The entire powder layering process was conducted as follows: sugar spheres, used as inert seeds, were poured into the pan, then intermittently treated with (1) a nebulized binder solution applied using spray guns until the bed was wet and tacky and (2) a finely dispersed drug powder until the bed was dry.
This process led to the formation of multiple layers of drug particles that adhere to one another due to capillary pressure and interfacial forces originating from the liquid phase, allowing the enlargement of the initial cores. It should be noted that, in principle, the process of powder layering can be continued until reaching the desired particle size. Intraparticellar solid bridges were formed after each wetting-powder cycle by the complete removal of water by a stream of warm air blown through the perforated sword system present in the GS coating equipment. The resulting final pellets can further be filmed by different polymers in order to obtain multiparticulate dosage forms with enteric or modified release properties.
Preformulatory Study
Sugar spheres consisting of sugar and starch, with a mean diameter of 600 μm, were chosen as inert seeds in order to obtain final pellets having dimensions compatible with the filling of hard gelatin capsules of intermediate size such as size n. 1. In order to maximize the interactions between drug and inert cores, a micronized ibuprofen powder with a mean diameter (by number) of 6 μm was chosen, resulting in a size-ratio of 1:100 between the drug particles and the inert cores.
The wet-ability of the drug powder was also considered. In fact, it is well known that successful interaction between the drug and the binder solution is greatly influenced by the wet-ability of the drug (as measured by the contact angle, which should be kept as low as possible). For instance, ibuprofen, being a hydrophobic compound, is characterized by an unfavorable wet-ability expressed by a contact angle of 70°. In order to reduce this value, a surface agent (sodium lauryl sulfate, SLS) was included in the formulation to aid the wetting of the drug. The results clearly indicate that a formulation including 0.75% (wt/wt) SLS was able to sharply reduce the contact angle to 0°, representing complete wetting of a solid surface.
The micronized ibuprofen powder was also characterized by a scarce flowability. To overcome this problem, colloidal silicone dioxide (CSD) was employed as a flow activator. CSD was found effective in improving the powder flow properties; in fact, the addition of 2% CSD resulted in a significant reduction of the Carr’s index from 52% to 29.2%. Drug and other excipients were mixed in a twin-shell mixer for 15 minutes. With respect to these values, it should be remembered that the Carr’s index is usually used to obtain information on the behavior of powders for tablet production, where the feeding of the die is accomplished by gravity and thus a very free-flowing powder is required. In the GS system, however, the powder is dosed by two synergic mechanical actions, the vibration applied to the powder feeding unit and the rotary movement of the helical conveyor. Under these conditions, powders with a Carr’s index up to 35%,40% also can be dosed very accurately, as proved by a preliminary set of experiments demonstrating that the powder dosing was always within 3% of the correct value.
Other important parameters to be considered in order to obtain optimal powder layering are the type and quantity of binder. The binder has to possess high adhesivity and an appropriate viscosity, to guarantee a good adhesion between sugar cores and drug particles, resulting in a high concentration of drug in the pellets. In the present study two different binders were assayed, namely hydroxypropyl cellulose (HPC) and polyvinylpyrrolidone (PVP K30). PVP is a water-soluble binder that allows a rapid dissolution of the final pellet. HPC, being a less water-soluble polymer, gave rise to slower dissolution rates.
Preliminary Experiments of Layered Pellet Production
On the basis of the data from the preformulatory study, an easily-wettable and sufficiently flowable powder was formulated, as well as a binding solution thatg was not too viscous.
In order to obtain pellets with optimal characteristics, a series of parameters were considered, such as (1) initial core load, (2) pan speed, (3) powder application rate, (4) type and position of the atomizers (spray guns), (5) atomization pressure, (6) air cap type, and (7) temperature of the bed.
Good drug layering yields were obtained by carefully adjusting the quantity of both the applied drug powder and the binder solution. Particularly, an excess of drug powder resulted in a high loss of drug through the exhaust system, powder caking on the pan walls, and formation of seedless drug agglomerates of various size. On the other hand, an excess of binder solution led to an over-wetted bed, causing the formation of sticky agglomerates between pellets and the wall of the pan.
Pan speed was also found to heavily influence the powder layering process. In fact, low pan rate rotation (such as 10 rpm) caused the agglomeration of the cores, while higher rotation rates such as 20 rpm allowed a good application of the powder on the core surface without aggregation problems.
Other operating parameters such as type of air cap, atomization pressure, and position of the spray guns were also considered. For instance, air cap was found to heavily influence both the diameter of the sprayed binder droplets and the spray angle. In the case of core pellets with sizes between 400 and 600 μm, optimal results were obtained using the size #4 air cap, which strongly reduced sticking and adhesion problems.
Finally, by modulating the temperature of the inlet air, it was possible to maintain a constant bed temperature during the application of both the binder solution and the powder, overcoming in this way the cooling effect due to evaporation of the binder solution solvent. When inlet air temperature was too high, an extreme drying process resulted, causing an elevated particle friction phenomenon that increased the percentage of product loss.
On the basis of the considerations above, the initial pelletization process was performed by preheating the cores to 34°C; afterwards, the powder containing the drug and the excipients were applied at a bed temperature between 34 and 35°C. At the end of the process, each batch was subjected to further drying for 5 minutes to remove the residual water. Initially, the binder of choice was HPC. The photomicrographs of the pellets prepared using HPC (batch #1) are reported. They show a quite uniform rough surface, but are not completely free of imperfections.
The resulting pellets were sticky to a certain degree, probably because of the low initial temperature of the cores and the high powder dosing rate. In an attempt to solve the problem, pellets were prepared with an increase in initial temperature of the cores from 35 to 38°C, but these pellets did not change substantially with respect to the previous samples.
As further improvement, the rate of powder application was reduced from 331 to 180 g/min to obtain a more gradual and homogeneous distribution of the powder. Notwithstanding this change, a marked stickiness of the pellets was again evident, together with a high degree of free drug particles in the pan. Nevertheless, the size and morphological characteristics of the isolated pellets were almost comparable with those of the previous preparations.
Finally, in order to promote the pellet separation into distinct units, 15% (wt/wt, with respect to the drug) talc was added to the drug powder as an anti-sticking agent (batch #4). Although talc is recognized as a hydrophobic excipient possibly resulting in reduced powder flow-ability and in delayed disintegration (not optimal for a rapid release formulation), it was maintained in the formulation to guarantee a high separation of the pellets. Obviously talc was used in association with SLS and CDS respectively representing efficient wetting and glidant agents that are able to guarantee sufficient flow of the powder. Unfortunately, the addition of talc did not completely solve the stickiness problem; in fact at the end of the process out of 100 units, 95 were single pellets and 5 were pellet agglomerations (5% of the pellets were part of pellet agglomerations).
As further modification of the formulation, the viscosity of the binder solution was decreased by reducing HPC in the binder solution from 5% to 3% (batch #5). The pellets produced indeed presented a strong reduction of the surface defects; in addition, an increase in recovery of up to 98% was obtained. The use of a more dilute binder solution and the presence of talc as an anti-sticking agent resulted in the separation of each pellet into individual units during the layering process.
In light of these results, the preparation of pellets was tentatively conducted employing an alternative, less viscous and sticky binder, an aqueous 7% (wt/wt) PVP K30 solution. The formulation of the first batch of pellets produced with PVP (batch #6) is reported. The scanning electron photomicrograph of both the surface and the section of the pellets clearly show that the particles present a relatively smooth surface and homogeneous morphological characteristics. The percentage of pellet recovery was satisfactory, namely 96.8%, but the obtained pellets contained some seedless drug particles. This problem was tentatively attributed to the too-low adhesion capacity of PVP (with respect to HPC), leading to the formation of small dispersed particles that induced the formation of seedless drug aggregates.
Pellet Preparation by Powder Layering Technique
At the end of the preliminary experiments, the formulation used for the preparation of batch #6 was tentatively chosen as standard for the successive layering cycles. The parameters and the technical data characterizing this powder layering process are reported. These operating conditions arose from the preparation of pellets containing 18.4% (wt/wt) of drug; however, because our final goal was to obtain pellets containing at least 600 mg of ibuprofen per gram of product, further analyses were performed, based on the application of successive layers of powder to the batch #6 pellets.
Three successive amounts of ibuprofen, consisting of 2 kg each, were applied to batch #6 pellets, using the powder layering technique described above. After each application step, the pellets were sieved to eliminate the small seedless aggregates and successively poured into the coating pan for the application of a further batch of drug. Three loads of ibuprofen were applied to batch #6, resulting in batch #6.1, batch #6.2 and batch #6.3. By the successive application of drug layers, the content of ibuprofen in the pellets gradually augmented up to 41.7% (wt/wt).
The scanning electron microscope analysis of batch #6.3 showed that the last applied powder layer possesses a higher porosity with respect to the previous layers. This phenomenon could take place for the following reasons: (1) low adhesion capacity of the binder solution; (2) unfavorable ratio between powder and binder; (3) inadequate drying time; and finally (4) possible interaction between water present in the binder solution and the underlying ibuprofen layer, which could lead to the reduction of the adhesion between drug particles.
In order to possibly solve this problem, the ratio between powder and binder, core temperature, and binder solution were modified. With respect to this latest parameter, the optimal composition of the binder for ibuprofen powder application was determined to be a mixture 1:1 (wt/wt) of aqueous 7% PVP and 3% HPC solutions, associating in this way the binding capacity of both the excipients. These experimental parameters were employed for the production of batch #7 pellets, to which (as previously described for batch #6) were successively applied three further loads of ibuprofen resulting in batch #7.1, #7.2, and #7.3. Using the parameters the final concentration of ibuprofen in the #7.3 pellets was 54.5%. As previously stated, these pellets can be conveniently metered into hard gelatin capsules, size number 00, resulting in a final 200 mg of active/capsule.
Enteric Coating
Pellets obtained following the operating conditions determined above, namely batch #7.3, were subjected to a film coating process using the acrylic polymer Eudragit L30D-55 in order to produce a gastro-resistant formulation. The final enteric coating has been also applied in a GS pan coating plant, by means of an air-spraying system and continuous drying. The drying air was flowed through immersed swords that force the drying air to flow across and within the core bed, ensuring a constant and effective heat exchange and a considerable reduction of time process. The increase of pellet weight after coating was 6%, while the pellet recovery was 96%. In order to test the enteric properties of the pellets, they were placed in a 0.1 N hydrochloride acidic solution at pH 1. In these conditions the pellets showed a disintegration time of more than 3 hours; conversely, when placed in a 0.05 M phosphate buffered solution at pH 6.8 (as suggested by USP XXIII) the pellets disintegrated within 20 minutes.
Finally, it has to be emphasized that during all the preparation steps, no sign of drug degradation was detectable. Moreover, the small percentage of residual moisture indicates that the system here described allows the production of highly stable formulations.
No comments:
Post a Comment