Upload
gregory-fletcher
View
213
Download
0
Tags:
Embed Size (px)
Citation preview
K S TanK S Tan, K H Leong, L Y Chung, M I Noordin, K H Leong, L Y Chung, M I NoordinK S TanK S Tan, K H Leong, L Y Chung, M I Noordin, K H Leong, L Y Chung, M I Noordin
Department of PharmacyDepartment of PharmacyFaculty of MedicineFaculty of MedicineUniversity of MalayaUniversity of MalayaKuala LumpurKuala Lumpur
2
Introduction
Optimization of pharmaceutical formulation
Conventional trial-and-error approach
Mathematical Modeling Method.
3
Introduction
Furosemide has a narrow absorption window (located at upper GI tract)1, 2 & 3.
Conventional oral formulation exhibits erratic bioavailability & unpredictable response4.
Furosemide
4
Introduction
Absorption window
5
Introduction
[Concept adapted from Reference 1]
Absorption window
o
6
Introduction
Gastroretentive dosage form prolongs retention time in stomach & permits continuous drug release to optimal absorption site1, 2 & 5.
7
Objectives
To optimize a formulation for furosemide characterized by a 12-hour gastroretentive and sustained release profile.
To demonstrate the usefulness of mathematical modeling method in optimization of formulation.
8
Methods: OverviewDetermination of 13 model formulations (Formulae A - M)
via simplex lattice design.
Preparation of tablets (Formulae A – M)
Tablet QC tests(Uniformity of weight, friability, tablet size, hardness)
In vitro dissolution tests (8 hours)•Enzyme-free simulated gastric fluid
(SGF) pH 1.2•USP paddle method (100 rpm)
•Temperature 37±0.5˚C• Sample buffered to pH 5.8 & Assayed with UV spectrophotometry at 278 nm
In vitro tablet swelling tests•Enzyme-free simulated gastric fluid
(SGF) pH 1.2•USP paddle method (100 rpm)
•Temperature 37±0.5˚C•Measurement of swelling tablet
diameter
(To be continued in next slide)
9
Methods: Overview
Data of in vitro tablet dissolution tests Data of in vitro tablet swelling tests
Multiple Linear Regression Analysis•Model-fitting
•Determination of best-fit models for individual response
Optimization of formulationDesign-Expert® integrate all models built and solve simultaneously to
search for optimal formulation based on the constraints imposed.
Verification of optimal formulation(In vitro dissolution tests & tablet swelling test)
Multiple Linear Regression Analysis•Model-fitting
•Determination of best-fit models for individual response
(Continued from previous slide)
10
Methods
Mixture experimental design
Tablet excipients: Iota-carrageenan,
Lambda-carrageenan
Acacia gum.
Simplex lattice design was employed to determine excipient composition of 13 model formulations.
Each 400 mg tablet contains 60 mg furosemide.
11
MethodsComposition of tablet excipients for 13 model formulations.
12
Results & Discussions Formula A
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula B
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula C
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula D
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula E
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula F
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
In Vitro Tablet Dissolution Profiles of 13 model formulations
Formula G
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula H
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula I
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula J
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula K
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula L
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula M
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
R
ele
ase
Formula C
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula H
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
(n = 6)
13
Results & DiscussionsFormula A
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula C
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula E
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula F
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
In Vitro Tablet Swelling Profiles of 13 model formulations
Formula H
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula I
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula J
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula K
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula L
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula M
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula G
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula B
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula D
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
(n = 6)
14
Results & Discussions
Formula B
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
Formula B
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula B: Dissolution Profile
Formula B: Tablet Swelling Profile
Formula D
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Re
lea
se
Formula D: Dissolution Profile
Formula D: Tablet Swelling ProfileFormula D
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
me
ter
(mm
)
15
Results & Discussions
Model-Fitting
The data of all response variables (tablet dissolution and swelling tests) for 13 formulations were fitted into various equations:
Linear model: Y = b1X1 + b2X2 + b3X3
Quadratic Model: Y = b1X1 + b2X2 + b3X3 + b12X1X2 + b13X1X3 +
b23X2X3
Special Cubic model:
Y = b1X1 + b2X2 + b3X3 + b12X1X2 + b13X1X3 +
b23X2X3 + b123X1X2X3
16
Results & DiscussionsModels for Drug Release & Tablet Swelling Profiles
17
Contour plots of individual response variable for in vitro tablet dissolution studies
Y30 min:% Drug released in 30 minutes Y1h:% Drug released in 1 hour Y1.5h:% Drug released in 1.5 hour Y2h:% Drug released in 2 hours
Y3h:% Drug released in 3 hours Y4h:% Drug released in 4 hours Y5h:% Drug released in 5 hours Y6h:% Drug released in 6 hours
Y7h:% Drug released in 7 hours Y8h:% Drug released in 8 hours
Y1.5h:% Drug released in 2 hour
18
Contour plots of individual response variable for in vitro tablet swelling studies
Z30min: Tablet diameter at 30th minZ15min: Tablet diameter at 15th min Z45min: Tablet diameter at 45th min Zih: Tablet diameter at 1st hour
Zi.5h: Tablet diameter at 1.5th hour Z2h: Tablet diameter at 2nd hour Z3h: Tablet diameter at 3rd hour Z4h: Tablet diameter at 4th hour
Z5h: Tablet diameter at 5th hour Z6h: Tablet diameter at 6th hour Z7h: Tablet diameter at 7th hour Z8h: Tablet diameter at 8th hour
19
Results & Discussions
Optimization of Formulation
Constraints imposed on:Drug release at 2hr (12-16%), 4hr (24-32%), 6hr (42-52%) & 8 hr (70-100%).
Tablet swelling: 13-19 mm (maximizing).
Optimized formula:
Excipients Excipient Composition (%)
ι-carrageenan, X154.44
λ-carrageenan, X221.11
Acacia gum, X324.45
20
Results & DiscussionsOptimized formulation
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
met
er (
mm
)
Tablet dissolution profile (A) and swelling profile (B) of optimal formulation predicted by the model.
B
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Rel
ease
d
A
21
Results & DiscussionsVerification of Optimal Formulation
0
4
8
12
16
20
0 2 4 6 8Time (Hour)
Dia
met
er (
mm
)
Observed response Predicted response
B
Tablet dissolution profile (A) and swelling profile (B) of optimal formulation (Comparing observed vs. predicted data)
0
20
40
60
80
100
0 2 4 6 8Time (Hour)
% D
rug
Rel
ease
d
Observed response Predicted response
A (Paired-samples T-test, p > 0.05) (Paired-samples T-test, p > 0.05)
22
Results & Discussions
0
20
40
60
80
100
0 2 4 6 8 10 12 14Time (Hour)
% D
rug
Rel
ease
The optimal formulation exhibits a zero-order release kinetic. (Fitted into Korsmeyer-Peppas model, n = 0.94)
In Vitro Tablet Dissolution Profiles
23
Results & Discussions
0
20
40
60
80
100
0 2 4 6 8 10 12 14Time (Hour)
% D
rug
Rele
ase
Commercial Product GRDF OF2
Commercial Product: furosemide 60 mg (Wakelkamp et al 1999)
GRDF: A gastroretentive dosage form, furosemide 60 mg developed by Klausner et al (2003)5
OF: The optimal formulation obtained in this study.
In Vitro Tablet Dissolution Profiles
24
Conclusions
Optimal formulation with desirable release profile & tablet swelling characteristics was obtained.
An efficient optimization process: omitting the cost- and time-consuming procedures as in the conventional trial-and-error approach.
Mathematical modeling permits the characterization of drug release kinetics during the optimization process.
Graphical optimization allows evaluation of excipient’s functionality in the dosage form.
25
References
1. Chawla, G, Gupta, P, Koradia, V & Bansal, AK 2003, ‘Gastroretention a means to address regional variability in intestinal drug absorption’, Pharmaceutical Technology, vol. 27, no. 7, pp. 50-68.
2. Davis, SS 2006, ‘Formulation strategies for absorption windows’, Drug Discovery Today, vol. 10, no. 4, pp. 249-257.
3. Rouge, N, Buri, P & Doelker, E 1996, ‘Drug absorption sites in the gastrointestinal tract and dosage forms for site-specific delivery’, International Journal of Pharmaceutics, vol. 136, pp. 117-139.
4. Ponto, LLB & Schoenwald, RD 1990, ‘Furosemide (frusemide): a pharmacokinetic/pharmacodynamic review (part I)’, Clinical Pharmacokinetics, vol. 18, no. 5, pp. 381-408.
5. Klausner, EA, Lavy, E, Stepensky, D, Cserepes, E, Barta, M, Friedmann, M & Hoffman, A 2003b, ‘Furosemide pharmacokinetics and pharmacodynamics following gastroretentive dosage form administration to healthy volunteers’, Journal of Clinical Pharmacology, vol. 43, pp. 711-720.
6. Wakelkamp, M, Blechert, Å, Eriksson, M, Gjellan, K & Graffner, C 1999, ‘The influence of frusemide formulation on diuretic effect and efficiency’, British Journal of Clinical Pharmacology, vol. 48, pp. 361-366.
26
27
28
Results & Discussions
Model Coefficient Y30min Y1h Y1.5h Y2h Y3h Y4h Y5h Y6h Y7h Y8h
Linear
SD 1.99 2.50 3.30 5.17 7.80 7.55 9.24 11.56 11.49 11.50
R2 0.4107 0.5764 0.7170 0.6991 0.7058 0.7435 0.7135 0.6716 0.6787 0.6223
Adjusted R2 0.2929 0.4917 0.6604 0.6390 0.6470 0.6921 0.6561 .0.6059 0.6145 0.5468
Predicted R2 -0.1500 0.1291 0.3717 0.3541 0.4497 0.5153 0.5124 0.4130 0.3826 0.3493
PRESS 77.25 128.48 241.32 573.19 1138.36 1078.04 1452.52 2387.25 2538.70 2278.71
Quadratic
SD 1.98 2.26 2.72 4.46 7.82 7.62 9.92 12.33 11.26 12.36
R2 0.5905 0.7579 0.8651 0.8428 0.7933 0.8171 0.7688 0.7383 0.7844 0.6948
Adjusted R2 0.2979 0.5850 0.7687 0.7305 0.6457 0.6865 0.6037 0.5514 0.6303 0.4768
Predicted R2 -1.2390 -0.3316 0.2168 -0.0622 -0.1154 0.0648 -0.1068 -0.1933 0.1959 -0.1896
PRESS 150.40 196.44 300.80 942.72 2307.11 2080.17 3297.06 4853.03 3306.65 4165.98
Special cubic
SD 2.09 2.33 2.61 4.57 8.02 7.17 9.61 9.55 7.73 9.35
R2 0.6104 0.7799 0.8932 0.8589 0.8135 0.8613 0.7688 0.8653 0.9127 0.8504
Adjusted R2 0.2207 0.5597 0.7864 0.7179 0.6271 0.7227 0.6037 0.7306 0.8254 0.7007
Predicted R2 -1.8123 -0.6640 0.0790 -0.3133 -0.3791 -0.0138 -0.1068 0.1293 0.6351 0.3314
PRESS 188.92 245.47 353.72 1165.50 2852.61 2255.02 3297.06 3541.14 1500.56 2341.21
Cubic
SD 1.53 1.90 1.97 4.01 9.03 8.31 11.53 11.34 10.73 11.85
R2 0.8952 0.9267 0.9696 0.9458 0.8817 0.9068 0.8660 0.9052 0.9160 0.8798
Adjusted R2 0.5809 0.7069 0.8785 0.7830 0.5267 0.6270 0.4642 0.6207 0.6639 0.5192
Predicted R2 -25.242 -25.7135 -11.4817 -11.0792 -21.5643 -16.1550 -33.0634 -13.4233 -11.7876 -15.7273
PRESS 1762.83 3940.73 4793.94 10720.08 46674.36 38158.03 101500 58657.11 52583.49 58577.50
Model-fitting Summary for Tablet Dissolution Profiles
29
Results & DiscussionsModel-fitting Summary for Tablet Swelling Profiles
Model Coefficient Z15min Z30min Z45min Z1h Z1.5h Z2h Z3h Z4h Z5h Z6h Z7h Z8h
Linear SD 0.27 0.32 0.28 0.28 0.39 0.61 0.54 1.15 2.46 4.19 2.64 2.72
R2 0.7941 0.8817 0.9163 0.9422 0.9403 0.8925 0.9390 0.8671 0.7438 0.6904 0.8721 0.8767
Adjusted R2 0.7530 0.8581 0.8996 0.9307 0.9284 0.8710 0.9268 0.8405 0.6926 0.6285 0.8466 0.8520
Predicted R2 0.6120 0.8096 0.8571 0.8897 0.8992 0.7876 0.8877 0.6836 0.3987 0.3534 0.7344 0.7816
PRESS 1.40 1.67 1.34 1.54 2.62 7.29 5.33 31.32 142.46 366.66 145.08 130.93
Quadratic SD 0.27 0.36 0.24 0.29 0.35 0.43 0.60 0.70 1.61 3.61 2.21 3.17
R2 0.8565 0.8980 0.9581 0.9584 0.9674 0.9615 0.9477 0.9658 0.9235 0.8392 0.9374 0.8826
Adjusted R2 0.7540 0.8251 0.9282 0.9286 0.9442 0.9340 0.9104 0.9414 0.8689 0.7244 0.8927 0.7987
Predicted R2 0.2257 0.6130 0.7697 0.7368 0.8895 0.8396 0.6958 0.7329 0.5035 0.4860 0.7743 -0.038
PRESS 2.80 3.40 2.16 3.67 2.87 5.50 14.44 26.44 117.63 291.50 123.28 622.22
Special cubic
SD 0.27 0.33 0.21 0.30 0.31 0.47 0.61 0.70 1.55 3.67 2.17 3.09
R2 0.8752 0.9263 0.9718 0.9602 0.9780 0.9616 0.9526 0.9706 0.9389 0.8577 0.9484 0.9047
Adjusted R2 0.7505 0.8526 0.9437 0.9204 0.9560 0.9232 0.9051 0.9412 0.8778 0.7155 0.8968 0.8093
Predicted R2 0.1188 0.6836 0.7845 0.6754 0.9091 0.7942 0.6440 0.6875 0.4416 0.4565 0.7673 -0.176
PRESS 3.18 2.78 2.02 4.53 2.36 7.06 16.90 30.93 132.28 308.24 127.12 704.94
Cubic SD 0.23 0.4 0.22 0.32 0.26 0.56 0.62 0.52 1.42 4.52 2.65 0.84
R2 0.9564 0.9444 0.9847 0.9780 0.9924 0.9724 0.9759 0.9918 0.9746 0.8919 0.9613 0.9965
Adjusted R2 0.8255 0.7777 0.9388 0.9120 0.9695 0.8896 0.9036 0.9672 0.8982 0.5677 0.8453 0.9860
Predicted R2 -7.0160 -7.5731 -1.1472 -2.0822 -0.4004 -3.5707 -2.3061 -0.2095 -2.9667 -13.896 -4.8503 -0.073
PRESS 28.97 75.40 20.15 43.02 36.39 156.86 156.94 119.73 939.72 8447.37 3195.22 643.30
30
Results & Discussions
Experimental dissolution data of optimal formula fitted into Korsmeyer-Peppas model.
Korsmeyer-Peppas model:
Mt
M∞
a
n
===
=
Cumulative amount of drug released at time tCumulative amount of drug at infinite timeConstant incorporating structural and geometric characteristics of the deviceRelease exponent, indicative of the mechanism of release.
31
Tan K S - Aug 2007
Tablet Swelling Profile: Optimal Formulation
0
4
8
12
16
20
0 2 4 6 8 10 12 14Time (Hour)
Dia
met
er (
mm
)
32
Tan K S - Aug 2007