Anal. Chem. 2003, 75, 5103-5115
Development and Evaluation of a Solid-Phase Microextraction Probe for in Vivo Pharmacokinetic Studies Heather L. Lord,† Russell P. Grant,‡,§ Markus Walles,† Bev Incledon,‡ Brian Fahie,| and Janusz B. Pawliszyn*,†
Department of Chemistry, University of Waterloo, Waterloo, ON N2L 3G1, Canada, Eli Lilly Canada Inc., 3650 Danforth Avenue, Scarborough, ON M1N 2E8, Canada, and Eli Lilly Corporate Center, Indianapolis, Indiana 46285
Simplified procedures and a reduction in sampling errors are important advantages for performing as many chemical analysis steps as possible at the site where a sample or subject is located. Solid-phase microextraction technology addresses this goal, and for our purposes, it also allows for in vivo monitoring of a dynamic living system with minimal disturbance of the system. Here we report the development of a solid-phase microextraction application for in vivo monitoring of circulating blood concentrations of benzodiazepines. A probe based on a polypyrrole extraction phase was developed and used for extraction of drug molecules directly from a peripheral vein with subsequent instrumental quantification. The probe provides good sensitivity and selectivity for the drugs versus the blood matrix, while eliminating the need to draw blood. After sampling, the extracted drugs are quantified by liquid chromatography-tandem mass spectrometry. The limit of detection of the method is ∼3-7 ng/mL for analysis of the benzodiazepines from whole blood, and the method is linear to 1000 ng/mL. The method was used to monitor the pharmacokinetic profiles of diazepam and its metabolites in dogs, and the results compared favorably with profiles determined by conventional methods. Because no blood and only very small amounts of drugs are removed, minimal disruption to the chemical balance in blood occurs. This approach offers the potential for reduced exposure to blood for analytical personnel, simplified, less disruptive sampling, and lower stress levels on animals for pharmacokinetic studies. It also allows for a significant reduction in animal usage for these studies, which is important both ethically and for improving data quality. The value of solid-phase microextraction (SPME) technology has been realized over the past decade in its inherent ability to simplify chemical analysis by combining sampling, sample preparation, and preconcentration to the extraction phase into one step with convenient introduction of extracted components into an * Corresponding author. E-mail:
[email protected]. † University of Waterloo. ‡ Eli Lilly Canada Inc. § Current address: Esoterix, 4301 Lost Hills Rd., Calabassas Hills, CA 91301. | Eli Lilly Corporate Center. 10.1021/ac0343230 CCC: $25.00 Published on Web 09/04/2003
© 2003 American Chemical Society
analytical instrument.1 Additionally, because the stationary extraction phase performs preconcentration, the use of solvents in these steps is obviated. A third strength of the technology lies in the fact that the extraction phase comes into concentration equilibrium with the chemicals in the surrounding sample matrix, rather than exhaustively extracting them as is done in conventional extraction. Above a certain sample size, sample volume is irrelevant as it does not impact results; therefore, it is not necessary to define a specific sample size for the analysis. Additionally, SPME will directly extract only unbound analyte. Where sample size is selected so that negligible depletion of the free fraction is achieved, dynamic systems containing both bound and unbound analyte are undisturbed. Finally, the technology is easily miniaturized to allow it to be used with both small living systems and microanalytical instruments. These strengths point to SPME as a promising tool for directly assaying chemical concentrations in vivo. The on-site SPME sampling procedure described here is a significant departure from conventional “sampling” techniques, where a portion of the system under study is removed from its natural environment and the compounds of interest are extracted and analyzed in a laboratory environment. There are several motivations for exploring these types of configurations. Primarily there is the desire to study chemical processes in association with the normal biochemical milieu of a living system, and there is often an impracticality or lack of availability frequently associated with size, of removing suitable samples for study from the living system. In pharmacokinetic studies with rodents (mice in particular), the limited blood volume results in a large number of animals being used to generate profiles with sufficient numbers of data points. If blood were not removed for analysis, smaller numbers of animals would be required and the data generated would be improved by reduced interanimal variation. As discussed above, because compounds of interest are not exhaustively removed from the investigated system, conditions can be devised where only a small proportion of the total free compounds are removed (negligible depletion), thus also avoiding a disturbance of the normal balance of chemical components. If significant depletion of the free fraction occurs, the probe directly extracts free drug, and the depletion in free concentration results in release of some of the bound fraction. The cycle continues until a new (1) Pawliszyn, J. Solid-Phase Microextraction: Theory and Practice; Wiley: New York, 1997.
Analytical Chemistry, Vol. 75, No. 19, October 1, 2003 5103
binding equilibrium is established. Ideally one aims to use a system where negligible depletion of the free fraction is achieved so that sample size does not impact results and the chemical balance in the system is undisturbed. This could be beneficial in the nondisruptive analysis of very small tissue sites or samples. The nondisruptive nature of the technique can be understood both in terms of not impacting sample volume during sampling, and where negligible depletion is achieved, of not impacting the balance of bound versus unbound drug in the matrix. With the current commercially available SPME devices, a stationary extraction polymer is coated onto a fused-silica fiber. The coated portion of the fiber is typically 1 cm long, and coatings have various thicknesses. The fiber is mounted into a stainless steel support tube and housed in a syringelike device for ease of use. Extractions are performed by exposing the extraction polymer to a sample for a predetermined time to allow sample components to come into equilibrium with the extraction phase. After extraction, the fiber is removed to an analytical instrument (typically a gas or liquid chromatograph) where extracted components are desorbed and analyzed. The amount of a component extracted is proportional to its concentration in the sample. To date, commercial devices have been used in some applications of direct in vivo analysis of living systems. For example, they have been applied for the in vivo analysis of pheromones and semiochemicals used in chemical communications by insects2,3 and frogs,4 respectively. In these cases, the living animals were noninvasively monitored over time, providing a convenient means to study complicated dynamic processes without interference. The current commercial devices do, however, have some limitations for in vivo analysis. First, the application to chemical analysis inside animals requires greater robustness in both the extraction phase and the supporting fiber core. In addition, most of the extraction phases currently available are better suited for more volatile and less polar compounds. Only one phase is particularly suitable for liquid chromatography (LC) applications (Carbowax/templated resin). Analytes of interest that typically circulate in living systems are less volatile and more polar and require LC analysis, so new or modified extraction phases are indicated. The overall dimension of the current device is typically too large for direct in vivo analysis and for direct interfacing to microanalytical systems, and the time required for the LC extraction phase to come into equilibrium with chemicals in a sample is relatively long (typically 1 h or more in a well-stirred sample). We have endeavored to address these limitations in our design of a new SPME probe for in vivo analysis of drug concentrations in the circulating blood of a living animal. To address the question of an extraction phase better suited to the application, we selected the general adsorbent polypyrrole (PPY). Extensive use in biosensor technology and our experience with it as an adsorbent for more polar compounds suggested its potential for in vivo extraction of drugs.5-7 Because polypyrrole can be prepared on a metal surface by electrodeposition from (2) Moneti, G.; Dani, F. R.; Pieraccini, G. T. S. Rapid Commun. Mass Spectrom. 1997, 11, 857-862. (3) Frerot, B.; Malosse, C.; Cain, A. H. J. High Resolut. Chromatogr. 1997, 20, 340-342. (4) Smith, B. P.; Zini, C. A.; Pawliszyn, J.; Tyler, M. J.; Hayasaka, Y.; Williams, B.; Caramao, E. B. Chem. Ecol. 2000, 17, 215-225. (5) Wu, J.; Pawliszyn, J. Anal. Chem. 2001, 73, 55-63. (6) Wu, J.; Pawliszyn, J. J. Chromatogr. 2001, 909, 37-52.
5104
Analytical Chemistry, Vol. 75, No. 19, October 1, 2003
aqueous solution, we selected a fine (127 µm) stainless steel wire as the supporting fiber. Polymer-coated wires can be autoclaved for sterilization and have sufficient robustness for exposure to blood flowing inside a vessel. The polymer is deposited as a thin (99.9%). There may be cases where a more sensitive detection or an extraction phase with higher affinity is required to perform the analysis in a system with a very low free concentration of drug. For the analysis described herein, however, the PPY probe described and the API 3000 MS/MS provided sufficient sensitivity for the analysis of diazepam and its metabolites in a pharmacokinetic study. EXPERIMENTAL SECTION The experimental work consisted of two phases. First, the SPME probe was prepared and evaluated for extraction characteristics. Second, a pharmacokinetic study was conducted in dogs, with drug concentrations in blood evaluated by both SPME probe and conventional blood draws for cross-validation. Preparation of SPME Probes. Stainless steel wires (grade T-304V, 0.005 in.) were from Small Parts Inc. (Miami Lakes FL). Lithium perchlorate (95%) and pyrrole (98%) were from Sigma/ Aldrich (Mississauga, ON, Canada). The quality of the pyrrole had a significant impact on the nature of the polymer. It was used as received for one month after opening and was stored refrigerated; the bottle was layered with nitrogen after each use. Distillation of the pyrrole prior to use resulted in a much less even polymer. PPY films were deposited onto the supporting electrode surface (stainless steel wire) by anodic oxidation of the pyrrole monomer in the presence of an aqueous electrolyte solution (counterion). A potentiostat/galvanostat (model 273, EG&G Princeton Applied Research, Oak Ridge, TN) controlled with version 4.1 software was used for the electrodeposition. The last 15 mm of the wires were coated potentiostatically at 0.8 V for 20 min. The placement of a silicon septum 15 mm from the end of the wire allowed for accurate control of coating length. The coating solution used was pyrrole (0.1 M) and lithium perchlorate (0.1 M) in ultrapure water and was prepared fresh daily. Coating was performed in a custom-designed 50-mL flow-through glass compartment. Coating solution was pumped through the compartment continuously to allow for one complete change of solution during each deposition (50 mL/20 min). The stainless steel wires were cut into 10.7-cm sections with a razor blade, and 2-4 cm at the end to be coated was etched with 400-grit silicon carbide polishing paper. Polishing with finer grit papers or chemical etching with acid resulted in polymers that were less well adhered. Wires were then sonicated in acetone until required to prevent accumulation of oxides or other contaminants on the wire surface. Immediately before use, the wires were rinsed briefly with water and were installed as the working electrode. The counter electrode consisted of a ∼20-cm section of platinum wire (0.65-mm o.d.) (7) Wu, J.; Lord, H.; Kataoka, H.; Pawliszyn, J. J. Microcolumn Sep. 2000, 255266.
formed into a coil of ∼1.5-cm diameter. The stainless steel wire was placed into the coating solution in the center of this coil. A calomel reference electrode was used. The polypyrrole coating thickness was somewhat variable (∼10%) but in all cases appeared to be
