Oliver Lester Saldanha1,2, Hannah Sophie Muti1,2, Heike I Grabsch3,4, Rupert Langer5,6, Bastian Dislich5, Meike Kohlruss7, Gisela Keller7, Marko van Treeck1,2, Katherine Jane Hewitt1,2, Fiona R Kolbinger2,8, Gregory Patrick Veldhuizen1,2, Peter Boor9,10, Sebastian Foersch11, Daniel Truhn12, Jakob Nikolas Kather13,14,15,16,17. 1. Department of Medicine III, University Hospital RWTH Aachen, Aachen, Germany. 2. Else Kroener Fresenius Center for Digital Health, Medical Faculty Carl Gustav Carus, Technical University Dresden, Fetscherstrasse 74, 01307, Dresden, Germany. 3. Pathology and GROW School for Oncology and Developmental Biology, Maastricht University Medical Center+, Maastricht, The Netherlands. 4. Pathology and Data Analytics, Leeds Institute of Medical Research at St James's, University of Leeds, Leeds, UK. 5. Institute of Pathology, Inselspital, University of Bern, Bern, Switzerland. 6. Institute of Pathology and Molecular Pathology, Kepler University Hospital, Johannes Kepler University Linz, Linz, Austria. 7. Institute of Pathology, TUM School of Medicine, Technical University of Munich, Munich, Germany. 8. Department of Visceral, Thoracic and Vascular Surgery, University Hospital and Faculty of Medicine Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany. 9. Institute of Pathology, University Hospital RWTH Aachen, 52074, Aachen, Germany. 10. Department of Nephrology and Immunology, University Hospital RWTH Aachen, 52074, Aachen, Germany. 11. Institute of Pathology, University Medical Center Mainz, Mainz, Germany. 12. Department of Diagnostic and Interventional Radiology, University Hospital RWTH Aachen, Aachen, Germany. 13. Department of Medicine III, University Hospital RWTH Aachen, Aachen, Germany. jakob-nikolas.kather@alumni.dkfz.de. 14. Else Kroener Fresenius Center for Digital Health, Medical Faculty Carl Gustav Carus, Technical University Dresden, Fetscherstrasse 74, 01307, Dresden, Germany. jakob-nikolas.kather@alumni.dkfz.de. 15. Pathology and Data Analytics, Leeds Institute of Medical Research at St James's, University of Leeds, Leeds, UK. jakob-nikolas.kather@alumni.dkfz.de. 16. Medical Oncology, National Center for Tumor Diseases (NCT), University Hospital Heidelberg, Heidelberg, Germany. jakob-nikolas.kather@alumni.dkfz.de. 17. Department of Medicine 1, University Hospital and Faculty of Medicine Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany. jakob-nikolas.kather@alumni.dkfz.de.
Abstract
BACKGROUND: Computational pathology uses deep learning (DL) to extract biomarkers from routine pathology slides. Large multicentric datasets improve performance, but such datasets are scarce for gastric cancer. This limitation could be overcome by Swarm Learning (SL). METHODS: Here, we report the results of a multicentric retrospective study of SL for prediction of molecular biomarkers in gastric cancer. We collected tissue samples with known microsatellite instability (MSI) and Epstein-Barr Virus (EBV) status from four patient cohorts from Switzerland, Germany, the UK and the USA, storing each dataset on a physically separate computer. RESULTS: On an external validation cohort, the SL-based classifier reached an area under the receiver operating curve (AUROC) of 0.8092 (± 0.0132) for MSI prediction and 0.8372 (± 0.0179) for EBV prediction. The centralized model, which was trained on all datasets on a single computer, reached a similar performance. CONCLUSIONS: Our findings demonstrate the feasibility of SL-based molecular biomarkers in gastric cancer. In the future, SL could be used for collaborative training and, thus, improve the performance of these biomarkers. This may ultimately result in clinical-grade performance and generalizability.
BACKGROUND: Computational pathology uses deep learning (DL) to extract biomarkers from routine pathology slides. Large multicentric datasets improve performance, but such datasets are scarce for gastric cancer. This limitation could be overcome by Swarm Learning (SL). METHODS: Here, we report the results of a multicentric retrospective study of SL for prediction of molecular biomarkers in gastric cancer. We collected tissue samples with known microsatellite instability (MSI) and Epstein-Barr Virus (EBV) status from four patient cohorts from Switzerland, Germany, the UK and the USA, storing each dataset on a physically separate computer. RESULTS: On an external validation cohort, the SL-based classifier reached an area under the receiver operating curve (AUROC) of 0.8092 (± 0.0132) for MSI prediction and 0.8372 (± 0.0179) for EBV prediction. The centralized model, which was trained on all datasets on a single computer, reached a similar performance. CONCLUSIONS: Our findings demonstrate the feasibility of SL-based molecular biomarkers in gastric cancer. In the future, SL could be used for collaborative training and, thus, improve the performance of these biomarkers. This may ultimately result in clinical-grade performance and generalizability.
Authors: Oliver Lester Saldanha; Philip Quirke; Nicholas P West; Jacqueline A James; Maurice B Loughrey; Heike I Grabsch; Manuel Salto-Tellez; Elizabeth Alwers; Didem Cifci; Narmin Ghaffari Laleh; Tobias Seibel; Richard Gray; Gordon G A Hutchins; Hermann Brenner; Marko van Treeck; Tanwei Yuan; Titus J Brinker; Jenny Chang-Claude; Firas Khader; Andreas Schuppert; Tom Luedde; Christian Trautwein; Hannah Sophie Muti; Sebastian Foersch; Michael Hoffmeister; Daniel Truhn; Jakob Nikolas Kather Journal: Nat Med Date: 2022-04-25 Impact factor: 87.241
Authors: Alec J Kacew; Garth W Strohbehn; Loren Saulsberry; Neda Laiteerapong; Nicole A Cipriani; Jakob N Kather; Alexander T Pearson Journal: Front Oncol Date: 2021-06-08 Impact factor: 6.244