Structure and Release Behavior in Controlled Drug Delivery ...people.math.gatech.edu/~chomp/workshop/saylor.pdf · Structure and Release Behavior in Controlled Drug Delivery Devices

Structure and Release Behavior in

Controlled Drug Delivery Devices

David M. Saylor

U.S. Food and Drug Administration

Computational Homology Workshop

Georgia Tech, Atlanta, GA

February 4, 2006

Contributors

Theory and Computation:

Chang-Soo Kim (FDA)

Jim Warren (NIST)

Experimental:

Dinesh Patwardhan (FDA)

Benita Dair (FDA)

Ken McDermott (FDA)

Homology:

Tom Wanner (GMU)

Controlled Drug Delivery• Drug is incorporated into a matrix material (polymer)

– diffusion barrier = reduced release rate

Time

Fra

ction D

rug

Deliv

ere

d

0.0

0.2

0.4

0.6

0.8

1 .0

“pure”

“controlled”

Supply a sustained safe

and effective dose of drug

to target media

dru

g

fra

ctio

n

1.0

0.0

drug +

matrix

pure

drugtarget

media

Controlled Delivery Devices

transdermal

patches

extended release

tablets

targeted

chemotherapydrug eluting

stent

Drug Eluting Stent (DES)

Coronary disease:

Future: other devices coated with CDD systems

Problem:restenosis

Solution:DES

traditional treatment =bare metal stent

DES Manufacture

bare metal stent

(316 stainless steel)

drug + polymer

dissolved in solvent

Paralene C

drug + polymer

polymer (optional)

Structure will depend on materials

and manufacture conditions

drug-richphase

- drug molecule

- polymer molecule

polymer-richphase

Drug-Polymer Microstructures

drugfraction

1.00.0

Microstructure Impacts Kinetics

Time

Fra

ction

Dru

g D

eliv

ere

d

0.0

0.2

0.4

0.6

0.8

1.0

Microstructure: spatial variation in

chemical and physical composition

40% drug

drugfraction

1.00.0

Example: Microstructure v. Kinetics

Wormuth, DeWitt, and Haugstad,

Polymer Preprints 2005, 46(2), 1222

increased phaseseparation

Elution of dexamethasone from thin coatings ofpoly(alkylmethacrylates) on stents

Regulatory / Manufacturing Needs1) Elucidate and quantify the influence of structural

variations on delivery kinetics

2) Identify the effect of process conditions on

structure development

Processing-Stucture-Response Framework

Delivery

Kinetics

Microstructure

Process

Conditions

Framework Application

Impact:

1) Remove empiricism from product development

2) Defined guidance for product regulation

Delivery

Kinetics

Microstructure

Process

Conditions

1) Select desired oracceptable range of

delivery kinetics

2) Determine requiredmicrostructure(s)

3) Identify processroutes to obtain

desired structure(s)

Objective:

Approach:

To develop tools that will enable the inter-relationships

between processing, microstructure, and delivery

kinetics to be quantified.

1) Theory/Computation

- develop theoretical and numerical tools to predict structural

evolution in CDD systems

- computational experiments (manufacture and deliver)

2) Laboratory

- fabricate CDD systems under different process conditions

- dissolve systems to characterize delivery kinetics

3) Quantitative Description of Microstructure

- computational homology





2) Laboratory





Theory

Overview:

• Thermodynamics provides driving force for evolution

• Kinetics governs speed of evolution

• Three (3) components: drug, polymer, and solvent

• Order parameter: amorphous or crystalline

• Diffusive transport phenomena (mixing and separation)

• Phase transformations (e.g. crystallization)

• Heterogeneous and homogenous nucleation

A set of partial differential equations, based on fundamental

materials chemistry and physics, that govern the evolution of

a system of materials.

System Specific Parameters

crystalline

amorphous

• Variables in equations are material system specific

thermodynamic and kinetic parameters.

• Determined by: experiment, chemical group theory, molecular

dynamics, educated guess

Thermodynamics

Material system:

polymer = PLA (biodegradable)

drug = Sirolimus

solvent (processing) = THF

solvent (delivery) = isopropanol

Application of TheoryNumerical methods used to solve the equations in space and

time yielding structural evolution:

Processing

drugsolvent

polymer

amorphous crystalline

t =

0t =

t*

composition order

Evolution During Delivery

drugsolvent

polymer

amorphous

crystalline

t = 0 t = t*

co

mp

ositio

no

rde

r

Computational ExperimentsCharacterize microstructure evolution during manufacturing:

• Drug:Polymer = 0.50:0.50

• Dissolve in Solvent

• Evaporate off Solvent

drugsolvent

polymer

amorphous

crystalline

co

mp

ositio

no

rde

r

Evaporation RateAlthough evaporation is not in the theory explicitly,

time of evolution is a qualitative measure

drugsolvent

polymer

“fast”

“slow”

evapora

tion

rate

Evaporation is Coming!drug : polymer : solvent = 15 : 15 : 70

Psolvent = P Psolvent = 0.5 P

composition order composition order

Drug LoadingAfter the same amount of evolution time (i.e. same

evaporation rate)

% drug

30%

50%

15%

drugsolvent

polymer

Process-Structure Relations

Extend over other variables (e.g. materials, temperature, etc.)

drug loading

15%

30%

50%

evapora

tion r

ate

“slo

w”

“fast”

drugsolvent

polymer

• 30% drug - “medium” evaporation rate

• dissolve composite in solvent (biodegradable polymer)

Evolution During Delivery

drugsolvent

polymer

Delivery Kinetics

drugsolvent

polymer

30% Drug Delivery KineticsMicrostructures formed with 30%

drug at varying evaporation rates:

drugsolvent

polymer

“medium”

“slow”

“fast”

Delivery Kinetics v. Composition

drugsolvent

polymer

Microstructures formed at “medium” evaporation

rate with varying compositions:

15%

30%

50%

Processing-Structure-Response

drug loading

15%

30%

50%

evapora

tion r

ate

“slo

w”

“fast”

delivery kinetics





2) Laboratory









2) Laboratory





Objective:

Approach:




Laboratory Experiments1) Sample fabrication

2) Microstructure characterization

3) Dissolution testing

polymerdrug

solvent

Sample Fabrication

Polymer = SIBS Drug = Tetracycline

Note: Material system is different from computational experiments

(e.g. polymer is non-biodegradable)

Process Variability Matrix:

Material System:Solvent = THF

dru

g:p

oly

mer

T(°C)40°C23°C

15:8

530:7

0

3 3

33

Microstructure v. Processing

100µm

30%

dru

g23°C

15%

dru

g40°C

drug particles

30% Drug Comparison

23°C (phase separated)

40°C (homogeneous)

Process-Structure-Response

23°C 40°C

15% drug 30% drug

23°C 40°C





2) Laboratory









2) Laboratory





Objective:

Approach:




Microstructure QuantificationMapping requires a statistical representation of

microstructure, i.e. a metric.

Delivery

KineticsMicrostructure

Process

Conditions

Require a metric for microstructure that is:

1) robust enough to distinguish between structures that yield

significantly different responses

2) relatively simple

Processing-Structure-Response

drug loading

15%

30%

50%

evapora

tion r

ate

“slo

w”

“fast”

delivery kinetics

Proposed Metric• Quantify topological invariants (i.e. Betti numbers) at different

levels of drug concentration:

Threshold:

25% drug

50% drug

75% drug

!0 = 17

!1 = 12

!0 = 15

!1 = 7

!0 = 15

!1 = 7

Topology “Signal”

• Map of !i per unit volume as a function of

threshold level:

!0 = 17

!1 = 12

!0 = 15

!1 = 7

!0 = 15

!1 = 7

Microstructure Classification• Topology "signal" can be used to quantify and classify

variations in microstructure

• Structures with topologically similar distributions of drug will

yield quantitatively similar "signals”, i.e. considered statistically

identical

drugsolvent

polymer

Evaporation v. Homology30% drug, variable evaporation rate:

“slow”

“fast”

evapora

tion

rate

Drug Loading v. Homology“medium” evaporation rate, variabledrug content:

15% drug

30% drug

50% drug

Homology v. Delivery KineticsMicrostructures formed with 30%

drug at varying evaporation rates:“medium”

“slow”

“fast”

Processing-Homology-Response

drug loading

15%

30%

50%

evapora

tion r

ate

“slo

w”

“fast”

delivery kinetics

Objective:

Approach:








2) Laboratory





Clinically Desired KineticsGoal: tailor delivery kinetics to a particular application

Desired or acceptable delivery kinetics depends

on the application:

time

Dru

g R

ele

ased cytotoxic

cytostaticmetabolic

rate

Framework Application

drug loading

15%

30%

50%

evapora

tion r

ate

“slo

w”

“fast”

delivery kinetics

cytostatic

Summary• Developed theoretical and computation tools to predict

microstructural evolution (processing & release conditions) in

controlled drug delivery systems.

• Complementary laboratory experiments are being conducted

to elucidate these same relationships and to provide validation

for the theory.

• A relatively simple microstructural metric based on

computational homology has been proposed to link

quantitatively the system microstructure with delivery kinetics

and processing routes.

• These tools can be used to build quantitative processing -

structure - response relations that can provide strict,

quantitative guidelines for device design and provide the basis

for product review decisions.

Appropriate Metric?Q. Is the proposed metric sufficient, i.e. does it provide

adequate resolution in microstructure space?

Q. Is there another relatively simple metric based on topological

measures that would be better?

Q. Is microstructure topology an appropriate measure for this

application?

vs.

e.g. !i(d) d

Structure and Release Behavior in Controlled Drug Delivery ...people.math.gatech.edu/~chomp/workshop/saylor.pdf · Structure and Release Behavior in Controlled Drug Delivery Devices

Documents