Deep Learning Model Using Stool Pictures for Predicting Endoscopic Mucosal Inflammation in Patients With Ulcerative Colitis.

INTRODUCTION: Stool characteristics may change depending on the endoscopic activity of ulcerative colitis (UC). We developed a deep learning model using stool photographs of patients with UC (DLSUC) to predict endoscopic mucosal inflammation. METHODS: This was a prospective multicenter study conducted in 6 tertiary referral hospitals. Patients scheduled to undergo endoscopy for mucosal inflammation monitoring were asked to take photographs of their stool using smartphones within 1 week before the day of endoscopy. DLSUC was developed using 2,161 stool pictures from 306 patients and tested on 1,047 stool images from 126 patients. The UC endoscopic index of severity was used to define endoscopic activity. The performance of DLSUC in endoscopic activity prediction was compared with that of fecal calprotectin (Fcal). RESULTS: The area under the receiver operating characteristic curve (AUC) of DLSUC for predicting endoscopic activity was 0.801 (95% confidence interval [CI] 0.717-0.873), which was not statistically different from the AUC of Fcal (0.837 [95% CI, 0.767-0.899, DeLong P = 0.458]). When rectal-sparing cases (23/126, 18.2%) were excluded, the AUC of DLSUC increased to 0.849 (95% CI, 0.760-0.919). The accuracy, sensitivity, and specificity of DLSUC in predicting endoscopic activity were 0.746, 0.662, and 0.877 in all patients and 0.845, 0.745, and 0.958 in patients without rectal sparing, respectively. Active patients classified by DLSUC were more likely to experience disease relapse during a median 8-month follow-up (log-rank test, P = 0.002). DISCUSSION: DLSUC demonstrated a good discriminating power similar to that of Fcal in predicting endoscopic activity with improved accuracy in patients without rectal sparing. This study implies that stool photographs are a useful monitoring tool for typical UC.

The Prognostic Utilities of DNA Mismatch Repair Status and KRAS and BRAF Mutation in Korean Colorectal Cancer Patients: The KASID Multicenter Study.

INTRODUCTION: KRAS, BRAF, and DNA mismatch repair (MMR) mutations aid clinical decision-making for colorectal cancer (CRC) patients. To ensure accurate predictions, the prognostic utilities of these biomarkers and their combinations must be individualized for patients with various TNM stages. METHODS: Here, we retrospectively analyzed the clinicopathological features of 904 Korean CRC patients who underwent CRC surgery in three teaching hospitals from 2011 to 2013; we also assessed the prognostic utilities of KRAS, BRAF, and MMR mutations in these patients. RESULTS: The overall frequencies of KRAS and BRAF mutations were 35.8% and 3.2%, respectively. Sixty-nine patients (7.6%) lacking expression of ≥1 MMR protein were considered MMR protein deficient (MMR-D); the remaining patients were considered MMR protein intact. KRAS mutations constituted an independent risk factor for shorter overall survival (OS) in TNM stage I-IV and stage III patients. BRAF mutations were associated with shorter OS in TNM stage I-IV patients. MMR-D status was strongly positive prognostic in TNM stage I-II patients. DISCUSSION/CONCLUSION: To our knowledge, this is the first multicenter study to explore the prognostic utilities of KRAS, BRAF, and MMR statuses in Korean CRC patients. Various combinations of KRAS, BRAF, and DNA MMR mutations serve as genetic signatures that affect tumor behavior; they are prognostic in CRC patients.

Incidence of colonoscopy-related perforation and risk factors for poor outcomes: 3-year results from a prospective, multicenter registry (with videos).

BACKGROUND AND AIMS: Perforation is a life-threatening adverse event of colonoscopy that often requires hospitalization and surgery. We aimed to prospectively assess the incidence of colonoscopy-related perforation in a multicenter registry and to analyze the clinical factors associated with poor clinical outcomes. METHODS: This prospective observational study was conducted at six tertiary referral hospitals between 2017 and 2020, and included patients with colonic perforation after colonoscopy. Poor clinical outcomes were defined as mortality, surgery, and prolonged hospitalization (> 13 days). Logistic regression was used to identify factors associated with poor clinical outcomes. RESULTS: Among 84,673 patients undergoing colonoscopy, 56 had colon perforation (0.66/1000, 95% confidence interval [CI] 0.51-0.86). Perforation occurred in 12 of 63,602 diagnostic colonoscopies (0.19/1000, 95% CI 0.11-0.33) and 44 of 21,071 therapeutic colonoscopies (2.09/1000, 95% CI 1.55-2.81). Of these, 15 (26.8%) patients underwent surgery, and 25 (44.6%) patients had a prolonged hospital stay. One patient (1.8%) died after perforation from a diagnostic colonoscopy. In the multivariate analysis, diagnostic colonoscopy (adjusted odds ratio [aOR] 196.43, p = 0.025) and abdominal rebound tenderness (aOR 17.82, p = 0.012) were independent risk factors for surgical treatment. The location of the sigmoid colon (aOR 18.57, p = 0.048), delayed recognition (aOR 187.71, p = 0.008), and abdominal tenderness (aOR 63.20, p = 0.017) were independent risk factors for prolonged hospitalization. CONCLUSIONS: This prospective study demonstrated that the incidence of colonoscopy-related perforation was 0.66/1000. The incidence rate was higher in therapeutic colonoscopy, whereas the risk for undergoing surgery was higher in patients undergoing diagnostic colonoscopy. Colonoscopy indication (diagnostic vs. therapeutic), physical signs, the location of the sigmoid perforation, and delayed recognition were independent risk factors for poor clinical outcomes in colonoscopy-related perforation.

Does precutting prior to endoscopic piecemeal resection of large colorectal neoplasias reduce local recurrence? A KASID multicenter study.

BACKGROUND: It would be expected that local recurrence could be reduced by performing precutting (with sufficient margins) prior to endoscopic piecemeal mucosal resection (EPMR). We explored the clinical outcomes and local recurrence after precutting EPMR of large colorectal neoplasias. METHODS: Between January 2005 and December 2015, in total, 223 patients with colorectal neoplasias ≥ 2 cm in diameter removed via EPMR in four tertiary hospitals were enrolled. The patients were divided into a precut EPMR group (n = 62) and a non-precut EPMR group (n = 161). We retrospectively evaluated clinical outcomes and factors associated with local recurrence. RESULTS: The mean total procedure time was significantly shorter in the non-precut EPMR group than in the precut EPMR group. However, the number of pieces, and the complete resection and recurrence rates, did not differ significantly [for the latter, precut 8.1% vs. non-precut 9.9%, P = 0.668]. The complete resection rate, number of pieces, and use of argon plasma coagulation (APC) were significantly associated with the local recurrence rate on univariate analysis. In the Cox's proportional hazards model, prophylactic APC [hazard ratio 0.307, 95% confidence interval (CI) 0.114-0.823; P = 0.019] and complete resection rate (odds ratio 0.083, 95% CI 0.011-0.655; P = 0.018) were significantly associated with the local recurrence rate. CONCLUSION: Precutting prior to EPMR did not significantly reduce the local recurrence rate or the number of resected pieces. Histologically complete resection, reducing the number of pieces, and prophylactic APC seem to be important in terms of reducing local recurrence.

Comparison of 1-Year Colectomy Risk Between the US and Korean Patients with Acute Severe Ulcerative Colitis: A Propensity Score Matching Analysis.

BACKGROUND: Colectomy risk after acute severe ulcerative colitis (ASUC) has not been compared between Eastern and Western countries. We compared the 1-year colectomy risk after ASUC between Korea and the USA. METHODS: Data on patients admitted for ASUC to five tertiary referral hospitals in Korea and Mount Sinai Hospital, New York, the USA, between January 2015 and January 2019 were reviewed retrospectively. For comparability between groups, a 1:1 propensity score matching (PSM) was performed. The primary outcome was colectomy, and secondary outcomes were mortality, readmission, and venous thromboembolism (VTE) within 1-year of the index admission for ASUC. The risk of each outcome was compared using Cox proportional hazards model in pre-matched cohort and Kaplan-Meier analysis with log-rank test in post-matched cohort. RESULTS: 290 ASUC patients were included in the study (121 Korea, 169 the USA). After PSM, 56 patients were selected in each group with no significant differences in baseline variables. At 1 year after ASUC, colectomy was less common in Korea than in the USA [3 (5.4%) vs. 24 (42.9%), p < 0.001]. The cumulative colectomy risk was significantly higher in the USA than in Korea in pre-matched cohort [adjusted hazard ratio 7.89, 95% confidence interval 3.23 to 19.22] and in post-matched cohort (log-rank p < 0.001), while there was no difference in cumulative risk of mortality, readmission, and VTE. CONCLUSION: Colectomy risk within 1 year of ASUC is significantly higher in the USA than in Korea. We observed no differences in mortality, readmission, and VTE between the two groups.

Low Bone Mineral Density in Young Patients Newly Diagnosed with Inflammatory Bowel Disease.

BACKGROUND: The prevalence and risk factors of low bone mineral density (BMD) in Asian patients newly diagnosed with inflammatory bowel disease (IBD) have not been fully suggested. AIMS: We aimed to examine the prevalence and risk factors of low BMD in young Korean patients newly diagnosed with IBD. METHODS: We prospectively enrolled 132 patients aged less than 50 years and newly diagnosed with IBD from six tertiary referral centers in Korea between November 2014 and April 2017. BMD was measured by dual-energy X-ray absorptiometry, and then the Z-score was determined. We defined low BMD as a Z-score ≤ - 1.0. RESULTS: Of 68 patients with ulcerative colitis (UC), 22 (32.4%) had low BMD. Also, of 64 patients with Crohn's disease (CD), 24 (37.5%) showed low BMD. Results from multivariate regression analysis identified the risk factors for low BMD as a high level of alkaline phosphatase (ALP) (≥ 140 U/L) (P = 0.010) in UC patients, and being underweight (body mass index ≤ 18.5 kg/m2) (P = 0.017) in CD patients. CONCLUSIONS: Our study showed that about one-third of newly diagnosed IBD Asian patients had low BMD. The clinical factors associated with low BMD were a high level of ALP in UC patients, and being underweight, in CD patients. Therefore, measurements of BMD in young patients should be considered at the diagnosis of IBD.

Fecal calprotectin is more accurate than fecal immunochemical test for predicting mucosal healing in quiescent ulcerative colitis: a prospective multicenter study.

Objective: Non-invasive stool tests, including the fecal immunochemical test (FIT) and fecal calprotectin (FC), are reliable biomarkers for mucosal healing (MH) in ulcerative colitis (UC). However, which fecal test is superior for predicting MH in inactive UC patients requires evaluation. We aimed to compare the accuracy of FIT and FC results for predicting MH in quiescent UC patients.Methods: This prospective, multicenter study was conducted at three tertiary hospitals. UC patients in clinical remission for at least three months underwent colonoscopy and MH was evaluated using the Mayo endoscopic sub-score (MES). The receiver operating characteristic (ROC) curve and cutoff value with the best accuracy for predicting MH were assessed.Results: Among 127 patients, 65 (51.2%) showed MH (MES = 0). The area under the curve (AUC) for predicting MH (MES = 0) was significantly higher for FC than for FIT (AUC 0.858 (95% confidence interval (CI) 0.784-0.913) vs. 0.707 (95% CI 0.620-0.784), p < .001); there was no difference when MH included MES = 1 (MES ≤ 1) (AUC 0.820 (95% CI 0.742-0.883) vs. 0.813 (95% CI 0.734-0.877), p = .891). When the cutoff value was 70 µg/g for FC and 10 ng/mL for FIT, the sensitivity, specificity, positive predictive value and negative predictive value were 89.2, 71, 76.3, and 86.3, respectively, for FC and 92.3, 50, 65.9, and 86.1, respectively, for FIT.Conclusion: FC is more accurate than FIT for predicting MH in quiescent UC patients. The superiority of FC might be related to the distinctive performance of FC in differentiating inflammatory levels, particularly in low-grade mucosal activity.

Sarcopenia Is a New Risk Factor of Nonalcoholic Fatty Liver Disease in Patients with Inflammatory Bowel Disease.

BACKGROUND AND AIM: Recently, sarcopenia has been proposed as an additional risk factor of nonalcoholic fatty liver disease (NAFLD), and there have been no studies in patients with inflammatory bowel disease (IBD). We aimed to analyze the clinical associations between sarcopenia and NAFLD in IBD patients. METHODS: From January 2004 to December 2017, a total of 488 IBD patients, with CT results, were classified according to the presence of NAFLD. Sarcopenia was assessed based on the muscle volume calculated by the total psoas muscle area in the third lumbar region divided by the square of the patient's height (m2). RESULTS: Among the 443 included patients, NAFLD was diagnosed in 49 patients (11.1%). Sarcopenia was noted in 34.9%; it was more common in the NAFLD group (51.0 vs. 33.0%; p = 0.019). In multivariate analysis, metabolic syndrome (odds ratio [OR], 8.63), hyperuricemia (OR, 4.66), small bowel resection (OR, 3.45), and sarcopenia (OR, 2.99) were significant risk factors of NAFLD in IBD patients. In addition, sarcopenia was an independent risk factor after adjustment for age, sex, and other metabolic factors (OR, 2.26). CONCLUSIONS: The prevalence of nonalcoholic fatty liver in IBD patients was 11.1%, and sarcopenia was an independent risk factor.

Impact of Endoscopists' Personality Traits on Adenoma and Polyp Detection Rates in Colonoscopy: A KASID Multicenter Study.

BACKGROUND: The personality traits of endoscopists have been suggested to affect the adenoma detection rate (ADR). We thus evaluated the relationship between endoscopists' personality traits and the ADR during colonoscopy using the Minnesota Multiphasic Personality Inventory-2 (MMPI-2). METHODS: In total, 1230 patients (asymptomatic and aged 50-80 years) who underwent screening or surveillance (≥ 5 years) colonoscopy were recruited from 13 university hospitals by 20 endoscopists between September 2015 and December 2017. We retrospectively measured the ADR, polyp detection rate (PDR), and number of adenomas per colonoscopy (APC). All 20 endoscopists completed all 567 true/false MMPI-2 items. RESULTS: The overall mean colonoscopy withdrawal time, PDR, ADR, and APC were 7.3 ± 2.8 min, 55%, 45.3%, and 0.97 ± 1.58, respectively. No significant difference was observed in the MMPI-2 clinical scales (e.g., hypochondriasis and psychasthenia), content scales (e.g., obsessiveness and type A character), or supplementary scales (e.g., dominance and social responsibility) between the high ADR group (ADR ≥45%, n = 10) and the low ADR group (ADR < 45%, n = 10). In multivariate logistic regression analysis, the ADR was associated significantly with patient age and sex. The ADR was related significantly to endoscopists' colonoscopy experience and the per-minute increase in the colonoscopy withdrawal time (OR 1.21, 95% CI 1.06-1.38, p = 0.005). In a logistic regression analysis adjusted for patient factors, the ADR was associated significantly with ego strength (OR 1.04, 95% CI 1.00-1.09, p = 0.044), as measured by the MMPI-2. CONCLUSIONS: With the exception of ego strength, the endoscopists' personality traits were not associated with adenoma or polyp detection.

Prevalence and Risk Factors of Gastric and Colorectal Cancer after Cholecystectomy.

BACKGROUND: Previous studies on the relationship between gastrointestinal (GI) cancer and cholecystectomy remain inconclusive. We aimed to evaluate this relationship, albeit particularly between cholecystectomy and gastric cancer or colorectal cancer (CRC), and the risk factors of cancer among individuals who have undergone cholecystectomy in Korea. METHODS: In total, 4,222 patients who underwent laparoscopic or open cholecystectomy at our institution between January 2006 and December 2013 were included. Patients who underwent cholecystectomy for gallbladder cancer or were undergoing surgery for GI, hepatic, or pancreatobiliary cancers were excluded, as were those who developed stomach cancer or CRC within a year of their cholecystectomy. The included patients were followed until July 20, 2020. The standardized incidence ratio (SIR) was used to calculate the relative risk of GI cancer in cholecystectomy patients. RESULTS: The median patient age (n = 3,588) at the time of cholecystectomy was 54.0 (range, 19-95) years, and the male-to-female ratio was 1:1.04. The median follow-up period after cholecystectomy was 15.0 (range, 0-146) months. We found a 108% greater risk of CRC (SIR, 2.08; 95% confidence interval [CI], 1.28-3.17) and 154% increased risk of CRC in females (SIR, 2.54; 95% CI, 1.16-4.84). Based on multivariate analysis, an age of > 60 years was a significant risk factor for GI cancer in cholecystectomy patients. CONCLUSION: Cholecystectomy may increase risk of CRC, especially in females. Age was considered a risk factor of GI cancers in patients with history of cholecystectomy.

Relationship between the endoscopic withdrawal time and adenoma/polyp detection rate in individual colonic segments: a KASID multicenter study.

BACKGROUND AND AIMS: Appropriate colonoscopy withdrawal times for individual colonic segments are not well known. The relationship between withdrawal time and adenoma detection rate (ADR)/polyp detection rate (PDR) in individual colonic segments was examined in this study. METHODS: This was a prospective observational study involving 724 patients who underwent colonoscopy screening or surveillance colonoscopy from October 2015 to February 2017 at 10 university hospitals. RESULTS: In the right side of the colon, the ADR (33.2% vs 13.7%, P < .001), PDR, serrated polyp detection rate, and number of adenomas per colonoscopy (APC) were significantly higher when the colonoscopy withdrawal time was ≥2 minutes compared with <2 minutes. When the withdrawal time was ≥4 minutes in the proximal colon and ≥3 minutes in the left segment of the colon, the ADR, PDR, and APC were significantly higher compared with withdrawal times of <4 minutes and <3 minutes, respectively. Multivariate analyses showed that the ADR was significantly associated with withdrawal times of ≥2 minutes in the right side of the colon (odds ratio [OR], 2.98; 95% confidence interval [CI], 1.72-5.15; P < .001), ≥4 minutes in the proximal colon (OR, 4.48; 95% CI, 3.15-6.36; P < .001), and ≥3 minutes in the left segment of the colon (OR, 2.92; 95% CI, 1.74-4.91; P < .001). CONCLUSIONS: The PDR and ADR appeared to be significantly increased when the withdrawal time was ≥2 minutes in the right-sided colon segment, ≥4 minutes in the proximal colon, and ≥3 minutes in the left-sided colon segment compared with shorter withdrawal times.

Changes in Lower Gastrointestinal Bleeding Presentation, Management, and Outcomes Over a 10-Year Span.

BACKGROUND: There are only limited data available on changes in the etiology, management, and clinical outcomes in patients with lower gastrointestinal bleeding over the past decade. STUDY: We compared 2 groups of consecutive patients hospitalized with lower gastrointestinal bleeding during 2 time periods: 2005 to 2007 (301 patients) and 2015 to 2017 (249 patients). RESULTS: Compared with the 2005 to 2007 group, the mean Charlson comorbidity index in the 2015 to 2017 group was higher (5.0±2.6 vs. 6.0±3.0, P=0.028), whereas the use of computerized tomographic angiography and small bowel capsule endoscopy was more common (12.9% vs. 58.1%, P<0.001, and 28.8% vs. 69.0%, P=0.031, respectively). In 2005 to 2007, ischemic colitis (12.0%) was the most common confirmed etiology of bleeding and diverticular bleeding the second most common (8.6%), whereas in 2015 to 2017, diverticular bleeding was the most common etiology (10.4%), followed by angiodysplasia (8.4%). Small bowel bleeding sources were confirmed more often in the 2015 to 2017 group (P=0.017). Endoscopic treatment was attempted in 16.6% of patients in 2005 to 2007 versus 25.3% in 2015 to 2017 (P=0.015). Higher rebleeding rates, longer hospitalization durations (4.6±4.3 vs. 5.5±3.4 d, P=0.019), and a higher proportion of patients needing a transfusion (62.0% vs. 78.4%, P=0.016) were noted in 2015 to 2017. CONCLUSIONS: Over a 10-year span, there were several notable changes: (1) more comorbidities in patients hospitalized for lower gastrointestinal bleeding; (2) marked increase in the use of computerized tomographic angiography and capsule endoscopy for diagnostic evaluation; and (3) longer hospitalization durations and greater need for blood transfusion, possibly reflecting the selection of sicker patients for in-patient management in 2015 to 2017.

Outcomes and Management Strategies for Capsule Retention: A Korean Capsule Endoscopy Nationwide Database Registry Study.

BACKGROUND: The most concerning complication of capsule endoscopy (CE) is capsule retention (CR) in the gastrointestinal (GI) tract; however, the clinical outcomes and management of patients with CR are still uncertain. AIMS: This study aimed to investigate the clinical outcomes and management of CR. METHODS: The outcomes of CR in multiple centers between October 2002 and June 2018 were retrospectively reviewed. Data on CE indication, findings, and management details were analyzed. RESULTS: A total of 2705 consecutive small-bowel CE procedures were performed. CR was detected in 20 cases (0.7%). The most common site of CR was the small bowel (19 cases), followed by the esophagus (one case). In patients who underwent CE, CR was detected in nine (0.6%) of 1397 patients with obscure GI bleeding. Further, CR occurred in 11 (6.5%) of 169 patients with Crohn's disease based on the final diagnoses after CE. Capsule retrieval was safely performed surgically in nine cases and endoscopically in six cases. The retained capsules dislodged after steroid treatment in two cases, whereas three cases of CR resolved without any intervention. In multivariate analysis, the development of abdominal symptoms after CR was a significant predictive factor for requiring endoscopic or surgical interventions for capsule extraction. CONCLUSIONS: This large multicenter study shows that CR is a rare complication with favorable clinical outcomes. Three-fourths of the patients with CR were managed with endoscopic or surgical intervention, which was required particularly in patients with abdominal symptoms after CR.

Risk of metachronous neoplasia on surveillance colonoscopy in young patients with colorectal neoplasia.

BACKGROUND AND AIMS: Few prior reports exist that address the appropriate colonoscopy surveillance interval for individuals <50 years old. We compared the risk of metachronous neoplasia between younger (20-49 years) and older (50-54 years) cohorts. METHODS: This multicenter retrospective cohort study compared the incidence of metachronous neoplasia in younger and older cohorts according to baseline risk stratification. Subjects were eligible if they underwent their first colonoscopy between June 2006 and May 2010 and had at least 1 or more surveillance colonoscopy up to June 2015. RESULTS: Among a total of 10,477 subjects who underwent baseline colonoscopy, 9722 were eligible after excluding 755 subjects. Of those 9722 subjects, 43% underwent surveillance colonoscopy. In the baseline high-risk adenoma group (n = 840), the 3-year risk of metachronous advanced neoplasia was 10.7% in the younger patients on screening colonoscopy and 8.9% in the older patients (P > .1). In the baseline low-risk adenoma group (n = 1869), the 5-year risk of metachronous advanced neoplasia was 4.9% in the younger patients on screening colonoscopy and 5.1% in the older patients (P > .1). Similarly, in the baseline no neoplasia group (n = 7013), the 5-year risk of metachronous advanced neoplasia was 4.1% in the younger patients on screening colonoscopy and 5.6% in the older patients (P > .1). CONCLUSIONS: Considering the similar risk of metachronous advanced neoplasia in younger and older individuals, we suggest a 3-year surveillance interval for high-risk adenoma and a 5-year surveillance interval for low-risk adenoma in young individuals without a strong family history.

An Adjusted Level of Adenoma Detection Rate is Necessary for Adults Below 50 Years Old.

BACKGROUND: Although adenoma prevalence is lower in younger people compared with screening-aged adults 50 years old and above, there is no adjustment recommendation for the target adenoma detection rate (ADR) in young people. Herein, we estimated a different target ADR for adults below 50 years old based on screening colonoscopy findings. MATERIALS AND METHODS: Asymptomatic, average-risk adults below 50 years old who underwent screening colonoscopy were enrolled at 12 endoscopy centers in Korea between February 2006 and March 2012. Screening colonoscopies were stratified into low or high ADR groups with ADR levels of 20% and 25%, respectively. RESULTS: The ADRs from 12 endoscopy centers ranged from 12.1% to 43.8% (median ADR, 24.1%) based on 5272 young adults receiving screening colonoscopies. Using 20% as an ADR level, the risks for metachronous adenoma and advanced adenoma were significantly higher in the low ADR group than the high ADR group (35.4% vs. 25.7%, P<0.001; 8.3% vs. 3.7%, P=0.001, respectively). However, using ADR level of 25%, the risk for metachronous neoplasia was similar in the high and low ADR groups in young adults according to screening colonoscopy. In subgroup analysis, similar findings were found in males, but not in females. CONCLUSIONS: Optimal target ADR may be different between younger and older populations, and the adoption of a 20% target ADR could be used as a performance indicator for young populations.

Predictors and clinical outcomes of follow-up loss in patients with inflammatory bowel disease.

BACKGROUND AND AIM: Nonadherence is a risk factor of disease worsening in inflammatory bowel disease (IBD). We analyzed the frequency, predictors, and clinical outcomes of patients with IBD who are lost to follow-up in outpatient clinics. METHODS: Medical records of 784 IBD patients visiting our IBD clinic between January 2010 and December 2015 were reviewed retrospectively. Overall, 285 newly diagnosed IBD patients who were followed up for at least 12 months were included in the analysis. RESULTS: For 285 IBD patients (161 ulcerative colitis and 124 Crohn's disease), the mean disease duration was 66.3 ± 34.0 months (7-137 months). Forty-two patients (14.7%; 27 ulcerative colitis and 15 Crohn's disease) were lost to follow-up. On multivariate regression analysis, travel time to clinic (odds ratio, 2.37; 95% confidence interval, 1.63-3.45; P = 0.01) and C-reactive protein levels at diagnosis (odds ratio, 0.63; 95% confidence interval, 0.43-0.68; P = 0.01) were significantly associated with follow-up loss. Among the 42 patients lost to follow-up, 36 (85.7%) revisited the clinic. The cause of revisit was disease flare-up in 22 patients (61.1%). Step-up treatment was needed in 15 patients (41.7%). Steroid was introduced in 14 patients (38.9%). Azathioprine and an antitumor necrosis factor agent were newly prescribed in three patients (8.3%) and one patient (2.8%), respectively. CONCLUSIONS: Follow-up loss rate for IBD patients in remission state was 14.7%, and the predictors were far from hospital and low C-reactive protein levels. Because most of follow-up loss patients experienced flare-up, clinicians need to try to encourage patients to keep their adherence.

Disease Activity Patterns Recorded Using a Mobile Monitoring System Are Associated with Clinical Outcomes of Patients with Crohn's Disease.

BACKGROUND: Usefulness of a mobile monitoring system for Crohn's disease (CD) has not been evaluated. We aimed to determine whether disease activity patterns depicted using a web-based symptom diary for CD could indicate disease clinical outcomes. METHODS: Patients with CD from tertiary hospitals were prospectively invited to record their symptoms using a smartphone at least once a week. Disease activity patterns for at least 2 months were statistically classified into good and poor groups based on two factors in two consecutive time frames; the degree of score variation (maximum-minimum) in each frame and the trend (upward, stationary, or downward) of patterns indicated by the difference in the mean activity scores between two time frames. RESULTS: Overall, 220 (82.7%) and 46 (17.3%) patients were included in good and poor groups, respectively. Poor group was significantly more associated with disease-related hospitalization (p = 0.004), unscheduled hospital visits (p = 0.005), and bowel surgery (p < 0.001) during the follow-up period than good group. In the multivariate analysis, poor patterns [odds ratio (OR) 2.62, p = 0.006], stricturing (OR 4.19, p < 0.001) or penetrating behavior (OR 2.27, p = 0.012), and young age at diagnosis (OR 1.06, p = 0.019) were independently associated with disease-related hospitalization. Poor patterns (OR 4.06, p = 0.006) and an ileal location (OR 5.79, p = 0.032) remained independent risk factors for unscheduled visits. Poor patterns (OR 15.2, p < 0.001) and stricturing behavior (OR 9.77, p = 0.004) were independent risk factors for bowel surgery. CONCLUSION: The disease activity patterns depicted using a web-based symptom diary were useful indicators of poor clinical outcomes in patients with CD.

Risk of advanced colorectal neoplasm in patients with more than 10 adenomas on index colonoscopy: A Korean Association for the Study of Intestinal Diseases (KASID) study.

BACKGROUND AND AIM: The number of patients with >10 adenomas is relatively small, and few studies have investigated the risk of colorectal neoplasm (CRN) in these patients. Thus, we aimed to investigate the risk of developing CRN in patients with >10 adenomas and to compare their risk with that of patients with 3-10 adenomas. METHODS: A retrospective multicenter cohort study that included 214 patients with >10 adenomas on index colonoscopy performed between January 2009 and December 2011, and underwent follow-up colonoscopy until December 2015 was conducted. The risk of developing advanced CRN (cancer or advanced adenoma with a diameter of ≥10 mm or with a villous component, or high-grade dysplasia) was investigated and compared with that in patients with 3-10 adenomas (n = 975). RESULTS: Among the 214 patients with >10 adenomas, the mean age was 62.9 years and the mean number of adenomas on index colonoscopy was 14.2. Overall, 57 patients (26.6%) developed an advanced CRN after a mean of 4.3 years from baseline colonoscopy. The respective 3- and 5-year cumulative risks of advanced CRN were 6.8% (95% confidence interval [CI] 2.9-10.7) and 28.7% (95% CI 20.8-36.5), higher than those in the group with 3-10 adenomas (n = 975, P = 0.001). Having >10 adenomas on index colonoscopy was an independent risk factor for developing advanced CRN (odds ratio 2.25, 95% CI 1.49-3.38). CONCLUSIONS: The risk of developing advanced CRN in patients with >10 adenomas was high and statistically higher than that in patients with 3-10 adenomas. Further prospective studies are needed to investigate whether a more intensive surveillance is needed in this group.

Risks of colorectal advanced neoplasia in young adults versus those of screening colonoscopy in patients aged 50 to 54 years.

BACKGROUND AND AIM: The role of screening or diagnostic colonoscopy to detect advanced neoplasia in young cohorts of age < 50 is unclear. This study compared the risk of colorectal neoplasia in a young age cohort against that in 50-54s screening cohort. METHODS: A multi-center retrospective study was conducted at 14 university hospitals to compare the detection rates of neoplasia and advanced neoplasia in screening or diagnostic colonoscopy in the young cohort of < 50s against those in screening colonoscopy in the 50-54s cohort. RESULTS: Among 10 477 eligible subjects, 9765 subjects were enrolled after excluding 712 subjects. Advanced neoplasia detection rates in the young screening cohort was significantly lower than that in the 50-54s screening cohort (5.9% vs 9.3%, P < 0.001). Compared with 50-54s screening cohort, the risk of advanced neoplasia was significantly reduced by 23%, 53%, and 54% in the 45-49s, 40-44s, and 20-39s screening cohorts, respectively. The detection rates of advanced neoplasia in the young diagnostic cohort was 5.0%, which was much lower than 11.8% in 50-54s screening cohort (P < 0.001). Compared with the 50-54s screening cohort, the risk of advanced neoplasia was significantly reduced by 50%, 66%, and 71% in the 45-49s, 40-44s, and 20-39s diagnostic cohorts, respectively. CONCLUSIONS: Colonoscopy to detect advanced neoplasia in young adults aged < 50 years should be reconsidered as their risk of advanced neoplasia on screening or diagnostic colonoscopy was much lower than those of 50-54s screening cohort; however, colonoscopy screening may be justified for high-risk 45-49s cohorts.

Comparison of 4-L Polyethylene Glycol and 2-L Polyethylene Glycol Plus Ascorbic Acid in Patients with Inactive Ulcerative Colitis.

BACKGROUND: Although colonoscopy preparation may cause symptom flares in patients with ulcerative colitis (UC), little is known about the standard preparation regimen in this population. AIM: We aimed to compare 4L polyethylene glycol (4L-PEG) with 2L polyethylene glycol plus ascorbic acid (2L-PEG-Asc) in quiescent UC patients. METHODS: Patients with inactive UC undergoing colonoscopy for surveillance or checkup of mucosal healing were prospectively enrolled at 5 tertiary hospitals. They were randomly assigned to 4L-PEG and 2L-PEG-Asc groups. The Boston Bowel Preparation Scale (BBPS) was used for the preparation quality. Symptoms were assessed using the Simple Clinical Colitis Activity Index (SCCAI) before colonoscopy, at 1 and 4 weeks after the procedure. RESULTS: Overall, 109 patients were included in the study (4L-PEG group 53, 2L-PEG-Asc group 56, the mean age at diagnosis 42.25 years, male 77). The quality of preparation was comparable between the groups (BBPS ≥ 6, 96.2 vs. 92.9%, p = 0.679). Although 26 patients (23.8%) had increased SCCAI scores within 4 weeks after colonoscopy, resulting in a medication dose-up or add-on in 3 patients (2.7%), the rise in scores was not different between the groups. No serious adverse events during preparation were observed in either group. However, the 2L-PEG-Asc group was more likely to be willing to repeat the preparation with the same agent than the 4L-PEG group (82.1 vs. 64.2%, respectively, p = 0.034). CONCLUSION: PEG-based regimens with different volumes are equally effective and safe in inactive UC patients. 2L-PEG-Asc is more acceptable in this population as indicated by the willingness for further usage.