optiPRM: A Targeted Immunopeptidomics LC-MS Workflow With Ultra-High Sensitivity for the Detection of Mutation-Derived Tumor Neoepitopes From Limited Input Material.

Personalized cancer immunotherapies such as therapeutic vaccines and adoptive transfer of T cell receptor-transgenic T cells rely on the presentation of tumor-specific peptides by human leukocyte antigen class I molecules to cytotoxic T cells. Such neoepitopes can for example arise from somatic mutations and their identification is crucial for the rational design of new therapeutic interventions. Liquid chromatography mass spectrometry (LC-MS)-based immunopeptidomics is the only method to directly prove actual peptide presentation and we have developed a parameter optimization workflow to tune targeted assays for maximum detection sensitivity on a per peptide basis, termed optiPRM. Optimization of collision energy using optiPRM allows for the improved detection of low abundant peptides that are very hard to detect using standard parameters. Applying this to immunopeptidomics, we detected a neoepitope in a patient-derived xenograft from as little as 2.5 × 106 cells input. Application of the workflow on small patient tumor samples allowed for the detection of five mutation-derived neoepitopes in three patients. One neoepitope was confirmed to be recognized by patient T cells. In conclusion, optiPRM, a targeted MS workflow reaching ultra-high sensitivity by per peptide parameter optimization, makes the identification of actionable neoepitopes possible from sample sizes usually available in the clinic.

Graph machine learning for integrated multi-omics analysis.

Multi-omics experiments at bulk or single-cell resolution facilitate the discovery of hypothesis-generating biomarkers for predicting response to therapy, as well as aid in uncovering mechanistic insights into cellular and microenvironmental processes. Many methods for data integration have been developed for the identification of key elements that explain or predict disease risk or other biological outcomes. The heterogeneous graph representation of multi-omics data provides an advantage for discerning patterns suitable for predictive/exploratory analysis, thus permitting the modeling of complex relationships. Graph-based approaches-including graph neural networks-potentially offer a reliable methodological toolset that can provide a tangible alternative to scientists and clinicians that seek ideas and implementation strategies in the integrated analysis of their omics sets for biomedical research. Graph-based workflows continue to push the limits of the technological envelope, and this perspective provides a focused literature review of research articles in which graph machine learning is utilized for integrated multi-omics data analyses, with several examples that demonstrate the effectiveness of graph-based approaches.

Molecular prediction of clinical response to anti-PD-1/anti-PD-L1 immune checkpoint inhibitors: New perspectives for precision medicine and mass spectrometry-based investigations.

Monoclonal antibodies (mAbs) acting as immune checkpoint inhibitors (ICIs) are among the most frequently used immunotherapies in oncology. However, precision medicine approaches to adapt the treatment to the patient are still poorly exploited. Given the risk of severe adverse reactions, predicting patient eligibility for ICI therapy represents a great asset for precision medicine. Today, the extended panel of mass spectrometric approaches, accompanied by newly developed sample preparation methods is a strategy of choice for responder and non-responder stratification on a molecular basis, and early detection of resistance. In this perspective article, we review the biodisposition of mAbs, the interest in molecular stratification of patients treated with these mAbs, and the possible analytical strategies to achieve this goal, with a major emphasis on mass spectrometric approaches.

MediMer: a versatile do-it-yourself peptide-receptive MHC class I multimer platform for tumor neoantigen-specific T cell detection.

Peptide-loaded MHC class I (pMHC-I) multimers have revolutionized our capabilities to monitor disease-associated T cell responses with high sensitivity and specificity. To improve the discovery of T cell receptors (TCR) targeting neoantigens of individual tumor patients with recombinant MHC molecules, we developed a peptide-loadable MHC class I platform termed MediMer. MediMers are based on soluble disulfide-stabilized ß2-microglobulin/heavy chain ectodomain single-chain dimers (dsSCD) that can be easily produced in large quantities in eukaryotic cells and tailored to individual patients' HLA allotypes with only little hands-on time. Upon transient expression in CHO-S cells together with ER-targeted BirA biotin ligase, biotinylated dsSCD are purified from the cell supernatant and are ready to use. We show that CHO-produced dsSCD are free of endogenous peptide ligands. Empty dsSCD from more than 30 different HLA-A,B,C allotypes, that were produced and validated so far, can be loaded with synthetic peptides matching the known binding criteria of the respective allotypes, and stored at low temperature without loss of binding activity. We demonstrate the usability of peptide-loaded dsSCD multimers for the detection of human antigen-specific T cells with comparable sensitivities as multimers generated with peptide-tethered ß2m-HLA heavy chain single-chain trimers (SCT) and wild-type peptide-MHC-I complexes prior formed in small-scale refolding reactions. Using allotype-specific, fluorophore-labeled competitor peptides, we present a novel dsSCD-based peptide binding assay capable of interrogating large libraries of in silico predicted neoepitope peptides by flow cytometry in a high-throughput and rapid format. We discovered rare T cell populations with specificity for tumor neoepitopes and epitopes from shared tumor-associated antigens in peripheral blood of a melanoma patient including a so far unreported HLA-C*08:02-restricted NY-ESO-1-specific CD8+ T cell population. Two representative TCR of this T cell population, which could be of potential value for a broader spectrum of patients, were identified by dsSCD-guided single-cell sequencing and were validated by cognate pMHC-I multimer staining and functional responses to autologous peptide-pulsed antigen presenting cells. By deploying the technically accessible dsSCD MHC-I MediMer platform, we hope to significantly improve success rates for the discovery of personalized neoepitope-specific TCR in the future by being able to also cover rare HLA allotypes.

Tertiary lymphoid structures and their association to immune phenotypes and circulatory IL2 levels in pancreatic ductal adenocarcinoma.

Pancreatic ductal adenocarcinoma (PDA) is usually unresponsive to immunotherapeutic approaches. However, tertiary lymphoid structures (TLS) are associated with favorable patient outcomes in PDA. A better understanding of the B cell infiltrate and biological features of TLS formation is needed to further explore their potential and improve patient management. We analyzed tumor tissues (n = 55) and corresponding blood samples (n = 51) from PDA patients by systematical immunohistochemistry and multiplex cytokine measurements. The tissue was compartmentalized in "tumor" and "stroma" and separately examined. Clinical patient information was used to perform survival analyses. We found that the mere number of B cells is not associated with patient survival, but formation of TLS in the peritumoral stroma is a prognostic favorable marker in PDA patients. TLS-positive tissues show a higher density of CD8+ T cells and CD20+ B cells and a higher IL2 level in the peritumoral stroma than tissues without TLS. Compartmental assessment shows that gradients of IL2 expression differ with regard to TLS formation: TLS presence is associated with higher IL2 levels in the stromal than in the tumoral compartment, while no difference is seen in patients without TLS. Focusing on the stroma-to-serum gradient, only patients without TLS show significantly higher IL2 levels in the serum than in stroma. Finally, low circulatory IL2 levels are associated with local TLS formation. Our findings highlight that TLS are prognostic favorable and associated with antitumoral features in the microenvironment of PDA. Also, we propose easily accessible serum IL2 levels as a potential marker for TLS prediction.

Co-Stimulatory Bispecific Antibodies Induce Enhanced T Cell Activation and Tumor Cell Killing in Breast Cancer Models.

Although T cell-recruiting CD3-binding bispecific antibodies (BiMAb) have been proven to be clinically effective for hematologic malignancies, the success of BiMAb targeting solid tumor-associated antigens (TAA) in carcinomas so far remains poor. We reasoned that provision of co-stimulatory BiMAb in combination with αTAA-αCD3 BiMAb would boost T cell activation and proliferative capacity, and thereby facilitate the targeting of weakly or heterogeneously expressed tumor antigens. Various αTAA-αCD3 and αTAA-αCD28 BiMAb in a tetravalent IgG1-Fc based format have been analyzed, targeting multiple breast cancer antigens including HER2, EGFR, CEA, and EpCAM. Moreover, bifunctional fusion proteins of αTAA-tumor necrosis factor ligand (TNFL) superfamily members including 4-1BBL, OX40L, CD70 and TL1A have been tested. The functional activity of BiMAb was assessed using co-cultures of tumor cell lines and purified T cells in monolayer and tumor spheroid models. Only in the presence of tumor cells, αTAA-αCD3 BiMAb activated T cells and induced cytotoxicity in vitro, indicating a strict dependence on cross-linking. Combination treatment of αTAA-αCD3 BiMAb and co-stimulatory αTAA-αCD28 or αTAA-TNFL fusion proteins drastically enhanced T cell activation in terms of proliferation, activation marker expression, cytokine secretion and tumor cytotoxicity. Furthermore, BiMAb providing co-stimulation were shown to reduce the minimally required dose to achieve T cell activation by at least tenfold. Immuno-suppressive effects of TGF-ß and IL-10 on T cell activation and memory cell formation could be overcome by co-stimulation. BiMAb-mediated co-stimulation was further augmented by immune checkpoint-inhibiting antibodies. Effective co-stimulation could be achieved by targeting a second breast cancer antigen, or by targeting fibroblast activation protein (FAP) expressed on another target cell. In tumor spheroids derived from pleural effusions of breast cancer patients, co-stimulatory BiMAb were essential for the activation tumor-infiltrating lymphocytes and cytotoxic anti-tumor responses against breast cancer cells. Taken together we showed that co-stimulation significantly potentiated the tumoricidal activity of T cell-activating BiMAb while preserving the dependence on TAA recognition. This approach could provide for a more localized activation of the immune system with higher efficacy and reduced peripheral toxicities.

Peripheral blood and tissue assessment highlights differential tumor-circulatory gradients of IL2 and MIF with prognostic significance in resectable pancreatic ductal adenocarcinoma.

Various reports have pointed out the potential of cytokines as diagnostic and prognostic biomarkers for pancreatic ductal adenocarcinoma (PDA). Nonetheless, the evidence is contradictory and the role of chronic inflammation and relationship between circulatory and corresponding tumoral cytokine levels remain elusive. Utilizing a broad array of cytokines, we identified two opposing parameters: serum levels of interleukin 2 (IL2) and macrophage migration inhibitory factor (MIF) are diagnostic and prognostic factors. While low IL2 levels are associated with PDA, they also relate to a favorable prognosis of patients. In contrast, high MIF levels are associated with PDA and simultaneously related to an unfavorable outcome. MIF levels are associated with the intratumoral density of M2 macrophages (CD163+). Focusing on the tumor-to-serum gradient, we unveiled a different pattern of compartmental cytokine expression between IL2 and MIF. Our findings indicate that an extra-tumoral source of IL2 exists in PDA patients leading to increased detectability in the circulatory system. In case of MIF, the tumor microenvironment is presumably the main site of production and thereby reflected by serum measurements. Taken together, our study describes IL2 and MIF levels as biomarker candidates for diagnosis and prognosis of PDA, highlighting the need for compartmental cytokine analyses. From the perspective of tumor immunobiology, we identify multiple inflammatory states (proposed as types I-III) and see that systemic chronic dysregulation, independent of tumor microenvironment, can be measured and is a possible tool for stratification. Thus, direct correlation of local cytokine levels to peripheral blood levels needs to be regarded with caution.

Association of circulating PLA2G7 levels with cancer cachexia and assessment of darapladib as a therapy.

BACKGROUND: Cancer cachexia (CCx) is a multifactorial wasting disorder characterized by involuntary loss of body weight that affects many cancer patients and implies a poor prognosis, reducing both tolerance to and efficiency of anticancer therapies. Actual challenges in management of CCx remain in the identification of tumour-derived and host-derived mediators involved in systemic inflammation and tissue wasting and in the discovery of biomarkers that would allow for an earlier and personalized care of cancer patients. The aim of this study was to identify new markers of CCx across different species and tumour entities. METHODS: Quantitative secretome analysis was performed to identify specific factors characteristic of cachexia-inducing cancer cell lines. To establish the subsequently identified phospholipase PLA2G7 as a marker of CCx, plasma PLA2G7 activity and/or protein levels were measured in well-established mouse models of CCx and in different cohorts of weight-stable and weight-losing cancer patients with different tumour entities. Genetic PLA2G7 knock-down in tumours and pharmacological treatment using the well-studied PLA2G7 inhibitor darapladib were performed to assess its implication in the pathogenesis of CCx in C26 tumour-bearing mice. RESULTS: High expression and secretion of PLA2G7 were hallmarks of cachexia-inducing cancer cell lines. Circulating PLA2G7 activity was increased in different mouse models of CCx with various tumour entities and was associated with the severity of body wasting. Circulating PLA2G7 levels gradually rose during cachexia development. Genetic PLA2G7 knock-down in C26 tumours only partially reduced plasma PLA2G7 levels, suggesting that the host is also an important contributor. Chronic treatment with darapladib was not sufficient to counteract inflammation and tissue wasting despite a strong inhibition of the circulating PLA2G7 activity. Importantly, PLA2G7 levels were also increased in colorectal and pancreatic cancer patients with CCx. CONCLUSIONS: Overall, our data show that despite no immediate pathogenic role, at least when targeted as a single entity, PLA2G7 is a consistent marker of CCx in both mice and humans. The early increase in circulating PLA2G7 levels in pre-cachectic mice supports future prospective studies to assess its potential as biomarker for cancer patients.

A nonviral, nonintegrating DNA nanovector platform for the safe, rapid, and persistent manufacture of recombinant T cells.

The compelling need to provide adoptive cell therapy (ACT) to an increasing number of oncology patients within a meaningful therapeutic window makes the development of an efficient, fast, versatile, and safe genetic tool for creating recombinant T cells indispensable. In this study, we used nonintegrating minimally sized DNA vectors with an enhanced capability of generating genetically modified cells, and we demonstrate that they can be efficiently used to engineer human T lymphocytes. This vector platform contains no viral components and is capable of replicating extrachromosomally in the nucleus of dividing cells, providing persistent transgene expression in human T cells without affecting their behavior and molecular integrity. We use this technology to provide a manufacturing protocol to quickly generate chimeric antigen receptor (CAR)-T cells at clinical scale in a closed system and demonstrate their enhanced anti-tumor activity in vitro and in vivo in comparison to previously described integrating vectors.

Dysregulated Host Response in Severe Acute Respiratory Syndrome Coronavirus 2-Induced Critical Illness.

BACKGROUND: Impaired immune response has been reported to be the cause of the development of coronavirus disease 2019 (COVID-19)-related respiratory failure. Further studies are needed to understand the immunopathogenesis and to enable an improved stratification of patients who are at risk for critical illness. METHODS: Thirty-two severely ill patients hospitalized with COVID-19 were recruited in our center at the University Hospital Heidelberg. We performed a comprehensive analysis of immune phenotype, cytokine, and chemokine profiling and leukocyte transcripts in patients with severe COVID-19 and compared critically ill patients who required mechanical ventilation and high-flow oxygen therapy and noncritically ill patient who received low-flow oxygen therapy. RESULTS: Critically ill patients exhibited low levels of CD8 T cells and myeloid dendritic cells. We noted a pronounced CCR6+ TH17 phenotype in CD4 central memory cells and elevated circulating levels of interleukin-17 in the critical group. Gene expression of leukocytes derived from critically ill patients was characterized by an upregulation of proinflammatory cytokines and reduction of interferon (IFN)-responsive genes upon stimulation with Toll-like receptor 7/8 agonist. When correlating clinical improvement and immune kinetics, we found that CD8 T-cell subsets and myeloid dendritic cells significantly increased after disconnection from the ventilator. CONCLUSION: Critical illness was characterized by a TH17-mediated response and dysfunctional IFN-associated response, indicating an impaired capacity to mount antiviral responses during severe acute respiratory syndrome coronavirus 2 severe infection.

Rapid MALDI-MS Assays for Drug Quantification in Biological Matrices: Lessons Learned, New Developments, and Future Perspectives.

Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) has rarely been used in the field of therapeutic drug monitoring, partly because of the complexity of the ionization processes between the compounds to be quantified and the many MALDI matrices available. The development of a viable MALDI-MS method that meets regulatory guidelines for bioanalytical method validation requires prior knowledge of the suitability of (i) the MALDI matrix with the analyte class and properties for ionization, (ii) the crystallization properties of the MALDI matrix with automation features, and (iii) the MS instrumentation used to achieve sensitive and specific measurements in order to determine low pharmacological drug concentrations in biological matrices. In the present hybrid article/white paper, we review the developments required for the establishment of MALDI-MS assays for the quantification of drugs in tissues and plasma, illustrated with concrete results for the different steps. We summarize the necessary parameters that need to be controlled for the successful development of fully validated MALDI-MS methods according to regulatory authorities, as well as currently unsolved problems and promising ways to address them. Finally, we propose an expert opinion on future perspectives and needs in order to establish MALDI-MS as a universal method for therapeutic drug monitoring.

Interrogating the microenvironmental landscape of tumors with computational image analysis approaches.

The tumor microenvironment is an interacting heterogeneous collection of cancer cells, resident as well as infiltrating host cells, secreted factors, and extracellular matrix proteins. With the growing importance of immunotherapies, it has become crucial to be able to characterize the composition and the functional orientation of the microenvironment. The development of novel computational image analysis methodologies may enable the robust quantification and localization of immune and related biomarker-expressing cells within the microenvironment. The aim of the review is to concisely highlight a selection of current and significant contributions pertinent to methodological advances coupled with biomedical or translational applications. A further aim is to concisely present computational advances that, to our knowledge, have currently very limited use for the assessment of the microenvironment but have the potential to enhance image analysis pipelines; on this basis, an example is shown for the detection and segmentation of cells of the microenvironment using a published pipeline and a public dataset. Finally, a general proposal is presented on the conceptual design of automation-optimized computational image analysis workflows in the biomedical and clinical domain.

A transplantable tumor model allowing investigation of NY-BR-1-specific T cell responses in HLA-DRB1*0401 transgenic mice.

BACKGROUND: NY-BR-1 has been described as a breast cancer associated differentiation antigen with intrinsic immunogenicity giving rise to endogenous T and B cell responses. The current study presents the first murine tumor model allowing functional investigation of NY-BR-1-specific immune responses in vivo. METHODS: A NY-BR-1 expressing tumor model was established in DR4tg mice based on heterotopic transplantation of stable transfectant clones derived from the murine H2 compatible breast cancer cell line EO771. Composition and phenotype of tumor infiltrating immune cells were analyzed by qPCR and FACS. MHC I binding affinity of candidate CTL epitopes predicted in silico was determined by FACS using the mutant cell line RMA-S. Frequencies of NY-BR-1 specific CTLs among splenocytes of immunized mice were quantified by FACS with an epitope loaded Db-dextramer. Functional CTL activity was determined by IFNγ catch or IFNγ ELISpot assays and statistical analysis was done applying the Mann Whitney test. Tumor protection experiments were performed by immunization of DR4tg mice with replication deficient recombinant adenovirus followed by s.c. challenge with NY-BR-1 expressing breast cancer cells. RESULTS: Our results show spontaneous accumulation of CD8+ T cells and F4/80+ myeloid cells preferentially in NY-BR-1 expressing tumors. Upon NY-BR-1-specific immunization experiments combined with in silico prediction and in vitro binding assays, the first NY-BR-1-specific H2-Db-restricted T cell epitope could be identified. Consequently, flow cytometric analysis with fluorochrome conjugated multimers showed enhanced frequencies of CD8+ T cells specific for the newly identified epitope in spleens of immunized mice. Moreover, immunization with Ad.NY-BR-1 resulted in partial protection against outgrowth of NY-BR-1 expressing tumors and promoted intratumoral accumulation of macrophages. CONCLUSION: This study introduces the first H2-Db-resctricted CD8+ T cell epitope-specific for the human breast cancer associated tumor antigen NY-BR-1. Our novel, partially humanized tumor model enables investigation of the interplay between HLA-DR4-restricted T cell responses and CTLs within their joint attack of NY-BR-1 expressing tumors.

Patterns of antibody responses to nonviral cancer antigens in head and neck squamous cell carcinoma patients differ by human papillomavirus status.

There have been hints that nonviral cancer antigens are differentially expressed in human papillomavirus (HPV)-positive and HPV-negative head and neck squamous cell carcinoma (HNSCC). Antibody responses (AR) to cancer antigens may be used to indirectly determine cancer antigen expression in the tumor using a noninvasive and tissue-saving liquid biopsy. Here, we set out to characterize AR to a panel of nonviral cancer antigens in HPV-positive and HPV-negative HNSCC patients. A fluorescent microbead multiplex serology to 29 cancer antigens (16 cancer-testis antigens, 5 cancer-retina antigens and 8 oncogenes) and 29 HPV-antigens was performed in 382 HNSCC patients from five independent cohorts (153 HPV-positive and 209 HPV-negative). AR to any of the cancer antigens were found in 272/382 patients (72%). The ten most frequent AR were CT47, cTAGE5a, c-myc, LAGE-1, MAGE-A1, -A3, -A4, NY-ESO-1, SpanX-a1 and p53. AR to MAGE-A3, MAGE-A9 and p53 were found at significantly different prevalences by HPV status. An analysis of AR mean fluorescent intensity values uncovered remarkably different AR clusters by HPV status. To identify optimal antigen selections covering a maximum of patients with ≤10 AR, multiobjective optimization revealed distinct antigen selections by HPV status. We identified that AR to nonviral antigens differ by HPV status indicating differential antigen expression. Multiplex serology may be used to characterize antigen expression using serum or plasma as a tissue-sparing liquid biopsy. Cancer antigen panels should address the distinct antigen repertoire of HPV-positive and HPV-negative HNSCC.

Antibody Responses to Cancer Antigens Identify Patients with a Poor Prognosis among HPV-Positive and HPV-Negative Head and Neck Squamous Cell Carcinoma Patients.

PURPOSE: The identification of high-risk patients within human papillomavirus (HPV)-positive and -negative head and neck squamous cell carcinoma (HNSCC) is needed for improved treatment and surveillance strategies. In this study, we set out to discover antibody responses (AR) with prognostic impact in HNSCC stratified by HPV status. EXPERIMENTAL DESIGN: A fluorescent bead-based multiplex serology assay on 29 cancer antigens (16 cancer-testis antigens, 5 cancer-retina antigens, and 8 oncogenes) and 29 HPV antigens was performed in samples of 362 patients with HNSCC from five independent cohorts (153 HPV positive, 209 HPV negative). A multivariable Cox proportional hazard model with bootstrapping (M = 1000) was used for validation of prognostic antibody responses. RESULTS: Antibody response to any of the cancer antigens was found in 257 of 362 patients (71%). In HPV-negative patients, antibody responses to c-myc, MAGE-A1, -A4, and Rhodopsin E2 (combined as ARhigh risk) were significantly associated with shorter overall survival. In HPV-positive patients, antibody responses to IMP-1 were discovered as a negative prognostic factor. ARhigh risk (HR = 1.76) and antibody responses to IMP-1 (HR = 3.28) were confirmed as independent markers for a poor prognosis in a multivariable Cox proportional hazard model with bootstrapping (M = 1000). CONCLUSIONS: We identified antibody responses to cancer antigens that associate with a dismal prognosis in patients with HNSCC beyond HPV-positive status. ARhigh risk may be used to detect HPV-negative patients with an extraordinarily bad prognosis. Most importantly, antibody response to IMP-1 may serve as a marker for a subgroup of HPV-positive patients who present with a poor prognosis similar to that in HPV-negative patients.

Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study.

BACKGROUND: For virtually every patient with colorectal cancer (CRC), hematoxylin-eosin (HE)-stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can extract prognosticators directly from these widely available images. METHODS AND FINDINGS: We hand-delineated single-tissue regions in 86 CRC tissue slides, yielding more than 100,000 HE image patches, and used these to train a CNN by transfer learning, reaching a nine-class accuracy of >94% in an independent data set of 7,180 images from 25 CRC patients. With this tool, we performed automated tissue decomposition of representative multitissue HE images from 862 HE slides in 500 stage I-IV CRC patients in the The Cancer Genome Atlas (TCGA) cohort, a large international multicenter collection of CRC tissue. Based on the output neuron activations in the CNN, we calculated a "deep stroma score," which was an independent prognostic factor for overall survival (OS) in a multivariable Cox proportional hazard model (hazard ratio [HR] with 95% confidence interval [CI]: 1.99 [1.27-3.12], p = 0.0028), while in the same cohort, manual quantification of stromal areas and a gene expression signature of cancer-associated fibroblasts (CAFs) were only prognostic in specific tumor stages. We validated these findings in an independent cohort of 409 stage I-IV CRC patients from the "Darmkrebs: Chancen der Verhütung durch Screening" (DACHS) study who were recruited between 2003 and 2007 in multiple institutions in Germany. Again, the score was an independent prognostic factor for OS (HR 1.63 [1.14-2.33], p = 0.008), CRC-specific OS (HR 2.29 [1.5-3.48], p = 0.0004), and relapse-free survival (RFS; HR 1.92 [1.34-2.76], p = 0.0004). A prospective validation is required before this biomarker can be implemented in clinical workflows. CONCLUSIONS: In our retrospective study, we show that a CNN can assess the human tumor microenvironment and predict prognosis directly from histopathological images.

Topography of cancer-associated immune cells in human solid tumors.

Lymphoid and myeloid cells are abundant in the tumor microenvironment, can be quantified by immunohistochemistry and shape the disease course of human solid tumors. Yet, there is no comprehensive understanding of spatial immune infiltration patterns ('topography') across cancer entities and across various immune cell types. In this study, we systematically measure the topography of multiple immune cell types in 965 histological tissue slides from N = 177 patients in a pan-cancer cohort. We provide a definition of inflamed ('hot'), non-inflamed ('cold') and immune excluded patterns and investigate how these patterns differ between immune cell types and between cancer types. In an independent cohort of N = 287 colorectal cancer patients, we show that hot, cold and excluded topographies for effector lymphocytes (CD8) and tumor-associated macrophages (CD163) alone are not prognostic, but that a bivariate classification system can stratify patients. Our study adds evidence to consider immune topographies as biomarkers for patients with solid tumors.

Multiplex bead-based measurement of humoral immune responses against tumor-associated antigens in stage II melanoma patients of the EORTC18961 trial.

Purpose: Determine the prognostic and predictive significance of tumor associated antigen (TAA)-specific serum antibodies in melanoma patients of a large adjuvant vaccination phase III trial. Patients and methods: Serum IgG antibodies were measured against a panel of 43 antigens by a bead-based multiplex assay in 970 stage II melanoma patients of the EORTC18961 trial, evaluating adjuvant ganglioside GM2-KLH/QS-21 vaccination versus observation. Primary end point was relapse-free survival (RFS). Patients' sera at baseline, after 12 and 48 weeks of study treatment and at the last available time point (at recurrence/remission) were evaluated. Results: Prognostic clinical variables are gender, surgical confirmation of lymph node-negative status, Breslow thickness and ulceration of the primary. Prognostic spontaneous antibody responses were associated with a significant dismal (GM2, Rhod_E2, SSX2) or good prognosis (CyclinB1, SCYE1v1) for RFS, distant metastasis-free (DMFS) or overall survival (OS). Predictive spontaneous antibody responses based on significant interaction with treatment were RhodN p = 0.02, Rab38 p = 0.04 for RFS, RhodE2 p = 0.006, Recoverin p = 0.04 for DMFS and RhodE2 p = 0.003; Recoverin p = 0.04, NA17.A p = 0.04, for OS respectively. The subgroups of patients according to antibody responses for RFS were determined for RhodN sero-negative (n = 849, HR = 1.07, p = 0.6); RhodN sero-positive (n = 121,HR = 0.42, p = 0.01) and Rab38 sero-negative (n = 682, HR = 1.12, p = 0.42), Rab38 sero-positive (n = 288, HR = 0.65, p = 0.04) patients respectively. Conclusion: We identified prognostic serum antibody responses against TAA in stage II melanoma patients. A set of antibody responses correlated with a beneficial outcome for GM2 vaccination.