`
`TECHNOLOGY CONFERENCE
`
`,
`
`,
`
`,, SflmguLngm 7 ,
`
`7 W,
`
`April 12th - 14th 1994
`
`19-SEP-1984 BLDSCBE§§§"7§“
`MRMCEUTICRL TECl-NOLOGY COHERENCE
`
` 5
`
`3
`
`'J
`
`Tuesday
`April 1 2th 1994
`
`Mylan v. Qualicaps, |PR2017-OO203
`QUALICAPS EX. 2043 -1/12 >
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 1/12
`
`

`

`A CHARACTERIZATION OF THREE HPMC SUBSTITUTION GRADES:
`
`RHEOLOGICAL PROPERTIES AND DISSOLUTION BEHAVIOUR
`
`MC. Bonferoni, S. Rossi, R. Sinistri, C. Caramella
`
`Department of Pharmaceutical Chemistry, University of Pavia, V.Ie Taramelli 12,
`
`27100 Pavia, Italy
`
`INTRODUCTION
`
`Hydroxypropylmethylcelluloses (HPMC) represent a wide famin of polymers: each
`
`of the three subtitution types described in the US Pharmacopeia (2906, 2910 and
`
`2208) is available in a wide variety of molecular weights and under different
`
`trademarks.
`
`The extensive and successful use of these polymers in pharmaceutical formulation
`
`justifies the efforts still recently made to thoroughly characterize them. Both the
`
`choice of the most suitable grade and the assessment of brand to brand variability
`
`and lot to lot reproducibility, require in fact the knowledge of the functionally
`
`relevant properties of the polymer.
`
`in particular, in hydrophilic matrix formulation,
`
`the influence of either the substitution type or the viscosity grade of the polymer on
`
`the physical properties of the gel layer that is formed around the matrix is worth
`
`investigation (1,2).
`
`Alderman described marked differences in drug release control between the three
`
`USP grades of HPMC.
`
`In his findings, the 2906 grade (Methocel® F) gave the
`
`fastest release, followed by the 2910 grade (Methocel® E) while the 2208 grade
`
`(Methocel® K) produced the slowest release. These differences were explained by
`
`considering the differing proportions of the hydrophobic methoxyl groups and of the
`
`hydrophilic hydroxypropyl groups that characterize the three polymer grades. The
`
`rate of polymer hydration was pointed out as a critical property towards controlled
`
`drug release (3).
`
`More recently, a few studies dealt with the characterization of the three grades of
`
`HPMC, with various results. V.S. Georgiannis et al. (4) compared for example the
`
`three grades of HPMC in the formulation of floating matrices. The behaviour
`
`observed agreed with the assumption of a K>E>F rank order of hydration rates. On
`
`the contrary, Mitchell et al. found that the three grades of HPMC performed similarly
`
`in hydrophilic matrices containing propranoiol hydrochloride. They studied the
`
`130
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 2/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 2/12
`
`

`

`influence of substitution type of HPMC by using several techniques. Only cloud
`point determination was able to differentiate between the three grades, while no
`evidence of differences could be found when hydration rates were directly
`measured (2, 5).
`
`Besides the rate of hydration or gelation, other polymer properties have been
`considered in the literature as possibly relevant to drug release control. Among
`these are, for example,
`the viscosity and the gel strength of the matrix outer gel
`layer.
`it sounds conceivable that, as already suggested by Alderman (3), the
`viscosity and gel strength of the outer layer affect its resistance to dilution and
`erosion; the effect on drug release will depend, of course, whether erosion is a
`limiting step or not. The relationship between gel strength and erosion has been
`recently studied by Mitchell et al. They used a penetrometer to investigate
`differences in gel strength between the three substitution grades of HPMC. which
`resulted to behave very similarly (2).
`
`Another approach to the characterization of the gelified HPMC is based on
`rheological measurements. Previous studies have pointed out that the simple
`quotation of the viscosity grade of the polymer (which is expressed by the viscosity
`of a 1% or 2% polymer solution) does not provide a sufficient description of its
`rheological properties (6). Rheological characterization should be able to describe
`the interactions that take place between polymer chains in the gel network and that
`are,
`in turn, responsible for disentanglement and erosion processes.
`in order to
`better characterize these interactions, in a recent study viscoelasticity analysis was
`used. The NaCMC behaviour was investigated and a close relationship was found
`between a series of viscoelastlc parameters and sensitivity to erosion of the
`polymer. ln particular, it was found that an increase in resistance of the polymer to
`dilution (from hydrated gel-like solutions) and to erosion (from matrices) was
`accompanied by an increase in creep viscosity and oscillation parameters (storage
`modulus G' and loss modulus G") of the polymer gels (7).
`Aim of the present work was therefore to assess whether this kind of rheological
`analysis allow to reveal differences between three substitution grades of the same
`viscosity grade of HPMC. Creep and oscillatory tests were performed on 5% and
`7% (w/w) polymer solutions. Erosion rate was moreover evaluated from tablets, and
`the dissolution rate from already hydrated polymers (dilution) was measured on 5%
`w/w samples.
`
`in order to assess to what degree the observed differences between the three
`polymers affected drug release, matrices containing 60% of polymer and either a
`very soluble drug (diprophylline) or an insoluble one (acetazolamide) were
`
`Mylan v. Qualicaps, |PR2017-00203
`‘3‘
`QUALICAPS EX. 2043 - 3/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 3/12
`
`

`

`prepared and tested for drug release. For matrix preparation,
`
`the same
`
`granulometric fraction of the three HPMC grades was used.
`
`EXPERIMENTAL
`
`MATERIALS
`
`Three grades of HPMC were tested: Methocel® K4M, Methocel® E4M and
`Methocel® F4M (Colorcon Ltd, Orpington, UK). As model drugs diprophylline
`
`(Proter, Milan,
`
`l) and acetazolamide (Sigma Chimica, Milan,
`
`l) were used. Lactose
`
`was USP XXII grade.
`
`METHODS
`
`Cloud point measurement
`
`Cloud point measurements were carried out according to Mitchell et al (8) on 2%
`
`w/v solutions in distilled water. Trasmittance % was read at 800 nm by means of a
`
`Spectracomp spectrophotometer (Advanced Products, Milan, l).
`
`Rheological studies
`
`Polymer solutions at 5% w/w and 7% w/w were prepared in distilled water, taking
`
`into account the content in water of the polymers. The samples were stored
`
`overnight in refrigerator to allow complete hydration and were analysed within 24 h
`
`from the preparation. Rheological analysis was performed with a Bohlin CS
`
`Rheometer (Bohlin Reologi, W. Pabish, Milan,
`
`I) equipped with a cone and plate
`
`system (CP 4/20). All measurements were conducted at 37 :t 2 °C. Dynamic
`
`(oscillatory) tests, at frequencies ranging between 0.1 and 4.0 Hz, and constant
`
`stress (creep) tests were carried out in the linear viscoelastic response range. From
`
`creep curves residual viscosity was calculated; from oscillation tests (3' (storage
`
`modulus) and G" (loss modulus) were obtained and tga (G"/G') was calculated.
`
`Erosion studies
`
`Erosion studies were performed by measuring the amount of HPMC dissolved both
`
`from gel-like solutions (dilution) and from tablets. All the tests were performed at
`37°C.
`
`The same 5% w/w solutions characterized for their rheological behaviour were
`
`centrifuged to remove entrapped air bubbles, and poured into cylindrical holders
`
`132
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 4/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 4/12
`
`

`

`(41 mm diameter and 12 mm height). These holders were placed in USP XXII
`
`vessels containing 500 ml desaerated water; paddle apparatus (at 2.5 cm above
`
`the samples) was used at 25 rpm.
`
`300 mg tablets of sieved fractions (105-180 pm) of HPMC were prepared by means
`
`of a hydraulic press for KBr discs (Perkin Elmer) equipped with a manometer, at 5
`
`tons for 1 minute; a flat punch of 13 mm diameter was used. The tablets were glued
`
`at the bottom of rotating discs. 500 ml of desaerated water in USP XXlI vessels
`
`were used; rotation speed was 100 rpm. Both in the case of erosion and of dilution
`
`tests, 5 ml samples were withdrawn at defined times and replaced with fresh
`
`medium. All samples were filtered before analysis. The amount of dissolved
`
`polymer was quantified by means of the anthrone method (9).
`
`Anthrone method
`
`The reagent was prepared by dissolving 50 mg of anthrone in a mixture of 28 ml
`
`water and 72 ml concentrated sulphuric acid. A mixture of 0.5 ml of sample and 2.5
`
`ml of
`
`reagent was heated 15 minutes in boiling water,
`
`refrigerated and
`
`spectrophotometrically read at 625 nm. Four standard solutions ranging from 0.1
`
`and 0.4 mg/ml were analysed together the samples and used to calculate a
`
`calibration line.
`
`Drug release studies
`
`Matrices (200 mg total weight) containing either diprophylline or acetazolamide
`
`were prepared by direct compression using an hydraulic press at 3 tons for 1 min,
`
`with a 10 mm convex punch. Sieved fractions (105-180 pm) of HPMC were used.
`
`The matrices contained 60% of polymer and either 40% of acetazolamide or 30% of
`
`diprophylline (10% of USP XXll lactose was used in this case as diluent).
`
`The release profiles were obtained in USP XXlI basket apparatus at 100 rpm, in
`
`desaerated water (1000 ml for diprophylline; 500 ml for acetazolamide). Automatic
`
`sampling was performed and absorbence was spectrophotometrically read at 273
`
`nm for diprophylline and at
`
`265 nm for acetazolamide (Spectracomp
`
`spectrophotometer; Advanced Products, Milan, l).
`
`RESULTS
`
`The values of cloud point measured for the three polymers (70 °C for HPMC K4M,
`
`57 °C for F4M and 54 °C for E4M) resulted in good agreement with those obtained
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 5/12
`
`‘33
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 5/12
`
`

`

`by Mitchell et al. (2) . They suggests that, for the three batches here examined, the
`K4M grade clearly differentiated from the other two grades, and that the F4M grade
`had slightly higher affinity for the hydration medium than the E4M grade.
`The creep viscosity values obtained from constant stress analysis (creep test) are
`given in Table l: for both concentrations, the K4M grade showed the highest and the
`E4M grade the lowest viscosity, whereas the F4M grade had intermediate viscosity
`values. The same rank order is observed in Table ll, where the results of the
`
`dynamic test at 0.1 Hz frequency are given for the 5% w/w solutions.
`
`Table l: Residual viscosity values (from creep test) (Pas) for 5% and 7% w/w
`
`solutions. mean (isd) of 3 replicates
`
`5%
`
`7%
`
`K4M
`
`E4M
`
`F4M
`
`314(i104)
`
`78 (-106)
`
`145 (:275)
`
`1200 (i 70 0)
`
`307 (a: 28 7)
`
`517 (i 38 8)
`
`Table ll: G' and G“ values (Pa) for 5% w/w solutions at 1 Hz frequency. mean (isd)
`
`of 3 replicates
`
`G'
`
`G"
`
`K4M
`
`E4M
`
`F4M
`
`47.7 (i 0.97)
`
`16.3 (i 0.14)
`
`24.8 (i 1.77)
`
`71.7 (:1: 1.70)
`
`45.5 (i 0.14)
`
`64.9 (i 2.55)
`
`The dependence of G' and G“ on frequency is illustrated in Figure 1 for 7% w/w
`solutions. From these profiles it is possible to determine the intersection point of G'
`and G" curves on the frequency axis. When the frequency is lower than the
`
`time enough in each oscillatory cycle for the
`there is
`intersection point,
`rearrangement of the polymer chains, and viscous response can develop and
`prevail over the elastic one. The frequency value at which the elastic response
`overcomes the viscous one can be different depending on the material and is
`
`usually lower for samples with more pronounced elastic character. The results
`given in Figure 1 show that K4M grade had intersection point at lower frequency,
`
`meaning higher elasticity, with respect to F4M and E4M.
`
`134
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 6/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 6/12
`
`

`

`a G,
`+ G'-
`
`1
`
`2
`
`3
`
`4
`
`frequency (Hz)
`
`G G.
`+ Gll
`
`1
`
`2
`
`3
`
`4
`
`frequency (Hz)
`
`_G_ G,
`+ G"
`
`1
`
`2
`
`a
`
`4
`
`frequency (Hz)
`
`K4M
`
`E4M
`
`HM
`
`1500
`
`1000
`
`500
`
`o
`
`0
`
`1 500
`
`1 000
`
`500
`
`O
`
`0
`
`1500
`
`1000
`
`500
`
`o
`
`o
`
`N
`a.
`
`Gn
`
`.
`
`aa
`
`.
`
`Figure 1: Dependence of G' and G" on frequency for the three HPMC grades. mean
`(isd) of 3 replicates
`
`Mylan v. Qualicaps, |PR2017-00203
`‘35
`QUALICAPS EX. 2043 - 7/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 7/12
`
`

`

`This result is strictly related to the dependence on frequency of the tangent of the
`
`phase angle between stress and strain (thor loss tangent); this is illustrated, for
`
`7% w/w solutions of the three polymers, in Figure 2. The loss tangent is the ratio
`
`between loss modulus G" and storage modulus (3': lower values of this parameter
`
`indicate that the elastic component is more pronounced with respect to the viscous
`
`one, The lowest values of tga that can be observed in Figure 2 at each frequency for
`
`the K4M grade (followed by F4M and E4M) means once more that the prevalence
`
`of elasticity over viscosity occurs according to the K>F>E order.
`
`tga
`
`frequency (Hz)
`
`Figure 2: Dependence of loss tangent (tga) on frequency for 7% w/w solutions of the
`
`three HPMC grades. mean (red) of 3 replicates
`
`The release profiles of the polymers from 5% w/w hydrated samples (dilution) are
`
`shown in Figure 3_ They are in agreement with the rheological characterization: it is
`
`conceivable that
`
`the same forces (weak chemical
`
`interactions and physical
`
`entanglements)
`
`that make the polymer chains to withstand strain during
`
`viscoelasticity tests, also hinder the polymer chains to leave the gel-like sample and
`to dissolve.
`
`The erosion profiles (from tablets) are illustrated in figure 4; although very close to
`
`each other, they show that the erosion from tablets follows a different order than
`
`from gel-like solutions. These differences can be attributed to the combined effect of
`
`hydration rate of the polymer and of the concentration gradient that is formed
`
`around the matrix. The cl0ud point of the K4M grade means in fact that this sample
`
`136
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 8/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 8/12
`
`

`

`has very high hydration rate; this can overcome the slow release of K4M chains
`
`from the gelified layer. Concerning the concentration of the polymer in the gel layer,
`
`as previous studies have pointed out (6.7), the dependence of the viscosity on
`
`concentration can be relevant for erosion phenomena.
`
`200
`
`mgdissolved
`
`150
`
`100
`
`50
`
`time (h)
`
`Figure 3: influence of HPMC grade on erosion profiles from 5% w/w gels (dilution).
`
`mean (isd) of 6 replicates
`
`mgdissolved
`
`time (h)
`
`Figure 4: Influence of HPMC grade on erosion profiles from matrices of polymer
`
`alone. mean (:sd)016 replicates
`
`Myrisa7n v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 9/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 9/12
`
`

`

`Figures 5 and 6 illustrate the release of diprophylline and of acetazolamide
`
`respectively. from matrices containing 60% ot HPMC.
`
`it
`
`is possible to observe that no substantial differences can be seen when the
`
`soluble diprophylline is used as model drug, and diffusion—controlled release is
`involved.
`
`100
`
`80
`
`60
`
`40
`
`20
`
`o
`
`O
`
`G K4M
`4- E4M
`-fl- F4M
`
`2
`
`4
`
`6
`
`8
`
`time (h)
`
`3
`U!
`
`m4
`
`’ E
`
`°\o
`
`Figure 5: Release profiles of diprophylline from matrices containing 60% of the
`
`three HPMC grades. mean (:sd) of 3 replicates
`
`°/oreleased
`
`0
`
`6
`
`12
`
`18
`
`24
`
`Figure 6: Release profiles of acetazolamide from matrices containing 60% of the
`
`three HPMC grades. mean (isd) of 3 replicates
`
`138
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 10/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 10/12
`
`

`

`in the case of the less soluble drug (acetazolamide), when erosion of the matrix is
`
`likely to play a more relevant role, differences between the behaviour of the three
`
`HPMC grades become more evident and seem to parallel the erosion behaviour of
`tablets of polymer alone.
`
`CONCLUSIONS
`
`to thouroughly describe the rheological
`Viscoelasticity tests are instrumental
`properties of a sample; they proved here to be sensitive enough to evidence
`differences, although very subtle, between materials having similar structure.
`Good parallelism has been found between the rheological behaviour of a gel and
`its rate of dilution. This confirms that the polymer~polymer and polymer-solvent
`interactions that are responsible for the gel network structure and for its sensitivity to
`erosion can be quite well described by a deep rheological analysis.
`The erosion from tablets resulted however very slow and the differences between
`
`the three grades could not be directly related to the rheological behaviour.
`The use of sieved fractions should have reduced the effect of particle size on
`porosity of the matrix and in turn on water intake. However, it must be taken in mind
`that other factors, such as the hydration rate of the polymer and its gradient of
`concentration along the outer gel layer, can play a role in matrix systems.
`Concerning drug release, the three HPMC grades controlled very similarly the
`release of diprophylline;
`in the case of poorly soluble acetazolamide, where
`erosion is likely to be more important, release rates show the same order as
`
`erosion rates from polymer tablets.
`
`Acknowledgements
`
`The authors wish to thank Dr. J.E. Hogan (Colorcon Ltd, Orpington, UK) for the kind
`gift of the Metl’iocel® samples.
`
`Work partially supported by MURST.
`
`REFERENCES
`
`1) CD. Melia: Hydrophilic matrix sustained release systems based on
`polysaccharide carriers. Crit. Rev. Ther. Drug Carrier Systems 8 (1991), 395-421
`
`13Iglylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 11/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 11/12
`
`

`

`2) K. Mitchell. J.L. Ford, D.J. Armstrong, P.N.C. Elliott, J.E. Hogan, C. Restron: The
`
`influence of substitution type on the performance of methylcellulose and
`
`hydroxypropylmethylcellulose in gels and matrices. lnt. J. Pharm. 100 (1993), 143-
`154
`
`3) DA. Alderman: A review of cellulose ethers in hydrophilic matrices for oral
`
`controlled release dosage forms. lnt. J. Pharm. Tech and Prod. Mfr, 5 (3) (1984), 1-9
`
`4) VS. Georgiannis, D.M. Rekkas, P.P. Dallas and NH. Choulis: Floating and
`
`swelling characteristics of various excipients used in controlled release technology.
`
`Proceedings of the 12th Pharmaceutical Technology Conference (1993), 54-73.
`
`5) K. Mitchell, J.L. Ford, D.J. Armstrong, P.N.C. Elliott, C. Rostron, J.E. Hogan: The
`
`influence of concentration on the release of drugs from gels and matrices
`
`containing Methocel®. Int. J. Pharm., 100 (1993), 155-163
`
`6) MC. Bonteroni, C, Caramella, M.E. Sangalli, U. Conte, R.M. Hernandez, J.L.
`
`Pedraz: Rheological behaviour of erodible polymers and drug release from
`
`erodible matrices. J. Controlled Rel, 18 (1992), 205-212
`
`7) MC. Bonferoni, S. Rossi, C. Caramella, U. Conte: Rheological properties and
`
`sensitivity to erosion of sodim carboxymethylcellulose. Proceedings of 6th
`
`International Conference on Pharmaceutical Technology (1992), 241-249
`
`8) K. Mitchell, J.L. Ford, D.J. Armstrong, P.N.C. Elliott, C. Rostron, J.E. Hogan: The
`
`influence of additives on the cloud point, disintegration and dissolution of
`
`hydroxypropylmethylcellulose gels and matrix tablets.
`233-242
`
`Int. J. Pharm., 66 (1990),
`
`9) D.L. Morris: Quantitative determination of carbohydrates with Dreywood's
`
`anthrone reagent. Science, 107 (1948). 254
`
`140
`
`Mylan v. Qualicaps, |PR2017-00203
`QUALICAPS EX. 2043 - 12/12
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2043 - 12/12
`
`