blends of synthetic and natural polymers as drug delivery systems for growth hormone

Biomaterials 16 (1995) 569474 0 1995 Elsevier Science Limited

Printed in Great Britain. All rights reserved 0142-9612/95/$10.00

Blends of synthetic and natural polymers as drug delivery systems for growth hormone

Maria Grazia Cascone, Bushra Sim* and Sandra Dowries” Department of Chemical Engineering, University of Pisa, Via Diotisalvi 2, 56723 Pisa, Italy; ‘The institute of Orthopaedics (XL), Brockley Hill, Stanmore, Middlesex NA7 4LP, UK

In order to overcome the biological deficiencies of synthetic polymers and to enhance the mechanical

characteristics of natural polymers, two synthetic polymers, poly(vinyl alcohol) (PVA) and poly(acrylic

acid) (PAA) were blended, in different ratios, with two biological polymers, collagen (C) and hyaluronic

acid (HA). These blends were used to prepare films, sponges and hydrogels which were loaded with

growth hormone (GH) to investigate their potential use as drug delivery systems. The GH release was

monitored in vitro using a specific enzyme-linked immunosorbent assay. The results show that GH can

be released from HA/PAA sponges and from HA/PVA and C/PVA hydrogels. The initial GH concentra-

tion used for sample loading affected the total quantity of GH released but not the pattern of release.

The rate and quantity of GH released was significantly dependent on the HA or C content of the

polymers.

Keywords: Drug delivery, growth hormone, collagen, hyaluronic acid

Received 12 November 1993; accepted 20 June 1994

Polymeric materials for use in biomedical applications should possess the appropriate physico-chemical and mechanical properties, and be suitable for fabrication into biomedical devices. Synthetic polymers are used in various areas of medicine, such as the cardiovascu- lar system (Dacron@ and Teflon@), orthopaedics (polyethylene, polypropylene and poly(methy1 methacrylate)), dentistry, ophthalmology (poly(methy1 methacrylate) and silicon polymers), artificial organs (cellulose and polyacrylonitrile) and drug delivery systems. The required mechanical properties of the polymers vary according to the specific application, for example, in the vascular field compliance matching of implant and host tissue is important whereas in orthopaedics (e.g. in polyethylene acetabular components) resistance to flexing, compression and abrasion is important. Many synthetic polymers that are already commercially available show physicochem- ical and mechanical properties comparable to those of the biological tissue that they are required to substi- tute, but are not sufficiently biocompatible. Whereas many biological polymers possess good biocompatibility, their mechanical properties are often inadequate. Improvements in the characteristics of synthetic biomaterials could be achieved by the addition of biological macromolecules such as fibrin, collagen, elastin and glycosaminoglycans. The resulting materials could combine the appropriate mechanical proper-

Correspondence to Dr S. Downes.

ties of the synthetic component with the biocompatibility of the biological componentlz2. These materials could be used in implantable devices to release biologically active substances in a controlled manner. Such additives could be growth factors, antibiotics or other therapeutic agents. Two different water-soluble synthetic polymers, poly(viny1 alcohol (PVA) and poly(acrylic acid) (PAA) were blended in different ratios with two different biological polymers, collagen and hyaluronic acid3. Each polymer component has advantageous properties. Collagen is widely distributed in living tissue, has low immuno- genicity and allows cell adherence4z5. Hyaluronic acid is able to influence cell adhesion, migration, aggrega- tion and proliferation6. In this work blends of collagen and hyaluronic acid with synthetic polymers were prepared according to the method of Giusti et ~1.~. The polymers were loaded with different concentrations of growth hormone (GH) and made into hydrogels, films and sponges. Using in vitro experiments the ability of these polymer systems to release GH was investigated and the release profile for each system was established.

MATERIALS AND METHODS

Growth hormone solution

In this study recombinant human growth hormone was used. A solution of 0.4 IU cm-3 GH was made by adding 4 units of GH powder (Serono, Italy) to 10 cmm3 0.99%

569 Biomaterials 1995, Vol. 16 No. 7

570 Drug delivery systems for growth hormone: M.G. C&scone et al.

sodium chloride solution. This solution was used to incorporate GH into the biomaterials used in this study.

Collagen/poly(vinyl alcohol) films

Poly(viny1 alcohol) (Aldrich, Dorset, UK) was used to make the polymers, the average molecular weight determined by viscometry was 114 000, with 100% hydrolysis and degree of syndiotacticity of 62% (determined by infrared analysis). Collagen (1 g) (Collagen Type III from calf skin, Sigma, Dorset, UK) was dissolved in 100 cmm3 of 0.5 M acetic acid in an ice bath, to obtain a final collagen solution of 1%. A 3% solution of PVA in distilled water was made, this was dissolved for 1 h at 120°C. These two solutions were blended in three different CIPVA ratios: 30170, 50150 and 70130; each blend was dispensed in three wells of a six-well plate (10 cm3 well-‘). One well was used as a control; to the other two wells GH was added in concentrations of 50 and 75 mIU GH per cm” collagen. Films were obtained by using the ‘casting solution’ method; the films were cast onto polystyrene and the solvent allowed to evaporate inside a cabinet flow hood. The films were then cross-linked by placing them in a desiccating chamber with 8% glutaraldehyde (GTA) vapour for 18 h at 37°C.

Hyaluronic acid/poly(acrylic acid) sponges

Solutions of 5% hyaluronic acid (HA) (supplied as sodium salt by Fidia Advanced Biopolymers SpA, Italy) and 5% PAA with molecular weight 250 000 (Aldrich) were prepared by dissolving in water at 50” C. The two solutions were blended in three different HAIPAA ratios: 20180, 40160 and 60140; each blend was dispensed in five wells of a 24-well plate (1.2 cm3 per well). The samples were then lyophilized. After cross-linking by thermal treatment at 130” C under vacuum for 24 h, GH was added. One sponge was used as a control and GH was added to the other four to produce final concentrations of 25, 50, 75 and 100 mIU GH per cm3 HA. The samples were lyophilized again and the release of GH was monitored in vitro.

Collagen/poly(vinyl alcohol) hydrogels

Collagen (1.5 g) was dissolved in 100 cm3 0.5 M acetic acid, in an ice bath, to obtain a final collagen solution of 1.5%. PVA (10 g) was added to 100 cm3 distilled water and dissolved in an autoclave for 1 h at 120°C to obtain a final concentration of 10% PVA. The two solutions were blended in three different CIPVA ratios: 30170, 20180 and 10190; each blend was dispensed in five wells of a 24-well plate (1.2 cm3 per well). One well was used as a control, and GH was added to the other four to produce final concentrations of 25, 511, 75 and 100 mIU GH per cm3 collagen. After GH addition, samples underwent eight cycles of freeze-thawing to obtain hydrogels. Each cycle, with the exception of the first one, consisted of 1 h at -20°C and 30 min at room temperature. The first cycle differed from the others due to a longer standing time at -20°C.

Hyaluronic acid/poly(vinyl alcohol) hydrogels

A 5% HA solution was made in distilled water at 50 C. PVA (5 g) was added to 100 cm3 distilled water and dissolved in an autoclave for 1 h at 120°C to obtain a final concentration of 5% PVA. The two solutions were blended in three different HAIPVA ratios: 30170, 20180 and 10190; each blend was dispensed in five wells of a 24-well plate (1.2 cm3 per well). One well was used as a control, and GH was added to the other four to produce final concentrations of 25, 50, 75 and 100 mIU GH per cm3 HA. After GH addition, samples underwent eight cycles of freeze-thawing to obtain hydrogels. Each cycle, with the exception of the first one, consisted of 1 h at -20’ C and 30 min at room temperature. The first cycle differed from the others due to a longer standing time at -20°C.

Elution studies

The GH eluates from the samples were monitored in vitro. The hydrogels and sponges were each placed in 3 cm” phosphate-buffered saline (PBS) while the films were placed in 4 cm” PBS in individual containers at 37’C. The elution fluid was removed at regular time intervals (every day for 7 d and then every 2 d for a further 7 d) and stored at -20 C. The eluate was replaced with PBS and the containers returned to 37’ C.

GH assay

Elution fluids were assayed for the GH using an enzyme-linked immunosorbent assay (ELISA), as previously described8. The microtitre plates were read using a 96-well fluorescent plate reader (MR 700 Dynatech Microplate Reader, West Sussex, UK). Optical density was measured at 490 nm with a reference wavelength of 650 nm. The standard used was 22K rhGh (recombinant human growth hormone, Novo-Nordisk, Gentofte, Denmark).

Scanning electron microscopy

Scanning electron microscopy studies of the surface and internal structures of the sponges and the hydrogels were made. The materials were dehydrated through an increasing series of alcohol solutions, critically point dried against carbon dioxide, sputter coated with gold and observed using a Joel T 300 scanning electron microscope.

RESULTS

Collagen/poly(vinyl alcohol) films

The eluants from the C/PVA films were analysed for GH using ELISA, but there was no detectable GH release.

Hyaluronic acid/poly(acrylic acid) sponges

Plots of the rates of GH release from samples prepared using the three different HAIPAA blends were linear. Figure 1 a shows the release of GH from the three blends of HAIPAA sponges loaded with 75 mIU GH

Biomaterials 1995, Vol. 16 No. 7

Drug delivery systems for growth hormone: M.G. &scone et al. 571

1 2 3 4

Square root time (days)

a 0.5

b


Figure 1 Release of growth hormone (GH) from hyaluronic acid/poly(acrylic acid) (HAIPAA) sponges loaded with: a, 75 mlU GH per cm3 HA; b, 50 mlU GH per cm3 HA. q , HAI PAA = 20180; +, HAIPAA = 40160; n , HAIPAA = 60140.

per cm3 HA and Figure lb shows the same blends loaded with 50 mIU GH per cm3 HA. The initial GH concentration did not affect the pattern of release, but increasing the GH concentration in the loading solution increased the total amount released. There was significantly less GH released from samples prepared using the 20180 (HAIPAA) blend than from samples prepared using the 40/60 (HA/PAA) blend (P < 0.05 paired t-test) and in turn the samples prepared using the 60140 (HA/PAA) blend (P < 0.05 paired t-test). Figure 2 shows the release curves for GH from sponges with the same HA content (60%) but differing GH concentrations and as expected there was more GH released from the sponges with higher initial GH concentrations.

Collagen/poly(vinyl alcohol) hydrogels

GH was released from the CIPVA hydrogels with a slow lag phase during the first 3 d followed by a

faster burst of GH release. The initial GH concentration did not affect the pattern of release but did, however, affect the total amount released. There was a direct relationship between the amount of GH incorporated in the hydrogels and the total amount of GH released. Figure 3 shows the release curves for GH from hydrogels with the same collagen content (20%) but differing GH concentrations. It was also observed that increasing the collagen content of the hydrogels loaded with the same amount of GH (25 mIU GH per cm3) increased the total amount of GH released (Figure 4). Samples from the SO/70 blend released significantly more GH (P < 0.05 paired t-test) than samples from the 20180 blend, which in turn released significantly more GH (P < 0.05 paired t-test) than samples from the 10160 blend.

e,i,

0.4

:_!fyyfy o’2 /

0.0 0 1 2 3 4


Figure 2 Release of growth hormone (GH) from hyaluro- nit acidlpoly(acrylic acid) sponges (60140) with different initial GH concentrations. -E-, GH = 2.5 mlU; +-, GH = 5mlU; 0 GH = 7.5 mlU; 0 GH = 10 mlU.

w._

0.2

0.1

0.0

0 1 2 3

Square root tlme (days)

Figure 3 Release of growth hormone (GH) from collagen/poly(vinyl alcohol) hydrogels (20/80) with different initial GH concentrations. +, GH=25 mlU; +-, GH = 50 mlU; 0, GH = 75 mlU; 0 GH = 100 mlU.


572 Drug delivery systems for growth hormone: M.G. Cascone et al.

Hyaluronic acid/poly(vinyl alcohol) hydrogels

During the first 3 d GH was released from the HA/ PVA hydrogels in a constant linear manner which subsequently reached a plateau (Figure 5). The 301

70 blend polymer released significantly more GH (P < 0.05 paired t-test) than the 20180 blend and these in turn released significantly more GH (P < 0.05 paired t-test) than the lo/90 blend. The initial GH concentration did not affect the pattern of release but did affect the total amount released. In the case of the lo/90 HA/PVA hydrogels, there was a direct relationship between the amount of GH incorporated in the hydrogels and the total amount of GH released (Figure 6).

o’4>

0.7

i

0.6 -I

s E 0.5 -

z

3 al Tl 0.4 - b

ii

0.3 -

0 1 2 3 4


Figure 6 Release of growth hormone (GH) from hyaluronic acid/poly(vinyl alcohol) hydrogels (lW90) with different initial GH concentrations. +X-, GH=25 mlU; f, GH =50 mlU; 0, GH = 75 ml& * GH = 100 mlU.

0.2 ! I I I


Figure 4 Release of growth hormone (GH) from collagen/ poly(vinyl alcohol) (C/PVA) hydrogels with three different collagen contents but the same GH concentration: -E-, C/ PVA = 10190; t, CIPVA = 20180; 0, CIPVA = 30170.

Figure 7 Scanning electron microscopy of the surface of collagen/poly(vinyl alcohol) (30:70) hydrogel. Bar = 10 pm.

“_L , . T I

0 1 2 3 4


Figure 5 Release of growth hormone (GH) from hyaluronic acid/poly(vinyl alcohol) (HA/PVA) hydrogels loaded with 25 mlU GH per cm3 HA. -D-, HA/PVA= 10190; t, HA/ PVA = 20180; 0, HAIPVA = 30170.


Scanning electron micrographs

Figure 7 shows the surface morphology of the CIPVA hydrogel with the ratio 30:70. The surface is fibrillar with a porous structure. Figure 8 shows the internal structure of the CIPVA hydrogel, ratio 30:70. The matrix is fibrillar with interconnecting pores. Figure 9 shows the surface of the HAIPAA 40:60 sponge. The surface of the material was smooth and porous with large clefts in the surface. Figure 10 shows the internal structure of the HA/PAA 40:60 sponge. The matrix shows sheets of smooth interconnecting material with large channels.

DISCUSSION

The use of biomaterials to deliver biologically active agents is an attractive concept because local adminis-

Drug delivery systems for growth hormone: M.G. Cascone et a/. 573

Figure 8 Scanning electron microscopy of the internal structure of collagen/poly(vinyl alcohol) (30:70) hydrogel. Bar=lOpm.

Figure 8 Scanning electron microscopy of the surface of hyaluronic acidlpoly(acrylic acid) (40:60) sponge. Bar = 100pm.

tration of certain therapeutic agents is often the most effective method of treatment. One example of such therapy is the local administration of antibiotics from bone cements for the prevention of deep wound sepsisg9 lo, Other examples include the delivery of growth hormone (GH) from bone cements’l’l’, ceramics13, degradable microspheres14 and wound dressings15. The development of controllable long-term effective release systems for the delivery of GH and other growth factors may improve wound healing and tissue repair.

The concept and development of ‘bioartificial polymeric materials’ was proposed by Giusti and co- workers16-18. Blends of HA/C with PVAIPAA have been studied for potential applications in the biomedical field’. We have investigated the potential use of such polymers for delivery systems of biologically active agents. It was found that the CIPVA films used in this study could not be used to deliver GH. We consider this to be due to the use of GTA as a cross-linking agent and

postulate that the GTA cross-linked not only collagen but also GH. If these films are to be used as drug delivery systems, it will be necessary to use a different cross- linking system. The dehydrothermal system requires a temperature of 120" C which is not practical as it would denature the GH. Alternative systems could be cross- linking by ultraviolet irradiation or by carbodiimide. The HAIPAA sponges proved to be an excellent delivery system for GH. The level of GH released was similar to that previously released from bone cement and it has been shown that this level of GH release can stimulate osteoblasts in an animal modellg.

It has been demonstrated that GH can be used to stimulate slow cycling germinal cells in the rat tibia1 growth plate”, chondrogenic cell line? and has a direct effect on bone cells in cultureZ2,23. Clearly, GH is a potent stimulant for bone and cartilage cells and the sponges described in this work could potentially be used as delivery systems. The release of GH was directly related to the initial amount incorporated in the sponges and increasing the HA content of the sponges increased the GH release. The scanning electron micrographs of both the surface and the internal structure of the sponges showed porous structures composed of sheets of the polymers with interconnecting channels; such a structure allows transport of GH from the polymers into the surround- ing environment. The hydrogels also proved to be useful for drug delivery but the release of GH from the C/PVA hydrogels was not linear during the first 3 d of the elution period. We assume this was because these hydrogels swelled in a non-linear fashion when first immersed in the buffer. The later release of GH was linear before levelling off, for all the hydrogels tested. Scanning electron micrographs of both the surface and the internal structure of the hydrogels showed a porous, fibrillar matrix which allowed the transport of additives through the matrix. As with the sponges, increasing the biological component (collagen or hyaluronic acid) increased the amount of GH released. Overall, there was more GH released from the HA/PVA hydrogels compared with the CIPVA hydrogels.

Figure 10 Scanning electron microscopy of the internal structure of hyaluronic acid/poly(acrylic acid) (40:60) sponge. Bar = 100 pm.


574 Drug delivery systems for growth hormone: M.G. Cascone et al.

In this work we have successfully combined natural and synthetic polymers to produce a new drug delivery system for growth factors. We are currently investigating the biocompatibility of these polymers for potential applications in cartilage and bone repair.

REFERENCES

1

2

3

4

5

6

7

8

9

10

11

Giusti P, Lazzeri L, Lelli L. Bioartificial polymeric materials: a new method to design biomaterials by using both biological and synthetic polymers. TRIP 1993; l(9): 261-267. Giusti P, Lazzeri L, Barbani N, Lelli L, De Petris S, Cascone MG. Blends of natural and synthetic polymers: a new route to novel biomaterials. Macromol Chem (in press]. Barbani N, Lazzeri L, Lelli L. Nuovi materiali polimerici bioartificiali costituiti da collagene e alcool polivinilico. Biomateriali1992; l/Z: 59-66.

Yannas IV. Collagen and gelatin in the solid state. ] Macromol Sci Rev Macromol Chem 1972; C?‘(l): 49-104. Rubin AL, Miyata T, Stenzel KH. Collagen: medical and surgical applications. J Macromol Sci Chem 1969;

A3(1): 1118-1123. Swann DA, Kuo J. Hyaluronic acid. In: Byrom D, ed. Biomaterials. Novel Materials from Biological Sources. New York: Stockton Press, 1991. Giusti P, Lazzeri L, Barbani N, et al. Hydrogels of poly(viny1 alcohol] and collagen as new bioartificial materials. J Mater Sci: Mater Med 1993; 4(6): 538-542. Pringle PJ, Di Silvio L, Hindmarsh PC, Matthews DR, Kurtz AB, Brook CGD. Analysis of trough serum growth hormone concentrations: comparison of an immunoradiometric assay and a sensitive ELISA for growth hormone. Clin Bndocrinol 1992; 37: 169. Buchholz HW. Prophylactic use of antibiotic-located cement: long term results. In: Marti RK, ed. Proceed- ings of a Symposium Amsterdam Experta Medica, Amsterdam, 1982. Downes S, Maughan PA. Mechanism of antibiotic release from poly(methyl)methacrylate bone cement. Clin Mater 1989; 4: 109-122. Downes S, Wood DJ, Malcolm A, Ali SY. Growth hormone in polymethylmethacrylate. Clin Orthop Rel Res 1990; 252: 294-298.

12

13

14

15

16

17

18

19

20

21

22

23

Downes S, Kayser MV, Blunn G, Ali SY. An electron microscopical study on the interaction of bone with growth hormone loaded bone cement. Cell Mater 1991; l(2): 171-176. Downes S, Di Silvio L, Klein CPAT, Kayser MV. Growth-hormone loaded bioactive ceramics. J Mater Sci: Mater Med 1991; 2: 176-180. Di Silvio L, Gurav N, Downes S. Gelatin spheres: a new delivery system for growth hormone. Biomaterials (in press). Andreatta-van Leyen S, Smith DJ, Bulgrin JP, Schafer IA, Eckert RL. Delivery of growth factor to wounds using a genetically engineered biological bandage. J Biomed Mater Res 1993; 27(g): 1201-1208. Giusti P, Ciardelli F, Pracella M, Mercogliano N, Barbani N. New bioartificial polymeric materials. In: Proceedings of 1st Pacific Polymer Conference, 12-15 December 1989, Pacific Polymer Federation, Maui, Hawaii, 137. Giusti P, Ciardelli F, Lazzeri L, Lelli L. Bioartificial (polymer based) materials. International Symposium Polymer Materials, IO-14 February 1991, Australia, Abstracts, 491. Giusti P, Lazzeri L. Bioartificial polymer materials for biomedical applications. 2nd Joint Soviet Union-Italy Meeting on Polymer Science, 9-13 September 1991,

Leningrad, Russia, Abstracts, 37. Downes S, Wood D, Malcolm AJ, Ali SY. Growth hormone in polymethylmethacrylate cement. Clin Orthop Rel Res 1990; 252: 294-298. Ohlsson C, Nilsson A, Isaksson 0, Lindahl A. Growth hormone induces multiplication of the slowly cycling germinal cells of the rat tibia1 growth plate. Proc Nat1 Acad Sci USA 1992; 89: 9826-9830. Ohlsson C, Nilsson A, Swolin D, Isaksson 0, Lindahl A. Establishment of a growth hormone responsive chondrogenic cell line from fetal rat tibia. Molec Cell Endocrinol 1993; 91: 167-175. Stracke H, Schulz A, Moeller D, Ross01 S, Schatz H. Effect of growth hormone on osteoblasts and demonstra- tion of Somatomedin-C/IGF-I in bone organ culture. Acta Endocrinol 1984; 107: 16-24. Schevan BA, Hamilton NJ, Fakkeldij TM, Duursma SA. Effects of recombinant human insulin-like growth factor-I and II and growth hormone on the growth of normal adult human osteoblast-like cells and human osteogenic sarcoma cells. Growth Regulation 1991; 1: 160-167.


blends of synthetic and natural polymers as drug delivery systems for growth hormone

Documents