Nanoparticle-Delivered Chemotherapy: Old Drugs in New Packages
Nanoparticle-Delivered Chemotherapy: Old Drugs in New Packages
Cytotoxic chemotherapies have a narrow therapeutic window, with high peaks and troughs of plasma concentration. Novel nanoparticle formulations of cytotoxic chemotherapy drugs can enhance pharmacokinetic characteristics and facilitate passive targeting of drugs to tumors via the enhanced permeability and retention effect, thus mitigating toxicity. Nanoparticle vehicles currently in clinical use or undergoing clinical investigation for anticancer therapies include liposomes, polymeric micelles, protein-drug nanoparticles, and dendrimers. Multiple nanoparticle formulations of existing cytotoxic chemotherapies are approved for use in several indications, with clinical data indeed showing optimization of pharmacokinetics and different toxicity profiles compared with their parent drugs. There are also many new nanoparticle drug formulations in development and undergoing early- and late-phase clinical trials, including several that utilize active targeting or triggered release based on environmental stimuli. Here, we review the rationale for nanoparticle formulations of existing or previously investigated cytotoxic drugs, describe currently approved nanoparticle formulations of drugs, and discuss some of the most promising clinical trials currently underway.
Introduction and Rationale for Nanoparticle Reformulations
While cytotoxic chemotherapies form a cornerstone of anticancer treatment, they universally have narrow therapeutic windows. Within the last 2 decades, novel formulations of chemotherapeutics using nanoparticle vehicles have been developed to optimize pharmacokinetics and improve tumor drug targeting, mitigating toxicity while maintaining efficacy. Multiple nanoparticle chemotherapy formulations are now approved as standard therapy, and even more agents are undergoing clinical trials.
Drugs encapsulated within or conjugated to nanoparticle vehicles display pharmacokinetics that are markedly different from those of free drugs. Drug concentration after administration depends on clearance (reflecting elimination rate) and volume of distribution (reflecting the size and number of body compartments available to a drug, based on characteristics such as hydrophobicity and size). Generally, by trapping cytotoxic drugs in long-circulating nanoparticles that slowly release the drug over time, the peaks and troughs of drug concentration are not as extreme as those seen with infusion of a free drug. Also, combinations of drugs can be encapsulated in a single nanoparticle to ensure combinatorial drug delivery at a defined ratio to yield a desired synergistic effect.
Additionally, trapping hydrophobic drugs like taxanes within nanoparticles with hydrophilic shells can eliminate the need for using emulsifying excipients, which have their own toxicities. These hydrophobic drugs otherwise require a vehicle of polyethoxylated castor oil (Kolliphor EL, formerly Cremophor EL)or polysorbate 80 that emulsify the water-insoluble drug and facilitate intravenous infusion. These excipients often cause hypersensitivity reactions upon infusion, so reformulation to avoid their use may be beneficial. Indeed, there are case reports of successful rechallenge with nanoparticle albumin–bound (nab)-paclitaxel in patients who previously had hypersensitivity reactions to conventional paclitaxel—although some patients truly do have allergic reactions to the taxane drug itself, and so there must be careful consideration of which patients may be safely rechallenged. Nevertheless, avoiding the need for these excipients may be another benefit of nanoparticle reformulations of hydrophobic drugs.
Optimizing Nanoparticle Properties to Enhance Biocompatibility
A nanoparticle vehicle must allow for sufficient duration of circulation without rapid clearance, and should facilitate passive targeting, in which the nanoparticles are preferentially deposited within tumors. To achieve this, a nanoparticle that is delivered intravenously must circulate within the blood while evading opsonization—which is nonspecific binding by plasma proteins, complement, or immunoglobulins that directs the bound particles to phagocytosis and clearance by the reticuloendothelial system (RES) of macrophages within the liver and spleen. The nanoparticles then need to successfully exit the intravascular space and enter the intratumoral extravascular space, where they must diffuse to the cancer cells, undergo internalization, and release their chemotherapy payload. The key factors that affect these capabilities include nanoparticle size and surface charge.
The size of nanoparticles determines their biodistribution, since particles of different sizes have differing abilities to evade clearance by the kidneys or by the RES. For example, particles smaller than 5 nm are rapidly filtered out by the kidneys, while particles over 200 nm accumulate within the spleen, liver, and lung. Additionally, nanoparticle vehicles passively target tumors rather than normal tissues because their size limits nonspecific extravasation. Tumors have leaky blood vessels with increased vascular permeability, mediated by high intratumoral levels of vascular endothelial growth factor (VEGF), nitric oxide, and other inflammatory mediators.[5,6] Nanoparticles are unable to exit the intravascular space in normal tissues, limiting their volume of distribution, but they are able to readily extravasate out of leaky tumor-associated vessels. This enhanced permeability and retention (EPR) effect thus passively targets tumors while decreasing drug delivery to nontarget tissues. Consequently, nanoparticle vehicles should have sizes under 200 nm to avoid clearance by the RES while still being able to take advantage of the EPR effect. However, the EPR effect requires the presence of aberrant tumor blood vessels and thus depends on the degree of vascularity and endothelial fenestration. Moreover, the EPR effect is modulated by treatment with antiangiogenic therapy, which normalizes tumor blood vessels but may decrease this tumor-specific vessel leakiness. Indeed, following antiangiogenic therapy, intratumoral penetration of larger nanoparticles (over 100 nm) decreased, while the penetration of smaller (12-nm) nanoparticles improved. The clinical utility of the effect of previous or concomitant anti-angiogenic therapies with nanoparticle chemotherapy formulations is an area of ongoing investigation.
Additionally, stromal components and interstitial fluid pressure within the tumor affect drug accumulation and diffusion. The interstitial fluid pressure within a tumor is uniformly elevated due to the generalized increased extravasation of fluids from the leaky vasculature and poor lymphatic drainage, and intratumoral drug transport thus occurs slowly, via diffusion. While decreased lymphatic drainage contributes to intratumoral nanoparticle retention, the stromal extracellular matrix can also restrict diffusion of nanoparticles. In pancreatic cancers, which have dense, fibrotic stroma, only 30-nm nanoparticles could penetrate beyond the immediate perivascular space. These tumor-specific characteristics likely help explain the differing efficacies of various nanoparticle chemotherapy formulations in different cancer types. Also, therapies that modulate the tumor microenvironment, such as recombinant human hyaluronidase, are likely to affect nanoparticle and drug diffusion through the stroma, and combinatorial or sequential treatment approaches continue to undergo investigation (for example, ClinicalTrials.gov identifier NCT02715804).
The surface charge of nanoparticles, called zeta potential, also impacts nanoparticle distribution and uptake. Cationic particles bind more avidly to negatively charged plasma proteins like albumin or complement proteins, and are consequently readily cleared by the RES; they thus have lower circulating duration and insufficient tumor deposition. Therefore, in vivo tumor deposition of nanoparticles is optimal with slightly anionic particles that measure 150 nm in diameter, and it is common to modify the surface characteristics of the nanoparticles by attaching polymer chains that confer more favorable properties. Polyethylene glycol (PEG) is a hydrophilic polymer that has a near-neutral charge, and decorating the surface of a nanoparticle with PEG moieties (pegylation) masks the surface charge to improve circulating half-life while hiding potentially antigenic epitopes and thus preventing immunologic reaction. PEG polymers are also directly conjugated to drugs to optimize pharmacokinetics for clinical use, as in the cases of filgrastim, interferon alfa-2b, and asparaginase. Pegylation is an important method of conferring “stealth” properties on nanoparticles to improve the duration of their circulation. The addition of other epitopes to facilitate molecularly directed targeting and to aid in shielding the nanoparticles from macrophages and immunoglobulins is also undergoing investigation.
Several nanoparticle platforms are currently in clinical use or are being investigated for use in cancer therapeutics (Figure). Liposomes are among the most commonly used vehicles, and most of the currently approved nanoparticle chemotherapy formulations are liposomal. Liposomes are comprised of a phospholipid bilayer, with cholesterol added for bilayer stabilization, surrounding an aqueous core. The lipid envelope is in a more permeable fluid state if the environmental temperature exceeds the melting temperature, and assumes a less permeable gel state at lower temperatures; therefore, an appropriate phospholipid composition with a melting temperature higher than body temperature is required to minimize nonspecific drug leakage from the liposomes. Hydrophobic drugs can be loaded within the lipid bilayer of the liposome, whereas hydrophilic drugs remain within the aqueous core.
Polymeric micelles consist of amphiphilic block polymers with a hydrophilic segment, often comprised of PEG, and a hydrophobic core; they consequently self-assemble into monolayer micelles. To increase stability and prevent overly rapid degradation and drug release, the polymers include moieties that are readily cross-linked—either within the hydrophilic region to form shell cross-linked micelles, or within the hydrophobic region to form core cross-linked micelles. Drugs can be either covalently linked to the polymers or physically trapped within the polymeric micelles, and drug release from the micelle occurs as the polymers degrade.
Additionally, the binding of chemotherapeutic drugs to proteins such as albumin has also facilitated the formation of nanoparticles. Drugs that are otherwise hydrophobic, such as paclitaxel, are rendered soluble in aqueous solutions through nanoparticle albumin binding, lessening the need for excipients such as polyethoxylated castor oil, while taking advantage of the EPR effect and of mechanisms of albumin transport across endothelial cells and uptake into tumor cells. Currently, nab-paclitaxel is the only compound in this class that has been approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA), although others are in development.
Finally, dendrimers are nanoparticles formed from repetitively branched polymers arising from a core. Dendrimers are symmetric and uniform in size, and because of their branched structures, they have a multitude of covalent binding sites at the nanoparticle surface. These sites can link to drugs, targeting antibodies, or to PEG in order to favorably modify the dendrimer surface properties to avoid uptake by the RES and to minimize nonspecific cell membrane toxicity. No dendrimer-bound chemotherapy formulations are currently approved, although clinical trials are underway.
Safety Considerations With Nanoparticle Therapies
The lipids and polymers utilized in nanoparticles currently in clinical use are generally accepted as being biocompatible and safe. However, novel nanoparticle vehicles must be evaluated for potential toxicity, since materials may have markedly different physicochemical characteristics when they are nanoparticle-sized; they thus may have very different effects on living cells than do their larger-sized versions, and can potentially result in toxicities even at low cumulative doses. For example, nonpegylated cationic vehicles cause general cell toxicity and increased genotoxicity in vivo. Also, early generations of dendrimers demonstrated unacceptable pulmonary, hematologic, and other toxicities in preclinical studies, although these toxicities can be eliminated by using more biocompatible polymers and by pegylation. The potential for adverse events and toxicities related to the nanoparticle vehicle itself must be considered, especially when conducting early-phase clinical studies of nanoparticle formulations.
Liposomal vehicles may increase the risk of a hypersensitivity response, similar to what is sometimes seen with excipients such as polyethoxylated castor oil, due to a non–immunoglobulin E–mediated reaction, called complement activation–related pseudoallergy (CARPA). This complement activation triggers mast cell degranulation in a subset of patients.[18,19] Properties of the liposomal nanoparticles and of the specific encapsulated drugs determine the likelihood of CARPA for a given therapeutic, with high negative surface charge and nonspherical particle shape contributing to higher rates of CARPA. For example, the loading of doxorubicin into liposomes caused deformation of the liposomes into an ellipsoid shape rather than a spherical shape, which contributed to greater rates of complement deposition and activation and higher rates of hypersensitivity reactions with pegylated liposomal doxorubicin. For drugs whose liposomal formulations confer an increased risk of mild hypersensitivity reactions compared with their conventional formulations—as is the case with doxorubicin—the increased risk must be outweighed by decreases in other severe toxicities, such as cardiotoxicity, when the liposomal formulation is used. These factors remain important considerations for liposomal drugs.
Also, though PEG is generally thought to be immunologically inert, emerging evidence suggests there may be an anti-PEG immunologic reaction in subgroups of patients. Neutralizing anti-PEG antibodies have been identified in small studies in 7% to 42% of normal human subjects, which may be attributable to the widespread use of PEG in many household consumer products. Small studies have also suggested an association between the presence of anti-PEG antibodies and decreased efficacy of pegasparaginase and PEG-uricase. There are also case reports of immediate and delayed hypersensitivity reactions to PEG. Given the importance of pegylation in making nanoparticles and other drugs more biocompatible, the clinical relevance of these antibodies needs to be further investigated.