Role of intratumoural bacteria in resistance to LAG-3 targeting therapeutics.

e14648 Background: Intratumoural bacteria play a pivotal role in the tumour microenvironment (TME) and may influence clinical trial outcomes. Here, we focus on LAG-3, lymphocyte activation gene-3, a receptor highly expressed in exhausted T cells. LAG-3 is the 'third' immune checkpoint inhibitor to be developed. While it improved progression-free survival in advanced melanoma, there has been limited success in colorectal, gastric and oesophageal cancer. There is no clear explanation for this variability. First, we hypothesise that intratumoural bacteria can express proteins that interact with the LAG-3 receptor. Of note, previous work has demonstrated that LAG-3 interacts with bacterial proteins. Second, we hypothesise that the differential prevalence of LAG-3-interacting bacteria across tumour types may explain clinical trial outcomes, i.e., they would be less prevalent in melanoma than others. Methods: Analysis was performed using BioCorteX's knowledge graph and proprietary engines, v20250128_100953. Metagenomic analysis was performed on 470 primary tumour samples across five tumour types (melanoma, colorectal, gastric, oesophageal, non-small cell lung). Sequence homology, complemented by 3D structural analysis, was used to evaluate the mimicry of bacterial proteins to human proteins that bind LAG-3. The prevalence and relative abundance of LAG-3-interacting bacteria were compared across tumour types and related to study outcomes. Results: Several human proteins known to interact with LAG-3 were found to have bacterial homologs expressed by a wide range of intratumoural bacterial species. We find that the presence of LAG-3-interacting bacteria differed significantly across tumour types. LAG-3-interacting bacteria were least prevalent in melanoma (4% vs 24%, p < 0.001), mirroring the clinical trial success. Conclusions: This study provides the first evidence that LAG-3-interacting bacteria in the TME may relate to clinical trial outcomes. More specifically, the only successful phase III trial of LAG-3 therapeutics appears to be in the tumour type with the lowest prevalence of LAG-3-interacting bacteria, melanoma. Understanding the role of intratumoural bacteria within the TME could unlock a new dimension to increase the likelihood of success for patients and oncology trials.