Preparation of candesartan and atorvastatin nanoparticles by solvent evaporation

Molecules 2012, 17, 13221-13234; doi:10.3390/molecules171113221 molecules ISSN 1420-3049 www.mdpi.com/journal/molecules Article Preparation of Candesartan and Atorvastatin Nanoparticles by Solvent Evaporation † Eliska Vaculikova 1,2,*, Veronika Grunwaldova 3,4, Vladimir Kral 3, Jiri Dohnal 1,5 and Josef Jampilek 1,5,* 1 Faculty of Pharmacy, University of Veterinary and Pharmaceutical Sciences, Palackeho 1/3, 612 42 Brno, Czech Republic 2 Nanotechnology Centre, VSB—Technical University of Ostrava, 17. listopadu 15/2172, 708 33 Ostrava, Czech Republic 3 Institute of Chemical Technology, Faculty of Chemical Engineering, Technicka 5, 166 28 Prague 6, Czech Republic 4 Institute of Inorganic Chemistry, Academy of Science, 250 68 Rez, Czech Republic 5 Research Institute for Pharmacy and Biochemistry, Lidicka 1879/48, 602 00 Brno, Czech Republic † Preliminary results were presented at The Sixteenth Electronic Conference on Synthetic Organic Chemistry (ECSOC-16, http://www.sciforum.net/presentation/1008), 1–30 November, 2012 (paper B1008). * Authors to whom correspondence should be addressed; E-Mails: [email protected] (E.V.); [email protected] (J.J.); Tel.: +420-541-562-925 (J.J.); Fax: +420-541-240-607 (J.J.). Received: 8 October 2012; in revised form: 31 October 2012 / Accepted: 2 November 2012 / Published: 6 November 2012 Abstract: The solubility, absorption and distribution of a drug are involved in the basic aspects of oral bioavailability Solubility is an essential characteristic and influences the efficiency of the drug. Over the last ten years, the number of poorly soluble drugs has steadily increased. One of the progressive ways for increasing oral bioavaibility is the technique of nanoparticle preparation, which allows many drugs to thus reach the intended site of action. Candesartan cilexetil and atorvastatin, belonging to class II of the biopharmaceutical classification system, were chosen as model active pharmaceutical ingredients in this study. Forty samples were prepared either by antisolvent precipitation/solvent evaporation method or by the emulsion/solvent evaporation technique with various commonly used surface-active excipients as nanoparticle stabilizers. All samples were analyzed by means of dynamic light scattering. The particle size of the OPEN ACCESS Molecules 2012, 17 13222 determined 36 nanoparticle samples was to 574 nm, whereas 32 samples contained nanoparticles of less than 200 nm. Relationships between solvents and excipients used and their amount are discussed. Based on the results the investigated solvent evaporation methods can be used as an effective and an affordable technique for the preparation of nanoparticles. Keywords: candesartan cilexetil; atorvastatin; nanoparticles; solvent evaporation; excipients; dynamic light scattering 1. Introduction For ensure the pharmacological activity of an active pharmaceutical ingredient (API), the solubility of the API in physiological liquids is required, so that the API can be available at the place of absorption. Solubility in various solvents is a characteristic property of a particular compound. The solubility of a compound in water correlates to a great extent with the solubility in physiological liquids and is the first limiting factor for good absorption and biodistribution. Contrary to these facts, over the last ten years, the number of poorly soluble drugs has steadily increased. Estimates suggest that 40% of the drugs in the pipelines have solubility problems. Literature states that about 60% of all drugs coming directly from synthesis nowadays are poorly soluble [1–3]. One of the progressive ways how to increase the solubility of an APIs is the preparation of drug nanoparticles. The technique of nanoparticle drug delivery allows many pharmacological agents to reach the desired site of action. APIs are either adjusted alone till nano size and administered in nanoparticle oral dosage forms or attached/incorporated into nanoparticles prepared from inert materials which serve as a universal drug delivery system. The advantages of nanotechnology are as follows: (i) increased bioavailability (quick dissolution; improved penetration through membranes); (ii) lower doses; (iii) lower toxicity; (iv) targeted biodistribution; (v) reduction of influence of food on variability; (vi) quicker development of formulations [2,4–7]. Nanoparticles of less than 200 nm are of practical importance [8–13]. A great problem is the insufficiently investigated possible toxicity of nanoparticles. The toxicity is dependent on the shape and surface properties of nanoparticles, because both can influence nanoparticle-cell interactions as well as the rate of penetration to cells. Among the various nanoparticle forms nanotubes were found to be one of the most toxic nanoparticle shapes [14–17]. A wide range of techniques have been developed for the preparation of nanomaterials. These methods are typically grouped into two categories: top-down (generally dispergation processes) [11–13,18–21] and bottom-up (generally precipitation processes) [11–13,18,22–24], whereas the latter is by far the most popular in the preparation of nanoparticles. In bottom-up methods, nanoparticles can be produced by crystallization/precipitation and solvent evaporation. Spray drying, evaporative precipitation into aqueous solution, microemulsions or supercritical fluid technology belong to the solvent evaporation methods. The liquid antisolvent (LAS) precipitation process is a noteworthy method that has been extensively studied. An excellent review dealing with this technique was published by Thorat et al. [25]. The current paper is aimed at verification of conditions of an effective Molecules 2012, 17 13223 and an affordable technique for the preparation of nanoparticles by solvent evaporation as was discussed recently [26]. A polar and nonpolar solvent were used in our research, therefore the exact principle of the applied solvent evaporation method is dependent on the water-based system, including or not an aqueous miscible organic solvent. The polar acetone (AC) and nonpolar dichloromethane (DCM) were chosen as the most suitable solvents for easy dissolution of the APIs, so two different possible mechanisms can be supposed for the nanoparticle synthesis. When API is dissolved in AC and then mixed with water containing a stabilizer, nanoparticles are formed spontaneously and immediately upon mixing. This method can be called antisolvent precipitation/solvent evaporation, and the procedure is in principle similar to the evaporative precipitation into aqueous solution [27,28] or the liquid antisolvent precipitation [25]. When the API is dissolved in DCM and then mixed with water containing stabilizers, an emulsion (o/w type) is formed; API is clustered by the excipient, which results in the encapsulation of the API into nano-vesicula. This combination of emulsification and solvent evaporation nanoparticle synthesis can be called emulsion/solvent evaporation [19,29]. The model APIs candesartan cilexetil (I) and atorvastatin calcium (II) were chosen as representatives of poorly aqueous soluble compounds belonging to class II drugs of the biopharmaceutical classification system (BCS). Drugs of the mentioned class are characterized by low aqueous solubility and high permeability [30]. Candesartan (2-ethoxy-1-({4-[2-(2H-1,2,3,4-tetrazol-5yl)phenyl]phenyl}methyl)-1H-1,3-benzodiazole-6-carboxylic acid) is an angiotensin II receptor antagonist used mainly for the treatment of hypertension. The prodrug candesartan cilexetil, see Figure 1, is marketed by AstraZeneca and Takeda Pharmaceuticals, commonly under the trade names Blopress®, Atacand®, Amias®, and Ratacand®. The use of a prodrug form increases the bioavailability of candesartan. Despite this, its absolute oral bioavailability is relatively poor (approx. 15%) [31,32]. Atorvastatin [(3R,5R)-7-[2-(4-fluorophenyl)-3-phenyl-4-(phenylcarbamoyl)-5-propan-2-ylpyrrol-1-yl]- 3,5-dihydroxyheptanoic acid] inhibits HMG-CoA reductase and thus causes a decrease of cholesterol in the body. Atorvastatin is used as a calcium salt, see Figure 1, and is marketed by Pfizer under the trade names Lipitor® or Sortis®. The low plasma concentration (approx. 12%) of atorvastatin is especially caused by an extensive first-pass metabolism in the liver, nevertheless the overall solubility of atorvastatin is strictly pH-dependent (many atorvastatin solid dosage forms are buffered, e.g., by carbonates), and administration with food produces a 25% reduction of atorvastatin absorption [33,34]. As mentioned, both APIs are BCS class II drugs, hence their oral bioavailability is solubility rate limited [30–32,35–37]. For enhancement of solubility of candesartan cilexetil various approaches can be used, such as pectin complexes [38], self-emulsifying drug delivery systems [39] or development of nanoparticle formulations [7,40]. The solubility of atorvastatin calcium can be enhanced, for example, using an amorphous API [41], by application of the liquisolid technique [42], formulation of self-microemulsifying drug delivery systems [43], utilization of drug-polymer interactions found due to physical mixing [44] or preparation of amorphous nanoparticles [45]. Various types of surface-active excipients were used as nanoparticle stabilizers and relationships between a substance, a solvent and a used excipient are discussed. Used excipients represent various classes of pharmaceutical adjutants (emulsifiers/viscosity modifiers/thickeners, nonionic or anionic surfactants) that can be utilized as solubility modifying compounds/nanoparticle stabilizers, such as Tween 80 (TW), sodium dodecyl sulfate (SDS), macrogol 6000 (PEG), sodium carboxymethyl Molecules 2012, 17 13224 cellulose (SCMC) and sodium salt of carboxymethyl dextran (SCMD). The main criteria for excipient selection were its pharmaceutical safety (all excipients are GRAS, Generally Recognized as Safe, substances) and their affordability. Based on a previous study 5% and 10% concentrations of each excipient were chosen [26]. The optimal concentration of surfactant is important for optimal particles wetting. If the concentration is too low, particles float on the surface. If the concentration is too high bubbles appear [46]. Figure 1. Structures of candesartan cilexetil as prodrug and atorvastatin calcium salt. N N N HN N N O O O OO O N N HO F OOH OH O candesartan cilexetil (I) 2 Ca 2 atorvastatin calcium (II) 2. Results and Discussion Both model APIs I and II dissolved in dichloromethane and acetone (2% concentration) were added to aqueous solutions (5%, 10% concentration) of excipients, i.e., eight samples were prepared with each excipient. The final API:excipient ratios were 1:2.5 (2%:5%), 1:5 (2%:10%). The systems were stirred for 10 min at 35 °C; afterwards the mixtures were transferred to an ultrasonic bath, where they were mixed again for 40 min, and simultaneously the organic solvent was evaporated (to final 10 mL sample volume) by self-warming of the ultrasonic bath. Then all the samples were characterized by dynamic light scattering [46]. All the results are presented in Tables 1–5 and Figures 2–8. Table 1. Particle size (x10, x90 [nm]) of APIs I, II and concentration [%] of Tween 80 in dichloromethane (DCM) or acetone (AC). All the presented results are reported as the medium value of four independent measurements. The results of nano-size samples are expressed as the mean ± SD (n = 4 measurements). The SDs of micro-size samples are not indicated due to the measurability range of Nanophox. Samples that contained nanoparticles <200 nm are bolded; nanoparticles <10 nm are indicated by asterisk. (S.No. = sample number). API/Solvent Tween 80 S.No. 5% S.No. 10% x10 x90 x10 x90 Particle size [nm] I/DCM 1 160 ± 4.8 219 ± 5.6 2 14 ± 0.4 16 ± 0.5 * I/AC 3 3183 6531 4 2 ± 0.1 3 ± 0.1 * II/DCM 5 97 ± 2.9 142 ± 4.3 6 145 ± 4.4 213 ± 6.4 II/AC 7 101 ± 3.0 111 ± 3.3 8 3 ± 0.1 4 ± 0.1 * Molecules 2012, 17 13225 Table 2. Particle size (x10, x90 [nm]) of APIs I, II and concentration [%] of sodium dodecyl sulfate in dichloromethane (DCM) or acetone (AC). All the presented results are reported as the medium value of four independent measurements. The results of nano-size samples are expressed as the mean ± SD (n = 4 measurements). The SDs of micro-size samples are not indicated due to the measurability range of Nanophox. Samples that contained nanoparticles <200 nm are bolded; nanoparticles <10 nm are indicated by asterisk. (S.No. = sample number). API/Solvent Sodium dodecyl sulfate S.No. 5% S.No. 10% x10 x90 x10 x90 Particle size [nm] I/DCM 9 90 ± 2.7 99 ± 3.0 10 2 ± 0.1 3 ± 0.1 * I/AC 11 4 ± 0.1 5 ± 0.2 * 12 1 ± 0.03 2 ± 0.1 * II/DCM 13 1 ± 0.03 2 ± 0.1 * 14 90 ± 2.7 99 ± 3.0 II/AC 15 2 ± 0.1 2 ±0.1 * 16 2 ± 0.1 4 ± 0.1 * Table 3. Particle size (x10, x90 [nm]) of APIs I, II and concentration [%] of macrogol 6000 in dichloromethane (DCM) or acetone (AC). All the presented results are reported as the medium value of four independent measurements. The results of nano-size samples are expressed as the mean ± SD (n = 4 measurements). The SDs of micro-size samples are not indicated due to the measurability range of Nanophox. Samples that contained nanoparticles <200 nm are bolded; nanoparticles <10 nm are indicated by asterisk. (S.No. = sample number). API/Solvent Macrogol 6000 S.No. 5% S.No. 10% x10 x90 x10 x90 Particle size [nm] I/DCM 17 2 ± 0.1 3 ± 0.1 * 18 2 ± 0.1 3 ± 0.1 * I/AC 19 2 ± 0.1 3 ± 0.1 * 20 156 ± 4.7 206 ± 6.2 II/DCM 21 1639 1804 22 5231 5755 II/AC 23 6 ± 0.2 8 ± 0.2 * 24 4 ± 0.1 6 ± 0.2 * Table 4. Particle size (x10, x90 [nm]) of APIs I, II and concentration [%] of sodium carboxymethyl cellulose in dichloromethane (DCM) or acetone (AC). All the presented results are reported as the medium value of four independent measurements. The results of nano-size samples are expressed as the mean ± SD (n = 4 measurements). The SDs of micro-size samples are not indicated due to the measurability range of Nanophox. Samples that contained nanoparticles <200 nm are bolded; nanoparticles <10 nm are indicated by asterisk. (S.No. = sample number). API/Solvent Sodium carboxymethyl cellulose S.No. 5% S.No. 10% x10 x90 x10 x90 Particle size [nm] I/DCM 25 11 ± 0.3 13 ± 0.4 26 2 ± 0.1 3 ± 0.1 * I/AC 27 1 ± 0.03 2 ± 0.1 * 28 32 ± 1.0 35 ± 1.1 II/DCM 29 401 ± 12 574 ± 17 30 1 ± 0.03 2 ± 0.1 * II/AC 31 6 ± 0.2 7 ± 0.2 * 32 27 ± 0.8 30 ± 0.9 Molecules 2012, 17 13226 Table 5. Particle size (x10, x90 [nm]) of APIs I, II and concentration [%] of sodium carboxymethyl dextran in dichloromethane (DCM) or acetone (AC). All the presented results are reported as the medium value of four independent measurements. The results of nano-size samples are expressed as the mean ± SD (n = 4 measurements). The SDs of micro-size samples are not indicated due to the measurability range of Nanophox. Samples that contained nanoparticles <200 nm are bolded; nanoparticles <10 nm are indicated by asterisk. (S.No. = sample number). API/Solvent Sodium carboxymethyl dextran S.No. 5% S.No. 10% x10 x90 x10 x90 Particle size [nm] I/DCM 33 2 ± 0.1 2 ± 0.1 * 34 1 ± 0.03 1 ± 0.03 * I/AC 35 3 ± 0.1 4 ± 0.1 * 36 39 ± 1.2 43 ± 1.3 II/DCM 37 2 ± 0.1 2 ± 0.1 * 38 9345 10281 II/AC 39 2 ± 0.1 3 ± 0.1 * 40 70 ± 2.1 77 ± 2.3 Figure 2. Dependence of particle size (x90 [nm]) of model APIs I, II on concentration [%] of Tween 80 in dichloromethane (DCM) or acetone (AC). (A) Samples are grouped according to APIs; (B) samples are grouped according to excipient percentage. For clarity sake, the values on y-axis are only to 300 nm. Figure 3. Dependence of particle size (x90 [nm]) of model APIs I, II on concentration [%] of sodium dodecyl sulfate in dichloromethane (DCM) or acetone (AC). (A) Samples are grouped according to APIs; (B) samples are grouped according to excipient percentage. For clarity sake, the values on y-axis are only to 100 nm. 0 50 100 150 200 250 300 I/DCM I/AC II/DCM II/AC X 90 [nm] API / Excipient concentration [%] 5% 10% 0 50 100 150 200 250 300 5% 10% X 90 [nm] Excipient concentration [%] / API I/DCM I/AC II/DCM II/AC A B 0 10 20 30 40 50 60 70 80 90 100 I/DCM I/AC II/DCM II/AC X 90 [nm] API / Excipient concentration [%] 5% 10% 0 10 20 30 40 50 60 70 80 90 100 5% 10% X 90 [nm] Excipient concentration [%] / API I/DCM I/AC II/DCM II/AC A B Molecules 2012, 17 13227 Figure 4. Dependence of particle size (x90 [nm]) of model APIs I, II on concentration [%] of macrogol 6000 in dichloromethane (DCM) or acetone (AC). (A) Samples are grouped according to APIs; (B) samples are grouped according to excipient percentage. For clarity sake, the values on y-axis are only to 300 nm. Figure 5. Dependence of particle size (x90 [nm]) of model APIs I, II on concentration [%] of sodium carboxymethyl cellulose in dichloromethane (DCM) or acetone (AC). (A) Samples are grouped according to APIs; (B) samples are grouped according to excipient percentage. For clarity sake, the values on y-axis are only to 100 nm. Figure 6. Dependence of particle size (x90 [nm]) of model APIs I, II on concentration [%] of sodium carboxymethyl dextran in dichloromethane (DCM) or acetone (AC). (A) Samples are grouped according to APIs; (B) samples are grouped according to excipient percentage. For clarity sake, the values on y-axis are only to 100 nm.  0 50 100 150 200 250 300 I/DCM I/AC II/DCM II/AC X 90 [nm] API / Excipient concentration [%] 5% 10% 0 50 100 150 200 250 300 5% 10% X 90 [nm] Excipient concentration [%] / API I/DCM I/AC II/DCM II/AC A B 0 10 20 30 40 50 60 70 80 90 100 I/DCM I/AC II/DCM II/AC X 90 [nm] API / Excipient concentration [%] 5% 10% 0 10 20 30 40 50 60 70 80 90 100 5% 10% X 90 [nm] Excipient concentration [%] / API I/DCM I/AC II/DCM II/AC A B 0 10 20 30 40 50 60 70 80 90 100 I/DCM I/AC II/DCM II/AC X 90 [nm] API / Excipient concentration [%] 5% 10% 0 10 20 30 40 50 60 70 80 90 100 5% 10% X 90 [nm] Excipient concentration [%] / API I/DCM I/AC II/DCM II/AC A B Molecules 2012, 17 13228 Figure 7. Dependence of particle size (x90 [nm]) of candesartan cilexetil (I) on concentration [%] of Tween 80 (TW), sodium dodecyl sulfate (SDS), macrogol 6000 (PEG), sodium carboxymethyl cellulose (SCMC), sodium carboxymethyl dextran (SCMD) in dichloromethane (DCM) or acetone (AC). For clarity sake, the values on y-axis are only to 50 nm. Figure 8. Dependence of particle size (x90 [nm]) of atorvastatin calcium (II) on concentration [%] of Tween 80 (TW), sodium dodecyl sulfate (SDS), macrogol 6000 (PEG), sodium carboxymethyl cellulose (SCMC), sodium carboxymethyl dextran (SCMD) in dichloromethane (DCM) or acetone (AC). For clarity sake, the values on y-axis are only to 50 nm. Figures 2–6 illustrate the dependence of particle size expressed as the cumulative distribution x90 [nm] of the APIs I, II on the concentration [%] of an individual excipient, whereas in Figures A samples are grouped according to individual APIs I, II, while in Figures B individual APIs are always separated according to the percentage of the excipient. The particle size x90 was used for evaluation of the method success, since this value represents 90% of the cumulative particle size distribution in the measured sample. The dispersity is a measure/degree of the homogeneity/heterogeneity of sizes of particles in a mixture/system. It is possible to see this feature on the width of the particle-size distribution, which is described as differences between cumulative distribution x10 and x90, see Tables 1–5. According to the results, when micro-size samples (3, 21, 22, 38) were eliminated, the average relation of the cumulative distribution x10/x90 ranged from 0.6 to 0.9. It is possible to suppose that DCM AC 0 10 20 30 40 50 TW 5% TW 10% SDS 5% SDS 10% PEG 5% PEG 10% SCMC 5% SCMC 10% SCMD 5% SCMD 10% 1 219 2 16 9 99 10 3 17 3 18 3 25 13 26 4 33 2 34 1 3 6531 4 3 11 512 2 19 3 20 206 27 2 28 35 35 4 36 43 Solvent X90 [nm] Excipient concentration [%] DCM AC 0 10 20 30 40 50 TW 5% TW 10% SDS 5% SDS 10% PEG 5% PEG 10% SCMC 5% SCMC 10% SCMD 5% SCMD 10% 5 142 6 213 13 2 14 99 21 1804 22 5755 29 574 30 2 37 2 38 10281 7 111 8 415 2 16 4 23 824 6 31 7 32 30 39 3 40 77 Solvent X 90 [nm] Excipient concentration [%] Molecules 2012, 17 13229 nanoparticles are spheres, because the size in dynamic light scattering represents the hydrodynamic diameter of the particle. All samples were dispersed by ultrasonics directly before the measurement to avoid possible re-agglomeration. Stabilization of the dispersed samples was achieved by surfactants and by the constant temperature. The measuring cell was equilibrated at 25 °C, so the Brown motion of nanoparticles is influenced just by their size. From Figures 2A–6A it can be stated that generally particle size is not dependent on the type of model API, but it is partially influenced by the type and concentration of the excipient utilized. Nevertheless, it can be supposed that in the case of candesartan cilexetil (I) smaller particles were found, especially when atorvastatin calcium (II) and SDS, PEG and SCMC in dichloromethane were used, as it is illustrated in Figures 7 and 8, where the dependences of the particle size of individual APIs I and II on the concentrations of individual excipients are shown. This fact is probably caused by the solvent used, because dichloromethane has less advantageous properties in comparison with acetone, as discussed below. Table 6 summarizes results of all the samples of nanoparticles under 900 nm size depending on solvents and the type and amount of excipients. As the aim of this contribution is specification of suitable conditions for nanoparticles preparation, in Table 6 generated nanoparticles are not divided according to used APIs. Table 6. View of formed samples of nanoparticles (≤900 nm) depending on solvents and type and amount of excipients. (conc. = concentration; excp. = excipient; dichloromethane = DCM; acetone = AC; Tween 80 = TW; sodium dodecyl sulfate = SDS; macrogol 6000 = PEG; sodium carboxymethyl cellulose = SCMC; sodium carboxymethyl dextran = SCMD). Excp. conc./type DCM Sum total Overall average x90 [nm] AC Sum total Overall average x90 [nm] 5% 10% 5% 10% number of nanop. samples number of nanop. samples TW 2 2 4 147 1 2 3 39 SDS 2 2 4 51 2 2 4 3 PEG 1 1 2 3 2 2 4 56 SCMC 2 2 4 148 2 2 4 18 SCMD 2 1 3 2 2 2 4 32 Sum total 9 8 17 351 9 10 19 148 Overall average x90 [nm] 117 42 160 70 80 16 41 57 30 29 After summation of all the results it can be concluded that from 40 prepared mixtures 36 samples contained nanoparticles (see Tables 1–5), from which 32 samples contained nanoparticles smaller than 200 nm (see Tables 1–5, bolded values). Nanoparticles under 10 nm were determined in 22 samples from 32, see Tables 1–5 (asterisked bolded values). Based on the results listed in Table 6 and Figures 7 and 8 it can be generally stated that the solvent used plays the crucial role in generation of nanoparticles. This fact was not so evident in the previous study, where only steroid-like compounds were investigated [26]. This effect of solvent was significant in the case of atorvastatin calcium (II), which is a salt and thus by its chemical nature absolutely