A novel panel of serum biomarkers was found to identify patients who will develop Crohn disease up to 5 years before diagnosis, according to a study published in Gastroenterology.
Serum samples were collected from patients before a diagnosis of Crohn disease (n= 200; mean age ± standard deviation [SD], 31.4±6.6 years) or ulcerative colitis (n=199; mean age ± SD, 28.9±5.2 years), as well as from individuals without inflammatory bowel disease (n=200; mean age ± SD, 28.5 4.8 years), between 1998 and 2013 as part of the US Defense Medical Surveillance System.
In each sample, the investigators measured the levels of antibodies against the yeast Saccharomyces cerevisiae (anti-S cerevisiae immunoglobulin A [IgA] or IgG) and the bacteria Escherichia coli (anti-E coli outer membrane porin C, anti-CBir1, anti-flagellin 2, anti-flagellin X, and perinuclear anti-neutrophil cytoplasmic antibodies) and those of 1129 other proteins.
They then used multivariate statistical analysis and modeling to predict patient disease status. Predictive performance at different points before diagnosis was evaluated using area under the receiver operating characteristic curves (AUCs). Biological pathways of differentially regulated proteins were identified following a comprehensive bioinformatics analysis and were used to validate the models.
Overall, 51 biomarkers were predictive of Crohn disease within the 5 years of diagnosis; the panel AUC was 0.76 and 0.87 within 5 years and 1 year of diagnosis, respectively.
Proteins included in the panel and predictive of the development of Crohn disease were involved in pathways involved with complement cascade, lysosomes, innate immune response, and glycosaminoglycan metabolism.
The biomarker panel used to predict a diagnosis of ulcerative colitis yielded low accuracy: AUC of 0.56 and 0.72 within 5 years and 1 year of diagnosis, respectively.
Limitations of the study highlighted by the authors included the use of only serum samples, imperfect demographic matching, and a lack of generalizability (mostly young, white, male, military personnel). Covariates known to affect disease risk, such as tobacco use, were unavailable.
“We believe our work represents an important foundation for future work in the preclinical phase of disease,” wrote the authors. “In the future, early identification of individuals at high-risk for disease development could allow for close monitoring and for the development of interventions to delay, attenuate or even halt disease initiation.”
Torres J, Petralia F, Sato T, et al. Serum biomarkers identify patients who will develop inflammatory bowel diseases up to 5 y before diagnosis [published online March 9, 2010]. Gastroenterology. doi:10.1053/j.gastro.2020.03.007