SITC Annual Meeting

Background Glioblastoma (GBM) is the most aggressive primary neuroepithelial tumor diagnosed in about 14,000 people in the United States each year. While immune-based cancer treatments have been successfully deployed in the clinic for hematological malignancies, treatment of solid tumors, like GBM, has proven more challenging. Targets for immunotherapy in GBM include PD-1, PD-L1, TIM-3 and LAG-3. A comprehensive understanding of the dynamic landscape of the tumor microenvironment (TME) will lead to identification of new biomarkers, development of new therapies and improved efficacy of current therapies.

The complex nature of the TME makes spatial multiplex immunohistochemistry (IHC) and immunofluorescence (IF) ideal for visualizing biomarker expression and cellular localization. The antibody’s specificity, sensitivity, and availability for multiplex IF is a critical success factor for TME characterization. Emerging automated staining and imaging platforms require qualified antibodies for accurate marker detection. Lunaphore COMET™ platform performs sequential immunofluorescence (seqIF™), which consists of sequential cycles of staining, imaging and elution.¹ COMET™ utilizes primary antibodies and fluorescently-conjugated secondary antibodies. We used seqIF™ to understand immune checkpoint marker expression in GBM.

Methods Formalin Fixed Paraffin Embedded (FFPE) tissue sections of human tonsil, kidney, and liver were stained with IHC validated antibodies from R&D Systems™, a Bio-Techne brand. Antigen retrieval and antibody incubation times were optimized on the COMET^TM platform.

Steps of Antibody Qualification on COMET^TM: 1) IHC validated antibodies are selected. 2) Specificity is confirmed by comparing staining on a positive and negative tissue sample, visual confirmation of subcellular localization, and overall expression pattern within the tissue. 3) Sensitivity is determined by staining tissues with different expression levels of the target protein. 4) Elution efficiency is determined by a secondary antibody only incubation using the integrated workflow on COMET™.

Results Using the antibody qualification process outlined above, we show tissue specific staining on liver endothelial cells and staining localized to sinusoids for Collagen IV. To further verify antibodies for use on Lunaphore COMET™, we show specific staining in additional tissues human tonsil and kidney. Importantly, we are able to simultaneously detect multiple immune checkpoint markers, like PD-L1, CTLA4, FASL, HLA-DRA, and IDO-1 in GBM.

Conclusions This study demonstrates a systematic qualification process for antibodies in multiplex immunofluorescence. We show successful staining of multiple immune checkpoint markers in GBM on Lunaphore COMET^TM. Rigorous antibody qualification for seqIF™ on COMET™ streamlines antibody selection and allows efficient TME profiling, facilitating a more comprehensive evaluation of solid tumors for biomarker discovery and immunotherapy development.

Acknowledgements We would like to thank the Lunaphore R&D team for their partnership on this project and the antibody qualification process.

Reference

1. Rivest François, et al. ‘Fully automated sequential immunofluorescence (seqIF) for hyperplex spatial proteomics.’ Scientific Reports 2023 Oct 9;13(1):16994.

Background Advances in spatial biology have enhanced the comprehension of signaling networks by allowing the investigation of tissue architectures and cellular interactions. Techniques such as multiplexed immunofluorescence (mIF) and RNA in situ hybridization (ISH) allow simultaneous detection of multiple protein and RNA biomarkers, therefore providing a comprehensive overview of cellular functions and signaling networks.

Combining spatial assays on the same tissue section is essential to increase our knowledge of tissue biology. In complex tissues such as tumor microenvironments or neural tissues, extracting precise information on cellular interconnections or neuronal connectivity and signaling activity is key for understanding the biological processes involved in development and disease.

Here, we present a novel fully automated approach that integrates the RNAscope™ HiPlexPro assay¹ and sequential immunofluorescence (seqIF™)² protocols for the co-detection of RNA and protein targets on the same tissue section on the COMET™ platform. The multiomics protocol was applied to mouse tissues and frozen sections for the first time, demonstrating the versatility and robustness of the approach.

Methods We employed the COMET™ platform, an advanced tissue staining and imaging system, to automate and integrate RNAscope™ and seqIF™ protocols for the simultaneous detection of RNA and protein biomarkers. The system ensures precise temperature control and reagent distribution, critical for maintaining the integrity of frozen sections.

Through this automated assay, protocols with up to three cycles of RNAscope™ detection (using four fluorescent channels per cycle) were combined with up to twelve cycles of seqIF™, detecting two protein markers each. Thus, resulting in a combined multiplexing capability of up to 12-plex RNA and 24-plex protein targets.

Results In this study, we demonstrated that the combination of RNAscope™ and seqIF™ protocols on COMET™ enables the simultaneous detection of RNA and protein biomarkers on sensitive frozen tissue sections, while ensuring high reproducibility and minimal user intervention.

Here, RNAscope™ probes targeting biomarkers relevant to neuronal function, including neurotransmitters and receptors, and other glial cells, were combined to selected protein markers profiling multiple cell types in their microenvironment, including several types of infiltrating immune cells (such as CD3+, F4/80+, CD11c+, or CD56+ cells).

Conclusions Our findings demonstrate the successful application of the combined RNAscope™ and seqIF™ protocols on the COMET™ platform to analyze delicate and high autofluorescence tissue sections and tissues of non-human origin. These results demonstrate the versatility and robustness of the approach and opens door to potential new applications in the immuno-oncology field, including biomarker and drug development.

References

1. Wang F, Flanagan J, Su N, et al. RNAscope: a novel in situ RNA Analysis platform for formalin-fixed paraffin-embedded tissues. Journal of Molecular Diagnostics 2012;14:22-29.

2. Rivest F, Eroglu D, Pelz B, et al. Fully automated sequential immunofluorescence (seqIF) for hyperplex spatial proteomics. Scientific Reports 2023;13,16994.

Background The immune system plays a critical role in combating several types of malignancies. Immunotherapies including checkpoint blockade therapies have utilized the innate ability of the immune system to identify and eliminate diseased cells. It has also been established that the profile of tumor-infiltrating lymphocytes (TILs) can to some extent predict patient response and overall survival. Traditionally, standard techniques such as flow cytometry, immunofluorescence have enabled successful assessment of tumor immune profile but identifying information about their activation states and cytotoxic effects has been challenging. Spatially visualizing the expression of soluble factors such as cytokines and chemokines along with immune cell markers can provide information about the immune cell composition and enable a comprehensive understanding of mechanisms underlying immune recruitment, infiltration and exclusion.

Methods To address this, we have developed a fully automated spatial multiomic protocol on the COMET™ that enables RNA detection using the RNAscope™ HiPlexPro assay combined with protein detection using sequential immunofluorescence (seqIF™, PMID: 37813886) to integrate same-section sequential detection of up to 12 RNAs followed by up to 24 proteins. This workflow allows the user to detect any RNA and protein target of interest by utilizing the vast catalog menu of RNAscope probes or generate a custom design for RNA targets and the use of standard, non-conjugated primary antibody for protein detection.

Results Here, we have demonstrated the precise spatial profiling of FFPE solid tumors through the detection of key cytokines indicative of activated T cells and macrophages by using cytokine RNA probes such as IFNG, IL-1B, TNFA, TGFB, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12 and IL-17 in combination with cell marker antibodies such as CD3,CD8,CD4, FOXP3 and CD68 .In addition, we were also able to visualize T cell recruiting chemokines and receptors such as CXCL10, CXCL9, CXCL13, CXCL12, CXCR3 and CXCR4, and spatially map tumor infiltrating T cells within tissue context and with subcellular resolution.

Conclusions The use of RNAscope HiPlexPro on COMET™ along with seqIF allows true multiomic analysis with simultaneous visualization of RNA and protein targets for immune cell profiling. Detecting cytokine and chemokine expression provides vital information about immune cell activation and recruitment, thereby increasing our understanding of phenomenon such as immune exclusion which is key in establishing predictive signatures for immunotherapies.

Background The ability to probe the tumor microenvironment (TME) at the single cell level is important for understanding of interactions between tumors and immune cells, particularly whether immune cells have infiltrated into tumor nests, and immune checkpoint marker presence. This requires assessing multiple biomarkers simultaneously for cell phenotyping in their spatial context. Here we demonstrate rapid design and optimization of a panel of antibodies for multiplexed immunohistochemical staining of a series of oral cancer tumor samples.

Methods A panel of biomarkers were selected to cover a range of immune cell biomarkers. Immunostaining was performed in a sequential immunofluorescence platform, the Lunaphore COMET (Lunaphore, Lausanne, Switzerland). Simultaneous imaging of serial dilutions of antibodies in the same staining run permitted the fine tuning of the panel to provide appropriate signal coverage more rapidly than with previous multiplex panel methods. Once the images were acquired, tumor nests were manually annotated, followed by application of the deep learning based StarDist segmentation to the tissue. Cellular phenotypes were identified based on marker thresholding, validated by visual inspection.

Results Validation and optimization of thirteen unique immune markers was completed in less than two weeks. The staining and imaging of twenty-five oral cancer tissue samples was completed in five weeks and whole-slide image data was exported for analysis. Tissue and cell segmentation permitted the quantification of biomarkers in a spatially resolved manner. Quantitative analysis revealed the presence of immune cells both surrounding and infiltrating tumor nests within oral cancer tissue sections. Breaking down the immune cell types into subcategories, the phenotypes assessed included tumor cell count (PanCK+), tumor cell proliferation (CK+PCNA+), functional Th (CD3+CD4+FoxP3-PD1-), Treg; immune suppression (CD3+FoxP3+), Activated Tc (CD3+CD8a+Granzyme B+), Memory Th cells (CD3+CD4+CD45RO+), Memory Tc cells (CD3+CD8+CD45RO+). The full table for marker phenotyping is shown in table 1, representative markers in figure 1A/B.

Conclusions Here, we present an end-to-end workflow for the rapid development of multiplexed immunostaining on clinical tissue samples. Sequential immunofluorescence staining permitted us to optimize antibody concentrations rapidly, enabling quick imaging turnaround. Quantitative image analysis was applied to identify discrete cells, define their localization relative to tumor nests, revealing areas of tumor cell proliferation and immune cell exclusion from the tumor center (figure 1A/B). The ability to provide quantitative readouts of spatial interactions allowed for a more refined understanding of tumor-immune cell interactions, permitting a more complete interrogation of the TME in a spatially resolved manner.

Background Understanding how the abundance, localization, and functional orientation of stromal cells in the tumor microenvironment influences high-grade serous ovarian cancer (HGSC) malignancy and patient survival remains largely unknown. Biomarkers to predict the development of minimal residual disease (MRD) or mechanisms of chemoresistance following primary treatment are currently unavailable. Spatial multi-omics can reveal insights into the tumor immune microenvironment (TIME). We combined multimodal MSI with spatial transcriptomics analysis (ST; Visium), multiplexed immunofluorescence (mIF; COMET, Lunaphore), and histology to investigate the cellular and molecular heterogeneity and mechanisms of intrinsic chemoresistance of HGSC.

Methods Eight primary cytoreductive surgery samples were collected from HGSC patients (4 MRD+, 4 MRD-), and a stack of serial FFPE sections with measurements conducted in the following order: ST with H&E, mIF with DAPI staining, and MSI with H&E. MSI-measured sections were analysed sequentially for metabolites, glycans, and peptides. DAN matrix was used for metabolite analysis. The matrix was removed post-measurement. For N-glycan imaging, sections underwent PNGaseF treatment. On-tissue tryptic digestion was performed for peptide analysis. All MSI measurements were conducted on a timsTOF fleX (Bruker). Afterwards, all sections were H&E stained and digitized.

Results This study demonstrates a multimodal data acquisition workflow for combined spatial multi-omics, and describes tooling and a cross-modality data structure for correlative, differential, and spatial multimodal data analysis.

An advanced integration pipeline was used to create a common coordinate system and match readouts across the measurement stack. This accounted for the different sections and spatial resolutions of the assays, enabling integrated analysis of the MSI, mIF and ST data, including spatial correlation between analytes across assays, multi-omics tissue segmentation and differential expression analysis. A web-based spatial multi-omics visualization tool enabled direct visual comparison of analytes and interactive browsing of data analysis results (figure 1).

Correlation analysis identified differences in molecular patterns between MRD+ and MRD- samples at different omic levels (e.g. glycomic and metabolomic). Cell types identified by COMET were analyzed for differential MSI profiles, e.g. tumor vs cancer-associated fibroblasts. A spatial analysis of the tumor-stroma interface revealed variations in the distributions of cell types relative to their location and distance to said interface.

Background Identifying biomarkers that predict patient response to immunotherapy is critical for optimizing treatment strategies and improving clinical outcomes. Despite the success of immunotherapy, a significant proportion of patients do not respond to treatment. Thus, there is an urgent need for more robust methods to differentiate responders from non-responders. In this study, we present a novel multiplex imaging-based immunotherapy panel and a comprehensive analysis pipeline to characterize the spatial distribution and function of immune cells and its application for spatial biomarkers detection in a cohort of immunotherapy and targeted therapy-treated melanoma patients.

Methods We designed a 28-plex panel to perform sequential immunofluorescence (seqIF™) on the COMET™ platform¹ to target key biomarkers associated with tumor microenvironment composition (TME), immune cell infiltration, and immune checkpoint pathways (figure 1). Utilizing Nucleai’s deep-learning-based multiplex imaging analysis pipeline,² we were able to identify 13 cell types, including 9 different immune cell populations, in addition to 10 cell state markers. Cells were assigned to the tumor area or TME and spatial features were calculated based on cell type, marker positivity, and area assignments (figure 2). We obtained pre-treatment biopsies from patients with known long-term response or rapid progression to immunotherapy combination treatment (Ipilimumab+Nivolumab) from the SECOMBIT Phase II Trial (NCT02631447)^{3 4} and profiled these samples using the aforementioned panel and analysis solution. We aimed to identify spatial biomarkers that can differentiate between long-term responders and non-responders to immunotherapy.

Results Our novel multiplex imaging panel and analysis pipeline demonstrated high balanced accuracy (> 0.8) and F1 scores (> 0.8) in cell typing and protein quantification for the majority of cell types and markers. This analysis pipeline enables the quantification of known biomarkers such as T cell activation states, T cell infiltration patterns, and TLS maturation. In addition, we explored several additional biomarkers such as receptor-ligand interactions of PD-1 and PD-L1, interactions between T cells and other immune populations, and stromal cells or tumor cells as additional biomarkers for their association with patient outcomes.

Conclusions Enabling the precise identification of cells and cellular states in immunotherapy-treated patients is critical for guiding personalized treatment strategies. The development of this multiplex imaging panel and deep learning pipeline highlights the potential of integrating multiplex imaging with AI analysis to enhance our understanding of treatment efficacy and resistance mechanisms, ultimately aiming to improve patient outcomes in clinical practice.

Acknowledgements The authors thank the patients and families who made this trial possible. Additionally, the authors acknowledge the clinical study teams and CRO who participated in the trial and in particular Paola Schiavo e Mariarita Arenella from CRT (Clinical Research Technology – Salerno). We thank Bristol-Myers Squibb (Princeton, NJ) and Array Biopharma Inc/Pfizer (Boulder, CO) for support. Moreover, the authors thank the participating investigators who did not enroll any patients and thus are not included as authors on the paper, Koelblinger P, Hafner C, Hoeller C (Austria), Weide B (Germany), Larkin J, Lorigan P (UK).

References

1. Rivest F, et al. Fully automated sequential immunofluorescence (seqIF) for hyperplex spatial proteomics. Sci Rep2023;13(1):16994.

2. Markovits E, et al. A novel deep learning pipeline for cell typing and phenotypic marker quantification in multiplex imaging. BioRxiv 2022.

3. Ascierto PA, et al. Sequencing of ipilimumab plus nivolumab and encorafenib plus 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.

Ethics Approval This study was designed in 2015 as a phase II, open-label randomized trial with no formal comparative test and a single-stage design for each arm. Patients were enrolled at 37 academic medical centers in 9 countries. The trial protocol was approved by the appropriate ethics body at each participating institution and is available in the Supplementary Information file. An independent data monitoring committee oversaw the trial. SECOMBIT is registered at ClinicalTrials.gov (NCT02631447). The study design and conduct complied with all current regulations regarding the use of human study participants and was conducted in accordance with the criteria set by the Declaration of Helsinki.

Background Neoadjuvant and perioperative immunotherapy with checkpoint inhibitors (ICIs), combined with chemotherapy (CT), have improved outcomes for patients with resectable non-small cell lung cancer (NSCLC). Pathologic tumor response has been utilized in clinical trials as a surrogate of clinical efficacy. Yet, over half of treated patients experience non-response, highlighting the need for novel markers to identify those who are most likely to derive therapeutic benefit.

Methods We aimed to dissect the spatial complexity of the tumor-associated immune landscape in lung tissues (n = 28) of 25 patients with a range of residual viable tumor (%RVT) following neoadjuvant chemoimmunotherapy (CT+ICI). We performed multimodal spatial transcriptomics (ST) with co-detection of 35 proteins (Visium CytAssist, 10X Genomics), as well as single-cell RNA-sequencing of fixed cells (scFFPE-seq, 10X Genomics) from consecutive sections of the same tissues. We also examined spatial patterns of 20-plex panel of immune and stromal markers at single-cell resolution in a subset of samples using sequential immunofluorescence (seq-IF)-based proteomics (COMET Lunaphore).

Results We generated and sequenced high-quality ST with protein co-detection and scFFPE libraries from 28 tissue samples. Preliminary analysis on tissues from 3 patients representing a range of%RVT – pathological complete response (pCR; 0% RVT), partial response (20% RVT), and no-response (100% RVT), identified heterogeneity in the immune response. We found robust gene-protein correlation, such as MS4A1 gene expression and CD20 protein levels, in our multimodal spatial-omics analysis. Intriguingly, the arrangement of B cells and their interaction with surrounding plasma cells (PCs) varied notably across patients. Unlike in the patients with partial or no response, pronounced tertiary lymphoid structure (TLS) formation and signatures were observed in the pCR patient, with the CXCL13-CXCR5 axis overlapping with the TLS profile, suggesting a functional interplay that may be crucial in mediating response to therapy. Intriguingly, in the pCR patient, PCs uniquely arrayed around dense B cell aggregates, marked by MS4A1, CR2, FCER1, and CXCL13. This spatial relationship between PCs and TLS was absent in the non-pCR patients.

Conclusions Our high-resolution and multimodal analysis identified spatially-resolved expression and patterns for B lineage cells that may be associated, and thus, underlie response of resectable NSCLC to neoadjuvant CT+ICI. Ongoing efforts on this expanding cohort will shed further light on transcriptional states and immunogenomic roles of B lineage cells in neoadjuvant CT+ICI-treated NSCLC.

Background Immune cell phenotyping of the tumor microenvironment is transforming cancer diagnostics and treatment by identifying novel predictive biomarkers for immunotherapy and powerful prognostic markers, thus providing deeper insights of disease mechanisms. Using our unique, clinical trial-ready mIF workflow, we demonstrate how custom multiplexed immunofluorescence (mIF) and digital image analysis offer an innovative method to investigate invasion in solid tumors by studying the tumor and the corresponding microenvironment.

Methods We have optimized a multiplex immunofluorescence assay on Lunaphore Comet™ for use with human control tissues and FFPE solid tumor specimens. Our comprehensive mIF panel includes immune cell markers CD3, CD4, CD8, FoxP3, CD56, CD20, CD68, CD11c, PD-L1, PD-1, CD45, Ki67 as an indicator of proliferation, and mismatch-repair proteins MLH1, MSH2, PMS2, MSH6. In addition, pan-Cytokeratin as tumor cell specific marker, α-SMA and FAP-α were used to delineate the tumor area including tumor-associated stroma. Following a rigorous and standardized approach to mIF panel establishment and validation, we established accuracy by comparingthe final mIF to single plex bright field immunohistochemistry as the ‘ground truth’. Furthermore, each individual target was independently validated for specificity regarding elution efficacy and epitope stability to repeated antibody elution. Automated image analysis and cellular phenotyping followed a workflow of custom Visiopharm™ apps. The tissue was segmented into tumor and non-tumor regions of interest by manual annotation. Additional parameters like TNM classification were also taken into consideration for data analysis.

Results The applied mIF immune cell panel and automated image analysis workflow identified a diverse range of immune cell phenotypes, their spatial relation to tumor cells and distribution within the activated stroma region.

Conclusions Our findings offer new insights into the complexity of the tumor microenvironment in solid tumors, highlighting significant roles played by both the immune system and the stromal cells on tumor cell invasion into the surrounding tissue. Our methodology not only deepens our understanding of tumor biology but also paves the way for new therapeutic strategies that target tumor microenvironment.

Background Multiplex immunofluorescence (mIF) has become fundamental for tumor microenvironment (TME) and biomarker research.¹ Therein, different marker expression levels often have biological meaning and the detection of markers expressed at low vs. high levels can provide crucial insights.² However, detecting antigens expressed at lower levels can be challenging and often benefits from signal amplification.³ Here, we demonstrate a new automated method enabling the amplification of individual markers within multiplex panels. This technique provides an adjustable level of amplification and an efficient elution for subsequent staining cycles.

Methods COMET™ is an automated platform that performs sequential immunofluorescence (seqIF™) assays.⁴ Based on this, the novel amplification method increases the number of detection antibodies per primary antibody in a cyclic manner leading to a stronger signal. The amplification complex can be eluted, enabling subsequent staining cycles while preserving tissue integrity. Formalin-fixed paraffin-embedded (FFPE) tissue sections from human colorectal and breast carcinoma were stained using a 26-plex panel including 20 markers in standard seqIF™ and 6 amplified markers, together covering most basic immune cell types, functions, and stroma compartments. The performance was assessed by comparing marker expression with and without amplification using Lunaphore HORIZON™ image analysis software as well as comparing amplified staining to single-plex chromogenic immunohistochemistry (IHC) for each marker.⁵ Finally, Hematoxylin and Eosin (H&E) staining was performed on slides retrieved from COMET™.

Results Our study demonstrated the successful integration of 6 amplified markers within a 20-plex seqIF™ panel. Amplified low-expressed markers could be reliably detected and their amplification produced a controllable signal intensity increase between low and high expressing cell populations, providing a higher dynamic range when compared to the unamplified seqIF™. Specifically, amplification enabled the detection of the full spectrum of PD-1 an PD-L1 expression intensity and identification of several low-expressing subtypes of regulatory immune cells such as Treg and macrophages within the TME. Furthermore, the staining patterns of amplified markers shows good correlation with chromogenic IHC staining, and elution of the amplification complex did not damage the tissue as demonstrated by the post-COMET™ H&E staining.

Conclusions Hyperplex seqIF™ panels integrating this novel amplification technique will enable the detection of markers that are expressed at low levels, which could not be robustly captured by non-amplified mIF, such as the full expression range of critical immune checkpoint markers. Analyzing a high number of markers simultaneously while allowing selective amplification will improve the profiling of immune and tumoral cells within their environment.

References

1. Francisco-Cruz A, Parra ER, Tetzlaff MT, Wistuba II. Multiplex immunofluorescence assays. Methods Mol Biol 2020;2055:467-495.

2. Berry S, Giraldo NA, Green BF, et al. Analysis of multispectral imaging with the AstroPath platform informs efficacy of PD-1 blockade. Science 2021;372(6547):eaba2609.

3. Stack EC, Wang C, Roman KA, Hoyt CC. Multiplexed immunohistochemistry, imaging, and quantitation: a review, with an assessment of tyramide signal amplification, multispectral imaging and multiplex analysis. Methods 2014;70(1):46-58.

4. Rivest F, Eroglu D, Pelz B, et al. Fully automated sequential immunofluorescence (seqIF) for hyperplex spatial proteomics. Sci Rep 2023;13(1):16994.

5. Taube JM, Akturk G, Angelo M, et al. The society for immunotherapy of cancer statement on best practices for multiplex immunohistochemistry (IHC) and immunofluorescence (IF) staining and validation. J Immunother Cancer2020;8(1):e000155.

Background Advances in the molecular understanding of carcinogenesis and immunology have transformed the development of targeted anticancer drugs addressing both the tumor and the immune system. The success of these therapies depends on the target associated signal transduction and/or interaction of different types of cells. Using our unique, clinical trial-ready mIF workflow, we demonstrate how custom multiplexed immunofluorescence (mIF) and digital image analysis offer an innovative method to categorize Her-2 expressing breast cancers by evaluating the tumor cells and the tumor microenvironment.

Methods We have optimized a multiplex immunofluorescence assay on Lunaphore Comet™ for use with human control tissues and FFPE solid tumor specimens. Among others, the mIF panel included standard immune cell markers CD3, CD4, CD8, FoxP3, CD56, CD20, CD68, CD11c, PD-L1, PD-1 and CD45. Pan cytokeratin, α-SMA and FAP-α were used to delineate the tumor area and investigate reactivity of the surrounding stroma. Diagnostic and selected Her-2 associated signal transduction markers included E-Cadherin, ER, PgR, PTEN, phospho-ERK1/2, phospho-AKT/PKB, ki67, CyclinD1, p27, EGFR, Her-2, phospho-Her-2, Her-3, and phospho-Her-3. Following a rigorous and standardized approach to mIF panel establishment and validation, we established accuracy by comparing the final mIF to single plex bright field immunohistochemistry as the ‘ground truth’. Furthermore, each individual target was independently validated for specificity regarding elution efficacy and epitope stability to repeated antibody elution. Automated image analysis and cellular phenotyping followed a workflow of custom Visiopharm™ apps. The tissue was segmented into tumor and non-tumor regions of interest by manual annotation.

Results Additional parameters like TNM classification were also taken into consideration for data analysis after image analysis. Statistical analyses employing unsupervised hierarchical 2D clustering helped us gain significant insights in Her-2 downstream signaling and breast cancer biology.

Conclusions For each patient population, the cluster analysis revealed the potential for sub-classification by means of molecular marker expression patterns, independent of clinical features. In the nearby future, protein expression analyses like the herein presented work will empower physicians to make more informed decisions ultimately leading to a more personalized and effective therapy.