11
DETERMINATION OF CRITICAL MANUFACTURING AND FORMULATION VARIABLES FOR A HYDROPHILIC MATRIX TABLET FORMULATION USING AN IN VITRO DISCRIMINATORY DISSOLUTION METHOD
P. Stark,l A. Kinahan,l S. Cunningham,l 1. Butler,l T. O'Hara,l A. Dunne,2 1. Connolly/ and 1. Devanel
IVIVR Co-operative Working Group lElan Corporation pic, Athlone, Co. Westmeath, Ireland
2University College Dublin Dublin, Ireland
1. INTRODUCTION
Formulation and manufacturing variables have been reported to influence the in-vitro release from hydrophilic matrix formulations. Studies on formulation and manufacturing variables have examined drug concentrationl-4 and particle sizel,3,4 polymer viscosity grade l-5 and particle size2 tablet compression force l,6 and formulation additives2,4,5,7.
A controlled release hydrophilic matrix tablet formulation of a Class I drugS containing Hydroxypropyl methylcellulose (HPMC) as the rate-controlling agent has been developed in this laboratory. A novel in-vitro discriminatory dissolution method has been developed for this product by examining formulations of differing HPMC concentrations9.
This in-vitro dissolution method has been used to identify parameters which influence the release of this drug from the HPMC matrix system.
This study is concerned with the determination of the critical manufacturing and formulation variables for a hydrophilic matrix tablet formulation using an in-vitro in-vivo relationship developed using a novel in-vitro discriminatory dissolution method. The experimental work was carried out in two stages. The first study was performed according to a Design of Experiments plan to indicate the critical variables for this formulation. The critical factors identified in the first study were then included in a second study in conjunction with another potentially critical formulation variable to produce products of different in-vitro profiles which were then evaluated in-vivo. Both experiments are discussed individually.
In Vitro--in Vivo Correlations, edited by Young et al. Plenum Press, New York, 1997 137
138 P. Stark et al.
2. MATERIALS AND METHODS - EXPERIMENT 1
2.1. Formulation Details
The hydrophilic matrix formulation evaluated in this study contains 60% Hydroxypropyl methylcellulose (HPMC) as the rate controlling agent. Formulation details of the hydrophilic matrix tablet are found in Table 1.
2.2. Experimental Design
A randomised Factorial Design of Experiments plan was used to determine critical manufacturing and formulation variables for the hydrophilic matrix formulation. The manufacturing and formulation parameters investigated were particle size of active, HPMC particle size, diluent grade and target tablet hardness. The particle size of the active, HPMC particle size and diluent grade factors were investigated at two levels using a 23 Factorial design which was replicated twice to give a total of 16 runs i.e. two runs for each combination of the factors evaluated. Each run was split into two and tabletted at low and high levels of tablet hardness to give a total of 32 runs in the experiment. Table 2 gives details of the experimental runs.
2.3. Method of Manufacture
2.3.1. Manufacture afTablet Blend. The active and excipients (except the lubricant) were blended for 10 minutes in a V-blender. Addition of the lubricant to the blend was followed by blending for 5 minutes and potency of the blend was determined. Two blends of 2kg were produced for each of the eight combinations of drug particle size, HPMC particle size and diluent grade resulting in 16 tablet blends in total.
2.3.2. Manufacture afTablets. Each tablet blend was split into two and one half tabletted at a target tablet hardness of 150N and the other half tabletted at a target tablet hardness of 250N on a Horn-Noack Rotary Tablet Press resulting in 32 tablet batches in total. This yielded two genuine replicates of each of the 16 combinations of the levels of the four factors. Tablet hardness was determined at set-up and at intervals through-out tableting using a Checkmaster hardness tester. Full details of the tablet batches manufactured and the associated factor levels may be found in Table 3.
Table 1. Formulation details of a hydrophilic matrix tablet
Component Function %w/w
Class I drug Active Hydroxypropyl methylcellulose (HPMC) CR polymer 60 Diluent Filler q.s Glidant Glidant <1% Lubricant Lubricant <1%
Hydrophilic Matrix Tablet Formulation
Table 2. Factors and levels at which they were investigated (experiment 1)
Factor
Class I drug particle size HPMC particle size Diluent grade Target Tablet Hardness
Low level
D90 = 160llm D90 = I 851lm D90 = 170llm 150 Newtons
2.4. In Vitro Dissolution Testing
High level
D90 = 2161lm D90 = 3041lm
D90 = 230llm 250 Newtons
139
One batch of each of the sixteen combinations of hydrophilic matrix tablets manu-factured were evaluated in-vitro using the discriminatory dissolution method according to a Balanced Incomplete Block Design of Experiments plan. In-vitro dissolution testing was carried out using a BioDis Apparatus (USP III) with a 20 mesh polypropylene screen, at 30 dips/minute in 250ml potassium di-hydrogen phosphate buffer (KH2P04, 0.32M) main-
Table 3. Tablet batches manufactured and the associated factor levels (experiment 1)
Tablet lot Blend lot Drug particle HPMC particle Target number number size size Diluent grade hardness (N)
PDI4082 PDI4070 High Low High 250 PDl4083 PDI4065 High High Low 250 PDI4084 PDI40n High High High 150 PDI4085 PDI4078 Low High Low 150 PDI4086 PDI4068 High Low High 250 PDI4087 PDI4067 Low Low Low 150 PDI4088 PDl4070 High Low High 150 PDI4089 PDI4065 High High Low 150 PDl4090 PDI4066 Low High Low 150 PDl4091 PDI4069 High High Low 150 PD14092 PDI4076 Low High High 150 PDl4093 PDl4069 High High Low 250 PDI4094 PDI4079 Low Low High 250 PDI4095 PDl4068 High Low High 150 PDI4096 PDl4067 Low Low Low 250 PDI4097 PDI4075 Low High Low 250 PDI4098 PDI4074 High Low Low 150 PDl4099 PDI4079 Low Low High 150 PDI4100 PD14072 High High High 150 PDI4101 PDI4073 Low Low High 150 PDI4102 PD14072 High High High 250 PDI4103 PDI4076 Low High High 250 PDI4104 PDI4080 High Low Low 150 PDl4105 PDI4075 Low High Low 150 PDI4106 PDI4078 Low High Low 250 PDl4107 PDI4080 High Low Low 250 PDI4108 PDI4071 Low Low Low 250 PDI4109 PDI4066 Low High Low 250 PDI4110 PDI4073 Low Low High 250 PDI4111 PDI4071 Low Low Low 150 PD14112 PDI40n High High High 250 PD14113 PDI4074 High Low Low 250
140 P. Stark et al.
tained at 37°C. Analysis of the dissolution samples was performed using a Waters 480 HPLC system with UV detection at 220nm.
3. RESULTS AND DISCUSSION - EXPERIMENT 1
Table 4 summarizes the in-vitro dissolution obtained for each of the batches examined ordered by AUC I-8 hours' The mean effects of each of the factors examined in Experiment 1 are discussed.
3.1. HPMC Particle Size
The in-vitro method indicated HPMC particle size as the primary factor associated with the in-vitro release of the active drug from the matrix tablets at a fixed HPMC concentration. Increasing the HPMC particle size from a low level to a high level resulted in a faster dissolution profile, as illustrated in Figure 1.
3.2. Tablet Hardness
Tablet hardness also was observed to be important with faster dissolution profiles associated with the lower tablet hardness as shown in Figure 2.
3.3. Active Particle Size and Diluent Grade
The main effects of particle size of the Class I drug and the diluent grade used were not shown, to be important using this method, as displayed in Figures 3 and 4, respectively.
Table 4. Batches tested using VSP III dissolution test ordered by AVe 1~8 hours
HPMC particle Tablet target Drug particle Lot No. AUCj.8hour.> size hardness Diluent grade size
PDI4108 441.16 Low High PHIOI Low PDI4094 442.29 Low High PHI02 Low PDI4113 446.30 Low High PHIOI High PDI4101 449.93 Low Low PHI02 Low PDI4111 450.38 Low Low PHIOI Low PDI4082 456.91 Low High PHI02 High PD14087 462.32 Low Low PHIOI Low PDI4095 467.93 Low Low PHI02 High PDI4103 478.31 High High PHI02 Low PDI4097 478.49 High High PHIOI Low PD14104 482.18 Low Low PHIOI High PDI4102 485.59 High High PHI02 High PDI4093 506.81 High High PHIOI High PD14084 509.84 High Low PHI02 High PDI4089 515.30 High Low PHIOI High PDI4085 528.95 High Low PHIOI Low PDI4090 531.74 High Low PHI02 Low
Hydrophilic Matrix Tablet Formulation
110
100
90
80
70
~ . 0 60
~ c . 50 0
,. 40 1
,al 20 I 10 1
O~
Profile
Time (hours)
Low HPMC PS
Error bars extend to 2 Standard Errors either side of meon
10
High HPMC PS
Figure 1. Mean in-vitro dissolution profiles of tablets containing low/high levels of HPMC particle size.
Profile:
Time (hours)
Low Hardness
Error bars extend to 2 Stondard Errors either side of meon
10
High Hardness
Figure 2. Mean in-vitro dissolution profiles of tablets manufactured at lowlhigh levels of tablet hardness.
141
142 P. Stark et aL
110
100
90
80
70
~ · 0 60 · ~ c · 50 u · ~
40
30
20
10
--~-------r.- -----,-- -, -------,---1 --T---,--,---,..-----, ,---~--,-~---,------.--
Profile:
Time (hours)
Low drug PS
Error bars extend to 2 Standard Errors eIther side of mean
10
High drug PS
Figure 3. Mean in-vitro dissolution profiles of tablets containing low/high levels of drug particle size.
110
100
90
80
70
60
50
40
30
20
10
Profile:
Time {hours}
Low Diluent PS
Error bors extend to 2 Stondard Errors either side of mean
10
High Diluent PS
Figure 4. Mean in-vitro dissolution profiles of tablets containing lowlhigh levels of diluent particle size.
Hydrophilic Matrix Tablet Formulation 143
3.4. HPMC Particle Size and Target Hardness
The HPMC particle size was demonstrated to have an effect at both levels of hardness i.e increasing HPMC particle size resulted in an accelerated dissolution profile for the low and high levels of hardness. Figure 5 indicates the faster dissolution profile for the combination high HPMC particle size/low tablet hardness. Conversely, tablets of low HPMC particle size and high tablet hardness had the slower dissolution profile. An interaction between HPMC particle size and target tablet hardness is the difference between the effect of HPMC particle size at high target hardness and the effect of HPMC particle size at low target hardness. This interaction was not considered to be important.
4. CONCLUSIONS - EXPERIMENT 1
HPMC particle size and tablet hardness were identified as potential critical manufacturing variables using the in-vitro discriminatory dissolution method, while the main effects of particle size of the active and of the diluent grade were not found to be important. The interaction between HPMC particle size and target tablet hardness was not considered to be important. At this stage, a confirmatory in-vivo study would normally be carried out using products which produced the slowest and fastest in-vitro dissolution profiles respectively. However, it was decided to first evaluate an additional factor which could potentially impact on the in-vivo performance of the product in a second experiment.
110
30
HPMC PS/Tablet Hardness
Time (hours)
--- Low HPMC/Low H High HPMC/Low H
Error bars extend to 2 Standard Errors either side of mean
Low HPMC/High H High HPMC/High H
10
Figure 5. Mean in-vitro dissolution profiles of tablets containing low/high HPMC particle size manufactured at lowlhigh tablet hardness.
144 P. Stark et aL
5. INTRODUCTION - EXPERIMENT 2
Prior to carrying out a confirmatory in-vivo evaluation of the critical manufacturing and formulation variables identified in Experiment 1, the chemical substitution of the HPMC was identified as another potentially important variable. Dahl et al. demonstrated that the in-vitro release ofNaproxen from a HPMC matrix tablet formulation was directly proportional to the hydroxypropyl substitution'o. Typically, the manufacturers supply HPMC with the methoxy concentration specification range of between 19% - 24% and hydroxypropoxy concentration range of 7% - 12%. The chemical structure of HPMC is shown in Figure 6. The extremes of the first experiment and this additional factor were evaluated together in this second experiment.
6. MATERIALS AND METHODS - EXPERIMENT 2
6.1. Experimental Design
Four batches of the hydrophilic matrix formulations were manufactured using HPMC material which had different ratios of methoxy substitution to hydroxypropyl substitution, denoted here as the MeO/HOPrO ratio. HPMC Lot Nos. C08909 and C07978 were considered to have high and low MeO/HOPrO ratios respectively as calculated in Table 5.
Four batches were manufactured which evaluated high and low methoxy/hydroxypropyl ratios. Based on the first experiment, high HPMC particle size was confounded with low target tablet hardness and low HPMC particle size with high target tablet hardness. Tablet manufacture and in-vitro dissolution testing was completed as described for Experiment 1. Formulation details for each of the batches manufactured are shown in Table 6.
7. RESULTS AND DISCUSSION - EXPERIMENT 2
7.1. In Vitro Evaluation
It can be seen from Figure 7 that changing the ratios ofmethoxy/hydroxypropyl substitution (MeO/HOPrO ratio) from a low to the high level results in faster dissolution pro-
~R
RO 0 o RO
CH2 I OR
OR I
R ~ ·H. ·CH, or ·CH, CH(OH) CH,
OR 0,--
n
Figure 6. The chemical structure of hydroxypropyl methylcellulose.
Hydrophilic Matrix Tablet Formulation 145
Table 5. The chemical substitution of hydroxypropyl methylcellulose
Hydroxypropyl methylcellulose Lot No. C08089 C07978 Manufacturers' specification
Methoxy Substitution (MeO) 22.9% 22.6% 19.0% - 24.0 % Hydroxypropyl Substitution (HOPrO) 9.6% 11.5% 7.0%- 12.0 % MEO/HOPrO ratio 2.39(high) 1.97(1ow)
files for both low HPMC particle size /high tablet hardness and high HPMC particle size /low tablet hardness batches. This effect was observed to be greater than the effects of the HPMC particle size or the tablet hardness. The in-vitro dissolution profiles are ranked from fastest to slowest in Table 7. From Figure 7, it can also be seen that the spread of the resultant dissolution profiles for the four batches were wider than the spread of dissolution results from the 40% - 80% HPMC batches previously evaluated in this laboratory9. These in-vitro results therefore indicated that batches manufactured using HPMC material with high or low extremes of methoxy/hydroxypropyl ratio would potentially be bio-inequivalent, assuming that the in-vitro test was indeed predictive of in-vivo performance.
7.2. In Vivo Evaluation
In-vivo evaluation examined the products which produced the fastest and slowest invitro profiles. A 60% HPMC product, evaluated in a previous studl, was also included as a reference. Its in-vitro dissolution profile was similar to the product with a low substitution ratio/ high HPMC particle size/low tablet hardness, ranked third in Table 7. The fourth product (ranked second in Table 7) had a high methoxy/hydroxypropyl substitution, a low HPMC particle size and a high tablet hardness and was included on the basis that the shape of its dissolution profile differed from that observed for the other lots examined. An immediate release product was included for deconvolution purposes.
The in-vivo study demonstrated that all products are bioequivalent to the reference product in terms of extent and rate of absorption with no significant difference in the Cmax or AVC reported between the treatments. Figure 8 illustrates the mean plasma profiles after administration. The in-vitro test is over-sensitive and thus produced false positives, while the formulation is more robust to the manufacturing and formulation variables evaluated than originally considered.
8. CONCLUSIONS - EXPERIMENT 2
The critical manufacturing variables of HPMC particle size, tablet hardness and chemical substitution of the HPMC do not impact on the in-vivo performance of this hydrophilic matrix formulation. These data show that novel in-vitro dissolution method is
Table 6. Tablet batches manufactured and the associated factor levels (experiment 2)
MeO/HOPrO ratio HPMC particle size Target hardness (N)
High Low 250 High High 150 Low Low 250 Low High 150
146
120
110
100
90
80
70
60
501
40 !
30 "
20 "I
10-
MeOHOPrO Ratio/HPMC PS/Hardness
Error bars extend to 2 Star.dard Errors either side of mean
H IL!H L IL IH
10 11
L IH IL H IH IL 40% HPMC - - -- 807.; HPMC
P. Stark et aL
12
Figure 7. Mean in vitro dissolution profiles of tablets containing low and high levels of MeO/HOPrO, HPMC particle size.
overselective to the formulation and manufacturing variables evaluated in these experiments. Further optimisation of this in-vitro dissolution system is on-going to optimise the selectivity of the in-vitro dissolution method.
9. CONCLUSIONS
The use of Design of Experiments plans has proved to be an effective strategy in determining the potential effect of critical formulation and manufacturing variables. A discriminatory in-vitro test method is required to screen products with different formulation and manufacturing variables in order to determine the most appropriate candidates for in-vivo biostudy. In-vivo evaluation of products which have different in-vitro profiles is required to be completed in order to confirm the predictive nature of the in-vitro test and to avoid erroneous assumptions which might lead to a overly restrictive manufacturing process.
Table 7. In-vitro rank order of tablet batches (experiment 2)
MeO/HOPrO ratio HPMC particle size Target hardness (N) Rank order
High Low High 2 High High Low Low Low High 4 Low High Low 3
Hydrophilic Matrix Tablet Formulation 147
600
500
~ 400 u g " 0 300 .~
;:,
" " u
" 200 0 u '" 8 i(J 100 ;:;:;
0
0 4 8
--- High / Low / High
. . . . . . High / High / Low
- . - . Low / Low / High
- - - - - - 60% Reference product
12 16
Time (hours)
PARAMETER
Log c""" HlHIL vs UUH HlHILvs 60% UUH vs 60%
Log AUC1nf HlHIL vs UUH HlHIL vs 60% UUHvs60%
Figure 8. Mean plasma profiles of tablets which had differing in vitro profiles.
9O%CI
82 - 100 83 - 102 92 - 113
84 - 107 84 - 106 88 - 112
10. REFERENCES
I. Ford, M.H. Rubinstein and J.E. Hogan, Formulation of sustained release promethazine hydrochloride tablets using hydroxpropyl methylcellulose matrices, lnt. J. Pharm., 24 (\ 985) 327-338.
2. Alderman, A review of cellulose ethers in hydrophilic matrices for oral controlled-release dosage forms, Int. J. Pharm. Technol. Prod. Manuf., 5 (1984) 1-9.
3. Ford, M.H. Rubinstein and J. Hogan, Propranolol hydrochloride and aminophylline release from matrix tablets containing hydroxpropyl methylcellulose, lnt. J. Pharm., 24 (\ 985) 339--350.
4. Ford, M.H. Rubinstein and J.E. Hogan, Dissolution of a poorly water soluble drug, indomethacin, from hydroxypropyl methylcellulose controlled release tablets, J. Pharm. Pharmacol., 37-supplement (\ 986) 33P.
5. Ford, M.H. Rubinstein, A. Changela and J.E. Hogan, Influence of pH on the dissolution of promethazine hydrochloride from hydroxypropyl methylcellulose controlled release tablets, J. Pharm, Pharmacol., 37-supplement (\ 986) 115P.
6. Korsmeyer, R. Gumy, E. Doelker, P. Buri and N.A. Peppas, Mechanisms of potassium chloride release from compressed, hydrophilic, polymeric matrices: Effect of entrapped air, J. Pharm. Sci., 72 (1983) 1189--1191.
7. Ford, M.H. Rubinstein, F. McCaul, J.E. Hogan and P.J. Edgar, Importance of drug type, tablet shape and added diluents on drug release kinetics from hydroxypropyl methylcellulose matrix tablets, Int. J. Pharm., 40 (1987) 223-234.
8. Amidon, H.L. Lennemas, v.P. Shah and J.R. Crison, A Theoretical Basis for a Biopharmaceutic Drug Classification: The Correlation of in Vitro Drug Product Dissolution and in Vivo Bioavailability, Pharm. Res., 12 (1995) 413-420.
9. C. Farrell, J. Butler, P. Stark H. Madden and J. Devane, The Development of a Novel In-Vitro Discriminatory Dissolution Method for a Class I Drug in a Matrix Tablet Formulation, in press.
10. Dahl, T. Calderwood, A. Bormeth, K. Trimble and E. Piepmeier, Influence of Physico-chemical Properties of Hydroxypropyl Methylcellulose on Naproxen Release from Sustained Release Matrix Tablets, J. Controlled Release, 14 (1990) \-\0.