Progressing the Development of AZD2811 Nanoparticles for Intravenous AdministrationRebecca Booth – Senior Scientist, Pharmaceutical Technology and DevelopmentDrug Delivery and Formulation Summit March 2019

Progressing the Development of AZD2811 Nanoparticles

Background

CMC Development

- formulation optimisation, physical & chemical composition

- developing a robust manufacturing process

- biopharmaceutics & critical quality attributes

References:

1. Mortlock AA, Foote KM, Jung FH et al., Discovery, synthesis, and in vivo activity of a new class of pyrazoloquinazolines as selective inhibitors of aurora B kinase. J Med Chem, 2007 50(9):2213-24.

2. Young HS, Shin E, Wang H, Nolan J, Low S et al, A novel in situ hydrophobic ion pairing (HIP) formulation strategy for clinical product selection of a nanoparticle drug delivery system. J Control Release. 2016; 229:106-119.

3. Ashton S, Taylor P, Curtis N et al. AZD1152HQPA AccurinTM nanoparticles inhibit growth of diffuse large B-cell lymphomas and small cell lung cancer. Cancer Res. 2015; 75:3102.

4. Ashton S, Song YH, Nolan J, Cadogan E, Murray J et al., Aurora kinase inhibitor nanoparticles target tumors with favorable therapeutic index in vivo. Sci Transl Med 2016; 325ra17-325ra17.

Targeting AURKB with AZD1152 (Barasertib) and AZD2811 Nanoparticle

•AZD1152 is a dihydrogen phosphate prodrug1

•In plasma, AZD1152 is rapidly converted into AZD2811 that shows up to 1000-fold higher selectivity for aurora kinase B

over aurora kinase A1

•AZD2811 is encapsulated in a novel nanoparticle delivery system2, 3, 4

AZD1152(barasertib, prodrug of AZD2811)

AZD2811(AZD1152-hQPA)

AZD2811 nanoparticle

PLA-PEG copolymer

matrix with PEG blocks

oriented toward particle

surface

Encapsulated AZD2811

3

AZD2811 Nanoparticle has the Potential to Overcome the Limitation

of AZD1152

AZD1152 demonstrated positive Proof of Concept in AML. Despite this, the program was stopped

due to:

• The 7-day infusion regimen limited the

indications to those where patients were already

hospitalised, such as AML patients

• The dose-limiting, mechanism-related bone

marrow toxicity meant that efficacious doses could

only be achieved in patients with haematological

cancers

These clinical limitations may be overcome by having:

• A slow release profile where a single infusion gives multiple days efficacious cover in an

outpatient setting

• An improved efficacy with the potential to establish efficacious doses in solid tumour

settings

LDAC: low-dose cytarabine

Reference: Kantarjian HM, Martinelli G, Jabbour EJ, et al. Stage I of a phase 2 study assessing the efficacy, safety, and tolerability of barasertib (AZD1152) versus low-dose cytosine arabinoside in

elderly patients with acute myeloid leukemia. Cancer. 2013; 119(14):2611-9

AZD1152 vs. LDAC in elderly patients unfit

for intensive chemotherapy (50 cf 25 pts)

4

ACCURINS®

5

• Drug encapsulated in PLA polymeric matrix with

PEG stealth layer

• Drug release via Diffusion

~100nm

Nanoparticles with Controlled Drug Release Rates

Stealth and

Protective

Layer

PEG• Hydration shell

• Protects against immune detection

• Long circulation time

Therapeutic

Payload

Broad range of therapeutic payloads• Small molecule, peptides, proteins,

nucleic acids

• No chemical modification

Controlled-

ReleasePolymer Matrix

Polylactic acid (PLA)• FDA approved polymer

• Non-covalently entraps payload

• Regulated release of payload over days

D a y s p o s t f i r s t d o s e

Tu

mo

ur

V

ol

um

e

(m

m3

)

0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0

0 . 0 0 0 1

0 . 0 0 1

0 . 0 1

0 . 1

1

1 0

P l a c e b o n a n o p a r t i c l e I V

B a r a s e r t i b ( A Z D 1 1 5 2 ) 2 5 m g / k g S C d 1 - 4

A Z D 2 8 1 1 n a n o p a r t i c l e 2 5 m g / k g I V d 1 / d 3

AZD2811 nanoparticle shows improved efficacy over AZD1152 at equivalent dose intensity in SCLC models

6

SC61 SCLC PDX

• Significant tumour regression with >80% tumours

still not visible 112 days after a single cycle of

AZD2811 nanoparticle

• Regressions are more durable than AZD1152

(barasertib)

99% regression (p<0.001);

9/10 mice with no tumour at day 39

85% regression at day 11

(p<0.001) then tumours re-grow

Ashton et al, AACR, 2017 - Development of AZD2811, an aurora kinase B inhibitor, incorporated into an AccurinTM nanoparticle for use in haematological and solid cancers

Progressing the Development of AZD2811 Nanoparticles

7

Manufacturing of AZD2811

Formulation optimisation & physicochemical characterisation

Formulation Optimisation of AZD2811 Nanoparticle

Goals• To increase drug loading to > 12% while

maintaining slow release rate

• Improve process efficiency

Benefits• Reduced number of vials for patients

• Reduced excipient load

• Reduced development costs/cost of goods

* Song et al, J Controlled Release, 2016, 229, 106

Hydrophobic Ion

Pairing Approach*

Drug loading 15%

Encapsulation

Efficiency

71%

Time 50%

release

120 hours

AZD2811 Drug Formulation & Product

9

• Presented as a frozen suspension for Ph1 & 2

• Lyophile will be investigated for commercial product

Component Function

AZD2811 API (~20% nanoparticle)

PLA-PEG Hydrophobic PLA (16kDa) to control

release

Hydrophilic PEG (5kDa) as stealth layer

Pamoic Acid Functional excipient, forms hydrophobic

ion pair with API

AZD2811:Pamoic 1:0.5

Main Components & Function of Nanoparticle

Understanding Physical & Chemical Composition

10

• Raman spectra confirms formation of AZD2811–pamoate 2:1 salt

Nanoparticles encapsulating

AZD2811 and pamoic acid

AZD2811

Pamoate

2:1 (premade salt)

AZD2811

Pamoic acid

Raman Shift cm -1

Nanoparticle Target Product Profile

• Surface characteristics

• Drug Release

rate/profile

• Particle size & PDI

• Morphology

• Free drug concentration

Process aids

Emulsion composition

Particle formation

Process controls (pressure, temp. etc)

Purification

Block architecture

PEG and PLA Mn

Homo-PLA, free PEG

Residual monomer, solvent

Production process

Potential Critical Quality AttributesPhysical & Chemical CompositionRelease-controlling Polymer

+/- targeting ligand

Drug loading

Counter-ion

Particle size distribution

Surface charge

Drug release rate

Nanoparticle Properties impact the in vivo

performance, both efficacy and safety

• Biodistribution – nanoparticle circulation time

and clearance, localized accumulation

• Release and clearance of drug

Critical material attributes

Size: ~100 ± 30 nm

Drug load ~ 15%

Similar in vitro

Release profile

Progressing the Development of AZD2811 Nanoparticles

12

Developing a robust manufacturing process

Manufacturing Process for AZD2811 nanoparticle

Size: ~100 ± 30

nm

Similar In vitro release profile

Drug Load: ~ 20%

Target Attributes

Fine Emulsion

dmax = ε-2/5 γ3/5 p-1/5

Energy density Surface tension Laplace pressure

Fill

finish

Sterilisation

Quench

Organic phase

(API, counter-ion,

polymer)

Aqueous phase

(water,

surfactant) Coarse emulsion Fine emulsion Nanoparticles dilute

Tangential Flow

Filtration

Nanoparticles

concentrate

The shear stress applied by the process must exceed theLaplace pressure, which is inversely proportional to dropletdiameter

Nanoparticle Manufacturing Process Control Aspects

14 IMED Biotech Unit I Option to include TA or IMED function

Process Step Input Parameters Key Outputs Impact

Coarse

Emulsification

Surfactant quantity

Energy

Temperature

Phase viscosities

Droplet size distribution

Emulsion stability

Increasing energy input → smaller droplets

Decreasing surface tension (more surfactant) →

smaller droplets

Smaller droplets → Increased Laplace pressure

Increasing temperature → less stable emulsion &

bigger droplets

Fine

Emulsification

Surfactant quantity

Energy

Temperature

Phase viscosities

Droplet size distribution

Emulsion stability

Drug loading

Drug release rate

Decreasing surface tension → smaller droplets

Smaller droplets → Increased Laplace pressure

Increasing temperature → less stable emulsion,

reduced drug loading

Droplet

stabilisation

Energy

Time

Sink conditions

Temperature/ Pressure

Droplet size distribution

Drug loading

Drug release rate

Temperature / pressure affects residual solvents,

drug loading

Rate of solvent removal impacts drug loading,

residual solvents

Clean &

concentrate

particles

Diavolumes

Transmembrane

pressure

Flow rate

Concentration

Free drug content

Solvent content

Drug loading

Concentration / Assay

More diavolumes→ reduced solvent / free drug

concentration, longer processing times and waste

Need to optimise transmembrane pressure,

concentration, flow rate, temp & diavolumes

Progressing the Development of AZD2811 Nanoparticles

Biopharmaceutics and Critical Quality Attributes

15

Conventional dosage forms IV Nanomedicines

Off-target tissue

Metabolism /

Elimination

Target tissue

Absorption

Free

drug

Distribution

ADME

Off-target tissue

Metabolism /

Elimination

Target tissue

Distribution

ADME

Release

• Efficacy/safety driven by systemic exposure = plasma

conc profile of free drug(target and off-target tissue concs in eqm with plasma conc)

• Plasma concentration limited by drug absorption

(unless IV)

➔ API sol, perm, DP disso

Biopharmaceutics of AZD2811

• Systemic exposure: 2 entities present released drug AND

encapsulated drug

• Different distribution patterns ➔ target exposure no longer

driven only by plasma conc➔ Focus on NP attributes that

can impact release and distribution

Biopharmaceutics potential Critical Quality Attributes

17

% Drug load per particle

Average size of nanoparticles (based on

measurement of 118 nanoparticles):

Mean: 57 nm

Standard deviation: 13 nm

Range: 18 – 80 nm

Particle size using Transmission Electron Microscope (TEM)

Particle size using Dynamic Light Scatter(DLS)

Particle size

using

nanoparticle

Tracking

Analysis

(NTA)

Particle Sizing of Nanoparticles

PEG Encapsulation

Hydrodynamic Radius

Radius of gyration

PEG Conformation

Radius of core

Shape

Stealth Layer Analysis- Quality and In Vivo performance

“Densities above a critical

value of approximately 20

PEG-5000 chains per 100

nm2 prolong circulation

times, irrespective of size*”

*Bertrand et al, Mechanistic understanding of in vivo protein corona formation on polymeric nanoparticles and impact on

pharmacokinetics, Nat. Comm, 2017, 8, 777

Stealth Layer Characterisation

• Surface PEG quantification with NMR• Quantify total PEG in/on nanoparticle by

• Quantify PEG in nanoparticle

• Subtraction = PEG on surface

• Determine size (Rg) via SANs and

therefore surface area

• Packing density: PEG/ 100 nm2

20

Neutron Scattering

Initial view surface PEG in AZD2811 Nanoparticles

21

Size Rg

Nm

(by SANS)

PEG

Encapsulation

% (by NMR)

PEG Density at

Surface

PEG mol/100

nm2

PEG Shell

Thickness nm

(by SANS)

AZD2811 NP 62 ~ 13 44 10 (PEG5000)

No large Q features

\ AMORPHOUS

core

Shell

SANS Experimental vs core-shell model

Cryo-EM

Courtesy Dr. Corinne Smith

Univ of Warwick

Similar Prolonged Plasma Profile for Formulation Variants

Plasma Profile from Similar Formulations with

Different Release Rates in Rat

Nanoparticle Properties

Amount of drug/ nanoparticle ~5 x104

Number of nanoparticles/ mg

of drug1.8 x 1013

Nanoparticle “plasma

circulation time half life” in rat~20 h

AZD2811 plasma half in man is ~30-50 hours

23

AZD2811 characterization of protein corona

Size-exclusion chromatography is used to recovery AZD2811 from plasma and

enable proteomics protein corona profiling from in vitro and in vivo samples as

previously reported (Hadjidemetriou et al. ACS Nano 2015, 9(8), 8142)

Adsorption of proteins on AZD2811

following systemic administration

AZD2811 Corona profiling has now been introduced in the

latest FDA guideline (2017) on Drug and biological

products containing nanomaterials, as the corona

signature might be relevant for

- nanomedicine risk assessment (CQA)

- platform fate in the body

- cellular uptake

https://www.regulations.gov/contentStreamer?documentId=FDA-2017-

D-0759-0002&attachmentNumber=1&contentType=pdf

Release Rate has little impact on Efficacy in SW620 Rat Model

In Vitro Release Biomarker (pHH3) Efficacy

In vivo studies in SW620 Rat Model

Further work to support Product Development

• Define acceptable drug product

attribute ranges to ensure equivalent

performance across clinical batches

• Define a robust, biorelevant,

discriminatory IVR (in vitro release)

method and assess IVIVR (in vitro in

vivo correlation)

• Look at impact of lyophilisation on

product performance and critical

quality attributes

• Process scale up to commercial scale

Variant 1

Variant 3

Variant 2

Formulation Variants with Different

Release Rates give similar PD &

Efficacy

In Summary

26

• AZD2811 is progressing through Phase I/II

• Significant progress in understanding the physical and chemical composition, the

manufacturing process, material attributes that affect the critical quality attributes

of AZD2811 nanoparticle

• Further work on-going to develop in vitro in vivo release correlation to support

product during later development and commercialisation

Acknowledgements

27

• Many many colleagues working in the nanomedicines field in AstraZeneca

Pharmaceutical Sciences, Oncology and Pharmaceutical Technology and

Development groups

• AZD2811 Project Team and sub teams

• External collaborators in particular those from ex BIND Therapeutics, University of

Warwick and ISIS.