Identifying Pharmacodynamic Protein Markers of Centrally Active Drugs in Humans: A Pilot Study in a Novel Clinical Model Sandeep T. Patil,*,† Richard E. Higgs,† John E. Brandt,† Michael D. Knierman,† Valentina Gelfanova,† Jon P. Butler,† AnnCatherine M. Downing,† Jill Dorocke,† Robert A. Dean,† William Z. Potter,‡ David Michelson,‡ Alan X. Pan,§ Stanford S. Jhee,| and John E. Hale† Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285, Merck Research Laboratories, West Point, Pennsylvania 19486, Lilly-NUS Centre for Clinical Pharmacology Pte, Ltd., Singapore 117597, and California Clinical Trials, Glendale, California 91206 Received July 26, 2006
Recognizing specific protein changes in response to drug administration in humans has the potential for significant utility in clinical research. In spite of this, many methodological and practical questions related to assessing such changes are unanswered. We conducted a series of clinical studies to assess the feasibility of measuring changes in proteins related to drug administration using a massspectrometry proteomics technique capable of quantifying hundreds of proteins simultaneously in cerebrospinal fluid (CSF) and plasma. Initially, the normal variability of proteins in these compartments was characterized in 16 healthy volunteers over a 2-week period. Drug-associated changes were subsequently assessed in the plasma and CSF proteomes of 11 subjects given atomoxetine, which served as a selective, centrally active probe to test the model. Protein levels in the CSF and plasma were unchanged between visits in the normal variability study. In contrast, statistically significant changes were detected in the CSF protein pattern after drug treatment. These studies suggest that identification of changes in the CSF proteome associated with the administration of centrally active drugs is feasible, and may be of value in the development of new drugs, as well as broader clinical research. Keywords: Proteomics • CSF • healthy volunteers • pharmacoproteomics • biomarkers • drug development • CNS • mass spectrometry • norepinephrine • DHPG
Introduction The importance of biomarkers that can help guide new drug development as well as disease management is widely accepted.1-6 A proven drug-activity biomarker can guide the dosing strategy in a clinical trial, confirm central penetration in the case of brain-specific neuropsychiatric compounds, provide new drug targets, and aid in patient selection that is consistent with a trial’s objectives. Such an approach can significantly reduce the time and cost of clinical research while concomitantly increasing patient safety and reducing the risk associated with the development of new therapies.6-8 Unfortunately, for many compoundssparticularly those aimed at pharmacologically novel targets in the central nervous system (CNS)s the available options to assess drug activity are limited, and an important focus of CNS drug research involves the de* To whom correspondence should be addressed. Sandeep Patil Ph.D., M.D., Lilly Corporate Center, Drop Code 1730, Indianapolis, IN 46285. Phone: (317) 433-8201. Fax: (317) 277-2123. E-mail:
[email protected]. † Lilly Research Laboratories, Lilly Corporate Center. ‡ Merck Research Laboratories. § Lilly-NUS Centre for Clinical Pharmacology Pte Ltd. | California Clinical Trials. 10.1021/pr0603710 CCC: $37.00
2007 American Chemical Society
velopment of techniques that can provide better information about central drug activity early in the clinical development process. Protein drug-activity biomarkers can be defined as specific protein changes in biological compartments that occur in response to drug administration. Mass spectrometry (MS)based proteomics offers a nontargeted way to identify protein or peptide biomarkers, making the technique particularly suitable for discovering drug-activity markers, since there is often no a priori knowledge of the particular proteins that are likely to change9 in response to drug administration.5,10 Mass spectrometry-based approaches can detect hundreds of proteins simultaneously, but a careful approach to clinical design and data analysis is important to maximize the chance of discovering meaningful drug-activity markers. In particular, the biological and temporal variability of the large set of proteins that can be simultaneously detected in human cerebrospinal fluid (CSF) and plasma is not well-characterized. Understanding proteomic variability under controlled conditions is critical to enable future clinical studies to be appropriately designed and to improve the contextual interpretation of data obtained from pharmacoproteomic studies.2 Journal of Proteome Research 2007, 6, 955-966
955
Published on Web 01/23/2007
research articles
Patil et al.
Figure 1. Base peak ion chromatograms for the first, middle, and last injections of the atomoxetine study plasma samples (a, c, and e) and their subsequent water blanks (b, d, and f). Total area under the base peak chromatogram for the water blanks was
