Journal of Controlled Release, 9 (1989)
`
`169-175 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands 169 GEL-MATRIX SYSTEMS EXHIBITING BIMODAL CONTROLLED RELEASE FOR ORAL DRUG DELIVERY A.C. Shah*, N.J. Britten, L.S. Olanoff and J.N. Badalamenti
`
`Pharmacy Research,
`
`The Upjohn Company, 30 1 Henrietta Street, Kalamazoo, Ml 4900 1 (U.S.A.) (Received June 9, 1988; accepted in revised form December 19, 1988)
`
`Certain types of hydroxypropyl methylcellulose ethers, when admixed with a therapeutic agent
`and compressed into a solid dosage form, are found to display a bimodal drug release profile. The
`bimodal profile is characterized by a rapid initial release of drug, followed by a constant rate of
`release, and then a second mode of fast drug release at the terminal phase. Release profiles can
`be selectively modified by varying the viscosity, concentration, and the combination of methyl-
`cellulose polymers. The mechanism of release appears to involve initial surface erosion, polymer
`gelation, a steady-state counter-current permeation of water and dissolved drug across the gel
`layer, dissolution of gel from the outer surface, and subsequent disintegration of the gel. A bimodal
`oral controlled release delivery system which produces an increased rate of drug release in the
`latter phases of dissolution, may offer some advantages over constant zero-order release systems
`for maintaining uniform drug levels in the body. Bimodal release profiles were obtained for as-
`pirin, ibuprofen, adinazolam, flurbiprofen, and other investigational drugs.
`
`INTRODUCTION
`
`An oral controlled release system which re- leases drug at a zero-order rate is often consid- ered an ideal system for maintaining constant drug levels. This is based on the assumption that drug absorption occurs rapidly through the en- tire GI tract, so that the rate of elimination dic- tates the rate at which the drug must release from the dosage form. However, for many drugs, absorption is moderately slow in the stomach, rapid in the proximal intestine, and it declines sharply in the distal segment of the intestine. This means that to maintain constant drug lev- els, the delivery system should release drug in such a way that it is able to compensate for the changing drug absorption pattern in the GI tract *To whom correspondence should be addressed. by increasing the drug release rate in regions of slow absorption. Thus, a release system with a variable rate of release may indeed be more de- sirable than a constant zero-order release sys- tem. The bimodal release system provides such a variable rate system. For drugs where gas- trointestinal absorption is uniformly rapid, serum concentrations can be produced which will closely reflect the differential tempered re- lease rates of the bimodal dosage form. Bimodal release is characterized by a rapid initial release, followed by a period of constant release, then a second phase of rapid drug re- lease. The bimodal release system can offer two major advantages over the other systems:
`
`it produces rapid drug release during the initial and later phase to provide rapid onset of action and to compensate for the relatively slow ab- sorption in the stomach and large intestine; (2 ) 0168-3659/89/$03.50 0 1989 Elsevier Science Publishers B.V.
`
`(1)
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 1/7
`
`

`

`170
`
`1 ]
`
`it can be used to design programmed pulse re- lease oral drug delivery systems for the thera- peutic agents that perform more effectively when drug levels undergo periodic changes. In this study, a group of hydroxypropyl methylcellulose polymers have been identified that exhibit bimodal release profiles when com- bined with therapeutic agent in a solid dosage formulation. Although some of these polymers have been widely used and studied [
`
`for their application in oral controlled release drug de- livery, their ability to provide bimodal drug re- lease has not been previously reported. MATERIALS AND METHODS Materials Bimodal hydroxypropyl methylcellulose ethers (HPMCs) which have been used in these studies include Metolose 65SH-50, 65SH-400, 65&H-1500,65SH-4000, Metolose 60SH-4000, 90SH-100, and 90SH-15,000 (Shin-Etsu Ltd., Japan) as well as Methocel F4-M (Dow Chem- ical Company, MI). The following non-bimodal HPMCs were tested: Methocel A4-C, A4-M, A15-LV, A15-C, E4-M, E5, E50, E15-LV, E50-LV, K4-M, K15- M, KlOO-M, and KlOO-LV (Dow Chemical Company). Metolose SM-1500, a methylcellu- lose, was evaluated as a release rate modifier. The drugs utilized were adinazolam, flurbipro- fen, aspirin, and ibuprofen. the active ingredient and length of drug release desired. Tablets were manufactured by either direct compaction of the mix on a carver press or on a rotary tablet compressor, after dry granulation. Dissolution apparatus Tablets were subjected to dissolution rate testing using an automated six-place rotating filter-stationary basket system [2], a water bath with Tecam C-400 circulator, a lo-place Master Flex pump, a Perkin-Elmer Lambda 3 UV-Visible spectrophotometer, and an IBM- AT computer. Dissolution procedure Dissolution studies were conducted in a me- dium of 0.05 A4 phosphate buffer, pH 7.2, tem- perature 37”C, and at a stirring speed of 300 rpm. Release rates were determined by contin- uous spectrophotometric analysis of the dis- solved drug. Viscosity measurements Viscosity testing was performed on selected HPMC samples. 2.0% solutions were prepared by dissolving the polymer in water at 60°C. Samples were refrigerated for 12 hours, then warmed to 20” C prior to analysis. Measure- ments were made using a Schott Gerate AVS 400 automated viscometer. Clinical study procedure Tablet preparation Tablet preparation consisted of through mix- ing of one or more of the hydroxypropyl meth- ylcellulose ethers with therapeutic agent and ingredients conventional in tablet making such as stearic acid, magnesium stearate, silicon dioxide, etc. The content of the bimodal poly- mer (s) comprised anywhere from 5 to 99% by weight of the total formulation, depending on Ten normal male volunteers were enrolled in a crossover trial to evaluate the pharmokinetics of the bimodal sustained release adinazolam and conventional formulations. Each dosing se- quence was separated by a one week interval. Venous blood specimens were collected by in- dividual venipunctures at 0,0.5, 1, 2,3, 4,6,8, 10, 12, 14, 16, 24 and 30 hours after dosing, to measure plasma drug concentration.
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 2/7
`
`

`

`RESULTS AND DISCUSSION Cellulose ethers are being utilized exten- sively in the design of oral controlled release dosage forms [ 31. In the present study, a series of hydroxypropylmethylcellulose ether poly - mers that provide bimodal and non-bimodal re- lease profiles from polymer-drug matrix tab- lets have been identified. As seen in Table 1, bimodal release was observed from some poly- mers supplied by Shin-Etsu Ltd., while most of the polymers obtained from Dow Chemical Company, with almost identical specifications, gave non-bimodal release profiles. It is not clear yet why these differences in release character- istics exist. Methods of manufacture, ionic composition, variations in the distribution of substituent groups, or distribution of molecular weight fractions may be the cause. Viscosity measurements were conducted on selected polymer lots. Experimental viscosity values were found to be well within the manufacturer’s specified ranges. Conventional oral sustained release formu- lations usually exhibit either a zero-order re- lease profile, where the rate of drug release is essentially constant, or a profile where the rate
`
`TABLE 1
`
`HPMC polymers exhibiting bimodal and non-bimodal
`release
`
`Bimodal
`
`Non-bimodal
`
`60SH-4000
`65SH-50
`65SH-400
`65SH-1500
`65SH-4000
`90SH-100
`90SH-15000
`F4-M
`
`A4-C
`A4-M
`AlB-LV
`A15-C
`E4-M
`E5
`E50
`E50-LV
`E15-LV
`K4-M
`K15-M
`KlOO-M
`KlOO-LV
`
`171
`
`of release decreases with time. Figure 1 com- pares declining, constant, and bimodal release profiles. Figure 2 illustrates the bimodal release of a controlled release aspirin tablet containing 40% Metolose 65SH-4000. The upper curve is the cumulative amount of drug released as a func- tion of time, while the bar graph comprises the same data presented as the rate of release as a function of time. A similar bimodal release pro- file is shown in Fig. 3 for an ibuprofen tablet made with 60% Metolose 65SH-4000. The bar graph clearly shows an initial rapid release of drug, followed by a period of constant release, and then a second mode of fast drug release, that are characteristic of bimodal release. The rate of drug release and the shape of the release profile can be selectively modified by varying the viscosity, concentration, and com- bination of hydroxypropyl methylcellulose used. Other excipients such as methylcellulose, so- dium carboxymethylcellulose, hydroxypropyl v , , , ,
`
`5
`
`10
`
`15 20 25 30
`HOURS
`
`35
`
`TIME
`
`Fig. 1. Comparison of bimodal release with constant (t)
`and declining (4)
`release profiles. A: Cumulative drug re-
`lease. B: Rate of release as a function of time.
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 3/7
`
`I
`

`

`100 1
`
`14 1
`
`0- 0 2 4 6 6 10 12 14 16 18
`HOURS
`
`6.
`
`
`
`_ 12
`
`21rLL___ 0
`
`5
`
`15
`10
`HOURS
`
`20
`
`25
`
`B.
`
`'0 2 4 8 8 10 12 14 16 18
`
`15
`10
`HOURS
`
`20
`
`25
`
`Fig. 2. Bimodal release profiles of 300 mg aspirin tablet. A: Fig. 3. Bimodal release profile of 60 mg ibuprofen tablet. A: Percent drug dissolved versus time. B: Bate of drug release Percent drug dissolved versus time. B: Bate of drug release (percent dissolved/hour) versus time. (percent dissolved/hour) as a function of time.
`
`cellulose, lactose, starch, binders, fillers, and
`disintegrating agents, may also be added to the
`formulation in alter the release rate. The effect
`of polymer viscosity on the time course of bi-
`modal release is demonstrated in Fig. 4. Gen-
`erally the use of higher viscosity polymers re-
`sults in slower drug release from the matrix
`tablets. In Fig. 5, the addition of increasing
`amount of methylcellulose (1500 cP) to a for-
`mulation containing Metolose 65SH-4000,
`shifts the bimodal profile to the left, increasing
`the rate of drug release.
`Bimodal release is evident in results of in ui-
`dissolution testing of 30 mg adinazolam tab-
`lets, prepared by incorporating 65% Metolose
`65SH-1500 and 25% Metolose SM-1500 with
`the drug (Fig. 6). Preliminary evaluation of
`these tablets in a clinical biopharmaceutical
`study suggests that the bimodal sustained re-
`lease is also evident in uitro. A comparison of
`serum levels in normal volunteers, as a result of
`oral administration of the bimodal tablets and
`
`80-
`
`60-
`
`40-
`
`20-
`
`.40% Metolose 65SH-50
`.40% Metolose 65SH-1500
`m40% Metolose 66SH-4000
`
`OY
`0
`
`5
`
`15
`10
`HOURS
`
`20
`
`25
`
`Fig. 4. Effect of polymer viscosity on drug release from 200 mg ansaid tablets.
`
`conventional tablets is given in Fig. 7. The bi-
`modal sustained release tablets provided an ini-
`tial serum drug level of approximately 20 ng/
`ml after an hour, which remained essentially
`constant fk seven hours. A subsequent rise was
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 4/7
`
`172
`tro
`A.
`

`

`UO
`
`1 * so 0
`
`60
`
`$
`
`B
`
`Fig. 7. Comparison of human serum levels of rapid release versus bimodal controlled release adinazolam tablets. Drug concentration in serum (ng/ml) as a function of time. c] = Conventional release tablet; 0 = controlled bimodal release tablet. initial gelation phase during which there is some erosion of the tablet matrix, which provides rapid dissolution. This phase is followed by counter-directional diffusion of dissolved drug and solvent through the polymer gel, substan- tial swelling, and considerable dissolution of the gel surface. Upon complete hydration of the polymer, ~sinte~ation of the gel-matrix oc- curs. The relative contribution and effects of each of the processes involved needs to be ac- cessed before a comprehensive description of the mechanism of bimodal release is possible. The kinetics of drug release from an erodible device, in the absence of diffusion, under per- fect sink conditions, have been presented by Hopfenberg [ 41 and by Baker and Lonsdale [ 51. Korsmeyer et al. [ 6,7 ] developed mathemat- ical models for solute release from noneroding swellable polymer systems, based on a drug dif- fusion coefficient which is dependent on the concentration of absorbed solvent, in a func- tional form consistent with the free volume theory of diffusion. Their findings showed an exponential dependence of the solute diffusion coefficient on pen&rant concentration. A mathematical model incorporating the dif- fusion of solvent in polymer, diffusion of dis- solved polymer in solution, and characteriza-
`
`~90% Metolose 65SH.4000
`055% Metolose 65SH-4000;
`35% ME~O~S~-l~
`.45% ~etolosa 65SH-4000:
`45% Metolose SM.1500
`
`0
`
`15
`10
`HOURS
`
`20
`
`25
`
`Fig. 5. Effect of rate modifier (Metolose SM-1500) on drug release from 30 mg deracyn tablets.
`._,_._._.-.......-.~...
`/
`
`A.100 1
`
`80
`
`6
`
`2
`
`5
`
`lb
`10
`HOURS
`
`20
`
`25
`
`Fig. 6. Bimodal release profile of 30 mg adinazolam tablet. A: Percent drug dissolved versus time. B: Bate of drug re- lease (percent dissolved/hour) as a function of time. seen at about nine hours, with nearly steady drug levels for the next five hours. Bimodal drug release from hydroxypropyl methylcelhrlose gel-matrix seems to involve an
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 5/7
`
`173
`5
`I
`

`

`174
`
`tion of the dissolution of polymer taking place at the interface separating the two diffusion processes, was provided by Tu [8], for swella- ble polymer systems in which diffusion is the predominate mechanism in the rate of dissolution. The rate of drug release by a simple diffusion erosion process has been described by Heller and Baker [9 1. In this equation the permeabil- ity, P, increases are an exponential function of time, due to progressive hydration of the gel matrix: BP,exp(Kt)C, ‘I2 -=-- t
`
`1
`
`Here nli, is the drug released at time t, A is the surface area, PO is the initial permeability of polymer to drug, K is the first order rate con- stant, C, is the initial concentration of drug in polymer, and t is time. Although this relationship should yield a bi- modal-type release profile, it does not take into account all the processes that occur in a swell- able eroding gel-matrix system. Diffusional drug release from polymer matri- ces involving moving boundaries has been in- vestigated by Rudolph [ 10,111, Peppas et al. [ 121, and Lee [ 13-161. Lee developed a refined integral method for the analysis of drug release from swellable eroding systems, which was ap- plied to the early stages of release, before the diffusion fronts begin to overlap the center. He characterized drug release with the quantitiy Ba/D, where B is the rate constant for surface erosion, a is the half-thickness of the polymer slab, and L) is the dif~sion coefficient of the drug in the polymer. This quantity is basically the erosion rate to matrix permeability ratio, which measures the relative contribution of the erosion and diffusion processes. The above described mathematical models, reported in the literature, do not, however, ad- equately describe all of the interrelated pro- cesses involved in the swellable erodible system which exhibits bimodal drug release. Individual processes are now being investigated, at all stages of the tablet’s dissolution, to further elu- cidate the physi~o-chemical principles which govern the mechanism of bimodal drug release. ACKNOWLEDGEMENT Human bioavailability data were provided by G. Peng. REFERENCES
`
`H. Lapidus and N.G. Lordi, Drug release from com-
`pressed hydrophilic matrices, J. Pharm. Sci., 57 (1986)
`1292-1301.
`AC. Shah, C.B. Peot and J.F. Ochs, Design and eval-
`uation of a rotating filter-stationary basket in vitro
`dissolution test apparatus. I. Fixed fluid volume sys-
`tem, J. Pharm. Sci., 62 f 1973) 671-677.
`U.S. Patent Numbers 4,389,39~ 4,369,172; 4,259,314;
`4,357,459; 4,226,849; 3,870,7X$ 3,065,143; and
`4,167,558.
`H.B. Hopfenberg, in: D.P. Paul and F.W. Harris
`(Eds.
`Controlled Release Polymeric Formulations,
`ACS Symposium Series Vol. 33, American Chemical
`Society, Washin~n~ DC!, 1976, pp. 26-32.
`R.W. Baker and H.K. Lonsdale, Erodible controlled
`release systems, Amer. Chem. Sot., Div. Org. Coat.
`Plast. Chem., Pap., 36 (1976) 235.
`R.W. Korsmeyer, S.R. Lustig and N.A. Peppas, Solute
`and penetrant diffusion in swellable polymers. I.
`Mathematical modeling, J. Poiym. Sci., Polym. Phys.
`Ed., 24 (1986) 395-408.
`R.W. Korsmeyer, Solute and penetrant diffusion in
`swellable polymers. II. Verification of theoretical
`models, J. Polym. Sci., Polym. Phys. Ed., 24 (1986)
`409-434.
`Y.O. Tu, A multiphase Stefan problem describing the
`swelling and dissolution of glassy polymer, Q. Appl.
`Math-, (1977) 269-285.
`J. Heller and R. Baker, Controlled Release of Bioac-
`tive Materials, Academic Press, New York, NY, 1980,
`p. 8.
`F.B. Rudolph, Diffusion in a multicomponent inho-
`mogeneous system with moving boundaries. I. Swell-
`ing at constant volume, J. Polym. Sci., Polym. Phys.
`Ed., 17 (1979) 1709-1718.
`F.B. Rudolph, Diffusion in multicomponent inhomo-
`geneous system with moving boundaries. II. Increas-
`ing or decreasing volume (swelling or drying), J. Po-
`lym. Sci., Polym. Phys. Ed., 18 (1980) 2323-2326.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 6/7
`
`) ,
`

`

`175
`
`12 N.A. Peppas, R. Gurny, E. Doelker and P. Buri, Mod- 15 P.I. Lee, Kinetics of drug release from hydrogel ma- eling of drug diffusion through swellable polymeric trices, J. Controlled Release, 2 (1985 1 277-288. systems, J. Membrane Sci., 7 (1980) 241-254. 16 P.I. Lee and N.A. Peppas, Prediction of polymer dis- 13 P.I. Lee, Diffusion release of a solute from a polymeric solution in swellable controlled release systems, J. matrix - Approximate analytical solutions, J. Mem- Controlled Release, 6 (1987) 207-215. brane Sci., 7 (1980) 255-275.
`P.I. Lee, Controlled drug release from polymeric ma- trices involving moving boundaries, in: D.H. Lewis (Ed. ) , Controlled Release of Pesticides and Pharma- ceuticals, Plenum, New York, NY, 1981, pp. 39-48.
`
`14
`
`Mylan v. Qualicaps, IPR2017-00203
`QUALICAPS EX. 2049 - 7/7
`
`