- Email: [email protected]
Effect of carbamazepine on viscoelastic properties and hot melt extrudability of Soluplus®
International Journal of Pharmaceutics 478 (2015) 232–239
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Effect of carbamazepine on viscoelastic properties and hot melt extrudability of Soluplus1 Simerdeep Singh Gupta a , Tapan Parikh a , Anuprabha K. Meena a , Nidhi Mahajan b , Imre Vitez b , Abu T.M. Serajuddin a, * a b
College of Pharmacy and Health Sciences, St. John’s University, 8000 Utopia Parkway, Queens, NY 11439, USA Catalent Pharma Solutions, 14 Schoolhouse Road, Somerset, NJ 08873, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 24 July 2014 Received in revised form 2 October 2014 Accepted 12 November 2014 Available online 13 November 2014
The purpose of this study was to apply viscoelastic properties of polymer and drug-polymer mixtures to determine processing conditions for the preparation of amorphous solid dispersion by melt extrusion. A poorly water-soluble drug, carbamazepine (CBZ), was mixed with Soluplus1 as the carrier. Torque analysis using a melt extruder was performed at 10, 20 and 30% w/w drug concentrations and the effect of barrel temperature was studied. Viscosity of the mixtures either at fixed temperatures with different angular frequencies or as a function of temperature with the same frequency was studied using a rheometer. The viscosity of Soluplus1 and the torque exerted on the twin screws decreased with the increase in CBZ concentration. The viscosity versus temperature plots for different CBZ concentrations were parallel to each other, without the drug melting transition, indicating complete drug-polymer miscibility. Thus, the drug-polymer mixtures could be extruded at temperature as low as 140 C with 10% w/w drug load, 135 C with 20% w/w drug and 125 C with 30% w/w drug, which were, respectively, 50 C, 55 C and 65 C below the melting point of 191 C for CBZ. The differential scanning calorimetry (DSC) and powder X-ray diffraction (XRD) analyses of the binary mixtures extruded at 125–150 C showed absence of crystalline drug. A systematic study of miscibility and extrudability of drug-polymer mixtures by rheological and torque analysis as a function of temperature will help formulators select optimal melt extrusion processing conditions to develop solid dispersions. ã 2014 Elsevier B.V. All rights reserved.
Keywords: Hot melt extrusion Rheology Viscosity Polymer Drug-polymer miscibility Torque analysis Soluplus1
1. Introduction For oral delivery, drug substances must dissolve in the gastrointestinal (GI) fluid before their absorption. Ensuring adequate aqueous solubility in the GI tract has, therefore, emerged as the most difficult biopharmaceutical challenge in the development of oral drug products for optimal clinical outcome (Li et al., 2005). Some of the enabling technologies applied to the development of poorly water-soluble drugs include salt formation, particle size reduction, solubilization, lipid-based drug delivery, solid dispersion, etc. Williams et al. (2013) published an excellent article reviewing these and other strategies to address low solubility of drugs. There are, however, practical challenges and
* Corresponding author at: Department of Pharmaceutical Sciences, College of Pharmacy and Health Sciences, St. John’s University, 8000 Utopia Parkway, Queens, NY 11439, USA. Tel.: +1 718 990 7822; fax: +1 718 990-1877. E-mail address: [email protected] (A.T.M. Serajuddin). http://dx.doi.org/10.1016/j.ijpharm.2014.11.025 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.
limits with each of these technologies. Among various approaches to formulate poorly water-soluble drugs, solid dispersion, where the drug is usually dispersed in a water-soluble amorphous carrier either molecularly or in the amorphous state (Leuner and Dressman, 2000; Vasanthavada et al., 2008), has emerged as the most promising one. After oral administration, the water-soluble matrix of the solid dispersion dissolves in the GI fluid, releasing the drug either in solution or as finely divided precipitate that redissolves rapidly. However, despite extensive research on the solid dispersion technology for over 50 years, only a very limited number of products based on solid dispersion principles have been available in the market. Manufacturing difficulties and stability issues have been the primary reasons for the limited application of the technology for commercialization of drug products (Serajuddin, 1999). One common method of preparing solid dispersion is to dissolve drug and carrier in an organic solvent and then removing the solvent by various evaporation techniques, including spray drying. It is, however, difficult to get a solvent that would simultaneously dissolve the water-insoluble drug and the water-
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
soluble carrier in a reasonably acceptable volume of liquid. Handling of large-scale volumes of solvents during the dosage form development and manufacturing is also a major issue. The prospect for the development of solid dispersion has greatly improved during the past decade due to the introduction of hot melt extrusion (HME) in the pharmaceutical field (Breitenbach, 2002). HME has many advantages over other technologies used for solid dispersion (Crowley et al., 2007; Ghosh et al., 2012; Shah et al., 2013; Lakshman et al., 2008). It is a continuous, less time-consuming process that can be scaled up relatively easily. It is also less prone to batch-to-batch variability than other processes. Also, being solventfree, it poses minimal environmental hazards. However, there are still many challenges in the development of new drug products by melt extrusion. Being a relatively new technology in the pharmaceutical field, the formulation and processing parameters of HME are not fully understood. In most of the published reports, it is not defined how the appropriate polymers, drug-polymer blends and the processing temperatures are selected. A good understanding of the material properties, such as the glass transition temperature (Tg) of polymers, the melt viscosity of polymers as well as drug-polymer blends and the drug-polymer miscibility is needed before the development of processing conditions for any new products. The drug-polymer mixtures should also be extrudable at as low a temperature as possible to minimize potential of degradation of drug, polymer or both. Recently, we applied rheology to study viscoelastic properties of various polymers having PVP (Gupta et al., 2014), cellulosic (Meena et al., 2014) and methacrylate (Parikh et al., 2014) backbones to investigate their suitability for melt extrusion. Only the neat polymers were used in these studies to generate a database relevant to physicochemical properties of different polymers relevant to melt extrusion. It was established that the temperature range where the melt viscosity of a polymer falls between 1000 and 10,000 Pa s was the most suitable temperature for the melt extrusion process. However, a drug formulation intended for melt extrusion is not polymer alone. Polymers are used along with drugs, which may influence the processing conditions. Therefore, the objective of the present investigation was to determine what effect the presence of drug as well as the drug concentration would have on viscoelastic properties and extrudability of a polymer as a function of temperature. Soluplus1, a polyvinyl caprolactam–polyvinyl acetate–polyethylene glycol graft copolymer, was chosen as a carrier for melt extrusion. A poorly water-soluble drug, carbamazepine (CBZ), was selected for the comparative study of its influence on the drug-polymer miscibility and the processing conditions. To determine extrudability of the polymer and the drug-polymer mixtures, the moment of force exerted onto the twin screws of the extruder, also known as torque, was measured. It was hoped that the viscosity and torque analyses would not only help in identifying processing conditions for melt extrusion, they would also provide valuable information on the miscibility of drug-polymer binary mixtures. To confirm miscibility of the mixtures in the products, differential scanning calorimetry (DSC) and powder X-ray diffraction (powder XRD) studies of the extrudates of various drug-polymer mixtures prepared at different temperatures were conducted and the results were compared with that of the physical blends. 2. Materials and methods
233
Fig. 1. Chemical structures of (a) carbamazepine and (b) Soluplus1.
2.2. Methods 2.2.1. Torque analysis by hot melt extrusion (HME) The torque analysis was performed using Process 11 Parallel Twin-Screw Extruder (Thermo Scientific Inc., Waltham, MA, USA) to determine (a) what would be the range of processing temperature in the melt extruder for a particular material, (b) how long would it take for the drug-polymer mixtures to reach a constant torque value, (c) what effect the drug concentration would have on the torque at a particular temperature, and (d) what would be the effect of temperature on torque at a fixed drug concentration. Only the neat polymer was initially used to determine the temperature range for melt extrusion of the polymer itself. The polymer was added into the extruder at the rate of 1 g/min at 100 rpm screw speed. A screw design with low, medium, and high shear geometry of the mixing elements was used. The barrel temperature was initially set at 200 C, and the torque generated in the twin screws was recorded at this temperature and at the decreasing temperature intervals of 5 C till the maximum limit of torque (100%) of 12 Nm in the extruder used reached and the rotation of screws stopped. The temperature that would give a torque value of 60% (7.2 Nm) and less in the extruder was used for further studies. To determine how long would it take to reach a constant torque value at a specific temperature, the torque analysis was conducted with and without the drug. For this purpose, 1 g of the neat polymer (Soluplus1) was added into the extruder at the barrel temperature of 150 C and the screw speed of 100 rpm every minute for three consecutive points and the resultant torque was recorded. Then, the mixture of 30% CBZ in Soluplus1 was added at 1 g/min starting from the fourth addition till a constant torque in presence of drug was obtained. At these conditions, the output through the 1.5 mm circular die at the end of the extrusion barrel was about 1 g/min. The effect of drug concentration on the torque generated was studied by feeding different drug-polymer mixtures (0, 10, 20 and 30% w/w drug) to the melt extruder at the rate of 1 g/min, screw speed of 100 rpm and the barrel temperature of 150 C consecutively for 6 min (time was based on the results obtained from the previous study). The torque value obtained at the steady state for each drug concentration was recorded and the results were plotted against the drug concentration. To study the effect of temperature, a constant drug concentration of 20% w/w was used for extrusion at different temperatures, and the torque values were plotted against it.
2.1. Materials Carbamazepine was purchased from Sigma–Aldrich Co. (St. Louis, MO, USA). Soluplus1 was donated by BASF Corporation (Tarrytown, New York, USA). Structures of the materials are given in Fig. 1.
2.2.2. Rheology The rheology of the samples to analyze viscoelatic properties was studied using Discovery Hybrid Rheometer 2 with oven heating assembly (DHR-2, TA Instruments, Newcastle, DE, USA). The method of the preparation of samples for analysis was
234
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
described earlier (Gupta et al., 2014). Essentially, 1 g of material was weighed and compressed into a slug using a 25 mm ejection die at 5000 pounds of pressure using a Carver press. The slug (2.2 mm thick) was placed in between two flat plates of 25 mm diameter after calibration of the zero gap. For the analysis of pure polymer, frequency sweep was conducted at 110–180 C in increments of 10 C, and the plots of resultant viscosity versus angular frequency were generated. For the analysis of drugpolymer physical mixtures, the slug was equilibrated at 120 C for about 2 min and the temperature flow was ramped up from 120 to 200 C at 5 C/min. Responses were obtained in terms of log viscosity (Pa s) versus temperature ( C). The results obtained at 150 C were correlated with the torque analysis by hot melt extrusion. 2.2.3. Differential scanning calorimetry (DSC) The thermal analysis of physical mixtures and melt-extruded samples was performed using a modulated DSC (DSC Q200, TA Instruments, Newcastle, DE, USA). For physical mixtures, drug and polymer were weighed and mixed to obtain 10, 20 and 30% w/w drug concentration. In case of melt extruded samples, extrudates were cooled to room temperature and milled before analysis. Usually, 5– 10 mg of sample was placed in an aluminum pan that was crimped with a lid on. It was then heated from 25 C to 210 C at the rate of 10 C/min. The scans were analyzed for the presence of Tg and any possible melting peaks after the extrusion process to determine whether the drug had converted completely to the amorphous form after extrusion. To confirm the results of DSC analysis, the extruded samples were also observed visually and under the polarized light microscope. The visual observation showed that the extrudates were clear, and, under the microscope, there was no birefringence due to the presence of any crystalline drug. 2.2.4. Powder X-ray diffraction (XRD) To obtain the powder XRD patterns of different melt extrudates, a Shimadzu 6000 powder X-ray diffractometer (Shimadzu, Tokyo, Japan) was used. Physical mixtures were prepared by separately weighing drugs and polymer and then mixing in 10, 20 and 30% w/ w concentration of drug. Melt extruded samples were triturated into fine powder before analysis. During the XRD analysis, monochromatic CuKa radiation source was operated at 40 kV and 30 mA, with the scanning rate (2 u)/min over the range of 10– 60 (2u). Results were analyzed for the presence of crystalline peaks in the final products. Preliminary studies in our laboratory using the powder X-ray diffractometer showed that the detection limit of crystalline drug for the equipment was 5%, below which the crystalline drug peaks could not be clearly differentiated from the baseline noise. Some of the amorphous samples were also observed under the polarized light microscope, where the lack of birefringence even at <5% confirmed that the materials were amorphous. 3. Results 3.1. Determination of processing temperature The temperature ranges at which the polymer or the drugpolymer mixtures could be extruded through a twin-screw extruder were determined by correlating torque values generated in the extruder with the viscoelastic properties of materials used. For initial correlation, only neat Soluplus1 without the addition of drug was extruded, and the effect of a change in temperature on the moment of force, or torque, generated on the twin screws of the extruder is plotted in Fig. 2. In these experiments, the feed rate, screw speed and screw design were maintained constant. The torque value at 200 C was about 1.5 Nm which was equivalent to
Fig. 2. Effect of melt extrusion processing temperature on the torque subjected on the twin screws by neat Soluplus1 (n = 3). There was a gradual increase in torque observed with a decrease in barrel temperature, indicating that the stiffness in polymer increased with a decrease in temperature.
about 13% of the torque capacity of equipment used. As the temperature was decreased, there was an increase in the torque value, indicating that the stiffness of the polymer increased. At 135 C, the torque value increased to 10.8 Nm (90%), above which the maximum torque limit (12 Nm = 100%) of the equipment was attained. At 150 C, the torque value was about 7 Nm, which was equivalent to 60% of the equipment limit. Above 180 C, the torque was <20% and the extrudate flowing the die at the end of the barrel was too soft to form spaghetti-like strands and, therefore, not suitable for extrusion. Based on these results, the temperature range of 150–180 C was considered optimal for the melt extrusion of Soluplus1. The next step in the determination of processing temperature was to ascertain what would be the viscosity of the polymer in the temperature range identified by torque analysis. Since Soluplus1 is a graft copolymer with viscoelastic properties, it behaves like a solid (elastic) with the storage modulus greater than the loss modulus at temperatures less than and close to its Tg (Gupta et al., 2014). With an increase in temperature, it behaves more like a viscous liquid. A frequency sweep was performed at various angular frequencies to determine the viscosity of the neat polymer at different temperatures, and the results are given in Fig. 3. At all temperatures, there was a decrease in viscosity observed with an increase in angular frequency. The complex viscosity of the polymer at 110 C at all frequencies was above 100,000 Pa s, which implied that the polymer was very stiff. With an increase in temperature, the complex viscosity value decreased. At 150 C, the viscosity dropped below 10,000 Pa s and was about 1000 Pa s at about 180 C. In the temperature range of 150–180 C, the polymer could be considered extrudable as it was free flowing under the applied stress and at the same time it was not so soft that it would reaggregate after extrusion. These results demonstrate that the processing temperature for a certain polymer or drug formulation may be identified by conducting rheology experiments and no elaborate experiments with the melt extruders are necessary. Thus, the viscoelastic analysis serves as a good predictor for the processing temperature during melt extrusion. A similar conclusion was also reached by Kolter and Grycke (2012) in a technical report of BASF Corporation. However, the study was limited to several BASF polymers only. No detailed reports validating the concept for other polymers as well as for drug-polymer mixtures has been reported in peer-reviewed journals. 3.2. Effect of drug on extrudability of polymer After a viscosity window of 10,000–1000 Pa s for the extrusion for Soluplus1 was defined, the effect of drug on the extrudability of
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
235
Fig. 3. Effect of angular frequency and temperature on the complex viscosity of neat Soluplus1. With an increase in the rate of shear subjected on the polymer (given by angular frequency), there was a decrease in complex viscosity observed at all temperatures. With an increase in temperature from 110 to 180 C, there was a drop in viscosity from about 200,000 Pa s till about 1000 Pa s at low frequency values.
the polymer was studied. This was done at 150 C, where the torque value of the neat Soluplus1 was about 7 Nm or 60% of the equipment limit. The results presented in Fig. 4 show that it took time for the added material to completely fill the barrel and give the torque value that represented the sample. It was, therefore, important to determine the number of additions in a given time period to get a constant torque value. The feed rate and the screw speed were kept constant. The study was conducted starting with the neat polymer and then in presence of CBZ. The extrusion was conducted till a constant torque value was obtained. It may be observed in Fig. 4 that the constant torque value with neat Soluplus1 was about 7.2 Nm at 150 C for the first three additions of material at 1 g/min. After the addition of drug-polymer mixture (30% carbamazepine) as the fourth sample, the torque gradually decreased and reached a value of 2.5 Nm when the eighth sample of the mixture was added. The torque value remained at about 2.4 Nm for next 4 additions. Between the fourth and the seventh additions of materials at 1 g/min, the extrusion barrel was filled partly with the neat polymer and partly with the drug-polymer mixture, and only when the entire neat polymer extruded out of the barrel and the barrel was completely filled with the drug-
The analysis of torque during melt extrusion was also conducted using different concentrations of carbamazepine at a constant temperature of 150 C. The analysis was performed first at 0% drug concentration (neat Soluplus1) to obtain a constant torque value, and the study was then continued by using increasing concentrations of CBZ in binary mixtures (10, 20, and 30% w/w). As shown in Fig. 5, there was a gradual decrease in torque values with the addition of drug and the increase in drug concentration. The constant torque recorded without the drug was about 7.2 Nm. After the addition of 30% drug-polymer mixtures, the torque decreased to 2.8 Nm. The decrease in torque with an increase in drug concentration was linear with R2 value of 0.99. These results demonstrate that the presence of CBZ can greatly influence the processing conditions of drug-
Fig. 4. Torque analysis of Soluplus1 in presence of 30% CBZ at 150 C. At initial sample additions without the drug, the torque was about 7.2 Nm. After addition of sample with 30% drug, there was a gradual decrease in torque observed till a constant value was attained at about 2.4 Nm. The constant torque sample number was used in the further analyses.
Fig. 5. Effect of drug concentration on the extrudability of drug-polymer mixtures at 150 C. There was a decrease in constant torque value after addition of 10% w/w drug-polymer mixture as compared to neat polymer. There was a further decrease in constant torque with an increase in drug concentration.
polymer mixture that the torque stabilized. The results presented in Fig. 4 demonstrate that the presence of CBZ has major impacts on the extrudability of Soluplus1 as it reduces the torque within the extruder barrel. 3.3. Effect of drug concentration on extrudability of drug-polymer mixtures
236
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
3.5. Effect of drug and temperature on viscoelastic properties of drugpolymer mixtures
Fig. 6. Effect of change of temperature on the extrudability of 20% w/w drugpolymer mixture. There was an increase in constant torque value with a decrease in barrel temperature. Below 125 C, the maximum torque exerted on the equipment was attained.
polymer mixtures by apparently acting as a plasticizer. The plasticizing effect may, however, be different for different drugs and certain drugs may not have any plasticizing effect at all. Further work is continuing in our laboratory to determine the effect of drug properties on plasticizing effects. 3.4. Effect of temperature on extrudability of drug-polymer mixtures Once the effect of drug concentration on the extrudability at a constant temperature of 150 C was determined, it was important to understand the effect of a change of temperature on the torque subjected on the screws. The result is given in Fig. 6. Drug-polymer binary mixture with 20% w/w drug concentration was used. The extrusion was first conducted at 150 C, which gave constant torque value of about 4 Nm, as shown for 20% drug load in Fig. 5. In subsequent experiments, lower barrel temperatures were used. There was a gradual increase in torque to 5.6 Nm with the decrease in temperature till 135 C. There was a sharp increase in torque to 9.9 Nm at 125 C, below which the extruder stopped. These results demonstrate that much lower processing temperature could be used for the melt extrusion of CBZ-Soluplus1 mixtures than that of the polymer itself.
The rheological analysis was also conducted to determine the effect of drug on the polymer behavior at high temperature. Fig. 7 gives the complex viscosity profile of CBZ-Soluplus1 mixtures as a function of drug concentration as well as temperature. For the neat Soluplus1, the viscosity was over 100,000 Pa s at the starting temperature of 120 C, and gradually decreased with the increase in temperature (Fig. 7). The viscosity was slightly under 10,000 Pa s at 150 C, and this temperature was, therefore, used for torque analysis mentioned earlier. The viscosity dropped below 1000 Pa s when the temperature increased over 180 C. With the addition of CBZ at increasing concentrations of 10, 20 and 30%, decreases in viscosity were observed at all temperatures, and the plots appeared to be approximately parallel to each other as well as that with that of the neat Soluplus1. If the viscosity of 10,000 Pa s is considered to be acceptable for melt extrusion, Fig. 7 shows that that the processing temperature could be decreased from 150 C for the neat Soluplus1 to 130 C in presence of 30% CBZ. At the extrusion temperature of 150 C, all the mixtures were in the extrudable viscosity range of 1000 10,000 Pa s. It may also be observed that there was no peak or bump for all three mixtures that may be attributed to the melting point of the drug substance (191 C). This indicates that carbamazepine was miscible with Soluplus1 at temperature below its melting and no crystalline material remained at the melting temperature. 3.6. Thermal and powder XRD analyses of melt extrudates The formation of solid dispersion was confirmed by the DSC analysis. Theoretically, when a drug is molecularly dispersed, only one Tg is obtained in the DSC scan (Qi et al., 2008). The results of the thermal analysis of CBZ and Soluplus1 physical mixtures and melt extrudates are shown in Fig. 8. When the physical mixtures of CBZ and the polymer were analyzed by DSC, endotherms corresponding to the melting temperature of drug were observed for all mixtures. In case of melt extruded samples at 150 C, there was no drug peak present, proving that the drug was distributed in the molecular or amorphous state within the polymer matrices. However, there were more than one Tg values observed at about
Fig. 7. Rheology of different concentrations of carbamazepine (CBZ) with Soluplus1. There was a decrease in viscosity observed with an increase in CBZ concentration at all temperatures. There was no drug melting transition observed at the melting point of drug (191 C).
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
237
Fig. 8. DSC of carbamazepine (CBZ)-Soluplus1 physical mixtures (PM) and hot melt extrudates (HME)
70 C and 130 C, indicating transition temperatures of a nonhomogenous mixture of the two components. The absence of drug peaks in all three CBZ-Soluplus1 binary mixtures extruded at 150 C indicates that solid dispersions without the presence any crystalline drug were formed. The conversion of drug from crystalline to their amorphous states were further confirmed by powder XRD studies (Fig. 9), where it was observed that the characteristic drug peaks were present in physical mixtures containing 10, 20 and 30% CBZ but not in the melt extrudates processed at 150 C. Fig. 9 also gives the powder XRD patterns of CBZ-Soluplus1 mixtures (20% CBZ) processed at 125 and 135 C. The characteristic peaks were absent in those samples as well. 4. Discussion Traditionally, processing temperatures and drug-carrier miscibility during HME were determined using thermal analysis by differential scanning calorimetry (DSC) (Qian et al., 2010). The presence of a single Tg confirmed the formation of a miscible system, where the efficiency of miscibility could be analyzed theoretically with the Gordon–Taylor equation (Liu et al., 2012).
Fig. 9. Powder XRD of carbamazepine (CBZ)-Soluplus1 physical mixtures (PM) and hot melt extrudates (HME)
Although having a single Tg for the drug-polymer mixture is a good predictor for drug-polymer miscibility and potential physical stability upon storage, it does not address whether the mixture can be passed or extruded through the melt extruder and, if so, at what temperature. Chokshi et al. (2005) suggested that rheological analysis for viscoelastic properties of polymers and drug-polymer mixtures could be better predictors for the processing conditions, such as extrusion temperature, motor load, etc., for melt extrudates. It has been shown in the present study that the temperature range for melt extrusion may be correlated with the viscosity of the materials used. The softened Soluplus1 within the viscosity range of 10,000–1000 Pa s could be easily processed through the extruder. This viscosity range corresponded to the processing temperature of 150–180 C. Although the Tg of Soluplus1 was 72 C (Gupta et al., 2014), the temperature had to be more than doubled to 150 C to obtain the optimal processing condition. Most pharmaceutical polymers are viscoelastic in nature, which implies that they have both solid-like and liquid-like traits at various temperatures. Theoretically, molecular motions increase when the temperature exceeds the Tg and the solid goes into the supercooled liquid state. However, physically, the material may still remain very stiff and, therefore, not be extrudable in the hot melt extruder. For example, at 110 C, which is 40 C above the Tg, the viscosity of Soluplus1 was over 100,000 Pa s. This corresponded to >12 Nm torque with the twin screws, and at this torque, 100% capacity of the extruder was reached and the screws stopped rotating (Fig. 2). Above 180 C, the polymer viscosity dropped below 1000 Pa s, which in terms of extrudability could be considered too fluid-like. At such a high temperature, the polymer chains completely disentangled and it behaved like a fluid. From an extrusion point of view, such fluid-like polymer is not acceptable. Thus, the results of the present investigation provide a systematic approach of setting up processing temperatures of drug-polymer mixtures during melt extrusion based on the viscoelastic properties of polymers and their formulations with drugs. Fig. 3 shows that the polymer viscosity of Soluplus1 was also influenced by the shear rate, where a decrease in complex viscosity with an increase in angular frequency was observed. Therefore, the viscosity measured by a rheometer may be applied to set up the processing conditions in a melt extruder, although the shear rate during melt
238
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
extrusion may often be higher depending on the experimental conditions used. If at all, the viscosity within an extruder could be somewhat lower than that measured by a rheometer, which could be favorable to the extrudability of the material. The most significant aspect of the present study is the determination of drug effect on processing conditions of melt extrudates. Usually, in the preparation of solid dispersions by HME, the processing temperature is raised above the melting point of drug to convert crystalline materials into amorphous forms (Li et al., 2014). There are, however, a few relatively recent reports in the literature where amorphous melt extrudates were prepared below melting temperatures of drugs (Qi et al., 2008; Abu-Diak et al., 2012; Raviña-Eirín et al., 2013). Li et al. (2014) attributed this phenomenon to the melting point depression of drugs in the drugpolymer mixtures. They indicated that the interaction between drugs and polymers could also be a factor. No systematic studies were, however, conducted how the optimal processing conditions in melt extruders for such drug-polymer mixtures can be determined. It is observed in the present investigation that when the carbamazepine- Soluplus1 mixtures with 10, 20 and 30% drug loads were extruded at 150 C, which is more than 40 C below the melting temperature of carbamazepine (191 C), clear extrudates with the complete conversion of drug to the amorphous form could be obtained (Figs. 8 and 9). This can be attributed to the miscibility of carbamazepine in the polymer at an elevated but much lower temperature than that of the melting point of the drug substance (Djuris et al., 2013). The complex viscosity of Soluplus1 decreased from over 200,000 Pa s at 110 C to 10,000 and lower at 150 C and higher temperature. Such a relatively lower viscosity helped in the mixing and dissolution of carbamzepine in the polymer leading to its molecular dispersion. Further, it may be observed in Fig. 5 that the presence of drug at a constant processing temperature did not increase the torque of polymer. Rather, there was a gradual and an almost linear decrease in torque observed with an increase in drug concentration. Thus, carbamazepine acted as a plasticizer for Soluplus1 effectively reducing its complex viscosity. As a result, the carbamazepine– Soluplus1 mixtures could be prepared at a temperature even much lower than that of the polymer alone. Li et al. (2014) reported that for a drug to plasticize a polymer, it is necessary that the drug is miscible with the polymer. The viscosity values for all concentrations of drug also decreased gradually with an increase in temperature in an almost parallel pattern (Fig. 7), indicating that the drug was homogenously mixed within the polymer. Because of the parallel nature of the plots with respect to pure polymer, it may be concluded that CBZ was in the dissolved state in the polymer from the start of the analysis. The absence of drug peak at the melting temperature of the drug also proves that the drug was dissolved in the polymer below its melting temperature. It should be mentioned here that the extrudable viscosity range of 10,000 to 1000 Pa s for the polymer and the drug-polymer mixtures in the present investigation was selected by conducting torque analysis using a laboratory scale melt extruder where the highest torque reached was 12 Nm. Since even higher torque values may be attained using larger production-scale melt extruders and more viscous materials may be extruded through such extruders, the question arises as to whether higher torque and a viscosity higher than 10,000 Pa s could be used in melt extrusion. During the preparation of amorphous solid dispersion by melt extrusion, the drug must dissolve in the polymer during the short dwell time within the extruder (usually <5 min). Since the dissolution rate of drug in a liquid medium is directly proportional to the diffusion coefficient of the medium, the drug may not dissolve in or be miscible with a polymer if the viscosity is too high. Therefore, the viscosity of 10,000–1000 Pa s and the corresponding temperatures where such viscosities may be achieved are reasonable guides for
the development of processing conditions for melt extrusion. The process developed by the laboratory scale equipment can thus be easily scaled to larger equipment.
5. Conclusions A solid dispersion formulation of a poorly water-soluble drug and a hydrophilic polymer can be prepared by hot melt extrusion. It is important to investigate the optimal processing conditions that will affect the final formulation. Analysis of the viscoelastic properties of a polymer, Soluplus1, at various temperatures was correlated with the torque analysis by hot melt extrusion. The results helped to determine an ideal extrusion window of Soluplus1 from 10,000 to 1000 Pa s at 150 to 180 C. When the temperature was kept constant at 180 C, there was a decrease in torque value from 7.2 Nm to 2.4 Nm observed in presence of 30% w/ w carbamazepine, illustrating the plasticization effect of the drug on the polymer. A systematic study of the effect of the concentration of carbamazepine, a poorly water-soluble drug, on the viscoelastic properties of Soluplus1 was done. There was a decrease in torque value during hot melt extrusion with an increase in drug concentration. The rheological analysis thus proved that the drug was miscible with the polymer at the processing temperature. Only the effect of drug concentration on viscoelastic properties and extrudability of a polymer has been studied in the present investigation. Once the effect of drug is determined, further studies may be conducted to determine the effects of added plasticizers, surfactants, etc., if any, used in a solid dispersion. This paper should provide guidance to formulators to take a methodical approach towards formulation of solid dispersions by hot melt extrusion by considering thermomechanical and torque analysis as characterization tools.
Acknowledgement This research was supported in part by a grant from Catalent Pharma Solutions, Somerset, NJ 08873, USA.
References Abu-Diak, O.A., Jones, D.S., Andrews, G.P., 2012. Understanding the performance of melt-extruded poly (ethylene oxide)–bicalutamide solid dispersions: characterisation of microstructural properties using thermal, spectroscopic and drug release methods. J. Pharm. Sci. 101, 200–213. Breitenbach, J., 2002. Melt extrusion: from process to drug delivery technology. Eur. J. Pharm. Biopharm. 54, 107–117. Chokshi, R.J., Sandhu, H.K., Iyer, R.M., Shah, N.H., Malick, A.W., Zia, H., 2005. Characterization of physico-mechanical properties of indomethacin and polymers to assess their suitability for hot-melt extrusion processs as a means to manufacture solid dispersion/solution. J. Pharm. Sci. 94, 2463–2474. Crowley, M.M., Zhang, F., Repka, M.A., Thumma, S., Upadhye, S.B., Kumar Battu, S., Martin, C., 2007. Pharmaceutical applications of hot-melt extrusion: part I. Drug Dev. Ind. Pharm. 33, 909–926. Djuris, J., Nikolakakis, I., Ibric, S., Djuric, Z., Kachrimanis, K., 2013. Preparation of carbamazepine–Soluplus1 solid dispersions by hot-melt extrusion, and prediction of drug-polymer miscibility by thermodynamic model fitting. Eur. J. Pharm. Biopharm. 84, 228–237. Ghosh, I., Vippagunta, R., Li, S., Vippagunta, S., 2012. Key considerations for optimization of formulation and melt-extrusion process parameters for developing thermosensitive compound. Pharm. Dev. Technol. 17, 502–510. Gupta, S.S., Meena, A., Parikh, T., Serajuddin, A.T.M., 2014. Investigation of thermal and viscoelastic properties of polymers relevant to hot melt extrusion, I: polyvinylpyrrolidone and related polymers. J. Excipients Food Chem. 5, 32–45. Kolter, K., Grycke, A., 2012. 2nd ed.. Hot-Melt Extrusion with BASF Polymers: Extrusion Compendium. BASF Corp., Germany, pp. 113–116. Lakshman, J.P., Cao, Y., Kowalski, J., Serajuddin, A.T.M., 2008. Application of melt extrusion in the development of a physically and chemically stable high-energy amorphous solid dispersion of a poorly water-soluble drug. Mol. Pharm. 5, 994–1002. Leuner, C., Dressman, J., 2000. Improving drug solubility for oral delivery using solid dispersions. Eur. J. Pharm. Biopharm. 50, 47–60.
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239 Li, S., He, H., Parthiban, L.J., Yin, H., Serajuddin, A., 2005. IV–IVC considerations in the development of immediate-release oral dosage form. J. Pharm. Sci. 94, 1396–1417. Li, Y., Pang, H., Guo, Z., Lin, L., Dong, Y., Li, G., Li, M., Wu, C., 2014. Interactions between drugs and polymers influencing hot melt extrusion. J. Pharm. Pharmacol. 66, 148–166. Liu, H., Zhang, X., Suwardie, H., Wang, P., Gogos, C.G., 2012. Miscibility studies of indomethacin and Eudragit1 E PO by thermal, rheological, and spectroscopic analysis. J. Pharm. Sci. 101, 2204–2212. Meena, A., Parikh, T., Gupta, S.S., Serajuddin, A.T.M., 2014. Investigation of thermal and viscoelastic properties of polymers relevant to hot melt extrusion, II: cellulosic polymers. J. Excipients Food Chem. 5, 46–55. Parikh, T., Gupta, S.S., Meena, A., Serajuddin, A.T.M., 2014. Investigation of thermal and viscoelastic properties of polymers relevant to hot melt extrusion, III: methacrylic acid copolymers and methacrylates. J. Excipients Food Chem. 5, 56–64. Qi, S., Gryczke, A., Belton, P., Craig, D.Q., 2008. Characterisation of solid dispersions of paracetamol and EUDRAGIT1 E prepared by hot-melt extrusion using thermal, microthermal and spectroscopic analysis. Int. J. Pharm. 354, 158–167.
239
Qian, F., Huang, J., Hussain, M.A., 2010. Drug–polymer solubility and miscibility: stability consideration and practical challenges in amorphous solid dispersion development. J. Pharm. Sci. 99, 2941–2947. Raviña-Eirín, E., Gómez-Amoza, J.L., Martínez-Pacheco, R., 2013. Utility of the hyperbranched polymer Hybrane S1200 for production of instant-release particles by hot melt extrusion. Drug Dev. Ind. Pharm. 39, 1107–1112. Serajuddin, A., 1999. Solid dispersion of poorly water-soluble drugs: early promises, subsequent problems, and recent breakthroughs. J. Pharm. Sci. 88, 1058–1066. Shah, S., Maddineni, S., Lu, J., Repka, M.A., 2013. Melt extrusion with poorly soluble drugs. Int. J. Pharm. 453, 233–252. Vasanthavada, M., Tong, W.Q., Serajuddin, A.T.M., 2008. Development of Solid Dispersion for Poorly Water-soluble Drugs, 2nd ed. Informa Healthcare, New York, pp. 149–184. Williams, H.D., Trevaskis, N.L., Charman, S.A., Shanker, R.M., Charman, W.N., Pouton, C.W., Porter, C.J., 2013. Strategies to address low drug solubility in discovery and development. Pharmacol. Rev. 65, 315–499.
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Effect of carbamazepine on viscoelastic properties and hot melt extrudability of Soluplus1 Simerdeep Singh Gupta a , Tapan Parikh a , Anuprabha K. Meena a , Nidhi Mahajan b , Imre Vitez b , Abu T.M. Serajuddin a, * a b
College of Pharmacy and Health Sciences, St. John’s University, 8000 Utopia Parkway, Queens, NY 11439, USA Catalent Pharma Solutions, 14 Schoolhouse Road, Somerset, NJ 08873, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 24 July 2014 Received in revised form 2 October 2014 Accepted 12 November 2014 Available online 13 November 2014
The purpose of this study was to apply viscoelastic properties of polymer and drug-polymer mixtures to determine processing conditions for the preparation of amorphous solid dispersion by melt extrusion. A poorly water-soluble drug, carbamazepine (CBZ), was mixed with Soluplus1 as the carrier. Torque analysis using a melt extruder was performed at 10, 20 and 30% w/w drug concentrations and the effect of barrel temperature was studied. Viscosity of the mixtures either at fixed temperatures with different angular frequencies or as a function of temperature with the same frequency was studied using a rheometer. The viscosity of Soluplus1 and the torque exerted on the twin screws decreased with the increase in CBZ concentration. The viscosity versus temperature plots for different CBZ concentrations were parallel to each other, without the drug melting transition, indicating complete drug-polymer miscibility. Thus, the drug-polymer mixtures could be extruded at temperature as low as 140 C with 10% w/w drug load, 135 C with 20% w/w drug and 125 C with 30% w/w drug, which were, respectively, 50 C, 55 C and 65 C below the melting point of 191 C for CBZ. The differential scanning calorimetry (DSC) and powder X-ray diffraction (XRD) analyses of the binary mixtures extruded at 125–150 C showed absence of crystalline drug. A systematic study of miscibility and extrudability of drug-polymer mixtures by rheological and torque analysis as a function of temperature will help formulators select optimal melt extrusion processing conditions to develop solid dispersions. ã 2014 Elsevier B.V. All rights reserved.
Keywords: Hot melt extrusion Rheology Viscosity Polymer Drug-polymer miscibility Torque analysis Soluplus1
1. Introduction For oral delivery, drug substances must dissolve in the gastrointestinal (GI) fluid before their absorption. Ensuring adequate aqueous solubility in the GI tract has, therefore, emerged as the most difficult biopharmaceutical challenge in the development of oral drug products for optimal clinical outcome (Li et al., 2005). Some of the enabling technologies applied to the development of poorly water-soluble drugs include salt formation, particle size reduction, solubilization, lipid-based drug delivery, solid dispersion, etc. Williams et al. (2013) published an excellent article reviewing these and other strategies to address low solubility of drugs. There are, however, practical challenges and
* Corresponding author at: Department of Pharmaceutical Sciences, College of Pharmacy and Health Sciences, St. John’s University, 8000 Utopia Parkway, Queens, NY 11439, USA. Tel.: +1 718 990 7822; fax: +1 718 990-1877. E-mail address: [email protected] (A.T.M. Serajuddin). http://dx.doi.org/10.1016/j.ijpharm.2014.11.025 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.
limits with each of these technologies. Among various approaches to formulate poorly water-soluble drugs, solid dispersion, where the drug is usually dispersed in a water-soluble amorphous carrier either molecularly or in the amorphous state (Leuner and Dressman, 2000; Vasanthavada et al., 2008), has emerged as the most promising one. After oral administration, the water-soluble matrix of the solid dispersion dissolves in the GI fluid, releasing the drug either in solution or as finely divided precipitate that redissolves rapidly. However, despite extensive research on the solid dispersion technology for over 50 years, only a very limited number of products based on solid dispersion principles have been available in the market. Manufacturing difficulties and stability issues have been the primary reasons for the limited application of the technology for commercialization of drug products (Serajuddin, 1999). One common method of preparing solid dispersion is to dissolve drug and carrier in an organic solvent and then removing the solvent by various evaporation techniques, including spray drying. It is, however, difficult to get a solvent that would simultaneously dissolve the water-insoluble drug and the water-
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
soluble carrier in a reasonably acceptable volume of liquid. Handling of large-scale volumes of solvents during the dosage form development and manufacturing is also a major issue. The prospect for the development of solid dispersion has greatly improved during the past decade due to the introduction of hot melt extrusion (HME) in the pharmaceutical field (Breitenbach, 2002). HME has many advantages over other technologies used for solid dispersion (Crowley et al., 2007; Ghosh et al., 2012; Shah et al., 2013; Lakshman et al., 2008). It is a continuous, less time-consuming process that can be scaled up relatively easily. It is also less prone to batch-to-batch variability than other processes. Also, being solventfree, it poses minimal environmental hazards. However, there are still many challenges in the development of new drug products by melt extrusion. Being a relatively new technology in the pharmaceutical field, the formulation and processing parameters of HME are not fully understood. In most of the published reports, it is not defined how the appropriate polymers, drug-polymer blends and the processing temperatures are selected. A good understanding of the material properties, such as the glass transition temperature (Tg) of polymers, the melt viscosity of polymers as well as drug-polymer blends and the drug-polymer miscibility is needed before the development of processing conditions for any new products. The drug-polymer mixtures should also be extrudable at as low a temperature as possible to minimize potential of degradation of drug, polymer or both. Recently, we applied rheology to study viscoelastic properties of various polymers having PVP (Gupta et al., 2014), cellulosic (Meena et al., 2014) and methacrylate (Parikh et al., 2014) backbones to investigate their suitability for melt extrusion. Only the neat polymers were used in these studies to generate a database relevant to physicochemical properties of different polymers relevant to melt extrusion. It was established that the temperature range where the melt viscosity of a polymer falls between 1000 and 10,000 Pa s was the most suitable temperature for the melt extrusion process. However, a drug formulation intended for melt extrusion is not polymer alone. Polymers are used along with drugs, which may influence the processing conditions. Therefore, the objective of the present investigation was to determine what effect the presence of drug as well as the drug concentration would have on viscoelastic properties and extrudability of a polymer as a function of temperature. Soluplus1, a polyvinyl caprolactam–polyvinyl acetate–polyethylene glycol graft copolymer, was chosen as a carrier for melt extrusion. A poorly water-soluble drug, carbamazepine (CBZ), was selected for the comparative study of its influence on the drug-polymer miscibility and the processing conditions. To determine extrudability of the polymer and the drug-polymer mixtures, the moment of force exerted onto the twin screws of the extruder, also known as torque, was measured. It was hoped that the viscosity and torque analyses would not only help in identifying processing conditions for melt extrusion, they would also provide valuable information on the miscibility of drug-polymer binary mixtures. To confirm miscibility of the mixtures in the products, differential scanning calorimetry (DSC) and powder X-ray diffraction (powder XRD) studies of the extrudates of various drug-polymer mixtures prepared at different temperatures were conducted and the results were compared with that of the physical blends. 2. Materials and methods
233
Fig. 1. Chemical structures of (a) carbamazepine and (b) Soluplus1.
2.2. Methods 2.2.1. Torque analysis by hot melt extrusion (HME) The torque analysis was performed using Process 11 Parallel Twin-Screw Extruder (Thermo Scientific Inc., Waltham, MA, USA) to determine (a) what would be the range of processing temperature in the melt extruder for a particular material, (b) how long would it take for the drug-polymer mixtures to reach a constant torque value, (c) what effect the drug concentration would have on the torque at a particular temperature, and (d) what would be the effect of temperature on torque at a fixed drug concentration. Only the neat polymer was initially used to determine the temperature range for melt extrusion of the polymer itself. The polymer was added into the extruder at the rate of 1 g/min at 100 rpm screw speed. A screw design with low, medium, and high shear geometry of the mixing elements was used. The barrel temperature was initially set at 200 C, and the torque generated in the twin screws was recorded at this temperature and at the decreasing temperature intervals of 5 C till the maximum limit of torque (100%) of 12 Nm in the extruder used reached and the rotation of screws stopped. The temperature that would give a torque value of 60% (7.2 Nm) and less in the extruder was used for further studies. To determine how long would it take to reach a constant torque value at a specific temperature, the torque analysis was conducted with and without the drug. For this purpose, 1 g of the neat polymer (Soluplus1) was added into the extruder at the barrel temperature of 150 C and the screw speed of 100 rpm every minute for three consecutive points and the resultant torque was recorded. Then, the mixture of 30% CBZ in Soluplus1 was added at 1 g/min starting from the fourth addition till a constant torque in presence of drug was obtained. At these conditions, the output through the 1.5 mm circular die at the end of the extrusion barrel was about 1 g/min. The effect of drug concentration on the torque generated was studied by feeding different drug-polymer mixtures (0, 10, 20 and 30% w/w drug) to the melt extruder at the rate of 1 g/min, screw speed of 100 rpm and the barrel temperature of 150 C consecutively for 6 min (time was based on the results obtained from the previous study). The torque value obtained at the steady state for each drug concentration was recorded and the results were plotted against the drug concentration. To study the effect of temperature, a constant drug concentration of 20% w/w was used for extrusion at different temperatures, and the torque values were plotted against it.
2.1. Materials Carbamazepine was purchased from Sigma–Aldrich Co. (St. Louis, MO, USA). Soluplus1 was donated by BASF Corporation (Tarrytown, New York, USA). Structures of the materials are given in Fig. 1.
2.2.2. Rheology The rheology of the samples to analyze viscoelatic properties was studied using Discovery Hybrid Rheometer 2 with oven heating assembly (DHR-2, TA Instruments, Newcastle, DE, USA). The method of the preparation of samples for analysis was
234
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
described earlier (Gupta et al., 2014). Essentially, 1 g of material was weighed and compressed into a slug using a 25 mm ejection die at 5000 pounds of pressure using a Carver press. The slug (2.2 mm thick) was placed in between two flat plates of 25 mm diameter after calibration of the zero gap. For the analysis of pure polymer, frequency sweep was conducted at 110–180 C in increments of 10 C, and the plots of resultant viscosity versus angular frequency were generated. For the analysis of drugpolymer physical mixtures, the slug was equilibrated at 120 C for about 2 min and the temperature flow was ramped up from 120 to 200 C at 5 C/min. Responses were obtained in terms of log viscosity (Pa s) versus temperature ( C). The results obtained at 150 C were correlated with the torque analysis by hot melt extrusion. 2.2.3. Differential scanning calorimetry (DSC) The thermal analysis of physical mixtures and melt-extruded samples was performed using a modulated DSC (DSC Q200, TA Instruments, Newcastle, DE, USA). For physical mixtures, drug and polymer were weighed and mixed to obtain 10, 20 and 30% w/w drug concentration. In case of melt extruded samples, extrudates were cooled to room temperature and milled before analysis. Usually, 5– 10 mg of sample was placed in an aluminum pan that was crimped with a lid on. It was then heated from 25 C to 210 C at the rate of 10 C/min. The scans were analyzed for the presence of Tg and any possible melting peaks after the extrusion process to determine whether the drug had converted completely to the amorphous form after extrusion. To confirm the results of DSC analysis, the extruded samples were also observed visually and under the polarized light microscope. The visual observation showed that the extrudates were clear, and, under the microscope, there was no birefringence due to the presence of any crystalline drug. 2.2.4. Powder X-ray diffraction (XRD) To obtain the powder XRD patterns of different melt extrudates, a Shimadzu 6000 powder X-ray diffractometer (Shimadzu, Tokyo, Japan) was used. Physical mixtures were prepared by separately weighing drugs and polymer and then mixing in 10, 20 and 30% w/ w concentration of drug. Melt extruded samples were triturated into fine powder before analysis. During the XRD analysis, monochromatic CuKa radiation source was operated at 40 kV and 30 mA, with the scanning rate (2 u)/min over the range of 10– 60 (2u). Results were analyzed for the presence of crystalline peaks in the final products. Preliminary studies in our laboratory using the powder X-ray diffractometer showed that the detection limit of crystalline drug for the equipment was 5%, below which the crystalline drug peaks could not be clearly differentiated from the baseline noise. Some of the amorphous samples were also observed under the polarized light microscope, where the lack of birefringence even at <5% confirmed that the materials were amorphous. 3. Results 3.1. Determination of processing temperature The temperature ranges at which the polymer or the drugpolymer mixtures could be extruded through a twin-screw extruder were determined by correlating torque values generated in the extruder with the viscoelastic properties of materials used. For initial correlation, only neat Soluplus1 without the addition of drug was extruded, and the effect of a change in temperature on the moment of force, or torque, generated on the twin screws of the extruder is plotted in Fig. 2. In these experiments, the feed rate, screw speed and screw design were maintained constant. The torque value at 200 C was about 1.5 Nm which was equivalent to
Fig. 2. Effect of melt extrusion processing temperature on the torque subjected on the twin screws by neat Soluplus1 (n = 3). There was a gradual increase in torque observed with a decrease in barrel temperature, indicating that the stiffness in polymer increased with a decrease in temperature.
about 13% of the torque capacity of equipment used. As the temperature was decreased, there was an increase in the torque value, indicating that the stiffness of the polymer increased. At 135 C, the torque value increased to 10.8 Nm (90%), above which the maximum torque limit (12 Nm = 100%) of the equipment was attained. At 150 C, the torque value was about 7 Nm, which was equivalent to 60% of the equipment limit. Above 180 C, the torque was <20% and the extrudate flowing the die at the end of the barrel was too soft to form spaghetti-like strands and, therefore, not suitable for extrusion. Based on these results, the temperature range of 150–180 C was considered optimal for the melt extrusion of Soluplus1. The next step in the determination of processing temperature was to ascertain what would be the viscosity of the polymer in the temperature range identified by torque analysis. Since Soluplus1 is a graft copolymer with viscoelastic properties, it behaves like a solid (elastic) with the storage modulus greater than the loss modulus at temperatures less than and close to its Tg (Gupta et al., 2014). With an increase in temperature, it behaves more like a viscous liquid. A frequency sweep was performed at various angular frequencies to determine the viscosity of the neat polymer at different temperatures, and the results are given in Fig. 3. At all temperatures, there was a decrease in viscosity observed with an increase in angular frequency. The complex viscosity of the polymer at 110 C at all frequencies was above 100,000 Pa s, which implied that the polymer was very stiff. With an increase in temperature, the complex viscosity value decreased. At 150 C, the viscosity dropped below 10,000 Pa s and was about 1000 Pa s at about 180 C. In the temperature range of 150–180 C, the polymer could be considered extrudable as it was free flowing under the applied stress and at the same time it was not so soft that it would reaggregate after extrusion. These results demonstrate that the processing temperature for a certain polymer or drug formulation may be identified by conducting rheology experiments and no elaborate experiments with the melt extruders are necessary. Thus, the viscoelastic analysis serves as a good predictor for the processing temperature during melt extrusion. A similar conclusion was also reached by Kolter and Grycke (2012) in a technical report of BASF Corporation. However, the study was limited to several BASF polymers only. No detailed reports validating the concept for other polymers as well as for drug-polymer mixtures has been reported in peer-reviewed journals. 3.2. Effect of drug on extrudability of polymer After a viscosity window of 10,000–1000 Pa s for the extrusion for Soluplus1 was defined, the effect of drug on the extrudability of
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
235
Fig. 3. Effect of angular frequency and temperature on the complex viscosity of neat Soluplus1. With an increase in the rate of shear subjected on the polymer (given by angular frequency), there was a decrease in complex viscosity observed at all temperatures. With an increase in temperature from 110 to 180 C, there was a drop in viscosity from about 200,000 Pa s till about 1000 Pa s at low frequency values.
the polymer was studied. This was done at 150 C, where the torque value of the neat Soluplus1 was about 7 Nm or 60% of the equipment limit. The results presented in Fig. 4 show that it took time for the added material to completely fill the barrel and give the torque value that represented the sample. It was, therefore, important to determine the number of additions in a given time period to get a constant torque value. The feed rate and the screw speed were kept constant. The study was conducted starting with the neat polymer and then in presence of CBZ. The extrusion was conducted till a constant torque value was obtained. It may be observed in Fig. 4 that the constant torque value with neat Soluplus1 was about 7.2 Nm at 150 C for the first three additions of material at 1 g/min. After the addition of drug-polymer mixture (30% carbamazepine) as the fourth sample, the torque gradually decreased and reached a value of 2.5 Nm when the eighth sample of the mixture was added. The torque value remained at about 2.4 Nm for next 4 additions. Between the fourth and the seventh additions of materials at 1 g/min, the extrusion barrel was filled partly with the neat polymer and partly with the drug-polymer mixture, and only when the entire neat polymer extruded out of the barrel and the barrel was completely filled with the drug-
The analysis of torque during melt extrusion was also conducted using different concentrations of carbamazepine at a constant temperature of 150 C. The analysis was performed first at 0% drug concentration (neat Soluplus1) to obtain a constant torque value, and the study was then continued by using increasing concentrations of CBZ in binary mixtures (10, 20, and 30% w/w). As shown in Fig. 5, there was a gradual decrease in torque values with the addition of drug and the increase in drug concentration. The constant torque recorded without the drug was about 7.2 Nm. After the addition of 30% drug-polymer mixtures, the torque decreased to 2.8 Nm. The decrease in torque with an increase in drug concentration was linear with R2 value of 0.99. These results demonstrate that the presence of CBZ can greatly influence the processing conditions of drug-
Fig. 4. Torque analysis of Soluplus1 in presence of 30% CBZ at 150 C. At initial sample additions without the drug, the torque was about 7.2 Nm. After addition of sample with 30% drug, there was a gradual decrease in torque observed till a constant value was attained at about 2.4 Nm. The constant torque sample number was used in the further analyses.
Fig. 5. Effect of drug concentration on the extrudability of drug-polymer mixtures at 150 C. There was a decrease in constant torque value after addition of 10% w/w drug-polymer mixture as compared to neat polymer. There was a further decrease in constant torque with an increase in drug concentration.
polymer mixture that the torque stabilized. The results presented in Fig. 4 demonstrate that the presence of CBZ has major impacts on the extrudability of Soluplus1 as it reduces the torque within the extruder barrel. 3.3. Effect of drug concentration on extrudability of drug-polymer mixtures
236
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
3.5. Effect of drug and temperature on viscoelastic properties of drugpolymer mixtures
Fig. 6. Effect of change of temperature on the extrudability of 20% w/w drugpolymer mixture. There was an increase in constant torque value with a decrease in barrel temperature. Below 125 C, the maximum torque exerted on the equipment was attained.
polymer mixtures by apparently acting as a plasticizer. The plasticizing effect may, however, be different for different drugs and certain drugs may not have any plasticizing effect at all. Further work is continuing in our laboratory to determine the effect of drug properties on plasticizing effects. 3.4. Effect of temperature on extrudability of drug-polymer mixtures Once the effect of drug concentration on the extrudability at a constant temperature of 150 C was determined, it was important to understand the effect of a change of temperature on the torque subjected on the screws. The result is given in Fig. 6. Drug-polymer binary mixture with 20% w/w drug concentration was used. The extrusion was first conducted at 150 C, which gave constant torque value of about 4 Nm, as shown for 20% drug load in Fig. 5. In subsequent experiments, lower barrel temperatures were used. There was a gradual increase in torque to 5.6 Nm with the decrease in temperature till 135 C. There was a sharp increase in torque to 9.9 Nm at 125 C, below which the extruder stopped. These results demonstrate that much lower processing temperature could be used for the melt extrusion of CBZ-Soluplus1 mixtures than that of the polymer itself.
The rheological analysis was also conducted to determine the effect of drug on the polymer behavior at high temperature. Fig. 7 gives the complex viscosity profile of CBZ-Soluplus1 mixtures as a function of drug concentration as well as temperature. For the neat Soluplus1, the viscosity was over 100,000 Pa s at the starting temperature of 120 C, and gradually decreased with the increase in temperature (Fig. 7). The viscosity was slightly under 10,000 Pa s at 150 C, and this temperature was, therefore, used for torque analysis mentioned earlier. The viscosity dropped below 1000 Pa s when the temperature increased over 180 C. With the addition of CBZ at increasing concentrations of 10, 20 and 30%, decreases in viscosity were observed at all temperatures, and the plots appeared to be approximately parallel to each other as well as that with that of the neat Soluplus1. If the viscosity of 10,000 Pa s is considered to be acceptable for melt extrusion, Fig. 7 shows that that the processing temperature could be decreased from 150 C for the neat Soluplus1 to 130 C in presence of 30% CBZ. At the extrusion temperature of 150 C, all the mixtures were in the extrudable viscosity range of 1000 10,000 Pa s. It may also be observed that there was no peak or bump for all three mixtures that may be attributed to the melting point of the drug substance (191 C). This indicates that carbamazepine was miscible with Soluplus1 at temperature below its melting and no crystalline material remained at the melting temperature. 3.6. Thermal and powder XRD analyses of melt extrudates The formation of solid dispersion was confirmed by the DSC analysis. Theoretically, when a drug is molecularly dispersed, only one Tg is obtained in the DSC scan (Qi et al., 2008). The results of the thermal analysis of CBZ and Soluplus1 physical mixtures and melt extrudates are shown in Fig. 8. When the physical mixtures of CBZ and the polymer were analyzed by DSC, endotherms corresponding to the melting temperature of drug were observed for all mixtures. In case of melt extruded samples at 150 C, there was no drug peak present, proving that the drug was distributed in the molecular or amorphous state within the polymer matrices. However, there were more than one Tg values observed at about
Fig. 7. Rheology of different concentrations of carbamazepine (CBZ) with Soluplus1. There was a decrease in viscosity observed with an increase in CBZ concentration at all temperatures. There was no drug melting transition observed at the melting point of drug (191 C).
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
237
Fig. 8. DSC of carbamazepine (CBZ)-Soluplus1 physical mixtures (PM) and hot melt extrudates (HME)
70 C and 130 C, indicating transition temperatures of a nonhomogenous mixture of the two components. The absence of drug peaks in all three CBZ-Soluplus1 binary mixtures extruded at 150 C indicates that solid dispersions without the presence any crystalline drug were formed. The conversion of drug from crystalline to their amorphous states were further confirmed by powder XRD studies (Fig. 9), where it was observed that the characteristic drug peaks were present in physical mixtures containing 10, 20 and 30% CBZ but not in the melt extrudates processed at 150 C. Fig. 9 also gives the powder XRD patterns of CBZ-Soluplus1 mixtures (20% CBZ) processed at 125 and 135 C. The characteristic peaks were absent in those samples as well. 4. Discussion Traditionally, processing temperatures and drug-carrier miscibility during HME were determined using thermal analysis by differential scanning calorimetry (DSC) (Qian et al., 2010). The presence of a single Tg confirmed the formation of a miscible system, where the efficiency of miscibility could be analyzed theoretically with the Gordon–Taylor equation (Liu et al., 2012).
Fig. 9. Powder XRD of carbamazepine (CBZ)-Soluplus1 physical mixtures (PM) and hot melt extrudates (HME)
Although having a single Tg for the drug-polymer mixture is a good predictor for drug-polymer miscibility and potential physical stability upon storage, it does not address whether the mixture can be passed or extruded through the melt extruder and, if so, at what temperature. Chokshi et al. (2005) suggested that rheological analysis for viscoelastic properties of polymers and drug-polymer mixtures could be better predictors for the processing conditions, such as extrusion temperature, motor load, etc., for melt extrudates. It has been shown in the present study that the temperature range for melt extrusion may be correlated with the viscosity of the materials used. The softened Soluplus1 within the viscosity range of 10,000–1000 Pa s could be easily processed through the extruder. This viscosity range corresponded to the processing temperature of 150–180 C. Although the Tg of Soluplus1 was 72 C (Gupta et al., 2014), the temperature had to be more than doubled to 150 C to obtain the optimal processing condition. Most pharmaceutical polymers are viscoelastic in nature, which implies that they have both solid-like and liquid-like traits at various temperatures. Theoretically, molecular motions increase when the temperature exceeds the Tg and the solid goes into the supercooled liquid state. However, physically, the material may still remain very stiff and, therefore, not be extrudable in the hot melt extruder. For example, at 110 C, which is 40 C above the Tg, the viscosity of Soluplus1 was over 100,000 Pa s. This corresponded to >12 Nm torque with the twin screws, and at this torque, 100% capacity of the extruder was reached and the screws stopped rotating (Fig. 2). Above 180 C, the polymer viscosity dropped below 1000 Pa s, which in terms of extrudability could be considered too fluid-like. At such a high temperature, the polymer chains completely disentangled and it behaved like a fluid. From an extrusion point of view, such fluid-like polymer is not acceptable. Thus, the results of the present investigation provide a systematic approach of setting up processing temperatures of drug-polymer mixtures during melt extrusion based on the viscoelastic properties of polymers and their formulations with drugs. Fig. 3 shows that the polymer viscosity of Soluplus1 was also influenced by the shear rate, where a decrease in complex viscosity with an increase in angular frequency was observed. Therefore, the viscosity measured by a rheometer may be applied to set up the processing conditions in a melt extruder, although the shear rate during melt
238
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239
extrusion may often be higher depending on the experimental conditions used. If at all, the viscosity within an extruder could be somewhat lower than that measured by a rheometer, which could be favorable to the extrudability of the material. The most significant aspect of the present study is the determination of drug effect on processing conditions of melt extrudates. Usually, in the preparation of solid dispersions by HME, the processing temperature is raised above the melting point of drug to convert crystalline materials into amorphous forms (Li et al., 2014). There are, however, a few relatively recent reports in the literature where amorphous melt extrudates were prepared below melting temperatures of drugs (Qi et al., 2008; Abu-Diak et al., 2012; Raviña-Eirín et al., 2013). Li et al. (2014) attributed this phenomenon to the melting point depression of drugs in the drugpolymer mixtures. They indicated that the interaction between drugs and polymers could also be a factor. No systematic studies were, however, conducted how the optimal processing conditions in melt extruders for such drug-polymer mixtures can be determined. It is observed in the present investigation that when the carbamazepine- Soluplus1 mixtures with 10, 20 and 30% drug loads were extruded at 150 C, which is more than 40 C below the melting temperature of carbamazepine (191 C), clear extrudates with the complete conversion of drug to the amorphous form could be obtained (Figs. 8 and 9). This can be attributed to the miscibility of carbamazepine in the polymer at an elevated but much lower temperature than that of the melting point of the drug substance (Djuris et al., 2013). The complex viscosity of Soluplus1 decreased from over 200,000 Pa s at 110 C to 10,000 and lower at 150 C and higher temperature. Such a relatively lower viscosity helped in the mixing and dissolution of carbamzepine in the polymer leading to its molecular dispersion. Further, it may be observed in Fig. 5 that the presence of drug at a constant processing temperature did not increase the torque of polymer. Rather, there was a gradual and an almost linear decrease in torque observed with an increase in drug concentration. Thus, carbamazepine acted as a plasticizer for Soluplus1 effectively reducing its complex viscosity. As a result, the carbamazepine– Soluplus1 mixtures could be prepared at a temperature even much lower than that of the polymer alone. Li et al. (2014) reported that for a drug to plasticize a polymer, it is necessary that the drug is miscible with the polymer. The viscosity values for all concentrations of drug also decreased gradually with an increase in temperature in an almost parallel pattern (Fig. 7), indicating that the drug was homogenously mixed within the polymer. Because of the parallel nature of the plots with respect to pure polymer, it may be concluded that CBZ was in the dissolved state in the polymer from the start of the analysis. The absence of drug peak at the melting temperature of the drug also proves that the drug was dissolved in the polymer below its melting temperature. It should be mentioned here that the extrudable viscosity range of 10,000 to 1000 Pa s for the polymer and the drug-polymer mixtures in the present investigation was selected by conducting torque analysis using a laboratory scale melt extruder where the highest torque reached was 12 Nm. Since even higher torque values may be attained using larger production-scale melt extruders and more viscous materials may be extruded through such extruders, the question arises as to whether higher torque and a viscosity higher than 10,000 Pa s could be used in melt extrusion. During the preparation of amorphous solid dispersion by melt extrusion, the drug must dissolve in the polymer during the short dwell time within the extruder (usually <5 min). Since the dissolution rate of drug in a liquid medium is directly proportional to the diffusion coefficient of the medium, the drug may not dissolve in or be miscible with a polymer if the viscosity is too high. Therefore, the viscosity of 10,000–1000 Pa s and the corresponding temperatures where such viscosities may be achieved are reasonable guides for
the development of processing conditions for melt extrusion. The process developed by the laboratory scale equipment can thus be easily scaled to larger equipment.
5. Conclusions A solid dispersion formulation of a poorly water-soluble drug and a hydrophilic polymer can be prepared by hot melt extrusion. It is important to investigate the optimal processing conditions that will affect the final formulation. Analysis of the viscoelastic properties of a polymer, Soluplus1, at various temperatures was correlated with the torque analysis by hot melt extrusion. The results helped to determine an ideal extrusion window of Soluplus1 from 10,000 to 1000 Pa s at 150 to 180 C. When the temperature was kept constant at 180 C, there was a decrease in torque value from 7.2 Nm to 2.4 Nm observed in presence of 30% w/ w carbamazepine, illustrating the plasticization effect of the drug on the polymer. A systematic study of the effect of the concentration of carbamazepine, a poorly water-soluble drug, on the viscoelastic properties of Soluplus1 was done. There was a decrease in torque value during hot melt extrusion with an increase in drug concentration. The rheological analysis thus proved that the drug was miscible with the polymer at the processing temperature. Only the effect of drug concentration on viscoelastic properties and extrudability of a polymer has been studied in the present investigation. Once the effect of drug is determined, further studies may be conducted to determine the effects of added plasticizers, surfactants, etc., if any, used in a solid dispersion. This paper should provide guidance to formulators to take a methodical approach towards formulation of solid dispersions by hot melt extrusion by considering thermomechanical and torque analysis as characterization tools.
Acknowledgement This research was supported in part by a grant from Catalent Pharma Solutions, Somerset, NJ 08873, USA.
References Abu-Diak, O.A., Jones, D.S., Andrews, G.P., 2012. Understanding the performance of melt-extruded poly (ethylene oxide)–bicalutamide solid dispersions: characterisation of microstructural properties using thermal, spectroscopic and drug release methods. J. Pharm. Sci. 101, 200–213. Breitenbach, J., 2002. Melt extrusion: from process to drug delivery technology. Eur. J. Pharm. Biopharm. 54, 107–117. Chokshi, R.J., Sandhu, H.K., Iyer, R.M., Shah, N.H., Malick, A.W., Zia, H., 2005. Characterization of physico-mechanical properties of indomethacin and polymers to assess their suitability for hot-melt extrusion processs as a means to manufacture solid dispersion/solution. J. Pharm. Sci. 94, 2463–2474. Crowley, M.M., Zhang, F., Repka, M.A., Thumma, S., Upadhye, S.B., Kumar Battu, S., Martin, C., 2007. Pharmaceutical applications of hot-melt extrusion: part I. Drug Dev. Ind. Pharm. 33, 909–926. Djuris, J., Nikolakakis, I., Ibric, S., Djuric, Z., Kachrimanis, K., 2013. Preparation of carbamazepine–Soluplus1 solid dispersions by hot-melt extrusion, and prediction of drug-polymer miscibility by thermodynamic model fitting. Eur. J. Pharm. Biopharm. 84, 228–237. Ghosh, I., Vippagunta, R., Li, S., Vippagunta, S., 2012. Key considerations for optimization of formulation and melt-extrusion process parameters for developing thermosensitive compound. Pharm. Dev. Technol. 17, 502–510. Gupta, S.S., Meena, A., Parikh, T., Serajuddin, A.T.M., 2014. Investigation of thermal and viscoelastic properties of polymers relevant to hot melt extrusion, I: polyvinylpyrrolidone and related polymers. J. Excipients Food Chem. 5, 32–45. Kolter, K., Grycke, A., 2012. 2nd ed.. Hot-Melt Extrusion with BASF Polymers: Extrusion Compendium. BASF Corp., Germany, pp. 113–116. Lakshman, J.P., Cao, Y., Kowalski, J., Serajuddin, A.T.M., 2008. Application of melt extrusion in the development of a physically and chemically stable high-energy amorphous solid dispersion of a poorly water-soluble drug. Mol. Pharm. 5, 994–1002. Leuner, C., Dressman, J., 2000. Improving drug solubility for oral delivery using solid dispersions. Eur. J. Pharm. Biopharm. 50, 47–60.
S.S. Gupta et al. / International Journal of Pharmaceutics 478 (2015) 232–239 Li, S., He, H., Parthiban, L.J., Yin, H., Serajuddin, A., 2005. IV–IVC considerations in the development of immediate-release oral dosage form. J. Pharm. Sci. 94, 1396–1417. Li, Y., Pang, H., Guo, Z., Lin, L., Dong, Y., Li, G., Li, M., Wu, C., 2014. Interactions between drugs and polymers influencing hot melt extrusion. J. Pharm. Pharmacol. 66, 148–166. Liu, H., Zhang, X., Suwardie, H., Wang, P., Gogos, C.G., 2012. Miscibility studies of indomethacin and Eudragit1 E PO by thermal, rheological, and spectroscopic analysis. J. Pharm. Sci. 101, 2204–2212. Meena, A., Parikh, T., Gupta, S.S., Serajuddin, A.T.M., 2014. Investigation of thermal and viscoelastic properties of polymers relevant to hot melt extrusion, II: cellulosic polymers. J. Excipients Food Chem. 5, 46–55. Parikh, T., Gupta, S.S., Meena, A., Serajuddin, A.T.M., 2014. Investigation of thermal and viscoelastic properties of polymers relevant to hot melt extrusion, III: methacrylic acid copolymers and methacrylates. J. Excipients Food Chem. 5, 56–64. Qi, S., Gryczke, A., Belton, P., Craig, D.Q., 2008. Characterisation of solid dispersions of paracetamol and EUDRAGIT1 E prepared by hot-melt extrusion using thermal, microthermal and spectroscopic analysis. Int. J. Pharm. 354, 158–167.
239
Qian, F., Huang, J., Hussain, M.A., 2010. Drug–polymer solubility and miscibility: stability consideration and practical challenges in amorphous solid dispersion development. J. Pharm. Sci. 99, 2941–2947. Raviña-Eirín, E., Gómez-Amoza, J.L., Martínez-Pacheco, R., 2013. Utility of the hyperbranched polymer Hybrane S1200 for production of instant-release particles by hot melt extrusion. Drug Dev. Ind. Pharm. 39, 1107–1112. Serajuddin, A., 1999. Solid dispersion of poorly water-soluble drugs: early promises, subsequent problems, and recent breakthroughs. J. Pharm. Sci. 88, 1058–1066. Shah, S., Maddineni, S., Lu, J., Repka, M.A., 2013. Melt extrusion with poorly soluble drugs. Int. J. Pharm. 453, 233–252. Vasanthavada, M., Tong, W.Q., Serajuddin, A.T.M., 2008. Development of Solid Dispersion for Poorly Water-soluble Drugs, 2nd ed. Informa Healthcare, New York, pp. 149–184. Williams, H.D., Trevaskis, N.L., Charman, S.A., Shanker, R.M., Charman, W.N., Pouton, C.W., Porter, C.J., 2013. Strategies to address low drug solubility in discovery and development. Pharmacol. Rev. 65, 315–499.