Indomethacin was generously gifted by Jagsonpal Pharmaceuticals, New Delhi (India), Sodium Alginate and hydroxypropylmethylcellulose (HPMC) K-100 was procured from Loba Chemie Pvt. Ltd., Mumbai, (India) and Dow Chemicals (USA), respectively. Agar was purchased from Hi-media Pvt. Ltd., Mumbai, (India). All other chemicals were of analytical grade.

The formulation variables and the physico-chemical characteristics of the various batches of the prepared implants are shown in Tables 1 and 2, respectively. Thickness, weight and drug content varied within ? 5%. Kunou et al^{6, 7}, studied the in vitro release from PLGA scleral implants of gancyclovir by incubating the implants in 2 ml of phosphate buffered solution in a shaking water bath at 37?. Balasubramaniam et al^{4, 8}, reported a static method for studying the in vitro release of indomethacin from film type intra ocular implants of indomethacin. As there is a lot of variation in the different methods used by various investigators, an attempt was made in the present study to evaluate the dissolution profiles of the prepared (compressed) implants using four different methods. In all the cases effect of parameters like particle size of the drug, HPMC concentration and calcium chloride concentration were studied.

Drug release from all the methods followed square root of time kinetics, as determined by correlation coefficient values. Further, the release exponent (n) values were also predominantly suggestive of matrix diffusion kinetics, excepting the agar diffusion method, wherein the ?n? values suggested the prevalence of an anomalous (erosion) mechanism in addition to the swelling controlled (matrix) diffusion.

Drug release was independent of the particle size of the drug from all the methods except the static method, where a significant difference (P<0.05) was observed (fig. 1). Fig. 2 shows that, an increase in the concentration of HPMC in the implants caused a significant decrease in the release rate of indomethacin. Though various factors are likely to be responsible for the decrease in drug release with an increase in HPMC concentration from all the methods, some general trend seems to appear. HPMC partially hydrates forming a pseudo gel which control swelling of the implant, subsequently overall dissolution rate and drug availability. Once the protective gel layer is formed, two rate mechanisms predominate⁹. Firstly, the pseudo gel permits additional water to penetrate into the device, extending the gel layer into the implant. Secondly, the outer gel layer fully hydrates and begins to be dissolved by the fluids. For sparingly soluble drugs like indomethacin, the dissolution rate is primarily dependent on diffusion (due to swelling) and to some extent on erosion, which in turn is dependent on viscosity of the gel. Thus increasing the concentration of HPMC increases the viscosity of the resultant gel, thus resulting in slower drug release. Cross-linking of sodium alginate involves the interaction between cations and ?G-residues? in sodium alginate, resulting in the formation of an ?egg box? structure¹⁰. (fig. 3), indicate that an increment of calcium chloride caused a significant decrease (P<0.05) in the amount of drug released from all the methods. With increasing concentration of the cross-linker (Ca ⁺⁺ ), more calcium ions will be available for cross-linking, which results in the formation of a dense calcium alginate matrix that can bring about a decrease in drug release.

Sodium alginate, swells on contact with aqueous dissolution medium resulting in an increase in the porosity of the matrix, thus facilitating the mobilization of the water molecules into the polymer matrix. The dissolution medium dissolves the calcium chloride incorporated in the sodium alginate implants, resulting in an in situ cross-linking of sodium alginate, which is an instantaneous phenomenon. In agar diffusion and continuous flow through methods, the amount of aqueous medium that contacted the implant was significantly less and the time of contact between the implant and the dissolution medium was comparatively higher than the other two methods, thus the cross-linking of the sodium alginate matrix in these two methods would be progressive and more uniform in comparison to the static method, wherein the presence of relatively more dissolution medium would have resulted in rapid, but incomplete cross-linking. In USP dissolution method the implant disintegrated upon swelling which resulted in formation of isolated cross-linked mass.

An increase in the concentration of agar from 1 to 2% resulted in a significant decrease (P<0.05) in the amount of drug released which may be a consequence of increase in the strength of the gel formed. The increment in gel strength decreases the amount of available aqueous medium, as the water molecules would be entrapped within the gel structure, resulting in decreased rate of diffusion of the drug from the implants to the surrounding medium. It has been reported earlier that the drug diffusion coefficients from HPMC and sodium alginate matrices are strongly dependent on the water content of the system¹¹.

Amongst the methods studied, the agar diffusion method and the continuous flow through method seem to have the potential to prolong the drug release from the compressed implant and also simulate to certain extent the in vivo conditions as far as the placement of the device is concerned. The static method, which was successfully utilized for evaluating drug release kinetics from film type implants^{4, 8} was not suitable for evaluating drug release from the compressed implant, since the mesh holder, in which the implant was placed acted as an impediment to the uniform swelling of the implant, resulting in irregular swelling during the course of the study.