Designing amorphous dispersion formulations for poorly soluble drugs - - PowerPoint PPT Presentation

SLIDE 1

Designing amorphous dispersion formulations for poorly soluble drugs

Ian Yates – Product Development Lead, Lonza Bend Tyler Clikeman – Senior Scientist, Product Development, Lonza Bend WEBINAR | May 23rd, 2019

SLIDE 2

• Lonza D

Dosage F Forms and Deliver ery S System ems (DFDS) Intro

• Proble

blem s statement de defini nitio ion a and f nd formulation s selection

• Amorpho

hous us s spray-drie ied di d dispe persio ion formul ulatio ion des esign

• Case s

e studies es

• Physical stability
• Chemical stability
• Correlating in vitro performance testing to in vivo data

Presentation Outline

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 3

3

Lonza DFDS Business Model

Feasibility Studies Commercial Manufacture Drug S Substance Inte termed ediates tes Drug s substances Drug P Product Inte termed ediates tes Drug P Products

Des esig ign

Small / Lab-Scale (non-GMP)

Dev evel elop

Clinical Scale

Manufacture

Commercial Scale

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 4

4

Specialized Focus Areas

Design gn Develop lop Manufacture e

Drug S Substance and I Inte termediate tes Drug P Product Concepts ts Ea Early –stage Clinical Trial Mat aterial als Clinical T Trial Materials Comme mmercial Supply

• Customized

ed A API Dev evel elopmen ent

• High

ghly P Pote tent A t API & Drug P g Products ts

• Addr

ddressi essing B Bioavailabi bility Challeng nges es

• Particle E

Eng ngineer eering

• Mod
• dif

ifying P Pharmacokin

• kinetics
• Multi

ti-particulate F e Formulations ns Pr Prod

• duct

Opt ptions ns

API / HAPI Drug Product Intermediate Soft Gelatin Capsules Tablets – IR, Osmotic, Matrix, Orally Dissolving Powder Multi-particulate Filled Capsules Liquid-filled Hard Capsules

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 5

Problem Statement Definition and Formulation Selection

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 6

6

70-80% of drugs in pharmaceutical pipeline are low solubility

Biopharmaceutical classification system

Our BA enhancement toolkit is geared towards addressing BCS II and IV compound challenges Depth in all enabling technologies used in addressing either BCS IIA, IIB, and IV compounds

• phase-appropriate equipment
• extensive track record
• predictive modeling & tools for tech selection

2008;7:255–270 IIA D Dissolution Rate L e Limited ed IIB S Solubility Limited

Butler, J., Dressman, J. J. Pharm. Sci., 2010

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 7

7

Goal is to efficiently arrive at product development with enabling approach

Problem statement definition guides technology choice

SDD LIPIDIC NXSTAL

Product C t Concept Mole

• lecula

lar P Prop

• pertie

ies Predictions Technolog

• gy &

y & For

• rmula

latio ion In v vit itro, in s silic lico, & i in v viv ivo testin ing Problem em S Statem emen ent

HME

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 8

8

Many Enabling Technologies Are Available for Bioavailability Enhancement

• Polymorphs
• Cocrystals
• Salts
• Cosolvents
• Surfactants
• Cyclodextrins
• Lipids:
• Oils
• SEDDS/SMEDDS
• Solid lipid pellets
• Solid lipid

nanoparticles

Amorphous Cr Crystal F Form rm Solv lvation ion Size R e Reducti tion

• Micronization
• Sub-micron crystals (100 to

800 nm)

• Nanocrystals (<100 nm)
• Solid dispersions
• SDD
• HME
• Lyophiles
• Drug/polymer

nanoparticles

• Layered beads,

nanoadsorbates

• Pure amorphous drug
• Molecular

modification

• Pro-drugs

API S I Sel electi tion

• Oral
• Parenteral
• Pulmonary

Route te of A Admin Lonza Bend Technologies in BA Enhancement

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 9

9

Important Considerations for Pre-formulation Assessment

Solu lubilit ility

1.

• 1. Crys

ystal alline A Aqueou

• us

2.

• 2. Amorphous A

Aqueous 3.

• 3. Crystallin

lline O Organic ic

Aqueous S Solu lubili ility C Challe llenge

1.

• 1. Lipophilic

ilicit ity/ y/Mic icelle lle p par artit itio ionin ing 2.

• 2. Melt

ltin ing p poin int/Crystal l l lattic ice energy ( (i.e .e. “ . “bri rick ck d dust”)

Permeabilit ility

1.

• 1. Molecular

lar D Descrip iptor

• rs

(e.g.

• g. M

MW, r rotatable b bonds, charge s sta tate te) ) 2.

• 2. Caco

co-2 3.

• 3. Perfu

fusion

Me Metabolis lism/ Ef Efflux Pharmac macokinetics

1.

• 1. Absolute B

BA 2.

• 2. BA dose d

dependence 3.

• 3. Food e

eff ffect 4.

• 4. Gastr

tric p pH H effect

Targ rget P Pro roduct P Pro rofile

1.

• 1. Clin

inic ical P l Phas ase 2.

• 2. Dose

se 3.

• 3. Dosing F

g Frequency 4.

• 4. In viv

ivo mod model ( l (e.g. r rat, dog, m , monkey, , human, e , etc.) c.)

Chemic ical S l Stabilit ility

1.

• 1. Labile f

functi tional g groups 2.

• 2. Forced d

degr gradati tion

Physic ical S l Stabil ilit ity

1.

• 1. Ther

ermal P Proper erties es (e.g .g. T Tm, T Tc, c, Tg Tg) 2.

• 2. Wat

ater U Uptak ake

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 10

10

Technology Mapping

Bioavailability Enhancement Map

Friesen et. al. Mol. Pharmaceutics, 5:6 (2008)1003-1019

Fraction Absorbed Classification System (FACS) Amorphous Dispersion Guidance Map

Williams et. al. Pharmacol. Rev., 65(2013), 315-499 Sugano and Terada, Pharm. Sci. 104:2777- 2788, 2015.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 11

11

Three Areas of Focus for Development of an Amorphous Dispersion

Performance Manufacture Stability

Performance:

• Problem statement identification
• Initial characterization through complementary

in vitro tests

• Biomodels to test hypotheses
• Inputs for in vivo results
• Refinement of in vitro tests
• Phase appropriate

Sta tability ty

• Prediction using thermal

properties

• Phase diagrams
• Accelerated stability

Manufact cturability

• Define solvent system
• Define key process parameters
• Scale-up considerations
• Enabling technologies for compounds

with poor organic solubility Early ly D Development Go Goals:

• 1. Learn as much as we can to deliver

the best formulation possible in a time and cost effective manner

• 2. Position program well for late stage

development

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 12

Amorphous Spray Dried Dispersion Formulation Design

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 13

13

Spray-Dried Dispersion – What Is It?

DRYIN YING C G CHA HAMBER

30 microns

Nozzle

THE HE PROCE CESS

FEED ED S SOLUTION Drug is dissolved with polymer in a common organic solvent. DRYIN YING G GA GAS

RE RESULTING S SDD

The resulting powder is a homogenous, stable, amorphous dispersion suitable for incorporation into oral dosage forms.

Pressure Nozzle

Initial Solution Droplet Hot Drying Gas Contacts Droplet Dried d SDD P DD Particle

Skinned Droplet

10 10-6 sec

10 10-2 se sec ~1 se 1 sec

Intensity (counts)

100 200 300 400 500 600 700 800 900

2-Theta - Scale

4 10 20 30 Amorphous SDD Bulk Drug

Intensity (counts)

100 200 300 400 500 600 700 800 900

2-Theta - Scale

4 10 20 30

Intensity (counts)

100 200 300 400 500 600 700 800 900

2-Theta - Scale

4 10 20 30 Amorphous SDD Bulk Drug

PXRD AN ANAL ALYSES

SDD

Bulk drug

SEM SEM TEM

THE P PRODU DUCT

RESULTIN ING G FORMU MULATION Homogeneous, stable, amorphous dispersion BIOAVAI AILAB ABILITY E ENHAN ANCED

• Dissolves rapidly
• Solubility increased
• Maintains super- saturation

in intestine MULTIPLE ORAL AL D DOSAG AGE FO FORMS S

• Tablets
• Capsules
• Powder in bottle
• CR dosage forms

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 14

14

SDD Dissolution Model

Several mechanisms

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 15

33

Problem Statement-specific Bioperformance in vitro Tools

Dissolu lutio ion

Fl Flux ux

Amorph phous us S Solubilit lubility Controlle lled T d Trans nsfer

• Amorphous “solubility”
• Precipitation risk
• Polymer selection
• Drug/polymer interaction
• Dissolution rate
• Precipitation rate
• Maximum apparent

concentration

• Speciation
• Clean measurement of “effective”

concentration

• Able to properly account for

micelle, colloid, and particle contribution to boundary layer diffusion and dissolution rate

• Dissolution rate
• Precipitation rate vs.

emptying rate

• Gastric precipitation
• “Book-end” for

formulation performance

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 16

16

Physical Stability of Spray Dried Dispersions Thermo modyna nami mics Kinet netics cs Expe perienc ence

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 17

Case Study #1: Modeling Physical Stability with a Chemically Stable SDD

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 18

18

• Low drug loading SDD: 15/85 API/HPMCAS-M
• Balance of manufacturability, performance, and

stability required accepting a small amount of crystallization over time

• Modeling showed that we could minimize

physical instability with packaging and storage

SDD characteristics

Performance Manufacture Stability

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 19

19

Modeling Physical Stability

Method:

• 1. Store SDD at accelerated T/%RH stability conditions Crystallize SDD at high

temperature/humidity conditions below Tg in ovens.

• 2. Measure resulting crystallinity from heat of fusion using fast DSC method.
• 3. Calculate initial rates of crystal growth (up to 10% crystalline drug).
• 4. Model rate of crystallization as a function of T, %RH, and/or Tg.

Model Rates of Crystallization Lnk vs. Tg/T Lnk = −Ea/RT + lnA + B(%RH) 1 2, 3 4

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 20

20

Modeling Physical Stability

Method:

• 1. Store SDD at accelerated T/%RH stability conditions Crystallize SDD at high

temperature/humidity conditions below Tg in ovens.

• 2. Measure resulting crystallinity from heat of fusion using fast DSC method.
• 3. Calculate initial rates of crystal growth (up to 10% crystalline drug).
• 4. Model rate of crystallization as a function of T, %RH, and/or Tg.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 21

21

Quantifying low levels of crystallinity in a low drug loading SDD

15/85 API/HPMCAS-M SDD

Very low levels of surface crystals are qualitatively detected by SEM, but quantitation is difficult.

<LOD by DSC and PXRD >LOQ by DSC, <LOQ by PXRD >LOQ by DSC, >LOQ by PXRD Crystalline growth on stability

DSC was able to detect intermediate levels of crystallinity with fast scan rate A significant amount of crystals were needed to quantitate by PXRD and long scan times were required

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 22

22

Measure rate of crystallization with DSC

Up to 10% crystalline API was used for initial rates 3 weeks 1 week

Method:

• 1. Store SDD at accelerated T/%RH stability conditions Crystallize SDD at high

temperature/humidity conditions below Tg in ovens.

• 2. Measure resulting crystallinity from heat of fusion using fast DSC method.
• 3. Calculate initial rates of crystal growth (up to 10% crystalline drug).
• 4. Model rate of crystallization as a function of T, %RH, and/or Tg.

melt

Tm 162 °C Tg 33 °C Tm/Tg 1.42 Heat of Fusion 99.7 J/g

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 23

23

Model crystallization at storage conditions

Method:

• 1. Store SDD at accelerated T/%RH stability conditions Crystallize SDD at high

temperature/humidity conditions below Tg in ovens.

• 2. Measure resulting crystallinity from heat of fusion using fast DSC method.
• 3. Calculate initial rates of crystal growth (up to 10% crystalline drug).
• 4. Model rate of crystallization as a function of T, %RH, and/or Tg.

Lnk vs. Tg/T Lnk = −Ea/RT + lnA + B(%RH)

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 24

24

Model using Lnk vs. Tg/T

• Accounts for humidity with Tg.
• Predicts 6.7 years of stability at 25 °C/60% RH open.
• Water can initiate a different reaction mechanism and cause a

different driving force for crystallization.

Below the Tg

• Same rate at 3 different temps with 3 different %RH

suggests strong correlation with Tg.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 25

25

Model using Lnk = −Ea/RT + lnA + B(%RH)

• Uses humidity modified Arrhenius equation that was developed for chemical stability.
• ASAPprime modeling software accounts for moisture uptake with packaging.
• Model accurately predicts 12 month stability data when 0% RH conditions are excluded.

2 g SDD in a 40 cc HDPE bottle with HIS

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

Parameter Results ln(A) 38.8 ± 0.2 Activation Energy, Ea (kcal/mol) 29.8 ± 2.3 Humidity Sensitivity Factor, B 0.064 ± 0.004

SLIDE 26

Case Study #2: Modeling Chemical Stability with a Physically Stable SDD

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 27

27

Modeling Chemical Stability with a Physically Stable SDD

N O NH2 R N H R

API Impurity hydrolysis

• 50/50 API/PVP-K30 SDD
• Low degradation specification (up to 0.6%)
• Humidity both increased molecular mobility by plasticizing

the SDD and introduced more water for the hydrolysis reaction.

• Physical changes occur above Tg and change mechanism
• Conditions near the Tg were required to measure

degradation within a reasonable timeframe (3 weeks)

• Degradation measured by HPLC

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 28

28

• Linear rates at each condition were fit to

the modified Arrhenius equation

Accelerated Stability Study

Para rameter Results ln(A) 28.6 ± 2.2 Activation Energy, Ea (kcal/mol) 22.5 ± 1.5 Humidity Sensitivity Factor, B 0.063 ± 0.005 R2 0.924

Lnk = −Ea/RT + lnA + B(%RH)

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 29

29

• Modeled 8 kg SDD in 10 liter LDPE double bag with 8-unit sieve desipaks
• Model showed that SDD could be stored in bags with > 10% desiccant, but double bags

in foil was more appropriate

Bulk SDD Storage

desipaks Wt% desiccant Probability

• f passing at

2 years (%) 2 1 3 13 2 5 41 3 8 73 4 10 91

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

desiccant estimation

SLIDE 30

30

comparison with long-term data

Capsule Stability

30 capsules in 30 cc HDPE HIS bottle, closed, no desiccant add desiccant

• Model accurately predicts 12 month stability data.
• 2 g desiccant can increase stability by 20 months
• Model was used to choose desiccant level for

future clinical packaging.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 31

31

Using model to understand stability outliers

• Prediction helped show that 9 month water value was an outlier.
• Additional processing steps and desiccant were added in order to

reduce starting water content and slow impurity formation.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 32

Case Study #3: Mechanistic Understanding of Belinostat Oral Absorption in Beagle Dogs

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 33

33

Case study - SDDs of belinostat dosed to dogs

+

Belinostat BCS II/IV pKa = ≥8 (acidic) LogP < 2 HPMCAS (weakly acidic) 25% active HPMCAS-M SDD Polyvinylpyrrolidone (neutral) Polyvinylpyrrolidone Vinyl Acetate (neutral) SDDs dosed to beagle dogs (n=4), fasted Dose: 50 mg Dosing vehicle: 0.5% Methocel A4M in H2O, 15 ml water rinse 25% active PVP K30 SDD 25% active PVP VA64 SDD Key belinostat attributes:

• High amorphous solubility in biorelevant media (>500 µg/mL).
• Amorphous solubility is impacted by the presence of polymer.
• Dissolution rate is a key driver for absorption and differs depending on

SDD formulation and testing method.

Stewart A, Yates I, et al. Mechanistic Study of Belinostat Oral Absorption from Spray Dried Dispersions. J. Pharm. Sci. (2018).

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 34

34

Belinostat apparent amorphous solubility depends upon dispersion polymer type

Belinostat BCS II/IV pKa = ≥8 (acidic) LogP < 2

Ilevbare, G. A. & Taylor, L. S. Cryst. Growth Des. 13, 3, 1497–1509 (2013).

Blank B Buffer ( (pH 2 2) 6. 6.7 7 mM mM SI SIF 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Amorphous Solubility (mg/ml)

Belinostat + HPMCAS-M Belinostat + PVP K30 Belinostat + PVP VA64

Amorphous solubility is defined as the onset of amorphous liquid-liquid phase

• separation. Presence of polymer influences the LLPS concentration.

6. 6.7m 7mM S SIF pH 6. 6.5

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 35

35

Evaluate belinostat dissolution performance using pH transfer test versus single medium test

In vitro Gastric In vitro Intestinal In vitro Intestinal In vivo Gastric In vivo Intestinal

HPMC MCAS-M SDD SDD 1. 1.3 0. 0.4 1. 1.5 1. 1.3 0. 0.8 PVP K30 S 30 SDD 1. 1.4 0. 0.4 1. 1.7 1. 1.4 0. 0.8 PVP VA64 S 64 SDD 3. 3.3 1. 1.0 4. 4.0 3. 3.3 2. 2.0 Assu ssumes: s:

• Fasted s

state

• 50

50 mL g gastri ric v volume

• 50 mL i

int ntestin inal v l volu lume

In vivo

pH 6.5 6.7 mM SIF 20 ml

Intestinal pH test (pH 6.5, 6.7 mM SIF) Gastric transfer test (pH 2 SGF  6.5, 6.7 mM SIF)

pH 2 SGF pH 6.5 6.7 mM SIF Add Concentrated SIF solution at t = 30 min 10 ml 20 ml

Dose/Volume/Solubility:

source: daviddarling.info

Non-sink Dose: 1000 µg/mL in SGF Non-sink Dose: 2000 µg/mL in SIF In situ fiber optic detection

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 36

36

Relative extents of dissolution between SDDs depends upon dissolution medium composition

0.0 0.5 1.0 1.5 30 60 90

Concentration (mg/mL) Time (min)

0.0 0.5 1.0 1.5 30 60 90 120

Concentration (mg/ml) Time (min) HPMCAS-M SDD PVP K30 SDD PVP VA64 SDD HPMCAS-M SDD PVP K30 SDD PVP VA64 SDD Intestinal pH test (pH 6.5, 7 mM SIF) M SDD > K30 SDD > VA64 SDD Gastric transfer (pH 2 SGF  6.5, 7 mM SIF) K30 SDD > M SDD ≈ VA64 SDD

Dashed lines represent the apparent amorphous solubility measured in SGF and SIF from the amorphous solubility assay (slide 30)

Dose: 1000 µg/mL (SGF), 500 µg/mL (SIF) Dose: 2000 µg/mL

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 37

37

Using amorphous solubility and dissolution data as key inputs to absorption model supports hypothesis of dissolution rate limited absorption

Amorphous solubility Dissolution rate/extent

In vitro inputs to model In silico predictions

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 38

38

Gastric → intestinal transfer test better rank orders SDDs with respect to in vivo performance in dogs

Gastric transfer (pH 2  pH 6.5, 7 mM SIF) Intestinal pH test (pH 6.5, 7 mM SIF) Sequential exposure to SGF and SIF at a more relevant dose/volume/solubility (dose number) is a better indicator for rank-ordering in vivo exposure from each SDD.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 39

39

Conclusions Belinostat Case Study

• Amorphous solubility of belinostat depends on polymer type.
• SGF/SIF transfer test a better indicator of in vivo performance.
• Used in vitro inputs to describe blood plasma profiles via absorption modeling.
• Rate-determining step to absorption: dissolution rate and extent achieved in the

stomach prior to transit down the GI tract.

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019

SLIDE 40

40

Thank you

Contact us: solutions@lonza.com

Yates Clikeman | Pharmaceutical Technology Webcast | May 23rd, 2019