A NEXT-GENERATION ARTIFICIAL INTELLIGENCE-DESIGNED ANTIBODY AGAINST TUMOR NECROSIS FACTOR-LIKE LIGAND 1A FOR INFLAMMATORY BOWEL DISEASE: CURRENT STATUS AND FUTURE PERSPECTIVES

Abstract

Inflammatory bowel diseases (IBD), which includes Crohn’s diseases and ulcerative colitis, it is a
chronic condition whose burden keeps growing worldwide, especially in Asia. Many patients still
respond poorly to current therapies, and none of the available drugs can reliably reverse intestinal
fibrosis. The TL1A-DR3 pathway has drawn attention because it drives both mucosal inflammation and
fibrotic tissue remodelling. Phase 2 trial of first-generation anti-TL1A antibodies, such as tulisokibart
and Afimkibart, have reported meaningful remission and endoscopic improvement. Their intravenous
dosing and potential immunogenicity, however, remain concerns. This review looks at the
epidemiology and socioeconomic burden of IBD, the immunological and genetic basis of the disease,
and the molecular biology of TL1A and its therapeutic blockade. We then focus on ABS-101, a fully
human anti-TL1A antibody created with Absci's generative AI platform and taken from target
identification to a clinical candidate in about 14 months. Preclinical work, including a 13-week primate
toxicology study, suggests high potency, low immunogenicity, a long half-life and subcutaneous
bioavailability. A first-in-human Phase 1 trial began in 2025. Although clinical efficacy in patients is
yet to be shown, ABS-101 illustrates how AI-guided design could speed up and streamline biologic
drug development in IBD.