A nose for success

Dan DiPietro and Kristin Woolley from SCO discuss the science behind Nastech's intranasal drug delivery technology

Nastech, a US leader in drug delivery, is still largely unknown on Wall Street. What attention has been paid to the company has largely focused on its lead product, apomorphine, which was partnered with Pharmacia before the latter's merger with Pfizer.However, Nastech is developing a platform technology for delivering large macromolecules intranasally and exploiting the natural anatomy and physiology of the olfactory system to circumvent the blood brain barrier to deliver drugs to the central nervous system (CNS).

Drug delivery is increasingly becoming an integral factor in the development of pharmaceutical products.

Once considered niche, and existing in the background of the pharmaceutical industry, companies specialising in drug delivery are receiving more attention from big pharmaceutical companies as well as the investment community.

There are several factors behind this phenomenon:

the well publicised imminent loss of key company drug patents, as well as the poor state of near-term pipelines has led to concerns about earnings growth for pharmaceutical manufacturers;

the intense competition seen in the pharmaceutical industry has greatly shortened the honeymoon period enjoyed by products in a particular therapeutic area;

poor patient compliance is seen as a serious impediment to successful disease control;

pharmacokinetic properties can be optimised by changing the delivery route strategy.

Intranasal drug delivery technologies generally address two goals: first, to leverage the patient-friendly and favourable pharmacoeconomic features of intranasal administration to deliver drugs systemically; and second, to exploit the physiology of the olfactory system to deliver neurologic drugs efficiently and specifically to the brain.

There is preliminary evidence that such technology can also be used to optimise absorption, both systemic and neurologic, of clinically important but pharmacologically unfavourable compounds. This article aims to explain the basis of core aspects of the technology at the anatomical and cellular level. To better understand the delivery approaches, familiarity with the structure of the nasal cavity is necessary.

There are three functionally and morphologically distinct regions in the nasal cavity - the olfactory, respiratory and vestibular regions, figures 1 and 2.

The vestibular region is located at the opening of the nasal cavity. It serves as a filter for airborne particles and is heavily populated with long hairs to serve this purpose.

The respiratory region is located further back from the nostrils and is the largest of the three regions. It serves as a removal system for particles that have been deposited on the mucous layer and is heavily populated with cilia to serve this purpose.

direct pathways

The region furthest back from the nostrils, constituting about 6% of the total surface area of the nasal cavity in humans, is the olfactory region. This region is free of airflow and is responsible for the sensory function of smell. It is directly adjacent to the bony cribriform plate, on the other side of which is the olfactory bulb, which is part of the CNS.

The axons of the olfactory neurons in this region penetrate the cribriform plate and synapse within the olfactory bulb, forming a direct pathway from the nasal mucosa to the brain.

The respiratory epithelium is populated by motile cilia, supporting columnar cells, goblet cells, which secrete the mucous layer, and basal cells. The olfactory epithelium is missing the motile cilia and goblet cells, but has Bowman's glands, which secrete the mucous layer, and primary olfactory neurons, which are not present in either of the other two nasal regions.

In order to utilise the nasal delivery route, the protective aspects, such as the thick, viscous, protease-rich mucosal layer, rapid mucociliary clearance, and lack of paracellular transport in the nasal cavity must be overcome.

The mucosal layer is 5-20mm thick and rich in proteases, which creates a problem for drug delivery as it serves both as a physical barrier to entry to the cellular space, and as a chemical barrier, resulting in proteolytic degradation in the case of protein-based therapeutics.

drug formulations

The 20min mucociliary clearance time is detrimental as it serves to remove the drug from the delivery pathway before it has had the opportunity to fully diffuse through the mucosal layer and contact the epithelial cell layer. Nastech has addressed the first of these problems by including protease inhibitors and absorption enhancers in its protein-based drug formulations.

The protease inhibitors prevent protein degradation while travelling through the mucous to the epithelial cells. Absorption enhancers are used to facilitate this travel and thereby overcome the mucocilliary clearance issue.

In all three nasal epithelial types, cells are joined to one another through tight junctions, figure 3. These are cell-to-cell adhesions found in epithelial and endothelial cell layers, which serve to regulate paracellular transport (transport between neighbouring cells).

When closed, these junctions stitch together the cells, prohibiting extracellular transport of water and solutes. Tight junctions consist of integral membrane proteins, claudins, occludins and JAMs (junctional adhesion proteins.

After the mucosa, the next hindrance to uptake by the blood or brain is the plethora of tight junctions between the epithelial cells inhibiting extracellular fluid flow.

Recent data has shown that tight junctions naturally open and close in response to intra and extracellular stimuli, selectively allowing the passage of solutes into the paracellular space. One example of such endogenous regulation is found in the TNF-alpha mediated migration of leukocytes during normal immune surveillance and inflammatory responses.

Mononuclear cells produce TNF-alpha, a 17kDa proinflammatory cytokine, which causes tight junctions to open, allow passage of the mono-nuclear cells and then rapidly reseal to reconstitute the epithelial barrier.

Other endogenous mediators of tight junction permeability include other cytokines (IL-1, IL-4, IL-13), TGF-alpha, IGF-I, IGF-II and VEGF. Additionally, certain solutes commonly added to drug formulations in currently marketed products have been shown to act as non-specific tight junction modulators (e.g. wheatgerm and chitosan).

One way to expand this approach is to build a sophisticated understanding of tight junction biology and develop proprietary drug-like molecules that directly and specifically regulate tight junctions in a manner mimicking natural processes.

Nastech employs a proprietary in vitro screening protocol (the EpiAirway model) to rapidly identify these small molecule tight junction modulators.

tight junctions open

The excipients in the formulation induce reversible opening of the tight junctions in order to allow an active drug to pass through the nasal epithelium. Depending upon the dose size and excipient used, the extent to which the tight junctions open can be regulated, allowing precise control over the size of the molecules allowed to enter the paracellular pathway.

This is important, as unregulated entry would result in a flooding of the immune system with nasally introduced antigens and could result in side effects - including allergic reactions and infection.

There is also evidence that tight junctions prevent the delivery of large macromolecules through the nasal mucosa. Additionally, small molecules that open tight junctions wide enough for the passage of large proteins have been identified.

By controlling the plume geometry of the spray, usually through mechanical modification of the actuator, the drug and excipients can be preferentially delivered to different parts of the nose, resulting in different final destinations - either to the bloodstream or directly into the CNS.

It is extremely important to control the site of drug delivery within the nose. Drugs exposed to the respiratory epithelium in the presence of tight junction modulators flow through the paracellular pathway and reach the capillary beds below the basal membrane.

By diffusion, these drugs are able to enter the bloodstream and promote systemic distribution. Drugs exposed to the olfactory epithelium in the presence of tight junction modulators also utilise paracellular transport to flow through the epithelium and past the basal membrane.

As the basal membrane of the olfactory epithelium is not as highly enervated with capillaries, most of the substance does not enter the systemic circulation.

Instead, the drug diffuses into the perineural sheaths of the neurons and travels in the perineural fluid (the fluid between the axon and the surrounding Schwann cell) until it passes through the cribriform plate and is mixed with the cerebral spinal fluid (CSF) in the arachnoid space. From here the substance can either drain into the cervical lymph nodes or pass through the pia membrane of the olfactory bulb and spread throughout the brain.

Drugs exposed to the olfactory epithelium in the absence of tight junction modulators can also travel to the CNS, by travelling through the neuron directly. The drugs are either taken up by pinocytosis or cell surface glycoprotein mediated adsorptive endocytosis.

The former method is nonspecific and targets the majority of the drug to lysosomes. These lysosomes undergo anterograde transport and are then targeted for degradation. Only a small portion of the contents of the lysosomes are released into the arachnoid space, so very little of the drug manages to successfully enter the brain through this pathway.

The latter method, mediated endocytosis, is more successful in transporting the drug through the nasal epithelium and into the CNS. This difference lies in the different types of lysosomes in which these drugs are packaged. After endocytosis, the drug is placed in neuronal Golgi-endoplasmic reticulum lysosomes (GERL), which are targeted for secretion.

These lysosomes undergo anterograde axoplasmic transport to the glomeruli in the olfactory bulb and are then exocytosed. This unique entry into the brain through the nasal cavity is particularly exciting because there is currently no other noninvasive route to the CNS due to the existence of the blood:brain barrier (BBB): a single layer of tile-like endothelial cells that line the inner surfaces of capillaries in the brain. These cells form tight junctions, with the result that no constituents of the blood, apart from water, can diffuse freely across the endothelial layer into or out of the paracellular space.

confounding barrier

Those elements required by the brain, such as glucose and other nutrients, are actively transported by the cells into the cerebral spinal fluid (CSF) and the thin layers of liquid that surround and bathe each brain cell.

This differs from non-neural transport between cells and capillaries in that non-neural capillary endothelial cells do not have tight junctions and so allow fairly large solutes to diffuse through the paracellular pathway.

The blood:brain barrier is a confounding factor in the development of effective neurologic drugs since high circulating levels of the active drug in the blood do not necessarily translate to high levels in the CNS.

If the CNS is the site of action, that can be a significant problem. Avoiding dilution in the systemic circulation and delivering drugs at higher concentration to the CNS would in many contexts be expected to improve efficacy.

The corollary of this is that often the toxicity associated with a neurologic drug is due to its presence in the systemic circulation. For example, the acetyl cholinesterase inhibitors for the treatment of Alzheimer's disease such as Aricept (Pfizer) and Exelon (Novartis), are associated with gastro-intestinal side effects.

This is most likely due to the activation of cholinergic neurons lining the gut. If it were possible to deliver an acetyl cholinesterase differentially to the brain, fewer GI side effects would be expected.

Work on the development of apomorphine has shown that its intranasal formulations, where appropriate, promote favourable CNS distribution. It has already been shown in a 12-patient clinical trial that the proprietary intranasal formulation of apomorphine can achieve 26.7-44.1% levels in the CSF relative to plasma, whereas subcutaneous administration produces 2.5-4.3%.

It is likely that promoting high CSF:plasma ratios will enhance the therapeutic index of CNS drugs. Again, the dose-limiting toxicity of many CNS drugs is GI distress due to secondary activity in peripheral nerves wiring the GI tract.

clinical trials

Additionally, this data would predict that protein therapeutics delivered directly and differentially to the central nervous system would reduce neutralising antibody production.

Nastech initiated Phase I trial intranasal formulation of Interferon beta, the active agent in multiple sclerosis treatments such as Avonex, Rebif and BetaSeron.

MS is caused by aberrant immune-mediated destruction of the insulating material lining axons essential for efficient impulse transmission. Interferon beta drugs are thought to work through down regulating the immune system and lessening autoimmune destruction of the myelin sheath lining axons in the CNS. Patients inject themselves intramuscularly or subcutaneously (1-3x/wk).

Interferon beta for the treatment of MS is therefore an ideal candidate for developing an intranasal formulation as:

it would spare patients the anxiety and inconvenience of regular self-injections;

it would theoretically allow high CSF levels to be achieved which could improve efficacy;

it would potentially minimise systemic exposure and thereby reduce side effects and minimise the production of neutralising antibodies.

A viable intranasal formulation of Interferon beta would be appealing to current manufacturers of the drug as there is newfound competition in the field. Biogen, Serono and Schering have all dedicated substantial resources to establishing a presence in the MS market.

They are each looking for features to further differentiate their product offerings and extend their reach. Complete Phase I results are expected this year, with safety, CSF: blood ratios and other pharmacological data.

The incidence of pure red blood cell aplasia in patients treated with Johnson & Johnson's Eprex has highlighted the immunogenic risk of injectable therapeutic proteins.

In the case of J&J's drug, in 1:10,000 patients the introduction of recombinant human erythropoetin (epo) resulted in the patients mounting an immune response against the protein resulting in neutralising anti-erythropoetin antibodies that targeted and destroyed both the recombinant protein and the patients' natural epo. Without epo, the patients were unable to produce red blood cells and were dependent upon blood transfusions for survival.

Epo is not the only recombinant protein harbouring autoimmune risk - other recombinant versions of naturally occurring human proteins carry this risk as well.

In fact, it has been documented that up to 80% of patients treated with IFNalpha2 produce autoantibodies to the protein.

Immunogenicity to recombinant human proteins can arise from the production, purification and storage procedures used in the manufacturing of the proteins, or directly from novel fusion sites created by joining two natural human proteins.

Nasal delivery has the potential to reduce or eliminate the immunogenic risk of introducing recombinant proteins, through the induction of mucosal tolerance.

Mucosal tolerance is a naturally occurring phenomenon that evolved in mammals to control autoreactive B and T cells that escaped elimination during the priming of the immune system in early age.

Though the process is not completely understood, the mechanism of action appears to be a combination of anergy (incomplete T cell activation), clonal deletion of antigen-specific T cells, and selective expansion of cells producing immunosuppressive cytokines such as IL-4, and TGF beta.

When incurred, mucosal tolerance allows the antigen to be repeatedly administered without inducing an immune response. For treatments that require chronic therapy or lengthy treatment regimens, such as epo, INF beta, and INF alpha, avoiding an immune response in the initial dosing periods and beyond would be a distinct advantage to intranasal delivery.

Terminology