Skin Permeation of Rosiglitazone from Transdermal Matrix Patches

Insulin resistance is a common feature characterizing the pathogenesis of Type 2 diabetes mellitus. Epidemiologic studies have found a strong association between insulin resistance and Type 2 diabetes (1). For example, more than 90% of African Americans, Hispanics, and nonhispanic whites with Type 2 diabetes in the United States were identified as insulin resistant (2). Rosiglitazone [(±)-5-[[4-[2-(methyl-2-pyridinylamino) ethoxy] phenyl] methyl]-2, 4-thiazolidinedione, (Z)-2-butenedioate (1:1)], a member of the thiazolidinedione class of antidiabetic agents, improves glycemic control because it has a high affinity for peroxisome proliferators-activated receptor-gamma (PPARγ). Thus, it regulates the transcription of certain insulin-responsive genes and improves insulin sensitivity (3). It controls glucose production, transport, and utilization. Although recent studies claim that the drug could be a risk for cardiac patients, it is still a good option for improving insulin sensitivity (4–6).

Rosiglitazone is a weak base with a molecular weight of 357.44 Da. As a maleate salt, 2 to 8 mg of it is administered in tablet form twice daily, with or without a combination of metformin. However, rosiglitazone is metabolized extensively by n-demethylation and hydroxylation, followed by conjugation with sulfate and glucuronic acid in the liver (7, 8). Therefore, alternative routes such as transdermal delivery may be a good choice to deliver the drug directly into systemic circulation through intact skin by bypassing the hepatic first-pass effect to reduce dose frequency by maintaining a prolonged therapeutic blood level of rosiglitazone.

A transdermal patch is a medicated adhesive patch placed on the skin to deliver drugs into the bloodstream (9). The drug has a log P value of 2.1, which contributes to its lipid solubility and hydrophilicity. The oral absorption of rosiglitazone is 99%. Clearance of the drug, in terms of excretion in breast milk, is unknown; renal excretion of metabolized drug is 64%, and feces show the presence of about 23% of the parent compound. The elimination half-life of the drug is 3–4 h, and total protein binding is 99.8% to albumin. The drug has the volume of distribution of 17.6 L. These factors suggest that the drug is a suitable candidate for transdermal drug-delivery systems (TDDS). The authors evaluated the efficacy of the formulations in streptozotocin (STZ)-induced diabetic rats and examined the pharmacokinetic parameters of drug administered through TDDS in humans. This study describes how to develop a matrix–type transdermal patch containing rosiglitazone maleate using pressure-sensitive adhesives (Duro-Tak 387-2516 and Duro-Tak 87-2852). A blend of Duro-Tak polymers can achieve the desired consistency and tackiness for a TDDS, without any gelling agents. Duro-Tak polymers are acrylate-polymeric solutions containing acrylic acid, methyl acrylate, and 2-ethylhexyl acrylate at different ratios. In the present study, transdermal matrix patches containing rosiglitazone maleate using combinations of polymers Duro-Tak 387-2516 and Duro-Tak 87-2852 in ratios of 4:5 and 4:6 were developed and evaluated in vitro and in vivo.

Materials

Rosiglitazone maleate was obtained as a gift sample from Sun Pharma (Mumbai). Duro-Tak 387-2516 (a semisolid polymer with boiling point of 60 °C, solid content of 41.5%, and viscosity of 4350 cp) and Duro-Tak 87-2852 (a semisolid polymer with a boiling point 65.5 °C, solid content of 33.5%, and viscosity of 2550 cp) were gifts from the National Starch and Chemical Company (Bridgewater, NJ). Polyvinyl alcohol was purchased from S.D. Fine-Chem (Boisar, India). Streptozotocin (STZ) was purchased from Sigma-Aldrich (Bangalore, India). All other chemicals used were of analytical reagent grade.

Methods

Drug-solubility study. Various percentages (10%, 20%, 30%, 40% v/v) of polyethylene glycol (PEG) in a normal saline or phosphate buffer have been reported as receptor fluids to study the skin permeation of drugs (10, 11). The mixtures provide biphasic characteristics of the sebum–sweat mixture, the main fluid responsible for in vivo skin permeation (12). The authors selected 20% v/v PEG in normal saline as a receptor fluid. The solubility of the drug in 20% v/v PEG in normal saline was determined by adding an excess amount (1 g) of drug in 2 mL of solvent mixture and keeping the flask containing the solution on a mechanical shaker for 24 h at 25 °C (13). These parameters were enough to provide a saturated solution of the drug. After 24 h, the solutions were transferred into test tubes and centrifuged at 2000 rpm for 30 min at room temperature. The solutions in the test tubes were allowed to stand for 30 min, and 1 mL of supernatant liquid was placed in a 100-mL volumetric flask and diluted with 100 mL of the solvent mixture (20% v/v PEG 400 in normal saline). After thorough mixing for 5 min, 5 mL solution from each volumetric flask was mixed with 5 mL of solvent mixture in a test tube. The mixture was filtered, and the absorbance was read at 318 nm with an ultraviolet–visible spectrophotometer (Genesys Thermoelectocorporation, Madison, WI) against a blank. The amount of drug dissolved was quantified from the calibration curve prepared previously.

Preparation of the patch. To prepare the backing membrane, polyvinyl alcohol (6% w/v) was added to warm (about 40 °C) double-distilled water, and the mixture was constantly stirred on a magnetic stirrer at 60 °C for 15–20 min to attain a homogeneous solution; 3 mL of homogeneous solution was then poured into hollow, glass, cylindrical mold (4 cm in height and 2.55 cm in internal diameter) on the end wrapped with aluminum foil. A smooth, uniform, transparent backing membrane was obtained after keeping the mold at 60 °C for 24 h (14). Backing membranes were removed from some of the molds, and their individual thicknesses were determined.

Initially, various combinations of the polymers were screened to prepare patches to achieve the appropriate consistency and physicochemical properties. The best two prepared formulations, in terms of the criteria mentioned above, are reported here. Blends of Duro-Tak 387-2516 and Duro-Tak 87-2852 were prepared in the volume ratio of 4:5 and 4:6, respectively, by dissolving the respective polymer mixtures in the appropriate volume ratios separately in solvent mixtures consisting of ethyl acetate, isopropyl alcohol, toluene, and n-hexane (12:6:1:1) (14). The volume ratio of the polymers and the solvents was 1:2. A homogeneous solution was made using a magnetic stirrer and a magnetic bead. For each patch, a separate mixture was prepared. Rosiglitazone maleate (12 mg per patch for all experiments other than the antidiabetic study in animals, which used 4-mg patches) was added to each mixture and stirred for 20 min until a homogeneous suspension was obtained; 3 mL of the homogeneous suspension was then cast on the prepared backing membrane and dried at room temperature for 24 h, which yielded a medicated matrix patch of rosiglitazone maleate (see Table I).

Table I: Ingredients and mean thickness of formulation I and II

Patch evaluation and characterization. Drug–excipient interaction. Infrared (IR) spectroscopy (Magna IR 750 series ll, Nicolet, Madison, WI) was carried out for a blend of Duro-Tak 387-2516 and Duro-Tak 87-2852, and for the mixture of drug and polymers. All three samples were placed between disks of sodium chloride windows and scanned over the region of –4000–400 cm^–1. Spectra were compared from computer-data sheets manually, and data were interpreted.

Scanning electron microscopy (SEM). The external morphology of the transdermal patches and the dorsal and ventral sides of the skin surface before the in vitro skin permeation study and 50 h after in vitro skin permeation were analyzed with a scanning electron microscope (JSM 6100, JEOL, Tokyo). The experimental samples were cut into small parts, mounted onto stubs, and coated with gold before SEM analysis (12).

Thickness measurement. The thickness of the backing membrane and of the whole patch (i.e., the adhesive matrix with drug plus backing membrane) were measured using digital calipers (Digmatic Masschieber, model CD-6 CS, Mitutoyo, Tokyo). The average thicknesses of the backing membrane and the whole patch were determined. The average thickness of the adhesive matrix containing rosiglitazone maleate was determined using the following equation:

Area of the patches. The diameter, D, of each patch was measured using a millimeter scale, and the area (π [D/2]²) of each patch was calculated.

Moisture content. The film was weighed and kept in desiccators containing anhydrous calcium chloride for 24 h (12). The film was weighed repeatedly until it became constant. The percent moisture content was determined with the following equation:

Moisture uptake. A weighed film kept in a dessicator was exposed to relative humidity of 75% (saturated solution of sodium chloride) in a dessicator (12). The film was weighed until it showed a constant weight. Percent moisture uptake was determined as follows:

In vitro drug-release study. An in vitro drug-release study was conducted using a US Pharmacopeia-type V dissolution apparatus (Paddle over disc, Electro-lab, Chennai, India). The patches were placed between the stainless-steel disks (of which one side was mesh and one side plate stainless-steel disc meant for transdermal study). The backing membrane side was attached with double-sided adhesive tape (cut same in area as the patch) to the stainless-steel plate so that release could occur from one side only. The drug-release study was carried out at 37 ± 0.5 °C and 100 ± 5 rpm in a dissolution jar holding 900 mL of 20% v/v PEG 400 in normal saline; 5 mL samples were withdrawn at various time intervals and replaced with 5 mL of 20% v/v PEG 400 in normal saline. Samples were analyzed with a UV–vis spectrophotometer at 318 nm against a blank (20% v/v PEG 400 in normal saline) using a validated method (15). The quantity of drug released over time was calculated from the calibration curve. The studies were conducted simultaneously by placing the patch with drug (i.e., the test) and the patch without the drug (i.e., the control) in separate baskets of the same dissolution apparatus. The difference between the test and control readings was the absorbance caused by the drug. In each case, the mean cumulative amount of drug released per square centimeter of patch was plotted against time.

In vitro skin-permeation study. A modified Keshary–Chien (40-mL capacity diffusion cell) was used for the in vitro skin-permeation study. A square section of excised human cadaver skin (R.G. Kar Medical College, Kolkata, India) was stripped of adhering fat and visceral debris. The skin was stored at -80 °C. The thawed skin was then tied with an adhesive tape so that its dorsal side faced upward. A measured portion of transdermal film was placed on the skin, keeping the backing membrane facing upward on the donor compartment. The donor compartment containing the skin and the patch was placed on the reservoir compartment of the diffusion cell containing 20% v/v PEG 400 in normal saline (10, 16). The temperature was maintained at 37 ± 0.5 °C using an air-circulating water jacket. The solution in the receiver compartment was continuously stirred using a magnetic bead (17).

One mL of sample was withdrawn at various time intervals and replaced with 1 mL of 20% v/v PEG 400 in normal saline. The absorbance of the samples was measured in a UV–vis spectrophotometer (Genesys, Thermoelectocorporation) at 318 nm, after appropriate dilution and filtration, against a blank (20% v/v PEG 400 in normal saline). To eliminate the interference of material leaching from the skin or the patch, a control diffusion cell with patch without drug was run in each experiment. The difference in absorbance between the test sample and the control sample was considered the absorbance caused by the drug at a particular time point. The mean cumulative amount of drug permeated per square centimeter of skin was plotted against the time.

In vivo study in animals. The animals used for in vivo experiments were adult Sprague–Dawley rats of either sex weighing 150–200 g. The animals were housed individually in polypropylene cages and were fed the standard pellet diets and water ad libitum. They were kept in a 12-h light–dark cycle at 25 ± 1 °C and 45–55% relative humidity. The in vivo experimental protocol was approved by the Institutional Animal Ethics Committee, Jadavpur University, Kolkata, India.

Experimental design. Animals were divided into seven groups of six rats each. Group I animals (i.e., the normal control) were treated with citrate buffer, pH 4.5. Diabetes was induced in the other rats. After an overnight fast, rats were made diabetic by a single intraperitoneal injection of STZ (Sigma Chemical Company, Mumbai, India) (60 mg/kg, i.p.) dissolved in cold citrate buffer (pH 4.5). After 24 h, all STZ-treated rats showed blood glucose levels between 250 and 350 mg/dL and were considered for further experiments. The diabetic condition of the animals stabilized for five consecutive days. On the sixth day, the experiment was started. Group II animals served as diabetic control. The hair on the backside of the rats in Groups III, IV, V, VI, and VII was removed with a depilatory cream (Anne French, Wyeth Limited, Hyderabad, India). Transdermal patches (each of 4.4 cm²) with Duro-Tak 387-2516 and Duro-Tak 87-2852 in ratios of 4:5 and 4:6, respectively, without drugs were applied to Group III and IV animals. Transdermal patches (4.4 cm² each) of Duro-Tak 387-2516 and Duro-Tak 87-2852 in ratios of 4:5 and 4:6, respectively, each containing 4 mg of rosiglitazone maleate, were applied to Group V and Group VI animals (see Figure 1). Group VII animals were given 5 mg of the drug solution per 1 kg of body weight orally using a round-tipped stainless-steel needle attached to a 1-mL syringe. All the animals were fasted overnight, but water was given ad libitum, blood was taken from a tail vein, and blood glucose was determined at intervals of 0, 12, 24, and 48 h using OneTouch glucometers (Accu-Check, Roche Diagnostics, Germany).

Figure 1 (ALL IMAGES ARE COURTESY OF THE AUTHORS)

In vivo study in humans. This study was performed after receiving approval from the Institutional Ethics Committee at Jadavpur University in Kolkata, India.

Selection of volunteers. Five male volunteers between 24 and 32 years old were selected. The volunteers weighed between 55 and 65 kg each. They had no significant chronic ailments and no history of hypersensitivity, cardiac malfunction, or contraindication to rosiglitazone. They did not receive any medications one month before or during the study, and had not been involved in any clinical trial during the previous six months. Each gave written consent and was selected on a random basis.

Application of the patch. Formulation I was selected for in vivo study on humans because it had the most suitable physicochemical properties and the best permeation profile. Patches were applied on the neck region behind the ear. The application area (see Figure 2) was not washed for the duration of the experiment (i.e., 48 h).

Figure 2

Collection of blood samples. Blood samples of 5 mL were withdrawn from each participant over a period of 48 h at predetermined time intervals of 2, 4, 8, 24, and 48 h. Because permission had been obtained to collect blood five times only from each volunteer in 48 h, the authors selected those time points to understand the initial and subsequent trend of the amount of drug in the blood during that period. The blood samples were withdrawn, collected in vials containing ethylene diamine tetra-acetic acid (10 mg/mL of blood), and plasma was separated.

Separation of plasma from blood samples. The blood samples collected were centrifuged for 10 min at 2000 rpm (Remi Equipment, Mumbai), and the supernatant (i.e., plasma) was pipetted into clean and dry test tubes and immediately frozen at —20 °C until further study was conducted.

Extraction of plasma. The authors quantified rosiglitazone in plasma samples using high-performance liquid chromatography (HPLC) according to a validated method (18). The authors filled 0.9 mL of plasma separated from each blood sample into 10-mL centrifuge tubes. Next, 0.1 mL of the internal standard (pioglitazone 1000 ng/mL) was added to each of those tubes so that the concentration of pioglitazone was 100 ng/mL in all the tubes. The tubes were vortexed for 1 min, then 200 μL of disodium tetraborate was added before the tubes were vortexed for an additional minute. Then 5 mL of dichloromethane was added to the tubes, which then were capped and shaken horizontally for 10 min. The tubes were centrifuged at 3000 rpm for 10 min at room temperature. The organic layer was then removed and evaporated to dryness under a stream of nitrogen at ambient temperature (Zymark Turbo evaporator, Zymark Corporation, San Diego, CA). The authors added 5 mL of a solvent mixture [n-hexane-dichloromethane (80:20 v/v)] to the residue and vortexed the tubes for 30 s. Then, 350 μL of phosphate buffer solution, pH 2.3, was added to the tube, and the solution was vortexed for 2 min and centrifuged at 3000 rpm for 10 min. After phase separation by centrifugation and subsequent removal of the top organic layer, the aqueous phase containing the extracted analytes was analyzed by HPLC (Perkin Elmer, Knauer, Berlin) by injecting 100 μL of the samples in the chromatographic system.

Preparation of mobile phase and buffer. The mobile phase consisted of methanol and mixed-phosphate buffer in the ratio of 30:70 (v/v). Mixed-phosphate buffer (10 mM; pH 2.6) was prepared by dissolving 1.41 g of dibasic sodium phosphate and 1.56 g of monobasic sodium phosphate in 800 mL of water, adjusting the pH of 2.6 with ortho-phosphoric acid, and diluting the solution to 1000 mL with water.

Monobasic potassium-phosphate solution 0.001 M (pH 2.3) was prepared by dissolving 0.136 g of monobasic potassium-phosphate in 800 mL of water, adjusting to pH 2.3 with 0.1 (N) HCl (v/v), and diluting the solution to 1000 mL with water. Disodium-tetra borate solution (0.02 M; pH 9.3) was prepared by dissolving 7.6 g of disodium tetra borate in 1000 mL of water. Methanol, dichloromethane, n-hexane, and water were of HPLC grade.

Preparation of standard stock solution. Standard stock solutions were prepared by dissolving 13.24 mg of rosiglitazone maleate (equivalent to 10 mg of rosiglitazone as free base; molecular weights of rosiglitazone maleate and rosiglitazone as free base are 473.52 and 357.44 g/mol, respectively) in acetonitrile in a 100-mL volumetric flask to yield a primary solution with a concentration of 100 μg/mL. Secondary and working standard solutions were prepared by dilution with the buffer solution, pH 2.3.

Internal standard stock solution was prepared by dissolving 10 mg of pioglitazone in acetonitrile in a 100-mL volumetric flask. This solution was further diluted with the buffer solution, pH 2.3, to yield a concentration of 10 μg/mL.

Preparation of standard plasma. Six separate test tubes, each containing 0.8 mL of drug-free plasma, 0.1 mL of working stock solution of rosiglitazone, and 0.1 mL of internal standard solution (i.e., pioglitazone) were used, which provided a concentration of 100 ng/mL, to yield concentrations of 25, 100, 250, 500, 750, and 1000 ng/mL of rosiglitazone. The extraction of plasma was carried out as described above.

Instrumentation. The chromatographic analysis was carried out on a BDS Hypersil C18 column (5-μm particle size, 250 mm X 4. 6-mm internal diameter, Knauer, Berlin) maintained at 30 °C. The analytes were eluted using a mobile-phase composition of 10 mM mixed-phosphate buffer, pH 2.6, and methanol (70:30, v/v) at a flow rate of 1 mL/min. The mobile phase was premixed, filtered through a 0.45-μm membrane filter (Millipak, Millipore, Billerica, MA), and degassed before use. The peaks were determined using a UV detector (Genesys, Thermoelectocorporation) set at a wavelength of 318 nm.

Calibration curve in plasma. The calibration curve of rosiglitazone was calculated in plasma. Different concentrations such as 25, 100, 250, 500, 750, and 1000 ng/mL of the drug were made in blank plasma. The peak area of drug and internal standard were noted, and the peak area ratio was calculated with the following formula:

Plasma-drug concentration was determined by using a calibration curve.

Results and discussion

Pressure-sensitive Duro-Tak adhesives are generally tacky, which enables them to adhere to the skin with gentle pressure upon application. A good pressure-sensitive adhesive for TDDS can be removed from the skin without leaving a residue (13, 19, 20). Duro-Tak 387-2516 and Duro-Tak 87-2852, in ratios of 4:5 and 4:6, respectively, were selected for formulations because the mixtures had acceptable tackiness and were easily removed without leaving residue on the skin.

A drug–excipient interaction study is an important preformulation study (21). Fourier transform IR (FTIR) spectroscopy has been widely used to identify the interactions between drug and excipient molecules at the level of functional groups. The authors evaluated drug–polymer interaction by analyzing the FTIR spectroscopic data. Figure 3A depicts the spectrum of pure drug rosiglitazone maleate. Figure 3B shows the spectrum of a blend of polymers, Duro-Tak 387-2516 and Duro-Tak 87-2852, in a ratio of 4:5. Figure 3(c) shows the spectrum of drug with a blend of the same polymers in a ratio of 4:5. The figures reveal the absence of a predominant variation in the IR spectra of drug in the Duro-Tak mixture [see Figure 3(c)] compared with those of the drug alone [see Figure 3(a)]. Minor changes in the peaks, mainly between 1500 and 400 cm^–1, could result from the formation of weak bonding such as hydrogen bonding, dipole moment, or bond formation resulting from van der Waals force. The depicted region provides the typical IR absorption frequencies that result from stretching vibrations of the functional groups such as weak –CH₂, out-of-plane bending of –CH=CH₂, out-of-plane –OH bending, –NH₂ and –N–H wagging, and the =CO group. Many of these groups are present both in the drugs and the polymers. Therefore, weak physical bonds may have formed because no major shifting of the peaks was noted.

Figure 3

The surface morphology of the patches and the dorsal and ventral surfaces of the skin samples, before the in vitro skin-permeation study and 50 h after the in vitro skin-permeation study, were scanned with an SEM. Figure 4(a) shows the external morphology of a patch before the in vitro skin permeation. Figure 4(b) shows the external morphology of the patch 50 h after in vitro drug-permeation studies. Figure 4(a) shows drug distribution in a transdermal patch. Figure 4(b) shows several drug particles, both large and small, that were not released. Figure 4(a) shows that the drug distribution was in clusters. Figure 4(b) shows small, holelike structures present on the matrix along with the drug clusters. The holes were probably to the result of the release of drug from the matrices. Furthermore, after 50 h of drug release, patches maintained their structure. The SEM of the morphology of the dorsal surface of the skin before permeation, and the morphology of the dorsal surface of the skin 50 h after in vitro drug permeation, did not vary. The same was true for the SEM of the ventral surfaces taken before permeation and 50 h after permeation study.

Figure 4

Various physiochemical tests were conducted, including the average thickness, the mean area of the patches, their moisture content, and moisture uptake capabilities.

The patches (i.e., Formulations I and II) were circular with an average diameter of 2.37 ± 0.04 cm (mean ± SD, n = 10) and had a mean area of 4.41 ± 0.17 cm² (mean ± SD, n = 10).

The mean thicknesses of the backing membrane, whole patch, and the drug–polymer matrix are shown in Table I. Thin patches were developed. However, Formulation I was thinner than Formulation II. That discrepancy may arise from their different polymeric-blend ratios.

The average percent moisture content of the formulations I and II were 2.13 ± 0.09% and 2.55 ± 0.13% (mean ± SD, n = 10) respectively. Low moisture content in the patches prevents them from forming a dry and brittle film (22).

For in vitro skin-permeation studies, various media such as normal saline, phosphate buffer, and 20% PEG 400 in normal saline were used (10, 12, 14 and 23). Of the media, 20% v/v in normal saline provided biphasic characteristics of in vitro receptor fluid, which is believed to be one of the best media for studying the in vitro skin permeation of a drug (12). The solubility of rosiglitazone maleate in 20% v/v PEG 400 in normal saline was 14.15 mg of the drug dissolved per mL of the media. This result indicated that 20% v/v PEG 400 in normal saline is suitable for in vitro release and skin-permeation studies of the drug.

A drug-release study was conducted to understand the release pattern of a drug from a patch to the surface of the skin. This study which predicts the availability of the drug on the skin surface for skin permeation. Drug release from each formulation was carried out in a USP-type V, dissolution apparatus using 20% v/v PEG 400 in normal saline at 37 ± 0.5 °C (12). The cumulative amount of drug released per cm² of patch was plotted against time for formulations I and II (see Figure 5). The true absorbance of the drug was measured by deducting the absorbance of the control sample from the absorbance of the test sample. The drug-release study showed that drug release sharply increased for 5 h to a mean cumulative amount of 1.5 mg/cm² from both formulations. The release patterns and rates changed afterward, and the drug released in comparatively slower rates till the study was continued. At 72 h, drug release was about 2 mg/cm² of patch for both formulations. The drug present on the patch-matrix surface and near the matrix surface initially released more drug. However, the duration of the release from the patches might be less because of the increased amount of time required for drug molecules to reach the patch surface through the entanglement of polymeric network of the patch matrix.

Figure 5

The in vitro skin-permeation study of a drug predicts the drug's in vivo skin-permeation performance. In the present study, in vitro skin permeation was carried out in a modified Keshary–Chein diffusion cell using cadaver skin. The cumulative amounts of the drug permeated through skin from each cm² of patch area was about 500 μg in the first 2.5 h in both formulations. Then, the skin permeation of the drug in both formulations slowly increased with a similar permeation pattern until the 50-h point (see Figure 6). A similar drug-permeation pattern was maintained in Formulation II until 72 h. The trend varied for Formulation I, where the drug permeated more between 50 and 70 h, compared with the drug in Formulation II. Drug permeation was high during the first few hours. This observation may result from the faster release of the drug (see Figure 5) from the patch surface and near to the patch surface in the patch matrix. During the first 10 h, drug permeation was fast, then it gradually slowed. The slowdown could have resulted from the availability of the drug on the skin surface as depicted in the drug-release data (see Figure 5).

Furthermore, the patches were evaluated on STZ-induced diabetic and normal-control rats. The patches were applied to the animals' backs (see Figure 1). The average blood-glucose levels of Group III and IV animals were similar to those in diabetic-control Group II animals. The plain patch without drug did not alter the glucose levels of animals in Group III and Group IV. Orally administered rosiglitazone maleate reduced Group VII diabetic animals' blood-glucose level to 140 mg/dl from 250 mg/dl in 12 h. However, the blood-glucose level rose to 220 mg/dl in 24 h. This rise necessitated the additional dose to be given to maintain the glucose level (see Figure 7). Again, it is worth mentioning that the glucose level was not reduced to the level of the control animals (i.e., Group I) upon administering the drug orally or transdermally. This suggests that additional therapy such as insulin analogues may be required because STZ damages islets cells irreversibly. Blood-glucose levels in Group V and Group VI animals were effectively reduced for 48 h to 150 mg/dl, which suggests that the patches controlled blood-glucose levels at least for 48 h in the diabetic rats. Formulation I reduced blood-glucose levels more effectively (i.e., in Group V animals) compared with Formulation II (i.e., in Group VI animals). Although numerical values did not differ greatly, the differences were persistent and statistically significant.

Figure 7

The location of the patch on humans is an important factor because the thickness of skin differs at different locations on the body. Different permeation rates were found at different locations when patch-type formulations of scopolamine hydrobromide were applied behind the ear and on the thigh (24). The steady-state permeation rate of this drug was 10 μg/cm²/h from the patch applied behind the ear, where average skin thickness is 0.084mm. The permeation rate was 4.7 μg/cm²/h from the patch applied on the thigh, where average skin thickness is 0.106 mm. Hence, the authors chose to apply the rosiglitazone transdermal patch to the neck region behind the ear.

The plasma-drug level was determined using HPLC. Figure 8(a) shows the chromatogram of pure drug and internal standard in solvent. Figure 8(b) represents the chromatogram of blank plasma. Figure 8(c) represents the chromatogram of rosiglitazone and pioglitazone in plasma for a standard sample. Figure 8(d) is a representative chromatogram of rosiglitazone in plasma after 8 h in humans, where pioglitazone was used as an internal standard.

Figure 8

Drug release from the experimental patches and subsequent penetration through the skin reached a drug concentration in human plasma of 106.45 ng/mL after 2 h, 258.36 ng/mL after 4 h, 460.08 ng/mL after 8 h, 232.63 ng/mL after 24 h, and 143.24 ng/mL after 48 h. The maximum concentration of the drug (C_max) in plasma was 460.08 ng/mL, and this level was reached in 8 h (T_max) (see Figure 9). An important limitation of the present study is that it was conducted only at a few predetermined time points, as permitted by the Institutional Ethics Committee, and based on those time points the C_max and T_max data were calculated. However, values might have varied if more time points had been considered. Thus, these results depict only the trend of drug skin-permeation patterns in humans, and the C_max and T_max of the formulations might not be accurate. The variation in drug concentration in the plasma among the volunteers may result from various factors such as variation in skin type and absorption characteristics. The study reveals that rosiglitazone can permeate through human skin and reach the blood circulation for at least 48 h from formulated patches.

Figure 9

Moreover, currently available antidiabetic agents such as sulfonylureas, meglitinides, and D-phenylalanine derivatives increase β-cell insulin secretion. Biguanides reduce hepatic-glucose production, and glucosidase inhibitors decrease the intestinal absorption of carbohydrates and insulin (25). None of the currently available antidiabetic agents increases insulin sensitivity. Thus, a patient can use any of the above mentioned antidiabetic agents along with a transdermal patch of rosiglitazone to combat the glucose level in Type 2 diabetes mellitus effectively.

Conclusion

Rosiglitazone maleate can be formulated into transdermal matrix patches suitably using Duro-Tak 387-2516 and Duro-Tak 87-2852 pressure-sensitive adhesive polymers without using any gelling agent. Furthermore, sustained-release delivery can avoid the risk of sudden high blood concentration of drug to avoid toxicity, and a prolonged blood level of the drug is suitable for the treatment of Type 2 diabetic patients.

Acknowledgment

This work was funded by the All Indian Council of Technical Education, New Delhi, India, under the Quality Improvement Program for doctoral degrees.

N. Damodharan is a senior research fellow, Gopa Roy is a senior research fellow, Soma Ghosh is a research associate, and Biswajit Mukherjee* is a reader, all in the department of pharmaceutical technology at Jadavpur University, Kolkata -700 032, West Bengal, India, tel: +91 33 24146677, fax +91 33 24146677, biswajit55@yahoo.com.

*To whom all correspondence should be addressed.

Submitted: Feb. 24, 2009. Accepted: Apr. 29, 2009.

References

1. E. Bonora et al., Diabetes 47 (10), 1643–1649 (1998).

2. S.M. Haffner et al., Diabetes 46 (1) , 63–69 (1997).

3. L.A. Werner and T.M. Travaglini, Pharmacotherapy 21 (9) , 1082–1099 (2001).

4. S.E. Nissen and K. Wolski, N. Engl. J. Med. 356 (24), 2457–2471 (2007).

5. S. Singh, Y.K. Loke, and C.D. Furberg, JAMA 298 (10), 1189–1195 (2007).

6. M. Hanefeld, Nat. Clin. Pract. Cardiovasc. Med. 4 (12), 648–649 (2007).

7. S.J. Baldwin, S.E. Clarke, and R.J. Chenery, Brit. J. Clin. Pharmacol. 48 (3), 424–432, (1999).

8. P.J. Cox et al., Drug Metab. Dispos. 28 (7), 772–780 (2000).

9. M. Segal, FDA Consum. 25 (8), 15–17 (1991).

10. P.P. Sarpotdar, J.L. Gaskill, and R.P. Giannini, J. Pharm. Sci. 75 (1), 26–28 (1986).

11. S. Proniuk , S.E. Dixon, and J. Blanchard, Pharm. Dev. Technol. 6 (3), 469–476 (2001).

12. B. Mukherjee et al., Eur J. Pharm. Biopharm. 59 (3), 475–483 (2005).

13. J. Fang et al., Biol. Pharm. Bull. 23 (11), 1357–1362 (2000).

14. B. Mukherjee et al., Pharm. Technol. 30 (3) , 146–163 (2006).

15. B. Mukherjee et al., J. Appl. Res. 5 (1), 96–108 (2005).

16. Y.W. Chien and K.H. Valia, Drug Dev. Ind. Pharm. 10 (4), 575–599 (1984).

17. N.S. Thomas and R. Panchagnula, Eur. J. Pharm. Sci. 18 (1), 71–79 2003.

18. B.L. Kolte et al., J. Chromatogr. B. Analyt. Technol. Biomed Life Sci. 788 (1), 37–44 (2003).

19. Y.W Chien, "Transdermal Therapeutic Systems," in Controlled Drug Delivery: Fundmentals and Applications, J.R. Robinson and V.H.L. Lee, Eds. (Marcel Dekker, New York, 2nd ed., 1987), pp. 524–552.

20. A.V. Pocius, "Adhesives," in Kirk-Othmer Encyclopedia of Chemical Technology, M. Howe-Grants, Ed. (Wiley-Interscience, New York, 1991), pp. 445–466.

21. A. Marini et al., J. Therm. Anal. Calorim. 73 (2), 547–561 (2003).

22. P. Arora and B. Mukherjee, J. Pharm. Sci. 91 (9), 2076–2089 (2002).

23. P.M. Satturwar, S.V. Fulzele, and A.K. Dorle, AAPS PharmSciTech. 6 (4), 49-54 (2005).

24. N.M Price et al., Clin. Pharmacol. Ther. 29 (3) , 414–420 (1981).

25. R.A. Codario, "Oral Agents for Type 2 Diabetes," in Type 2 Diabetes, Pre-Diabetes, and the Metabolic Syndrome: The Primary Guide to Diagnosis and Management, R.A. Codario, Ed. (Humana Press, Totowa, NJ, 2005), pp. 75–92.