Collection, storage, and filtration of in vivo study samples using 96-well filter plates to facilitate automated sample preparation and LC/MS/MS analysis.

Abstract

The benefits of high-throughput bioanalysis within the pharmaceutical industry are well established. One of the most significant bottlenecks in bioanalysis is transferring in vivo-generated study samples from their collection tubes during sample preparation and extraction. In most cases, the plasma samples must be stored frozen prior to analysis, and the freeze/thaw (F/T) process introduces thrombin clots that are capable of plugging pipets and automated liquid-transfer systems. A new approach to dealing with this problem involves the use of Ansys Captiva 96-well 20-microm polypropylene filter plates to collect, store frozen, and filter plasma samples prior to bioanalysis. The samples are collected from the test subjects, and the corresponding plasma samples are placed directly into the wells of the filter plate. Two Duoseal (patent pending) covers are used to seal the top and bottom of the plate, and the plate is stored at down to -70 degrees C. Prior to sample analysis, the seals are removed and the plate is placed in a 96-well SPE manifold. As the plasma thaws, it passes (by gravity or mild vacuum) through the polypropylene filter into a 96-well collection plate. A multichannel pipet or automated liquid-transfer system is used to transfer sample aliquots without fear of plugging. A significant advantage of this approach is that, unlike other methods, issues related to incomplete pipetting are virtually eliminated. The entire process is rapid since thawing and filtering take place simultaneously, and if a second F/T cycle is required for reanalysis, it is not necessary to refilter the samples (additional clotting was not observed after three F/T cycles). This technique was tested using monkey, rat, and dog plasma and sodium heparin and EDTA anticoagulants. To assess the possibility of nonspecific binding to the polypropylene filter, a variety of drug candidates from diverse drug classes were studied. Validation data generated for two Lilly compounds from distinct classes, before and after filtering, are presented in this paper as practical examples of this technique. While LC/MS/MS is the primary method of bioanalysis in our laboratory, the technique presented in this paper is applicable to other forms of detection as well.