P1344 A serum antibody profile against environmental and microbial antigens distinguishes patients with newly diagnosed inflammatory bowel disease. A Latin-American inception cohort

Abstract Background Inflammatory bowel disease (IBD) results from a complex interaction between environment, microbiota, and host immune response. PhIP-seq (Phage ImmunoPrecipitation sequencing) explores this interaction by detecting 244,000 host antibodies (Ab) against environmental and microbial antigens. We aimed to compare the antibody repertoire detected at IBD diagnosis against non-IBD controls and assess its discriminatory capacity in a geographical region experiencing a rising incidence of IBD. Methods Chilean adults undergoing ileocolonoscopy for symptoms suggestive of IBD were recruited, and blood samples were collected for PhIP-seq analysis. The diagnosis of IBD was based on established endoscopic and histological findings. Participants with normal endoscopy and histology were considered non-IBD controls. The proportion of antibodies in each group was compared using Fisher’s exact test and p-values were corrected for multiple testing using Benjamini-Hochberg method (q-value). Corrected and uncorrected p-value are shown. Furthermore, machine learning models were developed to predict IBD status based on the presence/absence of antibodies and/or clinical factors. Leave-one-out cross-validation was performed to assess the predictive power of classifiers trained with different subsets of data, for which receiver operating characteristic (ROC) curves were built. Area under the ROC curve (AUC) and the 95% confidence intervals (CI) for the predictions were calculated using bootstrapping. SHAP analysis was then used to identify the most informative data and their respective contributions to individual classifications. Results A total of 40 IBD patients and 40 controls were analysed. Clinical data is summarized in Table 1. In total, 13,106 antibodies were detected. Higher frequencies of antibodies against Cytomegalovirus and Limosilactobacillus reuteri were observed in IBD, while antibodies against Helicobacter pylori, Streptococcus pneumoniae, enteroviruses, and Aegilops tauschii predominated in controls (uncorrected p-value 0.05). A subset of 10 antibodies displaying the highest SHAP values showed a high ability to discriminate IBD patients, outperforming clinical factors and yielding a classifier with a mean AUC of 0.94 and an (95% CI 0.89, 0.99, Figure). Conclusion A distinct and abnormal humoral response to common antigens distinguish between Latin-American new-onset IBD patients and non-IBD controls. These antibody responses may be contributing to the emergence and worldwide spread of IBD across different geographical regions. Clinicaltrials.gov NCT06756100; FONDECYT #1230757. References: 1. Kaplan GG, Windsor JW. The four epidemiological stages in the global evolution of inflammatory bowel disease. Nat Rev Gastroenterol Hepatol. 2021;18(1):56-66. 2. Kotze PG, Underwood FE, Damião AOMC, et al. Progression of Inflammatory Bowel Diseases Throughout Latin America and the Caribbean: A Systematic Review. Clinical Gastroenterology and Hepatology. 2020;18(2):304-312 3. Vogl T, Klompus S, Leviatan S, et al. Population-wide diversity and stability of serum antibody epitope repertoires against human microbiota. Nat Med. 2021;27(8):1442-1450. Conflict of interest: Dr. Hernandez-Rocha, Cristian: No conflict of interest Riffo-Vicencio, Bárbara: No conflict of interest Vargas, Celeste: No conflict of interest Miranda-Contreras, Gabriel: No conflict of interest Reyna-Blanco, Carlos S: No conflict of interest Sepúlveda, Ignacia: No conflict of interest Kato, Sumie: No conflict of interest Pérez, Tamara: No Pavez, Carolina: No conflict of interest Chahuan, Javier: No conflict of interest Alvarez-Lobos, Manuel: No conflict of interest Vogl, Thomas: No conflict of interest Garrido, Daniel: No conflict of interest Saa, Pedro A: No conflict of interest