Enhancing Calprotectin's Predictive Power as a Biomarker of Endoscopic Activity in Ulcerative Colitis: A Machine Learning Use Case.

Ulcerative colitis (UC) is a chronic inflammatory bowel disease characterized by periods of exacerbation and remission, making disease monitoring and management challenging. Endoscopy, the gold standard for assessing disease activity and severity, involves invasive procedures and is associated with patient discomfort and risks. Using machine learning (ML) to combine fecal calprotectin with other clinical or biological tests can significantly enhance the non-invasive prediction of endoscopic disease activity (EDA) in UC. Aim: To prove that by fusing fecal calprotectin with other clinical data into an ML model, the performance of the non-invasive prediction of EDA can be significantly improved. Methods: We conducted a prospective, observational, single-center study encompassing 103 patients diagnosed with UC. We employed multilayer perceptron models as the core ML algorithm for predicting EDA. For the constructed models, we utilized the varImp function from the caret library in R to assess the significance of each variable in predicting the outcome. Results: Calprotectin as a sole predictor obtained an accuracy of 70% and an area under the curve (AUC) of 0.68. Combining calprotectin with the list of selected predictors that were fed to the MLP models improved accuracy and the AUC. The accuracy of the algorithm on the test set was 85%. Similarly, the AUC increased to 0.93. This is the first study to propose the use of calprotectin as a predictor in an ML model to estimate UC endoscopic disease activity. Conclusion: The deployment of this ML model can furnish doctors and patients with valuable evaluation of endoscopic disease activity which can be highly beneficial for individuals with UC who need long-term treatment.

Accurate, Robust, and Scalable Machine Abstraction of Mayo Endoscopic Subscores From Colonoscopy Reports.

BACKGROUND: The Mayo endoscopic subscore (MES) is an important quantitative measure of disease activity in ulcerative colitis. Colonoscopy reports in routine clinical care usually characterize ulcerative colitis disease activity using free text description, limiting their utility for clinical research and quality improvement. We sought to develop algorithms to classify colonoscopy reports according to their MES. METHODS: We annotated 500 colonoscopy reports from 2 health systems. We trained and evaluated 4 classes of algorithms. Our primary outcome was accuracy in identifying scorable reports (binary) and assigning an MES (ordinal). Secondary outcomes included learning efficiency, generalizability, and fairness. RESULTS: Automated machine learning models achieved 98% and 97% accuracy on the binary and ordinal prediction tasks, outperforming other models. Binary models trained on the University of California, San Francisco data alone maintained accuracy (96%) on validation data from Zuckerberg San Francisco General. When using 80% of the training data, models remained accurate for the binary task (97% [n = 320]) but lost accuracy on the ordinal task (67% [n = 194]). We found no evidence of bias by gender (P = .65) or area deprivation index (P = .80). CONCLUSIONS: We derived a highly accurate pair of models capable of classifying reports by their MES and recognizing when to abstain from prediction. Our models were generalizable on outside institution validation. There was no evidence of algorithmic bias. Our methods have the potential to enable retrospective studies of treatment effectiveness, prospective identification of patients meeting study criteria, and quality improvement efforts in inflammatory bowel diseases.

Our accurate pair of models automatically classify colonoscopy reports by Mayo endoscopic subscore and abstain from prediction appropriately. Our methods can enable large-scale electronic health record studies of treatment effectiveness, prospective identification of patients for clinical trials, and quality improvement efforts in ulcerative colitis.

Platelet/Albumin ratio and plateletcrit levels are potential new biomarkers for assessing endoscopic inflammatory bowel disease severity.

BACKGROUND: Endoscopy is currently recognized as the gold standard for assessing inflammatory bowel disease (IBD) severity. However, because the procedure is costly and invasive, endoscopy is not suitable for frequently monitoring intestinal inflammation. In this study, our aim was to identify noninvasive, low cost, and convenient biomarkers for identifying endoscopic IBD activity. METHODS: In total, 246 patients with IBD (131 with Ulcerative colitis (UC) and 115 with Crohn's disease (CD)) and 369 healthy controls were recruited for this retrospective study. IBD activity was evaluated using endoscopic and clinical examinations. The potential of several inflammatory biomarkers, including platelets (PLT), plateletcrit (PCT), albumin (ALB), highly sensitive C-reactive protein (hs-CRP), erythrocyte sedimentation rate (ESR), and platelet-to-albumin ratio (PLT/ALB) to assess endoscopic IBD activity was evaluated using receiver operating characteristic (ROC) analyses. RESULTS: PLT/ALB ratio, PLT, ALB, and hs-CRP levels were correlated with Mayo scores in UC patients, while PCT, PLT, fibrinogen (FIB), PLT/ALB ratio, hs-CRP, and ESR levels were correlated with Simple Endoscopic Scores for CD (SES-CD) in CD patients. ROC analyses showed that the area under the curve (AUC) value for the PLT/ALB ratio (0.705) was greater than hs-CRP (0.607) and ESR (0.552) values in UC patients. The AUC value for PCT (0.779) was greater than hs-CRP (0.698) and ESR (0.746) values in CD patients. CONCLUSION: PLT/ALB ratio and PCT biomarkers were the most appropriate of all tested inflammatory biomarkers for assessing endoscopic IBD activity in UC and CD patients, respectively.

Validating the Simplified Endoscopic Mucosal Assessment for Crohn's Disease: A Novel Method for Assessing Disease Activity.

BACKGROUND: To demonstrate treatment efficacy in Crohn's disease (CD), regulatory authorities require that trials include an endoscopic remission/response end point; however, standardized endoscopic assessment of disease activity, such as the Simple Endoscopic Score for Crohn's Disease (SES-CD), is not typically recorded by clinicians in practice or outside of clinical trials. The novel Simplified Endoscopic Mucosal Assessment for Crohn's Disease (SEMA-CD) was developed to be easy to use in routine clinical practice and as a trial end point. We conducted a study to assess and validate the reliability and feasibility of SEMA-CD as a measure of endoscopic disease activity. METHODS: Pre- and post-treatment ileocolonoscopy videos of pediatric (n = 36) and adult (n = 74) CD patients from 2 ustekinumab clinical trials were each scored with SEMA-CD by 2 to 3 professional central readers, blinded to clinical history and other video scorings; the correlation between SEMA-CD and SES-CD previously completed during the trials was assessed. Sensitivity to change, inter- and intrarater reliability, and comparative ease of scoring were also assessed. RESULTS: The SEMA-CD strongly correlated with SES-CD (Spearman ρ = 0.89; 95% confidence interval, 0.86-0.92). Pre- to post-treatment changes in SEMA-CD vs in SES-CD were strongly correlated, and the correlation remained strong between the scores when compared by study population (pediatric, adult), disease severity, and video quality. Intra- and inter-rater reliability were good, and SEMA-CD was rated easier than SES-CD to score 63.0% of the time, although slightly more difficult than SES-CD to score <1.0% of the time. CONCLUSIONS: The SEMA-CD is reliable, reproducible, sensitive to change, and easy to use in both pediatric and adult patients with CD.

Predictive models for endoscopic disease activity in patients with ulcerative colitis: Practical machine learning-based modeling and interpretation.

Background: Endoscopic disease activity monitoring is important for the long-term management of patients with ulcerative colitis (UC), there is currently no widely accepted non-invasive method that can effectively predict endoscopic disease activity. We aimed to develop and validate machine learning (ML) models for predicting it, which are desired to reduce the frequency of endoscopic examinations and related costs. Methods: The patients with a diagnosis of UC in two hospitals from January 2016 to January 2021 were enrolled in this study. Thirty nine clinical and laboratory variables were collected. All patients were divided into four groups based on MES or UCEIS scores. Logistic regression (LR) and four ML algorithms were applied to construct the prediction models. The performance of models was evaluated in terms of accuracy, sensitivity, precision, F1 score, and area under the receiver-operating characteristic curve (AUC). Then Shapley additive explanations (SHAP) was applied to determine the importance of the selected variables and interpret the ML models. Results: A total of 420 patients were entered into the study. Twenty four variables showed statistical differences among the groups. After synthetic minority oversampling technique (SMOTE) oversampling and RFE variables selection, the random forests (RF) model with 23 variables in MES and the extreme gradient boosting (XGBoost) model with 21 variables in USEIS, had the greatest discriminatory ability (AUC = 0.8192 in MES and 0.8006 in UCEIS in the test set). The results obtained from SHAP showed that albumin, rectal bleeding, and CRP/ALB contributed the most to the overall model. In addition, the above three variables had a more balanced contribution to each classification under the MES than the UCEIS according to the SHAP values. Conclusion: This proof-of-concept study demonstrated that the ML model could serve as an effective non-invasive approach to predicting endoscopic disease activity for patients with UC. RF and XGBoost, which were first introduced into data-based endoscopic disease activity prediction, are suitable for the present prediction modeling.

Assessment of endoscopic response using pan-enteric capsule endoscopy in Crohn's disease; the Sensitivity to Change (STOC) study.

BACKGROUND: The potential use of pan-enteric capsule endoscopy (pan-CE) to evaluate mucosal changes during treatment has not been evaluated. We aimed to assess the ability of pan-CE to measure changes in mucosal disease activity before and after starting biologic treatment in Crohn's disease patients. METHODS: In this two-center prospective study, patients with clinical and biochemical signs of active Crohn's disease underwent pan-CE before and 8 to 12 weeks after treatment initiation with infliximab, adalimumab or vedolizumab. Endoscopic disease activity was assessed using the simple endoscopic score for Crohn's disease (SES-CD) and the Crohn's disease endoscopic index of severity (CDEIS), expanded with two segments (i.e., the jejunum and pre-terminal ileum). Occurrence of endoscopic remission (i.e., absence of ulcers) and endoscopic response (i.e., 50% decrease in SES-CD and CDEIS scores compared to baseline) was assessed and the standardized effect size was calculated. RESULTS: Twenty-two patients (50% females) completed the study. Endoscopic remission was observed in 6 out of 22 (27%) patients and 13/22 patients (59%) showed an endoscopic response for both the SES-CD and CDEIS score. Median SES-CD and CDEIS scores decreased from 16.0 (IQR 10.0 - 24.0) to 6.0 (IQR 2.8 - 12.0, p = .001) and from 7.1 (IQR 4.6 - 11.2) to 3.0 (IQR 0.9 - 6.0, p = .001), respectively. The standardized effect size was 1.44 and 1.24 for the SES-CD and CDEIS, respectively. No adverse events related to pan-CE were reported. CONCLUSION: Pan-CE was a useful technique to assess changes in mucosal disease activity in Crohn's disease patients.

Platelet-to-lymphocyte percentage ratio index: a simple non-invasive index to monitor the endoscopic activity in Crohn's disease.

BACKGROUND: Recent evidence has shown that the complete blood count (CBC) is abnormal in patients with Crohn's disease (CD). We aimed to investigate an effective CBC parameter and explore its impact on disease activity in a large CD cohort. METHODS: We performed a retrospective analysis of patients with established CD who underwent clinically indicated endoscopy at four tertiary centres in China between 2016 and 2020. Individual variables of the Simple Endoscopic Score for CD, CBC parameters, C-reactive protein (CRP) levels, erythrocyte sedimentation rate, and faecal calprotectin (FC) were independently reviewed by different investigators. The hold-out method was used to verify the predictive power of the established model. RESULTS: Data from a total of 1388 endoscopic procedures performed for 882 eligible CD patients were available with routine blood parameters and related indicators. The model using platelet-to-lymphocyte percentage ratio (PLpR) had high accuracy for identifying patients in endoscopic remission (ER), with an area under the curve (AUC) of 0.785 [95% confidence interval (CI): 0.784-0.787], which was comparable with that for CRP (AUC: 0.775, 95% CI: 0.774-0.777). Notably, the AUC of PLpR was significantly higher than that of CRP in patients with colonic disease and with a history of surgery. Moreover, after combining the FC with PLpR, the AUC value of FC + PLpR increased up to 0.892 (95% CI: 0.890-0.894) for identifying ER. CONCLUSIONS: We explored an index (PLpR) to identify CD patients in ER based on platelet and lymphocyte percentage from the CBC. PLpR helped evaluate the degree of disease activity and monitor the therapeutic response.

Validation and Investigation of the Operating Characteristics of the Ulcerative Colitis Endoscopic Index of Severity.

BACKGROUND: The Ulcerative Colitis Endoscopic Index of Severity (UCEIS) is a novel instrument to evaluate endoscopic disease activity. It has been demonstrated to outperform the more widely used Mayo endoscopic score (MES) in predicting long-term prognosis, including the need for colectomy. Despite its potential benefits, many clinicians still prefer to use MES because its operating characteristics are better defined and its grades are more readily applicable to clinical decision-making. The aims of our study were to quantify the UCEIS cutoff most closely associated with the need for treatment escalation and to perform a validation exercise using MES and clinical, biochemical, and histological measures of disease activity. METHODS: Endoscopies performed in UC patients between November 2016 and January 2018 were retrospectively reviewed. Agreement between the UCEIS and MES was quantified using Kappa (κ) statistics. A UCEIS cutoff for treatment escalation was calculated using chi-square, receiver operating characteristic curve, and area under the curve (AUC) analyses. The Pearson correlation coefficient was used to compare linear relationships between UCEIS and clinical (Simple Clinical Colitis Activity Index [SCCAI]), biochemical (C-reactive protein [CRP]), and histological (Nancy Histological Index [NHI]) activity. RESULTS: Two hundred one (56%) procedures documented both UCEIS and MES, demonstrating substantial agreement (κ = 0.713; P < 0.001). Treatment was escalated after 199 (56%) procedures. Receiver operating characteristic curve analysis of need for treatment escalation showed the highest sensitivity and specificity for UCEIS ≥4 (0.80 and 0.93, respectively; AUC, 0.93). Of 170 patients with a UCEIS ≥4, treatment was escalated in 159 (94%), but not for 11 (6%). Of 185 patients with a UCEIS ≤3, 40 (22%) were escalated, whereas 145 (78%) were not (P < 0.001). UCEIS correlated strongly with NHI (0.723; P < 0.001), moderately with SCCAI (0.671; P < 0.001), and weakly with CRP (0.279; P < 0.001). CONCLUSIONS: A UCEIS ≥4 was significantly associated with treatment escalation. This cutoff could therefore be used to support clinical decision-making based on endoscopic findings. Strong and moderate correlations were found between UCEIS and histological and clinical disease activity, respectively, whereas a weak correlation was found with CRP.10.1093/ibd/izy325_Video_1 izy325.video1 5849933952001.

Understanding Endoscopic Disease Activity in IBD: How to Incorporate It into Practice.

Endoscopic assessment of disease activity is an essential part of clinical practice in inflammatory bowel disease (IBD) and is used for diagnosis, prognosis, monitoring for dysplasia and increasingly for the evaluation of mucosal or endoscopic response to therapy. Recently, mucosal or endoscopic healing has emerged as a key goal of therapy as it has been found that patients who achieve endoscopic remission have improved outcomes compared to those who do not, and this may be independent of their clinical disease activity. However, there is currently no validated definition of mucosal healing and there are numerous endoscopic scoring systems proposed to define endoscopic activity and response to therapy in both ulcerative colitis and Crohn's disease. This article will discuss the most common endoscopic scores used to measure endoscopic disease activity in IBD, the pros and cons of each of these scoring systems and proposed definitions for endoscopic response or remission that exist for each. In addition, the role of endoscopy in prognosticating the disease course is discussed and how endoscopy can be utilized as part of a "treat-to-target" treatment strategy where endoscopy results direct decisions regarding medical strategies in clinical practice is highlighted.

Endoscopic scoring systems for inflammatory bowel disease: pros and cons.

Endoscopy plays a pivotal role for diagnosis and assessment of disease activity and extent in patients with inflammatory bowel diseases. International guidelines recommend the use of endoscopic scoring systems for evaluation of the prognosis and efficacy of medical treatments. Ideal scoring systems are easy to use, reproducible, reliable, responsive to changes, and validated in different clinical settings in order to guide therapeutic strategies. However, currently available endoscopic scoring systems often appear as complex for routine endoscopy and suffer from insufficient interobserver agreement and lack of formal validation which often limit their use in clinical trials. Here, we describe the role of endoscopic scoring systems in inflammatory bowel diseases focusing on pros and cons in the era of advanced endoscopic imaging and mucosal healing.