Toxicity, biodegradation and elimination of polyanhydrides

Abstract

Although originally developed for the textile industry, polyanhydrides have found extensive use in biomedical applications due to their biodegradability and excellent biocompatibility. Polyanhydrides are most commonly synthesized from diacid monomers by polycondensation. Efficient control over various physicochemical properties, such as biodegradability and biocompatibility, can be achieved for this class of polymers, due to the availability of a wide variety of diacid monomers as well as by copolymerization of these monomers. Biodegradation of these polymers takes place by the hydrolysis of the anhydride bonds and the polymer undergoes predominantly surface erosion, a desired property to attain near zero-order drug release profile. This review examines the mode of degradation and elimination of these polyanhydrides in vivo as well as the biocompatibility and toxicological aspects of various polyanhydrides.

Introduction

Interest in synthetic biodegradable polymers has grown tremendously during the past two decades. Synthetic biodegradable polymers are attractive candidate materials for short-term medical applications, like sutures, drug delivery devices, orthopedic fixation devices, wound dressings, temporary vascular grafts, stents, different types of tissue engineered grafts, etc. [1]. The development of implantable drug delivery systems is perhaps the most widely investigated application of biodegradable polymers. Due to their transient nature, biodegradable polymers do not require surgical removal after their intended application and thus can circumvent some of the problems related to the long-term safety of non-degradable implanted devices.

Biodegradable polymers are those that degrade both in vitro and in vivo into products that are either normal metabolites or into products that could be completely eliminated from the body with or without further metabolic transformations. Biodegradation actually involves two complementary processes, degradation and erosion. Degradation refers to the bond cleavage during which polymer chains are cleaved to low molecular weight fractions and erosion refers to the physical phenomena such as dissolution and diffusion of these low molecular weight fractions from the polymer matrix [2]. The degradation of polymers can take place passively by hydrolysis or actively by enzymatic reaction. The hydrolytically labile polymers are mostly preferred as implants due to their minimal site-to-site and patient-to-patient variation. Polymer erosion can be of two types, bulk erosion and surface erosion [3]. In ideal bulk erosion, material is lost from the entire polymer volume at the same time due to water penetrating the bulk. In this case the erosion rate depends on the total amount of the material. In surface erosion, material is lost from the polymer matrix surface only. These are generally hydrophobic polymers or polymers containing a hydrophobic component wherein water cannot penetrate easily into the bulk. In ideal surface erosion, the erosion rate will be proportional to the surface area.

The most extensively investigated hydrolytically labile polymers for controlled drug delivery applications are hydrophilic polyesters such as poly(glycolic acid), poly(lactic acid) and their copolymers. However, these polymers undergo bulk erosion and the rate of drug release was found to be by diffusion control rather than by polymer erosion rate. Therefore, these matrices do not show well-defined drug release kinetics, particularly for water-labile low molecular weight drugs and high molecular weight peptides and proteins. Also, the stability of water-labile drugs may be compromised in these matrices due to prior interaction of drug with water before release.

Due to the problems associated with bulk eroding polyesters, researchers have focused their effort on polymers that degrade by surface erosion. The surface eroding polymers are highly preferred for drug delivery applications due to the accurate predictability of the erosion process [4]. Currently, polyanhydrides [5] and poly(orthoesters) [6] are the best-known examples of polymers that can be fabricated into surface eroding devices. In surface eroding polymers, the drug is delivered by surface erosion of the polymer matrix and not by diffusion from the interior of the matrix. This contributes to both near zero-order release of the drug as well as enhanced stability of the drug in the polymer.

Polyanhydrides are among the most reactive and hydrolytically unstable polymers developed for controlled drug delivery application [5]. The first application of polyanhydrides as a bioerodable matrix for controlled drug delivery was reported by Rosen et al. in 1983 [7]. Polyanhydrides constitute the only class of surface eroding polymers approved for clinical trial use by the Food and Drug Administration (FDA). Consequently, extensive research has been focused on developing novel polyanhydride-based polymers [8]. Aliphatic polyanhydrides degrade in a few days, while some aromatic polyanhydrides degrade over several years [9]. Several aliphatic–aromatic copolymers have been developed since then with intermediate rates of degradation for a variety of biomedical applications. Thus, the polyanhydrides can be tailored to obtain desired degradation properties via their chemical synthesis [9].

An implantable drug delivery matrix, in addition to providing a predictable kinetics of drug release to maintain therapeutic levels, should also be highly biocompatible. In the case of biodegradable polymers, apart from the potential problems of toxic contaminants or additives leaching from the implant, the toxicity of the degradation products and subsequent metabolites have to be taken into consideration. Therefore, biodegradable polymers should satisfy more stringent requirements in terms of their biocompatibility than non-degradable materials. Extensive biocompatibility evaluations have been conducted with most of the polyanhydrides developed for biomedical applications. Detailed reports on the toxicity evaluations have been published for the most commonly used copolymers such as poly(carboxyphenoxypropane–sebacic acid) P(CPP-SA). This paper initially discusses the influence of polyanhydride structure on degradability, mode of degradation in vitro and reviews the in vivo degradation, elimination, biocompatibility and toxicology of various polyanhydrides used for biomedical applications.

Biodegradation of polymers involves cleavage of hydrolytically sensitive bonds in the polymer that leads finally to polymer erosion. The rate of degradation of biodegradable polymers depends primarily on the reactivity of the hydrolytically sensitive bonds present in it. Polyanhydrides are composed of sparingly water-soluble diacid monomers connected to each other by anhydride bonds. The anhydride bond is hydrolytically very labile and readily splits in the presence of water into two carboxylic acids.

Degradation of the polyanhydride, being a hydrolytically triggered process, depends on the rate of water uptake into the polymer matrix and pH of the surrounding medium. The rate of water uptake depends mainly on parameters like hydrophobicity and the crystallinity of the matrix, the porosity of the matrix and the surface area/volume of the matrix. The higher the hydrophobicity, the lower the water permeability of the matrix. The highly hydrophobic polyanhydrides exhibit ideal surface erosion, since the rate of hydrolytic degradation at the surface will be much faster than the rate of water penetration into the bulk of the matrix.

Some of the important parameters for measuring the degradation of a polymer are molecular weight change, loss of mechanical strength, sample weight loss and change in geometry. All these properties are related but need not necessarily obey the same kinetics. Thus, the molecular weight loss of polyanhydrides can be substantial during the first 12 h in phosphate buffer solution without apparent loss of weight or change in geometry [3].

Leong et al. studied the effect of hydrophobicity of the matrix on the degradation rate of polyanhydrides [9]. They used poly(carboxyphenoxypropane) P(CPP) as the hydrophobic polymer where the aromatic groups imparted hydrophobicity. The hydrophobicity of P(CPP) was subsequently reduced by incorporating various proportions of a hydrophilic monomer, sebacic acid (SA). The degradation studies were performed in 0.1 M phosphate buffer (pH 7.4) at 37 °C. Among P(CPP) and P(CCP-SA) polymers of different composition, the more hydrophobic polymers, P(CPP) and P(CPP-SA) in a molar ratio of 85:15 displayed constant erosion kinetics over 8 months, which can be extrapolated to over 3 years. The degradation rate increases as the composition of sebacic acid (SA) in the copolymer increases. An increase of degradation rate to 800 times was observed when the sebacic acid concentration reached 80%. Therefore, in addition to other parameters, the hydrophobicity of the matrix significantly affects the rate of degradation of polyanhydrides by affecting the rate of water permeability. This was further corroborated by the fact that increasing the number of methylene groups in the backbone of poly(carboxyphenoxy alkane) P(CPA) increases the hydrophobicity of the matrix. A study on the effect of methylene groups incorporated showed that, when the number of methylene groups increases from one to six, the erosion rates underwent a decrease of three orders of magnitude [9]. The effect of chain length on the rate of polymer degradation was also evaluated in vitro by Domb and Nudelman [10] in 0.1 M phosphate buffer at 37 °C, wherein they studied polymers based on natural diacids with general structure –[OOC–(CH₂)_x–CO]_n, where x is between 4 and 12. Polymers based on poorly water-soluble long chain diacids (seven to ten methylenes) lost 20% weight within 48 h, while the short aliphatic chain polymer (four to six methylenes) lost 70% weight during the same period and underwent total hydrolysis within a week as evidenced from IR spectra. This study indicated that the number of methylene groups could be altered to tailor the degradation properties of polyanhydrides based on natural acids.

Mathiowitz et al. studied the effect of matrix porosity on water permeability and degradation of polyanhydrides [11]. They showed that polyanhydride microspheres P(CPP-SA) produced by melt encapsulation were very dense and eroded slowly, whereas when the same polymers were formed into microspheres by solvent evaporation, the microspheres eroded rapidly as they were relatively more porous and hence more water permeable.

Later, the effect of matrix thickness on the rate of degradation of various polyanhydrides, like poly(fatty acid dimer–sebacic anhydride)[P(FAD-SA)] [12], poly(sebacic anhydride) P(SA) [2] and P(CPP-SA) [13], was investigated. In all these cases it was shown that matrices of different thickness with same surface area were found to have similar erosion rates; however, thicker devices generally exhibited longer periods of erosion.

Since hydrolysis of the anhydride bond is base catalyzed, the pH of the surrounding media can significantly affect the rate of degradation of polyanhydride matrices. Polyanhydrides in general degrade more rapidly in basic media than in acidic media [9]. At pH 7.4, pure poly(carboxyphenoxypropane) P(CPP) degrades in about 3 years and the rate of degradation increases with increase in pH, with the polymer degrading in just over 100 days at pH 10.0. The same trend in degradation rate was found for another hydrophobic polyanhydride, P(FAD-SA) having molar ratio of 50:50 [14]. At very acidic pH values, many of the polyanhydrides virtually do not erode at all [9].

The erosion of polymer matrices depends on processes such as rate of degradation, swelling, porosity and ease of diffusion of oligomers and monomers from the matrices. High-performance liquid chromatography (HPLC) assay of the degradation products of polyanhydrides, such as P(CPP-SA) [7], showed that these polymers degrade to their corresponding monomers. The diffusion of oligomers and monomers formed by polymer degradation in turn depends on pH of the surrounding medium and the solubilities of these compounds in the medium. Since anhydrides are cleaved into carboxylic acids, the solubilities of these degradation products will be higher at higher pH and hence erosion will be higher at higher pH [12]. At low pH, these degradation products will be in their unionized form and hence their solubilities will be lower. Domb and Nudelman [10] reported the effect of monomer solubility on in vitro erosion of a homologous series of aliphatic polyanhydrides. The rate of polymer erosion decreases with decrease in the solubilities of the corresponding monomers. Similarly, Seidel et al. evaluated the effect of monomer solubilities on the erosion of poly(anhydride-co-imide) matrix [15]. Using poly(trimellitylimidoglycine-co-1,6-bis(carboxyphenoxy)hexane] P[(TMA-gly:CPH) (30:70)] with monomer solubilities of 6.49×10⁻² mol/l for trimellitylimidoglycine (TMA-gly) and of less than 10⁻⁵ mol/l for p-carboxyphenoxy hexane (CPH) in phosphate buffer at pH 7.4, TMA-gly was completely removed by diffusion through the polymer matrix and only the CPH monomer remained in the device.

Elevated temperatures can accelerate the erosion of polyanhydrides. Studies done at 37 and 60 °C for the erosion of poly(carboxyphenoxy propane) P(CPP) indicated much faster erosion at the higher temperature. In addition to this, Kost et al. reported a strong dependence of degradation rates of polyanhydrides on the application of ultrasound [16]. Up to 5-fold reversible increases in degradation rate was observed for polyanhydrides with the use of ultrasound. This property could potentially be used for the external regulation of polymer degradation in an in vivo setting.

It has been found that the initial molecular weight of polyanhydrides does not have a significant effect on the rate of degradation of the polymeric matrices. Dang et al. reported the degradation of P(CPP-SA) with 20:80 compositions having molecular weight in the range 20,000–110,000 Da in 0.1 M phosphate-buffered saline at 37 °C [17]. Despite the differences in the initial molecular weight of these polymers, weight loss profiles, as well as molecular weight decrease profiles, of all these polymers showed similar trends, as molecular weights of all samples decreased to less than 10 kDa after 1 day of incubation. This was further corroborated by following the morphology of the degrading polyanhydrides. Dang et al. used SEM to visualize the morphological changes during in vitro and in vivo degradation of P(CPP-SA) wafers loaded with the chemotherapeutic agent carmustine (BCNU) [18]. They used polyanhydrides with two different molecular weights 48 and 72 kDa. Even though the lower molecular weight wafers became porous at an earlier time compared to the 72 kDa polymers, the trend in surface morphology changes were the same and very similar surface morphologies were observed after 1 day of incubation in phosphate-buffered saline for both types of wafers.

Since there are no specific enzymes for anhydride bond cleavage, the degradation rate of polyanhydrides is unaffected by enzymes [10].

Thus, a number of parameters including, polymer composition, fabrication techniques, size and shape of the polymer discs, pH of the surrounding media and solubility of the degradation products affects rate of polyanhydride degradation and erosion. Therefore, it follows that the rate of degradation of polyanhydrides can be modulated by carefully tuning one or more of these parameters.