Yinyin Yuan
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (18)
Background Improved cancer risk stratification is needed to differentiate high-risk individuals with Barrett’s esophagus (BE) from low-risk populations to reduce overtreatment and improve outcome. The evolution of BE towards adenocarcinoma is likely driven by a combination of genomic and microenvironmental factors, yet existing predictive models rarely integrate both using routine specimens. Method We developed BEACON (Barrett Esophagus DNA content Abnormality and immune ecology for Cancer Outcome), a spatially aware framework predicting DNA content abnormalities and characterizing immune spatial ecology from routine histopathology. First, using 777 BE biopsies with flow cytometry-based DNA content data scanned at two institutions, we trained and tested DACOR (DNA content abnormality recognition), a multi-instance learning model that predicts DNA content abnormalities from histopathology. Next, complementary models for cell classification and tissue segmentation enabled spatial immune ecology metric computations. Lastly, a logistic regression model integrated molecular immune ecological features and epithelial morphology for cancer risk stratification. Results DACOR achieved 0.825 AUC in the test cohort for DNA content abnormality prediction. DNA content abnormal regions exhibited increased lymphoplasma cellular inflammation versus normal regions (p=0.006). Patients classified as DNA content abnormal by DACOR demonstrated increased cancer progression (p=0.0001). Among patients with DNA content abnormality, cancer progressors exhibited increased plasma cell clustering adjacent to abnormal epithelium compared to non-progressors. The integrated risk classification model stratified DNA content abnormal patients into high- and low-risk groups with 0.817 AUC. Conclusion BEACON spatially integrates molecular abnormality with immune spatial ecology to stratify BE patients by cancer progression risk using routine pathology images. This scalable, explainable approach could improve clinical decision-making and reduce unnecessary surveillance in low-risk patients.
Year:
2025
Beyond tertiary lymphoid structures, a significant number of immune-rich areas without germinal center-like structures are observed in non–small cell lung cancer. Here, we integrated transcriptomic data and digital pathology images to study the prognostic implications, spatial locations, and constitution of immune rich areas (immune hotspots) in a cohort of 935 patients with lung cancer from The Cancer Genome Atlas. A high intratumoral immune hotspot score, which measures the proportion of immune hotspots interfacing with tumor islands, was correlated with poor overall survival in lung squamous cell carcinoma but not in lung adenocarcinoma. Lung squamous cell carcinomas with high intratumoral immune hotspot scores were characterized by consistent upregulation of B-cell signatures. Spatial statistical analyses conducted on serial multiplex IHC slides further revealed that only 4.87% of peritumoral immune hotspots and 0.26% of intratumoral immune hotspots were tertiary lymphoid structures. Significantly lower densities of CD20 + CXCR5 + and CD79b + B cells and less diverse immune cell interactions were found in intratumoral immune hotspots compared with peritumoral immune hotspots. Furthermore, there was a negative correlation between the percentages of CD8 + T cells and T regulatory cells in intratumoral but not in peritumoral immune hotspots, with tertiary lymphoid structures excluded. These findings suggest that the intratumoral immune hotspots reflect an immunosuppressive niche compared with peritumoral immune hotspots, independent of the distribution of tertiary lymphoid structures. A balance toward increased intratumoral immune hotspots is indicative of a compromised antitumor immune response and poor outcome in lung squamous cell carcinoma. Significance: Intratumoral immune hotspots beyond tertiary lymphoid structures reflect an immunosuppressive microenvironment, different from peritumoral immune hotspots, warranting further study in the context of immunotherapies.
Year:
2025
Beyond tertiary lymphoid structures, a significant number of immune rich areas without germinal center-like structures are observed in non-small cell lung cancer. Here, we integrated transcriptomic data and digital pathology images to study the prognostic implications, spatial locations, and constitution of immune rich areas (immune hotspots) in a cohort of 935 lung cancer patients from the TCGA. A high intratumoral immune hotspot score, which measures the proportion of immune hotspots interfacing with tumor islands, was correlated with poor overall survival in lung squamous cell carcinoma but not in lung adenocarcinoma. Lung squamous cell carcinomas with high intratumoral immune hotspot scores were characterized by consistent upregulation of B cell signatures. Spatial statistical analyses conducted on serial multiplex immunohistochemistry slides further revealed that only 4.87% of peritumoral immune hotspots and 0.26% of intratumoral immune hotspots were tertiary lymphoid structures. Significantly lower densities of CD20+CXCR5+ and CD79b+ B cells and less diverse immune cell interactions were found in intratumoral immune hotspots compared to peritumoral immune hotspots. Furthermore, there was a negative correlation between the percentages of CD8+ T cells and T regulatory cells in intratumoral but not in peritumoral immune hotspots, with tertiary lymphoid structures excluded. These findings suggest that the intratumoral immune hotspots reflect an immunosuppressive niche compared to peritumoral immune hotspots, independent of the distribution of tertiary lymphoid structures. A balance towards increased intratumoral immune hotspots is indicative of a compromised anti-tumor immune response and poor outcome in lung squamous cell carcinoma.
Year:
2024
A and B, Spatial neighborhood of immune and tumor cells ( A and B ) and between MGUS and NDMM ( B ). The P * indicate P values after multiple testing correction using the BH method. The points represent the mean and the bars are 95% CIs, indicating uncertainty. C, Sample images showing an increased number of BLIMP1 + cells in the neighborhood with CD8 + on NDMM samples (NDMM example shown here is the same image as Fig. 5L ) compared with MGUS samples. D–I, Clustered or dispersed pattern of immune and tumor cells in BM trephine sample. Box plots showing the difference in NND and Z score between NDMM and posttreatment for CD8 + cells ( D ), BLIMP1 + cells ( E ), and FOXP3 − CD4 + cells ( F ). Box plots showing the difference in NND and Z score between NDMM and MGUS for CD8 + cells ( G ) and BLIMP1 + cells ( H ), and between male and female for BLIMP1 + cells ( I ). The unit of NND is μmol/L. The Z score shows the significance of the difference between the NND distribution for a given cell type from a complete spatial random distribution and the observed NND (Supplementary Materials and Methods).
Year:
2024
A and B, Spatial neighborhood of immune and tumor cells ( A and B ) and between MGUS and NDMM ( B ). The P * indicate P values after multiple testing correction using the BH method. The points represent the mean and the bars are 95% CIs, indicating uncertainty. C, Sample images showing an increased number of BLIMP1 + cells in the neighborhood with CD8 + on NDMM samples (NDMM example shown here is the same image as Fig. 5L ) compared with MGUS samples. D–I, Clustered or dispersed pattern of immune and tumor cells in BM trephine sample. Box plots showing the difference in NND and Z score between NDMM and posttreatment for CD8 + cells ( D ), BLIMP1 + cells ( E ), and FOXP3 − CD4 + cells ( F ). Box plots showing the difference in NND and Z score between NDMM and MGUS for CD8 + cells ( G ) and BLIMP1 + cells ( H ), and between male and female for BLIMP1 + cells ( I ). The unit of NND is μmol/L. The Z score shows the significance of the difference between the NND distribution for a given cell type from a complete spatial random distribution and the observed NND (Supplementary Materials and Methods).
Year:
2024
Studying bone physiology using MoSaicNet. A , Proportion of different compartments of BM trephine digital images. B–E, One stacked bar represents a sample. Box plots showing the difference in %bone between samples from NDMM and posttreatment ( B ), MGUS and NDMM ( C ), different age groups (median age = 58.0 years; D ), and gender groups ( E ). F, Scatter plot showing the number of bone superpixels in 17 clusters from MGUS, NDMM, and posttreatment samples. The size of the dots represents the percentage of superpixels. The color represents the number of slides in each cluster. G, Correlation of percentage of superpixels in each cluster between different patient groups. A point represents a cluster. H, Scatter plot of slide-level heterogeneity of bone features measured by features variance (Supplementary Materials and Methods). A point represents a patient/slide. I and J, Box plots showing differences in bone density heterogeneity between NDMM and posttreatment ( I ), and between MGUS and NDMM ( J ). K–N, Box plots showing the difference in bone thickness between samples from NDMM and posttreatment ( K ), MGUS and NDMM ( L ), and different age groups (median age = 58.0 years; M ) and gender ( N ).
Year:
2024
Performance evaluation of MoSaicNet and AwareNet deep learning models: A , The ROC curves and AUC values of the MoSaicNet superpixel classifier. The values in brackets indicate the 95% CI. B, Two-dimensional mapping of superpixels using MoSaicNet learned 200-dimensional features after dimensionality reduction by UMAP. C, The ROC curves and AUC values of single-cell classifier model on separately held test data. The values in brackets indicate the 95% CI. D, UMAP features visualization of deep learned features by AwareNet single-cell classifier CNN. E and F, Validation of AwareNet model using correlation of density of CD8 + ( E ) and CD4 + ( F ) cells in panel 1 and panel 2.
Year:
2024
Bone marrow trephine biopsy is crucial for the diagnosis of multiple myeloma. However, the complexity of bone marrow cellular, morphologic, and spatial architecture preserved in trephine samples hinders comprehensive evaluation. To dissect the diverse cellular communities and mosaic tissue habitats, we developed a superpixel-inspired deep learning method (MoSaicNet) that adapts to complex tissue architectures and a cell imbalance aware deep learning pipeline (AwareNet) to enable accurate detection and classification of rare cell types in multiplex immunohistochemistry images. MoSaicNet and AwareNet achieved an AUC of >0.98 for tissue and cellular classification on separate test datasets. Application of MoSaicNet and AwareNet enabled investigation of bone heterogeneity and thickness as well as spatial histology analysis of bone marrow trephine samples from monoclonal gammopathies of undetermined significance (MGUS) and from paired newly diagnosed and posttreatment multiple myeloma. The most significant difference between MGUS and newly diagnosed multiple myeloma (NDMM) samples was not related to cell density but to spatial heterogeneity, with reduced spatial proximity of BLIMP1 + tumor cells to CD8 + cells in MGUS compared with NDMM samples. Following treatment of patients with multiple myeloma, there was a reduction in the density of BLIMP1 + tumor cells, effector CD8 + T cells, and regulatory T cells, indicative of an altered immune microenvironment. Finally, bone heterogeneity decreased following treatment of patients with multiple myeloma. In summary, deep learning–based spatial mapping of bone marrow trephine biopsies can provide insights into the cellular topography of the myeloma marrow microenvironment and complement aspirate-based techniques. Significance: Spatial analysis of bone marrow trephine biopsies using histology, deep learning, and tailored algorithms reveals the bone marrow architectural heterogeneity and evolution during myeloma progression and treatment.
Year:
2024
Density of immune T cells and plasma cells in MGUS, NDMM, and posttreatment samples. A–G, Box plots showing the difference in density of FOXP3 + CD4 + ( A ), the density of CD8 + ( B ), the density of FOXP3 − CD4 + ( C ), FOXP3 + CD4 + :FOXP3 − CD4 + ratio ( D ), FOXP3 + CD4 + :CD8 + ratio ( E ), density of BLIMP1 + ( F ), and CD8 + :BLIMP1 + ratio ( G ) between paired NDMM samples and posttreatment samples ( n = 10 pairs). H–J, Box plot showing the difference in density of FOXP3 + CD4 + ( H ), the density of BLIMP1 + cells ( I ), and CD8 + :BLIMP1 + cells ( J ) between MGUS and NDMM samples ( n = 19). K and L, Sample images showing the reduction of the density of FOXP3 + CD4 + and CD8 + cells ( K ) and BLIMP1 + cells ( L ) at posttreatment compared with paired NDMM samples. The cell density is presented per 1 mm 2 tissue area.
Year:
2024
Density of immune T cells and plasma cells in MGUS, NDMM, and posttreatment samples. A–G, Box plots showing the difference in density of FOXP3 + CD4 + ( A ), the density of CD8 + ( B ), the density of FOXP3 − CD4 + ( C ), FOXP3 + CD4 + :FOXP3 − CD4 + ratio ( D ), FOXP3 + CD4 + :CD8 + ratio ( E ), density of BLIMP1 + ( F ), and CD8 + :BLIMP1 + ratio ( G ) between paired NDMM samples and posttreatment samples ( n = 10 pairs). H–J, Box plot showing the difference in density of FOXP3 + CD4 + ( H ), the density of BLIMP1 + cells ( I ), and CD8 + :BLIMP1 + cells ( J ) between MGUS and NDMM samples ( n = 19). K and L, Sample images showing the reduction of the density of FOXP3 + CD4 + and CD8 + cells ( K ) and BLIMP1 + cells ( L ) at posttreatment compared with paired NDMM samples. The cell density is presented per 1 mm 2 tissue area.
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Collaborators (16)
Simon P Castillo
Senior Research Fellow
University College London
Kwee Yong
University College London
Charles Swanton
Chief Clinician
-
Mariam Jamal-Hanjani
Group Leader and Thoracic Oncologist
University College London
Matthew Stachler
Associate Professor
University of California, San Francisco
Hanyun Zhang
Garvan Institute of Medical Research
Selvaraju Veeriah
University College London
Carlo Maley
Professor
Arizona State University
Katey Enfield
Assistant Professor
The University of British Columbia
Helen Byrne
Professor of Mathematical Biology
University of Oxford
Shan E Ahmed Raza
Associate Professor
University of Warwick
Manuel Rodriguez-Justo
Consultant & Professor Research Pathology
University College London
Khalid AbdulJabbar
Institute of Cancer Research
William Grady
Fred Hutchinson Cancer Research Center
John Le Quesne
Professor of Molecular Pathology
University of Glasgow
Erik Sahai
The Francis Crick Institute
Make your next move.
Prepare your application
Review drafts before you send them.