A DoE study to understand partition rate in a biphasic system

Purpose In a biphasic system, a drug molecule may partition from an aqueous phase to an immiscible organic phase in a process that is thought to be similar to passive diffusion of drugs across biological membranes. A Design of Experiments (DoE) model was used to explore the effects of pH, organic solvent type, excipients and biorelevant media on the partitioning rate. Methods Minitab 17 was used to design a 62 ½ fractional factorial model with center points. Only individual and 2-level interactions were analyzed and the alias structure showed that these were only confounded by 4-level interactions and above. 38 experiments were required, which were randomized in their run order. The Sirius inForm was used to run single sector biphasic dissolution experiments on indomethacin and dipyridamole. Here, the media was automatically added to the experiment vessel and equilibrated at 37°C. The sample was prepared as a 10 mM stock solution using DMSO and then automatically added to the aqueous layer using a needle. A UV dip probe was used to collect UV absorbance data from the organic layer, which was converted to concentration and plotted against time. The area under the curve was integrated and entered into the Minitab DoE model for analysis. Results The rate of partition was positively correlated with lipophilicity for dipyridamole, but was negatively correlated for indomethacin. The negative correlation was ascribed to a precipitation event of indomethacin in the aqueous phase that occurred at low pH. The addition of biorelevant components and excipients was negatively correlated with partition rate, except for sodium tauroucholate which was positively correlated with ionized dipyridamole. The less polar organic solvent was more sensitive to changes in lipophilicity than the more polar solvent. Conclusion Clear trends were observed with five of the six factors tested here. By understanding how each factor can affect the partition rate, it may be possible to optimise the in vitro system for correlation with in vivo bioavailability data.