This study comprehensively investigated the utility of our novel hydrogel-forming MN arrays in transdermal drug delivery. Mechanical properties, in vitro/in vivo permeation of model drug substances, combined effect with iontophoresis (ITP), biocompatibility, and safety in human subjects were investigated. The methods for each study are detailed below.
Poly(methylvinylether/maelic anhydride) (Gantrez AN-139) was provided by ISP, Guildford, UK. Poly(ethyleneglycol) 10,000 daltons and all drug substances were obtained from Sigma-Aldrich, Poole, Dorset, UK. All other chemicals were of analytical reagent grade.
2.2. Preparation and Mechanical Testing of MN
Aqueous blends containing 15% w/w poly(methylvinylether/maelic acid) (PMVE/MA) and 7.5% w/w poly(ethyleneglycol) 10,000 (PEG) were utilized to fabricate MN by using laser-engineered silicone micromould templates.10, 15, 16 Optimum polymeric composition was determined in our previous hydrogel work.11–13 MN were crosslinked (esterification reaction) by heating at 80 °C for 24 h.11–13 Mechanical properties of these hydrogel MN (with 600 μm height, 300 μm width at base, 150 μm interspacing, and 3 × 3 arrays) were determined as reported previously.8, 10, 15, 16
2.3. In Vitro Permeation Studies
The ability of the novel hydrogel-forming MN in enhancing and controlling transdermal drug delivery was investigated by using six hydrophilic solute molecules of increasing molecular weight: fluorescein-isothiocyanate labelled bovine serum albumin (FITC-BSA), insulin, methylene blue (MB), caffeine (CF), theophylline (TP), and metronidazole (MZT) with molecular weights of ≈67,000, ≈6000, 319.85, 194.19, 180.17, 171.15 Daltons, respectively. For studies involving high-molecular weight molecules, a pore-forming agent (NaHCO3) was added to the hydrogel formulation. Adhesive patches, as stated below, containing the above molecules at defined loadings, were then attached to the upper baseplates of hydrogel-forming MN, with the novel composite system termed as “integrated MN” (Figure 2a). Permeation was then investigated from this integrated MN system and compared with that of the adhesive patches alone.
Figure 2. Fabrication and mechanical properties of novel hydrogel MN arrays. a) The method of fabrication of the integrated hydrogel-forming MN system from aqueous polymeric blends using custom laser-engineered silicone micromoulds. b) Percentage reduction in height of hydrogel-forming MN following axial load, Means ± S.D, n = 45. c–e) Digital microscopic images of hydrogel-forming MN following axial compression; where c) MN were subjected to a force of 0.05 N per needle, d) MN were subjected to a force of 0.36 N per needle, and e) MN were subjected to a force of 1.42 N per needle. f,g) Digital microscope images of hydrogel-forming MN subjected to transverse forces; f) before application and g) after application of 0.51 N per needle. h) A histogram depicting the percentage number of holes created in the stratum corneum of neonatal porcine skin in vitro following application of a range of insertion forces. i) A digital image of methylene blue staining showing 100% MN penetration of the stratum corneum of neonatal porcine skin in vitro, following application of an insertion force of 0.03 N per needle or greater. j) Forces required to break the MN base-plates and the angle of bending of the MN base-plates at break, Means ± SD, n = 6. The scale bar in (i) represents 200 μm. *P < 0.05 and ** P < 0.001. Error bars in b, h, and j indicate standard deviations.
Download figure to PowerPoint
Adhesive patches were prepared, by using a casting method, from aqueous blends of 10% w/w PMVE/MA and 5% w/w tripropyleneglycol methyl ether (TPM).28 Required amounts of solute molecules were either added directly into the aqueous blends or dissolved in 0.01 M HCl prior to addition to the aqueous blends. An aliquot (2.7 g) of aqueous blend was cast into silicone moulds (30 × 30 mm2) and dried under a gentle air flow for 48 h. Attaching the formed adhesive patches with gentle pressure to hydrogel-forming MN yielded the integrated MN system.
Transdermal permeation was investigated in vitro across dermatomed neonatal porcine skin (300–400 μm thickness, previously shown by us to be the optimum skin model for prediction of in vivo performance of dissolving polymeric MN)29 by using a modified Franz-cell setup, as described previously.10, 15, 16 Integrated MN were applied to the dermatomed skin by using a custom spring-activated applicator16 at a force of 11.0 N per array. Adhesive patches were applied using gentle pressure. Sample volumes of 300 μL were withdrawn, in triplicates, and analysed by the methods below. The release medium for FTIC-BSA was 0.05% w/w SDS in phosphate buffered saline (PBS) at pH 7.4, for insulin 0.1 M Tris buffer at pH 10 (due to stability issues) and for the other drug molecules PBS at pH 7.4 was used.
2.4. Analysis of Drug Permeation
Receptor medium samples for insulin, TP and MZT were analysed using validated analytical methods reported previously.10, 15, 30 FITC-BSA was quantified using a gradient HPLC method (Agilent Technologies 1200 Series, Stockport, UK) in which the separation was performed on a C4 (4.6 mm × 50 mm, 5 μm) analytical column (Symmetry 300, Waters Ireland, Dublin). The mobile phase gradient consisted of 0.1% triforoacetic acid (TFA) in acetonitrile (ACN) and 0.1% TFA aqueous solution. The gradient linearity changed from 20:80 (ACN: TFA) to 90:10 in 8 min and continued for 1.0 min at the ratio of 90:10. The reversal to the initial conditions was attained within 1.0 min, followed by a 3.0 min re-equilibration period. The total analysis run time was 13.0 min. The injection volume was 20 μl and elution flow rate was 1.0 ml min−1. FITC-BSA was detected using a fluorescence detector with excitation and emission wavelengths set at 490 nm and 520 nm, respectively. Data were processed using Agilent Rapid Res software. CAF was analyzed following modification of a literature method.31 Briefly, an isocratic HPLC method was used with a reverse phase C-18 (Waters Spherisorb 5 μm ODS 4.6 × 150 mm, Waters Ireland, Dublin) analytical column. The column was thermostated at 35 °C. The mobile phase consisted of 90% of 0.52% glacial acetic acid and 10% of 50% ACN: 50% tetrahydrofuran. The flow rate was 1.0 ml min−1. Sample injection volume was 50 μl. UV detection was performed at 273 nm. Data were processed using Agilent Rapid Res software. MB was analysed at 664 nm using a UV microplate reader (Powerwave XS, Bio-Tek Instruments Inc., Minooski, USA).
2.5. In Vitro ITP Studies
We further investigated the combined effect of integrated MN and ITP on in vitro permeation of TP, MB, fluorescein sodium (FS), insulin and FITC-BSA. In this case, TP, MB and FS were loaded at a concentration of 3 mg cm−2 in the adhesive patches. FITC-BSA and insulin were loaded at concentration of 2.5 and 5 mg cm−2, respectively. The methodology for this study is similar to that stated above. For studies involving application of an electrical current, an Ag electrode was used as the anode and an Ag-AgCl electrode used as the cathode. The delivery electrode was placed directly on top of the integrated MN, whilst the return electrode was placed into the receiver medium via the side arm of the Franz cell. A commercially available power supply (Phoresor II, Iomed, Lake City, FL, USA) was used to deliver a current of 0.5 mA for a period of 6 h for in vitro studies. A sample volume of 300 μl was collected at predetermined time intervals and was assayed by the analytical methods described above. For samples containing FS, UV-spectroscopy was used at a λmax of 497 nm.
2.6. In Vivo Studies
Delivery of insulin, FITC-BSA and MZT from the integrated MN (with 600 μm height, 300 μm width at base, 50 μm interspacing, & 19 × 19 arrays) was investigated in a rat model. For studies involving insulin, diabetic Sprague-Dawley rats were selected. The methodology was similar to that in our previous publication, which involved soluble MN.10 Positive controls were performed by subcutaneously injecting bovine insulin solution in PBS pH 7.4 at a dose of 0.2 IU per animal. Integrated MN based on adhesive patches containing 5 mg cm−2 insulin were applied to the shaved skin (Gillette Mach 3 Power, Procter & Gamble, Hampshire, UK) on the rats' backs by using the spring-activated applicator at a force of 11.0 N per array. Skin integrity was confirmed using transepidermal water loss measurements made using a Deflin Vapometer (Deflin Technologies Kuopio, Finland) Following application, the integrated MN were held in place for 12 h with the help of a self-adhesive silicone sheet.
Blood samples were collected at pre-defined time intervals over a 12 h period by lateral tail vein prick and blood glucose levels (BGL) were measured using a glucometer (Accu-Check Aviva, Roche Ltd., Mannheim, Germany). Blood glucose levels were expressed as the percentage of initial BGL and calculated values were plotted against time to obtain blood glucose level-time profiles. Cmax, denoting maximum% decrease in BGL, was calculated by subtracting the lowest% BGL from 100, as described elsewhere.32 Tmax, denoting time required to achieve Cmax, was also reported.
Integrated MN containing FITC-BSA at 2.5 mg cm−2 loading in patches, were applied to healthy rats in an analogous way to insulin-loaded patches. A 0.25 ml aliquot of blood was drawn from the tail vein and collected into heparinised tubes (Microvette CB300, Sarstedt, Leicester, UK) at pre-defined time intervals over a period of 24 h. Plasma samples were obtained by centrifuging the collected blood samples at 10 000 rpm for 10 min. All samples were analysed, within 2.0 h of collection using spectrofluorimetry at an excitation wavelength of 490 nm and emission wavelength of 520 nm.
In vivo delivery of MZT from the integrated MNs was also investigated in rat models. The concentration of MZT versus time data obtained using the dried blood spot (DBS) method reported previously30 was subjected to non-compartmental analysis (WinNonlin version 2.1, Pharsight, USA) to determine pharmacokinetic parameters, namely, area under curve (AUC), Cmax and Tmax. The steady state blood concentration (Css) of MZT after application of integrated MNs was calculated by using the equation:
For in vivo studies, involving application of an electrical current, a circular PVC ring (diameter 2.0 cm) was secured onto the rats back at the site of integrated hydrogel MN or adhesive patch application, using cyanoacrylate glue (Loctite Ltd, Dublin Ireland). The Ag electrode anode was then placed on top of the integrated MN or adhesive patch and held in place using a layer of double-sided adhesive tape (Henkel Ltd, Cheshire, UK) and an additional layer of Micropore tape (3M Ireland, Dublin). A second PVC ring (diameter 2.0 cm) was then secured at an adjacent site, ≈ 2 cm away from the site of MN application. An aliquot of PBS pH 7.4 (1.0 ml) was then placed into the centre of this second PVC ring and an Ag-AgCl electrode was attached at the site using Micropore tape. The Phoresor™ II was used to deliver a current of 0.5 mA for a 2 h period after which the integrated MN and/or adhesive insulin/FITC-BSA patch were removed.
All animal experiments throughout this study were conducted according to the policy of the Federation of European Laboratory Animal Science Associations (FELASA) and The European Convention for the protection of vertebrate Animals used for Experimental and Other Scientific Purposes, with implementation of the principle of the 3Rs (replacement, reduction, refinement).
2.7. Biocompatability Evaluation
Biocompatability of PEG-crosslinked PMVE/MA-based hydrogel MN materials was performed in three different cell-lines using an indirect test with two suspended cell-lines; fibroblasts (Balb/3T3) and keratinocytes (NRERT-1)24 and a 3D keratinocyte organotypic raft culture. In all three cases, an MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] assay was used to determine any cytotoxic effects.
To depict conditions pertaining to actual in vivo application of MN, a Franz-cell setup was used to expose only the MN and not the baseplate to the extraction medium. A Silescol membrane was sandwiched between the receptor and donor compartment of the Franz-cells and the hydrogel-forming MN were applied in a similar manner as stated above. The receptor compartment was filled with 12.0 ml of DMEM (Dulbecco's modified Eagle's medium) containing 4.5 g L−1 glucose, 2.0 mM L-glutamine, 10% foetal calf serum, 1.0 IU ml−1 penicillin and 1.0 μg ml−1 streptomycin. After 24 h of application, the DMEM solution was collected and sterile filtered. The medium (100%, 50% and 10% diluted in fresh DMEM) was exposed to cell monolayers of Balb3T3 and NRERT-1 keratinocytes at 1 × 104 cells ml−1 in the 96-well culture plate for 24 hrs. Following 24 h incubation at 37 °C, the test extracts were assessed for cell viability by means of the quantitative MTT assay at 540 nm (Biotrak II Plate Reader, Amersham Biosciences, Cambridge, UK). A total of six replicates were studied for each extract, including control (sterile DMEM media) in the same plate. In 3D keratinocyte organotypic raft type culture studies, the 3D keratinocyte tissues were pre-incubated (at 37 °C and 5% CO2) for 1.0 h in assay medium before starting the experiment. Prior to dosing with MN extract the assay medium was renewed. An aliquot (50 μL) MN extract (undiluted) was applied atop the tissues in triplicate and incubated for 24.0 h. At the end of the exposure period, the tissue cultures were gently rinsed with PBS, incubated in fresh medium for a further 24.0 h and finally placed into plates containing 0.5 mg ml−1 of MTT solution at 1.0 ml cm−2 cultured tissue (4.0 ml). Following 3.0 h incubation and a rinse with PBS, the reduced MTT was extracted by submerging the tissues in 2.0 ml of isopropanol and shaking for 2.0 h. The absorbance of the extraction solutions was again measured at 540 nm. The mean absorbance was calculated for each tested sample (undiluted MN extract). The mean result of the negative control (H2O) tissues was set to represent 100% viability.
2.8. In Vitro 3D Skin Irritancy Test
In vitro skin irritancy tests were performed by using a 3D reconstructed human skin model (EpiSkin, Skin Ethic Laboratories, Lyon, France). The assay endpoint for irritancy measurements was IL-1α (Interleukin-1 alpha). An aliquot (10 μl) of sterile-filtered hydrogel extract, obtained as described above, was exposed to the reconstructed epidermis of the EpiSkin skin model. A negative (PBS) and positive (5% w/v sodium dodecyl sulphate, SDS) control were also included in the protocol. In each case, the epidermis of EpiSkin skin model was exposed at room temperature, according to the manufacturer's recommendations. Following exposure for 24 h, EpiSkin™ constructs were washed with PBS, and transferred to wells that contained fresh maintenance medium, and allowed to recover for 42 h at 37 °C, 5% CO2. The IL-1α content of the recovery medium was assessed using a human IL-1α specific ELISA (Thermo Scientific/Pierce, IL, USA).
2.9. Human Volunteer Studies
In this study, gamma-sterilised hydrogel-forming MN were applied to the ventral forearm skin of six healthy human volunteers (3 men and 3 women, aged between 23 and 31 years), with no pre-existing skin conditions. They were asked not to wash or apply any cosmetic formulations on the ventral forearm during the study period. The School of Pharmacy's Ethical Committee, Queen's University Belfast, approved this study.
Hydrogel-forming MN arrays were attached to waterproof plasters (Elastoplast, Beiresdorf, Hamburg, Germany) using double-sided adhesive tape. The spring loaded applicator was used to apply the MN-containing plasters Each volunteer was subjected to four interventions, a control plaster (without MN), a plaster containing MN baseplate only, a plaster containing hydrogel-forming MN (11 × 11 array) and a plaster containing hydrogel MN (19 × 19 array), as shown.
The study was conducted at a controlled room temperature of 20 °C and a relative humidity of 45 ± 5%. The subjects were acclimatized in this room for 15 min prior to the start of the measurements. This study involved the application of waterproof plasters, as described above, to the ventral forearm on three different occasions; 1) Patches applied and removed immediately (i.e., 0 h treatment group); 2) Patches applied and removed after 2 h (i.e., 2 h treatment group), and 3) Patches applied and removed after 24 h (i.e., 24 h treatment group).
Before application of the patches the ventral forearm were cleansed with a sterile swab (Boots Pharmaceuticals, Nottingham, UK) in each case. Four square areas (≈1.5 cm2) were marked on the ventral forearm of each subject using a ballpoint pen. The square areas were located at equidistance on each forearm. Before the application of the plasters, TEWL and clinical photographs of skin were recorded, as described below. Following application of the plasters, the applicator was activated against each plaster to ensure consistent MN penetration in each volunteer (confirmed using OCT, as described previously).16 Immediately upon application, and after removal of the plasters (i.e., in 0, 2 and 24 h treatment groups), the VAS, TEWL and any change in skin colour was recorded in all the four interventions, as follows;
2.12. Clinical Scoring
Clinical scoring was used to determine irritation threshold in all study volunteers. Clinical scores were based on visual inspection following guidelines of the International Contact Dermatitis (ICD) Research Group and the North American Contact Dermatitis Group. Clinical photographs of the skin before and after plaster application were captured using a Nikon D40X digital camera equipped with a Nikon AF-S VR Micro-Nikkor 105 mm f/2.8G IF-ED lens and a Nikon Close-up Speedlight Commander Kit R1C1 with two wireless remote Speedlight SB-R200 flash units (Nikon UK Ltd, Kingston, UK) mounted on a ring on either side of the lens and using aperture priority. The camera was fixed on a tripod and was adjusted to know height so that the distance between the ventral forearm and the camera lens were consistent throughout the study. Images were then printed on high quality photographic paper and were evaluated and scored blindly by two-experienced dermatologists. A clinical score for each test site was assigned, using the ICD scores. ICD scores were rated as 0, 0.5, 1, 2 or 3, which represent either: a negative, a barely perceptible macular erythema, a mild erythema, a moderate-intense uniform erythema or an intense erythema, vesiculation or erosion, respectively.36, 37 Optical coherence tomographic (OCT) studies, which allowed measurement of depth of MN insertion and permitted study of in-skin swelling, were conducted as described previously.16