© FIMM - Institiute for Molecular Medicine Finland www.fimm.fi Disclosures

› Research collaborations: .Labcyte Inc (liquid handling) .Pfizer, Roche .IMI-Predect public-private project (10 companies)

› Off-label use of drugs will be discussed

2 Genomic and molecular landscape in AML

Number of patient samples

23 significantly mutated genes in 200 AML patient samples

The Cancer Genome Atlas (TCGA) dataset Panoramic view of AML, Chen and Chen; Nature Genetics, June 2013

Genomic and Epigenomic Landscapes of Adult De Novo , New England Journal of Medicine 2013

3 Example of a translational gap in acute myeloid leukemia

• AML genomes Novel therapy sequenced targets and Many <50% survival • Clonal therapeutics leukemia -Resistance evolution (e.g. FLT-3, patients common (10% characterized DNMTs) die due to survival) • Increasing disease/patient- disease -Classical data on specific and/or molecular molecular treatment used (from pathogenesis markers toxicity 1960s) with • Stem cell toxicity niche -Targeted treatments not available -Mutational profile often not helpful

4 US National Academy of Sciences 2012 5 Infrastructure and collaborative environment for individualized systems medicine in AML

Finnish Hematology Society

Genomic & molecular profiles Kimmo Porkka National Drug sensitivity testing Satu Mustjoki Clinic Biobank Data integration & models Mika Kontro

Krister Wennerberg Tero Aittokallio Individualized and Caroline Heckman Jonathan Knowles improved therapy Olli Kallioniemi Why AML? Sampling over the course of disease Cancer always accessible Ex-vivo functional drug testing easier 6 Systems medicine to optimize drugs and drug combinations for individual AML patients

• Repeated sampling at diagnosis, therapy response, relapse and drug resistance

• Integration of genomic and molecular profiles and high-throughput ex-vivo drug testing for each patient (real-time)

• Feedback therapeutic insights back to the clinic for tailoring personalized therapy or guiding clinical trials with new agents

• Molecular and functional analysis of drug resistant samples after novel therapies 7 Biobanking all leukemia patients across the country: On average 46 sample aliquots per patient

8 Samples across time points provide molecular insights on disease progression in patients, including drug resistance

DiagnosisRelapse 1 Treatment 1 Relapse 2 Relapse 3

Drug testing Drug testing Drug testing Exome seq Exome seq Exome seq Exome seq p-proteomics RNA seq RNA seq RNA seq p-proteomics p-proteomics p-proteomics 9 Individualized systems medicine (ISM)

System medicine approach: - Data integration - Repeated sampling - Feedback to clinic - Learning system Pemovska et al. Cancer Discovery, on-line Sep 20, 2013 Rapid data analysis, integration and interpretation to provide therapeutic insights to the clinic

Drug sensitivity testing Exome sequencing

Integration & Interpretation & Feedback to clinic (4 days to 2 weeks)

Gene expression Phosphoproteomics Fusion genes

11 High throughput screening infrastructure: - “repurposed” from functional genomics/drug discovery to drug sensitivity testing for individualized medicine

BeckmanCoulter integrated robotic system Labcyte Access Workstation – fully automated assays – nl acoustic dispensing

Plate readers Micro- 96 384 1536 Imaging arrays Flow cytometry Genomics

Wennerberg, Östling et al. HTB facility at FIMM 12 The most direct way to ascertain patient treatment: test drugs individually on patient cells Drug sensitivity and resistance testing (DSRT) workflow

Leukemia cells 4 day turnaround isolated from Fluorescent dyes blood or bone (HCS) marrow Measure cell growth/survival/ p- 30 x 106 cells proteins and calculate response 37°C, 72 h

Pre-plated drug collection (1, 10, 100, 1000 and 10 000 nM) 13 Drug sensitivity and resistance testing with dose-response curves for each drug

Detailed dose-response curves Probes for all oncology drugs and many 66 emerging cancer compounds for individual patient samples Approved 137 Investigati onal 103

14 (+)JQ1 Bimatoprost GDC-0941 Mechlorethamine Omacetaxine Roscovitine Tosedostat 2-methoxyestradiol BKM-120 Danusertib GDC-0980 Megestrol OSI-027 Ruboxistaurin Abiraterone Dasatinib Gefitinib OSI-906 ABT-751 BMS-754807 Mepacrine Ruxolitinib UCN-01 Afatinib Goserelin S-trityl-L-cysteine Uracil mustard Allopurinol Bosutinib Dexamethasone Gossypol Saracatinib Brivanib Dexrazoxone GSK1120212 Methoxsalen Patupilone Selumetinib Vandetanib Alvespimycin Bryostatin 1 Hydroxyurea Methylprednisolone Pazopanib Serdemetan Vargatef Alvocidib Doramapimod MGCD-265 PD 0332991 Sirolimus Amifostine Cabozantinib Dovitinib Midostaurin SNS-032 Aminoglutethimide CAL-101 Imatinib Sorafenib Vemurafenib EMD1214063 Imiquimod PF-00477736 Sotrastaurin VER 155008 Amonafide Canertinib Indibulin PF-04691502 Streptozocin Enzastaurin Iniparib MK-2206 PF-04708671 Sunitinib Anastrozole Erlotinib MK1775 Pilocarpine Tacrolimus Estramustine MLN8237 Tamoxifen Vismodegib AS703026 JNJ-26481585 Mocetinostat Plerixafor Tandutinib Volasertib AT9283 Cediranib Everolimus Lapatinib Motesanib Tanespimycin Axitinib Exemestane Lasofoxifene Navitoclax Ponatinib Tarenflurbil XAV-939 AZ 3146 Fasudil Lenalidomide Prednisolone Tasocitinib XL147 CI-994 Finasteride Lestaurtinib Neratinib Prednisone XL765 AZD1480 Fingolimod Letrozole Nilotinib PRIMA-1(Met) Temsirolimus YM155 AZD4547 FK-866 Leucovorin calcium Nilutamide Zolendronic acid AZD7762 Levamisole PX-866 Thalidomide AZD8055 Clomifene Linifanib Nutlin-3 Quizartinib Thioguanine Barasertib Crizotinib NVP-AUY922 RAF265 CUDC-101 Flutamide LY2157299 NVP-BEZ235 Raloxifene Foretinib LY2603618 NVP-LDE225 RDEA119 Tivozanib BI 2536 Fulvestrant LY2784544 Obatoclax Regorafenib BIIB021 Galiellalactone Masitinib Ridaforolimus Toremifene 15 Identification of cancer-specific drug responses

› Kill cancer cells more than normal cells (cancer-selective response) › We calculate DSS scores by comparing cancer’s response to controls (Healthy bone marrow)

100 control AML 75

50 Patient-specific cancer-

% viability selective drug response 25

0

-9 -8 -7 -6 -5 Conc (log M) High DSS value => cancer cells have higher sensitivity to the drug 16 Summary of ex-vivo drug testing data from 27 chemorefractory AML patients

Most frequent bioactive drugs in 27 AML patients

17 Guidance for personalized I II III IV V 600_1 252_5 252_1 560_1 560_9 600_2 1690 4701001 medicine and 1993 2095 1064 1145_3 800 393_1 393_3 1145_6 1145_1 718 1280 1497 600_3 252_3 252_2 1867 784_2 784_1 250 Disease stage DD D*RR D*DR D*RR D R R R D R R R R R R R R R D R Navitoclax clinical Ruxolitinib

$" & + '& + % &' $ $ $ $ $ $ $ $ $ $ $ $ $ $ $

Dexamethasone $" &+( ' ! $ ++) MEKi development of PI3K/mTORi $" )))*& &# Quizartinib emerging drugs Sunitinib DasatinibLink between MEK Topoisomeraseand JAK IIi inhibition sensitivity in DSRT Correlating drug TP53 RUNX1subtype II sensitivity with genomic DNMT3A NRAS/KRAS data: MLL-X fusions • Predictive biomarkers WT1 ETV6-NTRK3 • Novel therapeutics KIT MLL fusions link to NUP98-NSD1 • Drug combinations FLT3-ITD MEK and PI3K NPM1 inhibitor sensitivity • Novel drug targets IDH1/IDH2 PTPN11 • Insights on drug resistance Activating FLT3 Adverse karyotype mutations link to FAB subtypes M2 M1 M2 M2M2M2M5 M5 M5 M5sensitivity M5M5M5M5 M5M5M5to FLT3 M1M1 inhibitors and dasatinib 18 Individualized systems medicine for AML to define optimal drugs for patients

Using drugs as biological probes => direct opportunity for translation 19 Drug sensitivity Cancer-selective drugs testing report: summary of drugs Ineffective drugs showing selective cancer efficacy in an individual AML patients based on ex vivo testing

20 Clinical data: A treatment-refractory AML patient receiving a novel combination of clinical drugs based on ex-vivo profiling: Complete clinical response and subsequent progression

Ex-vivo drug response data after relapse

21 Clinical implementation of ex-vivo drug-testing data in chemorefractory AML patients

Leukemia‐selective responses observed in ex‐vivo drug testing Drugs available for human use (on‐ or off‐label)

14/17 (81%): ex‐vivo guided treatment possible

10 patients received personalized combinatorial treatments designed according to predictions

2 x CRi , 2 x leukemia free (hypoplasia) 40% response rate

Drug combinations for patients: - Clo-Dasatinib-Vin - Sorafenib-Clo - Dasatinib-Sunitinib-Temsirolimus - Azacytidine-Sunitinib 22 Kontro et al, EHA 2013 Individualized systems medicine for AML to define optimal drugs for patients

 Molecular profiling to understand drug resistance and tx failure  Drug-testing to understand drug (cross)-resistance and new vulnerabilities 23 Assesment of treatment response: clonal evolution by next-gen seq after treatment with targeted drugs

- Genomics as a measure of drug response in patients - Understand resistance mechanisms and evolution of multiple subclones => The clones causing clinical resistance often pre-exist before therapy

24 Evidence of vulnerability to previously ineffective drugs after resistance has emerged in patients => An opportunity to design curative drug combinations

New drug vulnerabilities = Therapeutic options for patients sample 1 sample 2 diagnosis relapse Drugs whose efficacy is cells after resistance not altered by the clinical therapy resistance Compare drug vulnerabilities after relapse Cross-resistance to other drugs Sensitivity of cancer Sensitivity of cancer cells before resistance

25 Understanding altered cell signaling after treatment

Kinase signalling networks predicted from drug response

26 Systems medicine to optimize drugs and drug combinations for individual AML patients

• Repeated sampling at diagnosis, therapy response, relapse and drug resistance

• Integration of genomic and molecular profiles and high-throughput ex-vivo drug testing for each patient (real-time)

• Feedback therapeutic insights back to the clinic for tailoring personalized therapy or guiding clinical trials with new agents

• Molecular and functional analysis of drug resistant samples after novel therapies 27 Future development Clinical: › Drug sensitivity testing in primary AML › Stratified, biomarker driven trials for new drugs in AML › Comparing treatment paradigms: Individualized vs. conventional clinical therapy Biological: › Progenitor / single cell assays › Mimicking the bone-marrow niche › Larger libraries, combinations to facilitate drug repositioning

28 Slide content not available for publication

WIN 2014 Symposium • 23-24 June • Paris • France Slide content not available for publication

WIN 2014 Symposium • 23-24 June • Paris • France Slide content not available for publication

WIN 2014 Symposium • 23-24 June • Paris • France Slide content not available for publication

WIN 2014 Symposium • 23-24 June • Paris • France Tissue slice cultures to assay short-term drug effects (e.g. DNA damage, phosphorylation events) Towards analysis of solid tumors?

Reprogrammed primary cell culture (Rock-inhibitor & feeder layers)

Primary prostate cancer organoid cultures in 3D

(Clevers protocol) 33 Precision Cancer Medicine Network FIMM HUCH/HRU Personalized Cancer Medicine Individualized Systems Medicine Kimmo Porkka Caroline Heckman Olli Kallioniemi Satu Mustjoki Jonathan Knowles Taija af Hällström Pekka Anttila Samuli Eldfors Henrik Edgren Mika Kontro Poojitha Kota Venkata Riikka Karjalainen Erkki Elonen Disha Malani Jarno Kivioja Hanna Koskela John Patrick Mpindi Ashwini Kumar Mette Ilander Astrid Murumägi Heikki Kuusanmäki Emma Anderson Päivi Östling Paavo Pietarinen Muntasir Mamun Majumder Maija Wolf Alun Parsons Jaakko Vartia Minna Suvela Technology Center Minna Lehto Janna Saarela Mervi Saari Chemical Systems Biology Evgeny Kulesskiy Kuopio University Central Hospital Krister Wennerberg Laura Turunen Raija Silvennoinen Tea Pemovska Anna Lehto Turku University Central Hospital Arjan van Adrichem Ida LIndenschmidt Tuija Lundán Pekka Ellonen Tampere University Central Hospital Computational Systems Biology Maija Lepistö Hannele Rintala Tero Aittokallio Sonja Lagström Tero Pirttinen Petteri Hintsanen Sari Hannula Marja Sankelo Agnieszka Szwajda Pirkko Mattila University of Bergen Bhagwan Yadav Aino Palva Bjørn Tore Gjertsen

34 © FIMM - Institiute for Molecular Medicine Finland www.fimm.fi