Symposium 2

Dr Joe Yeong Poh Sheng

Group Leader, Institute of Molecular and Cell Biology, A*STAR

Department of Anatomical Pathology, Singapore General Hospital

Dr. Joe Yeong's main research focus is to understand and overcome the resistance of immunecheckpoint blockade immunotherapy. As an immuno-pathologist, his key vision is to bridge between immunologists and pathologists to better harness the advances of immunotherapy and further beyond. He is the pioneer in automation of quantitative multiplex immunohistochemistry, using clinical autostainers to study and quantitate tumour immune microenvironment in clinical samples, and has published > 100 papers in this field. His works on cancer immunology are included in multiple National Medical Research Council funded studies as well as pharmaceutical industry sponsored projects (>14 million dollars since 2017). He also served as a committee member in the World Immunotherapy Council, Society for Immunotherapy of Cancer (SITC) and is one of the organizers for its 2019 & 2023 WIC Global Symposium as well as multiplex IF expert consensus meeting 2022. He also serves as Program Chair of one of the largest AI medical Imaging conferences, CLINICCAI-MICCAI. He is also having editorial role of Elsevier (Immunoinformatics), SLAS Technology (Journal), Frontiers, World Scientific (Chief Editor) and Pathogens. He serves as a Secretary (Executive) in Singapore Society of Oncology – Cancer Immunotherapy Consortium, Co-lead in Education/Diagnostic of Singhealth Duke-NUS Cell Therapy Centre as well as Advisor (Spatial Technology), Cancer Discovery Hub, National Cancer Centre. In 2023, he co-founded World Immunotherapy Council Asia for promoting tumor immunology and advancing cancer immunotherapy education, information and research across Asia.

Session:

Bespoke Healthcare (Efforts to Tailor the Healthcare Journey for Patients)
12 April 2024, 1045 - 1215hrs, L1-S3

Presenting Title:

Immunopathological Signature: Integrating AI, Advancing Cancer Screening, and Refining Precision Medicine

Immunological profiling and surveillance are essential components of cancer research. Our lab focuses on its application in artificial intelligence (AI), cancer screening, and precision medicine. In AI, we advanced our capabilities by integrating morphology, transcriptome and proteomics to perform single-cell phenotyping and spatial analysis of tumor-immune interactions. Leveraging of deep learning to predict immune marker such as CD3 and CD8 positivity in H&E-stained tissue, with an automated workflow that enhances ground truth mapping precision at tissue and cellular levels. We encapsulated these deep learning models as H&E 2.0, a concept for next generation of digital pathology. A publicly accessible H&E 2.0 image database for interactive visualisation of various markers has been developed, along with iCellSight, which offers code-free interactive visualization and analysis for in-situ high-plex cellular protein data.

For cancer screening, we found a correlation between mutations in tumor suppressor genes with increased cancer risk in a nested case-control study of healthy elderly Asian women that have 17 years of follow-up data. This association was observed above a particular threshold of VAF in the peripheral blood. Moreover, from a CAR-T therapy trial, we devised the DETECT method as an unbiased proteomics approach for low molecular weight protein enrichment and identified different peptides fingerprint between responders and non-responders.

In precision medicine, our multi-step pipeline for longitudinal blood-based immune-monitoring in an HCC immunotherapy trial identified several cytokines as potential predictors for treatment response and 3-year overall survival. Furthermore, our automated mRNA mapping on whole tissue slides outperformed manual method in efficiency, consistency, reliability, and reproducibility. The universal multi-omics pipeline which integrates spatial transcriptomics, proteomics and diagnostic H&E assays on a single tissue section achieved single-slide multi-omics data with high-resolution tissue morphology images.

Return to Speakers >