Sunday, September 3, 2017 | 2:45-5:15 PM
Amsterdam, The Netherlands
For more information on this session, please click here.
Computational Pathology and Interoperability to advance Digital Pathology
Esther Abels, MSc, Philips Digital Pathology Solutions
Objectives: Digital transformation in pathology is slowly progressing. We present how the Digital Pathology Association (DPA) will use regulations, standards and technology to enable advancing digital and computational pathology.
The Digital Pathology Association (DPA) has the mission to advance digital pathology by bringing clarity to the regulatory pathway for digital pathology including its evolution and creating awareness thereof, and working towards the development and adoption of standards as well as promoting interoperability in digital pathology for clinical use. Whole Slide Imaging systems have been in clinical use for primary diagnosis in Europe for several years now. The adoption of Digital Pathology is slowly progressing. The benefits and opportunities digital pathology can bring to improve efficiency and better inform decision making is when the actual transformation to digital will start to take its effect. Digital Pathology is comprised of a set of hardware, software and algorithms working together to forming a digital process in the pathology lab. This opens doors to exciting and timely developments of image analytics, deep learning and computational pathology, which could enable the lab and the pathologist to become more effective in enhancing diagnosis, safety, quality and efficiency. Digital Pathology can also provide advantages to Pharma companies enabling them to use WSI platforms to discover biomarkers and new drug applications, and develop companion or complementary diagnostics to ultimately improve patient health. The underlying prerequisite to have this take effect is the compatibility of various hardware, software and algorithm platforms. Scientific evidence has to be put forward to demonstrate this. It is the goal of the DPA to provide clarity and guidance on the required type of evidence required for regulatory purposes. Regulations are present to protect public health and at the same time ensure health care providers and patients will have access to innovative disruptive technologies such as Digital Pathology. The evolution, progress and future regulatory steps involved in this Digital Pathology transformation will be presented.
Automated Image Analysis of Immunohistochemistry as a Biomarker for Immuno-oncology (I-O)
Michael Montalto, PhD, Bristol-Myers Squibb
Immune checkpoint modulators are rapidly changing the way cancer is treated. Biomarkers that can predict response to this class of drugs are becoming increasingly important in current pathology practice. Several immunohistochemistry (IHC) assays for PD-L1 are approved by the FDA as companion and complimentary diagnostics for various immune checkpoint inhibitors. This presentation will briefly review the promises and challenges of these assays for determining therapeutic eligibility, as well as explore other IHC tests that have promise as biomarkers for immune-modulatory agents. The potential role of automated image analysis algorithms to augment IHC testing in the clinic will be highlighted in the context of biomarker and companion test development for IHC markers, as well as the challenges to bringing such tests to market.
Integrated pathology informatics: The UHN experience
Sylvia Asa, MD, PhD, University Health Network
Informatics is a key requirement for the provision of high quality pathology. This presentation will describe the development of a multisite pathology informatics platform using a sophisticated laboratory information system and whole slide imaging for histology and immunohistochemistry, integrated with electron microscopic images, data captured from flow cytometers, as well as integrated cytogenetics and molecular diagnostics. This platform provides access to a model of subspecialty pathology for patients in numerous geographical locations in a health care system that is responsible for services across a large geographic area, and allows reporting of every specimen by the right pathologist at the right time. It facilitates pathologist participation in multidisciplinary case conferences, including virtual presentations in different locations. Discrete data collection using a synoptic reporting module and capture of quality assurance activities and workload measurement provide additional examples of benefits of this electronic approach to pathology. Integrated pathology informatics builds the foundation for big data collection and high quality personalized and precision medicine.
QuPath: A Flexible, Extensible, Open Source Platform for Digital Pathology
Peter Bankhead, Philips
Objective Quantitative image analysis in digital pathology has huge potential to improve the speed, objectivity and reproducibility of whole slide analysis and biomarker interpretation. However, developing, validating, applying and sharing novel algorithms remains difficult. We have created a novel, open source platform to provide a comprehensive panel of tumor identification and high-throughput biomarker evaluation tools to the community, with an extensible design to enable new ideas and algorithms to be shared freely. Methods QuPath has been developed as a cross-platform Java application. We applied it to score a range of biomarkers in tissue microarrays within the CRUK Accelerator programme on immuno-oncology markers, and in the Northern Ireland Molecular Pathology Laboratory for other biomarker panels. Results Across a cohort of 300 breast cancer patients, analysis with QuPath showed high concordance with a pathologist’s manual scoring for ER, PR, HER2, Ki67 and p53. QuPath analysis of CD3, CD8, PD-L1 and p53 within a cohort of 660 colon cancer patients also robustly identified significant associations between biomarker expression and disease specific survival within this cohort. Conclusions QuPath offers a highly flexible toolkit for digital pathology and scoring a diverse range of biomarkers. It is available open source with documentation at https://qupath.github.io.