Refine
Year of publication
Document Type
- Article (23)
- Doctoral Thesis (1)
- Preprint (1)
Has Fulltext
- yes (25)
Is part of the Bibliography
- no (25)
Keywords
- Hodgkin lymphoma (2)
- Machine learning (2)
- Prostate cancer (2)
- classical Hodgkin lymphoma (2)
- gene expression (2)
- Affymetrix (1)
- Artificial intelligence (1)
- B-cell immunology (1)
- B-cell lymphoma (1)
- BATF3 (1)
Institute
Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) is an indolent lymphoma, but can transform into diffuse large B cell lymphoma (DLBCL), showing a more aggressive clinical behavior. Little is known about these cases on the molecular level. Therefore, the aim of the present study was to characterize DLBCL transformed from NLPHL (LP-DLBCL) by gene expression profiling (GEP). GEP revealed an inflammatory signature pinpointing to a specific host response. In a coculture model resembling this host response, DEV tumor cells showed an impaired growth behavior. Mechanisms involved in the reduced tumor cell proliferation included a downregulation of MYC and its target genes. Lack of MYC expression was also confirmed in 12/16 LP-DLBCL by immunohistochemistry. Furthermore, CD274/PD-L1 was upregulated in DEV tumor cells after coculture with T cells or monocytes and its expression was validated in 12/19 cases of LP-DLBCL. Thereby, our data provide new insights into the pathogenesis of LP-DLBCL and an explanation for the relatively low tumor cell content. Moreover, the findings suggest that treatment of these patients with immune checkpoint inhibitors may enhance an already ongoing host response in these patients.
In pathology, tissue images are evaluated using a light microscope, relying on the expertise and experience of pathologists. There is a great need for computational methods to quantify and standardize histological observations. Computational quantification methods become more and more essential to evaluate tissue images. In particular, the distribution of tumor cells and their microenvironment are of special interest. Here, we systematically investigated tumor cell properties and their spatial neighborhood relations by a new application of statistical analysis to whole slide images of Hodgkin lymphoma, a tumor arising in lymph nodes, and inflammation of lymph nodes called lymphadenitis. We considered properties of more than 400, 000 immunohistochemically stained, CD30-positive cells in 35 whole slide images of tissue sections from subtypes of the classical Hodgkin lymphoma, nodular sclerosis and mixed cellularity, as well as from lymphadenitis. We found that cells of specific morphology exhibited significant favored and unfavored spatial neighborhood relations of cells in dependence of their morphology. This information is important to evaluate differences between Hodgkin lymph nodes infiltrated by tumor cells (Hodgkin lymphoma) and inflamed lymph nodes, concerning the neighborhood relations of cells and the sizes of cells. The quantification of neighborhood relations revealed new insights of relations of CD30-positive cells in different diagnosis cases. The approach is general and can easily be applied to whole slide image analysis of other tumor types.
Bioinformatics analysis quantifies neighborhood preferences of cancer cells in Hodgkin lymphoma
(2017)
Motivation Hodgkin lymphoma is a tumor of the lymphatic system and represents one of the most frequent lymphoma in the Western world. It is characterized by Hodgkin cells and Reed-Sternberg cells, which exhibit a broad morphological spectrum. The cells are visualized by immunohistochemical staining of tissue sections. In pathology, tissue images are mainly manually evaluated, relying on the expertise and experience of pathologists. Computational quantification methods become more and more essential to evaluate tissue images. In particular, the distribution of cancer cells is of great interest.
Results Here, we systematically quantified and investigated cancer cell properties and their spatial neighborhood relations by applying statistical analyses to whole slide images of Hodgkin lymphoma and lymphadenitis, which describes a non-cancerous inflammation of the lymph node. We differentiated cells by their morphology and studied the spatial neighborhood relation of more than 400,000 immunohistochemically stained cells. We found that, according to their morphological features, the cells exhibited significant preferences for and aversions to cells of specific profiles as nearest neighbor. We quantified differences between Hodgkin lymphoma and lymphadenitis concerning the neighborhood relations of cells and the sizes of cells. The approach can easily be applied to other cancer types.
Introduction and Objective: Identifying patients that benefit from cisplatin-based adjuvant chemotherapy is a major issue in the management of muscle-invasive bladder cancer (MIBC). The purpose of this study is to correlate “luminal” and “basal” type protein expression with histological subtypes, to investigate the prognostic impact on survival after adjuvant chemotherapy and to define molecular consensus subtypes of “double negative” patients (i.e., without expression of CK5/6 or GATA3).
Materials and Methods: We performed immunohistochemical (IHC) analysis of CK5/6 and GATA3 for surrogate molecular subtyping in 181 MIBC samples. The mRNA expression profiles for molecular consensus classification were determined in CK5/6 and GATA3 (double) negative cases using a transcriptome panel with 19.398 mRNA targets (HTG Molecular Diagnostics). Data of 110 patients undergoing radical cystectomy were available for survival analysis.
Results: The expression of CK5/6 correlated with squamous histological subtype (96%) and expression of GATA3 was associated with micropapillary histology (100%). In the multivariate Cox-regression model, patients receiving adjuvant chemotherapy had a significant survival benefit (hazard ratio [HR]: 0.19 95% confidence interval [CI]: 0.1–0.4, p < 0.001) and double-negative cases had decreased OS (HR: 4.07; 95% CI: 1.5–10.9, p = 0.005). Double negative cases were classified as NE-like (30%), stroma-rich (30%), and Ba/Sq (40%) consensus molecular subtypes and displaying different histological subtypes.
Objectives: To analyze the performance of radiological assessment categories and quantitative computational analysis of apparent diffusion coefficient (ADC) maps using variant machine learning algorithms to differentiate clinically significant versus insignificant prostate cancer (PCa). Methods: Retrospectively, 73 patients were included in the study. The patients (mean age, 66.3 ± 7.6 years) were examined with multiparametric MRI (mpMRI) prior to radical prostatectomy (n = 33) or targeted biopsy (n = 40). The index lesion was annotated in MRI ADC and the equivalent histologic slides according to the highest Gleason Grade Group (GrG). Volumes of interest (VOIs) were determined for each lesion and normal-appearing peripheral zone. VOIs were processed by radiomic analysis. For the classification of lesions according to their clinical significance (GrG ≥ 3), principal component (PC) analysis, univariate analysis (UA) with consecutive support vector machines, neural networks, and random forest analysis were performed. Results: PC analysis discriminated between benign and malignant prostate tissue. PC evaluation yielded no stratification of PCa lesions according to their clinical significance, but UA revealed differences in clinical assessment categories and radiomic features. We trained three classification models with fifteen feature subsets. We identified a subset of shape features which improved the diagnostic accuracy of the clinical assessment categories (maximum increase in diagnostic accuracy ΔAUC = + 0.05, p < 0.001) while also identifying combinations of features and models which reduced overall accuracy. Conclusions: The impact of radiomic features to differentiate PCa lesions according to their clinical significance remains controversial. It depends on feature selection and the employed machine learning algorithms. It can result in improvement or reduction of diagnostic performance.
Distinct immune patterns of hepatocellular carcinoma (HCC) may have prognostic implications in the response to transarterial chemoembolization (TACE). Thus, we aimed to exploratively analyze tumor tissue of HCC patients who do or do not respond to TACE, and to identify novel prognostic biomarkers predictive of response to TACE. We retrospectively included 15 HCC patients who had three consecutive TACE between January 2019 and November 2019. Eight patients had a response while seven patients had no response to TACE. All patients had measurable disease according to mRECIST. Corresponding tumor tissue samples were processed for differential expression profiling using NanoString nCounter® PanCancer immune profiling panel. Immune-related pathways were broadly upregulated in TACE responders. The top differentially regulated genes were the upregulated CXCL1 (log2fc 4.98, Benjamini–Hochberg (BH)-p < 0.001), CXCL6 (log2fc 4.43, BH-p = 0.016) and the downregulated MME (log2fc −4.33, BH-p 0.001). CD8/T-regs was highly increased in responders, whereas the relative number of T-regs to tumor-infiltrating lymphocytes (TIL) was highly decreased. We preliminary identified CXCL1 and CXCL6 as candidate genes that might have the potential to serve as therapeutically relevant biomarkers in HCC patients. This might pave the way to improve patient selection for TACE in HCC patients beyond expert consensus.
In gastric cancer (GC), there are four molecular subclasses that indicate whether patients respond to chemotherapy or immunotherapy, according to the TCGA. In clinical practice, however, not every patient undergoes molecular testing. Many laboratories have used well-implemented in situ techniques (IHC and EBER-ISH) to determine the subclasses in their cohorts. Although multiple stains are used, we show that a staining approach is unable to correctly discriminate all subclasses. As an alternative, we trained an ensemble convolutional neuronal network using bagging that can predict the molecular subclass directly from hematoxylin–eosin histology. We also identified patients with predicted intra-tumoral heterogeneity or with features from multiple subclasses, which challenges the postulated TCGA-based decision tree for GC subtyping. In the future, deep learning may enable targeted testing for molecular subtypes and targeted therapy for a broader group of GC patients. © 2022 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.
Background: Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) usually presents in middle aged men and shows an indolent clinical behavior. However, up to 30% of the patients present a secondary transformation into aggressive diffuse large B cell lymphoma (DLBCL). The aim of the present study was to characterize morphology and immunophenotype of this kind of DLBCL in detail and compare it with conventional DLBCL.
Methods: Morphology and immunophenotype of 33 cases of NLPHL with simultaneous or sequential transformation into DLBCL were investigated. These cases were compared with 41 de novo DLBCL in Finnish men.
Results: The majority of cases exhibited different immunophenotypes in the NLPHL and the DLBCL components. The immunophenotype of the DLBCL secondary to NLPHL was heterogeneous. However, BCL6, EMA, CD75 and J-chain were usually expressed in both components (≥73% positive). Overall, the NLPHL component was more frequently positive for EMA, CD75 and J-chain than the DLBCL component. In contrast, B cell markers, CD10 and BCL2, were more frequently expressed and were expressed at higher levels in the DLBCL component than in the NLPHL component. In the independent series of de novo DLBCL 4 cases could be identified with a growth pattern and immunophenotype that suggested that they had arisen secondarily from NLPHL.
Conclusions: The morphology and immunophenotype of DLBCL arisen from NLPHL is heterogeneous. Further characterization of the particular molecular features of this subgroup is warranted to be able to better identify these cases among conventional DLBCL.
The mechanisms involved in malignant transformation of mature B and T lymphocytes are still poorly understood. In a previous study, we compared gene expression profiles of the tumor cells of Hodgkin lymphoma (HL) and anaplastic large cell lymphoma (ALCL) to their normal cellular counterparts and found the basic leucine zipper protein ATF-like 3 (BATF3) to be significantly upregulated in the tumor cells of both entities. To assess the oncogenic potential of BATF3 in lymphomagenesis and to dissect the molecular interactions of BATF3 in lymphoma cells, we retrovirally transduced murine mature T and B cells with a BATF3-encoding viral vector and transplanted each population into Rag1-deficient recipients. Intriguingly, BATF3-expressing B lymphocytes readily induced B-cell lymphomas after characteristic latencies, whereas T-cell transplanted animals remained healthy throughout the observation time. Further analyses revealed a germinal center B-cell-like phenotype of most BATF3-initiated lymphomas. In a multiple myeloma cell line, BATF3 inhibited BLIMP1 expression, potentially illuminating an oncogenic action of BATF3 in B-cell lymphomagenesis. In conclusion, BATF3 overexpression induces malignant transformation of mature B cells and might serve as a potential target in B-cell lymphoma treatment.
Background: Prostate cancer is a major health concern in aging men. Paralleling an aging society, prostate cancer prevalence increases emphasizing the need for efcient diagnostic algorithms.
Methods: Retrospectively, 106 prostate tissue samples from 48 patients (mean age,
66 ± 6.6 years) were included in the study. Patients sufered from prostate cancer (n = 38) or benign prostatic hyperplasia (n = 10) and were treated with radical prostatectomy or Holmium laser enucleation of the prostate, respectively. We constructed tissue microarrays (TMAs) comprising representative malignant (n = 38) and benign (n = 68) tissue cores. TMAs were processed to histological slides, stained, digitized and assessed for the applicability of machine learning strategies and open–source tools in diagnosis of prostate cancer. We applied the software QuPath to extract features for shape, stain intensity, and texture of TMA cores for three stainings, H&E, ERG, and PIN-4. Three machine learning algorithms, neural network (NN), support vector machines (SVM), and random forest (RF), were trained and cross-validated with 100 Monte Carlo random splits into 70% training set and 30% test set. We determined AUC values for single color channels, with and without optimization of hyperparameters by exhaustive grid search. We applied recursive feature elimination to feature sets of multiple color transforms.
Results: Mean AUC was above 0.80. PIN-4 stainings yielded higher AUC than H&E and
ERG. For PIN-4 with the color transform saturation, NN, RF, and SVM revealed AUC of 0.93 ± 0.04, 0.91 ± 0.06, and 0.92 ± 0.05, respectively. Optimization of hyperparameters improved the AUC only slightly by 0.01. For H&E, feature selection resulted in no increase of AUC but to an increase of 0.02–0.06 for ERG and PIN-4.
Conclusions: Automated pipelines may be able to discriminate with high accuracy between malignant and benign tissue. We found PIN-4 staining best suited for classifcation. Further bioinformatic analysis of larger data sets would be crucial to evaluate the reliability of automated classifcation methods for clinical practice and to evaluate potential discrimination of aggressiveness of cancer to pave the way to automatic precision medicine.