Applying Quality by Design for Extended Release Hydrophilic Matrix Tablets

Understanding the effect of excipients’ material attributes on the final drug product is integral to quality by design (QbD). The authors examine the effect and interaction of variations in the material properties of hypromellose on powder flow, the physical attributes of tablets, and in vitro drug-release profiles from two model formulations of extended-release hydrophilic matrix tablets using QbD principles.

Quality by design (QbD) is a systematic approach to designing and developing pharmaceutical formulations and manufacturing processes to ensure predefined product quality (1). In the case of hydrophilic matrix tablets, it is important to consider potential variability in material attributes of the rate-controlling polymer in addition to variability in the API properties and processing conditions (2–4). This proactive and enhanced understanding supports efficient pharmaceutical product development.

This study examines the effect and interaction of variations in hypromellose physicochemical properties on powder flow, the physical attributes of tablets, and in vitro drug-release profiles from two model formulations of extended-release (ER) hydrophilic matrix tablets using QbD principles. This article presents a QbD approach to determine the effect of material attributes on both the physical properties and in vitro drug-release performance of the matrix tablets.

The excipient hypromellose United States Pharmacopeia (USP) substitution type 2208 (Methocel K15M Premium CR, Dow Chemical) was used as the rate-controlling polymer for two case studies with a soluble drug (propranolol hydrochloride [HCl]) and slightly soluble drug (theophylline). Normal variation of Methocel material attributes (apparent viscosity, percent hydroxylpropoxyl (HP) substitution, and particle size) was studied at polymer concentrations of 15% w/w and 30% w/w. The study demonstrated consistent physical properties for direct-compression blends and subsequent tablet cores, irrespective of the Methocel concentration or drug included. In vitro drug release, however, showed greater sensitivity to material-attribute variability at lower polymer concentration.

The importance of QbD

QbD is a systematic approach to pharmaceutical development that results in increased quality and reduced costs. QbD means designing and developing formulations and manufacturing processes to ensure predefined product quality (1). Adoption of QbD principles for new-chemical-entity and generic-drug products is becoming an expectation by regulatory agencies to better ensure that high-quality medicines are available to the end-user, namely the patient. Building quality into drug products by design also benefits developers. Successful first-cycle approval, reduction of postapproval changes, and the potential of real-time release could offset initial investment associated with QbD implementation.

Importantly, enhanced understanding of the product and manufacturing process also can lead to the elimination of production rejects and recalls due to quality issues. Before FDA introduced QbD into the chemistry, manufacturing, and controls (CMC) review process in 2004, the amount of product waste due to manufacturing mistakes was reported to be as high as 50% (5). Clearly, for the end-user, the patient, drug-product recalls associated with quality issues, and potential shortages of medicines are a risk to health. For the manufacturer, these problems can lead to severe financial penalties due to loss of market share and even litigation. Needless to say, adverse publicity also can erode consumer confidence and damage a manufacturer's reputation.

The foundations of QbD for drug-product development are contained within the International Conference on Harmonization (ICH) quality guideline ICH Q8 (R2) Pharmaceutical Development (R2) (6). This guideline for pharmaceutical development includes "determining the critical quality attributes (CQA) of the drug substance (and) excipients and selecting the type and amount of excipient to deliver drug product of the desired quality" (6). This determination is of particular importance for designing drug products for ER applications, where the performance of the rate-controlling excipient is crucial to precisely deliver the required amount of drug over time. Typically, for ER technologies, such as hydrophilic matrices, barrier membrane-coated multiparticulates and osmotic delivery systems, the dose of the drug within a single unit is much greater than in an immediate-release product. Understanding the primary rate-controlling excipients' physiochemical properties (i.e., material attributes) is important to ensure robustness of the finished product and to mitigate any risk of batch-to-batch variability and/or potential premature drug release that could impact the patient.

Hydrophilic matrix products

Hydrophilic matrices are a well-established ER delivery platform due to their flexibility in delivering a wide range of drugs, relatively simple manufacturing, and generally good product stability and shelf-life. The majority of marketed hydrophilic matrix products use high-viscosity hypromellose (HPMC) as the rate-controlling polymer. HPMC polymers are semisynthetic materials derived from cellulose with chemical modification to add both the methoxyl (CH3–O–) and hydroxypropoxyl (CH₃CHOHCH₂–O–) functional groups. In addition to the type and distribution of these functional groups, the polymer molecular weight (measured indirectly by apparent viscosity) and particle size are key material attributes that could affect drug-product manufacturability and performance.

Methocel for hydrophilic matrix applications uses two types of chemical substituent groups signified by either "E" or "K" designations (7). Methocel E chemistry is the USP substitution type 2910; K chemistry is the substitution type 2208. The number that follows the chemistry designation identifies viscosity in millipascal-seconds (mPa·s), measured at 2% weight/volume aqueous solution at 20 °C. The letter "M" is used to represent a multiplier of 1000.

Along with the polymer, ER matrix formulations typically consist of the API, filler, binder, glidant, and lubricant. Other functional ingredients also may be added, such as additional polymers to modify the release rate, buffering agents to mitigate the effects of pH-dependent drug solubility, stabilizers, and surfactants. Commonly, a matrix-tablet formulation also will be film-coated with a conventional immediate-release coating or may be coated with a functional modified-release coating system.

Accordingly, the matrix formulation can be designed to influence the mechanism and rate of drug release. The design can include polymer type and concentration, drug solubility and dose, polymer-to-drug ratio, filler type and concentration, polymer-to-filler ratio, the particle size of the drug and polymer, and the shape of the matrix (8–12). Drug solubility is an important factor in determining the mechanism of drug release from hypromellose hydrophilic matrices (i.e., diffusion, diffusion and erosion, or erosion) and guides the selection of other excipients as well as the viscosity and chemistry grade of the hypromellose.

Nevertheless, as the principal rate-controlling excipient, it is important to assess the criticality of both polymer concentration and the effect of material-attribute variation (within the manufacturer's sales-specification limits) on the final drug-product quality. This knowledge is important to justify development of a robust formulation and to set an appropriate control strategy for consistent manufacture of a high-quality finished product.

Materials and methods

Two case studies were designed to investigate the influence of material attributes: the percent HP substitution, viscosity, and particle size on the functional performance of hydrophilic matrix-tablet formulations (2–3).

The rate-controlling polymer in the model formulations was Methocel K15M Premium CR (USP substitution type 2208). The designation of "15M" describes a relatively high-viscosity material, and the "CR" grade is designed for controlled-release applications.

Polymer concentration can be an important factor for matrix robustness. Two polymer concentrations, therefore, were evaluated: 30% w/w, which has been shown to produce robust formulations, and 15% w/w, which was considered relatively low and could result in performance differences of the hydrophilic matrix tablet associated with variability in the material attributes.

For these case studies, Methocel K15M Premium CR batches were carefully selected. Six of the batches were selected on the basis of having two out of three material attributes (percent HP, particle size, and apparent viscosity) within the nominal manufacturer sales-specification values, with the third property at the "high" or the "low" extremes of the normal sales-specification range. In addition, one batch had all three properties close to the nominal specification values, denoted as "center point" (see Table I). A total of 14 matrix formulations (seven each for 15% and 30% w/w polymer concentration) were prepared. The Methocel K15M Premium CR batches used in these studies will be referred to by the "batch name" listed in Table I.

Table I: Physiochemical properties of hypromellose (Methocel K 15 Premium CR , Dow Chemical) batches.

The methoxyl substitution content could be considered another material attribute for Methocel that may affect the robustness of the formulation. Prior assessment of the methoxyl content variation (from the manufacturer's sales-specification) showed this to be precisely controlled, and therefore, it was not considered to be a significant variable and was excluded from the study.

Propranolol hydrochloride ER model formulations

For the first study, the model API was propranolol HCl (soluble drug, 50 mg/mL, 160-mg dose). The formulation is detailed in Table II.

Table II: Extended-release model formulation containing propranolol hydrochloride as the active ingredient.

Tablet preparation procedure. Propranolol HCl, hypromellose, and microcrystalline cellulose were passed through an ASTM #30 mesh (600 µm) screen and mixed in a four-quart V blender (Model B Lab Blender, Patterson–Kelley) at 26 rpm for 10 min. Magnesium stearate was screened through an ASTM #40 mesh (400 µm) screen and added to the powder mixture, followed by blending for an additional 3 min. The final powder mixtures were compressed at 5–20 kN (compaction pressure of 70–280 MPa) using an instrumented 10-station rotary tablet press (Piccola, RIVA) at 20 rpm using standard round 9.52-mm concave tooling and a tablet weight of 350 mg.

The formulated powder blends were analyzed for bulk and tapped densities using a VanKel density tester (Model 10705, Varian), flowability using a flow tester (Sotax FT 300, Sotax), and loss on drying (LOD) using an infrared (IR) moisture balance (Model IR-200, Denver Instrument). Tablet weight, breaking force, diameter, and thickness were measured with an automated tablet tester (Multicheck V, Erweka). Tablet friability was measured using a VanKel friabilator (Varian) at 100 revolutions and 25 rpm. A dissolution study was performed using an USP Apparatus II, 100 rpm, with sinkers, and 1000 mL of a pH 6.8 phosphate buffer. Propranolol release was detected at a wavelength of 289 nm using a ultraviolet (UV)-visible spectrophotometer (Agilent 8453, Agilent Technologies) fitted with quartz flow cells of a 2-mm path length.

The similarity factor (f₂), which is a measurement of the similarity in the percentage of dissolution between two curves, was calculated by comparing the high versus the low end of the selected physicochemical property. Two dissolution profiles are considered similar when the f₂ value is > 50. In addition, the release exponent (n) and release-rate constant (k) were calculated by fitting the dissolution data to the Power Law equation (M_t/M_inf) = k tⁿ , where M_t is the amount of drug released at time t; M_inf is the amount of drug released over a very long time, which corresponds in principle to the initial loading; k is the kinetic constant; and n is the release exponent (12).

Theophylline ER model formulations

In the second study, the model API was theophylline anhydrous (slightly soluble drug, 8.3 mg/mL, 160-mg dose). The formulation is detailed in Table III.

Table III: Extended-release model formulation containing theophylline as the active ingredient.

Tablet-preparation procedure. Theophylline, hypromellose, lactose, and fumed silica (Cab-O-Sil, Cabot) were passed through an ASTM #30 mesh (600 µm) screen and mixed in a four-quart V blender (Patterson-Kelley) at 26 rpm for 10 min. Magnesium stearate was screened through an ASTM #40 mesh (400 µm) screen, added to the powder mixture, followed by blending for a further 3 min. The final powder blends were compressed at 15 kN (210 MPa) using an instrumented 10-station rotary tablet press (Piccola, RIVA) at 20 rpm using a standard round 9.52-mm concave tooling and a tablet weight of 350 mg.

All blends were analyzed for bulk and tapped density using a VanKel density tester (Varian) and LOD (Model IR-200, Denver Instrument). Tablets were examined for physical properties, including weight variation, thickness, and hardness as well as friability. Drug release was measured using an USP Apparatus II (VK 7000, Varian) at 100 rpm with sinkers and 1000 mL of deionized water at 37 ± 0.5 °C. Theophylline release was detected at a wavelength of 272 nm using a UV-visible spectrophotometer (Agilent 8453, Agilent Technologies) fitted with quartz flow cells of a 2-mm path length. The similarity factor (f₂) was calculated by comparing the high versus the low end of the selected physicochemical property. In addition, the release exponent (n) and release-rate constant (k) were calculated by fitting the dissolution data to the Power Law equation (11).

Results

Propranolol hydrochloride ER model formulations. The results indicated that at 30% polymer concentration, all propranolol blends exhibited comparable bulk/tapped density and powder flow. All matrix tablets had comparable hardness, tensile strength, and friability values. Similar results were observed for all formulations with 15% w/w polymer concentration, indicating that the material attributes (i.e., percent HP, particle size, and viscosity) of Methocel K15M CR had minimal or no influence on the physical properties of the formulated powder blends or tablets. All matrices showed low friability (≤ 0.06%) and consistent content uniformity (97.8–101.5%).

Propranolol HCl release was slower when polymer concentration increased from 15% to 30% w/w (see Figures 1–3). At both 15% and 30%, drug-release profiles were similar (f₂ = 63 and 68, respectively) despite variations in Methocel viscosity (see Figure 1). Use of higher polymer concentration (30% w/w) resulted in lower tablet-to-tablet variability as indicated by the error bars.

Figure 1: Propranolol hydrochloride release profiles-effect of viscosity.

The effect on drug release of the percent HP substitution of hypromellose on the drug-release profiles is shown in Figure 2. Here too, at both 15% and 30% polymer concentration, the drug-release profiles were similar (f₂ = 82 and 91, respectively) despite variations in percent HP content.

Figure 2: Propranolol hydrochloride release profiles-effect of percent hydroxypropoxyl

The effect of Methocel particle size on the drug-release profiles is shown in Figure 3. At 30% polymer concentration, the drug-release profiles were very similar (f₂ = 95) despite variations in particle size. At 15% polymer concentration, however, the batch with the larger particle size (low percentage through 230 mesh) gave a faster and dissimilar (f₂ = 46) drug-release profile compared with the batch with the finer particle size (high percentage through 230 mesh) of the polymer. In addition, tablet-to-tablet variability was higher in the formulation containing the coarser particle size in comparison to the center point and fine particle-size formulations. All formulations produced good results fitting to the Power Law equation (R² > 0.99). The release exponent (n) was in the range of 0.59–0.63 for 30% w/w polymer formulations and 0.48–0.56 for 15% w/w polymer formulations, indicating drug release mainly by diffusion (11).

Figure 3: Propranolol hydrochloride release profiles-effect of particle size.

Higher polymer concentration may decrease sensitivity of the formulation to minor variations in raw materials or the manufacturing process. The potential for particle-size variability to influence in vitro drug release was shown to be negated when higher concentration of Methocel K15M CR was used.

Theophylline ER model formulations. Study results indicated that comparable physical properties were obtained for theophylline powder blends and compressed tablets at both 15% and 30% polymer concentration. All matrices showed low tablet-weight variation (1.0–1.9%), low friability (≤ 0.14%), and consistent content uniformity (94.8–100.0%).

Theophylline release rates were lower when polymer concentration was increased from 15% to 30% (w/w) as shown in Figure 4. At both 15% and 30% polymer concentrations, drug-release profiles were similar (f₂ > 50) despite variations in Methocel viscosity, percent HP substitution, and particles size. Results for all formulations fit to the Power Law equation (R² > 0.99). The release exponent (n) was in the range of 0.50–0.62 for 30% w/w polymer formulations and 0.39–0.48 for 15% w/w polymer formulations, indicating that diffusion is the principal mechanism of drug release (13).

Figure 4: Theophylline release profiles–effect of particle size (n = 6; drug dissolution using USP

The linear-regression model also was applied to examine the relationship between drug-release response (i.e., release constant (k), release exponent (n) or time for 80% drug release [T_80%])and predictor variables (i.e., viscosity, percent HP, and particle size measured by percent through 230 mesh). Results indicated statistically an insignificant relationship (p value > 0.1).

Conclusion

The study demonstrated that evaluation of hypromellose materials-attribute variability on matrix formulation robustness can be readily determined. It was shown that material-attribute-variability effects can be dependent upon the rate-controlling polymer concentration. This observation has important implications for designing Methocel-based ER matrices.

Results indicate that the ranges studied for viscosity, percentag of HP, and particle size of Methocel K15M Premium CR had no significant effect on the physical properties of propranolol HCl or theophylline formulation blends and tablets. This finding is important for direct-compression processing because blend properties, such as flow and compactability, can impact CQA, such as content uniformity for a matrix formulation.

For both model formulations, the drug-release profiles from Methocel matrices were slower when the polymer concentration was increased from 15% to 30% w/w. At 30% polymer concentration, the drug-release profiles of propranolol HCl were similar (f₂ > 68) despite variations in viscosity, percent HP, and particle size. At 15% w/w polymer concentration, the drug-release profiles of propranolol HCl were similar (f₂ > 63) despite variations in viscosity and percent HP substitution; therefore, for these case studies, both material attributes were noncritical.

An early indication of risk associated with material-attribute variability is an important factor in formulation design and the subsequent manufacturing-process selection. The formulator can develop an enhanced understanding by building quality into its drug product by evaluating material attributes and "designing out" variability effects. Development of poorly designed and understood products can lead to manufacturing and cost inefficiencies, including customized excipient specifications and batch selection as well as producing out-of-specification drug products. The approach, presented in this study, provides a useful starting point for identifying and managing excipient material-attribute criticality when developing drug products through QbD strategies.

References

1. L.X. Yu, Pharm. Res. 25(4), 781–791 (2008).

2. T.D. Cabelka et al., "Application of Quality by Design Principles to the Formulation of aHydrophilic Matrix Tablet of a High Dose/High Solubility Drug," presented at the AAPS Annual Meeting (Los Angeles, 2009).

3. T.D Cabelka et al., "Application of Quality by Design Principles to the Formulation of Extended-Release Propranolol Hydrochloride Hydrophilic Matrix Tablets," presented at the AAPS Annual Meeting (New Orleans, 2010).

4. H. Deng et al., "Application of Quality by Design Principles to the Formulation of Extended-Release Theophylline Hydrophilic Matrix Tablets," presented at the CRS Annual Meeting and Exposition (National Harbor, Maryland, 2011).

5. A.S. Rathore and H. Winkle, Nature Biotechnol. 27 (1) 26–34 (2009).

6. ICH, ICH Q8(R2) Pharmaceutical Development (2009).

7. USP 30–NF 25, "USP Monographs: Hypromellose," online, www.uspnf.com, accessed May 10, 2008.

8. J.E. Hogan, Drug Dev. Ind. Pharm. 15 (6), 975–1999 (1989).

9. P. Gao et al., J. Pharm. Sci. 85 (7), 732–740 (1996).

10. M. Levina and A.R. Rajabi-Siahboomi, J. Pharm. Sci. 93 (11), 2746–2754 (2004).

11. K. Mitchell et al., Int. J. Pharm. 66 (1–3), 233–242 (1990).

12. A.R. Rajabi-Siahboomi, A. Nokhodchi, and M.H. Rubinstein, "Compaction Behavior of Hydrophilic Cellulose Ether Polymers," in Pharm. Technol. Yearbook: Tableting and Granulation (1998), pp. 32–40 .

13. R.W. Korsmeyer et al., Int. J. Pharm. 15 (1), 25–35 (1983).