Improving surveillance and early detection of outbreaks. Frank M. Aarestrup

Transcription

1 Improving surveillance and early detection of outbreaks Frank M. Aarestrup

2 INFECTIOUS DISEASES Direct cause of 22% of all global deaths (15 million), huge burden on society and economy Expected to increase; e.g. AMR alone causing 10 million deaths in 2050 (cancer 8.2 million in mainly older people) Dynamics are changing and new diseases/variants emerge Demographic change, population density, anti vaccine, AMR, climate change, etc. Effects of interventions are difficult to predict due to complexity Rapid flexible response Public health, diagnostic, vaccine development and clinical response depend on global capacity for disease surveillance Rapid sharing, comparison and analysis of data from multiple sources and using multiple methodologies open science in a controlled way

3 Box plots of the median time between estimated outbreak start and various outbreak milestones for a subset of WHO-confirmed outbreaks, Emily H. Chan et al. PNAS 2010;107:

4 Assay development and validation timeline according to FIND Assumption 1: immediate sharing of sequences and strains Assumption 2: access to high quality clinical material Assumption 3: staff and funds available 4 Courtesy of Catharina Boehme, FIND, at Wellcome trust meeting October 2016

5 The surveillance pyramid Only a minor proportion of cases are recognized by routine surveillance Notified Cases diagnosed Persons tested Persons seeking health care Ill individuals Infected individuals

6 Reduce impact by improving: - Early detection & control - Treatment - Prediction & prevention Karesh et al., 2012

7 Prediction through hot spot surveillance? Massive underreporting from developing countries

8 What the world needs Real-time data on occurences of all infectious agents everywhere Pathogen independent Detection of related clusters in time and space Possibility to observe trends in species/clones/virulence/amr Ability to rapidly compare between all types of data Transfer of information to those who need to: Take public health respose Develop tests and vaccines Clinical decisions There can be no real-time surveillance without real-time data sharing

9 Possible solutions Easy sharing of data publicly, privately, securely Central repository? Hot spot surveillance Sewage (combines many people/animals) Climate (prediction of locations/time) Social data (syndromic) Generic tools Enables comparison between data Diagnostic preparedness networks and biobanks Verification of information Collaborative mechanism for rapid sharing, validation and implementation Standard legal and administrative framework

10 Relevant networks and activities Acronym Focus Funding EURLs Detection of specific zoonotic pathogens EU PANDEM Capacity building for crisis management EU PREPARE Clinical research during epidemics EU COMPARE Platform for rapid exchange of information EU MIRRI Microbial Archive EU - ESFRI EVA-g Virus archive EU - ESFRI ERINHA BSL4-infrastructures EU - ESFRI Elixir Bioinformatic tools and databases EU - ESFRI EVDlabnet Expert laboratory support National ISARIC Network for research National GHSI Network for management National EDCARN Network National GMI Network for NGS sharing National PNI FBP cluster analyses National WHO & OIE CCs Specific issues and pathogens National PREDICT Hot-spots for sampling zoonoses USAID GLOPID-R Funding in emergencies Adapted from EMERGE M. Koopmans and C. Reusken

11 PREDICT COMPARE PNI EVDlabnet MIRRI EVA-g ISARIC ERINHA WHO CC OIE RL EU RL PREPARE Clinical studies Sampling Detection Pathogen characterisation Test, treatment Vaccines Industry Datasharing Management action Funding GMI, ENA, Elixir GLOPID-R GHSI PANDEM

12 PANDEM Surveillance, Situational Awareness & Decision Support Specific solutions requiring further investment: Enhance capacity to analyse and visualise data by developing visual analytic tools for multiple users Gather, process and analyse data at community level to measure and track societal impact Carry out predictive modelling to examine likely scenarios and the impact of countermeasures Supporting the development and integration of tracking and tracing and the next generation of laboratory services

13 NGS advantages Laboratory diagnostics increasingly rely on genomic information RNA / DNA are common across pathogens, therefore, methods to analyse pathogen genomes are potentially universal NGS capacity is developing fast, and costs are becoming competitive NGS provides a universal language that can be easily shared and harnessed for early detection and comparisons across disciplines and domains If the technology keeps developing, less equipped labs may leapfrog

14 Clinical: Diagnosis of common pathogens Evaluation of syndromes Public health: Evaluation of syndromes Tracking of pathogens Virulence mapping Research: Reservoir discovery Vaccine (escape) Resistance traits Transmissibility markers 1. Everyone needs to be able to sequence and get data analysed 2. Common bottlenecks (protocols for matrices/pathogens, bioinformatics, ICT, etc.) 3. Agreed mechanisms for data sharing, give and take, some protection 4. Open source solutions This project has received funding from the European Union s Horizon 2020 research and innovation programme under grant agreement No

15 Ultra-short update Decision support algoritms for risk based sampling SOPs for sample processing and sequencing Analytic workflow for clinical decision and outbreak detection Datahubs and procedues for sharing (and analysing) NGS data Barriers +2 million bioinformatic analyses conducted Pilot-studies on: AMR prediction, flu transmission, virulence factors, phylogeny, hosp transmission, etc. Hot-spot surveillance

16 Why a central public repository? A common language Avoid a lot of data transfer (petabytes) Allowing easy transfer between levels of access including public (immediate access in war-time) Allow access to bioinformatics for the frontline Include the less advanced Allowing for constantly improving the analytic pipelines

17 ECDC / EFSA / COMPARE pilot March May 2018 RESTful API or Analysis Notebook DTU uploader or Webin Web Portal Data Hub Discovery API FTP ENA Member state Analyses and sharing This project has received funding from the European Union s Horizon 2020 research and innovation programme under grant agreement No

18 Global sewage surveillance Hendriksen et al. (in prep) 9 March Tb, large diversity + 30 million genes

19 AMR

20 Prediction of the World, based on global data (World bank, Flight, trade, AMU) Normal map Map according to population size and AMR Can explain 85%-90% of the variation by available data Trade and travel of limited importance

21 Barriers Scientific The selfish scientist Analytic tools, unstructured data Speed Solutions Scientific Funding bodies/journals Deep learning Bioinformatics Technical Security Access in peace and war Speed Technical Zero knowledge proof / Block-chain Central repository HPC, central repositiory Traditional networks Focus on the sample PEARL (politic, ethic, adm., reg., legal) Access to data from hosp., lab., individuals PEARL Sewage Un-traditional sources Searchable but not accessible

22 COMPARE vision: one system serves all Guiding principles: - Cross sector, cross domain, open source (not commercial) - Interaction with the rest of the world (all inclusive) - Data for action (actionable outputs) - Central repository (ENA, DDJ, NCBI) (bring the tools to the data) There can be no real-time disease detection & surveillance without real-time data sharing This project has received funding from the European Union s Horizon 2020 research and innovation programme under grant agreement No