Potential of thermally induced phase separation as a porogen technique has been studied in an effort to produce a surgical implant suitable for cell transplantation. Emphasis has been placed on the liquid-liquid phase separation of solutions of amorphous poly DL-lactide and semicrystalline poly L-lactide in an 87/13 dioxane/water mixture. The related temperature/composition phase diagrams have been set up by turbidimetry, and the possible occurrence of a gel has been discussed. Freeze-drying of some phase-separated polylactide solutions can produce flexible and tough foams with an isotropic morphology. Interconnected pores of 1-10 microns in diameter are expected to result from the spinodal decomposition of the polylactide solutions with formation of co-continuous phases. Thermodynamics of the polymer/solvent pair has a decisive effect on the final macroporous foams, as shown by the dependence of their porosity, density, porous morphology, and mechanical behavior on molecular weight and crystallinity of polylactide and concentration of the original solutions. On the basis of the foam characteristics, potential of the liquid-liquid phase separation (spinodal decomposition) has been compared with the solid/liquid phase separation (solvent crystallization) as a porogen technique.

Freeze-drying of polylactide solutions in 1,4-dioxane has been studied as a way to produce microcellular foams. The thermally induced phase separation has been studied in relation to several processing and formulation parameters. The effects of polymer concentration, chain stereoregularity, polymer molecular weight and cooling rate have been investigated in connection with the porous morphology and the physico-mechanical characteristics of the final foams. As a rule, bundles of channels are formed with a diameter of ~100 µ.m. They have a preferential orientation that fits the cooling direction. A porous substructure (~10 µm) is observed in the internal walls of the tubular macropores. Variations in this general porous morphology-and particularly in the porosity, density, solvent residue, mechanical resistance and degree of regularity in the spatial organization of pores-have been observed when polymer concentration in 1,4-dioxane and polylactide stereoregularity are changed. As expected, cooling rate has a strong effect on the foam morphology, which is essentially controlled by the solvent crystallization. Pores are nothing but the fingerprints of 1,4-dioxane crystallites.

This article reports the production of a surgical implant meeting several specific requirements such as biocompatibility, biodegradability, macroporosity, and flexibility. Porosity was controlled by an original method consisting of the aggregation of calibrated poly-D,L-lactide microparticles. The size of the interstices between the aggregated microspheres was in a direct relationship to the microsphere diameter. A first approach was based on coating the microspheres with poly(vinyl alcohol) followed by chemically crosslinking the coating layers that were in mutual contact. This method was disregarded because of the acute cytotoxicity of glutaraldehyde used as the crosslinking agent, the absence of macroporosity, and the complete lack of flexibility. A physical technique of aggregation was then tested, which relied on the plasticization of poly-D,L-lactide microspheres with triethylcitrate to the point where microspheres strongly adhered to each other when they were in contact. This method has proved to be straightforward and definitely superior to the chemical approach, particularly with respect to cytotoxicity, control of macroporosity, and flexibility. A polymer support was thus successfully which was biodegradable, macroporous( interconnected pores of 10-100 microns in diameter), and flexible. This potential medical device is presently being used for neuronal transplantation in the central nervous system.

No abstract

The process of microencapsulation of proteins by double emulsion/evaporation in a matrix of polylactide (PLA) can be divided into three successive steps: first, an aqueous solution of the active compound is emulsified into an organic solution of the hydrophobic coating polymer; second, this primary water-in-oil emulsion (w/o) is dispersed in water with formation of a double water-oil-water emulsion (w/o/w); third, the organic solvent is removed with formation of solid microparticles. This paper focuses on the effect of primary emulsion stability on the morphology and properties of polylactide microparticles loaded with bovine serum albumin (BSA) used as model drug. Depending on the stability of the primary emulsion, the internal structure of microparticles can be changed from a multivesicular to a matrix-like structure. Similarly, the average porosity can be controlled in a range from a few tenths of a micron to ca. 20 to 30 microns. This morphology control could find potential applications not only for the controlled drug delivery but also for the production of microporous particles intended for some specific applications, such as cell culture supports and chromatographic matrices. Although, the interplay of several processing parameters (polymer precipitation rate, polymer coprecipitation with interfacial compounds such as protein or surfactant, stirring rate...) may not be disregarded, this study also indicated that a high loading of a hydrophilic drug can only be expected from a stable primary emulsion. When the stability of the primary emulsion is such as to prevent formation of macropores (> 10 microns), the total pore volume is close to that of the originally dispersed aqueous drug solution.