with the known association between CF andgastrointestinal malignancies, including colon cancer. CF should be recognized as a geneticcolon cancer syndrome. Adult patients with CF should begin screening colonoscopic exami-nations before age 40 and subsequent surveillance interval should be , 5 years. Finally,standard colonoscopy preparations are rarely adequate in CF patients, and they requiremore rigorous purgative protocols.

535

Neoplasia Yield of Repeated Immunochemical FOBT Following a NegativeScreening SigmoidoscopyCarlo Senore, Paola Armaroli, Arrigo Arrigoni, Paola Cassoni, Franco Ferrero, AdrianoGiacomin, Mauro Giovanardi, Orietta Giliani, Nereo Segnan

BACKGROUND: The combination of fecal immunochemical test (FIT) and sigmoidoscopy(FS) has been proposed as a potentially effective screening strategy for colorectal cancer(CRC). AIMS. To assess the detection rate (DR) of advanced adenomas (AAd) and CRCand the stage distribution of screen-detected (SD) CRCs among people examined withsigmoidoscopy (FS) who were offered biennial FIT starting 2 years after the FS. METHODS: