Search

The role of principal component analysis in the selection of pharmaceutical formulations is presented. The objective and the procedure of the analysis are discussed in detail. The technique was successfully applied to a system consisting of 10 response variables (tablet properties). Analysis of the results showed that the first component (dissolution) and components one and two together (dissolution and disintegration) contributed 95.4 and 99.3%, respectively, to the overall information about the formulations and that eight of 10 response parameters contributed nothing further to the overall information. The results obtained from this method of analysis may be found useful for achieving economy in both cost and time of measuring responses. Principal component analysis also provides a basis for understanding the underlying mechanism of the system under consideration.

This investigation was designed to determine the in vitrorelease of indomethacin from suppository bases and the in vivobioavailability in rabbits. Suppositories containing 25 mg of indomethacin were made by the fusion method with theobroma oil, esterified fatty acids (C10-C18), and polyethylene glycol 1000. To produce an exact dosage form, a formula for the determination of the displacement value was derived, and it was found that theobroma oil > esterified fatty acids (C10-C18) > polyethylene glycol 1000. The suppository hardness was determined by using appropriate apparatus and it was found that the esterified fatty acids (C10-C18) allowed the formation of more brittle suppositories. The release rates were determined with the USP dissolution apparatus, with or without cellophane membrane, and it was found that polyethylene glycol 1000 > esterified fatty acids (C10-C18) > theobroma oil. The bioavailability of indomethacin after rectal administration was greater with polyethylene glycol base. Significant correlation was obtained during the first 45 min between the in vitrorelease (dialyzing tubing) and the in vivobioavailability.

Two motorized and two manually powered instruments were used to measure the breaking strength of tablets of three different shapes, each of two sizes and each at three different target breaking strength values. The experimental results were statistically analyzed. The uniformity among breaking strength testers was compared and these results are presented. A high correlation among devices was observed, allowing for the derivation of regression equations which could be used to predict precisely the breaking strengths from one instrument to another. Manually operated instruments tended to yield lower and more variable values than those of their electrically powered counterparts. Tablet shape and size were also found to affect the uniformity and consistency of tablet breaking strengths.