AGA Abstracts
dates and characteristics of all antecedent colonoscopies. For the purposes of this analysis, we defined a PCCRC as a CRC with a prior colonoscopy >6 months but <60 months prior to date of diagnosis. For PCCRC cases, clinical characteristics were extracted from available medical records by a health record analyst. An abstract of each case was reviewed by the study authors according to a structured algorithm (see Figure 1). Cases were excluded if they were from a high-risk group (IBD, genetic syndromes), if there was a database error (e.g. flexible sigmoidoscopy included), or if the medical records were insufficient to make a determination. Path to care issues (e.g. severe intercurrent illness, patient refusal, loss to follow-up, long surgical wait) were classified separately. RESULTS: 1278 patients >39 years of age were diagnosed with CRC in 2013 and had a colonoscopy identified by the database linkage. 119 patients were identified with a total of 154 colonoscopies performed >6 months but < 5 years prior to CRC diagnosis. 77 colonoscopies were excluded for reasons other then colonoscopy quality (See Table 1). 15 of these were associated with an incorrect followup recommendation based on colonoscopy findings and polyp pathology. Of the remaining 77 colonoscopies, 48 were done 6-36 months prior to diagnosis. 29 were done 36-60 months prior to diagnosis. CONCLUSIONS: Analysis of PCCRC outside screening programs requires an algorithmic approach to consistently identify clinical associations and root causes. Reduction in PCCRC incidence will occur only through improvements in system-wide clinical pathways in addition to colonoscopy quality. Cases excluded for reasons other then colonoscopy quality
Male gender and FIT Hb concentration could be predictors of risk of colorectal cancer and thus used to prioritize colonoscopy in patients with suspected advanced neoplasia, both in screening and in symptomatic patients.
Su1756 EFFICACY AND TOLERABILITY OF VARIOUS BOWEL PREPARATIONS IN PATIENTS WITH DIABETES: A RANDOMIZED CONTROLLED TRIAL Mohammad F. Madhoun, Khadija Chaudrey, Sian S. Chisholm, Aftab Ahmed, Belinda Frost, William M. Tierney BACKGROUND Diabetes mellitus (DM) has been shown to be an independent risk factor of having poor bowel preparation. Bisacodyl is a stimulant laxative that might help with the underlying problem of colonic dysmotility associated with diabetes. AIM We hypothesized that the addition of bisacodyl to split dose bowel preparation (SDBP) would improve the quality of bowel preparation among diabetics. METHODS: Adult outpatients aged 18 to 80 years undergoing colonoscopy were recruited to the study. One hundred and eighty six diabetic patients were randomly assigned to 1 of 3 treatment arms; 1) conventional 4 liters of polyethylene glycol electrolyte lavage solution (PEG-ELS) (conventional), 2) splitdose of 4 liters PEG-ELS (SDBP), and 3) split-dose of 4 liters PEG-ELS preceded by bisacodyl 10 mg (SDBP-B). The primary outcome measure was the bowel cleansing quality using Boston bowel preparation scale (BBPS). Endoscopists were blinded to the preparation used. The secondary outcome measures were safety and patient tolerability. RESULTS: There were no differences among the groups with regards to age, indication, duration of DM, using of insulin, the presence of diabetic neuropathy, nephropathy or retinopathy or the use of narcotics. There was a trend of better bowel preparation quality among the SDBP and SDBP-B compared to the conventional group but failed to reach statistical significance (≥7 BBPS; 55% vs. 63% vs. 63%, p=0.6), (≥6 BBPS; 79% vs. 90% vs. 82%, p=0.3). In terms of safety and tolerability, there were no differences among the three groups. (Table) CONCLUSION: Addition of bisacodyl to SDBP does not improve the quality of bowel preparation in diabetic patients. Further efforts are needed to optimize bowel preparation in this population.
Su1755 FECAL HEMOGLOBIN CONCENTRATION, A GOOD PREDICTOR OF RISK OF ADVANCED COLORECTAL NEOPLASIA IN SYMPTOMATIC AND ASYMTOMATIC PATIENTS Mercedes Navarro, Ignacio Omella, Patricia Carrera, Gonzalo Hijos, Pilar Roncales, Federico Sopena, Juan Jose' Puente, Angel Ferrandez, Angel Lanas Su1757 Background. Periodical fecal immunochemical testing (FIT) is a cost-effective strategy in colon cancer screening programs. FIT is also used as a diagnostic test in symptomatic patients. Some studies suggest the risk of advance neoplasia (AN) defined as advanced adenoma and cancer, increases when the concentration of FIT Hemoglobin (Hb) is higher, but data are scarce. Aim. To determine the association between FIT Hb concentracion and advanced neoplasia detected in colonoscopy in two different populations. Methods. The outcomes of colonoscopies performed after a positive FIT (> 117 ng/ml) (Sentinel Gold test) result were analyzed in patients included within a population-based CRC screening program (screening group) and, as diagnostic evaluation in symptomatic patients (symptomatic group). The study was performed between January 1st 2014 and October 31th 2016 in a tertiary hospital. Continuous variables were reported as medians with interquartile ranges whereas qualitative variables were expressed as frequencies and percentages. The positive predictive value (PPV) at arbitrary fecal hemoglobin concentrations was calculated for AN. A logistic regression analysis was performed to determine the independent association of sex, age and FIT quartiles with the detection of AN. Results. 2742 colonoscopies (57.8% men) were performed; 1515 (53.5%) in the CRC screening program. Patients in the screening group were younger (65.0±3.3 vs 66.2±13.4 years, p<0.001) and more frequently male (61.5% vs 53.3%, p<0.001) compared to the symptomatic group. Colonoscopy found more frequently neoplastic lesions in the screening vs the symptomatic group (61.9% vs 44.8% p<0.001). Median Hb concentration was significantly higher in patients with AN compared with patients without AN: 2469 ng/mL (Interquartile range (IR) 530-6624 ng/mL) vs 303ng/mL, (IR 181744 ng/mL), p<0.001 in the screening group, and 4103 ng/mL, (IR 581-12282 ng/mL) vs 421 ng/mL (IR 205.8-1739.0 ng/mL) (p<0.001). The risk of AN increased significantly in both groups according to FIT Hb concentration in the Quartil 3 (range 431-1956 ng/ml) OR (95% IC): 3.43 (2.36 - 4.99) and Quartil 4 (> 1957 ng/mL): 9.77 (6.85-13.94). Males and patients included the screening group showed higher Odds of presenting AN (Table) and higher positive predictive values for AN in positive FIT tests (range 17% - 40.2%; 117 to > 1000 ng/mL) compared to those in the symptomatic group (18% - 32.7%). Conclusions.
AGA Abstracts
COLORECTAL CANCER SCREENING DETECTS TUMOURS AT EARLIER STAGES IMPROVING CLINICAL OUTCOMES Jan Kral, Mirko Jakovlevicˇ, Jan Knot, Jan Bauman, Monika Vladarova, Zdenka Kralova, Radka Buresova, Vladimir Kojecky, Tomas Grega, Michal Stepan, Jan Kotyza, Radka Stepanova, Julius Spicak BACKGROUND: Colorectal cancer (CRC) remains a major health burden. Screening is recommended and considered to be beneficial, nevertheless the evidence is rather scanty. OBJECTIVE: The goal of our national multicentre prospective observational study was to compare the characteristics of CRC disease between those detected by screening (screening group) and CRC diagnosed by other ways (non-screening group). According to National population-based program screening was defined as a primary screening colonoscopy or colonoscopy after a positive FOBT in an average risk population. METHODS: Between March 2013 and September 2015 we enrolled 265 patients (73 in the screening group, 192 in the non-screening group) in 12 centers across the Czech Republic. The screening and non-screening groups were compared for pathology status and clinical characteristics and outcomes. RESULTS: Comparing the screening group to the non-screening group, stages 0, I, and II were observed in 70 % of the screening group vs. 49 % in the non-screening group (p<0.001), while stages III and IV were found in 30 % vs 51 % (p<0.001) of these groups, respectively. Metastatic CRC (M1) was observed in 55 screening group patients vs. 139 non-screening group patients (p<0.001). Palliation treatment was indicated in two patients in the screening group compared to 23 patients in the non-screening group (p= 0.018). Radical surgery could not be performed due to advanced CRC in 7 % of the screening group patients vs. 32 % non-screening group patients (p<0.066), and resection margins R1+R2 were ascertained in 2 % vs 14 % (p< 0.05), respectively. CONCLUSION: Colorectal
S-548