In general, polymer-based drug delivery systems can be broadly classified into physically and chemically controlled systems. Diffusion-controlled drug release is an example of a physical mechanism that is based on the diffusion of a drug molecule through a certain coating or matrix, mostly of polymeric nature. Besides the benefit that the drug is fixed on the implant’s surface and relatively protected against mechanical stress, the thickness and type of polymeric coating allow a wide range of control over the drug release characteristics [111, 112]. In a further iteration of this approach, diffusion-controlled LDD systems can be divided into matrix and membrane systems, where the drug is either directly incorporated within or surrounded by the polymeric coating. Amongst others, these polymer-based LDD systems can be obtained by dip-coating and spray-coating techniques where the intended polymer and drug are either handled in the same solution, which might be a mixture of two solvents if needed, or separately for matrix and membrane systems, respectively. Polymers applied are rather hydrophobic to retard water interpenetration and thereby guarantee sustained drug release. If degradation of the device is desired, hydrolyzable polymers as polyesters with slower degradation than drug diffusion rate can be used for these systems. Table 15.5 provides an overview of polymers used in controlled drug release. Chemically controlled LDD systems offer another strategy for drug release. In these systems, drug release is achieved by degradation or swelling of the drug carrier or by cleavage of chemically bound drugs or biomolecules. For degradation-controlled systems polyanhydrides and poly(ortho esters), possessing highly labile groups that ensure rapid hydrolysis of polymer chains encountering water molecules and retarded water permeation by designing the polymers with hydrophobic monomer units, are used [113]. In contrast for LDD-systems, controlled via swelling, more hydrophilic polymers as e.g. hyaluronic acid are required. However, both systems, diffusion- and chemically controlled, provide drug release independently of their need, possibly leading to ineffectiveness of the LDD system at time of interest. Having this in mind, it would be highly beneficial if the drug is delivered by a system that senses the signal caused by disease, judges the magnitude of the signal and then acts to release the right amount of drug in response [114]. Here, stimulus-responsive hydrophilic polymeric networks as poly(acrylic acid) or poly(N-isopropyl acrylamide) hydrogels with swelling properties dictated by changes in pH [115, 116] or temperature [117–119], respectively, came into discussion. The induction of hydrogel swelling by such physical and chemical stimuli allows control over release of incorporated or underlying drugs. However, the cited stimuli are rather non-specific. An alternative are specific chemo-responsive hydrogels, which are designed to contain internal non-covalent interactions based on pendant ligands and complementary receptors [120]. These crosslinks can be broken by soluble ligand or receptor competitors, which are specific for certain diseases and can diffuse into the gel displacing the internal affinity crosslinks. Subsequently, the degree of crosslinking decreases and drugs can be released. Besides glucose-responsive hydrogels with concanavalin A—glucose affinity crosslinks [121], antigen-responsive hydrogels, where antibody–antigen pairs, chemically grafted to the hydrogel network serve as reversible crosslinkers, are discussed in literature. Several model systems have been established [120, 122, 123], while the best studied is probably the one of Miyata et al. [124–127], which is based on immunoglobulin G (IgG) antigens and corresponding antibodies grafted onto a poly(acrylamide)-(PAAm-)hydrogel, forming affinity crosslinks cleavable in the presence of free antigen. In an own study, we provided a thorough in vitro characterization of PAAm-hydrogels regarding drug permeability with the purpose of exploring the applicability of such hydrogels for implant-associated LDD systems [128]. Within this study, we were furthermore able to demonstrate an antigen-triggered release of bovine serum albumin as model drug from a model implant-associated LDD system. The continuation of work will now focus on the real implementation of such stimulus-responsive hydrogels to implant surfaces and the use of natural biocompatible polymers, particularly those in the poly(saccharide) family (e.g. hyaluronic acid and chitosan).

LDD entered the field of cardiology with the development of drug-eluting stents (DES) which revolutionized the treatment of coronary artery disease. Stents can generally be described as tubular scaffolds from a wire mesh or a slotted tube, which are either self-expanding or balloon-expandable. Typical biomaterials for stent manufacture are stainless steel and cobalt-chromium alloys for balloon-expandable stents, and nickel-titanium alloys (nitinol) for self-expanding stents [129]. DES are specialized vascular stents which allow the delivery of drugs in a controlled manner to arterial wall with the purpose to reduce or prevent in-stent restenosis as process of enhanced smooth muscle cell proliferation. As described the most DES consist of a polymeric matrix as drug carrier, which controls the drug release [130, 131]. Drug carriers used on the initial commercially available DES were permanent polymers as a blend of PEVA and PBMA in ratio of 50/50 % (w/w) (Cypher stent, Johnson & Johnson, USA) and the copolymer SIBS (Taxus stent, Boston Scientific, USA). These polymeric coating materials however raised the question about long-term biocompatibility. Therefore, the activities in DES development headed for more biocompatible and biodegradable materials as PLA [38, 132, 133] and copolymers, such as PLGA [134, 135], which are fully metabolized to carbon dioxide and water (Krebs cycle). These polyester-based drug reservoirs have the advantage that they degrade during or after the drug release so that they do not cause long-term foreign body reactions. In addition, degradable polymer blends of P(4HB) and PDLLA have been developed. Recently several design approaches to provide temporary absorbable stents, which may allow for restoration of vasomotion and long-term positive remodeling of the stented vessel [14], with an LDD-function are discussed. The first clinically available stent is equipped with a PDLLA-coating (BVS, Abbott, USA).

Due to the low abundance of available functional groups on the surface of most polymers, their surface activation is often a precursor to performing biofunctionalization in the sense of chemically attaching a bioactive compound. Techniques that modify surface properties by introducing random, non-specific groups or by coating the surface are less useful in bioconjugation to polymer surfaces. Used surface activation techniques should in contrast be tailored to introduce a specific functional group, which might be achieved amongst others by wet- and plasma-chemical processes, detailed in the following and in Table 15.6.

Wet–chemical surface activation. In wet-chemical surface modification, a material is treated with liquid reagents to generate reactive functional groups on the surface. In contrast to plasma and other energy source surface modification techniques, this method has a general higher penetration depth and therefore modification of three-dimensional scaffolds within pores is often feasible in a simple way [136]. For instance, oxygen-containing moieties were introduced to PE [137, 138] and PP by treatment with chromic acid and potassium permanganate in sulfuric acid [139, 140]. Carboxyl and hydroxyl groups were generated by base and acid hydrolysis of PMMA [141], and PCL [142] with concentrated sodium hydroxide and sulfuric acid. One has to note that hydrolysis time of polyesters needs to be thoroughly optimized in order to retain the properties of the bulk polymers. Subsequent conversion of the functional groups to isocyanate groups by treatment with 4,4′-methylenebis(phenyl isocyanate) allowed formation of terminal amino groups by addition of ammonia. In an own study, we have already applied this activation procedure for the immobilization of acetylsalicylic acid and vascular endothelial growth factor (VEGF) to PCL [142]. Generation of primary amines is also possible by wet-chemical treatment without the intermediate step of hydrolysis. For instance aminolysis with various diamines including hydrazine hydrate, 1,6-hexanediamine, ethylenediamine, and N-aminoethyl-1,3-propanediamine, as well as lithiated diamines [7, 19, 22–28] has been applied to introduce amines to PMMA [143, 144], PUR [145], PLLA [146, 147], and PLGA [148]. The simple these methods are, they are however non-specific and the degree of surface functionalization might therefore not be repeatable between polymers of different molecular weight, crystallinity, or tacticity. Moreover, hazardous chemical waste might be generated [149]. A clean alternative are ionized gas treatments, first of all plasma-chemical modifications, described in the following.

Plasma–chemical surface activation. In contrast to the above described wet-chemical methods plasma treatments can provide modification of the top nanometer of a polymer surface without using solvents or generating chemical waste and with less degradation and roughening of the material than many wet-chemical treatments [136]. The type of inserted functional groups can be varied by selection of the plasma gas (Ar, N₂, O₂, H₂O, CO₂, NH₃) and operating parameters (pressure, power, time, gas flow rate) [136]. For instance, surfaces of PCL [142, 150], PLLA [147], PA [18], PE [137], and PET [151] were provided with oxygen containing functional groups by oxygen plasma exposition. Moreover, carbon dioxide plasma could be used to introduce carboxyl groups to PP [152], PS [153], and PE [152]. The generation of amine groups was in addition performed by ammonia plasma to PTFE [154] and PS [153]. In own studies we treated PCL [142] and PLLA [147] by ammonia plasma and qualified the generation of amine groups via fluorescent labelling. Results reveal less functional groups on these surfaces in comparison to chemically activated surfaces, which is in good agreement with the well-known low plasma penetration depth of only 10 nm. However, no correlation with the efficiency of subsequent immobilization of biomolecules could be observed [155]. Besides a standalone technique, plasma treatments are often used as a precursor to other surface modification techniques, as for example plasma activation followed by silanization, as in detail described in the following.

Surface activation via silanization. Although surface modification using silane monolayers is primarily applied on glass and silicone substrates as a means to couple an organic polymer, this is certain to be an emerging field in research for polymer surface modification [136]. Because the target surface functional group in glass or silicon silanization is a hydroxyl, silanization can be applied to polymer surfaces, which have been hydroxylated by for instance treatment with oxygen plasma. As organosilanes with various end functionalities including amine, epoxy and vinyl groups are available, silanization offers the potential of modifying polymer surfaces with a bunch of functional groups for different application. Moreover, self assembled monolayers, produced by silanization, provide a nearly crystalline and hence more defined surface functionalization than typical wet-chemical or ionized gas functionalization techniques [156]. In own studies we used organsilanes with epoxy and amine head groups for the chemical attachment of hyaluronic acid to PA [18] and VEGF [142] and antibodies [147] to polyester surfaces, respectively.

Among the numerous reported methods for immobilizing biomolecules to surfaces physical adsorption via van der Waals or electrostatic interactions, physical entrapment within hydrogels, ligand–receptor pairing (as in biotin–avidin) and covalent attachment via cleavable or non-cleavable bonds are most commonly used (Fig. 15.1). In dependence of the applied immobilization protocol a short-term and long-term localization of the biomolecule on the implant surface can be achieved [110].

Thus, biofunctionalization can be either used to provide the implant with a drug delivery or a stable novel surface functionality. In an own study, we could show this wide range of modification possibilities by VEGF immobilization and delivery from PLLA surfaces [157]. Observed release profiles differed in dependence of the applied immobilization method. Fastest release was observed for VEGF entrapped in a PAA hydrogel, followed by physically adsorbed VEGF and covalently attached VEGF via hydrolyzable bonds with the slowest release rate. The release of physically adsorbed or entrapped VEGF is diffusion-controlled and starts immediately after insertion into the elution medium. Chemical attachment however affords the cleavage of covalent bonds prior to liberation of biomolecules. This was achieved in that case by the formation of ester bonds between the side chain hydroxyl groups of VEGF and N,N-disuccinimidyl carbonate-activated surface amino groups on the PLLA surface.

As detailed above most of biofunctionalization methods afford a previous surface activation, which might be combined with the grafting of an intermediary spacer between the surface and the biomolecule for multiplication of available functional groups on the surface. Here, multifunctional spacers as poly(ethylenimine) or dendrimers with a wide range of terminal functional groups in defined quantity offer a way to increase this surface functionality [157, 158]. In addition, maintenance of biological activity is very important for hydrophobic polymer surfaces as these are often associated with non specific adsorption and denaturation. Besides the integration of hydrophilic PEG spacer [159], which effectively shield the compound from these issues, we investigated in an own study the inclusion of the peptide spacer (GGAP)₄ on PLLA surfaces. By this we reached efficient prevention of non-specific protein adsorption and enhanced bioactivity of subsequently covalently attached anti-CD34 antibodies [147].

A prerequisite for successful biomolecule immobilization is of course moreover that the specific functionality imparted to the inert surface must be compatible with the reactive sites on the compound to be covalently attached to that surface. While for adsorption processes the introduction of charged groups or simple modification of hydrophilic properties should be enough, covalent immobilization affords defined functional groups, including thiol, aldehyde, carboxylic acid, hydroxyl, and primary amine groups. These can be directly used for biomolecule attachment. However, most cases apply additional crosslinking agents, which can link the bioactive compound directly to the functionalized substrate (zero-length crosslinkers), or introduce themselves a spacer of several angstroms. Moreover, they can be stable or introduce hydrolyzable bonds in order to allow for biomolecule delivery from the surface. A detailed description of these agents and the associated chemistries can be found in Bioconjugate Techniques, by Hermanson [160].

In the field of cardiology, intensive research on biofunctionalization of polymers has been done since decades to develop novel surface designs for all types of endovascular implants in order to modulate their behavior in contact with blood and improve their bio-and hemocompatibility. Even prior to the development of DES this issue has been addressed by the establishment of stents, being coated with the anticoagulant heparin (e.g., Palmaz-Schatz stent [161, 162], Wiktor stent [163], and Jostent [164]), fibrin [165], cellulose [166], and phosphorylcholine (BiodivYsio stent [167]). Nowadays more attention is however drawn to the biofunctionalization of the polymer coating enabling activation of a cascade of biological processes that eventually regenerate or replace a functioning endothelium, referred to as prohealing approach. Materials promoting in situ endothelialization of vascular implants by capturing in vivo circulating endothelial progenitor cells (EPC) with the capacity to differentiate into mature endothelial cells (EC) and repopulate areas of vascular injury and endothelial disruption [168, 169]. The selection of the optimal target on the EPC surface in combination with an ideal capture molecule immobilized on the implant surface, which is decisive for the success of in situ endothelialization, is however still a challenge. A stainless steel-coated stent with a polysaccharide matrix covered by murine monoclonal anti-human antibodies directed towards the EPC surface marker CD34 is probably the most studied “EPC capture stent” nowadays. This Genous R-stent, developed by the company Orbus Medical Technologies in collaboration with Kutryk’s group, has achieved CE approval and is commercially available since 2010. Although CD34 is not a specific marker of EPC, as only a small portion of the CD34-positive cells are actually EPC [170], first human clinical investigations of this technology demonstrated its safety and feasibility for the treatment of de novo coronary artery disease [171] and the long-term promotion of significant late regression of neointimal hyperplasia [172, 173]. Unfortunately, in spite of these initial promising clinical outcomes, following randomized clinical trials comparing the Genous R-stent with BMS (GENIUS STEMI) [174] and DES (TRIAS) [175–177] evidenced disappointing results with regard to late luminal loss. These results let to a general discussion of the Genous’ clinical efficacy. As reaction to this discussion, novel surface modification strategies in order to increase availability of immobilized compounds and thereby biofunctionality of the modified surface and more selective capture molecules are searched. In this context, further antibodies than CD34 (e.g., anti-KDR) [178], against EPC selected aptamers [179], and materials mimicking the extra cellular matrix (ECM), are in focus of current research. From the plethora of insoluble and soluble macromolecules contained within ECM, the fibrillar protein collagen is probably the most commonly used coating on for example commercially available polyester vascular grafts [180]. Furthermore, besides the use of integer ECM components, various ECM peptide sequences as functional domains of proteins, glycoproteins, and proteoglycans have been isolated and grafted on biomaterials to control cell behavior, for example, REDV [181], PHSRN [182], RGD, and GRGDSP from fibronectin [183], laminin-derived recognition sequences, IKLLI, IKVAV [184] LRE, PDSGR, RGD, and YIGSR, and collagen type I derived sequence, DGEA [185]. However, these peptides, nicely reviewed by De Mel et al. [186], are not cell-selective as they bind to integrins, present on many cell types. To overcome this risk of low cell selectivity, Veleva et al. [187] used a combinatorial peptide library and phage display technique to isolate peptides, being more selective for human blood outgrowth endothelial cells (HBOEC). Indeed, selectivity of terpolymers, covalently coupled with these peptides, could be demonstrated for HBOEC in serum-free medium. The application of a coating combing both, an integer ECM component with functional peptide domains, might hence be very promising with regard to the creation of an optimal cell environment and thereby suppression of the foreign body reaction.

Cell interactions with polymers are usually studied using cell culture techniques. In vitro evaluation of surface materials by directly seeding endothelial cells (ECs) and smooth muscle cells (SMCs) onto the biopolymers represents a common procedure to assess biocompatibility and cytotoxicity as well as cell morphology [203–205]. In this context, most in vitro studies analyzing the impact of material on tissue cells involve experiments under static conditions. However, in healthy vessels especially ECs are exposed to the blood stream. This pulsatile, laminar blood flow exerts shear stress on vascular endothelium, which induces anti-apoptotic signals and maintains an anti-inflammatory and non-thrombotic endothelial phenotype compared to ECs cultured under static conditions [206, 207]. Cell culture experiments represent a model in which one can study specific mechanisms involved in the biologic responses of blood and tissue cells to materials. During the last decades a variety of techniques has been developed to investigate vascular biology under hemodynamic conditions.

Parallel Plate Flow Chamber. The parallel plate flow chamber (PPFC) is a design most frequently used to study vascular cells under flow conditions. Here, cell culture media is circulated through the chamber at adjustable flow rates creating a defined laminar flow (Fig. 15.2).

A typical flow chamber consists of a polycarbonate corpus (bottom), a silicone gasket, which defines height and length of the flow channel (side walls) and a glass cover slip (top), which can be replaced by polymer foils. Under static cell culture conditions, the coverslip can be seeded with the cells of interest prior to the flow experiment. Additionally, these flow chambers are designed to allow fluorescence microscopic observation of a perfused cell layer. In this context PPFCs are suitable to measure the kinetics of cell attachment, detachment, and rolling on surfaces under flow conditions. Therefore, a range of studies refers to the adhesion of leukocytes and platelets on endothelial cells [208, 209]. Recently, Busch et al. [193] published a study which assessed a range of endothelial parameters in interaction with different stent materials using an in vitro perfusion system. However, these kind of experiments are limited by the rectangular geometry of the flow channel and do only allow for the investigation of cell monolayers over several hours.

Live Cell Imaging. To investigate the influence of the substrate regarding endothelial cell morphology in particular, it might be advisable to culture endothelial cells under flow conditions for several days. For that purpose, a specially designed sticky–slide (ibidi) can be utilized [210]. Different substrates can be mounted to the bottomless channel slide, which has been developed for microscopic applications. The slide can then be connected to the ibidi pump system and kept sterile for several days in a conventional incubator. Additionally, for example the cytoskeleton can be visualized and changes in morphology due to the substrate can be monitored by microscopic methods.

Tube Chamber System. To overcome the limitations of 2D cell culture models vessel-simulating bioreactors have been developed. Typically, a 3D bioreactor consists of a graft with an individual culture media reservoir to allow a transversal gas and media exchange between the perfused graft lumen and the outer chamber volume which simulates the interstitial tissue liquid. The porous graft (e.g. silicone, PTFE) is seeded with cells and the culture media is perfused through the construct. This procedure allows the in vitro cultivation of endothelial cells under fluid flow and results in a synthetic vessel [210, 211]. A special modification of a vessel-simulating flow-through cell has been developed by Neubert et al. and is particularly suitable for the evaluation of drug release and distribution from drug-eluting stents [212–214].

To investigate cell material interactions, cultured cells are seeded on the polymer surfaces of interest and the extent of cell adhesion, proliferation, and viability is measured by standardized assays. Although in vitro experiments do not entirely reproduce the whole range of cellular responses to the implanted material, these methods provide a level of quantification which cannot be easily obtained in vivo.

Cell adhesion. For most applications the adhesion of cells is of fundamental interest being the initial event of cell attachment. A simple method to quantify adherent cells is to incubate cells over a surface for a period of time subsequently followed by the detachment of loosely adherent cells by gently washing the surface. Remaining cells can be labeled by fluorescent dyes and quantified using fluorescent measurements. Cell adhesion assays are also frequently used to assess monocyte or platelet adhesion on polymeric surfaces [215, 216]. In this context Hezi-Yamit et al. [217] did show that polymer hydrophilicity should be considered as a parameter to assess the biocompatibility of polymer surfaces. They showed that hydrophobic polymers such as PBMA or SIBS promote the adhesion of inflammatory activated monocytes while more hydrophilic polymers (e.g. PC (Phosphorylcholine) polymer) lead to less pro-inflammatory responses.

Cell proliferation. Endothelial growth is an essential factor after implantation. A continuous and healthy endothelial layer only prevents inflammatory and thrombotic events on the vessel wall. Besides, a healthy endothelium is crucial for regulating smooth muscle cell growth [218, 219]. Cell proliferation can be assessed by commercially available BrdU ELISA assay kits. On the basis of the study by Busch et al. [193] proliferation of ECs on hydrophobic permanent polymers (PEVA, PBMA) as well as biodegradable polymers [PLLA, P(4HB)] is reduced, whereas the polymeric blend PLLA/P(4HB) promotes EC growth. Data from this study do also show that SMC proliferation is influenced in the same way as EC growth and none of the polymers tested is able to enhance endothelial proliferation by simultaneously reducing smooth muscle cell growth.