SITC Annual Meeting

Background
Protein-protein interactions (PPIs) act as a crucial communication system between cells and their local surrounding environment to maintain cellular homeostasis. The dysregulation of this process is involved in affecting the course of several pathogenesis, including cancer. Notably, the interaction of Programmed cell Death Ligand 1 (PD-L1) on tumor cells binding to the Programmed cell Death Protein 1 (PD-1) receptor on immune cells is a key mechanism employed by cancer cells to evade the immune response. Regardless of the success of PD-1/PD-L1 checkpoint inhibitor immunotherapies, a correct patient stratification for these therapies has been challenging. The spatial investigation of PD-1/PD-L1 and other PPIs in the multiomics data framework holds the potential to deeply reveal the interplay of cellular components for a better understanding and forecasting of the clinical outcomes. 

Methods
We developed a fully automated multiplex workflow on the COMET™ platform capable of staining and imaging protein-protein interactions, mRNA and protein biomarkers on the same FFPE tissue section, at subcellular resolution. RNAscope™ HiPlex Pro and sequential immunofluorescence (seqIF™), for RNA and protein marker detection, respectively, were combined with a novel assay exploiting oligonucleotide-conjugated antibodies and the specificity and sensitivity of RNAscope™ technology for efficient protein proximity detection.  

Results
We showed, as a proof-of-concept for the assay, the co-detection of PD-1 and PD-L1 interaction, combined with RNA targets and protein biomarkers in their spatial context in FFPE human tissues. The PPI signal was validated by overlapping the corresponding sequential immunofluorescence staining from the single antibodies for PD-1 and PD-L1 on the same section. Additionally, the cell phenotyping of the surrounding region was investigated by a panel of RNA targets and protein biomarkers for multiple immune and stromal cells.   

Conclusions 
The integration of PPI assays on the multiomics workflow of the COMET™ platform is a powerful strategy to investigate protein interactions directly in their spatial settings. The flexibility of the technique in the choice of the markers under investigation opens a large potential for new biomarker discovery and validation of new clinically relevant interactions, for improving patient stratification and development of novel immunotherapies.

Background
The tumor microenvironment (TME) is a complex and dynamic ecosystem playing a crucial role in tumor development, progression, immune evasion, and therapeutic response. Integrating protein expression with complementary outputs, such as mapping cytokine and chemokine-expressing cells, is crucial to unveiling functional cell states and mechanisms of cancer progression [1,2]. However, accurately visualizing secreted molecules while preserving spatial context still remains a significant challenge in spatial biology. In this study, we employed an automated hyperplex multiomics approach to simultaneously detect protein and RNA expression, providing an in-depth cellular profiling of the TME across a variety of cancer types.  

Methods
We analyzed a formalin-fixed paraffin-embedded Tissue Microarray (TMA) comprising specimens of multiple human cancer types, namely prostate cancer, lung cancer, breast cancer, colorectal cancer, melanoma, and lymphoma. A TMA section was stained and imaged on the COMET™ automated platform, integrating RNAscope™ [3] HiPlex Pro and sequential immunofluorescence (seqIF™) [4]. This multiomics approach enabled concomitant detection of up to 12 RNA and over 50 protein targets on the same tissue section. The resulting image was analyzed using the HORIZON™ software to reveal in situ single-cell level features and spatially map their distribution.  

Results
The high-plex proteomic panel enabled detailed characterization of the TME composition, including tumor cells, a large variety of immune cell subsets, cancer-associated fibroblasts, vasculature, and markers of cell state and activation, including cell division and immune checkpoints. The combined detection of transcripts encoding secreted molecules, such as cytokines and chemokines, provided insight into local immune activity and cell-cell communication within the tumor and surrounding stromal compartments. 

Conclusions
This automated multiomics workflow enables comprehensive analysis of the TME across multiple cancer types, while significantly reducing the experimental turnover and sample consumption. Multiomics spatial profiling at single-cell resolution opens new opportunities to explore the cellular interactions at the tumor-immune interface and identify functional states relevant to immunotherapy and disease progression.

References

1. Ferri-Borgogno S et al. Molecular, Metabolic, and Subcellular Mapping of the Tumor Immune Microenvironment via 3D Targeted and Non-Targeted Multiplex Multi-Omics Analyses. Cancers (Basel). 2024 Feb 20;16(5):846. doi: 10.3390/cancers16050846. 
2. Yi M et al. Targeting cytokine and chemokine signaling pathways for cancer therapy. Signal Transduct Target Ther. 2024 Jul 22;9(1):176. doi: 10.1038/s41392-024-01868-3. 
3. Wang F et al. RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn. 2012 Jan;14(1):22-9. doi: 10.1016/j.jmoldx.2011.08.002.  
4. Rivest F et al. Fully automated sequential immunofluorescence (seqIF) for hyperplex spatial proteomics. Sci Rep. 2023. 13(1):16994. doi: 10.1038/s41598-023-43435-w. 

Background
There is an urgent need for more robust methods to differentiate immunotherapy responders from non-responders. In this study, by utilizing a novel multiplex imaging (MI)-based panel and analysis pipeline, we found that the spatial distribution and function of immune cells correlate with response in a cohort of immunotherapy-treated melanoma patients.   

Methods
We designed a 28-plex sequential immunofluorescence panel (seqIF™) on the COMET™ platform [1] to target key biomarkers associated with tumor microenvironment composition (TME), immune cell infiltration, and immune checkpoint pathways. Utilizing Nucleai’s deep-learning-based analysis pipeline [2], we profiled 42 pre-treatment biopsies from the SECOMBIT trial [3-5] (NCT02631447). We identified 15 cell types, including 10 different immune cell populations, and 10 cell state markers. Following cell typing, cellular neighborhoods were designated as previously described [6] and were assigned to tumor or stromal areas. By analyzing 1,941 spatial features within each treatment arm, we identified the top 3 spatial features associated with progression-free survival (PFS), overall survival (OS), and prolonged clinical benefit.  To perform the survival analysis, each spatial feature was binarized by proximity to the 0.25 and 0.75 quantiles, and the survival of the resulting groups was compared using the logrank test (Figure 1).  

Results
In Arm A (MAPK inhibitor, MAPKi, until progression followed by immune checkpoint blockade, ICB), several spatial interactions were significantly associated with patient outcomes. The spatial features associated with good MAPKi outcome showed activation of the anti-tumor immune response, including enhanced antigen presentation and PD-L1 positivity within CD8 T-cells and antigen-presenting cells, and ICOS positivity within CD4 T-cells. The high content of stromal cells was associated with a worse outcome. 

In Arm B (ICB until progression followed by MAPKi), PD-1+ CD8 T-cells presence in the tumor invasive margin and their interaction with PD-L1+ CD4 T-cells were associated with improved PFS and OS. High proportions of TCF1+ CD4 Tregs interactions with macrophages and T-cell interactions with endothelial cells were associated with worse outcome. 

In Arm C (8 weeks of MAPKi, followed by ICB until progression, followed by MAPKi), PD-L1+ and VISTA+ antigen-presenting cells’ interaction with CD4+ and CD8+ T-cells within the tumor invasive margin was associated with better outcome. In contrast, CD8+ T-cells and CD4+ T-cells interaction with CD163+ macrophages in the outer TME was associated with a worse outcome.  

Conclusions
Our data demonstrates that area-specific immune niches contribute to the success or failure of immunotherapy response, highlighting the importance of spatial biology in predicting outcomes.

References

1. Rivest F, et al. Fully automated sequential immunofluorescence (seqIF) for hyperplexspatial Sci Rep. 2023, 13(1):16994. 
2. Markovits E, et al. A novel deep learning pipeline for cell typing and phenotypic marker quantification in multiplex imaging,bioRxiv2022; 
3. Ascierto PA, et al. Sequencing of Ipilimumab Plus Nivolumab and Encorafeni bPlus Binimetinib for Untreated BRAF-Mutated Metastatic Melanoma (SECOMBIT): A Randomized, Three-Arm, Open-Label Phase II Trial. J Clin Oncol. 2023 Jan 10;41(2):212-221. 
4. Ascierto PA, et al. Sequential immunotherapy and targeted therapy for metastatic BRAF V600 mutated melanoma: 4-year survival and biomarkers evaluation from the phase II SECOMBIT trial. Nat Commun. 2024, 15(1):146 
5. Ascierto PA, et al. Sequencing of checkpoint of BRAF/MEK inhibitors on brain metastasis in Melanoma. NEJM Evid 2024;3(10)
6. Schürch, C. M. et al. Coordinated Cellular Neighborhoods Orchestrate Antitumoral Immunity at the Colorectal Cancer Invasive Front. Cell182, 1341-1359.e19 (2020).

Background
Glioblastoma (GBM), known as one of the most aggressive primary neuroepithelial tumors, presents significant challenges in diagnosis and treatment [1]. Traditional immunohistochemistry often falls short in capturing the complexity of the tumor microenvironment (TME) within GBM. With an increasing emphasis on biomarker discovery and targeting complex diseases like GBM, spatial biology has emerged as a fundamental approach, revealing multiple biomarkers while preserving spatial tissue information. The pivotal task of transforming information-rich multiplex images into crucial quantitative data remains a significant hurdle. In this study, we present an innovative workflow for generating and analyzing hyperplex images.   

Methods
We used the COMET™ platform, which performs fully automated spatial multiomics on tissue sections of human GBM samples.  On COMET™, we integrated the RNAscope™ HiPlex Pro technology [2] with a sequential immunofluorescence (seqIF™) approach [3] to detect RNA and proteins on a single tissue section simultaneously. A multiomics panel of 12 RNAs and 24 proteins was used to study immune and cancer cell populations and their cellular activity within the TME.  Multiplex immunofluorescence images were analyzed with the HORIZON™ software.   

Results
Deep-cell phenotyping was performed on the multiomics datasets using HORIZON™ to identify multi-dimensional insights about the sample. A batch analysis was performed on multiple regions of interest to apply the same analysis workflow, encompassing AI-driven cell segmentation, followed by RNA signal detection, feature extraction, and supervised classification. Additionally, spatial evaluations of cellular phenotypes of interest unveiled intricate organizational relationships within the TME and provided valuable hypotheses-driven insights into the sample.  

Conclusions
This work demonstrates how spatial multiomics offers a comprehensive view of the TME, advancing our understanding of GBM pathology. Leveraging the single-cell resolution of spatial multiomics and an intuitive image analysis approach will pave the way for researchers to seamlessly navigate the complex immune and cancer networks within the TME.

References

1. Schaff LR, Mellinghoff IK. Glioblastoma and Other Primary Brain Malignancies in Adults: A Review. JAMA. 2023 Feb 21;329(7):574-587.
2. Wang F et al., RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn. 2012 Jan;14(1):22-9.
3. Rivest F et al, Fully automated sequential immunofluorescence (seqIF) for hyperplex spatial proteomics. Sci Rep. 2023 Oct 9;13(1):16994.