Development and Validation of Machine Learning Models in Prediction of Remission in Patients With Moderate to Severe Crohn Disease.

Importance: Biological therapies have revolutionized inflammatory bowel disease management, but many patients do not respond to biological monotherapy. Identification of likely responders could reduce costs and delays in remission. Objective: To identify patients with Crohn disease likely to be durable responders to ustekinumab before committing to long-term treatment. Design, Setting, and Participants: This cohort study analyzed data from 3 phase 3 randomized clinical trials (UNITI-1, UNITI-2, and IM-UNITI) conducted from 2011 to 2015. Participants (n = 401) were individuals with active (C-reactive protein [CRP] measurement of ≥5 mg/L at enrollment) Crohn disease who received ustekinumab therapy. Data analysis was performed from November 1, 2017, to June 1, 2018. Exposures: All included patients were exposed to 1 or more dose of ustekinumab for 8 weeks or more. Main Outcomes and Measures: Random forest methods were used in building 2 models for predicting Crohn disease remission, with a CRP level lower than 5 mg/dL as a proxy for biological remission, beyond week 42 of ustekinumab treatment. The first model used only baseline data, and the second used data through week 8. Results: In total, 401 participants, with a mean (SD) age of 36.3 (12.6) years and 170 male (42.4%), were included. The week-8 model had a mean area under the receiver operating characteristic curve (AUROC) of 0.78 (95% CI, 0.69-0.87). In the testing data set, 27 of 55 participants (49.1%) classified as likely to have treatment success achieved success with a CRP level lower than 5 mg/L after week 42, and 7 of 65 participants (10.8%) classified as likely to have treatment failure achieved this outcome. In the full cohort, 87 patients (21.7%) attained remission after week 42. A prediction model using the week-6 albumin to CRP ratio had an AUROC of 0.76 (95% CI, 0.71-0.82). Baseline ustekinumab serum levels did not improve the model's prediction performance. Conclusions and Relevance: In patients with active Crohn disease, demographic and laboratory data before week 8 of treatment appeared to allow the prompt identification of likely nonresponders to ustekinumab without the need for costly drug-level monitoring.

Predicting Corticosteroid-Free Biologic Remission with Vedolizumab in Crohn's Disease.

Background and Aims: Vedolizumab (VDZ) is effective for Crohn's disease (CD) but costly and is slow to produce remission. Early knowledge of whether vedolizumab is likely to succeed is valuable for physicians, patients, and insurers. Methods: Phase 3 clinical trial data on VZD for CD were used to predict outcomes. Random forest modeling on the training cohort was used to predict the outcome of corticosteroid-free biologic remission at week 52 on the testing cohort. Models were constructed using baseline data, or data through week 6 of VDZ therapy. Results: The clinical trial included 594 subjects who received VDZ with baseline active inflammation [elevated C-reactive protein (>5 mg/L)]. Subjects with missing predictor variables (N = 120) or missing outcome data (N = 2) were excluded to produce a modeling dataset of 472 subjects. The Area Under the Receiver Operating Characteristic curve (AuROC) for corticosteroid-free biologic remission at week 52 using baseline data was only 0.65 (95% CI: 0.53 - 0.77), but was 0.75 (95% CI: 0.64 - 0.86) with data through week 6 of VDZ . Patients predicted to be in corticosteroid-free biologic remission at week 52 by the model achieved this endpoint 35.8% of the time, whereas patients predicted to fail only succeeded 6.7% of the time. Conclusions: An algorithm using laboratory data through week 6 of VDZ therapy was able to identify which CD patients with baseline inflammation would achieve corticosteroid-free biologic remission on VDZ at week 52. A majority of patients can be identified by week 6 as very unlikely to achieve remission.

Machine Learning Algorithms for Objective Remission and Clinical Outcomes with Thiopurines.

BACKGROUND AND AIMS: Big data analytics leverage patterns in data to harvest valuable information, but are rarely implemented in clinical care. Optimising thiopurine therapy for inflammatory bowel disease [IBD] has proved difficult. Current methods using 6-thioguanine nucleotide [6-TGN] metabolites have failed in randomized controlled trials [RCTs], and have not been used to predict objective remission [OR]. Our aims were to: 1) develop machine learning algorithms [MLA] using laboratory values and age to identify patients in objective remission on thiopurines; and 2) determine whether achieving algorithm-predicted objective remission resulted in fewer clinical events per year. METHODS: Objective remission was defined as the absence of objective evidence of intestinal inflammation. MLAs were developed to predict three outcomes: objective remission, non-adherence, and preferential shunting to 6-methylmercaptopurine [6-MMP]. The performance of the algorithms was evaluated using the area under the receiver operating characteristic curve [AuROC]. Clinical event rates of new steroid prescriptions, hospitalisations, and abdominal surgeries were measured. RESULTS: Retrospective review was performed on medical records of 1080 IBD patients on thiopurines. The AuROC for algorithm-predicted remission in the validation set was 0.79 vs 0.49 for 6-TGN. The mean number of clinical events per year in patients with sustained algorithm-predicted remission [APR] was 1.08 vs 3.95 in those that did not have sustained APR [p < 1 x 10-5]. Reductions in the individual endpoints of steroid prescriptions/year [-1.63, p < 1 x 10-5], hospitalisations/year [-1.05, p < 1 x 10-5], and surgeries/year [-0.19, p = 0.065] were seen with algorithm-predicted remission. CONCLUSIONS: A machine learning algorithm was able to identify IBD patients on thiopurines with algorithm-predicted objective remission, a state associated with significant clinical benefits, including decreased steroid prescriptions, hospitalisations, and surgeries.

Matrix stiffness corresponding to strictured bowel induces a fibrogenic response in human colonic fibroblasts.

BACKGROUND: Crohn's disease is characterized by repeated cycles of inflammation and mucosal healing which ultimately progress to intestinal fibrosis. This inexorable progression toward fibrosis suggests that fibrosis becomes inflammation-independent and auto-propagative. We hypothesized that matrix stiffness regulates this auto-propagation of intestinal fibrosis. METHODS: The stiffness of fresh ex vivo samples from normal human small intestine, Crohn's disease strictures, and the unaffected margin were measured with a microelastometer. Normal human colonic fibroblasts were cultured on physiologically normal or pathologically stiff matrices corresponding to the physiological stiffness of normal or fibrotic bowel. Cellular response was assayed for changes in cell morphology, α-smooth muscle actin staining, and gene expression. RESULTS: Microelastometer measurements revealed a significant increase in colonic tissue stiffness between normal human colon and Crohn's strictures and between the stricture and adjacent tissue margin. In Ccd-18co cells grown on stiff matrices corresponding to Crohn's strictures, cellular proliferation increased. Pathologic stiffness induced a marked change in cell morphology and increased α-smooth muscle actin protein expression. Growth on a stiff matrix induced fibrogenic gene expression, decreased matrix metalloproteinase, and proinflammatory gene expression and was associated with nuclear localization of the transcriptional cofactor MRTF-A. CONCLUSIONS: Matrix stiffness, representative of the pathologic stiffness of Crohn's strictures, activates human colonic fibroblasts to a fibrogenic phenotype. Matrix stiffness affects multiple pathways, suggesting that the mechanical properties of the cellular environment are critical to fibroblast function and may contribute to auto-propagation of intestinal fibrosis in the absence of inflammation, thereby contributing to the intractable intestinal fibrosis characteristic of Crohn's disease.

CARD-024, a vitamin D analog, attenuates the pro-fibrotic response to substrate stiffness in colonic myofibroblasts.

Intestinal fibrosis is one of the major complications of Crohn's disease (CD) for which there are no effective pharmacological therapies. Vitamin D deficiency is common in CD, though it is not known whether this is a contributing factor to fibrosis, or simply a consequence of the disease itself. In CD, fibrosis is mediated mainly by activated intestinal myofibroblasts during remodeling of extracellular matrix in response to wound healing. We investigated the effects of CARD-024 (1-alpha-hydroxyvitamin D5), a vitamin D analog with minimal hypercalcemic effects, on the pro-fibrotic response of intestinal myofibroblasts to two fibrogenic stimuli: TGFß stimulation and culture on a physiologically stiff matrix. TGFß stimulated a fibrogenic phenotype in Ccd-18co colonic myofibroblasts, characterized by an increase in actin stress fibers and mature focal adhesions, and increased αSMA protein expression, while CARD-024 repressed αSMA protein expression in a dose-dependent manner. Culture of colonic myofibroblasts on physiological high stiffness substrates induced morphological changes with increased actin stress fibers and focal adhesion staining, induction of αSMA protein expression, FAK phosphorylation, induction of fibrogenic genes, and repression of COX-2 and IL-1ß. CARD-024 treatment repressed the stiffness-induced morphological features including stellate cell morphology and the maturation of focal adhesions. CARD-024 repressed the stiffness-mediated induction of αSMA protein expression, FAK phosphorylation, and MLCK and ET-1 gene expression. In addition, CARD-024 partially stimulated members of the COX-2/IL-1ß inflammatory pathway. In summary, CARD-024 attenuated the pro-fibrotic response of colonic myofibroblasts to high matrix stiffness, suggesting that vitamin D analogs such as CARD-024 may ameliorate intestinal fibrosis.

Intestinal fibrosis is reduced by early elimination of inflammation in a mouse model of IBD: impact of a "Top-Down" approach to intestinal fibrosis in mice.

BACKGROUND: The natural history of Crohn's disease follows a path of progression from an inflammatory to a fibrostenosing disease, with most patients requiring surgical resection of fibrotic strictures. Potent antiinflammatory therapies reduce inflammation but do not appear to alter the natural history of intestinal fibrosis. The aim of this study was to determine the relationship between intestinal inflammation and fibrogenesis and the impact of a very early "top-down" interventional approach on fibrosis in vivo. METHODS: In this study we removed the inflammatory stimulus from the Salmonella typhimurium mouse model of intestinal fibrosis by eradicating the S. typhimurium infection with levofloxacin at sequential timepoints during the infection. We evaluated the effect of this elimination of the inflammatory stimulus on the natural history of inflammation and fibrosis as determined by gross pathology, histopathology, mRNA expression, and protein expression. RESULTS: Fibrogenesis is preceded by inflammation. Delayed eradication of the inflammatory stimulus by antibiotic treatment represses inflammation without preventing fibrosis. Early intervention significantly ameliorates but does not completely prevent subsequent fibrosis. CONCLUSIONS: This study demonstrates that intestinal fibrosis develops despite removal of an inflammatory stimulus and elimination of inflammation. Early intervention ameliorates but does not abolish subsequent fibrosis, suggesting that fibrosis, once initiated, is self-propagating, suggesting that a very early top-down interventional approach may have the most impact on fibrostenosing disease.

Prior Helicobacter pylori infection ameliorates Salmonella typhimurium-induced colitis: mucosal crosstalk between stomach and distal intestine.

BACKGROUND: Helicobacter pylori infection is associated with a lower risk of chronic autoimmune diseases including inflammatory bowel disease (IBD). H.pylori modulates the gastric immune response, decreasing the local inflammatory response to itself. In mice, chronic Salmonellatyphimurium infection induces colitis similar to Crohn's disease, characterized by inflammation, which progresses toward fibrosis. The aim of this study was to determine whether prior H. pylori infection acts at a distance to modulate the immune response of S.typhimurium-induced colitis. METHODS: Mice were infected with the mouse-adapted strain of H. pylori (SS1), followed by infection with S.typhimurium. The effect of H. pylori on colitis was determined by gross pathology, histopathology, cytokine response, and development of fibrosis in the cecum. Gastritis and systemic immune response was measured in response to infection. RESULTS: H.pylori suppresses the Th17 response to S.typhimurium infection in the mouse cecum, but does not alter the Th2 or T-regulatory response or the development of fibrosis. H. pylori infection induces IL-10 in the mesenteric lymph nodes, suggesting an extragastric mechanism for immunomodulation. H. pylori / S.typhimurium coinfection decreases inflammation in both the cecum and the stomach. CONCLUSIONS: This study demonstrates a potential mechanism for the negative association between H. pylori and IBD in humans. H. pylori represses the lower gastrointestinal tract Th17 response to bacterially induced colitis via extragastric immunomodulatory effects, illustrating immunological crosstalk between the upper and lower gastrointestinal tract.

Transient or persistent norovirus infection does not alter the pathology of Salmonella typhimurium induced intestinal inflammation and fibrosis in mice.

Murine noroviruses (MNV) are currently the most prevalent viruses infecting mouse research colonies. Concurrent infection of research mice with these viruses can dramatically alter the experimental outcome in some research models, but not others. In this report, we investigated the effect of MNV1 and MNV4 on a murine model of intestinal inflammation and fibrosis induced by Salmonella typhimurium infection in C57BL/6 mice. Subsequent co-infection of these mice with MNV1 or MNV4 did not lead to major changes in histopathology, the inflammatory response, or the fibrotic response. Thus, MNV does not substantially alter all gastrointestinal research models, highlighting the importance of investigating potential alterations in the research outcome by MNV on an individual basis. We hypothesize that this is particularly important in cases of research models that use immunocompromised mice, which could be more sensitive to MNV infection-induced changes.