Novel Approaches to the Quantitative Determination of Drugs
Jack Henion, Jerry Zweigenbaum, Katja Heinig, and Timothy Wachs, Analytical Toxicology, New
York State College of Veterinary Medicine, Cornell University, 927 Warren Dr., Ithaca, NY 14850
There continues to be significant pressures placed upon analysts within the pharmaceutical
industry to provide even shorter response times for needed analytical information. Modern
requirements for a healthy “pipeline” of new drugs places demands upon drug discovery,
development and all other aspects of the pharmaceutical industry that were believed to be
unrealistic just five years ago. Modern LC/MS and LC/MS/MS techniques now play a major role in
our ability to speed up the analytical aspects of pharmaceutical discovery and development. One
key to improved analytical capabilities is our ability to creatively use these new techniques. This is
important for early PK and toxicology studies, Caco 2 absorption studies, in vitro incubation
studies, N-in-one dosing experiments, and a host of studies from clinical trials in early and late
drug development. In addition, there are other areas for high-volume analyses such as those
resulting from combinatorial chemical library syntheses.
The focus of this paper is to explore and demonstrate new strategies for the use of LC/MS/MS in
high sample throughput situations. These include pushing the limits of sample preparation,
LC/MS/MS analytical run times, and data management. Results will be shown which demonstrate
the sample preparation and LC/MS/MS analysis of over 2000 biological samples within 24 hours
as well as an example of the analysis of 240 samples in one hour which could lead to the eventual
capability to analyze 5760 samples within 24 hours.
Sample preparation was done via liquid-liquid extraction in 1.1 mL deep-well 96-well format plates.
This parallel sample preparation strategy allows for the preparation of many more samples than
is customary by conventional approaches. A Tomtec robot was used for aliquoting and
dispensing samples and solvents while the analyst manually transported the plates from the
corresponding workstations which included the Tomtec robot, a hematology mixture, centrifuge,
and nitrogen blow-down device. The reconstituted extracts contained in the 96-well plate format
were placed directly on a Shimadzu autosampler for direct analysis by LC/MS/MS techniques
using selected reaction monitoring (SRM). A total of four similar autosamplers was used to
provide injections every 30 sec. A control box constructed in our laboratory provided
communication between the PE SCIEX API 3000 mass spectrometer system and the four
autosamplers. These autosamplers were multiplexed in sequence to provide 30 second
sequenced injections which are not presently possible from most commercial autosampler
systems. A single Shimadzu HPLC pump was used for this work. The targeted compounds were
selective estrogen receptor modulators (SREM’s) which included raloxifene, tamoxifen and its 4hydroxy metabolite, nafoxidine, and idoxifene. The internal standard for these studies was D5idoxifene. Standard curves included seven standards analyzed in duplicate at the beginning and
end of each plate which ranged from 10 ng/mL to 10,000 ng/mL in plasma with three quality
control (QC) samples analyzed in replicates of six randomized within each plate.
One analyst was able to prepare 21 96-well plates of biological sample extracts within one day
and analyze these samples (2016) within 24 hours on the next day. The analysis of results
showed that acceptable precision (10%) and accuracy (15%) for the targeted analytes was
achieved when the analyte had a corresponding stable isotope internal standard. The acceptance
criteria could be met for compounds eluting near the internal standard, but not for those with
retention times distant from the internal standard. The system performance was acceptable
throughout the entire 24-hour period. Thus chromatographic performance and system sensitivity
were maintained through the run while HPLC column pressures remained within acceptable
tolerances. These results significantly improve the current sample analysis throughput of
bioanalytical LC/MS/MS capabilities and suggest that even higher sample throughput capabilities
may be possible in the future.
We thank PE SCIEX for the generous loan of the API 3000 mass spectrometer system and
SmithKline Beecham for a generous supply of Idoxifine and its stable isotope internal standard.
JZ thanks the Eastman Kodak Company for support of his studies at Cornell.
Figure 1. SRM LC/MS Chromatogram of SERMs showing reproducibility and
ruggedness for analysis of over 2000 samples in 24 hours. Compounds are 1) raloxifene,
2) 4-hydroxytamoxifen, 3) nafoxidine, 4) tamoxifen, and 5) idoxifene.