Ole-Johan Skrede1, Sepp De Raedt1, Andreas Kleppe1, Tarjei S Hveem2, Knut Liestøl1, John Maddison2, Hanne A Askautrud2, Manohar Pradhan2, John Arne Nesheim2, Fritz Albregtsen1, Inger Nina Farstad3, Enric Domingo4, David N Church5, Arild Nesbakken6, Neil A Shepherd7, Ian Tomlinson8, Rachel Kerr4, Marco Novelli9, David J Kerr10, Håvard E Danielsen11. 1. Institute for Cancer Genetics and Informatics, Oslo University Hospital, Oslo, Norway; Department of Informatics, University of Oslo, Oslo, Norway. 2. Institute for Cancer Genetics and Informatics, Oslo University Hospital, Oslo, Norway. 3. Department of Pathology, Division of Laboratory Medicine, Oslo University Hospital, Oslo, Norway; Institute of Clinical Medicine, University of Oslo, Oslo, Norway. 4. Department of Oncology, University of Oxford, Oxford, UK. 5. Wellcome Centre for Human Genetics, University of Oxford, Oxford, UK; National Institute of Health Research Oxford Biomedical Research Centre, Oxford University Hospitals NHS Foundation Trust, John Radcliffe Hospital, Oxford, UK. 6. Department of Gastrointestinal Surgery, Oslo University Hospital, Oslo, Norway; Institute of Clinical Medicine, University of Oslo, Oslo, Norway; KG Jebsen Colorectal Cancer Research Centre, Oslo, Norway. 7. Gloucestershire Cellular Pathology Laboratory, Cheltenham General Hospital, Cheltenham, UK. 8. Institute for Cancer Genetics and Informatics, Oslo University Hospital, Oslo, Norway; Edinburgh Cancer Research Centre, University of Edinburgh, Edinburgh, UK. 9. Institute for Cancer Genetics and Informatics, Oslo University Hospital, Oslo, Norway; Department of Informatics, University of Oslo, Oslo, Norway; Research Department of Pathology, University College London Medical School, London, UK. 10. Nuffield Division of Clinical Laboratory Sciences, University of Oxford, Oxford, UK. 11. Institute for Cancer Genetics and Informatics, Oslo University Hospital, Oslo, Norway; Department of Informatics, University of Oslo, Oslo, Norway; Nuffield Division of Clinical Laboratory Sciences, University of Oxford, Oxford, UK. Electronic address: hdaniels@labmed.uio.no.
Abstract
BACKGROUND: Improved markers of prognosis are needed to stratify patients with early-stage colorectal cancer to refine selection of adjuvant therapy. The aim of the present study was to develop a biomarker of patient outcome after primary colorectal cancer resection by directly analysing scannedconventional haematoxylin and eosin stained sections using deep learning. METHODS: More than 12 000 000 image tiles from patients with a distinctly good or poor disease outcome from four cohorts were used to train a total of ten convolutional neural networks, purpose-built for classifying supersized heterogeneous images. A prognostic biomarker integrating the ten networks was determined using patients with a non-distinct outcome. The marker was tested on 920 patients with slides prepared in the UK, and then independently validated according to a predefined protocol in 1122 patients treated with single-agent capecitabine using slides prepared in Norway. All cohorts included only patients with resectable tumours, and a formalin-fixed, paraffin-embedded tumour tissue block available for analysis. The primary outcome was cancer-specific survival. FINDINGS:828 patients from four cohorts had a distinct outcome and were used as a training cohort to obtain clear ground truth. 1645 patients had a non-distinct outcome and were used for tuning. The biomarker provided a hazard ratio for poor versus good prognosis of 3·84 (95% CI 2·72-5·43; p<0·0001) in the primary analysis of the validation cohort, and 3·04 (2·07-4·47; p<0·0001) after adjusting for established prognostic markers significant in univariable analyses of the same cohort, which were pN stage, pT stage, lymphatic invasion, and venous vascular invasion. INTERPRETATION: A clinically useful prognostic marker was developed using deep learning allied to digital scanning of conventional haematoxylin and eosin stained tumour tissue sections. The assay has been extensively evaluated in large, independent patient populations, correlates with and outperforms established molecular and morphological prognostic markers, and gives consistent results across tumour and nodal stage. The biomarker stratified stage II and III patients into sufficiently distinct prognostic groups that potentially could be used to guide selection of adjuvant treatment by avoiding therapy in very low risk groups and identifying patients who would benefit from more intensive treatment regimes. FUNDING: The Research Council of Norway.
RCT Entities:
BACKGROUND: Improved markers of prognosis are needed to stratify patients with early-stage colorectal cancer to refine selection of adjuvant therapy. The aim of the present study was to develop a biomarker of patient outcome after primary colorectal cancer resection by directly analysing scanned conventional haematoxylin and eosin stained sections using deep learning. METHODS: More than 12 000 000 image tiles from patients with a distinctly good or poor disease outcome from four cohorts were used to train a total of ten convolutional neural networks, purpose-built for classifying supersized heterogeneous images. A prognostic biomarker integrating the ten networks was determined using patients with a non-distinct outcome. The marker was tested on 920 patients with slides prepared in the UK, and then independently validated according to a predefined protocol in 1122 patients treated with single-agent capecitabine using slides prepared in Norway. All cohorts included only patients with resectable tumours, and a formalin-fixed, paraffin-embedded tumour tissue block available for analysis. The primary outcome was cancer-specific survival. FINDINGS: 828 patients from four cohorts had a distinct outcome and were used as a training cohort to obtain clear ground truth. 1645 patients had a non-distinct outcome and were used for tuning. The biomarker provided a hazard ratio for poor versus good prognosis of 3·84 (95% CI 2·72-5·43; p<0·0001) in the primary analysis of the validation cohort, and 3·04 (2·07-4·47; p<0·0001) after adjusting for established prognostic markers significant in univariable analyses of the same cohort, which were pN stage, pT stage, lymphatic invasion, and venous vascular invasion. INTERPRETATION: A clinically useful prognostic marker was developed using deep learning allied to digital scanning of conventional haematoxylin and eosin stained tumour tissue sections. The assay has been extensively evaluated in large, independent patient populations, correlates with and outperforms established molecular and morphological prognostic markers, and gives consistent results across tumour and nodal stage. The biomarker stratified stage II and III patients into sufficiently distinct prognostic groups that potentially could be used to guide selection of adjuvant treatment by avoiding therapy in very low risk groups and identifying patients who would benefit from more intensive treatment regimes. FUNDING: The Research Council of Norway.
Authors: Andreas Kleppe; Ole-Johan Skrede; Sepp De Raedt; Knut Liestøl; David J Kerr; Håvard E Danielsen Journal: Nat Rev Cancer Date: 2021-01-29 Impact factor: 60.716
Authors: Dordi Lea; Martin Watson; Ivar Skaland; Hanne R Hagland; Melinda Lillesand; Einar Gudlaugsson; Kjetil Søreide Journal: Cancer Immunol Immunother Date: 2021-01-13 Impact factor: 6.968