Asymmetric membrane capsules for osmotic drug delivery. I. Development of a manufacturing process.

This paper describes a novel non-disintegrating polymeric capsule and a manual and a semi-automatic process developed for its manufacture. The capsule wall was made by a phase inversion process in which the membrane structure was precipitated on stainless steel mold pins by dipping the mold pins into a coating solution containing a polymer-solvent-nonsolvent system followed by dipping into a quench solution. The resulting asymmetric membrane wall was composed of a thin dense region supported on a thicker porous region. The asymmetric membrane capsules can be filled with a blend of the active agent and excipients for use in osmotically modulated controlled drug delivery applications.

Asymmetric membrane capsules for osmotic drug delivery II. In vitro and in vivo drug release performance.

In a previous paper, we described asymmetric membrane capsules and a phase inversion process for manufacturing them. In this paper, we describe the in vitro and in vivo drug release characteristics from these capsules. The capsule formulations were developed with model drugs to understand the variables that influenced drug release. Studies were also conducted to understand the drug release mechanism and it was shown that osmotic drug delivery was possible with asymmetric membrane capsules.

Perforated coated tablets for controlled release of drugs at a constant rate.

Tablets with a central hole and a water-impermeable coating were prepared. These perforated coated tablets (PCTs) dissolve and release drug through the central hole only. In vitro release of the model drugs sodium benzoate and benzamide from PCTs occurred at a constant rate up to 80% release. The zero-order release rate varies with hole size, drug solubility, drug concentration, diluent solubility, and binder concentration. These results demonstrate that the PCT design can be used to prepare drug delivery devices which release at controllable constant rates.

Calcium accelerates cholesterol phase transitions in analog bile.

Analog bile supersaturated with cholesterol was constituted, filtered and divided into equal portions containing no calcium or calcium, 2.5-15 mM. Aliquots were removed over the next 48 h and filtrates analyzed for cholesterol, bile acid and lecithin. Calcium accelerated cholesterol loss from solution in a dose-related fashion.

Acceleration of cholesterol phase transitions in analog bile by chlorpromazine, pentobarbital and ethinyl estradiol.

The effects of incorporation of chlorpromazine, pentobarbital and ethinyl estradiol on the maintenance of cholesterol supersaturation was studied in bile analogs. Bile solutions were initially supersaturated and microscopically clear. Chlorpromazine and pentobarbital were almost totally solubilized; ethinyl estradiol was poorly solubilized. Chlorpromazine and pentobarbital in concentrations of 5 and 10 mg/ml rapidly (less than 3-5 h) diminished cholesterol in bile filtrates compared to controls; ethinyl estradiol did so at a concentration of 1 mg/ml but less rapidly (24 h). Bile acid and lecithin concentrations, over time, did not differ significantly between groups. The results indicate that drug interactions with bile constituents, without causing their precipitation, can alter the maintenance of cholesterol supersaturation and phase transitions in bile.

Penetration of guinea pig skin by acyclovir in different vehicles and correlation with the efficacy of topical therapy of experimental cutaneous herpes simplex virus infection.

Inadequate penetration of antiviral agents through the stratum corneum of the skin may be one of the limiting factors in the topical therapy of recurrent cutaneous herpes simplex virus infections in humans. In vitro studies of the penetration of the nucleoside analog acyclovir (ACV) through guinea pig skin demonstrated a marked increase in drug flux when ACV was formulated in dimethyl sulfoxide (DMSO), compared with water or polyethylene glycol (PEG) as the vehicle. To examine whether the increased transcutaneous flux of ACV effected by DMSO was meaningful in vivo, topical 5% ACV in DMSO was evaluated for the treatment of cutaneous herpes simplex virus infection in guinea pigs and compared with topical 5% ACV in PEG. When compared with infection sites treated with the vehicle alone, ACV in DMSO produced a greater percent reduction than did ACV in PEG in median lesion number (8 versus 58%; P less than 0.001), median lesion area (35 versus 73%; P = 0.001), and median lesion virus titer (21 versus 84%; P = 0.08). We conclude that DMSO is a highly effective vehicle for topical administration of ACV and is superior to PEG in our model. Careful choice of vehicle and consideration of transcutaneous penetration may be important for realization of the full potential of topical antiviral therapy in humans.

Free fatty acid-induced platelet aggregation: studies with solubilized and nonsolubilized fatty acids.

The aggregation response of washed porcine platelets to the sodium salts of stearic, oleic, palmitic, and myristic acids was analyzed turbidometrically. The fatty acids were prepared as aqueous suspensions and as taurocholate- or albumin-solubilized systems. The final concentration of fatty acid in the platelet preparation varied between 70 and 600 microM. This range was within or below the normal physiological limits of 300-1200 microM. Platelet aggregation was observed with both the suspended and taurocholate-solubilized fatty acids. The extent of platelet aggregate formation increased with the fatty acid concentration and chain length. With the exception of stearate, the taurocholate-solubilized fatty acids were more active than the suspensions. Albumin-solubilized fatty acids were devoid of platelet aggregating activity. Particle-size analysis of the solubilized fatty acids indicated that fatty acid precipitation had occurred subsequent to the addition of taurocholate-solubilized fatty acids to the platelets. This precipitation did not occur with the albumin-solubilized systems, suggesting that the fatty acids must assume a particulate physical state to induce aggregation. Platelet aggregation induced by fatty acids was not inhibited by 80 nM epoprostenol, 75 microM alprostadil, or 150 microM indomethacin. This finding indicated that the fatty acid-induced platelet aggregation was independent of cyclic AMP-related calcium shift, cyclooxygenase-arachidonate, or granular nucleotide release mechanisms.

Progestin permeation through polymer membrane V: Progesterone release from monolithic hydrogel devices.

Progesterone release from monolithic devices prepared from various copolymers of poly(hydroxyethyl methacrylate) and poly(methoxyethoxyethyl methacrylate) or poly(methoxyethyl methacrylate) was examined. In general, plots of the fraction of drug released versus (time)1/2 were linear during the early stages of drug release. This behavior is similar to that found for drugs released from hydrophobic polymers such as polydimethyl siloxane. However, for some release curves using the hydrogels, a breakpoint appeared during the early stages of drug release. These breakpoints were due to the effects of water absorption by these polymers. From analyses of permeability coefficients, it was demonstrated that release rates also were dependent on the initial drug load and the equilibrium water content of the polymer. These conclusions were verified from cross-membrane diffusion studies on films depleted of their initial drug load. In conjunction with this work, the aqueous solubility of progesterone was determined by several methods; an average value of 38 micrograms/ml at 23 degrees was obtained.

Progestin permeation through polymer membranes IV: Mechanism of steroid permeation and functional group contributions to diffusion through hydrogel films.

Hydrogel films were prepared from hydroxyethyl methacrylate, both with (Film II) and withouth (Film I) 5.25 mole% of ethylene glycol dimethacrylate. Permeation, diffusion, and partition coefficients for progesterone, testosterone, nandrolone, norethindrone, 17 alpha-hydroxyprogesterone, estradiol, and hydrocortisone were determined. A solute permeation model was proposed based on the separation of a domain (B) composed of "bulk-like" water and a doman (A) composed of polymer, interfacial water, and bound water present in the films. The separate contributions from the "pore" and "solution-diffusion" mechanisms to the total permeability were calculated from the model. Steroid permeabilities through Films I and II were analyzed in accordance with this model. Permeation of Film II occurred via the solution-diffusion mechanism. Permeation of Film I occurred predominately by the pore mechanism with a small contribution (approximate 20%) from the solution-diffusion mechanism. The latter contribution was dependent on the solubility of the solute within the A domains of the hydrogel film. Functional group contributions to permeation of Film II were ascribed to either steric or hydrogen bonding effects.

Progestin permeation through polymer membranes III: Polymerization solvent effect on progesterone permeation through hydrogel membranes.

Hydrogels prepared from poly(hydroxyethyl methacrylate) are biocompatible and highly permeable to low molecular weight solutes. Permeation rates can be varied by altering the cross-linker concentration or using copolymers; the latter are chosen to alter the hydrogel equilibrium hydration. These factors suggest that hydrogels are good candidates for controlled-release drug delivery devices. Hydrogels may be synthesized using various temperatures, initiators (nature and concentration), and solvents (nature and concentration). This study demonstrated that progesterone permeation through poly(hydroxyethyl methacrylate) films is independent of polymerization solvent (nature and concentration) for the solvents, water, ethanol, and tert-butyl alcohol. The importance of hydrogel equilibrium hydration in progesterone permeation is emphasized.

Progestin permeation through polymer membranes I: diffusion studies on plasma-soaked membranes.

The potential of several commercially available polymeric materials for use in controlled-release drug delivery devices was investigated. Progesterone was used as a model hydrophobic drug. The progesterone permeation rates through polydimethylsiloxane, two polyether urethanes, a hydroxyethyl methacrylate, a polyether urethan--polydimethylsiloxane blend, and a cellulosic membrane were determined. The permeabilities were obtained on nonsoaked membranes and on membranes soaked in plasma for varying times. The purpose of the plasma soaks was to examine the effects of lipid absorption and degradative processes within the membrane on progesterone permeability. This study identified several polymers that show potential for use in controlled-release drug delivery devices. The plasma treatment studies showed that several polymers may not be acceptable. The plasma soak studies were interpreted in terms of the mechanisms of drug permeation through the membranes.

Progestin permeation through polymer membranes II: diffusion studies on hydrogel membranes.

The potential use of hydrogels in controlled-release drug delivery systems for contraceptive steroids was investigated. The permeabilities, diffusion coefficients, and partition coefficients for progesterone were determined for hydrogels made from hydroxyethyl methacrylate containing varying amounts of ethylene glycol dimethacrylate and tetraethylene glycol dimethacrylate. In addition, copolymers of hydroxyethyl methacrylate with methoxyethyl methacrylate and methoxyethoxyethyl methacrylate were investigated. The results were interpreted in terms of the mechanisms of permeation of progesterone through the hydrogels. This study showed that progesterone permeated these membranes primarily through loose pores in the hydrogel network except at high concentrations of the cross-linker, ethylene glycol dimethacrylate, where dissolution and diffusion of the progesterone in the polymer network was the dominant mechanism.

Light-scattering studies on bile acid salts II: pattern of self-association of sodium deoxycholate, sodium taurodeoxycholate, and sodium glycodeoxycholate in aqueous electrolyte solutions.

The pattern of self-association of the bile salts sodium deoxycholate, sodium glycodeoxycholate, and sodium taurodeoxycholate was investigated in aqueous electrolyte solutions by the light-scattering technique. The turbidity of the bile salt solutions was obtained over the concentration range of 0-20 mg/ml at 25 degrees. These data were analyzed according to a monomer-micellar equilibrium model and a stepwise association model. Comparison of the light-scattering data with these models suggests that the monomer-micellar model may be inappropriate. Analysis of the data according to the stepwise association model suggests that the dihydroxy bile salts associate to form dimers, trimers, and tetramers in addition to a larger aggregate which varies in size depending on the degree of conjugation of the bile salt.

Solubilization of napthalene by sodium cholate and pattern of self-association of sodium cholate in 0.15 M sodium chloride.

Naphthalene solubility was determined in aqueous 0.15 M NaCl containing sodium cholate in the 0-0.05 M concentration range at 25 +/- 0.1 degrees. Sodium cholate tends to self-associate in aqueous solutions. Most often, the association pattern has been described in terms of a monomer-micellar model in which it is assumed that no association occurs below the critical micelle concentration. By comparison of the experimental solubilization curve with curves calculated on the basis of the monomer-micellar model, it was shown that this model is inappropriate for the self-association pattern of sodium cholate. The solubility data were consistent with a model that assumes that sodium cholate associates to form dimers, trimers, and higher aggregates with an average aggregation number of 7.63. Model calculations suggest that naphthalene is solubilized by dimers and higher aggregates. Solubilization of naphthalene by trimers appears to be negligible.

Light-scattering studies on bile acid salts I: Pattern of self-association of sodium cholate, sodium glycocholate, and sodium taurocholate in aqueous electrolyte solutions.

The pattern of association of the trihydroxy bile salts in aqueous electrolyte solutions was investigated utilizing the light-scattering technique. The turbidity of the bile salts sodium cholate, sodium taurocholate, and sodium glycocholate was determined over the concentration range of 0-25 mg/ml at 25 degrees. For sodium cholate, the concentration of the supporting electrolyte was varied from 0.15 to 0.5 M. For all bile salts in 0.15 M electrolyte, the turbidity was determined in sodium fluoride, sodium chloride, sodium bromide, and sodium iodide. Comparison of the light-scattering data with amonomer-micellar model showed that qualitative agreement was obtained; however, quantitative agreement could not be achieved. Further examination of the data showed that the light-scattering results were in good agreement with a model that includes dimers, trimers, and a higher aggregate containing approximately eight monomeric units.

Solubilization as a method for studying self-association: solubility of naphthalene in the bile salt sodium cholate and the complex pattern of its aggregation.

Solubilization of uncharged, slightly soluble solutes is shown to be a useful approach for investigating patterns of self-association. The solubility of naphthalene in aqueous solutions of sodium cholate was determined over the concentration range of 0-0.20 mole/liter at 25 degrees. Bile salts such as sodium cholate have many detergent-like properties and exhibit hydrophobic self-association in aqueous solutions. It has become cutomary to describe this aggregation using the model of micelle formation. The naphthalene solubility data show that the CMC for sodium cholate is not well defined. Comparison with solubilization in a typical micelle-forming system, sodium decanesulfonate, shows clearly that sodium cholate does not resemble a micelle-forming system. Further examination of the solubility data in terms of mutual association of naphthalene with aggregate species shows that the self-association of sodium cholate is not consistent with the formation of (a) only large micelles containing 10 or more monomers, (b) only dimers, (c) dimers and large micelles, and (d) any unique oligomer or multimer. A complex pattern of association, including the formation of dimers and one or more higher oligomers, is indicated.