The value of Biobanks for personalized medicine in acute Leukemia

Hematology Research Unit Helsinki

Biobanks and Personalized Medicine: The FHRB Biobank Concept

Kimmo Porkka Helsinki University Hospital Comprehensive Cancer Center

Department of Hematology and


University of Helsinki

Biobanking as a resource for biomedical research, 9-13 Feb 2015


Population-based vs. disease/research-oriented biobanks

• Population-based biobanks

• Large patient numbers, “simple” sample collection

• Pathology samples as the backbone

• Focused, disease-specific biobanks

• Smaller patient numbers

• Extensive sample collection/patient

• Multiple collection instances (diagnosis, remission, relapses)

• Need for both, not competitive, complementary


Some challenges for (cancer) biobanks

• Site and temporal variation inherent to cancer

• Repeated sampling, multisite sampling

• Clinical annotation of biological/biobank samples

• No uniform, national/international quality registers for clinical care and research

• Annotation of drug response phenotypes


Example: selecting biobank samples for drug research from clinical annotation data

• Commonly, new drugs are tested in a population of patients resistant to conventional therapy

• Search criteria: pick biobank samples from those patients who have first responded to conventional therapy, but then lost the response (secondary resistance)

• Current patient records/charts (FIN): not possible

• Other hospital databases (lab, pathology etc): not possible/difficult, indirect

• Only feasible with structured, disease-specific evaluation of therapy outcome


Individualized systems medicine for leukemia

Individualized and improved therapy

Biobank Profiling and drug testing technologies

Research

Clinics

Example of a disease-specific population-based biobank

Finnish Hematology Registry and Biobank


FHRB – owners and partners

• Finnish Red Cross Blood Service

• FIMM – Institute for Molecular Medicine Finland

• Finnish Association of Hematology

• Contractual partners: all hospital districts treating hematological patients (n=21)


Finnish Hematology Registry and Biobank - FHRB

• National registry and biobanking effort that aims to include all university and regional medical centers by 2015 in a population-based manner.

• High-quality samples from all patients with a hematological disease diagnosis.

• Bone marrow aspirates and blood samples collected at diagnosis, remission and relapse(s). Skin biopsy taken at first sampling.

• Samples processed and coded by the Finnish Red Cross Blood Service and then stored at FIMM as viable frozen mononuclear cells, frozen plasma and DNA.

• Researchers (incl. pharma) apply to the FHRB board for use of samples.

• Data returned to FHRB database


Flow chart

Sample processing Finnish Red Cross Blood Service

(centralized)

Biobanking Clinical registry

FIMM

FAH

Patient consent

Sample, data collection • 30 mL blood, 30 mL bone marrow, skin biopsy (opt.)

• at diagnosis, remission (opt.), relapses

• all university and central hospitals (n=21) Hospitals

Experiments, discovery

Research groups, pharma

Clinical applications

Diagnostics Imaging Targeted therapies

FHRB


Clinical registry

Biobank

+ FHRB =


FHRB clinical quality register & sample inventory: real-time data entry



FHRB: Samples biobanked/patient

50 sample tubes/patient/sampling: all samples stored in LN2


FIMM: centralized LN2 based sample

storage system

• Current capacity >1M samples (2 ml

cryotubes); expandable

• Samples stored in LN2 vapour phase

(-180C)

• Trained personnel, audited processes

• IT-systems

• Sample management

• Storage conditions surveillance

• Safety procedures

13


FHRB-project: biobanking timeline

• Started in Dec 5, 2011; currently all newly-diagnosed and relapse patients sampled (acute leukemias, MDS, myeloma, MPN, CLL, CML)

• All University hospitals: Helsinki, Turku, Tampere, Oulu, Kuopio collecting samples

• All other hospitals treating hematological patients by the end of 2015

• 300-500 patients/year (50 000 samples/year)


FHRB is part of routine patient care

• A biobank lab request

– 453 €/sample (+157 € for Fridays)

• Data registry

• Sampling, transport

• Sample processing

• Biobanking (6-7 yrs)

• Costs included in the daily care of the patients (laboratory budget)

• ”good care of a hematological patient”

• Cost of one PCR test

• <3% of university clinic lab budget


Continuous quality control: cell viability

FHRB operational group quality control round September 2013


Status 04.FEB.2015

http://www.hematology.fi/fhrb-status

873

Total 39 799 tubes in biobank


FHRB – a new biobank

• Finnish biobank act September 1, 2013

• TUKIJA (National Committee on Medical Research Ethics) approval November 2013

• VALVIRA (National Supervisory Authority for Welfare and Health) approval July 14, 2014

• Transfer of FHRB project samples to FHRB biobank (October 2014).

• First national, disease-focused biobank in Finland

• Accepting applications for samples autumn 2015 – Academic research groups – Pharma


FHRB - organization

Board of owners

Steering group

Executive officer Operative unit

Scientific advisory board

Administrative headquarters

Administrative and legal responsibility

Operative responsibility

Hospitals

Sample applications


Infrastructure and collaborative environment required to practice individualized systems medicine in acute myeloid leukemia (AML)

Genomic & molecular profiles Drug sensitivity testing

Data integration & models

Individualized and improved therapy

National Biobank

Clinics

Why AML? Sampling over the course of disease

Cancer always accessible Ex-vivo functional drug testing easier

Finnish Hematology Association


AML: survival from relapse by age

Rowe, J. M. et al. Blood 2010;116:3147-3156


Genomic and molecular landscape in AML

22

23 significantly mutated genes in 200 AML patient

samples

The Cancer Genome Atlas (TCGA) dataset

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

Number of patient samples

Panoramic view of AML, Chen and Chen; Nature Genetics, June 2013


Molecular diversity: a challenge for drug development

100 newly diagnosed AML patients

AML genotype 1

AML genotype 2

AML genotype 3

AML genotype 5

AML genotype 6

AML genotype 7

AML genotype 98

AML genotype 99

AML genotype 100 …

New drug 1: Phase I-II (n=10-40)

Response rate 2/100= 2% Study fail End development



Individualized Systems Medicine Overview

Systems medicine approach:

- Data integration

- Repeated sampling

- Feedback to clinic

- Learning system

FHRB biobank

Pemovska, Kontro et al. Cancer Discovery, 2013


Leukemia sample work flow

Hematology Clinic FIMM

1 h 1 day

Biobank

DSRT

4 days

NGS

2-3 weeks

Proteomics

2 days

FIMM

Sample collection • Bone marrow aspirate

• Peripheral blood

• Skin biopsy

Sample processing • Mononuclear cell separation

• Protein lysates

• DNA extraction

• RNA extraction

Sample analysis • Drug screening

• Phospho-protein analysis

• Whole genome/exome sequencing

• RNA sequencing


Augmenting molecular drug discovery/repurposing

• Genome/transcriptome-driven

Molecular profiling of leukemic cells

Disease-associated genomic variants

Molecularly targeted therapies

- Rarely directly clinically actionable

- 3-10+ years

Drug response profiling of leukemic cells

› Drug-response phenotype-driven

- Often directly clinically actionable

- Unbiased - Drug repurposing

- 1-5+ years

Disease-associated drug responses

Molecularly dissection of drug response

Molecularly targeted therapies


1408

Sele

ctiv

e D

SS

−10

010

20

Dexam

eth

asone

Meth

ylp

rednis

olo

ne

Ponatin

ib

AZ

D8055

Pre

dnis

olo

ne

FK

−866 H

Cl

Axitin

ib

OS

I−027

Vin

cris

tine

Moceto

tinosta

t

Plic

am

ycin

GD

C−

0980

Prim

a−

1 M

et

Nera

tinib

PF

−04691502

Dan

usertib

Tra

metin

ib

TA

K−

733

Entin

osta

t

Clo

fara

bin

e

Lesta

urtin

ib

PX

−866

Alv

ocid

ib

Dovitin

ib

Dasatin

ib

Masitin

ib

Sara

catin

ib

Dasatin

ib

Alis

ertib

SB

743921

AZ

D 7

762

Vandeta

nib

AT

9283

MK

1775

SN

S−

032

Bosutin

ib

PF

477736

UC

N−

01

Criz

otin

ib

Afa

tinib

Drug sensitivity and resistance testing

300 dose response curves

Oncology

collection

5 conc. 10,000-

fold conc. range

37°C,

72 h

Assay & detection

reagents

(CellTiter-Glo for viability)

Leukemic

patient cells

Calculate activity

Compare to controls

Novel biological

understanding of the

disease

Find responding

cells/patients.

Repositioning

Translate

directly to

patients

-9 -8 -7 -6 -5

0

25

50

75

100 control

AML

Conc (log M)

% v

iab

ility

4 day turnaround


Assembling a concise and comprehensive functional screening collection

• Currently 306 oncology-focused active substances/drugs – Approved small molecule oncology substances – Oncology-related approved substances – Major investigational oncology compounds – Probe compounds with unique and cancer-relevant activities

• 135 approved • 171 investigational

• 65 conventional chemotherapeutics • 20 hormone therapy drugs • 119 kinase inhibitors • 33 epigenetic/differentiating drugs • 55 other targeted drugs • 10 immunosuppressants


Unsupervised hierarchical clustering creates

functional taxonomy of patients AML patient samples can be divided into

five broad groups

Controls and patient samples

Dru

gs

693

105

4

10

67

134

6

153

4

207

1

220

0

199

3

209

5

106

4

114

5_

3

800

393

_1

393

_3

114

5_

6

114

5_

1

718

128

0

149

7

600

_3

252

_3

252

_2

186

7

784

_2

784

_1

250

600

_1

252

_5

252

_1

560

_1

560

_9

600

_2

169

0

BE

RG

208

0

Temozolomide Thio-TEPA Thioguanine Bryostatin 1 Enzastaurin Vatalanib Cediranib Plerixafor Tamoxifen Prima-1 Met Fasudil Lenalidomide Vemurafenib Chlorambucil Dexamethasone Prednisolone Methylprednisolone Linsitinib Navitoclax Tipifarnib Azacitidine Gefitinib Doramapimod Erlotinib Dovitinib Gandotinib Sotrastaurin Ruboxistaurin Capecitabine YM155 Pentostatin Crizotinib Obatoclax Decitabine Tretinoin Bexarotene Bortezomib Clofarabine Cladribine Midostaurin UCN-01 Serdemetan Carboplatin Uracil mustard Selumetinib Pimasertib Trametinib Refametinib Tacrolimus Tofacitinib Ruxolitinib 2-methoxyestradiol Gemcitabine Olaparib Motesanib Rucaparib Tacedinaline Nilotinib Plicamycin Cytarabine Ponatinib Lestaurtinib Temsirolimus Tanespimycin Alvespimycin BIIB021 NVP-AUY922 Entinostat Panobinostat Belinostat CUDC-101 Vorinostat AZD8055 PF-04691502 OSI-027 BMS-754807 Dactolisib MK-2206 GS-1101 Regorafenib Pictilisib Methotrexate Floxuridine Everolimus Sirolimus Imatinib Saracatinib Dasatinib Masitinib Tivozanib Axitinib Tandutinib Sorafenib Pazopanib Canertinib Fluorouracil Busulfan Vandetanib Amonafide Mitoxantrone Foretinib Etoposide Bleomycin MGCD-265 Sunitinib Daunorubicin Valrubicin Doxorubicin Idarubicin Teniposide Dactinomycin Quizartinib Fludarabine Cyclophosphamide Streptozocin Iniparib Fingolimod Nelarabine Raloxifene Vismodegib Exemestane Mitomycin C Carfilzomib

D

19

93

20

95

10

64

11

45

_3

39

3_

1

39

3_

3

80

0

11

45

_1

11

45

_6

71

81

28

0

14

97

60

0_

3

25

2_

2

25

2_

3

18

67

78

4_

1

78

4_

2

25

060

0_

1

252

_5

25

2_

1

56

0_

1

56

0_

9

60

0_

21

69

04

70

10

01

NavitoclaxRuxolitinibDexamethasoneMEKiPI3K/mTORiQuizartinib

Topoisomerase IIi

II III IV VI

FLT3-ITD

WT1

TP53

NRAS/KRASDNMT3A

PTPN11

RUNX1

NUP98-NSD1

MLL-X fusions

ETV6-NTRK3

NPM1

M1M2FAB subtypes M2 M2M2M2 M5M5 M5 M5 M5M5M5M5 M5M5M5 M1M1

KIT

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

$" & +( ' ! $ ++)

$ " ) ) ) *& & #

DasatinibSunitinib

IDH1/IDH2

Figure 3. Pemovska et al

Adverse karyotype

D D* R R D* D R D* R R D R R R D R R R R R R R R R D RDisease stage



Challenge with big data management, rapid integration and interpretation to give feedback to the clinic

Drug sensitivity testing Exome sequencing

Gene expression Phosphoproteomics Fusion genes

Integration & Interpretation &

Feedback to clinic (4 days to 2 weeks)

Hematology Research Unit Helsinki 09/02/2015

In personalized medicine, the patient (person) is at the center


DRUG REPURPOSING FOR CHEMOREFRACTORY AML

Example I


Patient case I: FHRB.600

• 50-year-old male, previously healthy

• AML FAB M5, normal karyotype, FLT3-ITD

• Failed 3 consecutive induction therapies

• Febrile, pancytopenia, WHO 3

• Molecular profiling: exome and transcriptome sequencing and targeted phospho-proteomics of serial samples

• DSRT: Selection of an off-label drug-combination as per compassionate use based on ex vivo drug screening results


FHRB.600: NUP98-NSD1 fusion

NUP98 (chr 11) NSD1 (chr 5)



600: 1st DSRT

0 5 10 15 20

Pazopanib - TKI (VEGFR)

Axitinib - TKI (VEGFR)

Sorafenib - TKI (B-Raf/VEGFR)

Temsirolimus - Rapalog*

Navitoclax - Bcl-2/Bcl-xL inhibitor

Alvespimycin - HSP90 inhibitor

Mitoxantrone - Topoisomerase II inhibitor

PF-04691502 - mTOR/PI3K inhibitor

Foretinib - TKI (VEGFR2/MET)

Quizartinib - Flt3 inhibitor

Cytarabine - Antimetabolite

Entinostat - HDAC inhibitor

NVP-AUY922 - HSP90 inhibitor

Lestaurtinib - Broad TKI

Dasatinib - Broad TKI*

Tivozanib - TKI (VEGFR)

Tanespimycin - HSP90 inhibitor

Sunitinib - Broad TKI*

BIIB021 - HSP90 inhibitor

Ponatinib - Broad TKI

sDSS

-9 -8 -7 -6 -5

0

25

50

75

100 control

AML

Conc (log M)

% v

iab

ility

Differ DSS

DSS: Drug sensitivity score =

(AUC/Total Area)*100/log(100-

min)

Ex vivo DSRT results

Pemovska, Kontro, et al. Cancer Discovery, 2013


Treatment response

0 10 20 30 40 500

20

40

60

0

1

2

3

Days from start of therapy

Bo

ne

ma

rro

w b

las

ts +

pro

mo

no

cy

te c

ou

nt

(%)

Ne

utro

ph

il co

un

t (10

9/L)

Dasatinib

Sunitinib

Temsirolimus

Infection

Sensitive sample Exome sequencing CNA RNAseq Expression Phophoproteomics

Resistant sample Exome sequencing CNA RNAseq Expression Phophoproteomics

Bone marrow blast cell %

Neutrophil count


0 5 10 15 200

5

10

15

20

600_2 sDSS

600_3 s

DS

S

BIIB021

Tivozanib

Tanespimycin

Ponatinib

Sunitinib

Dasatinib

Lestaurtinib

Entinostat

Cytarabine

Alvespimycin

Pazopanib

PF-04691502

Quizartinib

AxitinibSorafenib

Temsirolimus

Bleomycin

VorinostatEtoposide

Valrubicin

Loss of ex vivo drug sensitivity

Pre treatment Post treatment



Pretreatment Postreatment



Temsirolimus

Dasatinib

Sunitinib


Clonal evolution

A B

C D

0 5 10 15 200

5

10

15

20

BIIB021

Tivozanib

Tanespimycin

Ponatinib

Sunitinib

DasatinibLestaurtinib

EntinostatCytarabine

Alvespimycin

Pazopanib

PF-04691502

Quizartinib

AxitinibSorafenib

TemsirolimusBleomycin

VorinostatEtoposide

Valrubicin

0 50 100 150 180 200 2200

20

40

60

80

100

0.0

2.5

5.0

Chemotherapy

600_2 Selective DSS

60

0_

3 S

ele

ctive

DS

S

Chemotherapy Das-Sun-Tem

600_0 600_2 600_3

%

10

E9

Chemotherapy Das-Sun-Tem Das-Tem

Bone marrow blasts (%)

Neutrophils (x10E9)

600_2 600_30

5

10

15

20

600_2 600_3

600_2 600_30

5

10

15

Dasatinib Sunitinib

Temsirolimus

Se

lective

DS

S

Se

lective

DS

S

Se

lective

DS

S

0

5

10

15

20

NUP98-NSD1

600_0 600_1 600_3600_2

90%

5%

5%

40%

30%

12%

18%

70%

10%

20%FLT3-ITD #1

WT1 #2

FLT3-ITD #2

WT1 #1 WT1 #4

WT1 #3

Pemovska, Kontro, et al. Cancer Discovery, 2013

Novel candidate combination therapy for NUP98-NSD1-driven AML


Functional investigation of NUP98-NSD1/FLT3 AML

Biobanked patient material: • 3 patients

Cell models: • Ba/F3 (murine, C3H, pro-B-cell,

BM)

o Empty vector

o NUP98-NSD1 variant 1


• 32D (murine, C3H, myeloblast-like, BM)

o Empty vector



• Balb/c BM Lin (-)

o Empty vector

o NUP98-NSD1 (pre)

o NUP98-NSD1/FLT3-ITD (pre)

o NUP98-NSD1/FLT3-ITD (M3)

• BL/6 BM Lin (-)

o Empty vector

o NUP98-NSD1/FLT3-ITD (M1)

Transgenic model

• In collaboration with Prof. Juerg Schwaller (Basel)


DRUG REPURPOSING – T315I MUTATED CML

Example II


Patient case: FHRB.1408 (CML BC T315I+)

• February 2012: male aged 35 diagnosed of CML in lymphatic blast phase

• Imatinib 600 mg monotherapy, hematologic response

• April 2012: hematologic relapse, switch to dasatinib 140 mg x1, no response. HD chemorherapy => partial response

• BCR-ABL1 KD sequencing: BCR-ABL1T315I 100%

• August 2012: AlloHSCT

• Current status (May 2014): BM in CR

• Exome sequencing (100x): missense mutations in

RUNX1, ABL1, NRK, ABCA13


20 most selective drugs

PI3K/mTOR KI

Namt inhibitor

VEGFR-1/2/3 TKI

mTORC2/mTORC1 inhibitor

p53-activating drug PI3K/mTOR KI

Multikinase inhibitor

EGFR/HER-2 inhibitor

PI3K/mTOR KI Aurora A/B/C KI MEK-inhibitor MEK-inhibitor

HDAC-inhibitor


0 5 10 150

5

10

15

20

25

RQ

-PC

R fo

r T

31

5I (

%)

Day

Clinical translation

Axitinib 5 mg x2

Confirmation of results from similar samples in the FHRB biobank Formal Phase I/II study; combination therapy; MOA – in partnership with Pfizer New, already approved drug for T315I-mutated leukemia in 2-3 years

Pemovska T, et al. Nature 2015 (in press)


Multiple myeloma sample analysis

DSRT ex vivo drug sensitivity

Exome sequence SNVs, CNVs

RNA sequence gene expression profile, expressed

fusion genes, small RNAs, lncRNAs

Methyl sequence epigenomic profile


CD138+ cell selection

DNA (CD138+ cells and skin biopsy)

RNA (CD138+ cells)

Drug sensitivity testing

Next generation DNA and RNA sequencing

Clinical data Finnish

Hematology Registry and

Biobank

Bone marrow aspirate

Biobanked samples


Conclusions

• Disease-specific deep biobanks are key components

in the development of personalized cancer care

• ISM strategy is one way to implement personalized

cancer therapy

• Novel drug candidates for difficult to treat cancers

• Drug repurposing: patient selection for formal Phase I-II

studies. Patenting opportunities

• Formal Phase II clinical study necessary to validate

clinical usefulness (in collaboration with

pharmaceutical companies and research consortia)


Tiina Vesterinen Kimmo Pitkänen Caroline Heckman Janna Saarela Kyösti Sutinen Olli Kallioniemi

Steering group Members Patients

Kari Aranko Tiina Wahlfors Henna Jalovaara Outi Huoponen Eeva Mainio

http://www.hematology.fi/fhrb

FHRB

Suomen Syöpäpotilaat ry Finnish Cancer Patient Association

Aimo Strömberg Mika Peltovaara Timo Laukkio


FIMM

Chemical Systems Biology

Krister Wennerberg

Tea Pemovska

Arjan van Adrichem

Computational Systems Biology

Tero Aittokallio

Petteri Hintsanen

Agnieszka Szwajda

Bhagwan Yadav

Technology Center

Janna Saarela

Evgeny Kulesskiy

Laura Turunen

Anna Lehto

Ida LIndenschmidt

Pekka Ellonen

Maija Lepistö

Sonja Lagström

Sari Hannula

Pirkko Mattila

Aino Palva

Leukemia Network Helsinki

HUCH/HRU Kimmo Porkka

Satu Mustjoki

Pekka Anttila

Mika Kontro

Erkki Elonen

Hanna Koskela

Mette Ilander

Emma Anderson

Paavo Pietarinen

Jaakko Vartia

Minna Lehto

Mervi Saari

Kuopio University Central Hospital

Raija Silvennoinen

Turku University Central Hospital

Tuija Lundán

Tampere University Central Hospital

Hannele Rintala

Tero Pirttinen

Marja Sankelo

University of Bergen

Bjørn Tore Gjertsen

Personalized Cancer Medicine

Caroline Heckman

Jonathan Knowles

Samuli Eldfors

Riikka Karjalainen

Jarno Kivioja

Ashwini Kumar

Heikki Kuusanmäki

Muntasir Mamun Majumder

Alun Parsons

Minna Suvela

Individualized Systems Medicine

Olli Kallioniemi

Taija af Hällström

Henrik Edgren

Poojitha Kota Venkata

Disha Malani

John Patrick Mpindi

Astrid Murumägi

Päivi Östling

Maija Wolf

The value of Biobanks for personalized medicine in acute Leukemia

Documents