Transcript Young Innovators 2009

INNOVATORS 2010
Novel Performance Evaluation of
Dry Powder Aerosols
Graduate Student Symposium Award in Manufacturing
Science and Engineering
Zhen Xu and Anthony J Hickey
University of North Carolina, Eshelman School of
Pharmacy, Chapel Hill, NC 27599
ABSTRACT
• Purpose: Performance evaluation of carrier
surface modified formulation using powder
aerosol deaggregation equation (PADE).
Young Innovators 2009
ABSTRACT
•
•
•
•
•
•
Methods: Lactose monohydrate (sieved (SV) and milled (ML)) were coated with stearic acid
(SA) using adsorption-coacervation method. The amount of surface coating was quantified by
gas chromatography.
The formulation blends were prepared at 2%(w/w).
Particle size distribution, morphology, and thermal properties of drug carriers, and mixtures
were examined.
Experimentally designed in vitro aerosolization was performed using standardized
entrainment tubes (SETs) with well-defined airflow parameters.
Formulations that pneumatically entrained were characterized by twin-stage liquid impinger
(TSLI) at a flow rate of 60 L/min.
The particle deaggregation exemplified by fine particle fraction (FPF) was correlated with
SET shear stress (ts) by applying PADE (non-linear and linear regression). Formulation
reproducibility across increasing shear stress can be determined when the hyperbolic curve
approaches a plateau. (1-4)
Young Innovators 2009
INTRODUCTION
•
•
•
•
•
The surfaces of pharmaceutical dry powder aerosols are heterogeneous.
In principle, the heterogeneity manifests on the particle surfaces should be similar to that
described in the surface adsorption theory, since the fundamental forces and surface energies
are described and characterized similarly.
The practical approach to describing energy heterogeneity (e.g. adsorption or protein binding)
is to average local thermodynamic quantities and treat them statistically.
PADE is an analogy to the surface adsorption theory. It led to a novel way of interpreting
particle interactions. (1-4)
PADE model:
k t
FPF
 d s
FPFmax 1  k d t s
( Eq.1)
ts
FPF

ts
FPFmax
Young Innovators 2009

1
( Eq.2)
k d ( FPFmax )
INTRODUCTION
•
As a shear force applies to the surface of carrier particles, drug particles are removed with
increasing difficulty because of the sites that they occupy until a saturation is reach, when no
drug particle can be removed at increasing shear.
•
PADE is mechanistically different form adsorption theory because it involves rapid system
volume expansion and changing boundary condition. (4)
Young Innovators 2009
MATERIALS
•
•
•
•
Respirable drug: micronized albuterol sulfate (AS).
Carrier: lactose monohydrate (RespitoseTM), surface-modified sieved (SV) and ML (ML)
batches
Surface active agent: stearic acid.
Other chemicals: anhydrous dichloromethane, anhydrous chloroform, petroleum ether, borotrifuoride (14%) in methanol, deionized water.
Young Innovators 2009
METHODS
•
•
•
•
•
•
•
•
The theoretical estimation for monolayer surface coating was performed based on the surface
area of lactose batches determined by nitrogen adsorption method.
Adsorption-coacervation for stearic acid (SA) coating. (5)
Interactive physical mixtures of AS and lactose batches (SV, ML, SV-0SA, ML-0SA, SVlowSA, ML-lowSA, SV-highSA, and ML-highSA) were prepared at drug concentration
2%(w/w).
Particle volume size distribution of drug and carriers was characterized by laser diffraction
Particle size and morphology was assessed by SEM at different magnification.
Thermal properties of drug, carriers, and mixtures were examined by DSC at scanning rate of
5 oC/min.
Experimentally designed in vitro aerosolization was performed using SETs.
Formulations that pneumatically entrained were characterized by TSLI at 60 L/min. PADE
was applied. Statistical analysis was carried out suing SigmaPlot software.
Young Innovators 2009
RESULTS
•
•
•
•
•
Jet-milled drug particles were in the respirable size range (D50 = 3.59). SV and ML had
similar D50 (59.7, 54.6 um), but significantly different span (1.10 vs. 3.13).
Non-spherical drug particles formed aggregates due to strong interfacial interactions at solidsolid interface. The primary particle size similar, but SV blend had much less fine
particle/agglomerates associated with the primary particles than ML.
Surface monolayer coverage were estimated to be 0.7 and 2.1 mg/g, respectively.
The adsorption-association isotherms obtained were similar for SV and ML. The surface
coating at two equilibrium concentrations were selected for performance studies.
DSC thermograms show reduced enthalpy for both endothermic peaks. The characteristic
exothermic peak at ~173 oC disappeared. No enhancement of the exothermic peak after
blending.
Young Innovators 2009
RESULTS
X3,000 AS
x5,000
X3,000 DSCG
x600
Young Innovators 2009
x600
RESULTS
Figure 2B
DSC Power Heat Flux
EEndothermic (down) Exothermic (up)
DSC Power Heat Flux
EEndothermic (down) Exothermic (up)
Figure 2A
SV
SV-0SA
SV-lowSA
SV-highSA
25
50
75
100
125
150
175
200
225
250
25
o
Temperature ( C)
Young Innovators 2009
ML
ML-0SA
ML-lowSA
ML-highSA
50
75
100
125
150
Temperature (oC)
175
200
225
250
RESULTS
•
The correlation plots using PADE non-linear and linear regression analyses showed
excellent correlation.
– Non-linear regression: Adjusted R2 = 0.8497-0.9957
– Linear regression: R2 = 0.9526-0.9993
•
Solvent treatment (SV-0SA, ML-0SA) resulted in negative influence on the
aerosolization performance (smaller FPFmax). For SV formulations, the
performance was dependent upon the amount of SA coated. Conversely, the
performance of ML formulations were less susceptible to the amount of SA coated.
•
ML formulations gave higher performance efficiency.
Young Innovators 2009
Figure 3A
Figure 3B
25
120
SV+AS
SV-0SA+AS
SV-lowSA+AS
SV-highSA+AS
100
RESULTS
ts/FPFTD*100
FPFTD (%)
20
15
10
5
80
60
40
20
0
0
0
2
4
6
8
10
2
Shear stress (ts) (N/m )
Figure 3C
12
14
0
2
4
6
8
10
Shear stress (ts) (N/m2)
12
14
12
14
Figure 3D
60
50
ML+AS
ML-0SA+AS
ML-lowSA+AS
ML-highSA+AS
50
40
40
ML+AS
ML-0SA+AS
ML-lowSA+AS
ML-highSA+AS
ts/FPFTD*100
FPFTD (%)
SV+AS
SV-0SA+AS
SV-lowSA+AS
SV-highSA+AS
30
30
20
20
10
10
0
0
2
4
6
8
10
Shear stress (ts) (N/m2)
0
12 Young14
Innovators 2009
0
2
4
6
8
10
2
Shear stress (ts) (N/m )
CONCLUSION

Significant differences were characterized between SV and ML on size distribution,
morphology, and surface energetics. SA was quantitatively coated on the surfaces of SV and
ML particles.

Physicochemical characterization indicated strong influence of solvent during coating.

The application of robust PADE, based on fundamental Langmuir theory, gave excellent
performance prediction (model well-tolerated), which allow rational design of dry powder
formulation instead of trial and error.

The reduced aerosolization could be explained by stripping of natural surface coating on
lactose monohydrate, whereas the artificial coating with SA could reduce surface
heterogeneity and improve aerosolization.

ML formulations performed better than SV formulations.
Young Innovators 2009
ACKNOWLEDGMENTS
• This research was funded by DMV-Fonterra
Excipients and Pfizer.
Young Innovators 2009
REFERENCES
1. Xu Z, Mansour, HM, Mulder T, McLean R, Langridge, J, Hickey, AJ, J Pharm Sci
2010, 99(8): 3398-3414.
2. Xu Z, Mansour HM, Mulder T, McLean R. Langridge J, Hickey AJ, J Pharm Sci
2010, 99(8): 3415-3429.
3. Mansour HM, Xu Z, Hickey AJ, J Pharm Sci, 2010, 99(8): 3430-3441.
4. Xu Z, Mansour HM, Mulder T, McLean R, Langridge J, Hickey AJ, J Pharm Sci
2010, 99(8): 3442-3461.
5. Hickey AJ, Jackson GV, Fildes FJ, J Pharm Sci, 1988, 77, 804-809.
Young Innovators 2009
BIOS/CONTACT INFO
• Zhen Xu
• PhD in Pharmaceutical Sciences, University of North Carolina at Chapel
Hill. Area: formulation and preformulation; aerosol drug delivery.
Graduated in August, 2010.
• Master of Sciences in Chemistry, Michigan State University, carbohydrate
chemistry. Graduated in August, 2004.
• [email protected]
• (919)-809-4174
• 6808 Blenheim Rd Apt D, Baltimore, MD 21212
Young Innovators 2009