Distinct clinical phenotypes for Crohn’s disease derived from patient surveys

Abstract

Background Defining clinical phenotypes provides opportunities for new diagnostics and may provide insights into early intervention and disease prevention. There is increasing evidence that patient-derived health data may contain information that complements traditional methods of clinical phenotyping. The utility of these data for defining meaningful phenotypic groups is of great interest because social media and online resources make it possible to query large cohorts of patients with health conditions. Methods We evaluated the degree to which patient-reported categorical data is useful for discovering subclinical phenotypes and evaluated its utility for discovering new measures of disease severity, treatment response and genetic architecture. Specifically, we examined the responses of 1961 patients with inflammatory bowel disease to questionnaires in search of sub-phenotypes. We applied machine learning methods to identify novel subtypes of Crohn’s disease and studied their associations with drug responses. Results Using the patients’ self-reported information, we identified two subpopulations of Crohn’s disease; these subpopulations differ in disease severity, associations with smoking, and genetic transmission patterns. We also identified distinct features of drug response for the two Crohn’s disease subtypes. These subtypes show a trend towards differential genotype signatures. Conclusion Our findings suggest that patient-defined data can have unplanned utility for defining disease subtypes and may be useful for guiding treatment approaches.

The Metabolic Role of MAP3K15: Genetic and Phenotypic Insights from the 23andMe Research Database and Genetics-Driven Recruitment

“…23andMe has conducted research studies in the past, where study participants were recruited on the basis of a condition they were already aware of or known genetic risk status disclosed to them in a 23andMe product report 2,33,34,35 . To further inform our understanding of the effect of MAP3K15 LoF on metabolic health, we established a protocol to recruit participants based on any type of genetics unknown to them, including the genotype(s) of a single variant or multiple variants, polygenic risk score, and others.…”

A Systematic Review of Artificial Intelligence and Machine Learning Applications to Inflammatory Bowel Disease, with Practical Guidelines for Interpretation

“…Of 78 studies included in the systematic review, the majority used supervised ML, with 4 articles employing unsupervised methods, 11–14 and 5 utilizing both supervised and unsupervised ML 15–19 for varied clinical applications. Many articles trialed different ML methods before selecting the optimal one, and some researchers implemented ML for multiple IBD applications.…”

“… 23 , 29 , 45 , 59 , 61 , 66 , 73 Crohn’s disease data (only) was used in 27 studies, 12 , 17 , 19 , 26–29 , 32 , 37–39 , 41 , 44 , 47–50 , 58 , 60 , 63 , 65 , 74–79 and UC data (only) was used in 15 studies, 25 , 40 , 42 , 46 , 52 , 55 , 59 , 61 , 62 , 64 , 68–70 , 80 , 81 with the remainder ( n = 36) using a mix of CD and UC data, or IBD data as one class. 11 , 13–16 , 18 , 20–24 , 30 , 31 , 33–36 , 43 , 45 , 51 , 53 , 54 , 56 , 57 , 66 , 67 , 71–73 , 82–88 Half of the research using UC-only data focused on predicting disease activity with endoscopy data, whereas the aims of ML classifications on CD data were varied. A breakdown of the method and classification task can be found in Figure 3 , which can be customized here (isstafford.github.io/review_ml_ibd_2021/).…”

Inflammatory Bowel Disease Treatments and Predictive Biomarkers of Therapeutic Response

“…For example, up to 30% of patients do not respond to initial therapy and even among initial responders, 13-46% lose response over time with estimates varying by treatment and disease subtypes [9], a percentage that can sometimes reach as high as 64% after treatment [10]. Therefore, a periodic assessment of IBD activity and disease severity is required to assess disease phenotype, including disease extent and severity in UC, as well as disease extent and disease behavior in CD, to provide a tailored therapy algorithm to every patient [5,[11][12][13].…”