Biological Activities of Polymers - American Chemical Society

---

8 Polymeric Delivery Systems for Macromolecules Approaches for Studying In Vivo Release Kinetics and Designing Constant Rate Systems ROBERT

1

LANGER ,

D E A N S. T.

1,2

HSIEH ,

and L A R R Y BROWN

Downloaded by FUDAN UNIV on December 25, 2016 | http://pubs.acs.org Publication Date: May 12, 1982 | doi: 10.1021/bk-1982-0186.ch008

Massachusetts Institute of Technology, Department of Nutrition and Food Science, Cambridge, M A 02139

Methods for quantitating i n vivo release kinetics and for designing constant-rate release systems for macromolecules are described. These systems are composed of ethylene-vinyl acetate copolymer and the incorporated drug. Approaches for directly quantitating i n vivo release kinetics have been established using the polysaccharide, i n u l i n . In v i t r o and i n vivo release kinetics of inulin from the copolymer system were nearly identical. An approach for achieving zero-order release kinetics was tested by constructing matrices i n the form of a hemisphere with all portions laminated with an impermeable coating except for a cavity i n the center face. Zero-order release kinetics of bovine serum albumin were observed for 60 days. Over the past 20 years research has been conducted i n 2 p r i mary areas concerning the use of polymers to improve the ways drugs are delivered to the body. One way has been to attach the drug to a polymer i n order to impart a different a f f i n i t y or s p e c i f i c i t y to the drug. The second approach has been to incorporate the drug inside a polymeric matrix to cause i t to be slowly released into the body. In this paper, two aspects of the second approach are reviewed for polymeric systems f o r delivering large molecular weight (M.W. > 600) drugs. Despite increasing advances i n the use of controlled release polymeric delivery systems (1-6)» i t has, i n general, proved d i f f i c u l t to develop such vehicles for the long-term administration of macromolecular drugs. This i s largely due to the fact that most biocompatible polymers such as silicone rubber are impermeable to molecules over 600 molecular weight (7). Recent studies i n our laboratory have demonstrated, however, that solvent casting of a variety of polymeric materials (ethylene-vinyl acetate 1

Also affiliated with Children's Hospital Medical Center, Department of Surgery, Boston, MA 02115. Also affiliated with Harvard Medical School, Boston, MA 02115. 2

0097-6156/82/0186-0095$5.00/0

© 1982 American Chemical Society Carraher and Gebelein; Biological Activities of Polymers ACS Symposium Series; American Chemical Society: Washington, DC, 1982.

Downloaded by FUDAN UNIV on December 25, 2016 | http://pubs.acs.org Publication Date: May 12, 1982 | doi: 10.1021/bk-1982-0186.ch008

96

BIOLOGICAL ACTIVITIES OF POLYMERS

copolymer, polyvinylalcohol, poly-2-hydroxyethyl-methacrylate i n the presence of powdered drug permits continuous release of macromolecules for over 100 days (§)• It appears that the powders cause a series of interconnecting channels to be formed in the polymeric matrix. These channels are large enough to permit macromolecular diffusion but tortuous enough to cause a slow and continuous release ( 9 ) . We have also examined factors affecting i n v i t r o release kinetics (10) and have demonstrated that the macromolecules released from the systems are biochemi c a l l y active i n v i t r o and biologically active i n vivo (11). The polymers used have also been shown to be biocompatible i n sensitive animal tissues (12). In this paper we review two areas of our most recent research that provide further information on these unique drug delivery systems. These are (1) the development of techniques for comparing i n vivo and i n v i t r o release kinetics; and (2) the development of approaches for achieving zero-order release kinetics. COMPARISON OF IN V^TRP AND

IN VIVO RELEASE KINETICS

In earlier studies i n which macromolecules were released from polymers i n vivo ( 1 1 ) . the substances tested were proteins or DNA which were metabolized. Thus, a direct comparison of i n vivo and i n v i t r o release was impossible. We therefore chose the polysaccharide inulin which has a molecular weight of 5200 daltons. Inulin was chosen as a marker because i t i s not metabolized, not excreted by the glomerulus, and i s neither reabsorbed or secreted by the kidney tubules. It i s not bound by plasma proteins and i s not toxic (13). Complete recovery of 3H-inulin (14) or C ^ - i n u l i n (15) i n urine has been observed from intravenous injection into rats, rabbits, dogs and humans. Thus, inulin provides an excellent marker for comparing i n v i t r o and i n vivo release kinetics because inulin recovered i n urine can be directly compared to inulin collected i n v i t r o . Methods• Polymer slabs containing inulin were made as follows: 1) inulin was diluted with non-radioactive inulin to a specific activity of 3.49 uCi/mg; 2) 586 mg of this inulin was suspended i n 15 ml of 5% ethylene-vinyl acetate copolymer i n methylene chloride; and 3) this suspension was cast i n a square 7 cm x 7 cm x 0.5 cm glass mold, dried and cut into nine 1 cm x 1 cm squares as described previously (10). In the in v i t r o kinetic experiments, 4 polymer squares were placed i n 5 ml of Phosphate Buffered Saline (PBS) at pH 7.4 and 37°C. Polymer slabs were transferred daily to fresh PBS and the amount of 3n-inulin released was assayed as follows: a 250 u l sample of PBS was added to 5 ml of Biofluour s c i n t i l l a t i o n cocktail and 0.5 ml of d i s t i l l e d water. The emulsion was vortexed and counted in a s c i n t i l l a t i o n counter. In the in vivo experiments, the remaining 5 squares were implanted subcutaneously in rats. The rats were housed in metabolic cages and urine was collected daily. A 250 u l urine sample was

Carraher and Gebelein; Biological Activities of Polymers ACS Symposium Series; American Chemical Society: Washington, DC, 1982.

8.

LANGER ET AL.

Polymeric

Delivery

Systems

97

placed into 5 ml of Biofluour s c i n t i l l a t i o n cocktail, vortexed and counted as above. Quenching of urine was corrected for using the channels ratio method (16). Results. Excellent correlation between i n v i t r o and i n vivo release rates was observed (Figure 1). As i n internal control, the inulin polymer squares were removed from the rats at the end of 450 hrs and rat urine analyzed 4.5 hrs later. As shown i n Figure 1, the recovery rate of H-inulin dropped 300 fold. This control confirmed that ^H-inulin i s rapidly cleared from the blood stream and excreted into the urine. Thus inulin i s an excellent marker for directly assessing release rates from the polymer implant.

Downloaded by FUDAN UNIV on December 25, 2016 | http://pubs.acs.org Publication Date: May 12, 1982 | doi: 10.1021/bk-1982-0186.ch008

3

APPROACHES TO ACHIEVING ZERO-ORDER RELEASE KINETICS Another limitation of the systems which were fabricated i n the shape of a slab i n earlier studies was that they released macromolecules at decreasing rates as a function of time. However, i n many cases, i t i s desirable to have constant release kinetics so that the body w i l l maintain a constant level of the released drug. The reason for the decreases i n release rates observed i n previous studies (10)» may be explained by considering a typical implant, i n the shape of a slab, as a model (17). Soon after implantation, the solid drug dissolves from the surface layer of the implant and then diffuses out of the implant. Since the drug doesn't have very far to travel to reach the surrounding media, release i s rapid. As time elapses, the surface layer becomes depleted, so drug from deeper within the implant must dissolve and diffuse to the surface. The drug now has a longer distance to travel, so release rates decrease. We thought i t would be possible to compensate for this phenomenon i f a polymer implant was designed i n such a way that an increased area of drug would be available for release as the distance from the release surface increased. A variety of shapes were analyzed from a theoretical standpoint (18); the best results were obtained with a hemispheric device laminated with an impermeable coating, except for a small cavity i n the center face (Figure 2). This can be envisioned as a small cantelope cut i n half, where the orange pulp of the melon i s the drug-polymer mixture. The melon half i s then coated everywhere except where the seeds were, so a l l the drug must be released through the small exposed section. In this case although the drug must s t i l l travel i n creased distances at a later point i n release time, this new shape compensates by providing more surface area of available drug as time elapses. To test this hypothesis, procedures were recently developed to construct hemispheric systems for the release of macromolecules (19) and these are discussed below. Methods. The molds used for fabricating hemisphere shaped systems for macromolecules were composed of glass and had hemis-

Carraher and Gebelein; Biological Activities of Polymers ACS Symposium Series; American Chemical Society: Washington, DC, 1982.

98

BIOLOGICAL ACTIVITIES OF POLYMERS

o

w

xlO -

o

3 O

Downloaded by FUDAN UNIV on December 25, 2016 | http://pubs.acs.org Publication Date: May 12, 1982 | doi: 10.1021/bk-1982-0186.ch008

o

T

xlO

POLYMER IMPLANT REMOVED

CL Q

o