modulated release of cyclosporine from soluble vinyl pyrrolidone–hydroxyethyl methacrylate...

Journal of Controlled Release 72 (2001) 1–11www.elsevier.com/ locate / jconrel

Modulated release of cyclosporine from soluble vinylpyrrolidone–hydroxyethyl methacrylate copolymer hydrogels

A correlation of ‘in vitro’ and ‘in vivo’ experimentsa a b,dÁlberto Gallardo , Fernando Fernandez , Alejandro Cifuentes ,

b c c´ ´Jose-Carlos Dıez-Masa , Paloma Bermejo , Mercedes Rebuelta ,a a ,*´ Ántonio Lopez-Bravo , Julio San Roman

a ´ Ínstituto de Ciencia y Tecnologıa de Polımeros, CSIC, Juan de la Cierva 3, 28006 Madrid, Spainb ´ Ínstituto de Quımica Organica, CSIC, Juan de la Cierva 3, 28006 Madrid, Spain

c ´Departamento de Farmacologıa, Facultad de Farmacia, Universidad Complutense, Madrid, SpaindInstituto de Fermentaciones Industriales, CSIC, Juan de la Cierva 4, 28006 Madrid, Spain

Received 12 April 2000; accepted 11 January 2001

Abstract

Soluble, uncrosslinked and high molecular weight copolymers of vinylpyrrolidone, VP, with 2-hydroxyethyl methacrylate,HEMA, prepared by free radical copolymerization, are proposed as supports for the modulated release of the immuno-suppressor cyclosporine. Two copolymeric systems with copolymer compositions f 50.52 (namely VP–HEMA 60–40) andVP

0.42 (VP–HEMA 40–60) have been prepared and tested in vitro and in vivo using rats as animal model. Micellarelectrokinetic capillary chromatography, MEKC, has been used for the simultaneous detection of the polymer reabsorptionand the drug release for the in vitro experiments. The composition and microstructural distribution of the copolymer systemcontrols the solubilization rate which modulates the in vitro release of the drug (with time profiles from a few days to severalweeks for the VP–HEMA 60–40 and 40–60, respectively) and the in vivo response that correlates with the previous in vitroresults: the more hydrophobic implant (VP–HEMA 40–60) reverts the immune response more slowly (2–4 weeks)compared to the more hydrophilic one (VP–HEMA 60–40, 1–2 weeks). 2001 Elsevier Science B.V. All rights reserved.

Keywords: Vinylpyrrolidone; 2-Hydroxyethyl methacrylate; Copolymers; Cyclosporine

1. Introduction the effects after oral administration depend on thevariable individual absorption or even on the diet of

The use of cyclosporine in immune suppression the patients.therapies is a common practice in modern medicine We have designed recently bioresorbable andand surgery, but there are secondary effects of great biocompatible non-crosslinked vinyl-pyrrolidoneimportance at medium and long term [1]. In addition (VP)–hydroxyethyl methacrylate (HEMA) copoly-

mers, which offer an interesting modulation of thecontrolled release of this and other drugs. The*Corresponding author.

É-mail address: [email protected] (J. San Roman). microstructural characteristics of this copolymer

0168-3659/01/$ – see front matter 2001 Elsevier Science B.V. All rights reserved.PI I : S0168-3659( 01 )00257-7

2 A. Gallardo et al. / Journal of Controlled Release 72 (2001) 1 –11

system correspond to a distribution of rich HEMA washed with the precipitant mixture and dried underchains and blocks of vinyl pyrrolidone units, depend- vacuum until constant weight was attained.ing on the initial composition of the medium used forthe preparation of the copolymer systems [2]. These 2.3. Characterizationmicrostructures provide an excellent control of thesolubilization of these materials, with a rate reg- Copolymer samples prepared with different com-ulated by the length and distribution of the VP- positions of VP and HEMA monomeric components

1blocks, which allows the design of system with fast were analysed by H NMR spectroscopy. The aver-or slow solubilization. The release of drugs in age composition of the systems was determined bygeneral and of cyclosporine in particular, is modu- comparison of the integrated intensities of charac-lated clearly by the solubilization rate and the teristic signals assigned to VP and HEMAcomposition of the VP–HEMA copolymer systems comonomeric units as described elsewhere [2]. The[3]. average molecular weight was determined by SEC in

In this paper we present some correlations of the solutions of dimethylformamide / lithium bromide.in vitro controlled release of cyclosporine and the in Details of the experimental conditions are givenvivo study of the behaviour of cotton-coated meshes elsewhere [2].containing films of cyclosporine, VP–HEMA co-polymer systems, implanted in the dorsal muscle of 2.4. Disc preparationrats for periods up to 4 weeks.

Loaded films were prepared over a cotton mesh of2.2 cm diameter (as support) by casting from a

2. Experimental water /ethanol (1 /1) solution. Polymer (120 mg) and30 mg of cyclosporine (20 wt.%) were used for all

2.1. Chemicals the compositions. The cotton mesh is used as supportof the polymeric matrix as well as immunogenic

Cyclosporine was a kind gift from Novartis Inc. factor as has been proposed in the literature [5].(Basel, Switzerland). HEMA (purchased from Fluka)was purified according to the literature [4]. VP 2.5. In vitro drug release(supplied by Fluka) was distilled under reducedpressure and used without further purification. 2,29- The in vitro experiments were carried out asAzobisisobutyronitrile, AIBN, was purified by frac- described in a previous paper [3], using micellartional crystallisation from methanol (m.p. 1048C). electrokinetic capillary chromatography, MEKC, asOther reagents were of extra pure grade and used as novel technique for the simultaneous detection ofpurchased. both the released drug and the reabsorbed polymer

[3]. Basically, the copolymer discs were settled2.2. Copolymerization between a teflon ring and a tap of a special incubat-

ing cell and immersed in saline solution with 10%Copolymerization reactions were performed in Tween 80; the cell was maintained at 378C in a

water /ethanol solutions at 5060.18C in Pyrex glass thermostated bath. At appropriate times, samplesampoules. Reactions were carried out in the absence were collected for analysis by MEKC (1 ml) andof oxygen by bubbling nitrogen twice for 30 min replaced by fresh medium (the same volume).before sealing the system. The monomer and initiator

22 21concentrations were 1.0 and 1.5310 mol l , 2.6. Animals and implantation procedurerespectively. The sealed ampoules were immersed ina water bath maintained at the polymerisation tem- Thirty male Wistar rats from 125 to 150 g weightperature. At 48 h, the ampoules were removed from were used. The surgical area was previously shavedthe bath and the contents were poured into a large and disinfected and the instrumentation sterilised.excess of acetone. The precipitated samples were After anaesthetising the rats with diethyl ether, an

A. Gallardo et al. / Journal of Controlled Release 72 (2001) 1 –11 3

incision between the omoplates and in the epithelial of the inflammatory reaction with the time togetherand connective tissues was done in order to access with the microscopical analysis of the healing pro-the dorsal muscle. The discs were placed over this cess using an optical microscope Zeiss MC63Amuscular tissue and the incision was closed with coupled to photographic equipment.sterilised suture. Finally the area was disinfected andprotected with a special device necklet-like to protectthe injury from the other animals.

3. Results and discussion2.7. Implant retrieval

Poly(vinyl pyrrolidone-co-hydroxyethyl methac-At appropriate times, the animals were anaesthet- rylate), VP–HEMA, is a well known biocompatible

ised and sacrificed by cervical dislocation. An inci- hydrogel with broad applications in the biomedicalsion in the implant area was done and the remaining area [6–8], mainly as crosslinked network. Probably,samples and surrounding tissue were collected and the most characteristic use is as contact lens supportstored in formaldehyde solution till analysis. material. Because of its biocompatibility and hydro-

philic nature, it has been proposed as an interesting2.8. Sampling carrier for drug delivery [9]. However, crosslinked

matrices are not resorbable. In this sense, we proposeThe animals were randomly assigned to four the use of not crosslinked VP–HEMA copolymers as

groups: blank (only the surgical procedure was an entirely soluble delivery system for the release ofperformed and no disc was implanted), control (in cyclosporine.this group discs with copolymers but without drug The VP–HEMA copolymeric system exhibits aloading were implanted) and two problem groups well-described different reactivity of the(the complete discs, cotton plus the two types of comonomeric pair, which provides copolymers withloaded polymers, VP–HEMA 60–40 and 40–60, a broad microstructural distribution and, in a limitingwere implanted). Samples were collected at 2, 7, 14 case, the formation of block copolymers or poly-and 28 days (n53 for each time in the problem meric systems with a predominant structure of thegroups, n52 in the control and n51 in the blank parent monomeric units [10–12]. This characteristicone). has been used positively for the preparation of

systems with a time-dependent solubilization, con-2.9. Histological analysis trolled by the microstructural distribution of co-

polymer chains prepared at high conversion fromThe remaining samples and surrounding tissue vinyl and acrylic monomers with very different

were analysed histologically by the fixation of cross- chemical reactivity.sections of the remaining implant and tissue with Two different copolymer systems prepared at highformaldehyde, stained with haematoxylin and eosin. conversion loaded with 20 wt.% of cyclosporineThe effects produced by the active discs implanted were used for the preparation of coated cottonwere qualitatively analysed following the evolution meshes. The average composition and molecular

Table 1Structural composition and number average molecular weight, M of the copolymer systems prepared at high conversionn

a 3Copolymer Average composition PVP M 3 10 Conversionn

system VP–HEMA mol.% mol.% wt.%

VP(60–40) 52–48% 30% 36 96VP(40–60) 42–58% 22% 40 95

a Statistical percentage of PVP long segments and homopolymer as a fraction of the VP-average composition quoted in the secondcolumn.


positions in vitro, following by kinetic electropho-resis the cumulative concentration of released cyclo-sporine in buffered solutions [3].

3.1. In vitro results

Fig. 1 shows the in vitro release profile of thecyclosporine for the system VP(60–40) and Fig. 2the data obtained in the same experimental con-ditions for the sample VP(40–60). The time-depen-dent concentration of cyclosporine in the solutiondepends on the composition of the correspondingcopolymer system. The VP–HEMA 60–40 systemFig. 1. In vitro cumulative release of cyclosporine for the VP–exhibits a higher burst effect which is related to theHEMA 60–40 system.hydrosoluble PVP content and in a second step acontrolled release during several days which isintimately related to the solubilization mechanism of

characteristics of the copolymer systems are quoted the residual HEMA-rich copolymer, which wasin Table 1. described in a previous paper [3]. The more hydro-

The VP(60–40) copolymer system gave a com- phobic 40–60 system, exhibits a more sustainedposition of 52 mol.% of VP, determined by NMR release during at least 4 weeks.spectroscopy whereas the system VP(40–60) gave avalue of 42 mol.% of VP units. Reactivity ratio 3.2. In vivo resultsdetermination (r 57.97 and r 50.08) and theHEMA VP

kinetic analysis of the reaction with conversion have The blank group, which was operated and manipu-been described previously [2]. According to the lated in the same conditions as the rest of the animalsstatistical calculations described in this previous but without implanting any material, did not showpaper, it can be considered that both systems con- any inflammation or immune response at the timestained 30 and 22% of PVP or long VP blocks. These described in Section 2 (up 1 month). Therefore, westructures are responsible for a controlled modulation can conclude that the applied technique is notof the release of cyclosporine which was demon- interfering in the in vivo process and that the surgicalstrated for copolymer samples with different com- procedure is not immunogenic itself.

The control group (to which unloaded discs,cotton plus polymers, were implanted) exhibited aclear inflammation and immune response. The his-tological analysis of cross-sections of the dorsalmuscle tissue retrieved after 2 days of implantation,Fig. 3, revealed an intense inflammatory reaction,with the presence of a great number of leukocytes,some granulomes associated probably to the stimula-tion of an encapsulation process, and oedema. Thisbehaviour appears also reflected in the micrographsof histological analysis at 1, 2 and 4 weeks ofimplantation (Fig. 3). This response validates the useof the cotton support as immunogenic model.

The discs based on the copolymer VP–HEMA60–40 (the more hydrophilic) shows at 2 days (Fig.Fig. 2. In vitro cumulative release of cyclosporine for the VP–

HEMA 40–60 system. 4) an aggressive immune reaction against the im-

A.

Gallardo

etal.

/Journal

ofC

ontrolledR

elease72

(2001)1

–115

Fig. 3. Microphotographs of the surrounding tissue of the control group at 2 (A) and 14 (B) days. →, Lymphocytes and inflammation; ←, oedema; ∧, granulation tissue.Original magnification 316, hematoxylin and eosin staining.


Fig. 4. Microphotographs corresponding to the VP–HEMA 60–40 implant at 2 days. →, Lymphocytes with inflammation; ←, oedema; .,fibrosis formation; ↓, angioproliferation; ,, microphocus of necrosis. Original magnification 316 (A and B) or 340 (C and D, D being adetail of the microphotograph B), hematoxylin and eosin staining.

plant, with presence of lymphocytes, oedema, an- 28 days (Fig. 9) most of the tissue is clean ofgioproliferation and a few microphocus of necrosis. inflammatory reaction which has to be related to theHowever, there are also fibrous areas of cicatrizial effect of the cyclosporine release.origin that could be related with the release ofcyclosporine. The effect of the drug is more clear at 3.3. In vitro–in vivo correlation7 and 14 days (Figs. 5 and 6, respectively), wherethe presence of healing tissue and areas of tissue There is a clear in vitro–in vivo correlation: thewithout any inflammation indicates that the system is more hydrophilic 60–40 copolymer system exhibitsable to revert the immune response. the faster response, with a burst effect in the in vitro

The slightly more hydrophobic implant based on experiments related with the high content in thethe VP–HEMA 40–60 copolymer system, shows a VP-rich component, followed by a controlled releaseslower response. At 7 days (Fig. 7) there are broad up to 10 days. On the other hand, the in vitro profilebands of inflammatory tissue, and lymphocytes are of the other system, the more hydrophobic 40–60,observed. However, at 2 weeks this immune reaction shows a small burst effect (because the amount ofhas been reverted to a great extent, although there VP-rich component has decreased, see Table 1) andare still broad areas of scar tissue as a consequence a sustained release up to several weeks. The in vivoof the intense previous immune reaction (Fig. 8). At results are in agreement with these in vitro release

A.

Gallardo

etal.

/Journal

ofC

ontrolledR

elease72

(2001)1

–117

Fig. 5. Microphotographs corresponding to the VP–HEMA 60–40 implant at 7 days. →, Infiltration band; ←, oedema. Original magnification 316, hematoxylin and eosinstaining.


Fig. 6. Microphotograph corresponding to the VP–HEMA 60–40 implant at 14 days. ., Incipient cicatrizial fibrosis; ←, oedema.Magnification 340, hematoxylin and eosin staining.

Fig. 7. Microphotograph corresponding the the VP–HEMA 40–60 implant at 7 days. ↓, Hemorrhage; →, inflammatory reaction; ←,oedema. Original magnification 316, hematoxylin and eosin staining.


Fig. 8. Microphotographs corresponding the VP–HEMA 40–60 implant at 14 days. ., Scar tissue; →, isolated inflammatory elements.Original magnification 316, hematoxylin and eosin staining.

10A

.G

allardoet

al./

Journalof

Controlled

Release

72(2001)

1–11

Fig. 9. Microphotographs corresponding to the VP–HEMA 40–60 implant at 28 days. ., Scar residues. Original magnification 316, hematoxylin and eosin staining.


´Dıez-Masa, Micellar electrokinetic chromatography appliedprofiles because the 60–40 system reverts the inflam-to copolymer systems with heterogeneous distribution,mation and the immune reaction in 1 week, while theMacromolecules 32 (1998) 610–617.

40–60 system achieves that in 2–4 weeks.´[3] A. Gallardo, F. Fernandez, P. Bermejo, M. Rebuelta, A.

As a conclusion, the histological analysis of ´ ´Cifuentes, J.C. Dıez-Masa, J. San Roman, Controlled releaseimplanted cyclosporine devices demonstrates that the of cyclosporine from VP–HEMA copolymer systems of

adjustable resorption monitorized by MEKC, Biomaterialsuncrosslinked VP–HEMA copolymers provide ex-21 (9) (2000) 915–921.cellent soluble and biocompatible systems for the

[4] R.J. Fort, T.M. Polyzoidis, Intrinsic viscosity–molecularcontrolled release of the immunosuppressor com-weight relationship for poly-2-hydroxyethyl methacrylate,

pound. It seems that a combination of copolymer Eur. Polym. J. 12 (1976) 685–689.systems prepared with different compositions would [5] W.C. Bowman, M.J. Rand, in: 2nd Edition, Textbook ofgive an optimized result with an adequate response at Pharmacology, Blackwell, 1980.

[6] M. Akashi, K. Takemoto, New aspects of polymer drugs,short and long times. These formulations could beAdv. Polym. Sci. 97 (1990) 107–146.specially interesting in the treatment of autoimmune

[7] C.L. Bell, N.A. Peppas, Biomedical membranes from hydro-cutaneous illness, as well as transdermal devices forgels and interpolymer complexes, Adv. Polym. Sci. 122

the controlled dosification of cyclosporine. (1995) 125–176.[8] R.J. Laporte, Hydrophilic Polymer Coatings for Medical

Devices, Technomic, Lancaster, 1997.´ ´[9] M.D. Blanco, R.M. Trigo, O. Garcıa, J.M. Teijon, ControlledAcknowledgements

release of cytarabine from poly(2-hydroxyethylmethacrylate-co-N-vinyl-2-pyrrolidone) hydrogels, J. Bioma-´Support from the Comision Interministerial deter. Sci. Polym. Edn. 8 (1997) 709–719.

´Ciencia y Tecnologıa, Grant MAT98-0964 and from [10] A.E. Hamielec, J.F. MacGregor, A. Pendilis, in: G.C.´the Comunidad Autonoma de Madrid CAM-8.4 / Eastmond, A. Ledwith, S. Russo, P. Sigwalt (Eds.), Co-

polymerization, Comprehensive Polymer Science, Vol. 3,0007/97 are gratefully acknowledged.Pergamon Press, Oxford, 1989, p. 17.

[11] F.L.M. Hautus, H.N. Linssen, A.L. German, Dependence ofcomputed copolymer reactivity ratios on the calculation

References method I. Effect of experimental setup, J. Polym. Sci. Polym.Chem. Ed. 22 (1984) 3487–3498.

[1] M.M.M. Hassan, M.A. Al-Yahya, in: K. Florey (Ed.), [12] A.D. Jenkins, in: S.L. Aggarwal, S. Russo (Eds.), Com-Analytical Profiles of Drug Substances, Vol. 16, Academic prehensive Polymer Science Second Supplement, ReactivityPress, London, 1987, pp. 146–206. in Radical Copolymerization, Pergamon Press, Oxford, 1996,

´[2] A. Gallardo, A.R. Lemus, J. San Roman, A. Cifuentes, J.C. p. 1.

modulated release of cyclosporine from soluble vinyl pyrrolidone–hydroxyethyl methacrylate...

Documents