Materials for Controlled Release of Local Anesthetics

Abstract

Controlled release systems for prolonged duration local anesthesia have long been an area of research interest, and now are entering clinical practice, in part driven by the opioid epidemic. We discuss the design considerations and material properties of systems for controlled release of local anesthetics, from relatively simple systems to covalent binding of drugs to materials and delivery triggered by external stimuli.

1. Introduction

Local anesthetics, such as used by a dentist to numb up one’s teeth, differ from general anesthetics as they are generally injected at the desired (local) site of action, for a localized effect. Such drugs work by interrupting the conduction of electrical impulses along nerves, preventing the sensation of pain (and other nerve functions such as control of movement) distal to the site of injection. Local anesthetics are effective, but have a relatively brief duration. There has long been interest in using drug delivery approaches to extend the duration of effect of a single injection to cover the duration of most perioperative pain, i.e., 2–3 days. Initially, that interest was primarily derived from the desire to improve the perioperative experience. There were also potential physiological advantages to prolonged local anesthesia. For example, if patients’ intercostal nerves could be numbed up after chest surgery, perhaps pain from the incisions would not impede breathing, and they would not require as much narcotics, which depress respiratory drive. Interest in prolonged duration of local anesthetics has been heightened by the opioid epidemic in the US. Therefore, there is a desire to develop non-opioid sustained release systems that would provide local anesthesia throughout the perioperative period – so called “opioid sparing” therapies.

As in other settings, the primary function of encapsulating local anesthetics has been to alter their pharmacokinetics. The maintenance of therapeutic levels of anesthetics at the nerve over an extended period will prolong the duration of effect. One subtle point is that it is not the total duration of release, but the duration of release above a given level that is critical. Thus, accelerating release can in some cases extend duration of effect. Control over release rates can also mitigate systemic toxicity.

2. Design Considerations

Ideally, formulations for attenuated release should require a single injection to achieve the full duration of effect, be easy to administer to patients, and not require general anesthesia or surgery to initiate. They should be biodegradable, have an acceptable inflammatory response, and cause no local tissue injury or systemic toxicity.

Sustained release systems may be used in most of the ways that unencapsulated local anesthetics are used. These include: infiltration, peripheral nerve block, plane/fascial block, and neuraxial block. In practice, they would likely be primarily used for relatively localized infiltration and peripheral nerve blocks. Use in the other blocks would either involve delivery of very large volumes – which might (but not necessarily) lead to systemic toxicity – or would ablate nerve conduction to and from a large part of the body for an extended period. The latter would also entail motor block, and potentially loss of bowel and bladder control, etc.

3. The Payload

The materials to be used with drug delivery systems for local anesthetics are in part dictated by the chemistry of local anesthetics themselves (Figure 1). In general, conventional local anesthetics (CLAs) are small molecules (~300 Da) with three main components: a variably substituted aromatic group, a tertiary amine, and an ester or amide linkage connecting the two. The aromatic end imparts hydrophobicity. The linkage determines whether the CLA is degraded by plasma cholinesterases (esters) or by the liver (amides). The tertiary amine is crucial in determining the solubility of the CLA. At acidic pH, it is protonated and water soluble. At alkaline pH, the tertiary amine is deprotonated and the CLA in its free base form is hydrophobic, practically insoluble in water but highly soluble in many organic solvents. CLAs properties are determined by the equilibrium between their hydrophilic and hydrophobic states. Most CLAs have pK_a values in the range 7.6–8.9, and are mostly protonated (hydrophilic) at physiological pH (7.4).

Figure 1.

Chemical structures amino-ester and amino-amide local anesthetics.

These properties are crucial for the drugs’ pharmacological functions. CLAs diffuse through body water from the site of administration to the axonal surface. (Their hydrophobicity helps them penetrate any intervening biological barriers.) There, they partition into the cell membrane in hydrophobic form, then partition into the relatively acidic inside of the cell in the protonated from. The voltage-gated sodium channel is then blocked from the inside of the cell.

This equilibrium between the hydrophilic and hydrophobic forms of these local anesthetics is important for the incorporation of CLAs in drug delivery systems. The polymeric particles discussed below are often created by a single emulsion system. The CLA free base is dissolved in an organic solvent along with a given polymer. The mixture is dispersed into droplets in water, forming an oil-water emulsion. CLA molecules near the droplet surface can partition into the water, leading to loss of drug, i.e., lower loading efficiency and decreased loading. Loss of drug can be minimized by alkalinizing the external aqueous phase of the emulsion, maintaining the CLAs in the free base form and thus encouraging them to remain partitioned within the organic solvent within the emulsion droplets.

4. Drug Delivery Systems for Conventional Local Anesthetics

Below we provide a few examples of types of materials used to provide prolonged local anesthesia using relatively conventional drug delivery systems (Figure 2). The CLAs selected are generally the most potent, given that only a finite amount mass can be injected in the clinical setting; the drug selected is often bupivacaine, a local anesthetic in common clinical use. Fortunately for some types of drug delivery systems, the potency of CLAs generally correlates with their hydrophobicity, though this is not true of all local anesthetics (the site 1 sodium channel blockers, such as tetrodotoxin, are extremely potent and very hydrophilic).

Figure 2.

Selected drug carriers for the delivery of local anesthetics.

4.1. Polymeric particles

Polymeric particles for delivering CLAs have been described for decades.^[1] In general, biodegradable and biocompatible polymers are used. These polymers vary but generally are relatively hydrophobic. We will use the α-hydroxy acids as examples of this class of polymers. These polymers are composed of lactic and glycolic acids connected by hydrolysable ester bonds, which render them biodegradable. The ratio of lactide to glycolide determines many of the properties of this polymer, with more lactide leading to greater hydrophobicity, slower degradation, slower drug release from microspheres, and greater stiffness. The polymers also can vary in molecular weight, with greater molecular weight associated with slower particle degradation and slower drug release. In many cases, the rate of effective release of local anesthetics (i.e., release resulting in nerve block) from particles made with such polymers is primarily dependent on diffusion from the matrix, with the particles often being present in tissue long after cessation of clinical effect of a nerve block.^[2]

While polymer properties can affect drug loading and release, particle size has a very marked impact. In general, larger particles have greater drug loading and slower release, due in part to their larger surface area-to volume ratio.

CLAs are usually encapsulated in such particles in the hydrophobic free base form. Polymeric particles can achieve very high loadings of CLAs:>70% w/w.^[3] In vitro, such particles, made with a high molecular weight PLGA, release bupivacaine for approximately 3 weeks at physiological pH. However, when injected at the sciatic nerve in rats, they provided 6–12 h of nerve block. Co-encapsulation of potentiating or synergistic compounds^[4] such as dexamethasone^[5] or tetrodotoxin^[6] increased duration of block to 3–4 days in the same animal model. Unfortunately, tissue exposure to high concentrations of CLAs for extended periods is associated with tissue injury.^[7]

4.2. Liposomes

Liposomes have also been used for decades to deliver CLAs. Liposomes are composed of lipid bilayers separated by aqueous layers. They can range from 40 nm–30 μm in diameter. They can be univesicular (usually when smaller) or multivesicular, and unilamellar or multilamellar. Since CLAs are amphiphilic, they can partition into the aqueous or lipid components.

Local anesthetics can be loaded into liposomes by two standard methods, passive loading or active loading (Figure 3).^[8] Passive loading uses a drug-containing aqueous solution to hydrate a dry lipid film or lipid cake. Active loading, also known as remote loading, uses a pH gradient or ionic gradient across the liposomal membrane to encapsulate drugs into liposomes. Active loading methods often result in higher loading efficiency and reduced drug leakage compared to passive loading.^[9]

Figure 3.

Schematic of drug loading into liposomes. a) In passive loading, drugs are encapsulated into liposomes upon hydration. b) In active loading, drugs are loaded into preformed liposomes driven by a transmembrane pH or ionic gradient.

Liposomes have entered local anesthetic clinical practice, in the form of “giant” liposomal bupivacaine under the trade name Exparel. Exparel was formulated using the DepoFoam technology, which encapsulates bupivacaine hydrochloride in multivesicular liposomes.^[10] The liposome bupivacaine formulation releases the drug over a desired period without altering the drug molecule. In rat studies, we have found Exparel has a duration of rat sciatic nerve block of 4–8 h.^[11] In humans it is claimed to provide much longer local analgesia.^[12] However, its advantages over non-liposomal bupivacaine have been disputed.^[13]

4.3. Hydrogels

Hydrogels are three-dimensional networks of polymer chains capable of holding large amounts of water molecules.^[14] These have been used extensively to prolong nerve blockade from CLAs, but the prolongation achieved is generally much shorter than that achieved by more hydrophobic systems. This is partly because water has such ready access to the CLAs, which are usually incorporated in their salt (protonated, hydrophilic) form, and the hydrogel is very porous, allowing rapid drug egress. Attempts to slow release from hydrogels have included: electrostatic interactions between gel matrix and CLAs (which are cationic when charged), for example, by using anionic polymers (e.g., hyaluronic acid); cross-linking the polymers of the hydrogel^[15] and other means of increasing viscosity – which slow diffusion.

4.4. Macroscopic systems

The above systems are injectable. To be injectable through a needle with clinically relevant dimensions, drug delivery systems must either be fluid and relatively non-viscous, or small (micro- or nano-scale). Larger systems can be designed,^[16] but would require implantation, which is much less desirable than injection in the majority of clinical applications.

4.5. Covalent modification

The duration of nerve blockade from a given formulation does not rely just on the duration of release, but on the duration of release above a given therapeutic level. However, upon injection near a nerve, most formulations release drug relatively rapidly. This “burst release” often is higher than the amount required for nerve blockade, and may contribute to tissue or systemic toxicity. Covalent linkage of the local anesthetic to a polymeric backbone^[17] can achieve prolonged release, provided that the linkage is labile (e.g., an ester) so that drug is released, ideally in its native form, in sufficient quantity to achieve nerve blockade but without burst release. With covalent modification, the primary determinants of drug loading and release differ from those for conventional drug encapsulation. Here, drug loading is determined by the ratio of the molecular weight of the local anesthetic to that of the polymer, and the rate of release depends on the rate of cleavage of the covalent linkage. Polymer properties can still have a marked effect on retention in tissue; the duration of tissue retention is generally the upper limit of the potential duration of clinical effect.

4.6. Pharmacological approaches

Co-encapsulation of drugs that have potentiating or synergistic effects^[4] can dramatically prolong duration of effect of CLAs (Table 1). Some of the adjuvantic drugs are local anesthetics themselves, such as the site 1 sodium channel blockers (e.g., saxitoxin^[18] and tetrodotoxin^[19]) and the transient receptor potential cation channel subfamily V member 1 (TRPV1) receptor agonist capsaicin.^[20] In general, the beneficial interaction is between compounds of different classes. Others adjuvants have no intrinsic local anesthetic effect, such as dexamethasone,^[5-6] or dexmedetomidine.^[21] Both of those molecules have a number of potential mechanisms of action, both are also vasoconstrictors. Local vasoconstriction can slow the uptake of drugs into the blood stream, maintaining the concentration of local anesthetics at the nerve for longer. Nerve block from microspheres co-delivering bupivacaine with dexamethasone^[6] lasted 3–4 days in rats, and 13 days in sheep.^[22] (Longer durations can be achieved in larger animals as a larger dose can be delivered with less toxicity.) Blocks lasting at least 4 days were achieved with this formulation in humans,^[23] albeit with problems with biocompatibility.^[7] Durations of nerve block lasting a week from a single injection in rats have been described with liposomes co-delivering saxitoxin and dexamethasone.^[18b]

Table 1.

Co-delivery of adjuvants with local anesthetics.

Adjuvant	Classification	Drug Carrier	Local Anesthetic	Refs.
Dexamethasone	Glucocorticoid receptor agonist	Microparticle	Bupivacaine	[5,22-24]
S1SCBs[a,b]	Site 1 sodium channel blocker	Microparticle	Bupivacaine	[6]
Dexamethasone	Glucocorticoid receptor agonist	Liposome	Bupivacaine	[25]
Dexmedetomidine	Alpha-2 adrenergic agonist	Liposome	Bupivacaine	[21b,25]
Clonidine	Alpha-2 adrenergic agonist	Hydrogel	Ropivacaine	[26]
Capsaicin[b]	TRPV1 agonist[c]	Liposome	Tetrodotoxin	[20b]

^[a]S1SCBs: site 1 sodium channel blockers.

^[b]S1SCBs and capsaicin are also local anesthetics.

^[c]TRPV1: transient receptor potential cation channel subfamily V member 1.

4.7. Triggered systems

Nerve block with CLAs generally entails motor block (i.e., immobility) along with the desired analgesia. Moreover, current single-administration prolonged duration local anesthesia systems provide a given block duration and intensity that cannot be modulated in response to changing patient needs. As research into systems that provide prolonged duration local anesthesia have progressed, and material science-driven fields such as remotely-triggered drug delivery systems have evolved, attention is being given to addressing some of these downsides of prolonged nerve block.

As noted above, formulations have been developed that can provide quite prolonged nerve block. Unfortunately, there is no way at present to discontinue nerve blockade should the patient so wish, or to moderate the degree of analgesia. An alternative would be to have a shorter but still prolonged duration of effect, followed by an extended period where the patient could trigger the system to provide additional analgesia.

One example of such a triggered system is liposomes that have been engineered to modulate the permeability of lipid bilayers in response to an energy source external to the body (Figure 4), so that on-demand drug delivery can be achieved. Liposomes have been modified with gold nanorods so that irradiation with near infrared (NIR) light would heat the nanorods via the photothermal effect, increasing lipid bilayer permeability, resulting in increased drug flux and local anesthesia.^[27] Similar results have been achieved with photodynamic approaches. In liposomes made with unsaturated lipids and incorporating photosensitizers, reactive oxygen species can be released upon irradiation, which peroxidizes the unsaturated lipids, increasing bilayer permeability and resulting in block.^[28] Similar results can be achieved with ultrasound^[29] and other types of energy sources. Modulation of the energy (irradiance and/or duration) of the energy source allows adjustment of the intensity and duration of local anesthesia. One problem with such approaches is that peripheral nerves can be many centimeters below the skin, and the energy of the trigger is attenuated as it travels to the nerve. Attenuation depends on distance, type of energy, and type of tissue.^[30] With some energy sources such as light, simply increasing the irradiance may be effective, but risks thermal injury. The sensitivity of the drug delivery system to the energy source is therefore critical,^[30] as is the potency of the payload drug^[21b] (a more potent payload will yield more effect per unit of irradiation).

Figure 4.

Light-triggerable local anesthesia in the rat footpad following injection of thermo-responsive liposomes coated with gold nanoparticles. Local anesthesia is presented as % maximum possible effect. Reproduced with permission from Ref. [27]. Copyright 2015 American Chemical Society.

Many drug delivery systems – for example, the polymeric microspheres and liposomes discussed above - constantly release drug. Duration of effect is determined by the point at which the rate of release drops below what is needed to maintain a therapeutic level. With triggerable systems, this constant release can be problematic. First, drug release happens whether or not the nerve block is desired, not only during triggered events. Second, the constant release of drug depletes the system of payload that could be used in triggered events. Both problems can be addressed by linking the drug to a polymer with a covalent linkage that is stable unless cleaved by an external energy source.^[31] Covalent linkage to a polymer inactivates the drug, so that there is no drug release or effect in the untriggered state, and release only happens when triggered.

5. Biocompatibility

Perineurial injection of drug delivery systems invariably results in inflammation, which will last as long as the drug delivery system is present (Figure 5).^[2b] This is a specific case of a generic response to injected biomaterials.^[32] It is not clear that such inflammation is hazardous.^[33] However, it is fairly clear that the encapsulated CLAs can be injurious to tissues such as nerves^[34] and adjacent muscles^[35]. Sustained release of CLAs can result in exposure of tissues to CLAs at high concentrations for extended periods, exacerbating their toxicity^[36] and the viability of the drug delivery system.^[7] For this class of compounds, biocompatibility of the system is crucial.

Figure 5.

Biocompatibility of bupivacaine-containing microparticles 4 days after injection at the rat sciatic nerve. Top panels: 60 μm PLGA microspheres (MS). Bottom panels: 60 μm lipid-protein-sugar particles (LPSPs), Reproduced with permission from Ref. [2b]. Copyright 2011 Wiley-VCH.

Although the tissue reaction to injected or implanted biomaterials has been studied for decades,^[32] it difficult to point to proven relations between material type and outcome, particularly in the perineurial space. The relation of material type to tissue injury per se (vs. inflammation) is even less well understood. In some tissues, there is some evidence to suggest that materials that degrade more rapidly might cause more tissue effect, as might some additives.^[37]

The specific location where materials are deposited is important. For example, PLGA microparticles injected in the peritoneum cause peritoneal adhesions;^[38] this is obviously not an issue in many other locations. Whether a tissue is healthy (e.g., how well vascularized it is) might affect its susceptibility to injury as well.

Particle size can affect how long materials are retained in tissues, both because larger particles will take longer to degrade and because smaller particles are more likely to leave the site of injection.^[38] Longer retention times entail longer durations of effects on tissues. Particle size also affects the type of inflammatory reaction. Particles that are too large to be phagocytosed will be walled off by a syncytium of monocytes, so-called giant foreign body cells.^[39] The total mass of material delivered likely has an effect (the greater the mass, the greater the inflammation).

6. Summary and Perspective

Despite decades of research into sustained release systems to prolong local anesthesia, few have entered clinical practice. The commercial success of Exparel underscores that there is a substantial unmet need for such formulations.

There are numerous compounds that either have local anesthetic properties, or that can potentiate local anesthesia. Their co-encapsulation can extend duration of effect, as seen above, or reduce local or systemic toxicity by reducing the amount of CLA needed to have a given effect. For many of these, encapsulation strategies will be similar to CLAs in that they are hydrophobic. One exception to this is the site 1 sodium channel blockers, such as tetrodotoxin and the saxitoxins, which are ultrapotent local anesthetics^[40] that bind a site on the voltage-gated sodium channel on the outside of the cell. These small molecules are very hydrophilic and – like many small hydrophilic molecules - are difficult to encapsulate. They are not currently in clinical use as local anesthetics.

Avoidance of adverse tissue reaction will be important. Using biocompatible biomaterials obviously will be helpful. Avoiding excessive local CLA levels will minimize tissue toxicity. It is not clear that selection of CLA to encapsulate causes a large difference to tissue toxicity.^[41] However, the site 1 sodium channel blockers, despite their considerable potential systemic toxicity, have very little tissue toxicity, even when encapsulated.^[18b]

To improve on-demand (triggered) local anesthetic systems, it will be important to maximize the depth at which systems can be triggered, without incurring thermal injury. One design problem for photosensitive systems is that wavelengths of light than can penetrate relatively deeply into tissue (e.g., NIR I, II) also have low energies and so cannot, for example, break photolabile covalent linkages. Co-delivering molecules such as photosensitizers (to produce reactive oxygen species to induce drug release events) may be complicated, depending on the system. Using upconversion-based approaches maybe be challenging since the requisite anti-Stokes shift (from tissue-penetrating wavelength to wavelength that is required to break bonds) may be large. Moreover, the materials that are used for upconversion, such as lanthanide-based nanoparticles, tend to be somewhat exotic in the world of local anesthetics. Multiphoton approaches such as second harmonic generation can also be used for triggered drug release using NIR light,^[42] but may require high irradiances and/or the use of expensive (femtosecond) lasers that are not in routine clinical use.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.