Prognostic implications of tumor-infiltrating T cells in early-stage endometrial cancer.

Patients with endometrial cancer differ in terms of the extent of T-cell infiltration; however, the association between T-cell subpopulations and patient outcomes remains unexplored. We characterized 285 early-stage endometrial carcinoma samples for T-cell infiltrates in a tissue microarray format using multiplex fluorescent immunohistochemistry. The proportion of T cells and their subpopulations were associated with clinicopathological features and relapse-free survival outcomes. CD3+ CD4+ infiltrates were more abundant in the patients with higher grade or non-endometrioid histology. Cytotoxic T cells (CD25+, PD-1+, and PD-L1+) were strongly associated with longer relapse-free survival. Moreover, CD3+ PD-1+ stromal cells were independent of other immune T-cell populations and clinicopathological factors in predicting relapses. Patients with high stromal T-cell fraction of CD3+ PD-1+ cells were associated with a 5-year relapse-free survival rate of 93.7% compared to 79.0% in patients with low CD3+ PD-1+ fraction. Moreover, in patients classically linked to a favorable outcome (such as endometrioid subtype and low-grade tumors), the stromal CD3+ PD-1+ T-cell fraction remained prognostically significant. This study supports that T-cell infiltrates play a significant prognostic role in early-stage endometrial carcinoma. Specifically, CD3+ PD-1+ stromal cells emerge as a promising novel prognostic biomarker.

Spatial immunoprofiling of the intratumoral and peritumoral tissue of renal cell carcinoma patients.

While the abundance and phenotype of tumor-infiltrating lymphocytes are linked with clinical survival, their spatial coordination and its clinical significance remain unclear. Here, we investigated the immune profile of intratumoral and peritumoral tissue of clear cell renal cell carcinoma patients (n = 64). We trained a cell classifier to detect lymphocytes from hematoxylin and eosin stained tissue slides. Using unsupervised classification, patients were further classified into immune cold, hot and excluded topographies reflecting lymphocyte abundance and localization. The immune topography distribution was further validated with The Cancer Genome Atlas digital image dataset. We showed association between PBRM1 mutation and immune cold topography, STAG1 mutation and immune hot topography and BAP1 mutation and immune excluded topography. With quantitative multiplex immunohistochemistry we analyzed the expression of 23 lymphocyte markers in intratumoral and peritumoral tissue regions. To study spatial interactions, we developed an algorithm quantifying the proportion of adjacent immune cell pairs and their immunophenotypes. Immune excluded tumors were associated with superior overall survival (HR 0.19, p = 0.02) and less extensive metastasis. Intratumoral T cells were characterized with pronounced expression of immunological activation and exhaustion markers such as granzyme B, PD1, and LAG3. Immune cell interaction occurred most frequently in the intratumoral region and correlated with CD45RO expression. Moreover, high proportion of peritumoral CD45RO+ T cells predicted poor overall survival. In summary, intratumoral and peritumoral tissue regions represent distinct immunospatial profiles and are associated with clinicopathologic characteristics.

Clonal heterogeneity influences drug responsiveness in renal cancer assessed by ex vivo drug testing of multiple patient-derived cancer cells.

Renal cell cancer (RCC) has become a prototype example of the extensive intratumor heterogeneity and clonal evolution of human cancers. However, there is little direct evidence on how the genetic heterogeneity impacts on drug response profiles of the cancer cells. Our goal was to determine how genomic clonal evolution impacts drug responses. Finding from our study could help to define the challenge that clonal evolution poses on cancer therapy. We established multiple patient-derived cells (PDCs) from different tumor regions of four RCC patients, verified their clonal relationship to each other and to the uncultured tumor tissue by genome sequencing. Furthermore, comprehensive drug-sensitivity testing with 460 oncological drugs was performed on all PDC clones. The PDCs retained many cancer-specific copy number alterations and mutations in driver genes such as VHL, PBRM1, PIK3C2A, KMD5C and TSC2 genes. The drug testing highlighted vulnerability in the PDCs toward approved RCC drugs, such as the mTOR-inhibitor temsirolimus, but also novel sensitivities were uncovered. The individual PDC clones from different tumor regions in a patient showed distinct drug-response profiles, suggesting that genomic heterogeneity contributes to the variability in drug responses. Studies of multiple PDCs from a patient with cancer are informative for elucidating cancer heterogeneity and for the determination on how the genomic evolution is manifested in cancer drug responsiveness. This approach could facilitate tailoring of drugs and drug combinations to individual patients.

Breast cancer outcome prediction with tumour tissue images and machine learning.

PURPOSE: Recent advances in machine learning have enabled better understanding of large and complex visual data. Here, we aim to investigate patient outcome prediction with a machine learning method using only an image of tumour sample as an input. METHODS: Utilising tissue microarray (TMA) samples obtained from the primary tumour of patients (N = 1299) within a nationwide breast cancer series with long-term-follow-up, we train and validate a machine learning method for patient outcome prediction. The prediction is performed by classifying samples into low or high digital risk score (DRS) groups. The outcome classifier is trained using sample images of 868 patients and evaluated and compared with human expert classification in a test set of 431 patients. RESULTS: In univariate survival analysis, the DRS classification resulted in a hazard ratio of 2.10 (95% CI 1.33-3.32, p = 0.001) for breast cancer-specific survival. The DRS classification remained as an independent predictor of breast cancer-specific survival in a multivariate Cox model with a hazard ratio of 2.04 (95% CI 1.20-3.44, p = 0.007). The accuracy (C-index) of the DRS grouping was 0.60 (95% CI 0.55-0.65), as compared to 0.58 (95% CI 0.53-0.63) for human expert predictions based on the same TMA samples. CONCLUSIONS: Our findings demonstrate the feasibility of learning prognostic signals in tumour tissue images without domain knowledge. Although further validation is needed, our study suggests that machine learning algorithms can extract prognostically relevant information from tumour histology complementing the currently used prognostic factors in breast cancer.

Spatial aspects of oncogenic signalling determine the response to combination therapy in slice explants from Kras-driven lung tumours.

A key question in precision medicine is how functional heterogeneity in solid tumours informs therapeutic sensitivity. We demonstrate that spatial characteristics of oncogenic signalling and therapy response can be modelled in precision-cut slices from Kras-driven non-small-cell lung cancer with varying histopathologies. Unexpectedly, profiling of in situ tumours demonstrated that signalling stratifies mostly according to histopathology, showing enhanced AKT and SRC activity in adenosquamous carcinoma, and mitogen-activated protein kinase (MAPK) activity in adenocarcinoma. In addition, high intertumour and intratumour variability was detected, particularly of MAPK and mammalian target of rapamycin (mTOR) complex 1 activity. Using short-term treatment of slice explants, we showed that cytotoxic responses to combination MAPK and phosphoinositide 3-kinase-mTOR inhibition correlate with the spatially defined activities of both pathways. Thus, whereas genetic drivers determine histopathology spectra, histopathology-associated and spatially variable signalling activities determine drug sensitivity. Our study is in support of spatial aspects of signalling heterogeneity being considered in clinical diagnostic settings, particularly to guide the selection of drug combinations. © 2018 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.

Chronic Activation of Innate Immunity Correlates With Poor Prognosis in Cancer Patients Treated With Oncolytic Adenovirus.

Despite many clinical trials conducted with oncolytic viruses, the exact tumor-level mechanisms affecting therapeutic efficacy have not been established. Currently there are no biomarkers available that would predict the clinical outcome to any oncolytic virus. To assess the baseline immunological phenotype and find potential prognostic biomarkers, we monitored mRNA expression levels in 31 tumor biopsy or fluid samples from 27 patients treated with oncolytic adenovirus. Additionally, protein expression was studied from 19 biopsies using immunohistochemical staining. We found highly significant changes in several signaling pathways and genes associated with immune responses, such as B-cell receptor signaling (P < 0.001), granulocyte macrophage colony-stimulating factor (GM-CSF) signaling (P < 0.001), and leukocyte extravasation signaling (P < 0.001), in patients surviving a shorter time than their controls. In immunohistochemical analysis, markers CD4 and CD163 were significantly elevated (P = 0.020 and P = 0.016 respectively), in patients with shorter than expected survival. Interestingly, T-cell exhaustion marker TIM-3 was also found to be significantly upregulated (P = 0.006) in patients with poor prognosis. Collectively, these data suggest that activation of several functions of the innate immunity before treatment is associated with inferior survival in patients treated with oncolytic adenovirus. Conversely, lack of chronic innate inflammation at baseline may predict improved treatment outcome, as suggested by good overall prognosis.

T-cell subsets in peripheral blood and tumors of patients treated with oncolytic adenoviruses.

The quality of the antitumor immune response is decisive when developing new immunotherapies for cancer. Oncolytic adenoviruses cause a potent immunogenic stimulus and arming them with costimulatory molecules reshapes the immune response further. We evaluated peripheral blood T-cell subsets of 50 patients with refractory solid tumors undergoing treatment with oncolytic adenovirus. These data were compared to changes in antiviral and antitumor T cells, treatment efficacy, overall survival, and T-cell subsets in pre- and post-treatment tumor biopsies. Treatment caused a significant (P < 0.0001) shift in T-cell subsets in blood, characterized by a proportional increase of CD8(+) cells, and decrease of CD4(+) cells. Concomitant treatment with cyclophosphamide and temozolomide resulted in less CD4(+) decrease (P = 0.041) than cyclophosphamide only. Interestingly, we saw a correlation between T-cell changes in peripheral blood and the tumor site. This correlation was positive for CD8(+) and inverse for CD4(+) cells. These findings give insight to the interconnections between peripheral blood and tumor-infiltrating lymphocyte (TIL) populations regarding oncolytic virotherapy. In particular, our data suggest that induction of T-cell response is not sufficient for clinical response in the context of immunosuppressive tumors, and that peripheral blood T cells have a complicated and potentially misleading relationship with TILs.

Cell Painting-based bioactivity prediction boosts high-throughput screening hit-rates and compound diversity.

Identifying active compounds for a target is a time- and resource-intensive task in early drug discovery. Accurate bioactivity prediction using morphological profiles could streamline the process, enabling smaller, more focused compound screens. We investigate the potential of deep learning on unrefined single-concentration activity readouts and Cell Painting data, to predict compound activity across 140 diverse assays. We observe an average ROC-AUC of 0.744 ± 0.108 with 62% of assays achieving ≥0.7, 30% ≥0.8, and 7% ≥0.9. In many cases, the high prediction performance can be achieved using only brightfield images instead of multichannel fluorescence images. A comprehensive analysis shows that Cell Painting-based bioactivity prediction is robust across assay types, technologies, and target classes, with cell-based assays and kinase targets being particularly well-suited for prediction. Experimental validation confirms the enrichment of active compounds. Our findings indicate that models trained on Cell Painting data, combined with a small set of single-concentration data points, can reliably predict the activity of a compound library across diverse targets and assays while maintaining high hit rates and scaffold diversity. This approach has the potential to reduce the size of screening campaigns, saving time and resources, and enabling primary screening with more complex assays.

Label-free prediction of cell painting from brightfield images.

Cell Painting is a high-content image-based assay applied in drug discovery to predict bioactivity, assess toxicity and understand mechanisms of action of chemical and genetic perturbations. We investigate label-free Cell Painting by predicting the five fluorescent Cell Painting channels from brightfield input. We train and validate two deep learning models with a dataset representing 17 batches, and we evaluate on batches treated with compounds from a phenotypic set. The mean Pearson correlation coefficient of the predicted images across all channels is 0.84. Without incorporating features into the model training, we achieved a mean correlation of 0.45 with ground truth features extracted using a segmentation-based feature extraction pipeline. Additionally, we identified 30 features which correlated greater than 0.8 to the ground truth. Toxicity analysis on the label-free Cell Painting resulted a sensitivity of 62.5% and specificity of 99.3% on images from unseen batches. We provide a breakdown of the feature profiles by channel and feature type to understand the potential and limitations of label-free morphological profiling. We demonstrate that label-free Cell Painting has the potential to be used for downstream analyses and could allow for repurposing imaging channels for other non-generic fluorescent stains of more targeted biological interest.

Stromal FAP Expression is Associated with MRI Visibility and Patient Survival in Prostate Cancer.

Some clinically significant prostate cancers are missed by MRI. We asked whether the tumor stroma in surgically treated localized prostate cancer lesions positive or negative with MRI are different in their cellular and molecular properties, and whether the differences are reflected to the clinical course of the disease. We profiled the stromal and immune cell composition of MRI-classified tumor lesions by applying multiplexed fluorescence IHC (mfIHC) and automated image analysis in a clinical cohort of 343 patients (cohort I). We compared stromal variables between MRI-visible lesions, invisible lesions, and benign tissue and assessed the predictive significance for biochemical recurrence (BCR) and disease-specific survival (DSS) using Cox regression and log-rank analysis. Subsequently, we carried out a prognostic validation of the identified biomarkers in a population-based cohort of 319 patients (cohort II). MRI true-positive lesions are different from benign tissue and MRI false-negative lesions in their stromal composition. CD163+ cells (macrophages) and fibroblast activation protein (FAP)+ cells were more abundant in MRI true-positive than in MRI false-negative lesions or benign areas. In MRI true-visible lesions, a high proportion of stromal FAP+ cells was associated with PTEN status and increased immune infiltration (CD8+, CD163+), and predicted elevated risk for BCR. High FAP phenotype was confirmed to be a strong indicator of poor prognosis in two independent patient cohorts using also conventional IHC. The molecular composition of the tumor stroma may determine whether early prostate lesions are detectable by MRI and associates with survival after surgical treatment. Significance: These findings may have a significant impact on clinical decision making as more radical treatments may be recommended for men with a combination of MRI-visible primary tumors and FAP+ tumor stroma.

Cell Morphological Profiling Enables High-Throughput Screening for PROteolysis TArgeting Chimera (PROTAC) Phenotypic Signature.

PROteolysis TArgeting Chimeras (PROTACs) use the ubiquitin-proteasome system to degrade a protein of interest for therapeutic benefit. Advances made in targeted protein degradation technology have been remarkable, with several molecules having moved into clinical studies. However, robust routes to assess and better understand the safety risks of PROTACs need to be identified, which is an essential step toward delivering efficacious and safe compounds to patients. In this work, we used Cell Painting, an unbiased high-content imaging method, to identify phenotypic signatures of PROTACs. Chemical clustering and model prediction allowed the identification of a mitotoxicity signature that could not be expected by screening the individual PROTAC components. The data highlighted the benefit of unbiased phenotypic methods for identifying toxic signatures and the potential to impact drug design.

Immune profiles in acute myeloid leukemia bone marrow associate with patient age, T-cell receptor clonality, and survival.

The immunologic microenvironment in various solid tumors is aberrant and correlates with clinical survival. Here, we present a comprehensive analysis of the immune environment of acute myeloid leukemia (AML) bone marrow (BM) at diagnosis. We compared the immunologic landscape of formalin-fixed paraffin-embedded BM trephine samples from AML (n = 69), chronic myeloid leukemia (CML; n = 56), and B-cell acute lymphoblastic leukemia (B-ALL) patients (n = 52) at diagnosis to controls (n = 12) with 30 immunophenotype markers using multiplex immunohistochemistry and computerized image analysis. We identified distinct immunologic profiles specific for leukemia subtypes and controls enabling accurate classification of AML (area under the curve [AUC] = 1.0), CML (AUC = 0.99), B-ALL (AUC = 0.96), and control subjects (AUC = 1.0). Interestingly, 2 major immunologic AML clusters differing in age, T-cell receptor clonality, and survival were discovered. A low proportion of regulatory T cells and pSTAT1+cMAF- monocytes were identified as novel biomarkers of superior event-free survival in intensively treated AML patients. Moreover, we demonstrated that AML BM and peripheral blood samples are dissimilar in terms of immune cell phenotypes. To conclude, our study shows that the immunologic landscape considerably varies by leukemia subtype suggesting disease-specific immunoregulation. Furthermore, the association of the AML immune microenvironment with clinical parameters suggests a rationale for including immunologic parameters to improve disease classification or even patient risk stratification.

CDX2 Loss With Microsatellite Stable Phenotype Predicts Poor Clinical Outcome in Stage II Colorectal Carcinoma.

Current risk factors in stage II colorectal carcinoma are insufficient to guide treatment decisions. Loss of CDX2 has been shown to associate with poor clinical outcome and predict benefit for adjuvant chemotherapy in stage II and III colorectal carcinoma. The prognostic relevance of CDX2 in stage II disease has not been sufficiently validated, especially in relation to clinical risk factors, such as microsatellite instability (MSI) status, BRAF mutation status, and tumor budding. In this study, we evaluated the protein expression of CDX2 in tumor center and front areas in a tissue microarrays material of stage II colorectal carcinoma patients (n=232). CDX2 expression showed a partial or total loss in respective areas in 8.6% and 10.9% of patient cases. Patients with loss of CDX2 had shorter disease-specific survival when scored independently either in tumor center or tumor front areas (log rank P=0.012; P=0.012). Loss of CDX2 predicted survival independently of other stage II risk factors, such as MSI status and BRAF mutation status, pT class, and tumor budding (hazard ratio=5.96, 95% confidence interval=1.55-22.95; hazard ratio=3.70, 95% confidence interval=1.30-10.56). Importantly, CDX2 loss predicted inferior survival only in patients with microsatellite stable, but not with MSI-high phenotype. Interestingly, CDX2 loss associated with low E-cadherin expression, tight junction disruption, and high expression of ezrin protein. The work demonstrates that loss of CDX2 is an independent risk factor of poor disease-specific survival in stage II colorectal carcinoma. Furthermore, the study suggests that CDX2 loss is linked with epithelial-to-mesenchymal transition independently of tumor budding.

Combined epithelial marker analysis of tumour budding in stage II colorectal cancer.

Tumour budding predicts survival of stage II colorectal cancer (CRC) and has been suggested to be associated with epithelial-to-mesenchymal transition (EMT). However, the underlying molecular changes of tumour budding remain poorly understood. Here, we performed multiplex immunohistochemistry (mIHC) to phenotypically profile tumours using known EMT-associated markers: E-cadherin (adherence junctions), integrin ß4 (ITGB4; basement membrane), ZO-1 (tight junctions), and pan-cytokeratin. A subpopulation of patients showed high ITGB4 expression in tumour buds, and this coincided with a switch of ITGB4 localisation from the basal membrane of intact epithelium to the cytoplasm of budding cells. Digital image analysis demonstrated that tumour budding with high ITGB4 expression in tissue microarray (TMA) cores correlated with tumour budding assessed from haematoxylin and eosin (H&E) whole sections and independently predicted poor disease-specific survival in two independent stage II CRC cohorts (hazard ratio [HR] = 4.50 (95% confidence interval [CI] = 1.50-13.5), n = 232; HR = 3.52 (95% CI = 1.30-9.53), n = 72). Furthermore, digitally obtained ITGB4-high bud count in random TMA cores was better associated with survival outcome than visual tumour bud count in corresponding H&E-stained samples. In summary, the mIHC-based phenotypic profiling of human tumour tissue shows strong potential for the molecular characterisation of tumour biology and for the discovery of novel prognostic biomarkers.

Immune cell constitution in bone marrow microenvironment predicts outcome in adult ALL.

As novel immunological treatments are gaining a foothold in the treatment of acute lymphoblastic leukemia (ALL), it is elemental to examine ALL immunobiology in more detail. We used multiplexed immunohistochemistry (mIHC) to study the immune contexture in adult precursor B cell ALL bone marrow (BM). In addition, we developed a multivariate risk prediction model that stratified a poor survival group based on clinical parameters and mIHC data. We analyzed BM biopsy samples of ALL patients (n = 52) and healthy controls (n = 14) using mIHC with 30 different immunophenotype markers and computerized image analysis. In ALL BM, the proportions of M1-like macrophages, granzyme B+CD57+CD8+ T cells, and CD27+ T cells were decreased, whereas the proportions of myeloid-derived suppressor cells and M2-like macrophages were increased. Also, the expression of checkpoint molecules PD1 and CTLA4 was elevated. In the multivariate model, age, platelet count, and the proportion of PD1+TIM3+ double-positive CD4+ T cells differentiated a poor survival group. These results were validated by flow cytometry in a separate cohort (n = 31). In conclusion, the immune cell contexture in ALL BM differs from healthy controls. CD4+PD1+TIM3+ T cells were independent predictors of poor outcome in our multivariate risk model, suggesting that PD1 might serve as an attractive immuno-oncological target in B-ALL.

Immune cell contexture in the bone marrow tumor microenvironment impacts therapy response in CML.

Increasing evidence suggests that the immune system affects prognosis of chronic myeloid leukemia (CML), but the detailed immunological composition of the leukemia bone marrow (BM) microenvironment is unknown. We aimed to characterize the immune landscape of the CML BM and predict the current treatment goal of tyrosine kinase inhibitor (TKI) therapy, molecular remission 4.0 (MR4.0). Using multiplex immunohistochemistry (mIHC) and automated image analysis, we studied BM tissues of CML patients (n = 56) and controls (n = 14) with a total of 30 immunophenotype markers essential in cancer immunology. CML patients' CD4+ and CD8+ T-cells expressed higher levels of putative exhaustion markers PD1, TIM3, and CTLA4 when compared to control. PD1 expression was higher in BM compared to paired peripheral blood (PB) samples, and decreased during TKI therapy. By combining clinical parameters and immune profiles, low CD4+ T-cell proportion, high proportion of PD1+TIM3-CD8+ T cells, and high PB neutrophil count were most predictive of lower MR4.0 likelihood. Low CD4+ T-cell proportion and high PB neutrophil counts predicted MR4.0 also in a validation cohort (n = 52) analyzed with flow cytometry. In summary, the CML BM is characterized by immune suppression and immune biomarkers predicted MR4.0, thus warranting further testing of immunomodulatory drugs in CML treatment.

Deep learning based tissue analysis predicts outcome in colorectal cancer.

Image-based machine learning and deep learning in particular has recently shown expert-level accuracy in medical image classification. In this study, we combine convolutional and recurrent architectures to train a deep network to predict colorectal cancer outcome based on images of tumour tissue samples. The novelty of our approach is that we directly predict patient outcome, without any intermediate tissue classification. We evaluate a set of digitized haematoxylin-eosin-stained tumour tissue microarray (TMA) samples from 420 colorectal cancer patients with clinicopathological and outcome data available. The results show that deep learning-based outcome prediction with only small tissue areas as input outperforms (hazard ratio 2.3; CI 95% 1.79-3.03; AUC 0.69) visual histological assessment performed by human experts on both TMA spot (HR 1.67; CI 95% 1.28-2.19; AUC 0.58) and whole-slide level (HR 1.65; CI 95% 1.30-2.15; AUC 0.57) in the stratification into low- and high-risk patients. Our results suggest that state-of-the-art deep learning techniques can extract more prognostic information from the tissue morphology of colorectal cancer than an experienced human observer.

ITGB1-dependent upregulation of Caveolin-1 switches TGFß signalling from tumour-suppressive to oncogenic in prostate cancer.

Caveolin-1 (CAV1) is over-expressed in prostate cancer (PCa) and is associated with adverse prognosis, but the molecular mechanisms linking CAV1 expression to disease progression are poorly understood. Extensive gene expression correlation analysis, quantitative multiplex imaging of clinical samples, and analysis of the CAV1-dependent transcriptome, supported that CAV1 re-programmes TGFß signalling from tumour suppressive to oncogenic (i.e. induction of SLUG, PAI-1 and suppression of CDH1, DSP, CDKN1A). Supporting such a role, CAV1 knockdown led to growth arrest and inhibition of cell invasion in prostate cancer cell lines. Rationalized RNAi screening and high-content microscopy in search for CAV1 upstream regulators revealed integrin beta1 (ITGB1) and integrin associated proteins as CAV1 regulators. Our work suggests TGFß signalling and beta1 integrins as potential therapeutic targets in PCa over-expressing CAV1, and contributes to better understand the paradoxical dual role of TGFß in tumour biology.

Systems pathology by multiplexed immunohistochemistry and whole-slide digital image analysis.

The paradigm of molecular histopathology is shifting from a single-marker immunohistochemistry towards multiplexed detection of markers to better understand the complex pathological processes. However, there are no systems allowing multiplexed IHC (mIHC) with high-resolution whole-slide tissue imaging and analysis, yet providing feasible throughput for routine use. We present an mIHC platform combining fluorescent and chromogenic staining with automated whole-slide imaging and integrated whole-slide image analysis, enabling simultaneous detection of six protein markers and nuclei, and automatic quantification and classification of hundreds of thousands of cells in situ in formalin-fixed paraffin-embedded tissues. In the first proof-of-concept, we detected immune cells at cell-level resolution (n = 128,894 cells) in human prostate cancer, and analysed T cell subpopulations in different tumour compartments (epithelium vs. stroma). In the second proof-of-concept, we demonstrated an automatic classification of epithelial cell populations (n = 83,558) and glands (benign vs. cancer) in prostate cancer with simultaneous analysis of androgen receptor (AR) and alpha-methylacyl-CoA (AMACR) expression at cell-level resolution. We conclude that the open-source combination of 8-plex mIHC detection, whole-slide image acquisition and analysis provides a robust tool allowing quantitative, spatially resolved whole-slide tissue cytometry directly in formalin-fixed human tumour tissues for improved characterization of histology and the tumour microenvironment.

Antibody-supervised deep learning for quantification of tumor-infiltrating immune cells in hematoxylin and eosin stained breast cancer samples.

BACKGROUND: Immune cell infiltration in tumor is an emerging prognostic biomarker in breast cancer. The gold standard for quantification of immune cells in tissue sections is visual assessment through a microscope, which is subjective and semi-quantitative. In this study, we propose and evaluate an approach based on antibody-guided annotation and deep learning to quantify immune cell-rich areas in hematoxylin and eosin (H&E) stained samples. METHODS: Consecutive sections of formalin-fixed parafin-embedded samples obtained from the primary tumor of twenty breast cancer patients were cut and stained with H&E and the pan-leukocyte CD45 antibody. The stained slides were digitally scanned, and a training set of immune cell-rich and cell-poor tissue regions was annotated in H&E whole-slide images using the CD45-expression as a guide. In analysis, the images were divided into small homogenous regions, superpixels, from which features were extracted using a pretrained convolutional neural network (CNN) and classified with a support of vector machine. The CNN approach was compared to texture-based classification and to visual assessments performed by two pathologists. RESULTS: In a set of 123,442 labeled superpixels, the CNN approach achieved an F-score of 0.94 (range: 0.92-0.94) in discrimination of immune cell-rich and cell-poor regions, as compared to an F-score of 0.88 (range: 0.87-0.89) obtained with the texture-based classification. When compared to visual assessment of 200 images, an agreement of 90% (κ = 0.79) to quantify immune infiltration with the CNN approach was achieved while the inter-observer agreement between pathologists was 90% (κ = 0.78). CONCLUSIONS: Our findings indicate that deep learning can be applied to quantify immune cell infiltration in breast cancer samples using a basic morphology staining only. A good discrimination of immune cell-rich areas was achieved, well in concordance with both leukocyte antigen expression and pathologists' visual assessment.