Keeping Nanomedicine on Target

Abstract

A principal purpose of targeting or triggering is to improve the active agents’ therapeutic indices by preferentially increasing their concentrations at the desired site of effect.

Ideally, a drug delivery system would deliver drugs only where they are needed, when they are needed, and to the degree that they are needed. Over the past many years, there has been increasing interest in targeted and triggered nanomaterials to achieve those goals. Targeted nanoparticles refer to those that can selectively accumulate in a targeted tissue relative to others. Triggered drug delivery systems achieve a drug delivery event (drug release or triggered targeting) in a tissue in response to a stimulus. Some systems rely on endogenous environmental properties such as pH, hypoxia, and/or enzyme activity, to provide the stimulus that causes the drug delivery event;¹ these are often referred to as “passive”. Others depend on external triggering sources, such as light, ultrasound, and chemicals.² Targeting of a tissue can also be achieved by attaching a specific ligand to the nanoparticle surface.³ Those systems that do not rely on simple tissue properties are often referred to as “active”. The goal is to increase the therapeutic effect of a given drug by maximizing the fraction of free drug accumulating (targeting) or being released (triggering) at the intended site of action, enhancing efficacy and minimizing toxicity. Despite a body of experimental data suggesting that such approaches could work, and some nanoencapsulated drugs having obtained FDA approval or entered clinic trials,⁴ there is still a lot of room for development.

Some diseases have tissue properties that enable targeting and triggering of drug release, so that nanomaterials show improved efficacy and reduced toxicity compared to conventional formulations. As an example of tissue properties driving targeting, the enhanced permeability and retention (EPR) effect has been used to target PEGylated liposomal doxorubicin (Doxil) to tumors. With EPR, the relatively leaky vasculature of tumors allows the preferential accumulation of drug-loaded nanoparticles of a given size on-target, improving efficacy in relation to toxicity.⁵ (The relevance to and reliability of EPR in human disease is debated.⁶) As an example of tissue properties enabling triggering, the mild acidity at tumor sites (and in specific cell compartments) has been used to trigger pH-responsive nanomaterials to degrade or dissolve at the site of interest, achieving high local drug concentrations.⁷

The fact that selectivity is not absolute remains a major challenge. Only a small percentage (~1%) of systemically administered nanoparticles reach the diseased site,⁸ and perhaps equally important, the amount reaching off-target sites and causing toxicity remains substantial, that is, the therapeutic index (the ratio of the toxic to the efficacious dose) has not been optimized. Therefore, it is still of great interest to develop novel strategies to enhance targeting.

Triggering by an external source can further enhance delivery to or release at a specific location, perhaps in the absence of tissue properties that might trigger such events, and can enable targeting even in the absence of a specific ligand. One approach is to decorate a nanoparticle with a ligand that works on most cell surfaces, such as arginine-glycine-aspartate (RGD) derivatives, or cell-penetrating peptides. Inactivation of the ligand by a photosensitive moiety renders that ligand phototriggerable, allowing accumulation of the nanoparticle in response to irradiation.⁹ As an example of application in vivo, irradiation of the eye increased accumulation of intravenously administered nanoparticles (and drug) modified with a cell penetrating peptide inactivated with a diethylamino-coumarin caging group in a murine model of choroidal neovascularization.¹⁰ In our experience, targeting, especially triggered targeting of this type, works much better in the context of an existing endogenous condition that enhances accumulation. In the preceding example, accumulation in the eye in the absence of the EPR-like state engendered by neovascularization was minimal.

External triggering can also provide temporal control of drug release, for example, to allow drug release to be actuated by a hand-held device held by a patient, on-demand, and to allow the patient to dial in exactly how much drug is released by adjusting the intensity of the stimulus. Liposomes loaded with local anesthetics were rendered photosensitive by decorating their surfaces with gold nanorods which would heat upon irradiation with near-infrared light, via surface plasmon resonance.¹¹ The heat would increase drug release from the liposomes, resulting in local anesthesia on demand, the intensity of which could be modulated by the intensity and duration of irradiation.

Applying energy sources for an extended period or at high irradiance could enhance the therapeutic effect but could also cause tissue injury.¹² Therefore, the sensitivity of nanoparticles to external stimuli has emerged as a key design feature, as has using triggering modalities with less attenuation as they travel through tissue.

While enhancing the triggerability of systems is important, excessive sensitivity can render them triggerable by ambient conditions (e.g., daylight), fever, airport scanners, and medical MRI. Also, often systems that are relatively easy to trigger also have more basal (i.e., untriggered) drug release. (For example, it is easier to trigger drug release from liposomes than from covalent drug-polymer conjugates, but liposomes also have greater basal drug release).¹³ Systems with high basal drug release also often have significant initial (“burst”) release, and subsequent basal release can lead to continuous unwanted depletion of drug that could have been used for triggered events.

Another challenge for externally stimulated systems is identifying where to apply the stimuli (e.g., the location of a tumor) or when nanoparticle accumulation reaches its maximum. Integration of traditional imaging technologies (e.g., guidance by ultrasonography in regional anesthesia¹⁴) with nanoimaging agents or theranostic nanoparticles could be helpful in these situations.

Triggerable and targeted systems can be enhanced by applications from other subfields of drug delivery. For example, covalent and noncovalent methods of minimizing untriggered drug release could minimize off target drug effect, improving the therapeutic index.

The road to translation for remotely triggered and targeted approaches is hampered by many factors. A common problem in the study of nanoparticulate delivery is the heterogeneity of different diseases or models; xenografted mouse models may not mimic human tumors; the EPR effect can vary in different tumors or within individual tumors. Consequently, efforts to understand the pathophysiology of each disease will remain important, as will the development of preclinical models that accurately reflect human disease. The discrepancy between human and animal may be particularly important in the context of external energy triggers, as the distances across which the energy has to be transmitted, and over which it can be attenuated or cause injury, are much greater. This scale issue highlights the importance of developing systems with high sensitivity to stimuli and stimuli with low attenuation. Stimuli that focus multiple beams at a given point may prove to be safer. Combinations of stimuli that are truly synergistic (a term that is very often used but rarely proven) may greatly enhance local accumulation and therefore the therapeutic index. However, combining stimuli and other improvements described above may involve more complex formulations and relatively exotic materials.

As technologies using these approaches proliferate, it may become important to demonstrate that they are better than others and that they improve outcomes. In particular, it will be important to demonstrate that the therapeutic index is meaningfully affected, using relevant therapeutic and toxic end points in animal models that accurately mimic human pathophysiology and pharmacology. Such data could facilitate translation by helping to mitigate possible regulatory hurdles created by, for example, the possible complexity and unusual materials of the formulations, or the fact that in some cases the system would be a combination of a device and a drug delivery system. Successful translation could have great effect in diseases where therapeutic effect is limited, especially if by toxicity (as in cancer), and also in conditions where the patient’s ability to control drug effect in real time would be greatly beneficial (as in pain).