Personalized Medicine Dr. Pablo Mentzinis, Director Government Relations, SAP SE Courtesy by Dominik Bertram, Marc von der Linden, Péter Adorján, SAP June 2016
1. Traditional medicine vs. personalized medicine
Personalized medicine is a paradigm change Understand and target the biological root cause FROM Descriptive Outcome based diagnosis Organ based Retrospective diagnosis Limitations for epidemiology Acute care Treatment for the average" patient TO Understand the disease mechanism Molecular diagnostics Fine molecular profile based groups Prospective diagnosis / Predisposition Environmental factors Prevention and early detection Individualized treatment Internal 3
Concept of traditional medicine The old way: Observe symptoms Classify disease Gold standard treatment Antibiotics High fever, cough, etc Infection High success rate Internal 4
Concept of traditional medicine A more complex disease: Breast Cancer Lumpectomy & Radiotherapy Lumps in breast Breast Cancer Low success rate, but sometimes effective Why? Internal 5
Concept of traditional medicine Treatment for the average patient is not effective for several subgroups Internal 6
Concept of personalized medicine A new approach: Determine individual root cause Targeted treatment BRCA1/2 Personalized prevention Gene Panel Breast Cancer HER2+ Personalized treatment Better, but not perfect Our understanding is incomplete Internal 7
Concept of personalized medicine As our understanding improves, therapy will become more individualized Individualized treatment Precise analysis of disease cause many different subtypes High response rate Internal 8
Concept of personalized medicine Targeting the individual root cause boosts treatment effectivity Internal 9
The challenge: handle biological complexity Molecular root cause Personalized prevention, diagnosis and treatment Cells Chromosomes DNA RNA Proteins Cellular Pathways Drugs METABOLOMICS PHARMACO- GENOMICS 10000s of cellular reactions Millions of compounds PROTEOMICS GENOMICS 3.3 billion base pairs (haploid) 3-4 million variants per individual ~20000 protein-coding genes ~10.000 genetic disorders (WHO) >5.000 known disorders (OMIM) TRANSCRIPTOMICS ~20000 protein-coding genes 53,000 non-coding RNAs 100000+ protein variants 160 million data points (2.4 GB) per sample 7.6 TB raw proteome data on ProteomicsDB.org Amount of Data and Value of Information Internal 10
2. Impact of personalized Medicine on healthcare
Why is it relevant? Low treatment efficacy Cancer: 75% Alzheimer`s: 70% Arthritis: 50% Diabetes: 43% Percentage of patients for whom drugs are ineffective. Depression: 43% Internal 12
The challenge: from data to actionable information Bridge the gap: advanced analytics, leverage network effects and digitize operations Number of nucleotides sequenced / $1,000 10,000,000,000 1,000,000,000 100,000,000 10,000,000 1,000,000 Moore s law 100,000 10,000 2000 2002 2004 2006 2008 2010 2012 2014 2016 Internal 13
Enabling personalized medicine Personalized medicine turns healthcare into an information problem of enormous magnitude Insight Gain precise insights from patient data Action Make healthcare processes more efficient SAP Personalized Medicine Solution Portfolio Network Create patient-centric healthcare networks Internal 14
Integrated view of omics data and patient information Empirical outcomes, symptoms Environmental factors & Lifestyle Wearable device data Genetics, genomics Transcriptomics Proteomics Metabolomics Microbiomics A true -omics integration and accessible analytics is required to create deeper insights Internal 15
Many actors in healthcare all work on their own Patient centric information backbone and collaboration remains a major a challenge GP Wearable devices Pharmacies Cyber physical systems Payers Patient Clinical laboratories Pharma/ R&D Chronic disease care centers Internal 16
Make healthcare processes more efficient Clinical systems: from time and data sink to easy access support and information gold mine SAP Patient Management application Patient experience and clinical delivery Ability to drive operational excellence from admission to bill Radical improvement through Modern software engineering User-centered design New technologies SAP Medical Research Insights Faster building and validation of hypotheses Patient cohorts and research Analytics (genome and patient data) Patient-trial matching SAP Foundation for Health Enablement of personalized medicine Ability to analyze massive volumes of structured and unstructured data (from patients, clinical, omics, third-parties, and so on) Secure collaboration and sharing Health Engagement Open environment for customers and partners to build care collaboration scenarios for: Engagement of the care network Motivation for behavioral change Prevention and risk detection Internal 17
Digitalization of healthcare EMR adoption model Stage 0 1 2 3 4 5 6 7 Cumulative capabilities All three ancillaries not installed Ancillaries Lab, Rad, Pharmacy All installed CDR, Controlled Medical Vocabulary, CDS, may have document imaging; HIE capable Nursing/Clinical documentation (flow sheets), CDSS (error checking), PACS available outside Radiology CPOE, Clinical Decision Support (clinical protocols) Full complement of radiology PACS Physician documentation (structured templates), full CDSS (variance & compliance), Closed Loop Medication Administration Complete EMR; CCD transactions to share data; Data warehousing; Data continuity with ED, ambulatory, OP 18
3. Solution portfolio
SAP Medical Research Insights In depth analytics enabled for clinical research Challenges addressed Build patient cohorts Match patients for trials Explore clinical outcomes Solution highlights Combines structured and unstructured patient data from disjointed data sources On the fly cohort analysis with an award winning UI Flexible model for comprehensive clinical and genetic data Internal 20
SAP Medical Research Insights SAP Medical Research Insights 2.0 GA release highlights Genomics data Analyzing Genomics variation of cohorts and single patients. Boxplot visualization Drill down to get more insights on the distribution behind the numbers Enhanced configuration Flexible adaptation of MRI to different users within your organization Internal 21
Clinical Measure Analytics Lab preview Turning quality assessment into immediate actions and excellence in patient care Effective clinical quality assessment is a challenge Complex quality rule sets are hard to implement and maintain Relevant clinical data are stored in multiple silos Labor intensive manual process to calculate quality measures Post hoc information no chance for early intervention Solution highlights On the fly dashboards for actionable measures and quality scores Flexible rule configuration and query engine Data sources are integrated in a clinical data warehouse Quality scores at multiple levels of granularity and time frames Scalable for hundreds of physicians and multi million patients Internal 22
Clinical Genomics Services Lab preview Generate actionable genetic insights in a clinical setting Enable physicians to determine pathogenicity of genetic variants Integrate genomic, clinical and wearable device data Efficient semi-automated genetic report generation Large scale genomic services enabled Workflow support: change alerts, knowledge sharing, audit trail, hand-overs and approvals Features highlights Functional annotation of whole genome sequencing data Patients like yours? Patient dashboard workflow support Clinical history timeline Interactive filtering and ranking of genetic variants Internal 23
Proteomics DB https://www.proteomicsdb.org A central repository for browsing and analyzing proteome data Offers a complete map of human proteins (proteome) to improve our understanding of physiological processes A public, free-of-charge platform powered by SAP HANA managing terra-bytes of human proteomics data Published in Nature May 2014 (selected as cover story) Internal 24
Proteomics DB 2015 release highlights Extended analytics and data exploration enabled Transcriptomics data Connecting proteomics and transcriptomics data to obtain more insights into molecular processes of the human organism Drug potency analysis See the effect of drugs on proteins and at the same time all potential side effects. Experimental planning Plan and store complex designs of experiments. Internal 25
Proteome tools Lab preview Translate proteome data into molecular and digital tools for drug discovery, personalized medicine, and life science research Synthesize and measure more then one million human peptides as a reference standard on https://www.proteomicsdb.org Turning the data into improved hardware, novel software approaches, and reagents for proteomics Provides a foundation for a new class of analytical approaches and algorithms High-class spectral libraries to improve methods and algorithms in proteomics Internal 26
Machine learning in proteomics Lab preview SAP-funded PhD / Postdoctoral research projects at Technische Universität München Drug sensitivity / resistance Make ProteomicsDB pharma-ready and enable the generation of drug sensitivity and resistance markers by considering additional model organisms, e.g. mouse, rat, and pig Multi-Omics biomarkers Leverage machine learning on top of several omics-data sources to enable molecular diagnostics and treatment decision support Proteomics analytics Combine the data, to be generated in ProteomeTools, with innovative identification algorithms to establish ProteomicsDB as a reference for protein identification Decoding of genomes Genomics Proteomics Decoding of proteomes Clinic Pharma Biotech. Crop Sci. Metabolomics Transcriptomics ADH5 MAP-K Internal 27
Medical Allround Care Service Solution Lab preview Health network = High quality patient care for chronic diseases Remove roadblocks for high quality care Low patient adherence No triggers for preventive intervention No patient centric view with multiple stakeholders Connect all stakeholders and share information Engage patients: mobile data exchange Alerts enable preventive measures Share medical data across stakeholders Pharmacy Clinic Doctor Dialysis center Data and services for medical systems and devices Patient-Front-Ends to support therapy in everyday life Patient Mobile apps Vital signs Platform solution based on SAP HANA Clinical data warehouse with unified ontologies Strict access right management, patient remains data owner Easy integration of applications and data services (e.g., NLP) Back-End / Middleware Secure storage of medical data incl. management of access privileges in SAP HANA Secure data platform at Charité Berlin Internal 28
Health Engagement Solution End-to-end solution concept open for partners Connected Physician portal devices Patient app Program manager SAP HANA Cloud Platform Researcher portal Internal 29