The Biomaterials: Silver Jubilee Compendium

Controlled release of biologically active compounds from bioerodible polymers

J. Heller    Polymer Sciences Department, SRI International, Menlo Park, CA94025, USA

Abstract

This article reviews the controlled release of biologically active agents by the erosion or chemical degradation of a polymer matrix into which the agent is incorporated. Chemically bound active agents and work on steroid release from cholesterol implants are not covered. The mechanisms of polymer erosion discussed are: cross-linked scission; hydrolysis, ionization or protonation of pendant groupts; backbone cleavage. Drug release studies are dealt with under each of these headings.

It is now generally recognized that the controlled release of biologically active agents to a local environment can be achieved by means of one of three general methodologies: (1) diffusion through a rate-controlling membrane, (2) use of osmotically regulated flow, and (3) release controlled by the erosion or chemical degradation of a matrix into which the active agent is incorporated¹.

Each of these methodologies offers certain unique characteristics which determine the design of specific therapeutic systems. Thus, methodology (1) allows construction of drug delivery devices that release therapeutic agents by zero order kinetics and where rate of delivery can be readily adjusted by changing the rate-limiting membrane and/or membrane thickness and area. Methodology (2) allows construction of devices that not only release their contents by zero order kinetics but are also able to sustain high delivery rates not normally available with membrane-moderated devices. Methodology (3) allows construction of drug delivery devices that have a predetermined life span and need not be removed from the site of action once their drug delivery role has been completed.

Drug release from bioerodible polymers finds use in both topical applications and systemic applications. Both uses demand that the polymer degrade to nontoxic products, and polymers used in systemic applications must also degrade to low-molecular-weight fragments that can be readily eliminated or metabolized by the body. In topical applications, retainment of high molecular weight of the degradation products is desirable, since in this way no unnecessary systemic absorption of the polymer will occur, and toxicological hazards are thus reduced.

The purpose of this article is to present a comprehensive review of methodology (3) where active agents are released to a surrounding aqueous environment by solubilization of the polymer matrix induced by the aqueous environment. The review is limited to devices in which the active agent is dissolved or dispersed in a polymer, and does not cover the important work in which the active agent is chemically bound to the polymer and is released to the surrounding medium by hydrolysis of a bond between the active agent and the polymer chain², ³; nor does it cover the extensive work of Kincl and coworkers on the release of steroids from cholesterol implants⁴.

For this review, it is convenient to systematize polymer erosion according to the three mechanisms shown in Figure 1, where ⊗ denotes a hydrolytically unstable bond⁵. In general terms, Mechanism I encompasses water-soluble polymers that have been insolubilized by hydrolytically unstable crosslinks; Mechanism II includes polymers that are initially water-insoluble and are solubilized by hydrolysis, ionization, or protonation of a pendant group; and Mechanism III includes hydrophobic polymers that are converted to small water-soluble molecules by backbone cleavage. Clearly, these three mechanisms represent extreme cases, and erosion by a combination of mechanisms is possible.

Figure 1 Schematic representation of degradation mechanisms;

⊗ denotes a hydrolytically unstable bond

A represents a hydrophobic substituent and

B → C represents hydrolysis, ionization or protonation

MECHANISM I

Solubilization by crosslink cleavage

In these systems, water-soluble polymers are insolubilized by means of hydrolytically unstable crosslinks. Consequently, the resulting matrix is highly hydrophilic and completely permeated by water. Since the active agent is in an aqueous environment, its water solubility becomes an important consideration, and compounds with appreciable water solubility will be rapidly leached out, independent of the matrix erosion rate.

There are two general applications in which erodible hydrogels are useful in the controlled delivery of active agents. In the first the active agent has extremely low water solubility, and in the second the active agent is a macromolecule that is entangled in the hydrogel matrix and cannot escape until a sufficient number of crosslinks have cleaved and matrix crosslink density has been reduced.

The first application is illustrated in Figure 2, which shows the release of a highly water-insoluble drug, hydrocortisone acetate, from a gelatin matrix crosslinked with formaldehyde⁵. As indicated by the first-order dependence, release is by diffusion with little contribution by matrix erosion. Because the drug is very water insoluble, useful release over many days is achieved. Such a device could be used when zero-order kinetics are not important and removal of the expended device is not convenient or desirable⁶.

Figure 2 Release of hydrocortisone acetate from a crosslinked gelatin matrix

Illustrative of the second application are many examples in which water-soluble macromolecules have been immobilized in hydrogels by physical entanglement⁷. However, the intent of most of these studies was to achieve long-term immobilization of enzymes or antigens; the slow diffusional escape and/or slow hydrolysis of the matrix with consequent liberation of the entrapped macromolecules was generally regarded as undesirable. Two studies, however, recognized the utility of erodible hydrogels for providing sustained delivery of entrapped macromolecules. Both studies took advantage of the hydrolytic instability of crosslinks formed in vinyl polymers by using N, N′-methyl-enebisacrylamide as a comonomer. Hydrolysis of the crosslinks proceeds as follows:

where –R– represents the vinyl polymer chain.

In the first study⁸, immobilized in a hydrogel prepared from acrylamide and 2%N, N′-methylenebisacrylamide. Slow release of insulin from the hydrogel was inferred because insulin-containing hydrogel implants sustained diabetic animals for at least a few weeks. It is not clear from the study how much insulin was released by diffusion and how much by cleavage of crosslinks.

In the second study⁹, α-chymotrypsin was immobilized in a hydrogel prepared from N-vinyl pyrrolidone and N, N′-methylenebisacrylamide. It was found that, by varying the N, N′-methylenebisacrylamide concentration from 0.1 to 1.0 w/w % with respect to N-vinylpyrrolidone, hydrogels with dissolution times of several days to practically insoluble could be prepared. However, release of α-chymotrypsin did not correlate well with hydrogel dissolution time, presumably because of diffusional escape.

To prevent diffusional escape, α-chymotrypsin was acylated with acryloyl chloride and then chemically incorporated in the hydrogel by copolymerization with N-vinyl pyrrolidone and N, N'-methylenebisacrylamide. Release of α-chymotrypsin from the resulting hydrogels was then found to correlate more closely with hydrogel dissolution times.

MECHANISM II

Solubilization by hydrolysis, ionization or protonation of pendant groups

Systems in this category include all polymers that are initially water-insoluble but become water-soluble as a consequence of hydrolysis, ionization, or protonation of pendant groups. Because no backbone cleavage takes place, the solubilization does not result in any significant changes in molecular weight.

The major emphasis in the development of these materials has been on enteric coatings. These are coatings designed to be insoluble in a certain pH environment, usually the stomach, and then to dissolve abruptly in an environment of a different pH, such as the intestines. Usually these polymers are applied to pills as protective coatings and do not produce steady, sustained release of therapeutic agents. However, by using mixtures of enteric coatings, each with a different disintegration time, it is possible to prolong the action of therapeutic agents¹⁰.

Literature on enteric coatings is much too voluminous for detailed review, but the coatings can be grouped into three categories according to their dissolution mechanism; (1) dissolution by side group hydrolysis, (2) dissolution by ionization of a carboxylic acid function and (3) dissolution by protonation of amine functions.

Dissolution by side group hydrolysis. Materials in this category are represented by copolymers of vinyl monomers and maleic anhydride:

where X is OR or H.

In the anhydride form, the polymers are water-insoluble, but on exposure to water they become soluble because of the hydrolysis of the anhydride group. A number of variables affect the rate of polymer dissolution and lag time before initiation of the dissolution process¹¹. In general, time before dissolution increases as the size of the alkyl substituent R in the vinyl ether portion of the copolymer increases, and decreases as the pH of the aqueous environment increases. The rate of polymer dissolution also increases as the pH of the aqueous environment increases.

Dissolution by ionization of carboxyl groups. Currently used enteric coatings represent this type, and can be represented generally as polyacids. While in unionized form they are water-insoluble, but on ionization of the carboxylic acid functions they become water-soluble.

The most widely used enteric coatings are based on cellulose acetate phthalate¹². They are insoluble in aqueous acidic media but, because of the free acid groups on the phthalate radical, dissolve in aqueous bases. Enteric coatings based on cellulose acetate succinate have also been described¹³.

Partially esterified copolymers of methyl vinyl ether and maleic anhydride or partially esterified copolymers of ethylene and maleic anhydride have also been investigated¹⁴–¹⁶. It was found that these materials characteristically exhibit a pH range above which they are soluble and below which they are insoluble. This pH range is quite sharp, about 0.25 pH units, and changes with the number of carbon atoms in the ester side group of the copolymers.

This behaviour can be understood readily by considering the number of ionized carboxyls necessary to drag the polymer chain into solution. With relatively small ester groups, a low degree of ionization is sufficient to solubilize the polymer; hence the dissolution pH is low. As the size of the alkyl group increases, so does the hydrophobicity, and progressively more ionization is necessary to solubilize the polymer, resulting in increasingly high dissolution pH. The same argument holds for polymers having the same ester grouping but different degrees of esterification. The higher the degree of esterification, the more hydrophobic the polymer and consequently the higher the dissolution pH.

Recently it has been shown¹⁷ that partially esterified copolymers of methyl vinyl ether and maleic anhydride can, in a constant pH environment, release hydrocortisone incorporated therein by excellent zero-order kinetics.

Figure 3 shows polymer dissolution rate and the rate of hydrocortisone release for n-butyl half-ester polymer films containing the dispersed drug. Each pair of points represents a separate device in which the amount of drug released by the device into the wash solution was determined by u.v. measurements and the amount of polymer dissolved was calculated from the total weight loss of the device. The excellent linearity of both polymer erosion and drug release over the lifetime of the device provides strong evidence for a surface-erosion mechanism and for negligible diffusional release of the drug. The latter result was independently verified by placing a drug-containing film in water at a pH low enough that no dissolution of the matrix took place and periodically analysing the aqueous solution for hydrocortisone. None was found over several days.

Figure 3 ), drug release; (Δ), polymer dissolution. Reproduced from J. Appl. Polym. Sci. 1978, 22, 1991 by permission of John Wiley and Sons Inc., New York

Figure 4 shows the effect of size of alkyl group on polymer erosion rate for a series of partial esters measured at pH 7.4. Because of the linear correlation between polymer erosion and drug release, distance eroded can be directly correlated with amount of drug released and was, in fact, derived from measuring drug release. All drug release rates again show excellent linearity and strong dependence on the size of the alkyl group. Since in all experiments drug depletion also coincided with total polymer dissolution, again it can be assumed that drug release and polymer erosion occur concomitantly.

Figure 4 Effect of size of ester group in half-esters of methyl vinyl ether-maleic anhydride copolymers on rate of erosion at pH 7.4. Reproduced from J. Appl. Polym. Sci. 1978, 22, 1991 by permission of John Wiley and Sons Inc., New York

The effect of pH on rate of polymer erosion and hence release of hydrocortisone dispersed in the n-butyl partial ester is shown in Figure 5. The date show a clear dependence of erosion rate and drug release on the pH of the eroding medium and, as expected, a progressive decrease in rate as the critical dissolution pH is approached.

Figure 5 Effect of erosion medium pH on erosion rate of half-esters of methyl vinyl ether-maleic anhydride copolymers. A = pH 5.5, B = pH 5.75, C = pH 6.0, D = pH 6.5, E = pH 7. Reproduced from J. Appl. Polym. Sci. 1978, 22, 1991 with permission of the copyright owner

The partial ester copolymer also has been used recently as a model for a bioerodible drug delivery system that releases a therapeutic agent in response to the presence of a specific external molecule¹⁸. In this model, hydrocortisone was incorporated into an n-hexyl half ester of a methyl vinyl ether-maleic anhydride copolymer and the polymer-drug mixture fabricated into discs. These discs were then coated with a hydrogel containing immobilized urease. In a medium of constant pH and in the absence of external urea, the rate of release of hydrocortisone was that normally expected for that polymer at the given pH. In the presence of external urea, ammonium bicarbonate and ammonium hydroxide were generated within the hydrogel, which accelerated polymer erosion and hence drug release. Drug delivery rate increase was proportional to the amount of external urea and was reversible; that is, when external urea was removed, the rate of drug release gradually returned to its original value. The effect of urea on polymer erosion and drug release is shown in Figure 6.

Figure 6 Rate of hydrocortisone release at 35 °C from a n-hexyl half-ester of a copolymer of methyl vinyl ether and mateic anhydride at pH 6.25 in the absence and presence of external urea. A = polymer + hydrogel, 10 −1  m urea; B = polymer + hydrogel, 10 −2  m urea; C = polymer + hydrogel, no urea. Reproduced from J. Pharm. Sci. 1979, 68, 919 with permission of the copyright owner

Date presented thus far show that the partial esterpolymer system is a useful polymer matrix for zero-order drug delivery; it shows the desired surface erodibility, and the erosion rate at any given pH can be adjusted by selecting the appropriate ester alkyl group.

The usefulness of this polymer system for delivery of a therapeutic agent in vivo has been investigated by fabricating hydrocortisone-containing films in small circles and placing them in the lower forniceal cul de sac of New Zealand rabbits⁵. Kinetic plots were obtained by placing weighed devices in eyes and removing the devices at desired time intervals. The amount of drug released was determined by measuring residual drug in the devices and subtracting that from the known original amount. The results expressed in terms of polymer erosion are shown in Figure 7.

Figure 7 In vivo release of hydrocortisone from ocular inserts in rabbits

Clearly, the devices are highly functional and release hydrocortisone by excellent zero-order kinetics. Furthermore, considering that each point represents a different device and a different rabbit eye, there is very little scatter, indicating a high degree of reproducibility.

Polymer dissolution by protonation of amine functions. Materials in this group are, in effect, reverse enteric coatings; they are insoluble in water and alkali but soluble in acids. Although no drug-release studies have been described, suggested uses have been as moisture-resistant medicament coatings that will release their contents in the stomach¹⁹, ⁹ and as veterinary medicament coatings to allow passage through the stomach of ruminants and subsequent release in the abomasum²⁰.

A material typical of this group is cellulose acetate N, N-diethylaminoacetate, prepared by the amination of cellulose acetate chloroacetate with diethylamine²¹. Another material was prepared by the addition of amines to crotonic acid esters of cellulose²².

MECHANISM III

Solubilization by backbone cleavage

This category includes all water-insoluble polymers that undergo hydrolytic backbone cleavage and are solubilized by conversion to small, water-soluble molecules.

A major driving force for the development of these materials was a search for absorbable sutures superior to catgut. From these studies have evolved two synthetic materials that were suitable for surgical implants: poly (lactic acid)²³, ²⁴ and poly (glycolic acid)²⁵. The first demonstration of the utility of poly(lactic acid) as a bioerodible implant capable of sustained release of a therapeutic agent was described about 10 years ago²⁶, and since then many publications have dealt with the sustained release of pharmaceuticals from bioerodible implants.

For this review, the literature will be discussed according to the type of pharmaceutical agent delivered: (1) delivery of narcotic antagonists, (2) delivery of contraceptive steroids, and (3) delivery of others.

Delivery of narcotic antagonists. The release of cyclazocine from poly(lactic acid)²⁷ and the release of Naltrexone from poly(lactic acid)²⁸ have been described in two studies. In the first study, composites of poly(lactic acid) and labelled cyclazocine were prepared and implanted in rats as 4 cm² films, as ground films having particle sizes falling within No. 25/35 sieves, and as 6.2 cm² poly (lactic acid)-cyclazocine composites enveloped in poly(lactic acid) containing no drugs. In vivo release was followed by monitoring the radioactivity of excreted urine.

The results, shown in Figure 8 were not those expected; cyclazocine is released by diffusion, and therefore release from small particles with their greater surface area should be much faster than release from films. It was postulated²⁷ that the in vivo results are influenced by varying degrees of inflammation and oedema. This may be reasonable, because in vitro release data show the expected much faster release from particles than from films. Since the polymer was said to bioerode in about 62 days and after 55 days, only 30% of the drug was released, a large burst in drug delivery should take place shortly thereafter.

Figure 8 ), film with drug sealed into poly (lactic acid) envelope. Reproduced with permission from J. Med. Chem. 1973, 16, 897 © American Chemical Society

In the second study²⁸, tritiated Naltrexone was incorporated into poly(lactic acid), and the composite was ground and then injected into rats. Drug release, as measured by urinary excretion, levels off at about 25% after about 30 days. Again, there was a considerable disparity between in vivo and in vitro release, this time attributed to tritium exchange in the body.

Release of cyclazocine microencapsulated in poly (lactic acid) has also been described.²⁹ However, because of macroscopic defects in the capsule wall, all cyclazocine was released in vivo between 14 and 17 days. The permeability of cyclazocine through solvent-cast poly(lactic acid) was measured to yield a value of 2.9 and 3.0 × 10−11 cm²/sec. Using an average of these values, it was calculated that a defect-free capsule should release 50% of its contents in about 28.5 months.

The in vitro release of Naltrexone and Naltrexone pamoate from poly (lactic acid) microcapsules has also been described³⁰. Again, the microcapsules contained defects, and about 50% of the contents were released after the first few hours of extraction.

Copolymers of lactic and glycolic acid have also been used in a bioerodible delivery system for Naltrexone and Naltrexone pamoate³¹. The in vitro release rates of Naltrexone from rods of 75/25 poly(lactic/glycolic acid) copolymer as functions of drug loading are shown in Figure 9. As expected, increase in drug loading results in increased delivery rates.

Figure 9 In vitro release of , 50% w/wn. base. Reproduced with permission from Life Sci 1975, 17, 1877 © 1975 Pergamon Press Ltd.

In vivo release of labelled Naltrexone from 90/10 poly (lactic/glycolic acid) beads is shown in Figure 10. The release was measured as urinary excretion; after about 90 days about 50% of the drug was accounted for. The remaining drug was assumed to have been excreted in the fæces. Other investigators³² found that Naltrexone is excreted in approximately equal amounts in the urine and in the fæces.

Figure 10 Cumulative percent of ³ H, implanted as ³ H-Naltrexone base (33% w/w) in 90/10poly (L(+)-lactic co-glycolic acid) beads, recovered in the urine of mice. Reproduced with permission from Life Sci 1975, 17, 1877 © 1975 Pergamon Press Ltd.

Delivery of contraceptive steroids. d-Norgestrel was incorporated into poly (lactic acid) and its release rate followed both in vitro and in vivo³³. It was found that films of the polymer containing 33% Norgestrel released the steriod at a relatively constant rate of 3 μg/day per cm² for over 80 days. When similar films were implanted subcutaneously in rats, the initial release was about 5.5 μg/day per cm²; by 80 days release had declined to about 3 μg/day per cm². Biodegradation of the matrix was slower than drug release.

In another study³⁴ ¹⁴C-labelled 20 wt% Norethisterone was incorporated into poly (L(+)-lactic acid) and the composite was cryogenically ground and separated into 90- to 180-μ particles and 180- to 250-μ particles. The 180- to 250-μ particles were overcoated with the benzene-soluble poly (dl-lactic acid). The in vitro release of the steriod from all three particle-size formulations is shown in Figure 11. Release from the larger particles was faster than that from the smaller particles, which is contrary to what might be expected because rate of diffusional release is directly proportional to surface area. Release from coated samples is considerably lower than that from uncoated samples. This is as expected; the steriod must diffuse through a rate-controlling membrane.

Figure 11 ) 180-250 μ; coated (using a 10% poly-DL-lactic acid/benzene solution). From Contraception 1976, 13, 375. Reproduced with permission of the copyright owner

Studies of in vivo release, as measured by urinary excretion, show an apparent zero-order delivery with an initial burst and another burst around day 90, which is attributed to dissolution of the polymer matrix. Comparison of actual in vivo release kinetics for all three samples is difficult because the steriod is excreted in both urine and fæces, and in the uncoated samples only the urinary steroid was measured. Furthermore, either because of cumulative error or errors in original specific activity, mass balance is poor. However, the results indicate fairly linear and sustained steroid release for about three months; a typical plot is shown in Figure 12.

Figure 12 Cumulative percent ¹⁴ C label from norethisterone excreted in vivo in the urine of rats from 90–180 μ particles of poly-L(+)-lactic acid containing 20% by weight norethisterone. From Contraception 1976, 13, 375. Reproduced with permission of the copyright owner

In vivo and in vitro release of progesterone, β-oestradiol, and dexamethasone from poly (lactic acid) beads and chips also has been described³⁵.

A considerable amount of work has been done on release of contraceptive steroids from crosslinked poly (dimethyl siloxane) implants³⁶. However, silicone rubber is not degradable, so the expanded device must be surgically removed. This is not desirable; consequently it is desirable to develop devices which would release steroids by membrane diffusion and the expended devices would then later bioerode.

-caprolactone and DL-lactic acid were investigated as potential bioerodible membranes-caprolactone) and silicone rubber have comparable maximum steady state fluxes of progesterone at 37 °C 2.2 × 10−10 and 0.6 × 10−10 g/cm sec, respectively). Poly (dl-lactic acid) had a maximum steady state flux of progesterone of 3.3 × 10−15 g/cm sec.

In another study-caprolactone, dl-lactic acid and glycolic acid in vitro and in vivo were determined. Detailed interpretation of the data is difficult because various factors, such as simple drug diffusion, polymer erosion, morphological changes in the polymer, and changes in concentration gradient across capsule walls, all contribute to the measured erosion rate.

The release of progesterone from bioerodible capsules based on glutamic acid/leucine copolymers has also been investigated³⁹. Devices were fabricated by first preparing progesterone-filled polypeptide rods and then coating the rods with unfilled polypeptide. Because of various fabrication difficulties, it was not possible to determine meaningful release rates. Bioerosion of the copolymers was extremely slow and complete erosion for a 50/50 copolymer was estimated to take about four years.

Delivery of other agents. Although the greatest emphasis of drug delivery from bioerodible polymers has centered around narcotic antagonists and contraceptive steroids, the methodology has been applied to other drugs.

A brief study described release of two anticancer drugs, cyclophosphamide and cis-dichlorodiammineplatinum, from poly (lactic acid)⁴⁰.

Another study⁴¹ described the release of the anti-malarial drug, 2, 4-diamino-6-(2-naphthylsulphonyl)-quinazoline (WR-158122) from a 25/75 poly (dl-lactic/glycolic acid) copolymer in mice. Measuring radioactivity excreted into urine and fæces, sustained release through 14 weeks, was demonstrated.

DEGRADATION MECHANISMS

The hydrolytic erosion of a solid polymer can be rationalized in terms of two extreme mechanisms. In one, referred to as homogeneous, the hydrolysis occurs at a uniform rate throughout the matrix. In the other, called heterogeneous, the process is confined to the surface of the device. Actual erosion can, of course, occur by some intermediate mechanism.

In a purely homogeneous process, the matrix will remain essentially intact until all parts reach some critical degree of reaction, at which point the matrix dissolves.

Drug release from such a matrix is complicated because it is a combination of diffusion and erosion. Diffusional drug release in the absence of erosion can be described by the Higuchi model⁴² which proposes that the drug is initially removed from the surface regions of the polymer, and consequently a progressively thicker drug-depleted layer forms adjacent to the surface of the device. Because the remaining drug must diffuse through a progressively thickening polymer membrane, the rate of drug release declines continuously.

In the presence of polymer erosion, the matrix progressively loosens up, and the permeability of the polymer to the drug increases with time. Consequently, the rate of drug release from polymers which erode either wholly or partially by bulk degradation at first shows the normal expected decline, but as the increasing polymer permeability gradually offsets this decline, the rate of drug release eventually accelerates.

The heterogeneous process is a much more desirable degradation mode because it will lead to zero-order drug release, provided that diffusional release of the drug is minimal and the overall shape of the device remains essentially constant, thus maintaining constant surface area. Furthermore, such a process avoids deterioration of mechanical properties, which can take place when random chain cleavage occurs in the bulk material.

In principle surface erosion can be achieved by constructing a highly hydrophobic polymer, so that water penetration and consequent bulk hydrolysis is much less probable than surface reaction. A much preferred approach is one in which a pH-sensitive reaction is selected and the interior of the matrix is buffered so that bulk hydrolysis is prevented and reaction can take place only at the surface of the device, where the buffer is neutralized.

The only polymers thus far described in detail that undergo true surface erosion are the partial esters of methyl vinyl ether/maleic anhydride copolymers¹⁷, ¹⁸. However, these represent a special case because solubilization does not involve backbone cleavage, and consequently the only polymer chains than can escape into the aqueous environment are at the surface of the device.

None of the described backbone-degradable polymers undergo surface erosion; rather, they undergo bulk erosion. This is not surprising, since most of them were originally designed as bioerodible suture materials and were not intended as drug carriers capable of releasing a drug by zero-order kinetics. However, work in progress here⁴³ and elsewhere⁴⁴, ⁴⁵ is devoted specifically to the development of bioerodible polymers that do undergo surface erosion and are capable of releasing incorporated drugs by zeroorder kinetics.

REFERENCES

1 Baker RW, Lonsdale HK. Chemtec. 1975;5:668.

2 Harris FW, Aulabaugh AE, Case RD, Dykes MK, Feld WA. Controlled Release Polymeric Formulations. In: Paul D.R., Harris F.W., eds. ACS Symposium Series No. 33. 1976:222.

3 Neogi AN, Allan GG. Controlled Release of Biologically Active Agents. In: Tanquary A.C., Lacey R.E., eds. Plenum Press; 1976:195.

4 Joseph AA, Hill JL, Patel J, Patel S, Kincl FA. J. Pharm. Sci. 1977;66:490.

5 Heller J, Baker RW. Paper presented at the University of Southern California Symposium on Applications of Polymers in Dentistry and Medicine. November 1973.

6 Hussain AA, Higuchi T, Shell JW. 1976 US Pat 3960 150.

7 Goldman R, Goldstein L, Katchalski E. Biochemical Aspects of Reactions on Solid Supports. In: Stark G.R., ed. Academic Press; 1971:13.

8 Davis BK. Experentia. 1972;28:348.

9 Torchilin VP, Tischenko EG, Smirnov VN, Chazov EI. J. Biomed. Mater. Res. 1977;11:223.

10 Ballard BE, Nelson E. Remington’s Pharmaceutical Science. 1975 Mack, Easton PA, Ch. 91.

11 Woodruff CW, Peck GC, Banker GS. J. Pharm. Sci. 1972;61:1916.

12 Wagner JG, Long S. J. Pharm. Sci. 1960;49:121.

13 Wilken Jr. LO, Kochhar MM, Bennett DP, Cosgrove FP. J. Pharm. Sci. 1962;51:484.

14 Lappas LC, McKeehan W. J. Pharm. Sci. 1962;51:808.

15 Lappas LC, McKeehan W. J. Pharm. Sci. 1965;54:176.

16 Lappas LC, McKeehan W. J. Pharm. Sci. 1967;56:1257.

17 Heller J, Baker RW, Gale RM, Rodin JO. J. Appl. Polym. Sci. 1978;22:1991.

18 Heller J, Trescony PV. J. Pharm. Sci. 1979;68:919.

19 Hiatt GD. 1960 US Pat. 2940901.

20 Farbenfabriken Bayer AG. 1965 Ger. Pat. 1 219 175.

21 Hiatt GD, Mench JW, Fulkerson B. Ind. Eng. Chem. Prod. Res. Develop. 1964;3:295.

22 Mench JW, Fulkerson B. Ind. Eng. Chem. Prod. Res. Develop. 1968;7:2.

23 Kulkarni RK, Pani KC, Neuman C, Leonard F. Arch. Surg. 1966;93:839.

24 Kulkarni RK, Moore EG, Hegyeli AF, Leonard F. J. Biomed. Mater. Res. 1971;5:169.

25 Frazza EJ, Schmitt EE. J. Biomed. Mater. Res. Symp. 1971;1:43.

26 Yolles S, Eldridge JE, Woodland JHR. Polym. News. 1970;1:9.

27 Woodland JHR, Yolles S, Blake DA, Helrich M, Meyer FJ. J. Med. Chem. 1973;16:897.

28 Yolles S, Leafe TD, Woodland JHR, Meyer FJ. J. Pharm. Sci. 1975;64:348.

29 Mason N, Thies C, Cicero TJ. J. Pharm. Sci. 1976;65:847.

30 Thies C. Controlled Release Polymeric Formulations. In: Paul D.R., Harris F.W., eds. 1976:190 ACS Symposium Series No. 33.

31 Schwope AD, Wise DL, Howes JF. Life Sci. 1975;17:1877.

32 Reuming RH, Malspeis L. Ohio State University reports on National Institute on Drug Abuse. 1974-1975 Contract No. H5M-42-73-182,

33 Jackanicz TM, Nash HA, Wise DL, Gregory JB. Contraception. 1973;8:227.

34 Anderson LC, Wise DL, Howes JF. Contraception. 1976;13:375.

35 Yolles S, Leafe T, Sartori M, Torkelson M, Ward L, Boettner F. Controlled Release Polymeric Formulations. In: Paul D.R., Harris F.W., eds. 1976:123 ACS Symposium Series No. 33.

36 Hillier SG, Jha P, Griffiths K, Lanmas KR. Contraception. 1977;15:473.

37 Pitt CG, Jeffcoat RA, Zweidinger RA, Schindler A. J. Biomed. Mater. Res. 1979;13:497.

38 Pitt, C.G., Gratzl, M.M., Jeffcoat, R.A., Zweidinger, R. and Schindler, A., J. Pharm. Sci., in press.

39 Sidman KR, Steban WD, Burg AW. Drug Delivery Systems. In: Gabelnick H.L., ed. 1976 DHEW Publication No. (NIH) 77-1238.

40 Yolles S, Leafe TD, Meyer FJ. J. Pharm. Sci. 1975;64:115.

41 Wise DL, McCormick GJ, Willet GP, Anderson LC. Life Sci. 1976;19:867.

42 Higuchi T. J. Pharm. Sci. 1961;50:874.

43 Heller, J. and Baker, R.W., Work in progress.

44 Choi NS, Heller J. 1978 US Pat. 4 093 709; 4131 648 (1978); 4 138 344 (1979).

45 Capozza RC, Schmitt EE, Sendelbeck LR. National Inst. Drug Abuse Research Monograph. In: Willette R.E., ed. 1976:76–296 DHEW Publication No. (ADM).

The response to the intramuscular implantation of pure metals

A. McNamara; D.F. Williams    Department of Dental Sciences, School of Dental Surgery, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, UK

Abstract

Discs of five high-purity metals, cobalt, nickel, copper, aluminium and lead have been implanted intramuscularly in rats and the response observed histologically for period up to 52 weeks. A reproducible but different response was observed with each metal. Whenever corrosion occurred, as with copper, nickel and some cobalt specimens, the implants became loose. In the absence of corrosion, the implants were firmly held within a more confined capsule. A minimal response was seen with lead, implying normally toxic metals do not necessarily induce excessive responses on implantation. Evidence was found, however, that some metals elicit an immune response whilst some, especially copper and nickel appear to render the host more susceptible to disease. The implants appear to have a profound effect on the immediate vasculature, are able to cause a prolonged polymorphonuclear response in the same way as bacteria, are associated with varying amounts of haemosiderin laden macrophages but not with giant cells. The animals appear to be able to deal with bacteria introduced at surgery without hindrance from the metal.

Ever since the early experimentation with metallic implants by Zierold¹, Hey-Groves² and others, there has been a steady rationalisation of the metals and alloys used clinically and the development of a concise list of recognised and acceptable materials. These are stainless steel of the 316 type, titanium and titanium – aluminium – vanadium alloy and some cobalt- based alloys such as the cobalt–chromium– molybdenum casting alloy, the wrought cobalt – chromium – nickel, cobalt – nickel – chromium – molybdenum and cobalt – chromium – nickel – iron – molybdenum alloys. The basis for this selection has been largely that of corrosion resistance and apparently good biocompatibility, although naturally mechanical properties have also been an important consideration.

The corrosion resistance of these alloys has been extensively investigated, both from in vitro experiments and in vivo observations³,⁴ and it is well established that titanium and its dilute alloy are essentially immune from corrosion in the physiological environment, the cobalt-based alloys are highly corrosion resistant with only a minimal susceptibility whilst stainless steel quite readily suffers crevice and pitting corrosion.

The tissue response has been less well investigated. There have been some experimental studies using various animal models⁵,⁶ and observations on biopsy or autopsy samples⁷,⁸ but little in the way of a detailed systematic study of the tissue response to metals. The general observation is that, in the absence of gross corrosion or wear there appears to be little difference in the reaction to the alloys mentioned above, implants becoming enveloped in a relatively thin fibrous capsule, with no ultra-structional features to distinguish the response to the different materials. There appears, therefore, to be little to choose between the various metals on the basis of this tissue response.

However, with increasing concern over the bio-compatibility of implant materials and especially in terms of the significance of corrosion, wear, hyper-sensitivity and carcinogenicity, it is important that a more detailed understanding of the effects of implanted metals be gained. A major programme of work is under way in the author's laboratory based on this objective and the present paper is concerned with part of this work. Although the ultimate aim must be related to the materials used clinically, with the exception of commercially pure titanium, these are all alloys with 3, 4 or even 5 major elements present. Since the tissue response will be largely dependent on the rate of release of metal ions (or particulate corrosion products) and the physiological activity of these ions, and since each element will be released at a different rate from a complex alloy and will have a different mechanism of toxicity, it would be difficult, if not impossible, to interpret the response to such alloys in terms of basic parameters until the precise effect of each constituent had been