Artificial intelligence to detect MYC translocation in slides of diffuse large B-cell lymphoma
- PDF / 1,231,218 Bytes
- 5 Pages / 595.276 x 790.866 pts Page_size
- 18 Downloads / 198 Views
BRIEF REPORT
Artificial intelligence to detect MYC translocation in slides of diffuse large B-cell lymphoma Zaneta Swiderska-Chadaj 1,2
&
Konnie M. Hebeda 1 & Michiel van den Brand 1,3 & Geert Litjens 1
Received: 11 May 2020 / Revised: 24 August 2020 / Accepted: 15 September 2020 # The Author(s) 2020
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. Keywords B-cell lymphoma . MYC . H&E . DLBCL . Deep Learning
Introduction Diffuse large B-cell lymphoma (DLBCL) is the most common type of B-cell lymphoma and includes a diversity of not yet completely characterized clinico-pathological lymphomas [1]. A subgroup of 5–15% of DLBCL shows MYC oncogene rearrangement. Especially when combined with a BCL2 and/or BCL6 rearrangement (high-grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangement), these DLBCL have a poor outcome when treated with standard R-CHOP chemotherapy [2], and may require a different treatment [3]. Recent Key points • The prediction of MYC translocation based on morphology is possible. This article is part of the Topical Collection on Quality in Pathology
studies on molecularly defined DLBCL subgroups confirmed the poor prognosis of a MYC rearrangement [4]. Therefore, currently a genetic test for MYC rearrangement and, if positive, for BCL2 and BCL6, is required for DLBCL patients for diagnosis, prognosis, and to guide therapy. Many pathological classifications are based on the fact that genetic changes in a tumor are reflected in aberrant transcription, changed protein expression, and often a characteristic morphology of the tumor cells or the tumor microenvironment. Several morphologic variants of DLBCL are recognized, but clinical relevance is not yet established [1]. We hypothesized that a trained computer algorithm will be able to predict MYC rearrangement from morphology on a standard hematoxylin and eosin (H&E)-stained slide of a DLBCL, thereby obviating molecular tests in the majority of cases that will be predicted to lack a MYC translocation.
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00428-020-02931-4) contains supplementary material, which is available to authorized users.
Methods * Zaneta Swiderska-Chadaj [email protected] 1
Department of Pathology, Radboud University Medical Center, Geert Grooteplein 10, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands
2
Faculty o
Data Loading...
