Multi-parameter cytometry technologies enable high-dimensional analysis of immune cell populations at single-cell resolution. Deep learning has been widely applied to these datasets, but existing methods often struggle with transferability across datasets due to technical variability, batch effects and identification of biologically relevant cell populations, limiting their utility in clinical research. We present dioscRi, a transferable deep learning framework that integrates a maximum mean discrepancy variational autoencoder for normalization and de-noising, enhancing cross-dataset compatibility. Changes in cell type proportions and marker expression are identified by structuring these features within biologically or empirically derived cell type hierarchies. These hierarchies are incorporated directly into an overlapping group LASSO model, improving the prediction of clinical outcomes. When applied to a coronary artery disease study, dioscRi recapitulated several known immune associations. Benchmarking across multiple datasets demonstrated dioscRi’s ability to transfer across cohorts with compatible marker panels and outperform existing methods on three of four datasets, establishing it as an interpretable tool for cytometry data analysis. Cytometry-based prediction models often fail to generalise across patient cohorts. Here, the authors present dioscRi, a framework that enables interpretable prediction of clinical outcomes from immune cell populations across independent cohorts.

Abstract Hepatocellular carcinoma (HCC) is a heterogeneous malignancy, requiring spatially resolved multi-omic approaches to advance therapeutic strategies. Spatial transcriptomics using the Xenium™ platform enables high-throughput mapping of hundreds to thousands of RNA targets within intact tissue architecture. While transcript-level insights provide critical context for understanding gene expression patterns, integrating proteomic data adds a complementary layer that enables direct validation of biomarker expression. Spatial proteomics technologies, such as Imaging Mass Cytometry™ (IMC™), complement transcriptomics by providing high-dimensional protein expression data at subcellular resolution. IMC leverages metal-tagged antibodies and laser ablation to simultaneously quantify over 40 protein markers with 5 orders of magnitude linear dynamic range, surpassing traditional immunohistochemistry and immunofluorescence. We demonstrate the feasibility and biological insights gained from applying IMC to same tissue sections previously processed with Xenium, integrating transcriptomic and proteomic data through computational co-registration. Formalin-fixed, paraffin-embedded HCC tissue sections were profiled using a custom Xenium v1 transcriptomic panel, followed by IMC with a 43-marker immuno-oncology themed antibody panel on the same section. IMC was also performed on serial sections without prior Xenium processing for performance comparison. Data integration was achieved using Xenium Explorer software, which employs a computational co-registration algorithm to align nuclei across modalities, enabling overlay of transcriptomic and proteomic biomarkers for spatial correlation analysis. IMC performed post-Xenium processing generated high-quality data comparable to IMC alone, preserving tumor and immune cell phenotyping capabilities. Both techniques localized macrophages, neutrophils, B cells, cytotoxic T cells, and T helper cells and their activation states within distinct tissue regions. Computational integration of transcriptomic and proteomic datasets revealed subpopulations of immune cells and activation states, as well as discrepancies between RNA and protein localization for several markers, underscoring the importance of multi-modal validation. This integrated approach provided a more nuanced view of HCC microenvironmental complexity. Overall, we demonstrate concurrent application of spatial transcriptomics and proteomics at the cellular level on the same tissue section. This integrated workflow, enabled by computational co-registration, delivers a multidimensional perspective of tumor biology and uncovers spatial relationships between unique cell populations with varying activation states offering novel insights into HCC heterogeneity and informing the development of precision therapeutic strategies. Citation Format: Atefeh Khakpoor, Qanber Raza, Merrin Mary Eapen, Dina Kazemi, Erin Coll, Liang Lim, Christina Loh, Nick Zabinyakov, Ling Qiao, Anna Di Bartolomeo, Helen McGuire, Jacob George, Ankur Sharma. Integrating spatial transcriptomics and Imaging Mass Cytometry™ for multi-omic mapping of hepatocellular carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 795.

Globally, regional and remote communities are burdened by both an increased prevalence and worse prognosis of many infectious and chronic diseases. However, largely owing to logistical challenges, these communities are under-represented in clinical trials and research studies. As individuals from rural communities experience unique environmental exposures and risk factors for disease, immune phenotyping data collected from metropolitan populations may not be broadly generalizable. To address this, we present a workflow that enables the inclusion of resource-limited sites in high-parameter mass cytometry studies. In this approach, whole blood (WB) or peripheral blood mononuclear cells (PBMCs) are collected, stained fresh for surface antigens, and cryopreserved at the collection site. Samples are then shipped to the central site for further processing, including neutrophil depletion, fixation, barcoding, intracellular staining, and data acquisition. Importantly, the WB staining approach does not require specialized equipment such as centrifuges and is therefore feasible to perform in a resource-limited environment. A support protocol details steps for data preprocessing and cleanup. We present example data demonstrating the application of this workflow to determine immune differences between eight patients with late-stage lung cancer and four healthy blood donors. Overall, this workflow may improve access to underserved communities and facilitate, for the first time, the scalability of immune phenotyping studies to harness geographically dispersed clinical centers. © 2026 The Author(s). Current Protocols published by Wiley Periodicals LLC. Basic Protocol 1: Preparation and staining of PBMCs for cytometry Basic Protocol 2: Preparation and staining of whole blood for cytometry Basic Protocol 3: Fixation, permeabilization, intracellular staining, and data acquisition for blood sample immunophenotyping Support Protocol: Data preprocessing and cleanup.

Abstract Improved cancer prognosis begins with the collection of high-quality data from clinical trials. However, execution of longitudinal immune phenotyping studies by flow cytometry is complex due to logistical challenges that could affect data quality. The challenges of immune monitoring are exacerbated when considering remote and rural communities in Australia, which are burdened by both an increased prevalence and worse prognosis of cancer. Paradoxically, these communities are often underrepresented in clinical trials. Our group has developed a novel blood-based “immune signature” that robustly predicts failure to make a clinical response to checkpoint therapies targeting the PD-1/PD-L1 pathway in melanoma and lung cancer. We have now taken these groundbreaking findings to implement in remote clinical settings, along with an expanded CyTOF™ panel for 50-plus-parameter analysis of cancer patient peripheral blood samples. Recent developments in CyTOF technology facilitate a highly simplified workflow of asynchronous sample collection and staining, compatible with current hospital laboratories, which typically run on unpredictable schedules, to remote settings, which are resource-poor and have limited clinical trial infrastructure. This is uniquely enabled by using a stable, dried-down cocktail of CyTOF antibodies and an easy-to-follow protocol that yields reproducible staining while minimizing sample required. Furthermore, this unique mass cytometry workflow provides flexibility and minimizes technical variation via sample barcoding and freezing of stained samples for shipment to a central site for batch acquisition. As such, this study is designed to demonstrate both clinical impact of the large panel and utility of CyTOF technology for a highly simplified and robust workflow for multi-site clinical trials. Here we present findings from the implementation of this novel workflow through a 100-sample pilot study at three sites across Australia, in a variety of clinical implementation setups. Remote asynchronous sample collection was performed to compare two surface staining approaches on i) PBMC versus ii) plasma-depleted whole blood to suit settings where centrifugation access was readily available or not, respectively. Stained samples were shipped to a central lab for processing. In the case of cryopreserved whole blood, bead-based granulocyte depletion was first performed, with all samples subsequently multiplex barcoded together for intracellular antibody staining of key functional markers. Analysis harmonization was achieved across sample collection sites and staining approaches. Overall, this workflow enables access to underserved communities, facilitating for the first time equity and scalability of immune phenotyping studies to harness truly geographically dispersed clinical centers. Citation Format: Natalie J. Smith, Michael Cohen, Lauren J. Tracey, Julie Alipaz, Christina Loh, David A. King, Neha Pulyani, Rebecca Auzins, Elin S. Gray, Sandra Taylor, Rajat Rai, Steven Kao, Barbara Fazekas de St Groth, Helen McGuire. Implementation of a novel immune monitoring strategy for multicenter clinical trials to enable equity in cancer prognosis for individuals from remote settings of Australia [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7723.

Imaging flow cytometry (IFC) is a powerful cell analytic tool that exploits multi-parameters in single-cell images to characterise cell phenotypes and fluorescence information. It enables in-depth analysis of cell signalling, DNA repair and marker localisation. However, conventional frame-based acquisition is bound to the triangle of imaging constraints—speed, resolution and sensitivity, which has become an everlasting challenge to overcome during development. Neuromorphic photosensors detect contrast changes in a scene via individual-firing pixels, characterising superior data efficiency, temporal resolution and fluorescence sensitivity. In this work, we have developed a neuromorphic imaging cytometer (NIC) to capture fast-moving cell events, curating the first neuromorphic cell dataset with human blood cells, endothelial cells and artificial particles. Recently, this sensor has been adopted to address the limitations in IFC with prominent results in diverse modalities and machine learning approaches. Such a dataset serves as a baseline of healthy cell groups for both diagnostic and research purposes. In addition, the rich spatial information derived from cell images has exceptional uses with deep learning (DL) approaches to automate cell analysis, classification, sorting and gating strategy. We also trained a lightweight model combining the convolutional block attention module with a spiking neural network (CBAM-SNN) to automate cell analysis and classification. The proposed architecture has achieved a promising performance of 97% accuracy and F1 score with a significant reduction in computation requirements. Combining the data sparsity in neuromorphic imaging with a lightweight DL model and operation platform can enable next-generation, AI-driven cytometry to deliver point-of-care diagnostic and research solutions.

Immune checkpoint inhibitors (ICIs) are currently the most effective treatment for late-stage lung cancer. However, most patients fail to mount a durable response, and the mechanisms underlying non-responsiveness remain elusive. Here, we apply a universal omics approach to identify correlates of non-responsiveness to ICIs in lung cancer. In-depth characterization of the plasma proteome was performed using the SomaScan™ platform on pretreatment and longitudinal blood samples from 40 ICI-treated patients, resulting in quantification of 10, 000 unique proteins. In parallel, comprehensive immune phenotyping was achieved with matched pretreatment PBMC for 90% of patients (36/40) through CyTOF™ technology, with a 38-plex panel describing over 150 immune populations in circulation. To enhance clinical outcome predictability of cellular and plasma-based immune relationships, cross-platform data integration was achieved with Stabl, a sparse, reliable omic biomarkers analysis strategy. Stabl identified key predictive features from both CyTOF and SomaScan technology, providing insight into immune deficits present in non-responsive patients. The combined model demonstrated a stronger capability to predict non-response from a pretreatment blood sample (AUROC = 0.79, p-value = 1.7e-2, Mann-Whitney non-parametric test) compared with individual platforms alone. Additionally, this multi-omic strategy is compatible with Imaging Mass Cytometry™ technology, unlocking another layer of clinical investigation via spatial biology. This study ultimately demonstrates both the clinical impact and utility of blood- and imaging-based CyTOF technology and the benefit of combining SomaScan and multi-omic-appropriate analysis approaches. Overall, robust treatment prediction, attributed to biomarker features, provides insights into mechanisms for non-response and highlights the potential for better treatment options in lung cancer. Helen M. McGuire, Natalie Smith, Michael Cohen, Julien Hedou, Grégoire Bellan, Xavier Durand, Erika L. Smith-Mahoney, Julie Alipaz, Jennifer Snyder-Cappione, Brice Gaudilliere, Christina Loh, David King, Michael Hinterberg, Clare Paterson, Barbara Fazekas de St Groth. Application of a comprehensive multi-omic immune profiling strategy achieves superior checkpoint immunotherapy response prediction in lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5004.

Background: Chronic wounds, such as diabetes-related foot ulcers, arise from delayed wound healing and create significant health and economic burdens. Macrophages regulate healing by shifting between pro- and anti-inflammatory phenotypes, known as macrophage polarization. Sex and diabetes can impair wound healing, but their influence on macrophage phenotype in skin tissue during wound healing remains unclear, which was investigated in this study using a novel two-sex diabetic mouse model. Methods: Diabetes was induced in male and female C57BL/6J mice using low-dose streptozotocin injections and high-fat diet feeding, with chow-fed mice as controls. After 18 weeks, each mouse received four circular full-thickness dorsal skin wounds. The macrophage phenotypes in wounded skin tissues at Day 0 and Day 10 post-wounding were analyzed using mass cytometry with manual gating and automated computational clustering. Results: Male diabetic mice exhibited more severe hyperglycemia and insulin resistance compared to females. Although diabetic mice did not display delayed wound healing, male mice had a greater proportion of total macrophages than females, especially a higher proportion of pro-inflammatory matrix metalloproteinase-9 (MMP-9)+ macrophages and a lower proportion of anti-inflammatory adiponectin receptor 1 (AdipoR1)+ macrophages in male diabetic mice compared to females, indicating an imbalanced polarization towards a pro-inflammatory phenotype that could result in poorer wound healing. Interestingly, computational clustering identified a new pro-inflammatory, pro-healing phenotype (Ly6C+AdipoR1+CD163–CD206–) more abundant in females than males, suggesting this phenotype may play a role in the transition from the inflammatory to the proliferative stage of wound healing. Conclusions: This study demonstrated a significant sex-based difference in macrophage populations, with male diabetic mice showing a pro-inflammatory bias that may impair wound healing, while a unique pro-inflammatory, pro-healing macrophage population more abundant in females could facilitate recovery. Further research is needed to investigate the role of these newly identified phenotypes in regulating impaired wound healing.

Improved cancer prognosis begins with the collection of high-quality data from clinical trials. However, the execution of longitudinal immune phenotyping studies by flow cytometry is complex, primarily due to logistical challenges that could affect data quality. The challenges of immune monitoring are exacerbated when considering remote and rural communities in Australia, which are burdened by both an increased prevalence and worse prognosis of cancer. Paradoxically, these communities are often underrepresented in clinical trials. Our group has developed a novel blood-based “immune signature” that robustly predicts failure to make a clinical response to checkpoint therapies targeting the PD-1/PD-L1 pathway in melanoma and lung cancer. These groundbreaking findings were achieved through applying a comprehensive 38-parameter immunophenotyping CyTOF™ panel to biobanked peripheral immune cells from cancer patients. We now aim to increase our clinical implementation in remote settings, with an expanded CyTOF panel of 50-plus-parameter analysis. Recent developments in CyTOF technology facilitate a highly simplified workflow of asynchronous sample collection and staining, compatible with current hospital laboratories that typically run on unpredictable schedules, to remote settings. This is uniquely enabled by using a dried-down cocktail of CyTOF antibodies that are stable, simple to use and yield reproducible staining while minimizing sample required. Furthermore, this workflow, unique to mass cytometry, provides flexibility and minimizes technical variation via sample barcoding and freezing of stained samples for shipment to a central site for batch acquisition. As such, this study is designed to ultimately demonstrate both clinical impact of the large panel and utility of CyTOF technology for a highly simplified and robust workflow for multi-site clinical trials. Here we present findings from the implementation of this novel workflow through a 100-sample pilot study. Remote asynchronous sample collection and surface staining was performed at five sites across Australia. Stained samples were then shipped to a central lab for multiplex barcoding and intracellular staining of several key functional markers. A combination of manual gating and automated high-dimensional clustering was performed in CellEngine, with principal component analysis of all samples demonstrating harmony across the sites. Overall, this workflow enables access to underserved communities, facilitating for the first time equity and scalability of immune phenotyping studies to harness truly geographically dispersed clinical centers. Natalie Smith, Michael Cohen, Julie Alipaz, Christina Loh, David King, Steven Kao, Sandra Taylor, Rajat Rai, Barbara Fazekas de St Groth, Helen McGuire. Implementation of a novel immune monitoring strategy for multicenter clinical trials to enable equity in cancer prognosis for individuals from remote settings of Australia [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2349.

With the increasing use of spectral flow and mass cytometry technologies, efficient single-cell phenotyping has become essential for the identification of complex cell populations. Traditionally, these populations are identified through manual gating, a time-consuming process that is also subject to variability and a lack of reproducibility. Here, we adapted a deep learning model [1] to devise an automated gating pipeline with the goal of enhancing the accuracy, speed, and reproducibility of immune cell gating. Our pipeline was evaluated on two mass-cytometry datasets [1], [2] (>5M cells, 35 markers), that had been previously gated by experts for comparison purposes. Phenotypic markers were used for identification, while the median expression of intracellular markers was used to determine the functional properties of populations. Accuracy and F1-score were used to compare classified populations against manually gated populations. A third dataset (15M events from 57 patients diagnosed with lung cancer) was used to compare the prediction of lung cancer progression after immunotherapy using properties based on manually vs. automatically gated populations. The automated gating pipeline achieved high overall accuracy comparable to expert manual gating (n=30 cell populations) (0.917 and 0.921 for dataset 1 and 2, respectively). Processing time was significantly reduced: <12min for both datasets (8CPUs, 8GB of RAM). High-level populations (n=10) were identified with excellent accuracy (e.g. Tcells, Bcells with f1-score of 0.993, 0.937 on the first dataset and 0.996, 0.934 on the second dataset). However, rare populations (n=20) showed higher discrepancies (e.g. intermediate monocytes, DC with f1-scores of 0.864, 0.816 and 0.678, 0.552 on dataset 1 and 2, respectively). Differences in identification scores had low effect on the functional properties of populations, as 91.4% of the functional properties defined from our pipeline were highly correlated (r>0.75) with those derived from manual gating. Further, the prediction of lung cancer progression after immunotherapy showed similar or improved results using functional properties based on automated identified populations (AUC=0.82) compared to manually gated populations (AUC=0.70). This automated approach eliminates operator biases and handles multiple markers simultaneously, offering reliable and efficient analyses for research and clinical applications. To explore discrepancies, future work will incorporate datasets annotated by multiple experts to assess inter-expert variability. 1. Blampey, Q. et al, A biology-driven deep generative model for cell-type annotation in cytometry. Brief Bioinform. 2023 Sep 20;. doi: 10.1093/bib/bbad260. 2. https://clinicaltrials.gov/study/NCT05523713 3. Ina A. Stelzer et al, Sci.Transl.Med.13, (2021). DOI:10.1126/scitranslmed.abd9898 Benjamin Waked, Grégoire Bellan, Xavier Durand, Alexandre Maillard, Franck Verdonk, Brice Gaudilliere, Helen McGuire, Natalie Smith, Christina Loh, David King, Dominique Blanchard, Julien Hedou. An automated and scalable pipeline for high-dimensional immune cell phenotyping in mass cytometry datasets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2500.

Introduction:Peripheral immune dysfunction may be critically involved in the pathophysiology of migraine. Some evidence supports a role for peripheral T cells, monocytes, and humoral factors including kynurenine metabolites and cytokines, however a comprehensive picture has yet to emerge. Objective:This study sought to undertake a systematic assessment of the immune changes in episodic and chronic migraine across phases of the migraine cycle. Methods:Migraine patients in different phases of the migraine cycle with a confirmed diagnosis of episodic or chronic migraine and age- and sex-matched healthy controls were recruited. Peripheral blood was assessed for circulating immune cells, plasma proteins, and kynurenine pathway metabolites in a cross-sectional case-control design. Data were acquired using high-dimensional approaches including proteomics, single-cell mass cytometry, and imaging flow cytometry. Results:Plasma proteins related to increased cell-cell adhesion and altered enzymatic activity were increased in migraine. The migraine prodrome displayed a strong and distinct proinflammatory phenotype defined by increased platelet-neutrophil aggregation, quinolinic acid production, and matrix metalloproteinase-9 expression. Migraine patients in the attack phase instead expressed higher levels of cytokine receptors and phosphorylated transcription factors in Th17 cells, monocytes, natural killer cells, and B cells. T cells were shifted to a mobilised, recirculating phenotype across all migraine phases. Episodic and chronic migraine patients were only distinguished by subtle changes in T-cell phenotype. Conclusion:Distinct proinflammatory peripheral signatures were detected between migraine phases, while few alterations distinguished episodic and chronic status. These data provide a resource that may aid in the identification of peripheral immune cells and mediators contributing to migraine attack onset.

Antibody titration is an important step in every cytometric workflow, with the goal being to determine antibody concentrations that ensure highly reproducible results. When aiming to compare antigen expression between samples using mean or median fluorescence intensity (MFI), reagents should be used at a saturating concentration so that unavoidable variations in staining conditions do not affect the fluorescence signal. The recommended concentrations of commercially available fluorophore-labeled monoclonal antibodies (mAbs) may not achieve plateau staining, and their saturating concentration may be too high to be experimentally useful. To address these common concerns, we present a novel method to achieve saturation of fluorophore-conjugated mAbs, by ‘spiking-in’ unlabelled antibody of the same clone. Here, we demonstrate the application of this workflow to human anti-CD3 (clone OKT3, mouse IgG2a) and anti-TCRαβ (clone IP26, mouse IgG1), two mAbs that do not achieve saturation at 2-fold above their commercially recommended concentrations. First, the saturating concentration of unlabelled (purified) OKT3 and IP26 was determined by detection with a fluorophore-labeled anti-mouse IgG (H + L) secondary antibody. Titration curves of unlabelled and labeled mAbs were compared for each clone to determine whether labeling had resulted in any loss in binding activity. Unlabelled antibody was then ‘spiked’ into the labeled antibody at varying ratios, and those that achieved saturation while maintaining an adequate fluorescence signal were identified. We demonstrate that antibody saturation can be achieved with an optimized mixture of labeled and unlabelled antibody, while maintaining a clear signal from the fluorophore. While this workflow has only been applied to OKT3 and IP26, it has potential applicability for any antibody clone for which both labeled and unlabelled preparations are available. This method has significance for robust comparison of biomarker expression when fluorophore labeled reagents do not reach saturation under standard staining conditions.

Approximately 50% of melanoma patients fail to respond to immune checkpoint blockade (ICB), and acquired resistance hampers long-term survival in about half of initially responding patients. Whether targeting BET reader proteins, implicated in epigenetic dysregulation, can enhance ICB response rates and durability, remains to be determined. Here we show elevated BET proteins correlate with poor survival and ICB responses in melanoma patients. The BET inhibitor IBET151, combined with anti-CTLA-4, overcame innate ICB resistance however, sequential BET inhibition failed against acquired resistance in mouse models. Combination treatment response in the innate resistance model induced changes in tumor-infiltrating immune cells, reducing myeloid-derived suppressor cells (MDSCs). CD4+ and CD8+ T cells showed decreased expression of inhibitory receptors, with reduced TIM3, LAG3, and BTLA checkpoint expression. In human PBMCs in vitro, BET inhibition reduced expression of immune checkpoints in CD4+ and CD8+ T cells, restoring effector cytokines and downregulating the transcriptional driver TOX. BET proteins in melanoma may play an oncogenic role by inducing immune suppression and driving T cell dysfunction. The study demonstrates an effective combination for innately unresponsive melanoma patients to checkpoint inhibitor immunotherapy, yet highlights BET inhibitors' limitations in an acquired resistance context.

Imaging flow cytometry (IFC) is an advanced cell-analytic technology offering rich spatial information and fluorescence intensity for multi-parametric characterization. Manual gating in cytometry data enables the classification of discrete populations from the sample based on extracted features. However, this expert-driven technique can be subjective and laborious, often presenting challenges in reproducibility and being inherently limited to bivariate analysis. Numerous AI-driven cell classifications have recently emerged to automate the process of including multivariate data with enhanced reproducibility and accuracy. Our previous work demonstrated the early development of neuromorphic imaging cytometry, evaluating its feasibility in resolving conventional frame-based imaging systems’ limitations in data redundancy, fluorescence sensitivity, and compromised throughput. Herein, we adopted a convolutional spiking neural network (SNN) combined with the YOLOv3 model (SNN-YOLO) to perform cell classification and detection on label-free samples under neuromorphic vision. Spiking techniques are inherently suitable post-processing techniques for neuromorphic vision sensing. The experiment was conducted with polystyrene-based microparticles, THP-1, and LL/2 cell lines. The network’s performance was compared with that of a traditional YOLOv3 model fed with event-generated frame data to serve as a baseline. In this work, our SNN-YOLO outperformed the YOLOv3 baseline by achieving the highest average class accuracy of 0.974, compared to 0.962 for YOLOv3. Both models reported comparable performances across other key metrics and should be further explored for future auto-gating strategies and cytometry applications.

Abstract Immunotherapies stimulate T cell function in the tumor microenvironment (TME). CD3+ and CD8+ T cell infiltration can predict disease recurrence in colorectal cancer (CRC) patients and has been validated as Immunoscore® (IS). The TME comprises multiple heterogeneous populations of immune cells, including T cells and cancer-associated fibroblasts (CAFs); CRC tumors are also morphologically diverse. Standard BioTools™ (SBI) has developed a novel whole slide Imaging Mass Cytometry™ technique that allows complete analysis of cell population heterogeneity in the TME using >40 markers. We hypothesised there are multiple heterogeneous subtypes of CAFs within CRC tumors and that they modulate T cell infiltration and patient outcome. In collaboration with SBI, we studied a cohort of CRC patients stratified based on T cell infiltrate and disease recurrence. We identified multiple CAF populations that interact with tumor-infiltrating T cell populations. Slides from 10 CRC patients with high or low CD3+CD8+ T cell infiltrate (IS) were stained with a panel encompassing markers for CAF and T cell heterogeneity, function, and metabolism. Slides were imaged using the Hyperion XTi™ Imaging System at SBI. Using a scan mode called Preview Mode, the whole slide was imaged in minutes to identify CAF and T cell phenotypic regions of interest. Next, the same slide was imaged at 1 μm resolution using Cell Mode (CM). Finally, for each sample, an additional serial section was imaged using a slightly lower resolution called Tissue Mode (TM). TM allowed us to image the samples in several hours instead of days while capturing the full heterogeneity of the entire tissue sample. We used clustering to identify phenotype clusters and neighbourhood analysis to study spatial enrichment. We also compared results between CM and TM. We report: Identification of heterogeneous CAF subtypes in CRC patients with high or low IS; determination of CAF populations that spatially interact with T cells in patients with high or low IS; and examination of consistent spatial interactions between CAF and T cells in CRC, and whether these interactions differ in patients with high or low IS. TM revealed several heterogeneous cellular populations that differed between patients with high or low IS; importantly, low IS tumors had more podoplanin+αSMA+ CAF populations and fewer podoplanin+αSMA- cell populations compared with high IS tumors. High IS tumors had more PD-1+CD8+ T cells compared with low IS tumors; these PD-1+CD8+ T cells spatially interacted with podoplanin+αSMA- CAF populations in high IS tumors. CAF-PD-1+CD8+ T cell interactions were absent from low IS tumors. Thus, CAF-PD-1+CD8+ T cell interactions are important in CRC patients with high IS, and these cellular interactions require further investigation. Citation Format: Rory M. Costello, Sonya Fenton, Helen McGuire, Liang Lim, David Howell, Qanber Raza, Jyh Yun Chwee, Roslyn Kemp. Predicting colorectal patient prognoses by functional characterisation of heterogeneous cell types and their spatial interaction using a new technique: Whole slide imaging mass cytometry [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 67.