RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Enhancements and Limits in Drug Membrane Transport Using Supersaturated Solutions of Poorly Water Soluble Drugs SHWETA A. RAINA,1 GEOFF G. Z. ZHANG,2 DAVID E. ALONZO,2 JIANWEI WU,2 DONGHUA ZHU,2 NATHANIEL D. CATRON,2 YI GAO,2 LYNNE S. TAYLOR1 1 2
Department of Industrial and Physical Pharmacy, College of Pharmacy, Purdue University, West Lafayette, Indiana Drug Product Development, Research and Development, AbbVie Inc., North Chicago, Illinois
Received 6 September 2013; revised 14 November 2013; accepted 25 November 2013 Published online 30 December 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23826 ABSTRACT: Amorphous solid dispersions (ASDs) give rise to supersaturated solutions (solution concentration greater than equilibrium crystalline solubility). We have recently found that supersaturating dosage forms can exhibit the phenomenon of liquid–liquid phase separation (LLPS). Thus, the high supersaturation generated by dissolving ASDs can lead to a two-phase system wherein one phase is an initially nanodimensioned and drug-rich phase and the other is a drug-lean continuous aqueous phase. Herein, the membrane transport of supersaturated solutions, at concentrations above and below the LLPS concentration has been evaluated using a side-by-side diffusion cell. Measurements of solution concentration with time in the receiver cell yield the flux, which reflects the solute thermodynamic activity in the donor cell. As the nominal concentration of solute in the donor cell increases, a linear increase in flux was observed up to the concentration where LLPS occurred. Thereafter, the flux remained essentially constant. Both nifedipine and felodipine solutions exhibit such behavior as long as crystallization is absent. This suggests that there is an upper limit in passive membrane transport that is dictated by the LLPS concentration. These results have several important implications for drug delivery, especially for poorly soluble C 2013 Wiley Periodicals, Inc. and the American Pharmacists Association J compounds requiring enabling formulation technologies. Pharm Sci 103:2736–2748, 2014 Keywords: absorption; amorphous; bioavailability; crystallization; diffusion; dispersion; liquid-liquid phase separation; passive; uptake
INTRODUCTION Low aqueous solubility of emerging drug candidates is a major hurdle in drug development. Many of the small molecules showing therapeutic activity against biological targets of interest have very low aqueous solubility either because they are lipophilic or have a high melting point, or because of the combined effect of these two factors. Therefore, there is a great deal of interest in formulating such compounds into dosage forms that can supersaturate.1–3 It has been noted that these lipophilic small molecules, when present at elevated supersaturations in aqueous media, can form colloidal drug aggregates that can subsequently coalesce leading to an increase in size, and/or crystallize.4–10 Colloidal aggregates of small molecules are routinely observed in the high-throughput enzymatic screens employed during drug discovery and are formed when concentrated aqueous solutions of drugs are generated by the solvent shifting method.7 In this method, a small aliquot of a concentrated solution of the drug, dissolved in an organic solution, is added to an aqueous solution to generate the desired concentration, often resulting in a highly supersaturated solution. A supersaturated solution is one where the concentration (or more Correspondence to: Yi Gao (Telephone: +847-937-9028; E-mail: [email protected]); Lynne S. Taylor (Telephone: +765-496-6614; Fax: +765-4946545; E-mail: [email protected]) David E. Alonzo’s present address is Formulation and Process Development, Gilead Sciences, Inc., Foster City, California. Donghua Zhu’s present address is Janssen R&D of Johnson and Johnson, China. Journal of Pharmaceutical Sciences, Vol. 103, 2736–2748 (2014)
C 2013 Wiley Periodicals, Inc. and the American Pharmacists Association
2736
rigorously, the chemical potential of the solute) exceeds that of a solution saturated with respect to the crystalline solid. These aggregates have attracted attention because they frequently lead to false positives in enzyme-based assays via nonspecific protein inhibition. Aggregation may also lead to false negatives as molecules in an aggregated state may be unable to interact with proteins. Compounds undergoing this phenomenon have been termed “promiscuous aggregators.” Studies by Doak et al.,11 Frenkel et al.,4 along with several other research groups have demonstrated that an array of active pharmaceutical ingredients (APIs) form nanosized aggregates. The formation of colloidal aggregates is concentration dependent and occurs only above a critical concentration termed the critical aggregate concentration. The formation of colloidal species has also been reported to occur following dissolution of amorphous solid dispersions (ASDs). One of the earliest studies reported the formation of colloidal species after dissolving a dispersion of polyvinyl pyrrolidone (PVP) and $-carotene.12 ASDs prepared with hydroxypropylmethyl cellulose acetate succinate (HPMCAS) are also reported to generate colloidal species.13 Other systems where colloid formation has been noted include felodipine and hydroxypropylmethlyl cellulose (HPMC),14 mangostin and PVP,15 and lopinavir/ritonavir formulations with copovidone.16 Recent studies have demonstrated that the formation of colloidal species upon dissolution of an amorphous dispersion only occurs when a critical concentration of dissolved drug has been achieved.17 The formation of these colloidal species has been postulated to be beneficial for oral drug delivery.13 It has been suggested that the underlying phenomenon causing the formation of colloidal aggregates, either from the solvent
Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
switch method, or from dissolution of an ASD, is liquid–liquid phase separation (LLPS).7,17 LLPS can also occur from other supersaturation generating systems for example during lipolysis of a self-microemulsifying drug delivery system18 or as a result of pH changes.19 Highly supersaturated solutions will undergo LLPS when a certain threshold concentration is exceeded whereby a homogeneous one-phase system separates into two liquid phases. LLPS is also referred to as oiling-out or liquid–liquid demixing.20–27 LLPS has been widely observed in a variety of systems including polymer blends, proteins, metals, and is often prevalent, albeit undesired, during indus¨ et al.29 reported LLPS for the trial crystallization.24,27,28 Svard water–vanillin system. Kiesow et al.30 reported LLPS during the crystallization of 4-4 -dihydroxydiphenylsulfone and used modeling to predict its occurrence. In the context of aqueous solutions of hydrophobic drugs with melting points above the experimental temperature, LLPS is the separation of a supersaturated drug solution into two liquid phases, one of which is drug rich and hydrophobic, that is, colloidal aggregates that predominantly consist of the drug, whereas the other phase is water rich and contains only a low concentration of drug.17 The two phases are in dynamic equilibrium, although both phases are thermodynamically metastable and the system is supersaturated. In formulated products, additives such as polymers, surfactants, complexing agents, and lipids are routinely used to improve solubility and dissolution profiles in vitro in the hope that this will enhance systemic drug concentrations in vivo. Often, extremely high-concentration enhancements are reported based on approaches that involve assaying the solution phase. This approach of measuring solution concentration can be problematic when attempting to predict in vivo performance because it does not discriminate between the two mechanisms of achieving enhanced solution concentrations: supersaturation versus solubility enhancement. True supersaturation occurs when there is an increase in the chemical potential of the solute relative to a saturated solution in which the solute chemical potential is the same as that of the crystalline solid. For example, polymers at low concentrations do not typically enhance equilibrium solubility,31 but enable supersaturated solutions to be generated and maintained after dissolving an amorphous formulation by inhibiting crystallization from supersaturated solutions. In contrast, micellar surfactant, pH adjustment, cyclodextrins, and cosolvents increase the equilibrium solubility of the crystalline solid (although supersaturation may be generated when the resultant system is diluted). As the crystalline solid is in equilibrium with the solution, it is clear that an increase in solution concentration can be achieved without an increase in the chemical potential of the solute. Because membrane transport is driven by the chemical potential gradient,32 it is important to consider the thermodynamic properties of the solution rather than the total solution concentration. Thus, it is well established that membrane transport can be decreased by the presence of micellar surfactant,33 cyclodextrins,34 but is increased by supersaturated solutions as long as crystallization is prevented.33,35–41 Although most of the studies on the relationship between membrane transport and supersaturation have been in the context of transdermal delivery, there is an increasing interest in exploiting supersaturated solutions to enhance oral absorption. Recent studies have demonstrated enhanced flux across intestinal membranes when perfused with supersaturated solutions.40–45 However, the impact of the colDOI 10.1002/jps.23826
2737
loidal species generated in highly supersaturated solutions has not been fully evaluated to date. Diffusion data from Alonzo et al.46 suggested that the flux reached a maximum when concentrations equal to or above the “amorphous solubility” were generated. Thus, at very high concentrations in the supersaturated regime, it appears that there may be a breakdown in the relationship between concentration and diffusion rate. The goal of the current study was to demonstrate that small organic molecules dissolved in aqueous solution at concentrations at and above their amorphous solubility undergo LLPS, and that the LLPS concentration corresponds to the maximum achievable diffusive flux for a supersaturated solution. Felodipine and nifedipine were selected as model drugs to evaluate the impact of supersaturation and LLPS on membrane transport. The approach of Corrigan et al.35 was utilized, whereby crystallization inhibitors were added to the donor solution to sustain supersaturation for sufficient time to enable diffusion measurements to be made; the polymer, hydroxypropyl methyl cellulose was used as the crystallization inhibitor. Phase diagrams were constructed to understand the concentrations at which liquid– liquid and liquid–solid phase transformations occurred.
MATERIALS Felodipine and nifedipine were purchased from Attix Pharmaceuticals (Toronto, Ontario, Canada) and Euroasia (Mumbai, India), respectively. HPMC Pharmacoat grade 606 was obtained from ShinEtsu (Shin-Etsu Chemical Company, Ltd., Tokyo, Japan). Dissolution media used in all experiments comprised 50 mM pH 6.8 phosphate buffer (ionic strength = 0.155 M) without or with predissolved polymer at a concentration of 100 :g/mL and 1 mg/mL for felodipine and nifedipine, respectively. Methyl alcohol was purchased from Pharmco Products, Inc., Brookfield, Connecticut. Molecular structures of the model compounds are shown in Figure 1. Regenerated cellulose membrane with a molecular weight cutoff (MWCO) of 6–8 K was obtained from Spectrum Laboratories, Inc. (Rancho Dominguez, California).
METHODS Crystalline Solubility Measurements Equilibrium crystalline solubility was determined using a modification of the shake flask method. An excess of crystalline felodipine and nifedipine was equilibrated in 20 mL scintillation vials with 50 mM pH 6.8 phosphate buffer in an agitating water bath (Dubnoff metallic shaking incubator; PGC Scientific, Palm Desert, California). Vials were wrapped in aluminum foil to protect the samples from light. Preliminary experiments indicated that equilibrium was reached by 48 h. Samples were agitated for 48 h at 20◦ C, 25◦ C, 30◦ C, 37◦ C, and 45◦ C before subjecting them to ultracentrifugation to separate excess solid from the supernatant (which is saturated with drug). An Optima L100 XP ultracentrifuge equipped with Swinging-Bucket Rotor SW 41 Ti (Beckman Coulter, Inc., Brea, California) was used and samples were centrifuged at 274,356g for 15 min. The supernatant obtained was diluted with acetonitrile: 50 mM phosphate buffer (50:50) in a 1:1 ratio and 200 :L samples were injected into an Agilent 1100 high-performance liquid chromatography system (Agilent Technologies, Santa Clara, California). The chromatographic separation was performed with Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
2738
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Figure 1. Molecular structures of model compounds.
a Zobrax Eclipse Plus C18 analytical column (2.1 × 150 mm2) ˚ (Agilent Technologies) using acetonitrile: 50 i.d., 5 :m, 100 A) mM phosphate buffer (70:30) as mobile phase. Flow through the column was maintained at 1 mL/min and the column was heated to 40◦ C. The pressure across the column was 28 bar and detection was carried out using ultraviolet (UV) absorbance at a wavelength of 360 nm. The total run time was 7 min and felodipine and nifedipine peaks appeared at 4.2 and 2.5 min, respectively. A standard curve was prepared using samples dissolved in methanol with an injection volume of 20 :L covering a concentration range of 0.1–50 :g/mL. Samples were analyzed in triplicate. The standard curve exhibited good linearity (r 2 > 0.998) over the specified concentration range. The equilibrium solubility of felodipine and nifedipine was also determined in the presence of 100 :g/mL and 1 mg/mL of HPMC in 50 mM pH 6.8 phosphate buffer (ionic strength = 0.155 M) using the method outlined above. Amorphous Solubility Estimations Amorphous solubility can be estimated by initially calculating the free energy difference between the crystal and supercooled liquid, for example, by using the Hoffman equation. This first estimate is then corrected for moisture sorption and ionization. Equation (1) shows the factors contributing to the solubility of a crystalline material: ln(x)solute = −
Hf R
−
Cp R
1 1 − T Tm 1−
Tm Tm + ln T T
− lnγ
(1)
where ln(x)solute is the natural log of the mole fraction solubility, Hf is the molar heat of fusion of the crystal, R is the universal gas constant, T is the temperature of interest, Tm is the melting temperature, and Cp is the heat capacity difference between the crystal and the liquid. The activity coefficient, γ accounts for nonideal solute–solvent interactions.47 Amorphous materials lack the long range order characteristic of their crystalline counterparts and hence Eq. (1) reduces to Eq. (2). ln(x)solute = − ln γ
(2)
After a series of approximations17,48 and derivations, the ratio of amorphous to crystalline solubility Fa /Fc can be written as Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
Eq. (3), using the Hoffman equation to estimate the free energy difference between the amorphous and crystalline material48 :
σa G = e RT = e σc
Hf T T Tm Tm RT
(
) (3)
Although the Hoffman equation provides a reasonable first approximation of the free energy difference between the amorphous and crystalline material,49 neither the impact of the degree of ionization of the molecules or the reduced activity of the amorphous solid resulting from water sorption are taken into account in Eq. (3). An improved estimate of amorphous solubility may be attained by using the novel approach developed by Murdande et al.50 a −μx σa (1 − α x ) [−I(a2 )] (μ2RT 2) e = e a σc (1 − α )
(4)
here R is the gas constant and T is temperature. α a represents fraction ionized for an amorphous solute, whereas α x represents the fraction ionized for the crystalline solute. The first term on the right hand side in the above equation represents differences in the ionization states in solutions equilibrated with amorphous versus crystalline forms. This ionization term is equal to 1 if the API is unionizable or is present in a buffered solution where ionization is minimal. Felodipine (pKa < 2)51,52 and nifedipine (pKa < 3.5)53 are unionized at pH 6.8 and thus this term reduces to 1 (for ionizable molecules or solute present at pH where it is ionized, this term is F
Between groups Within groups Total Between groups Within groups Total
9.73E-8 5.40E-7 6.37E-7 1.27E-7 5.52E-6 5.65E-6
2 12 14 2 9 11
4.87E-8 4.50E-8
1.08
0.37
6.33E-8 6.13E-7
0.103
0.90
Between groups Within groups Total
2.73E-5 2.73E-4 2.99E-4
2 6 8
1.36E-5 4.55E-5
0.30
0.75
Null hypothesis: means of all levels are equal. Alternate hypothesis: the means of one or more level are different. At the 0.05 level, the population means are not significantly different for felodipine at 25◦ C and 37◦ C and nifedipine at 37◦ C. a Data shown compare means from 5, 10, 25, 50, and 100 :g/mL groups. b Data shown compare means from 10, 25, 50, and 100 :g/mL groups. c Data shown compare means from 72, 108, and 145 :g/mL groups.
important to note that other variables such as temperature, viscosity, and timescale of experiments also impact metastable zone width. The spinodal represents the supersaturation where the solution is unstable and there is no free energy barrier for the nucleation of a new phase. It has been observed that the distance between the spinodal and the binodes is small for organic molecules.17 For relatively slow crystallizers like felodipine,54 where the nucleation rate is slow relative to the rate of supersaturation generation, it is clear that supersaturations that exceed the amorphous solubility can be achieved leading to LLPS (see Figs. 4a and 4b). In contrast, the kinetics of crystallization of nifedipine, a faster crystallizer54 (in the absence of a crystallization inhibitor), is sufficiently rapid that the coexistence curve cannot be reached; nifedipine crystallizes at concentrations just below the coexistence curve. However, the presence of a nucleation inhibitor, HPMC, inhibits crystallization, enabling higher supersaturations to be generated, promoting the LLPS of nifedipine when the amorphous solubility is exceeded. The excellent agreement between the predicted amorphous solubility and the concentration at which LLPS is observed experimentally suggests that the metastable zone for LLPS is small, in agreement with previous studies.17 As stated earlier, both the drug-rich and drug-lean phases are in equilibrium with each other and the drug molecules are at equivalent chemical potentials and thus have the same thermodynamic activity. Therefore: aDR = aDL = CDR γ DR = CDL γ DL
(9)
where aDR , CDR , and γ DR represents the activity, concentration, and activity coefficient in the drug-rich phase, respectively. Similarly, aDL , CDL , and γ DL represent the activity, concentration, and activity coefficient in the drug-lean phase. The maximum concentration that can be achieved in the drug-lean phase (i.e., bulk aqueous solution) is CDL . For concentrations between the crystal and amorphous solubilities, assuming that the activity coefficient term is constant across this concentration range (a reasonable assumption for these highly dilute solutions), the activity of the drug can be estimated using Eq. (12): α=
Ci CS
(10)
where Ci is the experimental concentration in the drug-lean phase, CS is the equilibrium crystalline solubility, and CDL is the maximum concentration in the drug-lean phase that can be achieved before LLPS occurs, that is, the amorphous solubility. Eq. (9) allows the observed flux versus concentration measurements shown in Figures 6a–6c to be explained on the basis of the thermodynamic activity. For the example of felodipine at 25◦ C, CDL can be taken as the predicted amorphous solubility, which corresponds to a supersaturation ratio of approximately 8. When Ci equals the equilibrium crystalline solubility, that is, CS , the supersaturation ratio is 1, and the thermodynamic activity equals 1. At all supersaturation ratios less than 8, an aqueous one-phase system persists with a thermodynamic activity of less than 8. At the amorphous solubility (Ci = CDL ),
Table 6. Comparison of LLPS Threshold Concentrations as Determined from Predicted Amorphous Solubility Values, UV Extinction Measurements, and Diffusion Studies
Felodipine Nifedipine
Temperature (◦ C)
Equilibrium Crystalline Solubility (:g/mL)
Predicted Amorphous Solubilitya (:g/mL)
LLPS Onset (UV Extinction) (:g/mL)
LLPS Onset (Diffusive Flux) (:g/mL)
25 37 37
0.74 ± 0.03 1.21 ± 0.09 10.00 ± 0.03
5.99 ± 0.01 8.70 ± 0.03 72.09 ± 1.76
6.01 ± 0.54 8.70 ± 0.70 72.85 ± 5.66
5.12 ± 0.00 11.07 ± 1.72 72.00 ± 0.00
n=3; errors indicate 1SD. a Predicted amorphous solubility with moisture correction. Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
DOI 10.1002/jps.23826
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Figure 7. Liquid–liquid phase separation and the free energy diagram. Stable and metastable refer to the stability of the liquid phases. Adapted from Deneau and Steele.23
that is, onset of LLPS, the thermodynamic activity is equal to 8 based on the use of the amorphous solubility as the upper limit. At concentrations greater than supersaturation ratios of 8, excess drug will form a drug-rich phase. Thus, irrespective of total solution concentration, the thermodynamic activity of felodipine always tends to 8 if the system is in the LLPS regime. In other words, in a liquid–liquid phase-separated system at increasing concentration, thermodynamic activity remains a constant. Thus, although it is often widely assumed that the activity term can be replaced by concentration, this assumption clearly breaks down when LLPS occurs whereby the solution thermodynamic activity is no longer proportional to the total concentration. Moreover, as diffusive flux is fundamentally driven by the activity gradient, the relationship between flux and concentration also breaks down when LLPS occurs. Hence, the only parameter in Eq. (6) upon which the flux depends is the activity. Figures 6a–6c clearly demonstrate that at and above threshold supersaturations corresponding to amorphous solubility, the flux approaches a constant, indicating that the thermodynamic activity has reached a constant, consistent with the thermodynamic properties of a phase-separated solution. This has important ramifications in the context of drug diffusion across a membrane and uptake into systemic circulation. As thermodynamic activity and flux approach a constant, the upper limit of drug absorption potential based on a passive absorption mechanism will be reached. Although concentrations at and above amorphous solubility will not yield a higher drug transport rate across biological membranes, a liquid–liquid phase-separated system will serve as a drug reservoir, providing an important potential advantage. Figure 9 compares the relative drug uptake into systemic circulation of supersaturating systems with different phase behavior. Consider a supersaturated system of a small molecule, having an apparent concentration exceeding the amorphous solubility, sitting in the gastric milieu in contact with a biological membrane devoid of any active transport processes such as influx or efflux for this particular molecule. This supersaturated system may undergo three scenarios: the system may crystallize immediately or over time, the system may persist as a one-phase supersaturated DOI 10.1002/jps.23826
2745
Figure 8. Schematic phase diagram showing metastable zone widths of LLPS and crystallization.
system (supersaturation < amorphous solubility), or the system may be present at a concentration above the amorphous solubility. When present at supersaturations greater than amorphous solubility, the system is in the LLPS regime and possesses maximum activity corresponding to maximum drug uptake potential. As the drug diffuses into the systemic circulation from the drug-lean or aqueous region of a LLPS system, drug from the drug-rich phase will replenish absorbed drug, acting as reservoir and maintaining the activity of the solution at the maximum value. This process is likely to be rapid relative to the dissolution of a crystalline material.19 A one-phase supersaturated system (concentration less than amorphous solubility) also possesses activity greater than a solution generated from the crystalline material that is, greater than 1. Although this system will exhibit higher drug uptake than a crystalline material, it will never attain the maximum flux associated with the amorphous solubility. Further, as drug from this one-phase system is depleted because of the drug uptake into systemic circulation, thermodynamic activity will drop resulting in decreasing flux and drug uptake. Thus, overall, the rate of drug transport in a LLPS system would be expected to be constant and much higher than either a crystallizing system or a onephase supersaturated solution in which the supersaturation is depleted by absorption. The use of crystallization inhibitory polymers is expected to convert a crystallizing system into a LLPS system, promoting drug uptake. Of course, the discussion presented herein is only relevant for drug uptake associated with passive diffusion and real life in vivo situations are much more complex.
CONCLUSIONS When present at elevated concentrations exceeding their amorphous solubilities, small molecules such as felodipine and nifedipine can undergo LLPS in aqueous solutions. As a consequence of this phenomenon, the thermodynamic activity of the system approaches a constant at and above the amorphous solubility. Hence, the corresponding passive diffusive flux across a membrane, which is driven by the thermodynamic activity, also approaches an upper limit or maximum. Thus, the drug Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
2746
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Figure 9. Comparison of passive diffusion and drug uptake of supersaturated systems that have undergone LLPS versus crystallization.
activities, and the corresponding passive diffusion membrane transport rates, that can be generated in aqueous media are limited by the amorphous activity, regardless of the mechanism by which supersaturation is generated. This conclusion has important theoretical and practical implications.
ACKNOWLEDGMENTS The authors would like to acknowledge AbbVie Inc. for providing research funding for this project as well as “The Dr. C. Wayne and Helen C. McKeehan Graduate Fellowship in Physical Pharmacy” for financial support for Shweta A. Raina. In addition, the authors would also like to thank Deliang Zhou, Abbvie Inc. for scientific discussion. Purdue University and AbbVie jointly participated in study design, research, data collection, analysis, and interpretation of data, writing, reviewing, and approving the publication. Shweta A. Raina is a graduate student and Lynne S. Taylor is a professor at Purdue University, and both have no additional conflicts of interest to report. David A. Alonzo is an employee at Gilead Sciences, Inc. and has no additional conflicts of interest to report. Donghua Zhu is an employee at Janssen R&D China of Johnson and Johnson and has no additional conflict of interest to report. Geoff G.Z. Zhang, Yi Gao, Jianwei Wu, Nathaniel D. Catron and Deliang Zhou are employees of AbbVie and may own AbbVie stock.
REFERENCES 1. Augustijns P, Brewster ME. 2012. Supersaturating drug delivery systems: Fast is not necessarily good enough. J Pharm Sci 101(1):7–9. 2. Brouwers J, Brewster ME, Augustijns P. 2009. Supersaturating drug delivery systems: The answer to solubility-limited oral bioavailability? J Pharm Sci 98(8):2549–2572. Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
¨ 3. Mullertz A, Ogbonna A, Ren S, Rades T. 2010. New perspectives on lipid and surfactant based drug delivery systems for oral delivery of poorly soluble drugs. J Pharm Pharmacol 62(11):1622– 1636. 4. Frenkel YV, Clark AD, Das K, Wang Y-H, Lewi PJ, Janssen PAJ, Arnold E. 2005. Concentration and pH dependent aggregation of hydrophobic drug molecules and relevance to oral bioavailability. J Med Chem 48(6):1974–1983. 5. Coan KED, Shoichet BK. 2008. Stoichiometry and physical chemistry of promiscuous aggregate-based inhibitors. J Am Chem Soc 130(29):9606–9612. 6. Coan KED, Maltby DA, Burlingame AL, Shoichet BK. 2009. Promiscuous aggregate-based inhibitors promote enzyme unfolding. J Med Chem 52(7):2067–2075. 7. Brick MC, Palmer HJ, Whitesides TH. 2003. Formation of colloidal dispersions of organic materials in aqueous media by solvent shifting. Langmuir 19(16):6367–6380. 8. Pacheco LF, Carmona-Ribeiro AM. 2003. Effects of synthetic lipids on solubilization and colloid stability of hydrophobic drugs. J Colloid Interface Sci 258(1):146–154. 9. Seidler J, McGovern SL, Doman TN, Shoichet BK. 2003. Identification and prediction of promiscuous aggregating inhibitors among known drugs. J Med Chem 46(21):4477–4486. 10. Feng BY, Simeonov A, Jadhav A, Babaoglu K, Inglese J, Shoichet BK, Austin CP. 2007. A High-Throughput Screen for Aggregationbased inhibition in a large compound library. J Med Chem 50(10):2385– 2390. 11. Doak AK, Wille H, Prusiner SB, Shoichet BK. 2010. Colloid formation by drugs in simulated intestinal fluid. J Med Chem 53(10):4259– 4265. 12. Tachibana T, Nakamura A. 1965. A methode for preparing an aqueous colloidal dispersion of organic materials by using water-soluble polymers: Dispersion of $-carotene by polyvinylpyrrolidone. KolloidZuZPolymere 203(2):130–133. 13. Friesen DT, Shanker R, Crew M, Smithey DT, Curatolo WJ, Nightingale JAS. 2008. Hydroxypropyl methylcellulose acetate succinate-based spray-dried dispersions: An overview. Mol Pharm 5(6):1003–1019. DOI 10.1002/jps.23826
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
14. Alonzo DE, Gao Y, Zhou D, Mo H, Zhang GGZ, Taylor LS. 2011. Dissolution and precipitation behavior of amorphous solid dispersions. J Pharm Sci 100(8):3316–3331. 15. Aisha AFA, Ismail Z, Abu-salah KM, Majid AMSA. 2012. Solid dispersions of "-mangostin improve its aqueous solubility through selfassembly of nanomicelles. J Pharm Sci 101(2):815–825. ¨ 16. Kanzer J, Hupfeld S, Vasskog T, Tho I, H¨olig P, Magerlein M, Fricker G, Brandl M. 2010. In situ formation of nanoparticles upon dispersion of melt extrudate formulations in aqueous medium assessed by asymmetrical flow field-flow fractionation. J Pharm Biomed Anal 53(3):359–365. 17. Ilevbare GA, Taylor LS. 2013. Liquid–liquid phase separation in highly supersaturated aqueous solutions of poorly water-soluble drugs: Implications for solubility enhancing formulations. Cryst Growth Des 13(4):1497–1509. 18. Sassene PJ, Knopp MM, Hesselkilde JZ, Koradia V, Larsen A, ¨ Rades T, Mullertz A. 2010. Precipitation of a poorly soluble model drug during in vitro lipolysis: Characterization and dissolution of the precipitate. J Pharm Sci 99(12):4982–4991. 19. Hsieh Y-L, Ilevbare G, Van Eerdenbrugh B, Box K, Sanchez-Felix M, Taylor L. 2012. pH-induced precipitation behavior of weakly basic compounds: Determination of extent and duration of supersaturation using potentiometric titration and correlation to solid state properties. Pharm Res 29(10):2738–2753. 20. Veesler S, Lafferr`ere L, Garcia E, Hoff C. 2003. Phase transitions in supersaturated drug solution. Org Proc Res Dev 7(6):983–989. 21. Veesler S, Revalor E, Bottini O, Hoff C. 2006. Crystallization in the presence of a liquid–liquid phase separation. Org Proc Res Dev 10(4):841–845. 22. Lafferr`ere L, Hoff C, Veesler S. 2003. Polymorphism and liquid– liquid demixing in supersaturated drug solution. Eng Life Sci 3(3):127– 131. 23. Deneau E, Steele G. 2005. An in-line study of oiling out and crystallization. Org Proc Res Dev 9(6):943–950. 24. Maeda K, Aoyama Y, Fukui K, Hirota S. 2001. Novel phenomena of crystallization and emulsification of hydrophobic solute in aqueous solution. J Colloid Interface Sci 234(1):217–222. 25. Tung H-H, Paul EL, Midler M, McCauley JA. 2008. In Critical issues in crystallization practice, crystallization crystallization of organic compounds: An industrial perspective. Hoboken, NJ, USA, John Wiley & Sons, Inc, pp 101–116. 26. Groen H, Roberts KJ. 2001. Nucleation, growth, and pseudopolymorphic behavior of citric acid as monitored in situ by attenuated total reflection Fourier transform infrared spectroscopy. The J Phys Chem B 105(43):10723–10730. 27. Duffy D, Cremin N, Napier M, Robinson S, Barrett M, Hao H, Glennon B. 2012. In situ monitoring, control and optimization of a liquid–liquid phase separation crystallization. Chem Eng Sci 77:112– 121. 28. Zhang F, Roosen-Runge F, Sauter A, Roth R, Skoda MWA, Jacobs R, Sztucki M, Schreiber F. 2012. The role of cluster formation and metastable liquid–liquid phase separation in protein crystallization. Faraday Dis 159:313–325. ˚ ¨ 29. Svard M, Gracin S, Rasmuson AC. 2007. Oiling out or molten hydrate—Liquid–liquid phase separation in the system vanillin— Water. J Pharm Sci 96(9):2390–2398. 30. Kiesow K, Tumakaka F, Sadowski G. 2008. Experimental investigation and prediction of oiling out during crystallization process. J Crystal Growth 310(18):4163–4168. 31. Raghavan SL, Kiepfer B, Davis AF, Kazarian SG, Hadgraft J. 2001. Membrane transport of hydrocortisone acetate from supersaturated solutions; the role of polymers. Int J Pharm 221(1–2):95–105. 32. Higuchi T. 1960. Physical chemical analysis of percutaneous absorption process from creams and ointments. J Soc Cosmet Chem 11:85–97. 33. Fischer SM, Brandl M, Fricker G. 2011. Effect of the non-ionic surfactant Poloxamer 188 on passive permeability of poorly soluble drugs across Caco-2 cell monolayers. Eur J Pharm Biopharm 79(2):416– 422. DOI 10.1002/jps.23826
2747
34. Iervolino M, Raghavan SL, Hadgraft J. 2000. Membrane penetration enhancement of ibuprofen using supersaturation. Int J Pharm 198(2):229–238. 35. Corrigan OI, Farvar MA, Higuchi WI. 1980. Drug membrane transport enhancement using high energy drug polyvinylpyrrolidone (PVP) co-precipitates. Int J Pharm 5(3):229–238. 36. Davis AF, Hadgraft J. 1991. Effect of supersaturation on membrane transport: 1. Hydrocortisone acetate. Int J Pharm 76(1–2):1–8. 37. Pellett MA, Davis AF, Hadgraft J. 1994. Effect of supersaturation on membrane transport: 2. Piroxicam Int J Pharm 111(1):1–6. 38. Santos P, Watkinson AC, Hadgraft J, Lane ME. 2011. Enhanced permeation of fentanyl from supersaturated solutions in a model membrane. Int J Pharm 407(1–2):72–77. 39. Frank KJ, Westedt U, Rosenblatt KM, H¨olig P, Rosenberg J, ¨ Magerlein M, Brandl M, Fricker G. 2012. Impact of FaSSIF on the solubility and dissolution-/permeation rate of a poorly water-soluble compound. Eur J Pharm Sci 47(1):16–20. 40. Kumprakob U, Kawakami J, Adachi I. 2005. Permeation enhancement of ketoprofen using a supersaturated system with antinucleant polymers. Biol Pharm Bull 28(9):1684–1688. 41. Frank KJ, Rosenblatt KM, Westedt U, H¨olig P, Rosenberg J, ¨ Magerlein M, Fricker G, Brandl M. 2012. Amorphous solid dispersion enhances permeation of poorly soluble ABT-102: True supersaturation vs. apparent solubility enhancement. Int J Pharm 437(1–2):288–293. 42. Miller JM, Beig A, Carr RA, Spence JK, Dahan A. 2012. A win-win solution in oral delivery of lipophilic drugs: Supersaturation via amorphous solid dispersions increases apparent solubility without sacrifice of intestinal membrane permeability. Mol Pharm 9(7):2009–2016. 43. Dahan A, Beig A, Ioffe-Dahan V, Agbaria R, Miller J. 2012. The twofold advantage of the amorphous form as an oral drug delivery practice for lipophilic compounds: Increased apparent solubility and drug flux through the intestinal membrane. AAPS J:1–7. 44. Ueda K, Higashi K, Limwikrant W, Sekine S, Horie T, Yamamoto K, Moribe K. 2012. Mechanistic differences in permeation behavior of supersaturated and solubilized solutions of carbamazepine revealed by nuclear magnetic resonance measurements. Mol Pharm 9(11):3023– 3033. 45. di Cagno M, Luppi B. 2013. Drug “supersaturation” states induced by polymeric micelles and liposomes: A mechanistic investigation into permeability enhancements. Eur J Pharm Sci 48(4–5):775– 780. 46. Alonzo D, Zhang GZ, Zhou D, Gao Y, Taylor L. 2010. Understanding the behavior of amorphous pharmaceutical systems during dissolution. Pharm Res 27(4):608–618. 47. Grant DJW, Higuchi T. 1990. Solubility behavior of organic compounds (Techniques of chemistry). New York, USA, John Wiley & Sons, p 21. 48. Hoffman JD. 1958. Thermodynamic driving force in nucleation and growth processes. J Chem Phys 29(5):1192–1193. 49. Marsac P, Konno H, Taylor L. 2006. A comparison of the physical stability of amorphous felodipine and nifedipine systems. Pharm Res 23(10):2306–2316. 50. Murdande SB, Pikal MJ, Shanker RM, Bogner RH. 2010. Solubility advantage of amorphous pharmaceuticals: I. A thermodynamic analysis. J Pharm Sci 99(3):1254–1264. 51. Qin X-Z, DeMarco J, Ip DP. 1995. Simultaneous determination of enalapril, felodipine and their degradation products in the dosage formulation by reversed-phase high-performance liquid chromatography using a Spherisorb C8 column. J Chromatogr A 707(2):245–254. 52. Videbæk LM, Kvist S, Mulvany MJ. 1995. Time course of action of amlodipine and felodipine in the rat is most rapid in small arteries. Eur J Pharm 274(1–3):109–115. 53. Motomura S, Wu Z, Hashimoto K. 1993. Lacidipine, a new longacting dihydropyridine calcium antagonist, has high vascular selectivity against all intracardiac variables. Heart Vessels 8(1):16–22. 54. Baird JA, Van Eerdenbrugh B, Taylor LS. 2010. A classification system to assess the crystallization tendency of organic molecules from undercooled melts. J Pharm Sci 99(9):3787–3806. Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
2748
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
55. Alonzo DE, Raina S, Zhou D, Gao Y, Zhang GGZ, Taylor LS. 2012. Characterizing the impact of hydroxypropylmethyl cellulose on the growth and nucleation kinetics of felodipine from supersaturated solutions. Crystal Growth Des 12(3):1538–1547. 56. Cussler EL. 1988. Diffusion: Mass transfer in fluid systems. New York, USA, Cambridge University Press, pp 184–185. 57. Sinko PJ. 2010. Martin’s Physical Pharmacy and Pharmaceutical Sciences. Baltimore, Maryland: Lippincott Williams & Wilkins.
Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
58. Qi S, Roser S, Edler K, Pigliacelli C, Rogerson M, Weuts I, Dycke F, Stokbroekx S. 2013. Insights into the role of polymer-surfactant complexes in drug solubilisation/stabilisation during drug release from solid dispersions. Pharm Res 30(1):290–302. 59. Yalkowsky SH. 2012. Perspective on improving passive human intestinal absorption. J Pharm Sci 101(9):3047–3050. 60. Barry BW. 2001. Novel mechanisms and devices to enable successful transdermal drug delivery. Eur J Pharm Sci 14:101–114.
DOI 10.1002/jps.23826
Enhancements and Limits in Drug Membrane Transport Using Supersaturated Solutions of Poorly Water Soluble Drugs SHWETA A. RAINA,1 GEOFF G. Z. ZHANG,2 DAVID E. ALONZO,2 JIANWEI WU,2 DONGHUA ZHU,2 NATHANIEL D. CATRON,2 YI GAO,2 LYNNE S. TAYLOR1 1 2
Department of Industrial and Physical Pharmacy, College of Pharmacy, Purdue University, West Lafayette, Indiana Drug Product Development, Research and Development, AbbVie Inc., North Chicago, Illinois
Received 6 September 2013; revised 14 November 2013; accepted 25 November 2013 Published online 30 December 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23826 ABSTRACT: Amorphous solid dispersions (ASDs) give rise to supersaturated solutions (solution concentration greater than equilibrium crystalline solubility). We have recently found that supersaturating dosage forms can exhibit the phenomenon of liquid–liquid phase separation (LLPS). Thus, the high supersaturation generated by dissolving ASDs can lead to a two-phase system wherein one phase is an initially nanodimensioned and drug-rich phase and the other is a drug-lean continuous aqueous phase. Herein, the membrane transport of supersaturated solutions, at concentrations above and below the LLPS concentration has been evaluated using a side-by-side diffusion cell. Measurements of solution concentration with time in the receiver cell yield the flux, which reflects the solute thermodynamic activity in the donor cell. As the nominal concentration of solute in the donor cell increases, a linear increase in flux was observed up to the concentration where LLPS occurred. Thereafter, the flux remained essentially constant. Both nifedipine and felodipine solutions exhibit such behavior as long as crystallization is absent. This suggests that there is an upper limit in passive membrane transport that is dictated by the LLPS concentration. These results have several important implications for drug delivery, especially for poorly soluble C 2013 Wiley Periodicals, Inc. and the American Pharmacists Association J compounds requiring enabling formulation technologies. Pharm Sci 103:2736–2748, 2014 Keywords: absorption; amorphous; bioavailability; crystallization; diffusion; dispersion; liquid-liquid phase separation; passive; uptake
INTRODUCTION Low aqueous solubility of emerging drug candidates is a major hurdle in drug development. Many of the small molecules showing therapeutic activity against biological targets of interest have very low aqueous solubility either because they are lipophilic or have a high melting point, or because of the combined effect of these two factors. Therefore, there is a great deal of interest in formulating such compounds into dosage forms that can supersaturate.1–3 It has been noted that these lipophilic small molecules, when present at elevated supersaturations in aqueous media, can form colloidal drug aggregates that can subsequently coalesce leading to an increase in size, and/or crystallize.4–10 Colloidal aggregates of small molecules are routinely observed in the high-throughput enzymatic screens employed during drug discovery and are formed when concentrated aqueous solutions of drugs are generated by the solvent shifting method.7 In this method, a small aliquot of a concentrated solution of the drug, dissolved in an organic solution, is added to an aqueous solution to generate the desired concentration, often resulting in a highly supersaturated solution. A supersaturated solution is one where the concentration (or more Correspondence to: Yi Gao (Telephone: +847-937-9028; E-mail: [email protected]); Lynne S. Taylor (Telephone: +765-496-6614; Fax: +765-4946545; E-mail: [email protected]) David E. Alonzo’s present address is Formulation and Process Development, Gilead Sciences, Inc., Foster City, California. Donghua Zhu’s present address is Janssen R&D of Johnson and Johnson, China. Journal of Pharmaceutical Sciences, Vol. 103, 2736–2748 (2014)
C 2013 Wiley Periodicals, Inc. and the American Pharmacists Association
2736
rigorously, the chemical potential of the solute) exceeds that of a solution saturated with respect to the crystalline solid. These aggregates have attracted attention because they frequently lead to false positives in enzyme-based assays via nonspecific protein inhibition. Aggregation may also lead to false negatives as molecules in an aggregated state may be unable to interact with proteins. Compounds undergoing this phenomenon have been termed “promiscuous aggregators.” Studies by Doak et al.,11 Frenkel et al.,4 along with several other research groups have demonstrated that an array of active pharmaceutical ingredients (APIs) form nanosized aggregates. The formation of colloidal aggregates is concentration dependent and occurs only above a critical concentration termed the critical aggregate concentration. The formation of colloidal species has also been reported to occur following dissolution of amorphous solid dispersions (ASDs). One of the earliest studies reported the formation of colloidal species after dissolving a dispersion of polyvinyl pyrrolidone (PVP) and $-carotene.12 ASDs prepared with hydroxypropylmethyl cellulose acetate succinate (HPMCAS) are also reported to generate colloidal species.13 Other systems where colloid formation has been noted include felodipine and hydroxypropylmethlyl cellulose (HPMC),14 mangostin and PVP,15 and lopinavir/ritonavir formulations with copovidone.16 Recent studies have demonstrated that the formation of colloidal species upon dissolution of an amorphous dispersion only occurs when a critical concentration of dissolved drug has been achieved.17 The formation of these colloidal species has been postulated to be beneficial for oral drug delivery.13 It has been suggested that the underlying phenomenon causing the formation of colloidal aggregates, either from the solvent
Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
switch method, or from dissolution of an ASD, is liquid–liquid phase separation (LLPS).7,17 LLPS can also occur from other supersaturation generating systems for example during lipolysis of a self-microemulsifying drug delivery system18 or as a result of pH changes.19 Highly supersaturated solutions will undergo LLPS when a certain threshold concentration is exceeded whereby a homogeneous one-phase system separates into two liquid phases. LLPS is also referred to as oiling-out or liquid–liquid demixing.20–27 LLPS has been widely observed in a variety of systems including polymer blends, proteins, metals, and is often prevalent, albeit undesired, during indus¨ et al.29 reported LLPS for the trial crystallization.24,27,28 Svard water–vanillin system. Kiesow et al.30 reported LLPS during the crystallization of 4-4 -dihydroxydiphenylsulfone and used modeling to predict its occurrence. In the context of aqueous solutions of hydrophobic drugs with melting points above the experimental temperature, LLPS is the separation of a supersaturated drug solution into two liquid phases, one of which is drug rich and hydrophobic, that is, colloidal aggregates that predominantly consist of the drug, whereas the other phase is water rich and contains only a low concentration of drug.17 The two phases are in dynamic equilibrium, although both phases are thermodynamically metastable and the system is supersaturated. In formulated products, additives such as polymers, surfactants, complexing agents, and lipids are routinely used to improve solubility and dissolution profiles in vitro in the hope that this will enhance systemic drug concentrations in vivo. Often, extremely high-concentration enhancements are reported based on approaches that involve assaying the solution phase. This approach of measuring solution concentration can be problematic when attempting to predict in vivo performance because it does not discriminate between the two mechanisms of achieving enhanced solution concentrations: supersaturation versus solubility enhancement. True supersaturation occurs when there is an increase in the chemical potential of the solute relative to a saturated solution in which the solute chemical potential is the same as that of the crystalline solid. For example, polymers at low concentrations do not typically enhance equilibrium solubility,31 but enable supersaturated solutions to be generated and maintained after dissolving an amorphous formulation by inhibiting crystallization from supersaturated solutions. In contrast, micellar surfactant, pH adjustment, cyclodextrins, and cosolvents increase the equilibrium solubility of the crystalline solid (although supersaturation may be generated when the resultant system is diluted). As the crystalline solid is in equilibrium with the solution, it is clear that an increase in solution concentration can be achieved without an increase in the chemical potential of the solute. Because membrane transport is driven by the chemical potential gradient,32 it is important to consider the thermodynamic properties of the solution rather than the total solution concentration. Thus, it is well established that membrane transport can be decreased by the presence of micellar surfactant,33 cyclodextrins,34 but is increased by supersaturated solutions as long as crystallization is prevented.33,35–41 Although most of the studies on the relationship between membrane transport and supersaturation have been in the context of transdermal delivery, there is an increasing interest in exploiting supersaturated solutions to enhance oral absorption. Recent studies have demonstrated enhanced flux across intestinal membranes when perfused with supersaturated solutions.40–45 However, the impact of the colDOI 10.1002/jps.23826
2737
loidal species generated in highly supersaturated solutions has not been fully evaluated to date. Diffusion data from Alonzo et al.46 suggested that the flux reached a maximum when concentrations equal to or above the “amorphous solubility” were generated. Thus, at very high concentrations in the supersaturated regime, it appears that there may be a breakdown in the relationship between concentration and diffusion rate. The goal of the current study was to demonstrate that small organic molecules dissolved in aqueous solution at concentrations at and above their amorphous solubility undergo LLPS, and that the LLPS concentration corresponds to the maximum achievable diffusive flux for a supersaturated solution. Felodipine and nifedipine were selected as model drugs to evaluate the impact of supersaturation and LLPS on membrane transport. The approach of Corrigan et al.35 was utilized, whereby crystallization inhibitors were added to the donor solution to sustain supersaturation for sufficient time to enable diffusion measurements to be made; the polymer, hydroxypropyl methyl cellulose was used as the crystallization inhibitor. Phase diagrams were constructed to understand the concentrations at which liquid– liquid and liquid–solid phase transformations occurred.
MATERIALS Felodipine and nifedipine were purchased from Attix Pharmaceuticals (Toronto, Ontario, Canada) and Euroasia (Mumbai, India), respectively. HPMC Pharmacoat grade 606 was obtained from ShinEtsu (Shin-Etsu Chemical Company, Ltd., Tokyo, Japan). Dissolution media used in all experiments comprised 50 mM pH 6.8 phosphate buffer (ionic strength = 0.155 M) without or with predissolved polymer at a concentration of 100 :g/mL and 1 mg/mL for felodipine and nifedipine, respectively. Methyl alcohol was purchased from Pharmco Products, Inc., Brookfield, Connecticut. Molecular structures of the model compounds are shown in Figure 1. Regenerated cellulose membrane with a molecular weight cutoff (MWCO) of 6–8 K was obtained from Spectrum Laboratories, Inc. (Rancho Dominguez, California).
METHODS Crystalline Solubility Measurements Equilibrium crystalline solubility was determined using a modification of the shake flask method. An excess of crystalline felodipine and nifedipine was equilibrated in 20 mL scintillation vials with 50 mM pH 6.8 phosphate buffer in an agitating water bath (Dubnoff metallic shaking incubator; PGC Scientific, Palm Desert, California). Vials were wrapped in aluminum foil to protect the samples from light. Preliminary experiments indicated that equilibrium was reached by 48 h. Samples were agitated for 48 h at 20◦ C, 25◦ C, 30◦ C, 37◦ C, and 45◦ C before subjecting them to ultracentrifugation to separate excess solid from the supernatant (which is saturated with drug). An Optima L100 XP ultracentrifuge equipped with Swinging-Bucket Rotor SW 41 Ti (Beckman Coulter, Inc., Brea, California) was used and samples were centrifuged at 274,356g for 15 min. The supernatant obtained was diluted with acetonitrile: 50 mM phosphate buffer (50:50) in a 1:1 ratio and 200 :L samples were injected into an Agilent 1100 high-performance liquid chromatography system (Agilent Technologies, Santa Clara, California). The chromatographic separation was performed with Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
2738
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Figure 1. Molecular structures of model compounds.
a Zobrax Eclipse Plus C18 analytical column (2.1 × 150 mm2) ˚ (Agilent Technologies) using acetonitrile: 50 i.d., 5 :m, 100 A) mM phosphate buffer (70:30) as mobile phase. Flow through the column was maintained at 1 mL/min and the column was heated to 40◦ C. The pressure across the column was 28 bar and detection was carried out using ultraviolet (UV) absorbance at a wavelength of 360 nm. The total run time was 7 min and felodipine and nifedipine peaks appeared at 4.2 and 2.5 min, respectively. A standard curve was prepared using samples dissolved in methanol with an injection volume of 20 :L covering a concentration range of 0.1–50 :g/mL. Samples were analyzed in triplicate. The standard curve exhibited good linearity (r 2 > 0.998) over the specified concentration range. The equilibrium solubility of felodipine and nifedipine was also determined in the presence of 100 :g/mL and 1 mg/mL of HPMC in 50 mM pH 6.8 phosphate buffer (ionic strength = 0.155 M) using the method outlined above. Amorphous Solubility Estimations Amorphous solubility can be estimated by initially calculating the free energy difference between the crystal and supercooled liquid, for example, by using the Hoffman equation. This first estimate is then corrected for moisture sorption and ionization. Equation (1) shows the factors contributing to the solubility of a crystalline material: ln(x)solute = −
Hf R
−
Cp R
1 1 − T Tm 1−
Tm Tm + ln T T
− lnγ
(1)
where ln(x)solute is the natural log of the mole fraction solubility, Hf is the molar heat of fusion of the crystal, R is the universal gas constant, T is the temperature of interest, Tm is the melting temperature, and Cp is the heat capacity difference between the crystal and the liquid. The activity coefficient, γ accounts for nonideal solute–solvent interactions.47 Amorphous materials lack the long range order characteristic of their crystalline counterparts and hence Eq. (1) reduces to Eq. (2). ln(x)solute = − ln γ
(2)
After a series of approximations17,48 and derivations, the ratio of amorphous to crystalline solubility Fa /Fc can be written as Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
Eq. (3), using the Hoffman equation to estimate the free energy difference between the amorphous and crystalline material48 :
σa G = e RT = e σc
Hf T T Tm Tm RT
(
) (3)
Although the Hoffman equation provides a reasonable first approximation of the free energy difference between the amorphous and crystalline material,49 neither the impact of the degree of ionization of the molecules or the reduced activity of the amorphous solid resulting from water sorption are taken into account in Eq. (3). An improved estimate of amorphous solubility may be attained by using the novel approach developed by Murdande et al.50 a −μx σa (1 − α x ) [−I(a2 )] (μ2RT 2) e = e a σc (1 − α )
(4)
here R is the gas constant and T is temperature. α a represents fraction ionized for an amorphous solute, whereas α x represents the fraction ionized for the crystalline solute. The first term on the right hand side in the above equation represents differences in the ionization states in solutions equilibrated with amorphous versus crystalline forms. This ionization term is equal to 1 if the API is unionizable or is present in a buffered solution where ionization is minimal. Felodipine (pKa < 2)51,52 and nifedipine (pKa < 3.5)53 are unionized at pH 6.8 and thus this term reduces to 1 (for ionizable molecules or solute present at pH where it is ionized, this term is F
Between groups Within groups Total Between groups Within groups Total
9.73E-8 5.40E-7 6.37E-7 1.27E-7 5.52E-6 5.65E-6
2 12 14 2 9 11
4.87E-8 4.50E-8
1.08
0.37
6.33E-8 6.13E-7
0.103
0.90
Between groups Within groups Total
2.73E-5 2.73E-4 2.99E-4
2 6 8
1.36E-5 4.55E-5
0.30
0.75
Null hypothesis: means of all levels are equal. Alternate hypothesis: the means of one or more level are different. At the 0.05 level, the population means are not significantly different for felodipine at 25◦ C and 37◦ C and nifedipine at 37◦ C. a Data shown compare means from 5, 10, 25, 50, and 100 :g/mL groups. b Data shown compare means from 10, 25, 50, and 100 :g/mL groups. c Data shown compare means from 72, 108, and 145 :g/mL groups.
important to note that other variables such as temperature, viscosity, and timescale of experiments also impact metastable zone width. The spinodal represents the supersaturation where the solution is unstable and there is no free energy barrier for the nucleation of a new phase. It has been observed that the distance between the spinodal and the binodes is small for organic molecules.17 For relatively slow crystallizers like felodipine,54 where the nucleation rate is slow relative to the rate of supersaturation generation, it is clear that supersaturations that exceed the amorphous solubility can be achieved leading to LLPS (see Figs. 4a and 4b). In contrast, the kinetics of crystallization of nifedipine, a faster crystallizer54 (in the absence of a crystallization inhibitor), is sufficiently rapid that the coexistence curve cannot be reached; nifedipine crystallizes at concentrations just below the coexistence curve. However, the presence of a nucleation inhibitor, HPMC, inhibits crystallization, enabling higher supersaturations to be generated, promoting the LLPS of nifedipine when the amorphous solubility is exceeded. The excellent agreement between the predicted amorphous solubility and the concentration at which LLPS is observed experimentally suggests that the metastable zone for LLPS is small, in agreement with previous studies.17 As stated earlier, both the drug-rich and drug-lean phases are in equilibrium with each other and the drug molecules are at equivalent chemical potentials and thus have the same thermodynamic activity. Therefore: aDR = aDL = CDR γ DR = CDL γ DL
(9)
where aDR , CDR , and γ DR represents the activity, concentration, and activity coefficient in the drug-rich phase, respectively. Similarly, aDL , CDL , and γ DL represent the activity, concentration, and activity coefficient in the drug-lean phase. The maximum concentration that can be achieved in the drug-lean phase (i.e., bulk aqueous solution) is CDL . For concentrations between the crystal and amorphous solubilities, assuming that the activity coefficient term is constant across this concentration range (a reasonable assumption for these highly dilute solutions), the activity of the drug can be estimated using Eq. (12): α=
Ci CS
(10)
where Ci is the experimental concentration in the drug-lean phase, CS is the equilibrium crystalline solubility, and CDL is the maximum concentration in the drug-lean phase that can be achieved before LLPS occurs, that is, the amorphous solubility. Eq. (9) allows the observed flux versus concentration measurements shown in Figures 6a–6c to be explained on the basis of the thermodynamic activity. For the example of felodipine at 25◦ C, CDL can be taken as the predicted amorphous solubility, which corresponds to a supersaturation ratio of approximately 8. When Ci equals the equilibrium crystalline solubility, that is, CS , the supersaturation ratio is 1, and the thermodynamic activity equals 1. At all supersaturation ratios less than 8, an aqueous one-phase system persists with a thermodynamic activity of less than 8. At the amorphous solubility (Ci = CDL ),
Table 6. Comparison of LLPS Threshold Concentrations as Determined from Predicted Amorphous Solubility Values, UV Extinction Measurements, and Diffusion Studies
Felodipine Nifedipine
Temperature (◦ C)
Equilibrium Crystalline Solubility (:g/mL)
Predicted Amorphous Solubilitya (:g/mL)
LLPS Onset (UV Extinction) (:g/mL)
LLPS Onset (Diffusive Flux) (:g/mL)
25 37 37
0.74 ± 0.03 1.21 ± 0.09 10.00 ± 0.03
5.99 ± 0.01 8.70 ± 0.03 72.09 ± 1.76
6.01 ± 0.54 8.70 ± 0.70 72.85 ± 5.66
5.12 ± 0.00 11.07 ± 1.72 72.00 ± 0.00
n=3; errors indicate 1SD. a Predicted amorphous solubility with moisture correction. Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
DOI 10.1002/jps.23826
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Figure 7. Liquid–liquid phase separation and the free energy diagram. Stable and metastable refer to the stability of the liquid phases. Adapted from Deneau and Steele.23
that is, onset of LLPS, the thermodynamic activity is equal to 8 based on the use of the amorphous solubility as the upper limit. At concentrations greater than supersaturation ratios of 8, excess drug will form a drug-rich phase. Thus, irrespective of total solution concentration, the thermodynamic activity of felodipine always tends to 8 if the system is in the LLPS regime. In other words, in a liquid–liquid phase-separated system at increasing concentration, thermodynamic activity remains a constant. Thus, although it is often widely assumed that the activity term can be replaced by concentration, this assumption clearly breaks down when LLPS occurs whereby the solution thermodynamic activity is no longer proportional to the total concentration. Moreover, as diffusive flux is fundamentally driven by the activity gradient, the relationship between flux and concentration also breaks down when LLPS occurs. Hence, the only parameter in Eq. (6) upon which the flux depends is the activity. Figures 6a–6c clearly demonstrate that at and above threshold supersaturations corresponding to amorphous solubility, the flux approaches a constant, indicating that the thermodynamic activity has reached a constant, consistent with the thermodynamic properties of a phase-separated solution. This has important ramifications in the context of drug diffusion across a membrane and uptake into systemic circulation. As thermodynamic activity and flux approach a constant, the upper limit of drug absorption potential based on a passive absorption mechanism will be reached. Although concentrations at and above amorphous solubility will not yield a higher drug transport rate across biological membranes, a liquid–liquid phase-separated system will serve as a drug reservoir, providing an important potential advantage. Figure 9 compares the relative drug uptake into systemic circulation of supersaturating systems with different phase behavior. Consider a supersaturated system of a small molecule, having an apparent concentration exceeding the amorphous solubility, sitting in the gastric milieu in contact with a biological membrane devoid of any active transport processes such as influx or efflux for this particular molecule. This supersaturated system may undergo three scenarios: the system may crystallize immediately or over time, the system may persist as a one-phase supersaturated DOI 10.1002/jps.23826
2745
Figure 8. Schematic phase diagram showing metastable zone widths of LLPS and crystallization.
system (supersaturation < amorphous solubility), or the system may be present at a concentration above the amorphous solubility. When present at supersaturations greater than amorphous solubility, the system is in the LLPS regime and possesses maximum activity corresponding to maximum drug uptake potential. As the drug diffuses into the systemic circulation from the drug-lean or aqueous region of a LLPS system, drug from the drug-rich phase will replenish absorbed drug, acting as reservoir and maintaining the activity of the solution at the maximum value. This process is likely to be rapid relative to the dissolution of a crystalline material.19 A one-phase supersaturated system (concentration less than amorphous solubility) also possesses activity greater than a solution generated from the crystalline material that is, greater than 1. Although this system will exhibit higher drug uptake than a crystalline material, it will never attain the maximum flux associated with the amorphous solubility. Further, as drug from this one-phase system is depleted because of the drug uptake into systemic circulation, thermodynamic activity will drop resulting in decreasing flux and drug uptake. Thus, overall, the rate of drug transport in a LLPS system would be expected to be constant and much higher than either a crystallizing system or a onephase supersaturated solution in which the supersaturation is depleted by absorption. The use of crystallization inhibitory polymers is expected to convert a crystallizing system into a LLPS system, promoting drug uptake. Of course, the discussion presented herein is only relevant for drug uptake associated with passive diffusion and real life in vivo situations are much more complex.
CONCLUSIONS When present at elevated concentrations exceeding their amorphous solubilities, small molecules such as felodipine and nifedipine can undergo LLPS in aqueous solutions. As a consequence of this phenomenon, the thermodynamic activity of the system approaches a constant at and above the amorphous solubility. Hence, the corresponding passive diffusive flux across a membrane, which is driven by the thermodynamic activity, also approaches an upper limit or maximum. Thus, the drug Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
2746
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Figure 9. Comparison of passive diffusion and drug uptake of supersaturated systems that have undergone LLPS versus crystallization.
activities, and the corresponding passive diffusion membrane transport rates, that can be generated in aqueous media are limited by the amorphous activity, regardless of the mechanism by which supersaturation is generated. This conclusion has important theoretical and practical implications.
ACKNOWLEDGMENTS The authors would like to acknowledge AbbVie Inc. for providing research funding for this project as well as “The Dr. C. Wayne and Helen C. McKeehan Graduate Fellowship in Physical Pharmacy” for financial support for Shweta A. Raina. In addition, the authors would also like to thank Deliang Zhou, Abbvie Inc. for scientific discussion. Purdue University and AbbVie jointly participated in study design, research, data collection, analysis, and interpretation of data, writing, reviewing, and approving the publication. Shweta A. Raina is a graduate student and Lynne S. Taylor is a professor at Purdue University, and both have no additional conflicts of interest to report. David A. Alonzo is an employee at Gilead Sciences, Inc. and has no additional conflicts of interest to report. Donghua Zhu is an employee at Janssen R&D China of Johnson and Johnson and has no additional conflict of interest to report. Geoff G.Z. Zhang, Yi Gao, Jianwei Wu, Nathaniel D. Catron and Deliang Zhou are employees of AbbVie and may own AbbVie stock.
REFERENCES 1. Augustijns P, Brewster ME. 2012. Supersaturating drug delivery systems: Fast is not necessarily good enough. J Pharm Sci 101(1):7–9. 2. Brouwers J, Brewster ME, Augustijns P. 2009. Supersaturating drug delivery systems: The answer to solubility-limited oral bioavailability? J Pharm Sci 98(8):2549–2572. Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
¨ 3. Mullertz A, Ogbonna A, Ren S, Rades T. 2010. New perspectives on lipid and surfactant based drug delivery systems for oral delivery of poorly soluble drugs. J Pharm Pharmacol 62(11):1622– 1636. 4. Frenkel YV, Clark AD, Das K, Wang Y-H, Lewi PJ, Janssen PAJ, Arnold E. 2005. Concentration and pH dependent aggregation of hydrophobic drug molecules and relevance to oral bioavailability. J Med Chem 48(6):1974–1983. 5. Coan KED, Shoichet BK. 2008. Stoichiometry and physical chemistry of promiscuous aggregate-based inhibitors. J Am Chem Soc 130(29):9606–9612. 6. Coan KED, Maltby DA, Burlingame AL, Shoichet BK. 2009. Promiscuous aggregate-based inhibitors promote enzyme unfolding. J Med Chem 52(7):2067–2075. 7. Brick MC, Palmer HJ, Whitesides TH. 2003. Formation of colloidal dispersions of organic materials in aqueous media by solvent shifting. Langmuir 19(16):6367–6380. 8. Pacheco LF, Carmona-Ribeiro AM. 2003. Effects of synthetic lipids on solubilization and colloid stability of hydrophobic drugs. J Colloid Interface Sci 258(1):146–154. 9. Seidler J, McGovern SL, Doman TN, Shoichet BK. 2003. Identification and prediction of promiscuous aggregating inhibitors among known drugs. J Med Chem 46(21):4477–4486. 10. Feng BY, Simeonov A, Jadhav A, Babaoglu K, Inglese J, Shoichet BK, Austin CP. 2007. A High-Throughput Screen for Aggregationbased inhibition in a large compound library. J Med Chem 50(10):2385– 2390. 11. Doak AK, Wille H, Prusiner SB, Shoichet BK. 2010. Colloid formation by drugs in simulated intestinal fluid. J Med Chem 53(10):4259– 4265. 12. Tachibana T, Nakamura A. 1965. A methode for preparing an aqueous colloidal dispersion of organic materials by using water-soluble polymers: Dispersion of $-carotene by polyvinylpyrrolidone. KolloidZuZPolymere 203(2):130–133. 13. Friesen DT, Shanker R, Crew M, Smithey DT, Curatolo WJ, Nightingale JAS. 2008. Hydroxypropyl methylcellulose acetate succinate-based spray-dried dispersions: An overview. Mol Pharm 5(6):1003–1019. DOI 10.1002/jps.23826
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
14. Alonzo DE, Gao Y, Zhou D, Mo H, Zhang GGZ, Taylor LS. 2011. Dissolution and precipitation behavior of amorphous solid dispersions. J Pharm Sci 100(8):3316–3331. 15. Aisha AFA, Ismail Z, Abu-salah KM, Majid AMSA. 2012. Solid dispersions of "-mangostin improve its aqueous solubility through selfassembly of nanomicelles. J Pharm Sci 101(2):815–825. ¨ 16. Kanzer J, Hupfeld S, Vasskog T, Tho I, H¨olig P, Magerlein M, Fricker G, Brandl M. 2010. In situ formation of nanoparticles upon dispersion of melt extrudate formulations in aqueous medium assessed by asymmetrical flow field-flow fractionation. J Pharm Biomed Anal 53(3):359–365. 17. Ilevbare GA, Taylor LS. 2013. Liquid–liquid phase separation in highly supersaturated aqueous solutions of poorly water-soluble drugs: Implications for solubility enhancing formulations. Cryst Growth Des 13(4):1497–1509. 18. Sassene PJ, Knopp MM, Hesselkilde JZ, Koradia V, Larsen A, ¨ Rades T, Mullertz A. 2010. Precipitation of a poorly soluble model drug during in vitro lipolysis: Characterization and dissolution of the precipitate. J Pharm Sci 99(12):4982–4991. 19. Hsieh Y-L, Ilevbare G, Van Eerdenbrugh B, Box K, Sanchez-Felix M, Taylor L. 2012. pH-induced precipitation behavior of weakly basic compounds: Determination of extent and duration of supersaturation using potentiometric titration and correlation to solid state properties. Pharm Res 29(10):2738–2753. 20. Veesler S, Lafferr`ere L, Garcia E, Hoff C. 2003. Phase transitions in supersaturated drug solution. Org Proc Res Dev 7(6):983–989. 21. Veesler S, Revalor E, Bottini O, Hoff C. 2006. Crystallization in the presence of a liquid–liquid phase separation. Org Proc Res Dev 10(4):841–845. 22. Lafferr`ere L, Hoff C, Veesler S. 2003. Polymorphism and liquid– liquid demixing in supersaturated drug solution. Eng Life Sci 3(3):127– 131. 23. Deneau E, Steele G. 2005. An in-line study of oiling out and crystallization. Org Proc Res Dev 9(6):943–950. 24. Maeda K, Aoyama Y, Fukui K, Hirota S. 2001. Novel phenomena of crystallization and emulsification of hydrophobic solute in aqueous solution. J Colloid Interface Sci 234(1):217–222. 25. Tung H-H, Paul EL, Midler M, McCauley JA. 2008. In Critical issues in crystallization practice, crystallization crystallization of organic compounds: An industrial perspective. Hoboken, NJ, USA, John Wiley & Sons, Inc, pp 101–116. 26. Groen H, Roberts KJ. 2001. Nucleation, growth, and pseudopolymorphic behavior of citric acid as monitored in situ by attenuated total reflection Fourier transform infrared spectroscopy. The J Phys Chem B 105(43):10723–10730. 27. Duffy D, Cremin N, Napier M, Robinson S, Barrett M, Hao H, Glennon B. 2012. In situ monitoring, control and optimization of a liquid–liquid phase separation crystallization. Chem Eng Sci 77:112– 121. 28. Zhang F, Roosen-Runge F, Sauter A, Roth R, Skoda MWA, Jacobs R, Sztucki M, Schreiber F. 2012. The role of cluster formation and metastable liquid–liquid phase separation in protein crystallization. Faraday Dis 159:313–325. ˚ ¨ 29. Svard M, Gracin S, Rasmuson AC. 2007. Oiling out or molten hydrate—Liquid–liquid phase separation in the system vanillin— Water. J Pharm Sci 96(9):2390–2398. 30. Kiesow K, Tumakaka F, Sadowski G. 2008. Experimental investigation and prediction of oiling out during crystallization process. J Crystal Growth 310(18):4163–4168. 31. Raghavan SL, Kiepfer B, Davis AF, Kazarian SG, Hadgraft J. 2001. Membrane transport of hydrocortisone acetate from supersaturated solutions; the role of polymers. Int J Pharm 221(1–2):95–105. 32. Higuchi T. 1960. Physical chemical analysis of percutaneous absorption process from creams and ointments. J Soc Cosmet Chem 11:85–97. 33. Fischer SM, Brandl M, Fricker G. 2011. Effect of the non-ionic surfactant Poloxamer 188 on passive permeability of poorly soluble drugs across Caco-2 cell monolayers. Eur J Pharm Biopharm 79(2):416– 422. DOI 10.1002/jps.23826
2747
34. Iervolino M, Raghavan SL, Hadgraft J. 2000. Membrane penetration enhancement of ibuprofen using supersaturation. Int J Pharm 198(2):229–238. 35. Corrigan OI, Farvar MA, Higuchi WI. 1980. Drug membrane transport enhancement using high energy drug polyvinylpyrrolidone (PVP) co-precipitates. Int J Pharm 5(3):229–238. 36. Davis AF, Hadgraft J. 1991. Effect of supersaturation on membrane transport: 1. Hydrocortisone acetate. Int J Pharm 76(1–2):1–8. 37. Pellett MA, Davis AF, Hadgraft J. 1994. Effect of supersaturation on membrane transport: 2. Piroxicam Int J Pharm 111(1):1–6. 38. Santos P, Watkinson AC, Hadgraft J, Lane ME. 2011. Enhanced permeation of fentanyl from supersaturated solutions in a model membrane. Int J Pharm 407(1–2):72–77. 39. Frank KJ, Westedt U, Rosenblatt KM, H¨olig P, Rosenberg J, ¨ Magerlein M, Brandl M, Fricker G. 2012. Impact of FaSSIF on the solubility and dissolution-/permeation rate of a poorly water-soluble compound. Eur J Pharm Sci 47(1):16–20. 40. Kumprakob U, Kawakami J, Adachi I. 2005. Permeation enhancement of ketoprofen using a supersaturated system with antinucleant polymers. Biol Pharm Bull 28(9):1684–1688. 41. Frank KJ, Rosenblatt KM, Westedt U, H¨olig P, Rosenberg J, ¨ Magerlein M, Fricker G, Brandl M. 2012. Amorphous solid dispersion enhances permeation of poorly soluble ABT-102: True supersaturation vs. apparent solubility enhancement. Int J Pharm 437(1–2):288–293. 42. Miller JM, Beig A, Carr RA, Spence JK, Dahan A. 2012. A win-win solution in oral delivery of lipophilic drugs: Supersaturation via amorphous solid dispersions increases apparent solubility without sacrifice of intestinal membrane permeability. Mol Pharm 9(7):2009–2016. 43. Dahan A, Beig A, Ioffe-Dahan V, Agbaria R, Miller J. 2012. The twofold advantage of the amorphous form as an oral drug delivery practice for lipophilic compounds: Increased apparent solubility and drug flux through the intestinal membrane. AAPS J:1–7. 44. Ueda K, Higashi K, Limwikrant W, Sekine S, Horie T, Yamamoto K, Moribe K. 2012. Mechanistic differences in permeation behavior of supersaturated and solubilized solutions of carbamazepine revealed by nuclear magnetic resonance measurements. Mol Pharm 9(11):3023– 3033. 45. di Cagno M, Luppi B. 2013. Drug “supersaturation” states induced by polymeric micelles and liposomes: A mechanistic investigation into permeability enhancements. Eur J Pharm Sci 48(4–5):775– 780. 46. Alonzo D, Zhang GZ, Zhou D, Gao Y, Taylor L. 2010. Understanding the behavior of amorphous pharmaceutical systems during dissolution. Pharm Res 27(4):608–618. 47. Grant DJW, Higuchi T. 1990. Solubility behavior of organic compounds (Techniques of chemistry). New York, USA, John Wiley & Sons, p 21. 48. Hoffman JD. 1958. Thermodynamic driving force in nucleation and growth processes. J Chem Phys 29(5):1192–1193. 49. Marsac P, Konno H, Taylor L. 2006. A comparison of the physical stability of amorphous felodipine and nifedipine systems. Pharm Res 23(10):2306–2316. 50. Murdande SB, Pikal MJ, Shanker RM, Bogner RH. 2010. Solubility advantage of amorphous pharmaceuticals: I. A thermodynamic analysis. J Pharm Sci 99(3):1254–1264. 51. Qin X-Z, DeMarco J, Ip DP. 1995. Simultaneous determination of enalapril, felodipine and their degradation products in the dosage formulation by reversed-phase high-performance liquid chromatography using a Spherisorb C8 column. J Chromatogr A 707(2):245–254. 52. Videbæk LM, Kvist S, Mulvany MJ. 1995. Time course of action of amlodipine and felodipine in the rat is most rapid in small arteries. Eur J Pharm 274(1–3):109–115. 53. Motomura S, Wu Z, Hashimoto K. 1993. Lacidipine, a new longacting dihydropyridine calcium antagonist, has high vascular selectivity against all intracardiac variables. Heart Vessels 8(1):16–22. 54. Baird JA, Van Eerdenbrugh B, Taylor LS. 2010. A classification system to assess the crystallization tendency of organic molecules from undercooled melts. J Pharm Sci 99(9):3787–3806. Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
2748
RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology
55. Alonzo DE, Raina S, Zhou D, Gao Y, Zhang GGZ, Taylor LS. 2012. Characterizing the impact of hydroxypropylmethyl cellulose on the growth and nucleation kinetics of felodipine from supersaturated solutions. Crystal Growth Des 12(3):1538–1547. 56. Cussler EL. 1988. Diffusion: Mass transfer in fluid systems. New York, USA, Cambridge University Press, pp 184–185. 57. Sinko PJ. 2010. Martin’s Physical Pharmacy and Pharmaceutical Sciences. Baltimore, Maryland: Lippincott Williams & Wilkins.
Raina et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:2736–2748, 2014
58. Qi S, Roser S, Edler K, Pigliacelli C, Rogerson M, Weuts I, Dycke F, Stokbroekx S. 2013. Insights into the role of polymer-surfactant complexes in drug solubilisation/stabilisation during drug release from solid dispersions. Pharm Res 30(1):290–302. 59. Yalkowsky SH. 2012. Perspective on improving passive human intestinal absorption. J Pharm Sci 101(9):3047–3050. 60. Barry BW. 2001. Novel mechanisms and devices to enable successful transdermal drug delivery. Eur J Pharm Sci 14:101–114.
DOI 10.1002/jps.23826
