POS0842 NON-INVASIVE COHERENT RAMAN IMAGING OF INVOLVED FOREARM SKIN REVEALS ALIGNED COLLAGEN IN DERMIS OF SYSTEMIC SCLEROSIS PATIENTS

Background

Systemic sclerosis (SSc) is an immune-mediated connective tissue disease with clinical hallmark of inflammation, vasculopathy and fibrosis resulting in abnormal collagen and intercellular matrix formation in the dermis and internal organ. Skin severity correlates with systemic complications and mortality in SSc.

Many skin assessment tools studied do not fulfilled all standards set by OMERACT which include criterion validity, construct validity, discrimination, responsiveness, reliability and feasibility.

Modified Rodnan skin score (mRSS) is commonly used as outcome measure in clinical trials of SSc. It is however, limited by inter-rater variability. Studies also showed that skin biopsy samples from body part of SSc patients with clinically normal mRSS had demonstrable pathological deposition of collagen. While repeat skin biopsy to track longitudinally, may be unacceptable to patients. Hence, ongoing research for non-invasive tools are encouraged.

Objectives

We follow through our previous work studying commercially available non-invasive imaging tools in SSc skin assessment by using Coherent-Raman scattering microscopy (CRS).

Methods

Skin biopsies of 4mm were obtained from SSc (n=3) and HC (n=1). Skin samples were collected at anterior surface of forearm from newly diagnosed diffuse SSc which were treatment naïve. Samples were frozen at -80c prior to analysis under CRS at skin depth of 150-200um. The laser setting was 3% pump (802nm) and 6% stokes (1045nm) to avoid tissue burn. 3 contrast methods used are Coherent anti-Stokes Raman scattering (CARS) for imaging of lipids, Second Harmonic Generation (SHG) for collagen and Two-Photon Fluorescence (TPF) for elastin fibers. We analysed collagen and elastin fibers in the skin samples which is disease relevant end products of fibroblast activation in SSc. Descriptive study of the skin CRS characteristics is reported.

Results

The SHG analysis revealed denser collagen fibers in dermal layer of diseased skin. It also appears that the collagen deposition occurs in more superficial layers of the skin. Peak of collagen curve were at depth of 110-130um HC vs 75-90um in SSc. Collagen fibres were more aligned in the SSc (Figure 1).

Figure 1.

Microscopy analysis of SSc (B) vs HC (A) skin. I: 3D image of biopsy samples with three contrasts. Red is CARS for lipid, blue is SHG for collagen and green is TPF for elastin fibers. II. SHG. Imaging revealed denser and more aligned collagen fibers in SSc.

Furthermore, TPF revealed larger number of disordered elastin fibres in the dermal layer of SSc than HC.

Conclusion

Nailfold capillaroscopy is used to define and stage micro-vasculopathy in SSc. Inflammatory pro-fibrotic processes on the other hand cause abnormal collagen and intercellular matrix formation in the dermis and internal organ.

With the demand of better tools in diagnosis of early SSc and therapeutic research, our work with CRS prove better in objective evaluation of skin changes at the molecular level. Here, we demonstrate that the SHG is altered in the early diffuse SSc skin with increased and aligned collagen in the SSc dermis compatible with mRSS score. Others have observed this alignment of the collagen, and we have published that SSc fibroblasts migrate/invade along aligned collagen and modify the underlying extracellular matrix, adding collagens I and III, cross-linking enzymes and other factors including TSP-1.

Our future work include:

1.Generating a computer module in defining pathological collagen level

2.Analysis of metabolites and pathological pathways in SSc

3.In-vivo trials with novel therapeutic peptides.

4.Lastly, manufacturing of non-invasive handheld device that is capable of diagnosis subclinical SSc and for outcome measure in clinical-therapeutic research

References

[1]Abignano G, et al. Virtual skin biopsy by optical coherence tomography: the first quantitative imaging biomarker for scleroderma. Ann Rheum Dis. 2013

[2]Ahmed Abdi B, et.al. Use of Patterned Collagen Coated Slides to Study Normal and Scleroderma Lung Fibroblast Migration. Sci Rep. 2017

Acknowledgements

We want to acknowledge our colleagues and patients in Royal Free Hospital, to make this project a success.

Disclosure of Interests

None declared

Delivery of peptides to the blood and brain after oral uptake of quaternary ammonium palmitoyl glycol chitosan nanoparticles

The clinical development of therapeutic peptides has been restricted to peptides for non-CNS diseases and parenteral dosage forms due to the poor permeation of peptides across the gastrointestinal mucosa and the blood-brain barrier. Quaternary ammonium palmitoyl glycol chitosan (GCPQ) nanoparticles facilitate the brain delivery of orally administered peptides such as leucine(5)-enkephalin, and here we examine the mechanism of GCPQ facilitated oral peptide absorption and brain delivery. By analyzing the oral biodistribution of radiolabeled GCPQ nanoparticles, the oral biodistribution of the model peptide leucine(5)-enkephalin and coherent anti-Stokes Raman scattering microscopy tissue images after an oral dose of deuterated GCPQ nanoparticles, we have established a number of facts. Although 85-90% of orally administered GCPQ nanoparticles are not absorbed from the gastrointestinal tract, a peak level of 2-3% of the oral GCPQ dose is detected in the blood 30 min after dosing, and these GCPQ particles appear to transport the peptides to the blood. Additionally, although peptide loaded nanoparticles from low (6 kDa) and high (50 kDa) molecular weight GCPQ are taken up by enterocytes, polymer particles with a polymer molecular weight greater than 6 kDa are required to facilitate peptide delivery to the brain after oral administration. By examining our current and previous data, we conclude that GCPQ particles facilitate oral peptide absorption by protecting the peptide from gastrointestinal degradation, adhering to the mucus to increase the drug gut residence time and transporting GCPQ associated peptide across the enterocytes and to the systemic circulation, enabling the GCPQ stabilized peptide to be transported to the brain. Orally administered GCPQ particles are also circulated from the gastrointestinal tract to the liver and onward to the gall bladder, presumably for final transport back to the gastrointestinal tract.

PEI-based vesicle-polymer hybrid gene delivery system with improved biocompatibility

Wider use of the transfection agent polymer polyethylenimine (PEI) in vivo has been hampered by its toxicity. In order to examine whether material combining properties of polymers and lipid type of carriers would have improved characteristics, four PEI derivatives were synthesised: The methylation of the branched PEI (25 kDa) created a permanently charged quaternary ammonium derivative. Acylation of these backbones using pendant palmitic acid chains created amphiphilic PEI variants which formed nanoparticles or vesicles. Finally hydrophilic groups were added to the polymer backbone by PEGylation. The materials were characterised and their in vitro and in vivo properties were tested. The modifications improved the materials biocompatibility markedly when compared to the starting material but also reduced transfection efficiency. The material bearing ammonium and palmitoyl groups was 10x less toxic while retaining about 30% of the transfection efficiency in vitro. After intravenous administration in a mouse model the materials also gave rise to GFP transgene expression in the liver. The synthetic strategy altered complex physicochemistry and improved biocompatibility while maintaining in vitro gene expression for most formulations. The strategy of combination of complementary properties of cationic lipids and polymers into a hybrid material may also be applicable to other materials.

Gene delivery with synthetic (non viral) carriers

Non-viral gene delivery involving the use of cationic polymer and cationic lipid based carriers still continues to enjoy a high profile due to the safety advantages offered by these systems when compared with viruses. However, there are still problems associated with the use of these agents, notably their comparatively low efficiency and the inability to target gene expression to the area of pathology. On intravenous administration gene expression is found predominantly in the first capillary bed encountered-the lung endothelium. The clinical use of non-viral gene delivery systems in cystic fibrosis or cancer has involved their direct application to the site of pathology due to the targeting difficulties experienced. For gene expression to occur genes must be transported to the interior of the cell nucleus and a number of biological barriers to effective gene delivery have been identified. These may be divided into extracellular such as the targeting barrier mentioned above and intracellular such as the need for endosomal escape after endocytosis and the inefficient trafficking of genes to the nucleus. Targeting ligands have been used with moderate success to overcome the targeting barrier while endosomal escape and nuclear targeting peptides are some of the strategies, which have been employed to overcome the problems of endosomal escape and nuclear trafficking. It is hoped that the next generation of carriers will incorporate mechanisms to overcome these barriers thus improving the efficacy of such materials.

Quaternary ammonium palmitoyl glycol chitosan--a new polysoap for drug delivery

A new polysoap, quaternary ammonium palmitoyl glycol chitosan (GCPQ, M(w)=178,000 g mole(-1)) with drug solubilising potential has been synthesised and characterised. In solution hydrophobic domains of GCPQ polymeric micelles were identified by the hypsochromic shift in the lambda(max) of methyl orange and by the increase in the ratio of the fluorescence emission intensity of the third and first pyrene vibronic peaks (I(3)/I(1)). At high aqueous concentrations (>10 mg ml(-1)) GCPQ presents as a gel which solubilises pyrene (2.5 mM, normal solubility in water approximately 2 microM) on probe sonication. Dilution of the gel to a liquid solution of polymeric micelles (< or =3.75 mg ml(-1) of GCPQ), results in the observation of fluorescent pyrene excimers (excited dimers) and a high excimer to monomer fluorescence ratio (I(E)/I(M)). However, attempts to solubilise pyrene at a concentration of 2.5 mM in a liquid solution of GCPQ (3.75 mg ml(-1)) results in a reduced I(E)/I(M) value and pyrene precipitation. Viscometry measurements show a more compact conformation for the polymer solubilising pyrene than the polymer alone. The polymer is non-haemolytic when present as the liquid solution and relatively non cytotoxic. In conclusion, a new biocompatible polysoap (potential drug solubiliser) is described which forms hydrophobic domains in solution and shows hysteresis in its solubilisation of pyrene.

Preliminary characterization of novel amino acid based polymeric vesicles as gene and drug delivery agents

The amino acid homopolymers, poly-L-lysine and poly-L-ornithine, have been modified by the covalent attachment of palmitoyl and methoxypoly(ethylene glycol) (mPEG) residues to produce a new class of amphiphilic polymers-PLP and POP, respectively. These amphiphilic amino acid based polymers have been found to assemble into polymeric vesicles in the presence of cholesterol. Representatives of this new class of polymeric vesicles have been evaluated in vitro as nonviral gene delivery systems with a view to finding delivery systems that combine effective gene expression with low toxicity in vivo. In addition, the drug-carrying capacity of these polymeric vesicles was evaluated with the model drug doxorubicin. Chemical characterization of the modified polymers was carried out using (1)H NMR spectroscopy and the trinitrobenzene sulfonic acid (TNBS) assay for amino groups. The amphiphilic polymers were found to have an unreacted amino acid, palmitoyl, mPEG ratio of 11:5:1, and polymeric vesicle formation was confirmed by freeze-fracture electron microscopy and drug encapsulation studies. The resulting polymeric vesicles, by virtue of the mPEG groups, bear a near neutral zeta-potential. In vitro biological testing revealed that POP and PLP vesicle-DNA complexes are about one to 2 orders of magnitude less cytotoxic than the parent polymer-DNA complexes although more haemolytic than the parent polymer-DNA complexes. The polymeric vesicles condense DNA at a polymer:DNA weight ratio of 5:1 or greater and the polymeric vesicle-DNA complexes improved gene transfer to human tumor cell lines in comparison to the parent homopolymers despite the absence of receptor specific ligands and lysosomotropic agents such as chloroquine.

Niosomes and polymeric chitosan based vesicles bearing transferrin and glucose ligands for drug targeting

PURPOSE

To prepare polymeric vesicles and niosomes bearing glucose or transferrin ligands for drug targeting.

METHODS

A glucose-palmitoyl glycol chitosan (PGC) conjugate was synthesised and glucose-PGC polymeric vesicles prepared by sonication of glucose-PGC/cholesterol. N-palmitoylglucosamine (NPG) was synthesised and NPG niosomes also prepared by sonication of NPG/ sorbitan monostearate/ cholesterol/ cholesteryl poly-24-oxyethylene ether. These 2 glucose vesicles were incubated with colloidal concanavalin A gold (Con-A gold), washed and visualised by transmission electron microscopy (TEM). Transferrin was also conjugated to the surface of PGC vesicles and the uptake of these vesicles investigated in the A431 cell line (over expressing the transferrin receptor) by fluorescent activated cell sorter analysis.

RESULTS

TEM imaging confirmed the presence of glucose units on the surface of PGC polymeric vesicles and NPG niosomes. Transferrin was coupled to PGC vesicles at a level of 0.60+/-0.18 g of transferrin per g polymer. The proportion of FITC-dextran positive A431 cells was 42% (FITC-dextran solution), 74% (plain vesicles) and 90% (transferrin vesicles).

CONCLUSIONS

Glucose and transferrin bearing chitosan based vesicles and glucose niosomes have been prepared. Glucose bearing vesicles bind Con-A to their surface. Chitosan based vesicles are taken up by A431 cells and transferrin enhances this uptake.

The effect of processing variables on the physical characteristics of non-ionic surfactant vesicles (niosomes) formed from a hexadecyl diglycerol ether

Niosomes are vesicles formed by self-assembly of non-ionic surfactants. In this investigation, the effects of processing variables, particularly temperature and sonication, on the physical characteristics and phase transitional behaviour of two niosomal systems based on a hexadecyl diglycerol ether (C(16)G(2)) have been studied. Systems containing C(16)G(2), cholesterol and poly-24-oxyethylene cholesteryl ether (Solulan C24) in the molar ratios 91:0:9 and 49:49:2 were prepared by aqueous dispersion of films, followed by examination of 5(6)-carboxyfluorescein entrapment, particle size and morphology. The thermal behaviour was examined using high sensitivity differential scanning calorimetry (HSDSC) and hot stage microscopy, while the effects of sonication were studied in terms of size and morphology, both immediately after preparation and on storing for 1 h at room temperature and 60 degrees C. Polyhedral niosomes were formed from systems containing C(16)G(2) and Solulan C24 alone, while cholesterol-containing systems formed spherical vesicles mixed with tubular structures; the polyhedral systems were found to have a larger particle size and higher CF entrapment efficiency. HSDSC studies showed the polyhedral systems to exhibit an endotherm at 45.4 degrees C and a corresponding exotherm at 39.1 degrees C on cooling which were ascribed to a membrane phase transition; no equivalent transition was observed for the cholesterol containing systems. Hot stage microscopy showed the polyhedral vesicles to convert to spherical structures at approximately 48 degrees C, while on cooling the spherical vesicles split into smaller structures and reverted to the polyhedral shape at approximately 49 degrees C. Sonication resulted in the polyhedral vesicles forming spherical structures which underwent a particle size increase on storage at room temperature but not at 60 degrees C. The study suggests that the polyhedral vesicles undergo a reversible transition to spherical vesicles on heating or sonication and that this morphological change may be associated with a membrane phase transition.

Liposomes encapsulating polymeric chitosan based vesicles--a vesicle in vesicle system for drug delivery

Drug delivery systems comprising vesicles prepared from one amphiphile encapsulating vesicles prepared from a second amphiphile have not been prepared previously due to a tendency of the bilayer components of the different vesicles to mix during preparation. Recently we have developed polymeric vesicles using the new polymer-palmitoyl glycol chitosan and cholesterol in a 2:1 weight ratio. These polymeric vesicles have now been encapsulated within egg phosphatidylcholine (egg PC), cholesterol (2:1 weight ratio) liposomes yielding a vesicle in vesicle system. The vesicle in vesicle system was visualised by freeze fracture electron microscopy. The mixing of the different bilayer components was studied by monitoring the excimer fluorescence of pyrene-labelled polymeric vesicles after their encapsulation within egg PC liposomes or hexadecyl diglycerol ether niosomes. A minimum degree of lipid mixing was observed with the polymeric vesicle-egg PC liposome system when compared to the polymeric vesicle-hexadecyl diglycerol ether niosome system. The polymeric vesicle-egg PC vesicle in vesicle system was shown to retard the release of encapsulated solutes. 28% of 5(6)-carboxyfluorescein (CF) encapsulated in the polymeric vesicle compartment of the vesicle in vesicle system was released after 4 h compared to the release of 62% of encapsulated CF from plain polymeric vesicles within the same time period.

The encapsulation of bleomycin within chitosan based polymeric vesicles does not alter its biodistribution

Polymeric vesicles have recently been developed from an amphiphilic chitosan derivative--palmitoyl glycol chitosan. Their potential as a drug delivery system was evaluated using the anti-cancer compound bleomycin as a model drug. Palmitoyl glycol chitosan (GCP41) was synthesised by conjugation of palmitoyl groups to glycol chitosan. Bleomycin-containing vesicles (669 nm diameter) were prepared from a mixture of GCP41 and cholesterol by remote loading. The vesicles were imaged by freeze-fracture electron microscopy and their in-vitro stability tested. Incubation of the larger vesicles with plasma in-vitro led to a reduction of mean size by 49%, a reaction not seen with control sorbitan monostearate niosomes (215 nm in size). They also showed a higher initial drug release (1 h), but GCP41 and sorbitan monostearate vesicles retained 62% and 63% of the encapsulated drug after 24h, respectively. The biodistribution of smaller vesicles (290 nm) prepared by extrusion through a 200-nm filter was also studied in male Balb/c mice. Encapsulation of bleomycin into polymeric vesicles did not significantly alter the pharmacokinetics of biodistribution of bleomycin in male Balb/c mice although plasma and kidney levels were slightly increased. It is concluded that the extruded GCP41 vesicles break down in plasma in-vivo and hence are unlikely to offer any therapeutic advantage over the free drug.

A non-covalently cross-linked chitosan based hydrogel

Hydrogels are normally formed by the covalent cross-linking of linear polymers. In the case of chitosan based hydrogels this cross-linking is often achieved with glutaraldehyde, glyoxal or other reactive cross-linking agents. Such hydrogel materials have limited biocompatibility and biodegradability. However by the attachment of hydrophobic palmitoyl groups to glycol chitosan, a water soluble chitosan derivative, we have produced a version of the amphiphilic vesicle forming polymer-palmitoyl glycol chitosan (Uchegbu et al., 1998, J Pharm Pharmacol 58, 453-458). The level of palmitoylation in this variant of the polymer (GCP11), as determined by proton neutron magnetic resonance spectroscopy, is 19.62+/-2.42% (n=4). GCP11 has been used to prepare soft, slowly eroding hydrogels suitable for drug delivery by simply freeze-drying an aqueous dispersion of the polymer. Non-covalent cross-linking to form the gel matrix is achieved by the hydrophobic interactions of the palmitoyl groups. The resulting material, as examined by scanning electron microscopy, is porous and may be hydrated to up to 20x its weight in aqueous media without any appreciable change in volume-transforming from an opaque to a translucent solid. The slow erosion of this material in aqueous environments gives a biodegradable and ultimately more biocompatible material than covalently cross-linked hydrogels. Unlike most chitosan-based gels, the gel is hydrated to 20x its weight at alkaline pH but only 10x its weight at neutral and acid pH. This is as a result of the gradual erosion of the gel at lower pH values. Hydration is also reduced from 20x the dry gel weight in water to 10x the dry gel weight in the presence of dissolved salts such as sodium chloride. GCP11 hydrogels have been loaded to 0.1% w/w with a model fluorophore, rhodamine B, by simply freeze-drying an aqueous dispersion of GCP11 in the presence of a solution of rhodamine B dissolved in either water or phosphate buffered saline (PBS, pH=7.4). The release of this model fluorophore was retarded by between 8 and 12% when PBS was contained in the gel in accordance with the hydration profiles. Rhodamine B release was also reduced by between 13 and 25% in the presence of acid as a result of the reduced solubility of rhodamine B at acid pH.

In vitro/in vivo characterisation of polyhedral niosomes

Non-ionic surfactant vesicles (niosomes) formed by a hexadecyl diglycerol ether (C16G2) and a series of polyoxyethylene alkyl ethers exhibit a variety of shapes dependent on their membrane composition. These surfactants form with an equimolar amount of cholesterol a mixture of largely spherical and tubular niosomes. In the absence of cholesterol, they form faceted polyhedral structures. The physicochemical and biological differences between polyhedral and spherical/tubular niosomes were studied. Polyhedral niosomes undergo a reversible shape transformation into spherical structures on heating above their phase transition temperature (Tm). The viscosity of polyhedral niosomes at room temperature is higher than their spherical counterparts due to their faceted and relatively rigid shape, and is more dependent on temperature due to shape transformation. At room temperature, polyhedral niosomes possess more rigid gel phase membranes and are less osmotically sensitive; however, they are more permeable because of a lack of or low levels of cholesterol in their membranes. Polyhedral niosomes loaded with luteinising hormone releasing hormone (LHRH), nonetheless, slow the release of drug compared to solution, albeit to a small extent.

Osmotic behaviour of polyhedral non-ionic surfactant vesicles (niosomes)

In addition to common spherical non-ionic surfactant vesicles (niosomes), disc-like, tubular, and polyhedral niosomes have also been reported. The permeability and osmotic activity of niosomes are important in determining their use as controlled-release drug-delivery systems. These properties have been compared for polyhedral niosomes prepared by hydrating a mixture of a hexadecyl diglycerol ether (C16G2), a poly(24)oxyethylene cholesteryl ether (Solulan C24), 91:9 or 98:2, and conventional spherical niosomes prepared from the same surfactants but with cholesterol. When subjected to osmotic gradients, polyhedral niosomes, the membranes of which are in the gel phase, swell and shrink less than their spherical counterparts and they are more permeable to the hydrophilic solute 5(6)-carboxyfluorescein. In 2 M NaCl the rate of release of carboxyfluorescein from polyhedral niosomes (both containing 9% Solulan C24) into either a hypotonic (water) or an isotonic medium (2 M NaCl) was low. This contrasted with similarly loaded spherical niosomes and polyhedral niosomes containing 2% Solulan C24, from which release was high in hypotonic media (e.g. water) but less in an isotonic medium (2 M NaCl). For both polyhedral and spherical niosomes encapsulating carboxyfluorescein (pKa = 6.4), release rates were higher at pH 8 than at pH 5. Polyhedral niosomes are thus, in general, less osmotically active than spherical niosomes because of their rigid but highly permeable membranes. The unusual polyhedral membrane impermeability to carboxyfluorescein co-entrapped with salt in hypotonic media is a function of Solulan C24 content, and is possibly a result of salting out of the polyoxyethylene chains; this is, therefore, a property that might be manipulated in the design of a drug-delivery system.

Preparation and in vitro/in vivo evaluation of luteinizing hormone releasing hormone (LHRH)-loaded polyhedral and spherical/tubular niosomes

Niosomes are vesicles formed by the self-assembly of nonionic surfactants in aqueous dispersions. They can entrap drugs and have been used experimentally as sustained drug delivery systems. Apart from conventional spherical niosomes, various types of vesicle ultrastructures can be formed by varying the composition of the vesicle membrane. Hexadecyl diglycerol ether (C16G2), cholesterol, and poly-24-oxyethylene cholesteryl ether (Solulan C24) in the ratio 91:0:9 gave polyhedral niosomes, whereas spherical and tubular niosomes are produced at a composition ratio of 49:49:2. The mean size of both polyhedral and spherical/tubular niosomes were within the range of 6 to 9 microm. Both types of vesicle were visualized by cryo-scanning electron microscopy. The properties of the two forms of niosomes were studied using luteinizing hormone releasing hormone (LHRH) as a model peptide. Analysis by high-performance liquid chromatography demonstrated high entrapment of LHRH acetate in polyhedral niosomes when prepared by remote loading methods using pH or (NH4)2SO4 gradients; in contrast, only low entrapment was achieved by passive loading methods (direct hydration at pH 7.4 or pH 3.0, dehydration-rehydration, and reversed-phase evaporation). In vitro studies demonstrated that both polyhedral and spherical/tubular niosomes were more stable in 5% rat skeletal muscle homogenate than in rat plasma. Also, polyhedral niosomes released more radiolabeled LHRH ([125I]LHRH) than spherical/tubular niosomes in both muscle homogenate and plasma. In clearance experiments in the rat, following intramuscular injection, both polyhedral and spherical/tubular niosomes gradually released [125I]LHRH into the blood, but some radioactivity remained at the injection site for 25 and 49 h, respectively. In contrast, [125I]LHRH in phosphate buffered saline was completely cleared from the injection site at 2 h. The release of drug is sustained by both niosome formulations, but spherical/tubular niosomes possess more stable membranes than polyhedral niosomes due to the presence of cholesterol.

Polymeric chitosan-based vesicles for drug delivery

A simple carbohydrate polymer glycol chitosan (degree of polymerization 800 approx.) has been investigated for its ability to form polymeric vesicle drug carriers. The attachment of hydrophobic groups to glycol chitosan should yield an amphiphilic polymer capable of self-assembly into vesicles. Chitosan is used because the membrane-penetration enhancement of chitosan polymers offers the possibility of fabricating a drug delivery system suitable for the oral and intranasal administration of gut-labile molecules. Glycol chitosan modified by attachment of a strategic number of fatty acid pendant groups (11-16 mol%) assembles into unilamellar polymeric vesicles in the presence of cholesterol. These polymeric vesicles are found to be biocompatible and haemocompatible and capable of entrapping water-soluble drugs. By use of an ammonium sulphate gradient bleomycin (MW 1400), for example, can be efficiently loaded on to these polymeric vesicles to yield a bleomycin-to-polymer ratio of 0.5 units mg(-1). Previously polymers were thought to assemble into vesicles only if the polymer backbone was separated from the membrane-forming amphiphile by a hydrophilic side-arm spacer. The hydrophilic spacer was thought to be necessary to decouple the random motion of the polymer backbone from the ordered amphiphiles that make up the vesicle membrane. However, stable polymeric vesicles for use in drug delivery have been prepared from a modified carbohydrate polymer, palmitoyl glycol chitosan, without this specific architecture. These polymeric vesicles efficiently entrap water-soluble drugs.

Polyhedral non-ionic surfactant vesicles

Large polyhedral (2-10 microns) non-ionic surfactant vesicles (niosomes) formed from mixtures of a hexadecyl diglycerol ether (C16G2), a cholesteryl poly-24-oxyethylene ether (solulan C24) and a low level of cholesterol are being investigated as slow-release systems for ophthalmic, subcutaneous or intramuscular administration. The phase-diagram of this three-component system has been constructed and these polyhedral vesicles are found to be in the gel (L beta) phase. Confocal laser-scanning microscopy was used to confirm the complex morphology of these vesicles. The thermo-responsive nature of release of entrapped carboxyfluorescein and nicotinamide adenine dinucleotide has been studied; release is increased with increase in temperature (37 degrees C) even though the polyhedral vesicles still maintain their polyhedral shape at this temperature. The results indicate that the thermo-responsive features of the niosomes are a result of reversible changes in bi-layer permeability caused by temperature-mediated alteration in the membrane-packing characteristics of the polyethoxylated cholesterol ether.

The effect of monomers and of micellar and vesicular forms of non-ionic surfactants (Solulan C24 and Solulan 16) on Caco-2 cell monolayers

Measurements of transepithelial electrical resistance (TEER), the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide) test and monitoring of poly(ethylene glycol) (PEG) transport have been used to study the effects of the non-ionic surfactants Solulan C24 and Solulan 16, either free in solution or as an integral part of niosome bi-layers, on intestinal epithelial cells from man (Caco-2 cell monolayers). The effects on epithelial integrity and on the transport of the hydrophilic drug metformin depend on the concentration of the surfactants. At concentrations above 1% the effect on TEER of the surfactant in niosomal form and free in solution were equivalent whereas cell viability was preserved to a higher concentration of Solulans when the Solulans were present in the niosomal form. It was concluded that the toxic effect of niosomes arises from free surfactant present in the niosome suspension.

Adverse drug events related to dosage forms and delivery systems

While some of the adverse events caused by the administration of medicines are specifically attributable to the drug molecule, a proportion arises because of the chemical, biological and physical nature of the formulation. The effects may be compounded by certain patient factors, an incomplete understanding of the behaviour of the formulation or the coadministration of other drugs. This review examines adverse drug reactions and other adverse events arising from the nature of the dosage form or formulation used. These adverse effects may be the result of local irritation/toxicity, hypersensitivity or allergic reactions, systemic effects from essentially local therapies, or idiosyncratic reactions in a small number of individuals. In certain cases where the exact nature of the formulation is unknown, adverse events cannot be attributed to any single ingredient. In addition, the total of all ingredients of a formulation, even where details of the formulation are clear, may give rise to abnormal behaviour of the formulation in vivo. Often the desired objective of a particular specialised formulation leads to an unforseen but related adverse effect, and in certain instances these events are completely unpredictable and at variance with the perceived objectives of the formulation.

The activity of doxorubicin niosomes against an ovarian cancer cell line and three in vivo mouse tumour models

Demonstration of the improved doxorubicin pharmacokinetics and tumoricidal activity, after a single intravenous dose of 10mg kg-1 doxorubicin sorbitan monostearate (Span 60) based niosomes in the mouse adenocarcinoma (MAC) tumour model (Uchegbu et al., 1995) preceded the present study in which the activity of doxorubicin C16G2 (a hexadecyl diglycerol ether) based niosomes was evaluated against naive and established MAC tumour models. C16G2 niosomes were equiactive with doxorubicin solution. It is concluded that while in some tumour models, niosomal formulations demonstrate some advantages over the free drug, caution is advocated in the extrapolation of these results. The activity of doxorubicin C16G2 and Span 60 niosomes was also studied against a human ovarian cancer cell line and its doxorubicin resistant subline. There was a slight reduction in the IC50 against the resistant cell line when the drug was encapsulated in Span 60 niosomes in comparison to the drug in solution. Taking into account the in-vitro release characteristics of the various niosomal formulations, it is concluded that the use of niosomal formulations against multidrug resistance shows sufficiently encouraging results to warrant further study.

Distribution, metabolism and tumoricidal activity of doxorubicin administered in sorbitan monostearate (Span 60) niosomes in the mouse

PURPOSE

Encapsulation of doxorubicin in niosomes was sought as a route to tumour targeting and improved tumoricidal through the alteration of doxorubicin pharmacokinetics and metabolism.

METHODS

Doxorubicin niosomes (10 mg kg-1 doxorubicin) prepared from sorbitan monostearate (Span 60), cholesterol and choleth-24 (a 24 oxyethylene cholesteryl ether) in the molar ratio 45:45:10 were administered intravenously to female NMRI mice bearing the MAC 15A subcutaneously implanted tumour. Plasma doxorubicin was fractionated by gel filtration and quantified by HPLC with fluorometric detection as niosome-associated doxorubicin and released doxorubicin. Tumoricidal activity of the formulation was assessed by the intravenous injection of 5 mg kg-1 and 10 mg kg-1 doxorubicin niosomes to male NMRI mice bearing a 6 day old MAC 15A tumour.

RESULTS

At least 90% of the plasma doxorubicin was associated with the niosome fraction 4 h after dosing, and 50% was still associated after 24 h. The clearance of doxorubicin released from the niosomes was about 10 fold greater than the clearance of niosomal doxorubicin (176.5 mL h-1 and 16.2 mL h-1, respectively). The area under the plasma level-time curve increased 6 fold when doxorubicin was administered in niosomes, compared to doxorubicin solution (66.0 micrograms.h mL-1 and 10.3 micrograms. h mL-1, respectively). The area under the tumour level time curve was increased by over 50% by the administration of doxorubicin in niosomes when compared to the drug administered in solution (58.6 micrograms.h mL-1 and 34.3 micrograms.h mL-1, respectively). There was no statistically significant difference between levels of the drug in the heart when niosomal doxorubicin or doxorubicin solution were administered. Doxorubicin metabolites, namely doxorubicinol and the aglycones doxorubicinone, doxorubicinolone and 7-deoxydoxorubicinone, were found associated with the niosomes in the plasma, possibly due to their adsorption to the vesicle surface once formed outside the niosome. Overall metabolite levels in the liver were increased when doxorubicin niosomes were administered compared to the drug in solution. A 5 mg kg-1 injection of doxorubicin niosomes produced a terminal mean tumour weight that was similar to that obtained from animals administered 10 mg kg-1 doxorubicin solution.

CONCLUSIONS

Modest tumour targeting was achieved by the delivery of doxorubicin in sorbitan monostearate niosomes, increasing the tumour to heart AUC0-24 ratio from 0.27 to 0.36 and a doubling of tumoricidal activity. The overall level of doxorubicin metabolites was also increased.

Drug distribution and a pulmonary adverse effect of intraperitoneally administered doxorubicin niosomes in the mouse

Niosomes (non-ionic surfactant vesicles) prepared from C16G2 (a hexadecyl-diglycerol ether), and loaded with doxorubicin, were administered intraperitoneally to male AKR mice at dose levels of 0, 2.5, 5.0, and 10.0 mg kg-1. Free drug was given at 10.0 mg kg-1 by the intraperitoneal route. At a dose level of 10.0 mg kg-1, peak doxorubicin levels in the central compartment were attained faster with the free drug than with the niosome formulation. However, the peak plasma levels were similar for the free drug and the niosome preparation at the 10 mg kg-1 dose level. With doxorubicin administered as the niosome preparation by the intraperitoneal route at 2.5, 5.0, and 10.0 mg kg-1, mean peak plasma concentrations of the drug showed a tendency to be dose-related although the differences were not significant. Over the 24 h period of the experiment, with doxorubicin at 10 mg kg-1, the niosome formulation delivered significantly more drug to the plasma compartment than the free drug (p < 0.05). When doxorubicin was given in niosomes at 2.5, 5.0 and 10.0 mg kg-1 by the intraperitoneal route, the resulting levels of doxorubicin in cardiac tissue were not dose related and the differences not significant and, although the mean peak cardiac-tissue concentration was higher in animals receiving the free drug at 10.0 mg kg-1 intraperitoneally than in mice given intraperitoneal doxorubicin niosomes at this dose level, the differences were again not significant. There were clinical signs of toxicity in mice given doxorubicin-containing niosomes intraperitoneally at 5.0 and 10.0 mg kg-1, and at post-mortem an accumulation of fluid in the pleural cavity was evident. These changes were not seen in mice dosed intraperitoneally with free drug at 10 mg kg-1, or in animals given doxorubicin niosomes intraperitoneally at 2.5 mg kg-1. In mice dosed intraperitoneally with doxorubicin niosomes at 12.0 mg kg-1 and at a dose volume of 0.2-0.4 mL, histological examination of the lungs demonstrated a congestion of the alveolar capillaries, and an increased number of acute inflammatory cells in the alveolar walls. There was no histological evidence of lung toxicity in mice dosed with doxorubicin niosomes at 12.0 mg kg-1 when the formulation was administered with the higher dose volume of 1.8-2.0 mL. Importantly there was no histological evidence of lung toxicity in mice dosed with empty niosomes intraperitoneally or with doxorubicin niosomes given intravenously at 12.0 mg kg-1.