Time Series AI Model for Acute Kidney Injury Detection Based on a Multicenter Distributed Research Network: Development and Verification Study.

Background: Acute kidney injury (AKI) is a marker of clinical deterioration and renal toxicity. While there are many studies offering prediction models for the early detection of AKI, those predicting AKI occurrence using distributed research network (DRN)-based time series data are rare. Objective: In this study, we aimed to detect the early occurrence of AKI by applying an interpretable long short-term memory (LSTM)-based model to hospital electronic health record (EHR)-based time series data in patients who took nephrotoxic drugs using a DRN. Methods: We conducted a multi-institutional retrospective cohort study of data from 6 hospitals using a DRN. For each institution, a patient-based data set was constructed using 5 drugs for AKI, and an interpretable multivariable LSTM (IMV-LSTM) model was used for training. This study used propensity score matching to mitigate differences in demographics and clinical characteristics. Additionally, the temporal attention values of the AKI prediction model's contribution variables were demonstrated for each institution and drug, with differences in highly important feature distributions between the case and control data confirmed using 1-way ANOVA. Results: This study analyzed 8643 and 31,012 patients with and without AKI, respectively, across 6 hospitals. When analyzing the distribution of AKI onset, vancomycin showed an earlier onset (median 12, IQR 5-25 days), and acyclovir was the slowest compared to the other drugs (median 23, IQR 10-41 days). Our temporal deep learning model for AKI prediction performed well for most drugs. Acyclovir had the highest average area under the receiver operating characteristic curve score per drug (0.94), followed by acetaminophen (0.93), vancomycin (0.92), naproxen (0.90), and celecoxib (0.89). Based on the temporal attention values of the variables in the AKI prediction model, verified lymphocytes and calcvancomycin ium had the highest attention, whereas lymphocytes, albumin, and hemoglobin tended to decrease over time, and urine pH and prothrombin time tended to increase. Conclusions: Early surveillance of AKI outbreaks can be achieved by applying an IMV-LSTM based on time series data through an EHR-based DRN. This approach can help identify risk factors and enable early detection of adverse drug reactions when prescribing drugs that cause renal toxicity before AKI occurs.

Development and Verification of Time-Series Deep Learning for Drug-Induced Liver Injury Detection in Patients Taking Angiotensin II Receptor Blockers: A Multicenter Distributed Research Network Approach.

OBJECTIVES: The objective of this study was to develop and validate a multicenter-based, multi-model, time-series deep learning model for predicting drug-induced liver injury (DILI) in patients taking angiotensin receptor blockers (ARBs). The study leveraged a national-level multicenter approach, utilizing electronic health records (EHRs) from six hospitals in Korea. METHODS: A retrospective cohort analysis was conducted using EHRs from six hospitals in Korea, comprising a total of 10,852 patients whose data were converted to the Common Data Model. The study assessed the incidence rate of DILI among patients taking ARBs and compared it to a control group. Temporal patterns of important variables were analyzed using an interpretable timeseries model. RESULTS: The overall incidence rate of DILI among patients taking ARBs was found to be 1.09%. The incidence rates varied for each specific ARB drug and institution, with valsartan having the highest rate (1.24%) and olmesartan having the lowest rate (0.83%). The DILI prediction models showed varying performance, measured by the average area under the receiver operating characteristic curve, with telmisartan (0.93), losartan (0.92), and irbesartan (0.90) exhibiting higher classification performance. The aggregated attention scores from the models highlighted the importance of variables such as hematocrit, albumin, prothrombin time, and lymphocytes in predicting DILI. CONCLUSIONS: Implementing a multicenter-based timeseries classification model provided evidence that could be valuable to clinicians regarding temporal patterns associated with DILI in ARB users. This information supports informed decisions regarding appropriate drug use and treatment strategies.

Validation of the simplified magnetic resonance index of activity by using DWI without gadolinium enhancement to evaluate bowel inflammation in Crohn's disease.

OBJECTIVES: To validate the modified simplified magnetic resonance index of activity (sMARIA) score using DWI on non-contrast magnetic resonance enterography (MRE) to evaluate active inflammation in patients with Crohn's disease (CD), compared to the original sMARIA scoring system, with and without contrast enhancement. METHODS: This retrospective study included 275 bowel segments from 55 CD patients who underwent ileocolonoscopy and MRE within a 2-week period. Two blinded radiologists evaluated original sMARIA on both conventional MRE (CE-sMARIA) and non-contrast MRE (T2-sMARIA). Modified sMARIA was then evaluated using non-contrast MRE, replacing ulcerations with DWI grades. Three scoring systems were compared for diagnostic accuracy of active inflammation, correlation with simple endoscopic score (SES)-CD, and interobserver reproducibility. RESULTS: The AUC of modified sMARIA for detecting active inflammation (0.863, 95% confidence interval [0.803-0.923]) was significantly higher than T2-sMARIA (0.827 [0.773-0.881], p = 0.017), and comparable to CE-sMARIA (0.908 [0.857-0.959], p = 0.122). CE-sMARIA, T2-sMARIA, and modified sMARIA all showed moderate correlation with SES-CD (r = 0.795, 0.722, and 0.777, respectively). Interobserver reproducibility of diffusion restriction (κ, 0.686 [0.602-0.770]) was significantly better than ulcers on conventional MRE (κ, 0.382 [0.212-0.552]; p = 0.001) and T2-weighted image (κ, 0.312 [0.034-0.590]; p = 0.012). CONCLUSIONS: Modified sMARIA using DWI can improve the diagnostic performance of sMARIA on non-contrast MRE, showing comparable performance to sMARIA using contrast-enhanced MRE. KEY POINTS: • DWI can improve the diagnostic performance of non-contrast magnetic resonance enterography (MRE) for assessing active inflammation in patients with Crohn's disease. • Modified simplified magnetic resonance index of activity (sMARIA) using DWI grades in place of ulcers showed comparable diagnostic performance to sMARIA using conventional MRE with contrast-enhanced sequences.

Risk Factors for Surgery in Patients with Intestinal Behçet's Disease During Anti-Tumor Necrosis Factor-Alpha Therapy.

PURPOSE: Behçet's disease (BD) is a chronic inflammatory immune-mediated disease involving multiorgan systems. Gastrointestinal (GI) manifestations of BD include abdominal pain, vomiting, GI bleeding, fistula formation, obstruction, and perforation that might require surgery. Recently, anti-tumor necrosis factor-alpha (anti-TNF-α) therapy has been shown to have favorable outcomes in patients with intestinal BD who are refractory to conventional therapy. This study sought to figure out the risk factors for undergoing surgery during anti-TNF-α therapy in patients with intestinal BD. MATERIALS AND METHODS: In this retrospective analysis of intestinal BD patients who were treated with anti-TNF-α, we collected the baseline patient data including comorbidities, clinical, endoscopic, and radiologic characteristics, and the Disease Activity Index for Intestinal Behçet's Disease at the time of anti-TNF-α initiation. Each potential risk factor was compared. For multivariate analysis, Cox regression was used. RESULTS: A total of 62 patients were considered eligible for analysis, and 15 of them (24.1%) underwent surgery. In univariate analysis, the presence of extraintestinal manifestation, such as joint symptoms and erythrocyte sedimentation rate (ESR), were significantly associated with surgery during therapy. In multivariate analysis, drug response within 4 weeks [hazard ratio (HR), 64.59], skin and joint manifestation (HR, 10.23 and HR, 6.22), geographic ulcer (HR, 743.97), and ESR >42.5 mm/h (HR, 9.16) were found to be factors predictive of undergoing surgery during anti-TNF-α therapy. CONCLUSION: We found five risk factors predictive of surgery in patients with intestinal BD receiving anti-TNF-α therapy, which can guide physicians in selecting appropriate patients between anti-TNF-α therapy and early surgery.

Tumor-Suppressive Effect of Metformin via the Regulation of M2 Macrophages and Myeloid-Derived Suppressor Cells in the Tumor Microenvironment of Colorectal Cancer.

Myeloid-derived suppressor cells (MDSCs) and M2 macrophages in the tumor microenvironment contribute to tumor progression by inducing immune tolerance to tumor antigens and cancer cells. Metformin, one of the most common diabetes drugs, has shown anti-inflammatory and anti-tumor effects. However, the effects of metformin on inflammatory cells of the tumor microenvironment and its underlying mechanisms remain unclarified. In this study, we investigated the effect of metformin on M2 macrophages and MDSCs using monocyte THP-1 cells and a dextran sodium sulfate (DSS)-treated ApcMin/+ mouse model of colon cancer. Metformin decreased the fractions of MDSCs expressing CD33 and arginase, as well as M2 macrophages expressing CD206 and CD163. The inhibitory effect of metformin and rapamycin on MDSCs and M2 macrophages was reversed by the co-treatment of Compound C (an AMP-activated protein kinase (AMPK) inhibitor) or mevalonate. To examine the effect of protein prenylation and cholesterol synthesis (the final steps of the mevalonate pathway) on the MDSC and M2 macrophage populations, we used respective inhibitors (YM53601; SQLE inhibitor, FTI-277; farnesyl transferase inhibitor, GGTI-298; geranylgeranyl transferase inhibitor) and found that the MDSC and M2 populations were suppressed by the protein prenylation inhibitors. In the DSS-treated ApcMin/+ mouse colon cancer model, metformin reduced the number and volume of colorectal tumors with decreased populations of MDSCs and M2 macrophages in the tumor microenvironment. In conclusion, the inhibitory effect of metformin on MDSCs and M2 macrophages in the tumor microenvironment of colon cancers is mediated by AMPK activation and subsequent mTOR inhibition, leading to the downregulation of the mevalonate pathway.

Rebleeding Rate and Risk Factors for Rebleeding after Device-Assisted Enteroscopy in Patients with Obscure Gastrointestinal Bleeding: A KASID Multicenter Study.

INTRODUCTION: The impact of device-assisted enteroscopy (DAE) on long-term rebleeding in patients with obscure gastrointestinal bleeding (OGIB) exhibiting detectable small-bowel lesions remains unclear. We investigated the long-term rebleeding rate and predictive factors for DAE in patients with OGIB. METHOD: Patients with OGIB with small bowel lesions detected through DAE were enrolled at three Korean tertiary hospitals. Predictive risk factors associated with rebleeding were analyzed using the Cox regression analysis. RESULTS: From April 2008 to April 2021, 141 patients were enrolled, including 38 patients (27.0%) with rebleeding. The rebleeding rates at 1, 2, and 3 years were 25.0%, 29.6%, and 31.1%, respectively. The Cox regression analysis revealed that multiple small-bowel lesions (hazard ratio [HR]: 2.551, 95% confidence interval [CI]: 1.157-5.627, p = 0.020), the need for more than five packed red blood cells (RBC) transfusions (HR: 2.704, 95% CI: 1.412-5.181, p = 0.003), and ulcerative lesions (HR: 1.992, 95% CI: 1.037-3.826, p = 0.039) were positively associated with rebleeding. Therapeutic interventions for patients with detectable lesions, overt bleeding (vs. occult bleeding), comorbidities, and medications were not associated with rebleeding. CONCLUSION: More than 25% of patients with OGIB having detectable small-bowel lesions had rebleeding. Patients with multiple lesions, a requirement of more than five packed RBC transfusions, and ulcerative lesions were associated with a higher risk of rebleeding.

Comparative effectiveness of second-line biological therapies for ulcerative colitis and Crohn's disease in patients with prior failure of anti-tumour necrosis factor treatment.

BACKGROUND: Therapeutic options for inflammatory bowel disease (IBD) have increased since the introduction of tumour necrosis factor (TNF) inhibitors a few decades ago. However, direct comparisons of the effectiveness of second-line biological agents in patients with ulcerative colitis (UC) and Crohn's disease (CD) are lacking. METHODS: Patients with UC or CD who experienced anti-TNF treatment failure and subsequently used vedolizumab, ustekinumab, or tofacitinib as a second-line drug were retrospectively recruited. The primary outcomes were the clinical remission rate at week 16 and the cumulative relapse rate 48 weeks after receiving induction therapy. RESULTS: A total of 94 patients with UC or CD experienced anti-TNF treatment failure and received vedolizumab (UC: 37; CD: 28), ustekinumab (CD: 16), or tofacitinib (UC: 13). The clinical remission rates were not significantly different between the vedolizumab and tofacitinib groups in UC patients (56.8% vs. 46.2%, p = 0.509). In CD patients, the clinical remission rates were not significantly different between the vedolizumab and ustekinumab groups (53.6% vs. 50.0%, p = 0.820). Moreover, the cumulative rates of clinical relapse were not significantly different between the vedolizumab and tofacitinib groups in UC patients and between the vedolizumab and ustekinumab groups in CD patients (p = 0.396 and p = 0.692, respectively). Safety profiles were also similar among the treatment groups in both UC and CD patients. CONCLUSIONS: After prior anti-TNF therapy failure, vedolizumab and tofacitinib in UC patients and vedolizumab and ustekinumab in CD patients were not significantly different in terms of the efficacy in inducing and maintaining a clinical response.

Comparison of Sensitivity Encoding (SENSE) and Compressed Sensing-SENSE for Contrast-Enhanced T1-Weighted Imaging in Patients With Crohn Disease Undergoing MR Enterography.

BACKGROUND. Long acquisition times for breath-hold contrast-enhanced (CE) T1-weighted imaging in MR enterography (MRE) protocols result in reduced image quality. OBJECTIVE. The purpose of this study was to compare CE T1-weighted imaging performed using sensitivity encoding (SENSE) and compressed sensing-SENSE (CS-SENSE) in terms of image quality and diagnostic performance for active inflammation in Crohn disease (CD). METHODS. This retrospective study included 41 patients (31 men, 10 women; mean age, 34 ± 12 [SD] years) who underwent MRE for known or suspected CD between June 2020 and September 2020. MRE was performed in one of two scanning rooms depending on scheduling availability. Per institutional protocol, in one room, the enteric phase was acquired using SENSE (acceleration factor, 3) and the portal phase was acquired using CS-SENSE (acceleration factor, 5); this order was reversed in the other room. Two radiologists independently assessed sequences for subjective image quality measures at the patient level and for active inflammation at the bowel-segment level. Mean image quality scores between readers were computed. Diagnostic performance for active inflammation was compared between SENSE and CS-SENSE using generalized estimating equations; a separate experienced radiologist reviewed the full MRE protocol to establish the reference standard. RESULTS. The mean acquisition time of CE T1-weighted imaging was 17.2 ± 1.1 seconds for SENSE versus 11.5 ± 0.8 seconds for CS-SENSE (p < .001). CS-SENSE scored significantly better than SENSE in overall image quality (4.2 ± 0.7 vs 3.7 ± 1.1; p = .02), motion artifacts (4.0 ± 0.8 vs 3.6 ± 1.2; p = .006), and aliasing artifacts (4.8 ± 0.4 vs 4.2 ± 0.6; p < .001). CS-SENSE scored significantly worse than SENSE in synthetic appearance (4.6 ± 0.5 vs 4.8 ± 0.4; p = .003). Contrast, sharpness, and blurring were not different between sequences (p > .05). For reader 1, CS-SENSE, compared with SENSE, showed a sensitivity of 86% versus 81% (p = .09), specificity of 88% versus 83% (p = .08), and accuracy of 87% versus 82% (p = .56). For reader 2, CS-SENSE, compared with SENSE, showed a sensitivity of 92% versus 79% (p = .006), specificity of 90% versus 98% (p = .16), and accuracy of 91% versus 86% (p = .002). CONCLUSION. Use of CS-SENSE for CE T1-weighted imaging in MRE protocols results in reduced scan times with reduced artifact and improved image quality. CLINICAL IMPACT. The benefits of CS-SENSE in MRE protocols may improve the diagnostic performance for active inflammation in CD.

Radiology plus ileocolonoscopy versus radiology alone in Crohn's disease: prognosis prediction and mutual agreement.

BACKGROUND/AIMS: The optimal tools for monitoring Crohn's disease (CD) are controversial. We compared radiology plus ileocolonoscopy and radiology alone in terms of prognosis prediction and evaluated the agreement between radiologic and ileocolonoscopic findings in patients with CD. METHODS: Patients with CD who were followed up with computed tomography enterography (CTE) or magnetic resonance enterography (MRE) alone or CTE or MRE plus ileocolonoscopy were retrospectively recruited. Time to relapse was investigated to evaluate the difference in prognosis using the log-rank and Cox regression tests, and the agreement between radiologic and ileocolonoscopic findings was determined using a kappa value. RESULTS: A total of 501 patients with CD in clinical remission who underwent CTE or MRE and/or ileocolonoscopy were analyzed. Of these, 372 (74.3%) patients underwent CTE or MRE alone and 129 (25.7%) patients underwent CTE or MRE plus ileocolonoscopy. The cumulative maintenance rate of clinical remission between the two groups was not significantly different (p = 0.526, log-rank test). In multivariate analysis, age <40 years (hazard ratio [HR], 2.756; 95% confidence interval [CI], 1.263 to 6.013) and a history of steroid use (HR, 2.212; 95% CI, 1.258 to 3.577) were found to independently predict an increased risk for clinical relapse in patients with CD in clinical remission. Radiologic and ileocolonoscopic findings had a moderate degree of agreement (κ = 0.401, -0.094 to 0.142). The comparison of agreement between radiologic and ileocolonoscopic findings was the highest in the anastomotic site (κ = 0.749, -0.168 to 0.377). CONCLUSION: Radiology plus ileocolonoscopy was not superior to radiology alone in predicting the prognosis of CD.

Continued Postoperative Use of Tumor Necrosis Factor-α Inhibitors for the Prevention of Crohn's Disease Recurrence.

Background/Aims: Many patients with Crohn's disease (CD) undergo intestinal resection during the disease course. Despite surgery, postoperative recurrence (POR) commonly occurs. Although postoperative use of tumor necrosis factor α (TNF-α) inhibitors is known to be effective in preventing POR, few studies have evaluated the effectiveness of continuing the same TNF-α inhibitors postoperatively in patients who received TNF-ɑ inhibitors before surgery. Methods: This retrospective observational study was performed in a single tertiary medical center. We retrospectively reviewed patients who had undergone the first intestinal resection due to CD and divided them into two groups: TNF-α inhibitor users in both the preoperative and postoperative periods, and TNF-α inhibitor users in only the preoperative period. We compared the clinical outcomes between these two groups. Results: In total, 45 patients who used TNF-α inhibitors preoperatively were recruited. Among them, TNF-α inhibitors were used postoperatively in 20 patients (44.4%). The baseline characteristics except age at diagnosis were similar in both groups. The rates of surgical and endoscopic recurrence were not different between the two groups, but the cumulative clinical recurrence rate was significantly lower in the postoperative TNF-α inhibitors group (log-rank p=0.003). In multivariate Cox regression analysis, postoperative TNF-α inhibitors use was significantly associated with a decreased risk of clinical recurrence (adjusted hazard ratio, 0.204; 95% confidence interval, 0.060 to 0.691; p=0.011). Conclusions: Continuing TNF-α inhibitors postoperatively in patients who were receiving TNF-α inhibitors before surgery significantly reduced the rate of clinical recurrence. For patients with CD who received TNF-α inhibitors preoperatively, continuing their use after surgery could be recommended.

Deep learning model for diagnosing gastric mucosal lesions using endoscopic images: development, validation, and method comparison.

BACKGROUND AND AIMS: Endoscopic differential diagnoses of gastric mucosal lesions (benign gastric ulcer, early gastric cancer [EGC], and advanced gastric cancer) remain challenging. We aimed to develop and validate convolutional neural network-based artificial intelligence (AI) models: lesion detection, differential diagnosis (AI-DDx), and invasion depth (AI-ID; pT1a vs pT1b among EGC) models. METHODS: This study included 1366 consecutive patients with gastric mucosal lesions from 2 referral centers in Korea. One representative endoscopic image from each patient was used. Histologic diagnoses were set as the criterion standard. Performance of the AI-DDx (training/internal/external validation set, 1009/112/245) and AI-ID (training/internal/external validation set, 620/68/155) was compared with visual diagnoses by independent endoscopists (stratified by novice [<1 year of experience], intermediate [2-3 years of experience], and expert [>5 years of experience]) and EUS results, respectively. RESULTS: The AI-DDx showed good diagnostic performance for both internal (area under the receiver operating characteristic curve [AUROC] = .86) and external validation (AUROC = .86). The performance of the AI-DDx was better than that of novice (AUROC = .82, P = .01) and intermediate endoscopists (AUROC = .84, P = .02) but was comparable with experts (AUROC = .89, P = .12) in the external validation set. The AI-ID showed a fair performance in both internal (AUROC = .78) and external validation sets (AUROC = .73), which were significantly better than EUS results performed by experts (internal validation, AUROC = .62; external validation, AUROC = .56; both P < .001). CONCLUSIONS: The AI-DDx was comparable with experts and outperformed novice and intermediate endoscopists for the differential diagnosis of gastric mucosal lesions. The AI-ID performed better than EUS for evaluation of invasion depth.

Risk of Diabetes in Subjects with Positive Fecal Immunochemical Test: A Nationwide Population-Based Study.

BACKGROUND: Positive fecal immunochemical test (FIT) results have been recently suggested as a risk factor for systemic inflammation. Diabetes induces inflammation in the gastrointestinal tract via several ways. We investigated the association between FIT results and the incidence of diabetes. METHODS: A total of 7,946,393 individuals aged ≥50 years from the National Cancer Screening Program database who underwent FIT for colorectal cancer (CRC) screening from 2009 to 2012 were enrolled. The primary outcome was newly diagnosed diabetes based on the International Classification of Disease 10th revision codes and administration of anti-diabetic medication during the follow-up period. RESULTS: During a mean follow-up of 6.5 years, the incidence rates of diabetes were 11.97, 13.60, 14.53, and 16.82 per 1,000 personyears in the FIT negative, one-positive, two-positive, and three-positive groups, respectively. The hazard ratios (HRs) for the incidence of diabetes were 1.14 (95% confidence interval [CI], 1.12 to 1.16; HR, 1.21; 95% CI, 1.16 to 1.27; and HR, 1.40; 95% CI, 1.28 to 1.55) in the one-positive, two-positive, and three-positive FIT groups compared with the FIT negative group, respectively. The effect was consistent in individuals with normal fasting blood glucose (adjusted HR 1.55 vs. 1.14, P for interaction <0.001). CONCLUSION: Positive FIT results were associated with a significantly higher risk of diabetes, suggesting that the FIT can play a role not only as a CRC screening tool, but also as a surrogate marker of systemic inflammation; thus, increasing the diabetes risk.

C-reactive protein is associated with postoperative outcomes in patients with intestinal Behçet's disease.

BACKGROUND: Patients with intestinal Behçet's disease (BD) frequently undergo intestinal resections, which significantly affects postoperative morbidity and mortality. The aim of this study was to identify the association between C-reactive protein (CRP) levels and postoperative outcomes in patients with intestinal BD who underwent surgical bowel resection. METHODS: Patients who were diagnosed with intestinal BD and underwent intestinal surgery due to BD at Severance Hospital between November 2005 and April 2018 were retrospectively investigated. Clinical relapse was defined as a disease activity index of BD (DAIBD) > 40, existence of newly added medications, re-hospitalization, or re-operation related to intestinal BD. The relationship between CRP level and postoperative outcomes was analyzed, and a receiver operating characteristic (ROC) curve was drawn to specify a cut-off value. RESULTS: Ninety patients with intestinal BD were included. Among them, 44 were male (48.9%), and the median age at diagnosis was 38 years (range, 11-69 years). The median total disease follow-up duration was 130 months (range, 3-460 months). Forty patients (44.4%) underwent laparoscopic surgery. A higher CRP level immediately after surgery was significantly associated with postoperative complications (OR 1.01, 95% CI 1.004-1.018, p < 0.01), re-operation (hazard ratio [HR] 1.01, 95% CI 1.005-1.020, p < 0.01), and re-admission (HR 1.01, 95% CI 1.006-1.017 p < 0.01). The ROC curve showed that CRP predicts the risk of postoperative complications (p < 0.01) at a cut-off value of 41.9% with a sensitivity of 60.0% and specificity of 67.7%. CONCLUSIONS: Postoperative CRP levels in patients with intestinal BD undergoing surgical resection were associated with postoperative outcomes.

Trajectories in glycated hemoglobin and body mass index in children and adolescents with diabetes using the common data model.

We evaluated trajectories of glycated hemoglobin (HbA1c) levels and body mass index z-scores (BMIz) for 5 years after diagnosis among Korean children and adolescents with type 1 diabetes (T1D) or type 2 diabetes (T2D) using the common data model. From the de-identified database of three hospitals, 889 patients < 15 years of age diagnosed with T1D or T2D (393 boys, 664 T1D patients) were enrolled. Diagnosis was defined as first exposure to antidiabetic drug at each center. Compared with T2D patients, T1D patients had lower BMIz at diagnosis (- 0.4 ± 1.2 vs. 1.5 ± 1.4, p < 0.001) and 3 months (- 0.1 ± 1.0 vs. 1.5 ± 1.5, p < 0.001), and higher HbA1c levels at diagnosis (10.0 ± 2.6% vs. 9.5 ± 2.7%, p < 0.01). After 3 months, HbA1c levels reached a nadir of 7.6% and 6.5% in T1D and T2D patients, respectively, followed by progressive increases; only 10.4% of T1D and 29.7% of T2D patients achieved the recommended HbA1c target (< 7.0%) at 60 months. T1D patients showed consistent increases in BMIz; T2D patients showed no significant change in BMIz during follow-up. Peri-pubertal girls with T1D had higher HbA1c and BMIz values. Achieving optimal glycemic control and preventing obesity should be emphasized in pediatric diabetes care.

Long-term Outcomes after the Discontinuation of Anti-Tumor Necrosis Factor-α Therapy in Patients with Inflammatory Bowel Disease under Clinical Remission: A Korean Association for the Study of Intestinal Disease Multicenter Study.

Background/Aims: Our study aimed to evaluate the long-term outcomes and risk factors for relapse after anti-tumor necrosis factor (TNF)-α cessation in inflammatory bowel disease (IBD) patients because they are not well established. Methods: A retrospective multicenter cohort study was conducted involving patients with Crohn's disease (CD) or ulcerative colitis (UC) from 10 referral hospitals in Korea who discontinued firstline anti-TNF therapy after achieving clinical remission. Results: A total of 109 IBD patients (71 CD and 38 UC) with a median follow-up duration of 56 months were analyzed. The cumulative relapse rates at 1, 3, and 5 years were 11.3%, 46.7%, and 62.5% for CD patients and 28.9%, 45.3%, and 60.9% for UC patients. Multivariable Cox analysis revealed that discontinuation owing to the clinician's decision was associated with lower risk of relapse (vs patient's preference: hazard ratio [HR], 0.13; 95% confidence interval [CI], 0.04 to 0.48; p=0.002) and adalimumab use was associated with higher risk of relapse (vs infliximab: HR, 4.42; 95% CI, 1.24 to 17.74; p=0.022) in CD patients. Mucosal healing was associated with lower risk of relapse (vs nonmucosal healing: HR, 0.12; 95% CI, 0.02 to 0.83; p=0.031) in UC patients. Anti-TNF re-induction was provided to 52 patients, and a response was obtained in 50 patients. However, 25 of them discontinued retreatment owing to a loss of response (n=15), the patient's preference (n=6), and other factors (n=4). Conclusions: More than 60% of IBD patients in remission under anti-TNF therapy relapsed within 5 years of treatment cessation. Anti-TNF re-induction was effective. However, half of the patients discontinued anti-TNF therapy, and 50% of these patients discontinued treatment owing to loss of response.

Prognostic Value of Terminal Ileal Inflammation in Patients with Ulcerative Colitis.

Background/Aims: Few studies have investigated terminal ileal lesions and their prognostic value in patients with ulcerative colitis (UC). We evaluated the clinical significance of these lesions as a prognostic factor in patients with UC who were in clinical remission. Methods: We retrospectively selected 567 of 4,066 colonoscopy reports that included positive findings from orificial observations of the terminal ileum (TI) and appendix in patients with UC. We finally recruited patients who were in clinical remission (n=204). We compared the clinical courses, including relapse and other prognostic parameters associated with UC, between the groups. Results: The baseline patient characteristics were not significantly different between patients with (n=69, TI+ group) and without TI lesions (n=135, TI- group), although there were more never-smokers in the TI+ group (n=57 [82.6%] in the TI+ group; n=86 [63.7%] in the TI- group; p=0.005). Of note, appendiceal orifice inflammation (AOI) was less frequently found in the TI+ group (14.5%) than in the TI- group (71.9%, p<0.001). The cumulative relapse rate was numerically higher in the TI- group, but it was not significantly different according to the Kaplan-Meier analysis (p=0.116). Multivariate Cox regression analysis also revealed advanced age at diagnosis as the most significant factor (adjusted hazard ratio, 0.964; 95% confidence interval, 0.932 to 0.998; p=0.037), but neither TI inflammation nor AOI were significantly associated with the cumulative relapse rate in patients with UC in clinical remission. Conclusions: For patients with UC in clinical remission, neither terminal ileal lesions nor AOI had significant clinical or predictive value for future relapse.

Proteinuria Is Associated with the Development of Crohn's Disease: A Nationwide Population-Based Study.

BACKGROUND AND AIMS: The impact of proteinuria and its severity on the incidence of inflammatory bowel disease (IBD) has not yet been studied. We aimed to determine the association between proteinuria measured by urine dipstick tests and the development of IBD. METHODS: This nationwide population-based study was conducted using the Korean National Health Insurance Service (NHIS) database. A total of 9,917,400 people aged 20 years or older who had undergone a national health examination conducted by the NHIS in 2009 were followed up until 2017. The study population was classified into four groups-negative, trace, 1+, and ≥ 2+-according to the degree of proteinuria measured by the urine dipstick test. The primary endpoint was newly diagnosed IBD, Crohn's disease (CD), or ulcerative colitis (UC) during the follow-up period. RESULTS: Compared with the dipstick-negative group, the incidence of CD significantly increased according to the degree of proteinuria (adjusted hazard ratio [aHR] with 95% confidence interval [CI], 1.01 [0.703-1.451], 1.515 [1.058-2.162], and 2.053 [1.301-3.24] in the trace, 1+, and ≥ 2+ dipstick groups, respectively; p for trend 0.007). However, there was no significant difference in the incidence of UC according to the degree of proteinuria (aHR with 95% CI, 1.12 [0.949-1.323], 0.947 [0.764-1.174], and 1.009 [0.741-1.373] in the trace, 1+, and ≥ 2+ dipstick groups, respectively; p for trend 0.722). In the subgroup analysis, dipstick-positive proteinuria independently increased the incidence of CD regardless of the subgroup. However, dipstick-positive proteinuria was associated with the risk of UC in those with diabetes mellitus and not in those without diabetes mellitus (aHR, 1.527 vs. 0.846; interaction p-value 0.004). The risk of CD was increased or decreased according to proteinuria changes but not associated with the risk of UC. CONCLUSION: Proteinuria, measured by the dipstick test, is strongly associated with the development of CD.