The Medication Safety Code initiative

Towards a global IT system for personalized medicine

Matthias SAMWALD, José Antonio MINARRO-GIMÉNEZ, Kathrin BLAGEC, Klaus-Peter ADLASSNIG

Medical University of Vienna, Austria

Drug safety and effectiveness vary drastically between patients with different genetic profiles

Up to 100,000 deaths and 2 million hospitalisations per year in the United States

Many therapeutics in development eventually fail to reach patients and costs of bringing a new drug to market are now > 1 billion Euros

Pharmacogenomic assays and treatment algorithms are becoming more and more numerous

Pharmacogenomic assays and treatment algorithms are becoming more and more numerous

Pharmacogenomic assays and treatment algorithms are becoming more and more numerous

Sequencing-based:PGRNseq

…

Microarray-based:23andMe

Affymetrix DMET chipFlorida/Stanford chip

…

We are creating a barrier-free system for storing and interpreting personal pharmacogenomic information: The Medication Safety Code

Current version contains data on ~15 pharmacogenes with clinically actionable variants

We are creating a barrier-free system for storing and interpreting personal pharmacogenomic information: The Medication Safety Code

Technology

• Compressed pharmacogenomic data in QR codes

• Can be readily decoded and interpreted with common mobile deviceso QR code readers come pre-installed on most devices; installation of

dedicated apps not requiredo Web-based decision support service gives dosing recommendations

based on up-to-date clinical guidelines

• All data are inside the QR codeo No central database required

(patients have full control over their data)o Decoding on client-side through Android app is possible

• Data remains anonymous

• Backed by a sophisticated knowledge base we createdo > 300 pharmacogenomic decision support rules as table and OWL

ontology

After scanning, the system presents a searchable, prioritized list of personalised pharmacotherapy guidelines

After scanning, the system presents a searchable, prioritized list of personalised pharmacotherapy guidelines

Alternatively, you can select drugs for which treatment recommendations are required and enter data manually

Recent addition: MSC app for offline-use available on Google Play for registered users

Preview: Formative user evaluations with clinical pharmacists

• Ongoing user evaluations with clinical pharmacists at University of Pittsburgh

• Preliminary results help optimize pharmacogenomic CDSo Guidelines from different sources slightly different, confusing to

users when both are presentedo Pharmacists really appreciate alternative treatment

recommendations (not only warnings)o Trustworthiness and traceability of recommendations are very

important

Preview: Checking the fidelity of (pharmaco)genomic tests for diverse patient populations

• Using full-genome data from > 1000 patients of diverse ancestries (1000 genomes dataset)

• Simulating the results one would yield with cheaper (array-based) pharmacogenomic tests and current haplotype definitions for these patientso Usually, these tests only check for certain ‘characters’ in the

genome sequence, while ignoring others

• Our unpublished results indicate that cheaper genetic tests are prone to produce false or questionable results quite frequently, and that these results differ based on ancestryo E.g., for certain genes and assays, patients with African or Asian

ancestries have higher risk of problematic pharmacogenomic assay results

Preview: Impact estimation of delivering pre-emptive pharmacogenomic testing and CDS to broad patient populations

Preview: Impact estimation of delivering pre-emptive pharmacogenomic testing and CDS to broad patient populations

• Impressive data on medications with associated pharmacogenomic guidelines from private practices in Germany alone: o Total net cost of 3,75 billion € per yearo 123 million prescriptions per year

• We are currently conducting a study with detailed prescription data from hundreds of millions of patients (IMEDS datasets, mostly from United States)

Vision

• Creating a barrier-free system for putting personalized medicine into the pockets of patients

• Making pharmacotherapy safer, more effective andless costly through pre-emptive genotyping

• Cooperation of stakeholders o Clinical institutions and medical professionalso Health insurance providers o Pharmaceutical companieso Genetic testing providerso Health IT companieso Patient organisations

• Enable bringing stratified/personalized therapeutics to market

Vision

• Towards a new synthesis: Personalized, computer-assisted pharmacotherapy Pharmacogenomics for pharmacokinetics and -dynamics Polypharmacy and clinical parameters (e.g., organ dysfunction) Personalisation based on sex / gender Age (pediatric and geriatric patients) Genetic risk markers for rare, severe adverse effects Somatic mutations in cancer therapy

Wrapping up:How does all that relate to you?

Join us onhttp://safety-code.org/

Wrapping up:How does all that relate to you?

• We are interested in evaluating the system and decision support functionality in various medical settings (as well as pharmacies) opportunity for research papers!

• We are looking for commercial partners to integrate MSC technology into their products opportunity for novel products for medication safety!

• European Research projects (e.g., PHC-24)

Thanks!

Local team (Medical University of Vienna)

Dr. Matthias Samwald (team lead, biomedical informatics & pharmacology)

Dr. Jose Antonio Miñarro-Gimenez (researcher and developer, computer science)

Kathrin Blagec (medical student)

Prof. Dr. Klaus-Peter Adlassnig (biomedical informatics)

International participants & supporters

Robert R. Freimuth (Mayo Clinic)

Richard Boyce (University of Pittsburgh)

Michel Dumontier (Stanford University)

Simon Lin (Marshfield Clinic)

Web

http://safety-code.org/

Funding

Austrian Science Fund (FWF): [PP 25608-N15]