Drug Studies Using Mouse Drug Studies Using Mouse ModelsModels

Melinda Hollingshead, D.V.M., Ph.D.Melinda Hollingshead, D.V.M., Ph.D.

Biological Testing BranchBiological Testing Branch

Developmental Therapeutics ProgramDevelopmental Therapeutics Program

Division of Cancer Treatment and DiagnosisDivision of Cancer Treatment and Diagnosis

NCINCI

Why should the taxpayers fund Why should the taxpayers fund this type of work?this type of work?

Cancer is the second leading cause of death in America. It accounts for one of every four deaths. More than 565,000 - that’s more than 1,500 people a day - die annually from cancer. Close to 1.4 million new cases are diagnosed each year. This estimate does not include

pre-invasive cancer or the more than 1 million cases of non-melanoma skin

cancer expected to be diagnosed annually.

Why do we use mouse Why do we use mouse models?models?

geneticsgenetics progressionprogression preventionprevention diagnosticsdiagnostics therapeuticstherapeutics physiologically complex vs in vitro studiesphysiologically complex vs in vitro studies cost effectivecost effective develop human clinical protocolsdevelop human clinical protocols

To prevent, diagnose and treat human disease

What models are available?What models are available?

Spontaneous Spontaneous Virus-inducedVirus-inducedTransgenicTransgenicKnock-out/inKnock-out/inInduced/carcinogensInduced/carcinogensTransplantedTransplanted

Spontaneous ModelsSpontaneous Models

Natural history is most “normal”Natural history is most “normal”Random tumorsRandom tumorsDifficult to predict occurrenceDifficult to predict occurrenceMonitor through lifespan ~ 2 yr in mice Monitor through lifespan ~ 2 yr in mice Large number of animals may be Large number of animals may be

required for a small experimentrequired for a small experimentTumors may be heterogeneousTumors may be heterogeneousSignificant animal holding space neededSignificant animal holding space needed

These allow the study of the biological history of natural disease. They can be applied to many types of studies.

Virus-Induced ModelsVirus-Induced Models

Classic models of virus-induced leukemia:Classic models of virus-induced leukemia: RauscherRauscher MoloneyMoloney LP-BM5LP-BM5 FriendFriend

Non-leukemic tumors including:Non-leukemic tumors including: Mammary tumors due to MMTV Mammary tumors due to MMTV Thymomas of AKR miceThymomas of AKR mice

High tumor occurrence rateHigh tumor occurrence rate Predictable time to tumor developmentPredictable time to tumor development Well-characterized disease states/natural historyWell-characterized disease states/natural history Do not accurately reflect the natural history of most Do not accurately reflect the natural history of most

human tumorshuman tumors

Early studies of cancer, both as a disease state and for therapeutic intervention, used virus-induced models. These models continue

to be applied in a variety of studies.

Transgenic ModelsTransgenic Models

Many available through commercial and collaborative Many available through commercial and collaborative arrangementsarrangements

http://emice.nci.nih.gov/emice/mouse_models http://mouse.ncifcrf.gov/ http://jaxmice.jax.org/query/f?p=205:1:1510299228434659165

Not all patents have expired - check with OTT regarding legalitiesNot all patents have expired - check with OTT regarding legalities Remember MTAs are needed to receive or ship material that may Remember MTAs are needed to receive or ship material that may

have patent or licensing rights. This also applies to mice. have patent or licensing rights. This also applies to mice.

Rate of tumor occurrence may be lowRate of tumor occurrence may be low Time to tumor occurrence may be difficult to predictTime to tumor occurrence may be difficult to predict Breeding schemes may be complex e.g., 3, 4 or more Breeding schemes may be complex e.g., 3, 4 or more

intermediate genetic crossesintermediate genetic crosses Genetic error(s) are generally well characterizedGenetic error(s) are generally well characterized Disease may follow a more natural course e.g., time to Disease may follow a more natural course e.g., time to

significant tumor burden may more accurately mimic the significant tumor burden may more accurately mimic the human; lesions likely relevant to their site of occurrencehuman; lesions likely relevant to their site of occurrence

Transgenic models aid our understanding of the genetics of cancer and are being pursued as models for intervention.

Induced/CarcinogensInduced/Carcinogens

EpithelialEpithelial tumorigenesistumorigenesis 7,12-dimethylbenz[a]anthracene [initiator mutagen] followed 7,12-dimethylbenz[a]anthracene [initiator mutagen] followed

by multiple applications of 12-by multiple applications of 12-OO-tetradecanoylphorbol-13--tetradecanoylphorbol-13-acetate [pro-inflammatory]acetate [pro-inflammatory]

GI tumorigenesisGI tumorigenesis 1,2 dimethlhydrazine-2-HCl1,2 dimethlhydrazine-2-HCl azoxymethaneazoxymethane

Sarcoma inductionSarcoma induction MethylcholanthreneMethylcholanthrene

Lung tumorigenesisLung tumorigenesis nitrosamine 4- (methyl-nitrosamino)-1-(3-pyridyl)-1-butanonenitrosamine 4- (methyl-nitrosamino)-1-(3-pyridyl)-1-butanone

These models aid studies of the local, systemic, and environmental factors that influence tumor susceptibility,

growth, and progression.

Induced/CarcinogensInduced/Carcinogens

Time to tumor development may be Time to tumor development may be variablevariable

Mimics some human diseases very Mimics some human diseases very wellwell

Requires technical work with known Requires technical work with known carcinogenic agents thus special carcinogenic agents thus special handling/equipment/facilities/training handling/equipment/facilities/training requiredrequired

Provides the entire natural history of Provides the entire natural history of tumor development to studytumor development to study

Transplanted ModelsTransplanted Models

Tissue SourceTissue Source Syngeneic - same inbred strain – e.g., B16 - same inbred strain – e.g., B16

tumors in C57Bl/6 micetumors in C57Bl/6 mice Allogeneic - same species, different strain - same species, different strain

(genetically diverse) – e.g., M5076 sarcomas in (genetically diverse) – e.g., M5076 sarcomas in athymic mice produces an allograathymic mice produces an allograftft

Xenogeneic - different species e.g., human, rat - different species e.g., human, rat or dog tumors grown in immunocompromised or dog tumors grown in immunocompromised mice produces a xenogramice produces a xenograftft

Implant SiteImplant Site OrthotopicOrthotopic HeterotopicHeterotopic

EndpointsEndpoints

These models are commonly applied in studies of genetics, diagnostics, and medical interventions.

Transplanted ModelsTransplanted Models

Easy to control time of tumor occurrenceEasy to control time of tumor occurrenceMany tumor types/lines availableMany tumor types/lines availableWell accepted modelsWell accepted modelsDo not accurately recapitulate human diseaseDo not accurately recapitulate human diseaseMetastatic lesions are difficult to findMetastatic lesions are difficult to findTumor growth rates may preclude multiple Tumor growth rates may preclude multiple

treatment cyclestreatment cycles

production In vitro studies

efficacy trials

Discovery & DevelopmentDiscovery &

Development

New therapy

Pharmacology & toxicology

• Human Tumors– Hollow fiber– Subcutaneous– Intravenous– Intraperitoneal– Orthotopic

• Mammary fat pad• Intracranial• Intrarenal• Intrahepatic• Intracecal

• Rodent Tumors– Subcutaneous– Intravenous– Intraperitoneal– Orthotopic– Metastatic– Transgenic– Knock-in/out– Induced– Spontaneous

In Vivo Efficacy ModelsIn Vivo Efficacy Models

• Tumor sources – Cells cultured in vitro– Serially passaged tumor – Cryopreserved tumor

Issues faced during Issues faced during Efficacy EvaluationsEfficacy Evaluations

ModelVehicle, formulation, stability

Dose, route and scheduleExperimental protocolEndpoints

What are you assessing? What are you assessing? Which type of model is most appropriate?Which type of model is most appropriate? Is the treatment designed to:Is the treatment designed to:

impact the tumor biochemically, e.g., impact the tumor biochemically, e.g., cytotoxiccytotoxic

impact the tumor genetically, e.g., impact the tumor genetically, e.g., modulatormodulator

impact the stroma e.g., vasculatureimpact the stroma e.g., vasculatureimpact the immune systemimpact the immune systemact as an adjuvantact as an adjuvantsynergize with known drugssynergize with known drugsinteract with specific proteinsinteract with specific proteins

ModelModel Selection Selection

Model Model SelectionSelection

A

0

200

400

600

800

1000

1200

1400

1600

1800

22 25 28 31 35 39 43 46 49 54 61

Day Post-Implantation

Med

ian T

umor

Weig

ht

B

-200

0

200

400

600

800

1000

1200

1400

1600

1800

13 19 22 26 29 33 36 39 46 50 53 56 60 63

Day Post-implantation

Med

ian

Tum

or W

eigh

t

B

-5 00

0

5 00

1 000

1 5 00

2 000

1 3 1 9 2 2 2 6 2 9 3 3 3 6 3 9 4 6 5 0 5 3 5 6 6 0 6 3

D a y P o st-im p la n t a tio n

100% sesame oil 45 mg/kg tamoxifen

22.5 mg/kg tamoxifen 11.25 mg/kg tamoxifen

estrogen independent

estrogen dependent

Experimental ProtocolExperimental Protocol

When will treatment start?When will treatment start?When will treatment end?When will treatment end?Will samples be collected for ex vivo Will samples be collected for ex vivo

evaluation?evaluation?Will tumors be monitored visually, by Will tumors be monitored visually, by

imaging techniques?imaging techniques?What will terminate the experiment, What will terminate the experiment,

i.e., what are the endpoints?i.e., what are the endpoints?How will the animals be randomizedHow will the animals be randomized

What is published/prior knowledge? What is published/prior knowledge? What is proposed/expected What is proposed/expected

mechanism?mechanism? How much exposure is required for How much exposure is required for

effect?effect? Is the material soluble/stable in Is the material soluble/stable in

aqueous solution?aqueous solution? What routes of administration are What routes of administration are

technically feasible?technically feasible? Options Options

What is the maximum tolerated dose (MTD)?What is the maximum tolerated dose (MTD)? IP, IV, SC, PO?IP, IV, SC, PO? QDx?; BIDx?; TIDx?QDx?; BIDx?; TIDx?

Dose, Route, ScheduleDose, Route, Schedule

Kaplan-Meier Plot of mice bearing intraperitoneal OVCAR-5 human ovarian cancer xenografts. Mice were treated with vehicle (PBS or lipid vehicle) with a therapeutic solubilized in each of the vehicles. Note that the lipid vehicle alone was as effective in improving survival as was the therapeutic prepared in the lipid vehicle and it was more effective than the therapeutic prepared in PBS.

0 20 40 60 800

20

40

60

80

100

PBS lipid vehicle

treatment in PBS treatment in lipid vehicle

day

Pe

rce

nt

su

rviv

al

EndpointsEndpoints

Tumor sizeTumor sizeWeight lossWeight lossTime to sacrificeTime to sacrificeImagingImagingPre-defined time of terminationPre-defined time of terminationTime post-treatmentTime post-treatment

PC-3 Tumor

12

0

10

0

80

60

40

20

0PS-341 (mg/m2)

0.9 1.8Vehicle

20

S A

cti

vit

y (

% V

eh

icle

)

120

100

80

60

40

20

0

0.3 0.9 3.0 9.0VCTLPS-341 (mg/m2)2

0S

Acti

vit

y (

% V

eh

icle

)

Mouse WBC

GENOMICSTumor RNA

BIOMARKER

SURROGATEBIOMARKER

DRUG ANALYSIS

Plasma

In Vivo StudiesIn Vivo Studies

IHC IHC Original tumor and each passageOriginal tumor and each passage Antibodies for specific gene productsAntibodies for specific gene products

Molecular analysis Molecular analysis genetics : NGS, RNA-Seq, microarray (LCM), RT-genetics : NGS, RNA-Seq, microarray (LCM), RT-

PCRPCR proteins/phosphoproteins: reverse arrays, IHCproteins/phosphoproteins: reverse arrays, IHC biomarkers: serum proteomicsbiomarkers: serum proteomics

Analysis of host responseAnalysis of host response immune system responseimmune system response

(syngeneic v.s. immunosuppressed)(syngeneic v.s. immunosuppressed) Angiogenesis (MFP v.s. sc)Angiogenesis (MFP v.s. sc)

Stem cell searchStem cell search OCT

IHC

Snap-Frozen (RNA)

Harvest and freeze cell suspension

Tools for AnalysisTools for Analysis

Colo-829 SC Tumors Collected Post Dose 4

Drug Vehicle Q12hx5D PO

Topotecan 5mg/kg QDx5 IP

ABT-888 25mg/kg Q12hx5D PO

Poly-ADP-ribose (Trevigen)

Min 20.0Max 400.04 min exp

Mou

se 5

Mou

se 9

Mou

se 1

1

Mou

se 1

3

Mou

se 1

5

Mou

se 1

8

AA

XR

2-7

#1

Mou

se 3

Mou

se 7

Mou

se 1

6

Mou

se 1

7

Mou

se 1

9

Mou

se 2

1

AA

XR

2-7

#1

Mou

se 1

Mou

se 2

Mou

se 4

Mou

se 6

Mou

se 8

Mou

se 1

0

AA

XR

2-7

#1

200

140

100

80

513 517 521

b-Actin

Topotecan 5mg/kg QDx5 IP

ABT-888 25mg/kg Q12hx5D PO

Poly-ADP-Ribose(Trevigen)Min 20.0

Max 400.04 min exp

Mou

se 2

6

Mou

se 2

7

Mou

se 3

0

Mou

se 3

4

Mou

se 3

6

AA

XR

2-7

#1

Mou

se 1

2

Mou

se 1

4

Mou

se 2

5

Mou

se 2

8

Mou

se 3

2

Mou

se 3

3

AA

XR

2-7

#1

519 523

Drug Vehicle Q12hx5D PO

Mou

se 2

0

Mou

se 2

3

Mou

se 2

4

Mou

se 2

9

Mou

se 3

1

Mou

se 3

5

AA

XR

2-7

#1

200

140

100

80

515

b-Actin

Colo-829 SC Tumors Collected Post Dose 10

Many Models – Many OptionsMany Models – Many Options

With few exceptions, every rodent model, even if conducted with hundreds of experimental

mice, represents a single patient. Interpreting the preclinical results and applying these

outcomes to human clinical trials continues to prove challenging for those charged with

translating the preclinical experience into viable drug candidates.

Statistically valid model assessing relevant endpoints on an optimal

schedule with clinically appropriate doses.

PROBLEMSPROBLEMS

The appropriateness of animal models to The appropriateness of animal models to identify, qualify and promote new identify, qualify and promote new therapies for cancer has been under therapies for cancer has been under review, and in some ways under attack, for review, and in some ways under attack, for many years. Continuing concerns about many years. Continuing concerns about the failure rate of agents being sent to the the failure rate of agents being sent to the clinic has led to a flurry of publications on clinic has led to a flurry of publications on the irreproducibility of published preclinical the irreproducibility of published preclinical data and their over-prediction of activity. data and their over-prediction of activity.

What can you do?What can you do? 1. use well powered animal studies1. use well powered animal studies 2. reproduce your own data2. reproduce your own data 3. have 2 separate operators generate the data3. have 2 separate operators generate the data 4. provide adequate details in publications4. provide adequate details in publications 5. don’t over-interpret your data5. don’t over-interpret your data 6. stage tumor studies correctly6. stage tumor studies correctly 7. don’t selectively use/present your data7. don’t selectively use/present your data 8. remember the clinical situation and what can be 8. remember the clinical situation and what can be

assessed in manassessed in man

Clinical Endpoints In ManClinical Endpoints In Man

Toxicity -Toxicity - W What harmful effects are induced?hat harmful effects are induced? Tumor response - Tumor response - Does the cancer respond to the treatment?Does the cancer respond to the treatment?

Biomarker modulation as a measure of the effect of a Biomarker modulation as a measure of the effect of a treatment that may correlate with a traditional clinical treatment that may correlate with a traditional clinical endpoint (PFS; TR) endpoint (PFS; TR)

Progression-free survival (stable disease)Progression-free survival (stable disease) Tumor regressionTumor regression Statistically significant improvement in survivalStatistically significant improvement in survival

Survival - Survival - how long does the person live?how long does the person live? Quality of life - Quality of life - how does the treatment affect a person's how does the treatment affect a person's

overall enjoyment of life and sense of well being?overall enjoyment of life and sense of well being?

Progression, Stable, RegressionProgression, Stable, Regression