Dataset Information

Artificial intelligence to detect MYC translocation in slides of diffuse large B-cell lymphoma.

ABSTRACT: In patients with suspected lymphoma, the tissue biopsy provides lymphoma confirmation, classification, and prognostic factors, including genetic changes. We developed a deep learning algorithm to detect MYC rearrangement in scanned histological slides of diffuse large B-cell lymphoma. The H&E-stained slides of 287 cases from 11 hospitals were used for training and evaluation. The overall sensitivity to detect MYC rearrangement was 0.93 and the specificity 0.52, showing that prediction of MYC translocation based on morphology alone was possible in 93% of MYC-rearranged cases. This would allow a simple and fast prescreening, saving approximately 34% of genetic tests with the current algorithm.

SUBMITTER: Swiderska-Chadaj Z

PROVIDER: S-EPMC8448690 | biostudies-literature | 2021 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Artificial intelligence to detect MYC translocation in slides of diffuse large B-cell lymphoma.

Swiderska-Chadaj Zaneta Z Hebeda Konnie M KM van den Brand Michiel M Litjens Geert G

Virchows Archiv : an international journal of pathology 20200926 3

In patients with suspected lymphoma, the tissue biopsy provides lymphoma confirmation, classification, and prognostic factors, including genetic changes. We developed a deep learning algorithm to detect MYC rearrangement in scanned histological slides of diffuse large B-cell lymphoma. The H&E-stained slides of 287 cases from 11 hospitals were used for training and evaluation. The overall sensitivity to detect MYC rearrangement was 0.93 and the specificity 0.52, showing that prediction of MYC tra ...[more]

PMID: 32979109

Similar Datasets

Project description:Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous and prevalent subtype of aggressive non-Hodgkin lymphoma that poses diagnostic and prognostic challenges, particularly in predicting drug responsiveness. In this study, we used digital pathology and deep learning to predict responses to immunochemotherapy in patients with DLBCL. We retrospectively collected 251 slide images from 216 DLBCL patients treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), with their immunochemotherapy response labels. The digital pathology images were processed using contrastive learning for feature extraction. A multi-modal prediction model was developed by integrating clinical data and pathology image features. Knowledge distillation was employed to mitigate overfitting on gigapixel histopathology images to create a model that predicts responses based solely on pathology images. Based on the importance derived from the attention mechanism of the model, we extracted histological features that were considered key textures associated with drug responsiveness. The multi-modal prediction model achieved an impressive area under the ROC curve of 0.856, demonstrating significant associations with clinical variables such as Ann Arbor stage, International Prognostic Index, and bulky disease. Survival analyses indicated their effectiveness in predicting relapse-free survival. External validation using TCGA datasets supported the model's ability to predict survival differences. Additionally, pathology-based predictions show promise as independent prognostic indicators. Histopathological analysis identified centroblastic and immunoblastic features to be associated with treatment response, aligning with previous morphological classifications and highlighting the objectivity and reproducibility of artificial intelligence-based diagnosis. This study introduces a novel approach that combines digital pathology and clinical data to predict the response to immunochemotherapy in patients with DLBCL. This model shows great promise as a diagnostic and prognostic tool for clinical management of DLBCL. Further research and genomic data integration hold the potential to enhance its impact on clinical practice, ultimately improving patient outcomes.

Project description:BackgroundMYC is a heterogeneously expressed transcription factor that plays a multifunctional role in many biological processes such as cell proliferation and differentiation. It is also associated with many types of cancer including the malignant lymphomas. There are two types of aggressive B-cell lymphoma, namely Burkitt lymphoma (BL) and a subgroup of diffuse large cell lymphoma (DLBCL), which both carry MYC translocations and overexpress MYC but both differ significantly in their clinical outcome. In DLBCL, MYC translocations are associated with an aggressive behavior and poor outcome, whereas MYC-positive BL show a superior outcome.MethodsTo shed light on this phenomenon, we investigated the different modes of actions of MYC in aggressive B-cell lymphoma cell lines subdivided into three groups: (i) MYC-positive BL, (ii) DLBCL with MYC translocation (DLBCLpos) and (iii) DLBCL without MYC translocation (DLBCLneg) for control. In order to identify genome-wide MYC-DNA binding sites a chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-Seq) was performed. In addition, ChIP-Seq for H3K4me3 was used for determination of genomic regions accessible for transcriptional activity. These data were supplemented with gene expression data derived from RNA-Seq.ResultsBioinformatics integration of all data sets revealed different MYC-binding patterns and transcriptional profiles in MYC-positive BL and DLBCL cell lines indicating different functional roles of MYC for gene regulation in aggressive B-cell lymphomas. Based on this multi-omics analysis we identified ADGRE5 (alias CD97) - a member of the EGF-TM7 subfamily of adhesion G protein-coupled receptors - as a MYC target gene, which is specifically expressed in BL but not in DLBCL regardless of MYC translocation.ConclusionOur study describes a diverse genome-wide MYC-DNA binding pattern in BL and DLBCL cell lines with and without MYC translocations. Furthermore, we identified ADREG5 as a MYC target gene able to discriminate between BL and DLBCL irrespectively of the presence of MYC breaks in DLBCL. Since ADGRE5 plays an important role in tumor cell formation, metastasis and invasion, it might also be instrumental to better understand the different pathobiology of BL and DLBCL and help to explain discrepant clinical characteristics of BL and DLBCL.

Project description:Primary diffuse large B-cell lymphoma of the central nervous system (CNS DLBCL) is a rare, aggressive subtype of DLBCL, the biology of which is poorly understood. Recent studies have suggested a prognostic role of MYC protein expression in systemic DLBCL, but little is known about the frequency and significance of MYC protein expression in CNS DLBCL. Hence, we investigated MYC protein expression profiles of CNS DLBCL and assessed the relationship between MYC expression and a variety of histopathologic, immunophenotypic, genetic, and clinical features. Fifty-nine CNS DLBCL diagnosed at our institution over the past 13 years were evaluated. The majority of cases (80%) showed centroblastic morphology, and 12 (20%) displayed a perivascular pattern of infiltration. According to the Hans criteria, 41 (69%) cases had a non-germinal center B-cell and 18 (31%) had a germinal center B-cell cell-of-origin (COO) phenotype. Mean MYC protein expression was 50% (median: 50%, range: 10-80%). Forty-three cases (73%) showed MYC overexpression (≥ 40%), and 35 (60%) showed MYC/BCL2 coexpression. MYC overexpression was seen in the single case harboring MYC translocation and in the cases showing increased copies of MYC (27%); however, no significant difference in mean MYC expression was seen between groups harboring or lacking MYC aberrations. In our series, age was associated with a significantly increased risk of death, and the perivascular pattern of infiltration was associated with a significantly increased risk of disease progression. Neither MYC expression (with or without BCL2 coexpression) nor other variables, including COO subtype were predictive of clinical outcome. Our findings indicate that the proportion of CNS DLBCL overexpressing MYC is higher compared to systemic DLBCL, and MYC overexpression appears to be independent of genetic MYC abnormalities. Thus, MYC expression and other immunophenotypic markers used for prognostication of systemic DLBCL might not apply to CNS DLBCL due to differences in disease biology.

Dataset Information

Artificial intelligence to detect MYC translocation in slides of diffuse large B-cell lymphoma.

Publications

Artificial intelligence to detect MYC translocation in slides of diffuse large B-cell lymphoma.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets