Development of a Simple, Serum Biomarker-based Model Predictive of the Need for Early Biologic Therapy in Crohn’s Disease

Con, Danny; Parthasarathy, Nina; Bishara, Maria; Luber, Raphael P.; Joshi, Neetima; Wan, Anna; Rickard, James A.; Long, Tony; Connoley, Declan; Sparrow, Miles; Langenberg, Daniel R. van; Vasudevan, Abhinav

doi:10.1093/ecco-jcc/jjaa194

Cited by 6 publications

(5 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In luminal gastroenterology, machine learning is gaining traction but its use has been relatively limited to automatic image recognition in endoscopy[ 8 - 11 ] as well as feature selection in genomic and microbiomics data[ 12 , 13 ]. Although there has been great interest in predicting clinical outcomes in CD such as response to therapeutics including biologics[ 14 - 18 ] and immunomodulators[ 19 , 20 ], studies investigating the utility of machine learning models for such predictive tasks have been more limited[ 21 - 23 ]. In particular, the utility of deep learning or ANNs specifically in clinical prediction of CD remains unknown[ 7 ].…”

Section: Introductionmentioning

confidence: 99%

Deep learning vs conventional learning algorithms for clinical prediction in Crohn's disease: A proof-of-concept study

Con¹,

Langenberg²,

Vasudevan³

2021

WJG

Self Cite

View full text Add to dashboard Cite

BACKGROUND Traditional methods of developing predictive models in inflammatory bowel diseases (IBD) rely on using statistical regression approaches to deriving clinical scores such as the Crohn's disease (CD) activity index. However, traditional approaches are unable to take advantage of more complex data structures such as repeated measurements. Deep learning methods have the potential ability to automatically find and learn complex, hidden relationships between predictive markers and outcomes, but their application to clinical prediction in CD and IBD has not been explored previously. AIM To determine and compare the utility of deep learning with conventional algorithms in predicting response to anti-tumor necrosis factor (anti-TNF) therapy in CD. METHODS This was a retrospective single-center cohort study of all CD patients who commenced anti-TNF therapy (either adalimumab or infliximab) from January 1, 2010 to December 31, 2015. Remission was defined as a C-reactive protein (CRP) < 5 mg/L at 12 mo after anti-TNF commencement. Three supervised learning algorithms were compared: (1) A conventional statistical learning algorithm using multivariable logistic regression on baseline data only; (2) A deep learning algorithm using a feed-forward artificial neural network on baseline data only; and (3) A deep learning algorithm using a recurrent neural network on repeated data. Predictive performance was assessed using area under the receiver operator characteristic curve (AUC) after 10× repeated 5-fold cross-validation. RESULTS A total of 146 patients were included (median age 36 years, 48% male). Concomitant therapy at anti-TNF commencement included thiopurines (68%), methotrexate (18%), corticosteroids (44%) and aminosalicylates (33%). After 12 mo, 64% had CRP < 5 mg/L. The conventional learning algorithm selected the following baseline variables for the predictive model: Complex disease behavior, albumin, monocytes, lymphocytes, mean corpuscular hemoglobin concentration and gamma-glutamyl transferase, and had a cross-validated AUC of 0.659, 95% confidence interval (CI): 0.562-0.756. A feed-forward artificial neural network using only baseline data demonstrated an AUC of 0.710 (95%CI: 0.622-0.799; P = 0.25 vs conventional). A recurrent neural network using repeated biomarker measurements demonstrated significantly higher AUC compared to the conventional algorithm (0.754, 95%CI: 0.674-0.834; P = 0.036). CONCLUSION Deep learning methods are feasible and have the potential for stronger predictive performance compared to conventional model building methods when applied to predicting remission after anti-TNF therapy in CD.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep learning vs conventional learning algorithms for clinical prediction in Crohn's disease: A proof-of-concept study

Con¹,

Langenberg²,

Vasudevan³

2021

WJG

Self Cite

View full text Add to dashboard Cite

show abstract

“…Biomarkers have the potential to predict outcomes and tailor therapeutic options to the individual, which are core components of modern IBD management algorithms and precision medicine. 17,26 Our study has highlighted the difference in biomarker dynamics between patients who required and avoided colectomy and demonstrated the potential utility of CAR and CLR early in the treatment course (day 3 post salvage) in predicting infliximab failure. With the expansion of efficacious treatments such as newer biologic therapies and small molecule inhibitors, [27][28][29][30] the ability to predict disease trajectory in ASUC should inform treatment choice-for example, a patient who is likely to require longer term colectomy despite infliximab salvage may warrant more proactive optimization of maintenance therapy with a biologic agent (a top-down approach), while patients with a lower risk of colectomy may only require conventional therapy.…”

Section: Discussionmentioning

confidence: 81%

Section: Discussionmentioning

confidence: 99%

Biomarker dynamics during infliximab salvage for acute severe ulcerative colitis: C-reactive protein (CRP)-lymphocyte ratio and CRP-albumin ratio are useful in predicting colectomy

et al. 2022

Self Cite

View full text Add to dashboard Cite

cerative colitis (ASUC) is a severe exacerbation of UC that warrants intravenous corticosteroids and may require emergency colectomy. 3 Corticosteroid failure necessitates salvage therapy with infliximab or ciclosporin which have demonstrated good short term efficacy. [4][5][6][7] However longer term prognosis after medical salvage remains guarded, where up to 50% of patients still proceed to colectomy in the absence of maintenance biologic therapy. [8][9][10] Conversely, a proportion of patients may have an initial period of significant symptoms, but develop only mild symptoms or even remission in the subsequent 10 years of

show abstract

“… 28 , 48 The improved AUC was 0.919 at week 28 and 0.887 at week 52, which exhibited an even stronger predictive ability than previous studies. 49 , 50 …”

Section: Discussionmentioning

confidence: 99%

Combined Use of Fecal Biomarkers in Inflammatory Bowel Diseases: Oncostatin M and Calprotectin

Cao¹,

Dai²,

Zhang³

et al. 2021

JIR

View full text Add to dashboard Cite

Background Fecal biomarkers have emerged as one of the most useful tools for clinical management of inflammatory bowel disease (IBD). Oncostatin M (OSM), like fecal calprotectin (FC), is highly expressed in the inflamed intestinal mucosa which may have potential usefulness. We aimed to evaluate the additional utility of these two fecal biomarkers for IBD diagnosis, activity, and prediction of infliximab response over FC alone. Methods In group 1, 236 IBD patients (145 Crohn’s disease, 91 ulcerative colitis), 50 disease controls, and 32 healthy controls were recruited for IBD diagnosis and activity. In group 2, baseline stool samples were collected from 62 patients to predict infliximab response at week 28 and 52. The performance of fecal biomarkers for IBD management was assessed by the area under the receiver operating characteristic curve (AUC). Results Fecal OSM and FC levels were increased in IBD patients and were positively correlated with clinical and endoscopic activity. Their combination showed a better ability for disease diagnosis (AUC = 0.93) and slightly improved the capability to identify mucosal healing (AUC = 0.923). Baseline OSM and FC levels were elevated in non-responders at week 28 and 52. The AUCs of OSM, FC, and their combination to predict therapeutic response were 0.763, 0.834, and 0.859 at week 28, 0.638, 0.661, and 0.704 at week 52, respectively. Combined use of fecal and blood biomarkers improved predictive accuracy with an AUC of 0.919 at week 28 and 0.887 at week 52. Conclusion In addition to FC, OSM is a novel fecal biomarker, and their combination is more beneficial for disease diagnosis and prediction of infliximab response but not for disease activity in IBD patients. Further larger-scale studies are required to confirm our findings.

show abstract

Development of a Simple, Serum Biomarker-based Model Predictive of the Need for Early Biologic Therapy in Crohn’s Disease

Cited by 6 publications

References 53 publications

Deep learning vs conventional learning algorithms for clinical prediction in Crohn's disease: A proof-of-concept study

Deep learning vs conventional learning algorithms for clinical prediction in Crohn's disease: A proof-of-concept study

Biomarker dynamics during infliximab salvage for acute severe ulcerative colitis: C-reactive protein (CRP)-lymphocyte ratio and CRP-albumin ratio are useful in predicting colectomy

Combined Use of Fecal Biomarkers in Inflammatory Bowel Diseases: Oncostatin M and Calprotectin

Contact Info

Product

Resources

About