β-cyclodextrin hinders PLGA plasticization during microparticle manufacturing

Accepted Manuscript β-Cyclodextrin Hinders PLGA plasticization during microparticle manufacturing Barbara Albertini, Nunzio Iraci, Aurélie Schoubben, Stefano Giovagnoli, Maurizio Ricci, Paolo Blasi, Carlo Rossi PII:

S1773-2247(15)00139-2

DOI:

10.1016/j.jddst.2015.07.022

Reference:

JDDST 83

To appear in:

Journal of Drug Delivery Science and Technology

Received Date: 24 April 2015 Revised Date:

26 July 2015

Accepted Date: 28 July 2015

Please cite this article as: B. Albertini, N. Iraci, A. Schoubben, S. Giovagnoli, M. Ricci, P. Blasi, C. Rossi, β-Cyclodextrin Hinders PLGA plasticization during microparticle manufacturing, Journal of Drug Delivery Science and Technology (2015), doi: 10.1016/j.jddst.2015.07.022. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT β-CYCLODEXTRIN HINDERS PLGA PLASTICIZATION DURING MICROPARTICLE MANUFACTURING

Barbara Albertini1§, Nunzio Iraci1§, Aurélie Schoubben1, Stefano Giovagnoli1, Maurizio Ricci1,

1

RI PT

Paolo Blasi2*, Carlo Rossi3**

Dipartimento di Scienze Farmaceutiche, Università degli Studi di Perugia, via del Liceo 1, 06123,

Scuola di Scienze del Farmaco e dei Prodotti della Salute, Università degli Studi di Camerino, Via

Sant’Agostino 1, 62032, Camerino, Italy. 3

M AN U

2

SC

Perugia, Italy.

Consorzio Interuniversitario Nazionale di Tecnologie Farmaceutiche Innovative, TEFARCO

EP

TE D

Innova, Viale delle Scienze 27/a campus – 43124, Parma, Italy.

§Barbara Albertini and Nunzio Iraci contributed equally to this work.

AC C

*Correspondence should be addressed to Paolo Blasi; e-mail: [email protected], phone +39 0737402289, fax: +39 0737637345. ** All of us are very much indebted to Prof. Dominique Duchêne for the great contribution given to the development of the cyclodextrin field and for the unbiased and enthusiastic way she raised the international importance of the Journal of Drug Delivery Science and Technology as Editor in Chief.

1

ACCEPTED MANUSCRIPT Abstract Macromolecules, such as poly(lactide-co-glycolide), have nowadays a pivotal role in drug delivery technologies. To have a precise and predictable control of the drug release kinetics in a delivery system, polymer physico-chemical characteristics must be deeply investigated together

RI PT

with all the other components of the device, including the active pharmaceutical ingredient/s. In fact, drug-polymer interaction may result in drastic changes of polymer characteristics. Plasticization, by changing polymer mechanical properties and diffusion coefficients through the

SC

matrix, may seriously compromise device performances.

We investigated the possibility to hinder the plasticizing effect of ketoprofen on poly(lactide-co-

cyclodextrin and then encapsulated in PLGA.

M AN U

glycolide) (PLGA) during microparticle preparation. To this aim, ketoprofen was included in β-

The formation of the inclusion complex was confirmed by X-ray and FT-IR data, and UV-VIS analysis showed that ~85% of ketoprofen was included in the cyclodextrins. Molecular dynamics

TE D

forecasted the enthalpy-driven formation of 4 inclusion complexes with hydrogen bonding playing an important role on their stability. PLGA microparticles were prepared using solvent diffusion/evaporation method with encapsulation efficiency higher than 60%, without polymer

EP

plasticization.

We proved that ketoprofen/β-cyclodextrin inclusion complex allowed to hinder the plasticizing

AC C

effect of ketoprofen on PLGA during microparticle preparation. The inclusion of ketoprofen within β-cyclodextrin did not allow to establish physical interactions with the polymer chains avoiding thus plasticization.

Keywords:

ketoprofen,

β-cyclodextrin,

poly(lactide-co-glycolide),

microparticles, molecular dynamics simulations

2

plasticizing

effect,

ACCEPTED MANUSCRIPT Introduction Biocompatible biodegradable polymers have, nowadays, a significant role in medical and pharmaceutical sciences. Among the different synthetic polymers investigated and proposed for drug delivery, homo- and heteropolymers prepared by condensation of lactic and/or glycolic acid

RI PT

are, at the moment, the most successful. Borne as absorbable materials for internal sutures in the seventies [1], these polyesters have been massively studied for drug delivery in the last 30 years. Tablets [2], pellets [3,4], microparticles [5], nanoparticles [6], slabs, and in situ forming devices [7]

SC

have been formulated using these macromolecules. Their application in regenerative medicine as scaffolding material and/or drug delivery systems is also under evaluation [8-11].

M AN U

Poly(lactide) (PLA) or poly(lactide-co-glycolide) (PLGA) are partially crystalline or completely amorphous polymers depending from their composition. Due to the amorphous state or the presence of amorphous regions in between crystallites, their glass transition temperature (Tg) becomes a fundamental parameter to monitor and to control (especially in totally amorphous materials). A

TE D

modification in Tg value during manufacturing may drastically change device performances (e.g., release kinetics and mechanisms, tensile strength). In the specific case of drug delivery systems, the diffusion coefficients of small molecules through the polymer matrix may increase of several

EP

orders of magnitude going from glassy to rubbery state [12]. Physical interaction of small organic molecules, macromolecules or even inorganic compounds

AC C

(e.g., active pharmaceutical ingredients, excipients) with the polymer chains may reduce or increase the Tg. A substance that mixed to a polymer reduces its Tg is generally called plasticizer while an antiplasticizer is a substance that provokes Tg increase [13,14]. Pharmaceutically relevant polymers may need addition of plasticizers to facilitate the manufacturing process or to obtain the desired characteristics [15,16]. However, in some cases, the plasticizing effect may be undesired [17,18]. It has been reported that different non-steroidal anti-inflammatory drugs, such as acetyl salicylic acid derivatives [19], ketoprofen [17,18,20], and ibuprofen [21-23], act as plasticizers when mixed 3

ACCEPTED MANUSCRIPT with hydrophilic or hydrophobic polymers. This effect limits not only the possibility to obtain the desired release kinetics, customizable choosing the right molecular weight, lactic/glycolic acid ratio (in the specific case of PLA/PLGA), end-chain, and device characteristics (e.g., size, porosity) but also the possibility to manufacture correctly such a device.

RI PT

Polymer plasticization has been largely studied to allow easier manufacturing, but the problems deriving from undesired plasticization (e.g., drug-polymer interaction), have been largely ignored in drug delivery science and technology. This is surely due to an incomplete understanding of the

SC

phenomenon and to inadequate theories and models available at the moment [24]. The addition of excipients to modulate drug release is a common practice [16,21] but few attempts to hinder

M AN U

polymer plasticization have been reported [21].

Ketoprofen loaded PLA/PLGA microparticles have been previously formulated using the solvent diffusion-evaporation technique. Due to ketoprofen and water [25] plasticizing effects, the production of PLA/PLGA microparticles was not possible with low-molecular-weight polymers

TE D

(low Tg value) using the mentioned method [18].

It was hypothesized that, by hindering the physical interaction between ketoprofen and polymer chains, plasticizing effect could be eluded during microparticle preparation. To test this hypothesis,

EP

ketoprofen was included in β-cyclodextrins (βCDs) and the inclusion complex was encapsulated in

AC C

low-molecular-weight PLA/PLGA.

4

ACCEPTED MANUSCRIPT 2. Materials and methods 2.1. Materials Polymers poly(D,L-lactide-co-glycolide) 50:50 RG 502 (Mw ̴10 kDa), RG 502 H (Mw ̴10 kDa), and poly(D,L-lactide) RG 202H (Mw

̴10 kDa) were purchased from Boehringer Ingelheim

RI PT

(Ingelheim, Germany). Ketoprofen was kindly supplied by Bidachem S.p.A. (Bergamo, Italy). βCDs were obtained from Fluka Chemical (Milan, Italy), KBr and (hydroxypropyl) methyl cellulose (viscosity of 80-120 cP, aqueous solution 2% w/v) were obtained from Sigma Aldrich

SC

Chemical Co. Ltd. (Milan, Italy). Dichloromethane and acetonitrile (HPLC grade) were purchased from J. T. Baker (Milan, Italy). Ultrapure water was obtained from Human Power 1 system (Human

available.

M AN U

Corporation, Caserta, Italy). All solvents used were of the highest purity grade commercially

2.2. Preparation and characterization of the inclusion complex

TE D

2.2.1. Inclusion complex preparation

Inclusion complex was prepared using the kneading method at two different βCD:ketoprofen stoichiometric ratios, 1:1 and 2:1. The two starting materials were mixed in a mortar and wetted

EP

with a minimum volume of ethanol/water (1:1 v/v) and kneaded for 10 minutes. The mixture was

AC C

exsiccated under vacuum for 3 days to obtain the inclusion complex as dry powder [26].

2.2.2. UV-Vis spectrophotometry The inclusion complexes were spectrophotometrically analyzed (λ, 254 nm) to quantify the free and total ketoprofen content using an Agilent 8453 spectrophotometer. To evaluate total ketoprofen content in the inclusion complex, 10 mg of powder were solubilized in 10 mL of a mixture of acetonitrile: water, 1:1 v/v. To determine the free ketoprofen, 10 mg of inclusion complex were dispersed in 10 mL of dichloromethane and the suspension obtained was filtered using Teflon filter

5

ACCEPTED MANUSCRIPT (0.2 µm-Lida, Kenosha). The solution obtained after filtration was analyzed to determine the free ketoprofen content [27].

2.2.3. Morphological analysis

RI PT

Morphological characterization was performed by scanning electron microscopy (SEM) using a Philips XL30 SEM (Philips Electron Optics, Heindoven, Netherlands). Samples were prepared by placing inclusion complex powder, physical mixtures, βCD and ketoprofen powder onto an

SC

aluminium stub covered by a carbon double sided adhesive disc. Samples were coated with gold prior to imaging (EMITECH K-550 sputter coater Ashford, Kent, UK). Coating was carried out at

M AN U

20 mA for 4 minutes.

2.2.4. Thermal analysis

Differential scanning calorimetry (DSC) data were acquired by a calorimeter DSC821e (Mettler

TE D

Toledo, Greifensee, Switzerland) equipped with a refrigerated cooling system. The system was calibrated by an indium standard. Samples (exactly weighed) were sealed in aluminium pans and submitted to a heating cycle from 25 °C to 100 °C at 10 °C/min. Data were elaborated with STARe

AC C

3 independent measures.

EP

software (Mettler Toledo, Griefen-see, Switzerland), and the results were expressed as the mean of

2.2.5. X-ray powder diffraction analysis X-ray powder diffraction spectra were recorded by a Philips 1710 powder X-ray diffractometer using Cu Kα radiation, a voltage of 40 kV and a 30 mA current. The diffractometer had a PW1820 goniometer and a graphite monochromator for diffracted rays. Spectra were recorded using the step scanning method (step size 0.03°) and elaborated with PC-APD software. Samples were prepared using the lateral loading method to minimize the preferential orientation.

6

ACCEPTED MANUSCRIPT

2.2.6. FT-IR analysis Samples were prepared by mixing the powder with KBr and compressing the mixture with a hydraulic press (Perkin Elmer, Beaconsfield, United Kingdom) (13-mm diameter die; compression

RI PT

force, 5 tons; time, 1 minute). FT-IR spectra were recorded on a Jasco FT-IR-410, 420 Herschel spectrometer (Jasco Corporation Tokyo, Japan) with an EasyDiffTM diffuse reflectance accessory.

SC

2.2.7. Molecular dynamics simulations

βCD and R- and S-ketoprofen were sketched using the Maestro interface (Maestro, version 9.9,

M AN U

Schrödinger, LLC, New York, NY, 2014) and submitted to mixed torsional/low-mode conformational sampling using MacroModel (MacroModel, version 10.5, Schrödinger, LLC, New York, NY, 2014). OPLS-2005 [28] was used as force field and GB/SA as solvation treatment [29]. Only global minima were retained from the conformational search and were used for the following

TE D

analysis. The 2 ketoprofen enantiomers were manually docked into βCD in both orientations and the resulting complexes, along with the unbound ketoprofen and βCD, were submitted to molecular dynamics studies. Simulation was run in explicit solvent, using the TIP4P water model [30] in a

EP

periodic boundary conditions orthorhombic box. Desmond Molecular Dynamics system (Desmond Molecular Dynamics System, version 3.9, D. E. Shaw Research, New York, NY, 2014. Maestro-

AC C

Desmond Interoperability Tools, version 3.9, Schrödinger, New York, NY, 2014) was used to set up and run the Molecular Dynamics (MD) simulations. The simulated environment was built using the system builder utility, with the structures being neutralized by Na+ and Cl- ions, which were added until a concentration of 0.15 M was reached. A series of minimizations and short MD simulations were carried out to relax the model system, by means of a relaxation protocol consisting of six stages: (i) minimization with the solute restrained; (ii) minimization without restraints; (iii) simulation (12 ps) in the NVT ensemble using a Berendsen thermostat (10 K) with non-hydrogen solute atoms restrained; (iv) simulation (12 ps) in the NPT ensemble using a Berendsen thermostat 7

ACCEPTED MANUSCRIPT (10 K) and a Berendsen barostat (1 atm) with non-hydrogen solute atoms restrained; (v) simulation (24 ps) in the NPT ensemble using a Berendsen thermostat (300 K) and a Berendsen barostat (1 atm) with non-hydrogen solute atoms restrained; (vi) unrestrained simulation (24 ps) in the NPT ensemble using a Berendsen thermostat (300 K) and a Berendsen barostat (1 atm). At this point, the

RI PT

systems were submitted to 5 ns (equilibration) + 50 ns (production) long MD simulations at a temperature of 300° K in the NPT ensemble using a Nose-Hoover chain thermostat and a MartynaTobias-Klein barostat (1.01325 bar). Energetics of the 50 ns MD trajectories was calculated with

SC

the MM-GB/SA approach [31] using macromodel (MacroModel, version 10.5, Schrödinger, LLC,

M AN U

New York, NY, 2014). Conformational entropies were estimated using the following equation:

Eq. 1

TE D

where kB is the Boltzmann constant and P is the probability (calculated from OPLS [30] potential energy) for each n conformer found during the MD trajectory, assuming a Boltzmann distribution.

EP

2.3. Microparticle preparation and characterization 2.3.1. Microparticle preparation

AC C

Microparticles were prepared using the solvent diffusion-evaporation method. βCD-ketoprofen inclusion complex was dispersed in the dichloromethane polymer (PLA or PLGA) solution and injected in a hydroxypropyl methylcellulose aqueous solution (2.5% w/v) maintained under stirring at 4 °C [17,32]. This emulsion (solid/oil/water) was then heated at 30 °C to allow the complete evaporation of dichloromethane and microparticle formation. After 3 hours at 30 °C, microparticle dispersion was cooled at 15 °C and filtered under nitrogen pressure through a 2.5 µm porosity filter (Whatman, United Kingdom), washed three times with 1L of ultrapure water and

8

ACCEPTED MANUSCRIPT dried under vacuum. Microparticles containing only CDs and ketoprofen were prepared in the same way. Each sample was prepared in triplicate.

2.3.2. Particle size and morphology characterization

RI PT

An Accusizer 770 (PPS Inc., Santa Barbara, CA, USA), equipped with a “Single Particle Optical Sensing” system, was used to estimate the mean particle diameter. Microparticles were suspended in 1 mL of deionized water and sonicated for 30 seconds before analysis. Analyses were performed

SC

in triplicate and the mean size was expressed as volume mean diameter (dmv) ± standard deviation. Microparticle external morphology was assessed by SEM. Samples were prepared as described

M AN U

above.

2.3.3. Determination of the ketoprofen content in the microparticles Ketoprofen content was determined by adding an amount of microparticles (exactly weighed) to 5

TE D

mL of acetonitrile. The sample was sonicated for 2 minutes and then 5 mL of deionized water were added. The solution, not perfectly clear, was filtered with polytetrafluoroethylene membrane filter (porosity 0.2 µm, Lida, Kenosha, WI, USA). Filtered solution was analyzed using UV-Vis

EP

spectrophotometry (λ, 254 nm) [33].

AC C

The encapsulation efficiency (EE) was calculated with the following equation.

Eq. 2

2.3.4. In vitro release studies In vitro release studies were carried out in centrifuge tubes containing 10 mL of 0.1 M phosphate buffer (pH 7.4) at 37 °C, ensuring the sink conditions. At predetermined times, samples were 9

ACCEPTED MANUSCRIPT centrifuged at 2500 rpm for 5 minutes and the supernatants were collected and replaced with fresh buffer. Samples were analyzed using UV-Vis spectrophotometry to determine the amount of ketoprofen released.

RI PT

3. Results 3.1. Preparation and characterization of inclusion complexes

The preparation method employed to produce inclusion complexes was selected for its easiness, low

SC

costs and good inclusion yield [26]. Quantitative analysis showed that 19.5 ± 8.4% (w/w) of the ketoprofen was included in βCD (βCD:ketoprofen molar ratio, 1:1). By using 2 moles of βCD

M AN U

and 1 of ketoprofen, an amount as high as 84.7 ± 1.7% of ketoprofen was included. Physical mixtures of the two components did not lead to the inclusion of ketoprofen in βCD. In fact, spectrophotometric analysis data showed that the whole amount of ketoprofen was just dispersed as powder and not included (data not shown). Figure 1a shows βCD crystalline powder with most of

TE D

crystals of about 100 µm and only few crystals of smaller size. Ketoprofen powder was composed of much smaller crystals (~ 10 µm) (Figure 1e). As expected, the 2 physical mixtures showed a different morphology (Figure 1b and 1f) with respect to the inclusion complexes (Figure 1c-d and

EP

1g-h). In the latter cases, it is not anymore possible to distinguish the βCD and ketoprofen crystals

AC C

still present in the physical mixtures.

βCD DSC data showed a broad endothermic event with the minimum around 130 °C related to the vaporization of crystallization water, generally estimated around 15% of the powder mass. Ketoprofen DSC data evidenced a very sharp endothermic peak around 95 °C previously ascribed to anhydrous crystals melting (Figure 2) [26]. Physical mixtures showed the overlapped endothermic events of both ketoprofen and βCD (Figure 2). The reduction of the endotherm peak intensity, ascribable to ketoprofen melting, was justified by its small amount in the blend. Inclusion complexes DSC data displayed a large endothermic event different from that of ketoprofen and βCD alone. However, part of the melting peak of ketoprofen is still visible in both complexes, 10

ACCEPTED MANUSCRIPT especially the 1:1 molar ratio complex (Figure 2), confirming the presence of an important amount of crystalline ketoprofen. The broadening of the endothermal event and the intensity reduction have been previously ascribed to the interaction between the two components [34]. DSC results were confirmed by the X-ray powder diffraction data (Figure 3). Diffraction patterns

RI PT

of the physical mixtures displayed the overlapping of ketoprofen and βCD patterns [35], while Xray powder diffraction data of the inclusion complexes showed a completely different pattern from that of the physical mixtures. These data allowed to assert that the formation of ketoprofen-βCD

indicated the presence of some free drug in the powder.

SC

complex occurred. Besides, the presence of low intensity reflections ascribed to ketoprofen crystals

M AN U

FT-IR data were used to confirm the inclusion complex formation. Ketoprofen and βCD FT-IR spectra had the characteristic absorption patterns reported for these compounds (Figure 1 supplementary material). Ketoprofen presents 2 interesting diagnostic absorptions at 1697 e 1655 cm-1, ascribed to carbonyl stretching vibrations of the carboxylic acid and of the ketone group of the

TE D

molecule arranged as a dimer [36-38]. The characteristic band of the carbonyl stretching of the acid group was not affected in the physical mixture (Figure 4), confirming the absence of interaction already observed from DSC and X-ray powder diffraction data. The FT-IR spectrum of the

EP

inclusion complex showed a shift of both peaks at higher wave number (1730 e 1667 cm -1, respectively). This effect is generally ascribable to hydrogen bond breaking of ketoprofen dimers

AC C

(following the formation of the inclusion complex) present in the crystal structure and to the formation of hydrogen bonds between a ketoprofen molecule and the hydroxyl groups of βCD [39,40].

3.2. Energetics of host-guest inclusion complex Racemic ketoprofen might form 4 different inclusion complexes with βCD. Both enantiomers (i.e., R-ketoprofen, S-ketoprofen) may be included in the βCD internal cavity with the carboxylic group interacting with the hydroxyl groups of the primary or secondary face. The 4 different inclusion 11

ACCEPTED MANUSCRIPT complexes have been named R-straight, R-reverse, S-straight, and S-reverse. Figure 5 shows the above mentioned inclusion complexes. Predicted ∆∆Gbinding (kcal/mol) for the 4 possible inclusion complexes, expressed as a difference

energies for R- and S-ketoprofen-βCD [41], are reported in Table 1.

RI PT

with the lowest value obtained (i.e., S-straight), together with the experimental binding free

The formation of the inclusion complex is mainly enthalpy-driven since entropy decreases in the complex respect to the unbound components (Figure 2 – supplementary material). In particular,

SC

the lowering of van der Waals (vdW) energies in the complexes reflects the interaction of the nonpolar portion of ketoprofen with the βCD hydrophobic cavity. On the other hand, to form the

M AN U

inclusion complex, ketoprofen and βCD pay an energy price in both bonded and solvation energies. These energy penalties are anyway rewarded by the vdW and electrostatic energies (Figure 3 – supplementary material). A significant part of electrostatic energy differences can be explained on the basis of the hydrogen bond occupancies (Table 2). Indeed, the most energetically favored

TE D

complexes are those that, during the molecular dynamics trajectories, make more hydrogen bonds between the βCD and ketoprofen carboxyl group.

EP

3.3 Polymer microparticle preparation and characterization Microparticle formulation parameters were preliminarily evaluated to investigate the effect of βCD

AC C

dispersion in the organic phase on particle characteristics. Spherical particles could be obtained by adding amount of CDs up to 1g, while the use of higher CD amounts generated non-spherical particles and aggregates (data not shown). The type (i.e., RG 202H, RG 502 H, RG 502) and amount of polymer, amount of dichloromethane and CDs had a non-negligible effect on particle size with the strongest influence of dichloromethane ratio (Table 1 - Supplementary material). By decreasing the amount of organic solvent (from 4.4 to 3.3 g), dmv increased of ~ 20 µm (Table 1 Supplementary material). The same trend was seen by increasing the amount of polymer (from 416 to 832 mg). This can be explained considering that the increase of polymer concentration 12

ACCEPTED MANUSCRIPT enhanced suspension viscosity. All formulations described above (Table 1 Supplementary material) were characterized by particles having a spherical shape and a smooth and non-porous surface (Figure 6 a). Microparticles containing the inclusion complex were prepared using the parameters previously

RI PT

described (Table 3). In analogy to what observed by embedding the sole βCD, particle dmv increased with the increase of polymer amount and with the decrease of methylene chloride volume. The polymer hydrophobic character (RG 502 H < RG 502 < RG 202 H) seems to affect the

SC

encapsulation efficiency: the higher the polymer hydrophobicity, the higher was the drug loading (Table 3; preparations 14, 15, and 16). Formulations 14, 15 and 16 showed the desired

M AN U

characteristics in terms of dmv and encapsulation efficiency and were further characterized [17]. SEM data did not show particular differences in external morphology between microparticles loaded with βCDs and those loaded with the inclusion complex (Figure 6). Release patterns from the 3 chosen formulations (i.e., 14, 15, and 16) showed substantial

TE D

differences primarily on the extent of the burst effects (14, 90% w/w; 15, 73% w/w; 16, 55% w/w). The high burst was surely due to the plasticizing effect of water [25] and ketoprofen [18] however the effect of βCD hydrophilicity has to be accounted as well [42]. Due to the presence of βCD that

EP

increases osmotic pressure within the matrix, polymer hydrophobicity (RG 502 H < RG 502 < RG 202 H) may plays a role during early stage micoparticle hydration and release.

AC C

DSC data of microparticles loaded with the inclusion complex (i.e., 14, 15, and 16) showed the absence of ketoprofen plasticization in the dry state. In fact, Tg values (14, 41.4±0.3 °C; 15, 41.6±0.5 °C; 16, 46.9±0.5 °C) did not differ significantly from the raw polymers.

13

ACCEPTED MANUSCRIPT 4. Discussion From the data herein reported, it is possible to conclude that the best inclusion complex was obtained by using 2:1 (βCD:ketoprofen) molar ratio albeit the whole ketoprofen was not fully complexed. The alternative of using higher amount of βCD to include the remaining ketoprofen was

RI PT

not investigated since the successive preparation of loaded microparticles was problematic. In fact, the inclusion complex powder has to be suspended in the organic phase containing the polymer and, the addition of an excessive amount of powder, leads to impediments during particle manufacturing

SC

(e.g., suspension injection, emulsion formation). High amount of powder forms non homogeneous microparticles and allows easy diffusion of the inclusion compound in the external aqueous phase

M AN U

with drug loss.

Polymeric microparticles containing ketoprofen were prepared using the solvent diffusionevaporation method [43], widely used in laboratory scale, conveniently modified for this specific application. In particular, the organic phase was not characterized by a homogeneous molecular

TE D

dispersion of drug and polymer, such as in the preparation of microparticles containing free ketoprofen, but was a homogeneous dispersion of the inclusion complex (βCD:ketoprofen, 2:1 molar ratio) in the polymeric solution. A multiple disperse system (solid-in-oil-in-water) was the

EP

starting point of composite microparticle production.

AC C

As hypothesized, the inclusion of ketoprofen molecules in the toroidal structure of βCD allowed the preparation of ketoprofen loaded microparticles with low-molecular-weight polyesters. Loaded microparticle DSC data consented to validate the hypothesis. Microparticles containing ketoprofenβCD inclusion complex had polymer Tg values superimposable to that of the raw materials, demonstrating that the inclusion in βCDs, by hindering the direct contact between ketoprofen and polymer chains, avoids plasticization during preparation. CDs have been used in pharmaceutical formulations mainly as solubilizers and/or stabilizers [44]. In fact, lipophilic molecules-CD inclusion complexes have higher apparent water solubility than the 14

ACCEPTED MANUSCRIPT free molecule [45]. This strategy has been used to increase the bioavailability of low-water soluble active pharmaceutical ingredients in different conventional forms [46,47]. CDs have been combined also with novel drug delivery systems (e.g., liposomes, microparticles for colon delivery) [48,49] or transformed, by ad hoc functionalization, in supramolecular devices for drug delivery [50,51]. Up

RI PT

to our best knowledge, this is the first report were βCD was employed to hinder the physical contact between a compound and a macromolecule to prevent or reduce plasticization.

Ketoprofen plasticization has important effects in the release kinetics. Previous studies evidenced

SC

that ketoprofen, loaded in PLA/PLGA microparticles, is mainly released by diffusion through the polymer matrix rather than by matrix degradation/erosion, thus polymer composition and molecular

M AN U

weight had a minor role in the duration of the release [17,18]. While plasticization is hindered during microparticle preparation and in dry microparticles by βCDs, the situation becomes more complex during the release. In vitro release data showed a high burst effect ascribed to ketoprofen plasticizing effect. In microparticles containing the inclusion complex, ketoprofen does not act as a

TE D

plasticizer when is included in βCDs but, once hydration takes place and ketoprofen is released from the inclusion complex, polymer plasticization and a fast diffusion occur. This reliable assumption is corroborated by molecular dynamics and previous data that showed a low energy of

EP

interaction in the complex [41]. In addition, once microparticle hydration takes place, the internal microenvironment becomes enough acid to completely protonated ketoprofen, a condition that

AC C

allows hydrogen binding with polymer backbone carboxyls. Due to the low interaction energy with the CD, its favorable interaction with the polymer chains, the osmotic pressure, and the concentration gradient, ketoprofen release is strongly favored. The effect of CDs on microparticle hydration and hydration rate should also play a role in the release kinetics. The interest of this finding is not limited to the preparation of microparticles by solvent diffusion evaporation but can be transferred to other preparation techniques, such as spray-drying, or to devices other than microparticles. Manufacturing methodologies that do not use water like emulsion technologies may use temperature to shape the polymer material (e.g., extrusion) [16] or to 15

ACCEPTED MANUSCRIPT evaporate the solvent. Since water (a strong plasticizer) or temperature are essential in these methodologies and can not be avoided, the hindering of unwanted plasticization due to ketoprofen or similar compounds makes feasible (or easier) procedures otherwise impossible.

RI PT

5. Conclusions

The inclusion of ketoprofen, a drug that acts as a polymer plasticizer, in the hydrophobic cavity of βCDs allowed to avoid the physical interactions between ketoprofen and polymer chains. This

SC

strategy allowed the preparation of ketoprofen-loaded microparticles using low-molecular-weight polymers, otherwise extremely challenging. The presence of βCDs in the polymer matrix affected

AC C

EP

TE D

a more hydrophobic polymer could reduce it.

M AN U

ketoprofen release from microparticles and, even though ketoprofen burst effect was not negligible,

16

ACCEPTED MANUSCRIPT References E.J. Frazza, E.E. Schmitt, A new absorbable suture, Biomed. Mater. Res. 5 (1971) 43–58.

[2]

H. Murakami, M. Kobayashi, H. Takeuchi, Y. Kawashima, Evaluation of poly(DL-lactide-coglycolide) nanoparticles as matrix material for direct compression, Adv. Powder Technology 11, (2000) 311-322.

[3]

R. Bodmeier, H.G. Chen, Evaluation of biodegradable poly(lactide) pellets prepared by direct compression, J. Pharm. Sci. 78 (1989) 819-822.

[4]

X. Zhang, U.P. Wyss, D. Pichora, B. Amsden, M.F.A. Goosen, Controlled release of albumin from biodegradable poly(DL-lactide) cylinders, J. Control. Release 25 (1993) 61-69.

[5]

J.W. Kostanski, B.C. Thanoo, P.P. DeLuca, Preparation, characterization and in vitro evaluation of 1- and 4-month controlled release orntide PLA and PLGA microspheres, Pharm. Dev. Technol. 5 (2000) 585-596.

[6]

M. Ricci, P. Blasi, S. Giovagnoli, L. Perioli, C. Vescovi, C. Rossi, Leucinostatin-A loaded nanospheres: characterization and in vivo toxicity and efficacy evaluation, Int. J. Pharm. 275 (2004) 61-72.

[7]

R.A. Jain, The manufacturing techniques of various drug loaded biodegradable poly(lactideco-glycolide) (PLGA) devices, Biomaterials 21 (2000) 2475-2490.

[8]

S. Giovagnoli, P. Blasi, G. Luca, F. Fallarino, M. Calvitti, F. Mancuso, M. Ricci, G. Basta, E. Becchetti, C. Rossi, R. Calafiore, Bioactive long-term release from biodegradable microspheres preserves implanted ALG-PLO-ALG microcapsules from in vivo response to purified alginate, Pharm. Res. 27 (2010) 285-295.

[9]

A. Atala, R. Lanza, R. Nerem, J.A. Thomson, Principles of Regenerative Medicine, second ed., Elsevier, New York, 2010.

EP

TE D

M AN U

SC

RI PT

[1]

AC C

[10] D.S. Kohane, R. Langer, Polymer biomaterials in tissue engineering, Pediatr. Res. 63 (2008) 487-491. [11] M.J.R. Virlan, D. Miricescu, A. Totan, M. Greabu, C. Tanase, C.M. Sabliov, C. Caruntu, B. Calenic, Current uses of poly(lactic-co-glycolic acid) in the dental field: a comprehensive review, J. Chem. 2015 (2015) article ID 525832. [12] O.J. Karlsson, J.M. Stubbs, L.E. Karlsson, D.C. Sundberg, Estimating diffusion coefficients for small molecules in polymers and polymer solutions, Polymer 42 (2001) 4915-4923. [13] S.P. Chamarthy, R. Pinal, Plasticizer concentration and the performance of a diffusioncontrolled polymeric drug delivery system, Coll. Surf. A: Physicochem. Eng. Aspects 331 (2008) 25-30. [14] G. Wypych, Introduction, in: G. Wypych (Ed.), Handbook of Plasticizers, ChemTec Publishing, Toronto - New York, 2004, pp. 1-5. 17

ACCEPTED MANUSCRIPT [15] E. Snejdrova, M. Dittrich, Pharmaceutically used plasticizers, in: M. Luqman (Ed.), Recent Advances in Plasticizers, InTech, Rijeka, Croatia, 2012, pp. 45-68. [16] C.L. Huang, T.W.J. Steele, E. Widjaja, F.Y.C. Boey, S.S. Venkatraman, J.S.C. Loo, The influence of additives in modulating drug delivery and degradation of PLGA thin films, NPG Asia Materials 5 (2013) e54.

RI PT

[17] M. Ricci, P. Blasi, S. Giovagnoli, C .Rossi, G. Macchiarulo, G. Luca, G. Basta, R. Calafiore, Ketoprofen controlled release from composite microcapsules for cell encapsulation: effect on post-transplant acute inflammation, J. Control. Release 107 (2005) 395-407.

SC

[18] P. Blasi, S. Giovagnoli, A. Schoubben, M. Ricci, C. Rossi, G. Luca, G. Basta, R. Calafiore, Preparation and in vitro and in vivo characterization of composite microcapsules for cell encapsulation, Int. J. Pharm. 324 (2006) 27-36.

M AN U

[19] C. Kunze, T. Freier, S. Kramer, K.P. Schmitz, Anti-inflammatory prodrugs as plasticizers for biodegradable implant materials based on poly (3-hydroxybutyrate), J. Mater. Sci. 13 (2002) 1051-1055. [20] P. Di Martino, E. Joiris, R. Gobetto, A. Masic, G.F. Palmieri, S. Martelli, Ketoprofenpoly(vinylpyrrolidone) physical interaction, J. Cryst. Growth 265 (2004) 302-308. [21] A. Fernández-Carballido, R. Herrero-Vanrell, I.T. Molina-Martınez, P. Pastoriza, Biodegradable ibuprofen-loaded PLGA microspheres for intraarticular administration effect of labrafil addition on release in vitro, Int. J. Pharm. 279 (2004) 33-41.

TE D

[22] C. De Brabander, G. Van Den Mooter, C. Vervaet, J.P. Remon, Characterization of ibuprofen as a nontraditional plasticizer of ethyl cellulose, J. Pharm. Sci. 91 (2002) 1678-1685.

EP

[23] C. Wu, J.W. McGinity, Non-traditional plasticization of polymeric films, Int. J. Pharm. 177 (1999) 15-27.

AC C

[24] K.L. Ngai, All standard theories and models of glass transition appear to be inadequate: missing some essential physics, in: S.J. Rzoska, V.A. Mazur (Eds), Soft Matter under Exogenic Impacts, Springer-Verlag GmbH, Berlin, Germany, 2007, p.p. 91-111. [25] P. Blasi, S.S. D’Souza, F. Selmin, P.P. DeLuca, Plasticizing effect of water on poly(lactideco-glycolide), J. Control. Release 108 (2005) 1-9. [26] P. Mura, M.T. Faucci, P.L. Parrini, S. Furlanetto, S. Pinzauti, Influence of the preparation method on the physicochemical properties of ketoprofen–cyclodextrin binary systems, Int. J. Pharm. 179 (1999) 117–128. [27] T. Van Hees, G. Piel, S.H. De Hassonville, B. Evrard, L. Delattre, Determination of the free/included piroxicam ratio in cyclodextrin complexes: comparison between UV spectrophotometry and differential scanning calorimetry, Eur. J. Pharm. Sci. 15 (2002) 347353.

18

ACCEPTED MANUSCRIPT [28] W.L. Jorgensen, D. S. Maxwell, J. Tirado-Rives, Development and testing of the OPLS allatom force field on conformational energetics and properties of organic liquids, J. Am. Chem. Soc. 118 (1996) 11225-11236. [29] W.C. Still, A. Tempczyk, R.C. Hawley, T. Hendrickson, Semi-analytical treatment of solvation for molecular mechanics and dynamics, J. Am. Chem. Soc. 112 (1990) 6127-6129.

RI PT

[30] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Klein, Comparison of simple potential functions for simulating liquid water, J. Chem. Phys. 79 (1983) 926-935. [31] J. Srinivasan, T.E. Cheatham, P. Cieplak, P.A. Kollman, D.A. Case, Continuum solvent studies of the stability of DNA, RNA, and phosphoramidate−DNA helices, J. Am. Chem. Soc. 120 (1998) 9401-9409.

M AN U

SC

[32] R. Bodmeier, R.W. McGinity, The preparation and evaluation of drug-containing poly(dllactide) microspheres formed by the solvent evaporation method, Pharm. Res. 4 (1987) 465471. [33] J.M. Rodrigues Júnior, K. de Melo Lima, C.E. de Matos Jensen, M.M. Gontijo de Aguiar, A. da Silva Cunha Júnior, The effect of cyclodextrins on the in vitro and in vivo properties of insulin-loaded poly (d,l-lactic-co-glycolic acid) microspheres, Artif. Organs 27 (2003) 492497. [34] K.H. Kim, M.J. Frank, N.L. Henderson, Applications of differential scanning calorimetry to the study of solid drug dispersions, J. Pharm. Sci. 74 (1985) 283-289.

TE D

[35] G. Bruni, A. Marini, V. Berbenni, R Riccardi, A novel method to obtain a β-cyclodextrin inclusion compound by solid state reaction: the ketoprofen case revisited, J. Incl. Phenom. Macro. 35 (1999) 517-530.

EP

[36] P. Blasi, A. Schoubben, S. Giovagnoli, L. Perioli, M. Ricci, C. Rossi, Ketoprofen poly(lactide-co-glycolide) physical interaction, AAPS PharmSciTech 8 (2007) Article 37.

AC C

[37] G.G. Liversidge, Ketoprofen, in: K. Florey (Ed.), Analytical Profiles of Drug Substances, Academic Press, New York, 1981, pp. 443-471. [38] P.P. Briard, J.C. Rossi, Kétoprofène, Acta Cryst. C46 (1990) 1036-1038. [39] P. Sancin, O. Caputo, C. Cavallari, N. Passerini, L. Rodriguez, M. Cini, A. Fini, Effects of ultrasound-assisted compaction on ketoprofen/eudragit® S100 mixtures, Eur. J. Pharm. Sci. 7 (1999) 207-213. [40] P. Mura, G.P. Bettinetti, A. Manderioli, M.T. Faucci, G. Bramanti, M. Sorrenti, Interactions of ketoprofen and ibuprofen with β-cyclodextrins in solution and in the solid state, Int. J. Pharm. 166 (1998) 189-203. [41] G. Marconi, E. Mezzina, I. Manet, F. Manoli, B. Zmabelli, S. Monti, Stereoselective interaction of ketoprofen enantiomers with β-cyclodextrin: ground state binding and photochemistry, Photochem. Photobiol. Sci. 10 (2011) 48-59. 19

ACCEPTED MANUSCRIPT [42] D. Bibby, N.M. Davies, I.G. Tucker, Mechanisms by which cyclodextrins modify drug release from polymeric drug delivery system, Int. J. Pharm. 197 (2000) 1-11. [43] S. Freitas, H.P. Merkle, B. Gander, Microencapsulation by solvent extraction/evaporation: reviewing the state of the art of microsphere preparation process technology, J. Control. Release 102 (2005) 313-332.

RI PT

[44] T. Loftsson, D. Duchêne, Cyclodextrins and their pharmaceutical applications, Int. J. Pharm. 329 (2007) 1-11. [45] S.V. Kurkov, T. Loftsson, Cyclodextrins, Int. J. Pharm. 453 (2013) 167-180.

SC

[46] V. Shukla, R. Masareddy, A. Anghore, F.V. Manvi, Influence of β-cyclodextrin complexation on ketoprofen release from matrix formulation, Int. J. Pharm. Sci. Drug Res. 1 (2009) 195202.

M AN U

[47] F.J. Otero-Espinar, J.J. Torres-Labandeira, C. Alvarez-Lorenzo, J. Blanco-Méndez, Cyclodextrins in drug delivery systems, J. Drug Del. Sci. Tech. 20 (2010) 289-301. [48] F. Maestrelli, N. Zerrouk, M. Cirri, N. Mennini, P. Mura, Microspheres for colonic delivery of ketoprofen-hydroxypropyl-β-cyclodextrin complex, Eur. J. Pharm. Sci. 34 (2008) 1-11. [49] F. Maestrelli, M.L. Gonzalez-Rodriguez, A.M. Rabasco, P. Mura, Preparation and characterization of liposomes encapsulating ketoprofen-cyclodextrin complexes for transdermal drug delivery, Int. J. Pharm. 298 (2005) 55-67.

TE D

[50] E. Memişoğlu, A. Bochot, M. Ozalp, M. Sen, D. Duchêne, A.A. Hincal, Direct formation of nanospheres from amphiphilic β-cyclodextrin inclusion complexes, Pharm. Res. 20 (2003) 117-125.

AC C

EP

[51] S. Daoud-Mahammed, C. Ringard-Lefebvre, N. Razzouq, V. Rosilio, B. Gillet, P. Couvreur, C. Amiel, R. Gref, Spontaneous association of hydrophobized dextran and poly-βcyclodextrin into nanoassemblies. Formation and interaction with a hydrophobic drug, J. Colloid Interface Sci. 307 (2007) 83-93.

20

ACCEPTED MANUSCRIPT

a

Predicted ∆∆Gbinding (kcal/mol)

R-reverse

±1.9

R-straight

±0.50

S-reverse

±1.03

S-straight

0

b

Experimental ∆Gbinding (kcal/mol)

-3.67

EP

TE D

M AN U

SC

-3.87

AC C

RI PT

Table 1. aPredicted ∆∆Gbinding (MM-GB/SA approach) for the 4 possible ketoprofen-β-cyclodextrin inclusion complexes. Values are expressed in kcal/mol as differences with respect to the lowest value obtained (S-straight). bExperimental free energies of binding for R- and S-ketoprofen-βcyclodextrin inclusion complexes according to Marconi et al [41].

ACCEPTED MANUSCRIPT Table 2. Hydrogen bond occupancies during the molecular dynamics trajectories.

βCD/water

ketoprofen-βCD

COOH/water

CO/water

COOH/βCD

CO/βCD

R-reverse

5.62

15.16

0.50

3.32

0.29

0.30

0.19

R-straight

5.44

14.91

0.73

3.26

0.32

0.58

0.15

S-reverse

5.57

14.67

0.48

3.23

0.32

0.34

0.13

S-straight

5.48

15.19

0.56

3.44

0.33

0.41

0.15

AC C

EP

TE D

M AN U

SC

RI PT

βCD/βCD

ACCEPTED MANUSCRIPT Table 3. Preparation parameters, mean size and encapsulation efficiency of microparticles loaded with the inclusion complex.

Mean volume

Preparation

Inclusion

Polymer

name

complex (mg)

Type/mg

7

924

502 H/416

4.4

8

462

502 H/832

4.4

9

462

502 H/832

2.2

10

924

202 H/832

3.3

11

924

502/832

4.0

12

924

502/624

13

924

502 H/832

14

924

502 H/416

15

924

502/416

16

924

202 H/416

CH2Cl2 (g)

Encapsulation efficiency (%)

RI PT

SD

60

12 ± 9

62

29 ±18

62

SC

19 ± 8

13 ± 21

63

14 ± 14

81

4.4

8±5

56

3.3

13 ± 7

81

3.3

9±5

58

3.3

12 ± 10

62

3.3

10 ± 7

74

M AN U

TE D EP

AC C

diameter (µm) ±

ACCEPTED MANUSCRIPT Figure captions

RI PT

Figure 1. Scanning electron microscopy photographs of β-cyclodextrin (a), βcyclodextrin:ketoprofen physical mixture (1:1 molar ratio) (b), β-cyclodextrin:ketoprofen inclusion complex (1:1 molar ratio) (c, d), ketoprofen (e), β-cyclodextrin:ketoprofen physical mixture (2:1 molar ratio) (f), β-cyclodextrin:ketoprofen inclusion complex (2:1 molar ratio) (g, h).

SC

Figure 2. Differential scanning calorimetry data of β-cyclodextrin, ketoprofen, physical mixtures, and inclusion complexes.

M AN U

Figure 3. X-ray powder diffraction data of β-cyclodextrin, ketoprofen, physical mixtures, and inclusion complexes.

Figure 4. Fourier transform-infrared spectroscopy data of ketoprofen, physical mixture, and inclusion complex in the spectral range 1600-1800 cm-1.

TE D

Figure 5. The possible inclusion complexes formed by β-cyclodextrin with R- and S-ketoprofen. βcyclodextrin is depicted in green sticks whereas the two ketoprofen enantiomers are depicted in yellow sticks.

AC C

EP

Figure 6. Scanning electron microscopy photographs of β-cyclodextrin loaded microparticles (sample 6, Table 1_supplementary material) (a), and inclusion complex loaded microparticles of formulations 14 (b), 15 (c), and 16 (d) (Table 3).

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Table 1_supplementary material. Preparation parameters and mean size of microparticles loaded with β-cyclodextrin.

Polymer β-cyclodextrin (mg)

CH2Cl2 (g)

Mean volume diameter (µm) ± SD

9±8

RI PT

Preparation name

Type/mg 1

832

502 H/416

4.4

2

832

502/416

4.4

3

832

502 H/416

3.3

4

832

502/416

2.2

5

416

502 H/416

2.2

13 ± 7

6

416

502 H/832

2.2

28 ± 18

SC

M AN U

TE D EP AC C

8±4

31 ± 23 22 ± 11

ACCEPTED MANUSCRIPT Figure captions_supplementary material Figure 1_SM. Fourier-transform infrared spectroscopy spectra of ketoprofen and β-cyclodextrin in the range 4000-700 cm-1.

RI PT

Figure 2_SM. Entropic (a) and enthalpic (b) energies estimations of unbound β-cyclodextrin summed to R- and S-ketoprofen (first two values respectively) and of the inclusion complexes (last four values). Error bars in panel b depict the standard deviations from the averaged values.

AC C

EP

TE D

M AN U

SC

Figure 3_SM. Van der Waals (a), stretching+bending+torsional (b), solvation (c) and electrostatic (d) energies averaged over 50 ns long molecular dynamics trajectories. Error bars depict the standard deviations from the averaged values. The first two values in each panel refer to the energy sums of unbound β-cyclodextrin and R- and S-ketoprofen, respectively. The following four values refer to the inclusion complexes.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT