While laboratories utilize many approaches to calibrate analytical assays, they always involve the analysis of one or more calibrators that have known levels of the analyte of interest. The response of the analyte of interest in test samples is directly or indirectly compared to the response of the same analyte in the calibrators to determine the concentration of the analyte in the test sample. Analyte calibration is burdened with limitations and potential bias: Calibrator preparation inherently contains a level of inaccuracy, which propagates to the quantitation of any unknown based on that calibration. Even small biases in the calibration will impact the quantitative accuracy of samples.
External calibration is dependent on the requirement that the analyte within the calibrator behaves and responds in a manner identical to the analyte in the test sample. Internal standards are typically employed to help correct and normalize for these differences. However, even with isotopic internal standards, the normalized extraction efficiency and instrument response may not be comparable due to differences between the matrix composition of the calibrators and the test samples. Additionally, calibrator analysis adds to assay overhead.
Figure 1 – a) Polyarc catalytic microreactor. b) Polyarc catalytic microreactor
in housing installed on an Agilent 7890A gas chromatograph.The Polyarc catalytic post-column reactor (Activated Research Company, Eden Prairie, MN) eliminates the need for external calibration curves. Gas chromatography column effluent is directed through the catalytic microreactor (Figure 1a and b) in line with a flame ionization detector (FID). In the presence of zero air and hydrogen, the reactor converts all carbon atoms in a molecule to methane, which is then detected by the FID. In comparison to a traditional methanizer, which only converts carbon dioxide and carbon monoxide to methane, the Polyarc performs a two-step reaction in which all carbon-containing compounds are first oxidized into carbon dioxide and subsequently reduced to form methane. Following the reaction in the catalytic reactor, the FID sees only methane; thus the response of the FID is proportional to the concentration of methane. The concentration of methane is in turn proportional to both the concentration of the analyte and the number of carbon atoms contained within a molecule of the analyte. Assuming the carbon content of the analyte is known, the concentration of the analyte can be calculated by comparing the FID response of the peak of interest to that of an internal reference standard of known carbon content fortified into the specific sample at a known concentration. The Polyarc allows quantitation without the need for external calibrators, and there is no variability resulting from external calibration.
Applications in drug monitoring and toxicology
Relatively complex and costly assay calibration procedures are used in drug monitoring and toxicology. Multipoint calibrators are commonly fortified into the specific sample type being analyzed. If multiple biological sample types such as blood, urine, saliva, plasma, or serum are being analyzed, calibration curves are typically prepared in each matrix type. Calibrators are then subjected to the same processing procedures as the test samples. Internal standards are used to normalize for extraction differences and variations in instrument performance between the test samples and calibrators.
To determine the feasibility of quantifying toxicology samples without the use of a calibration curve, a Polyarc catalytic microreactor was incorporated into the flow of a direct-inject capillary GC-FID method for the analysis of ethanol and other volatile alcohols. The catalytic microreactor was installed on an Agilent 7890A GC equipped with an FID and an Agilent 5973 mass selective detector with an electron impact ionization source (Agilent Technologies, Santa Clara, CA). Zero air and H2 were plumbed into the heated catalytic reactor in line with the FID. Next, the GC system was configured to mimic a previously validated direct-inject capillary GC-FID procedure for the analysis of ethanol in blood, plasma, and urine. Chromatographic analysis was performed using an Agilent DB-5ms UI column (30 m × 0.25 mm × 1 µm) isothermally held at 35 °C for 5 minutes prior to a brief bakeout, and helium was used as both the carrier gas and FID makeup gas. The column effluent was split at a ratio of 1:10 between a 5973A mass spectrometer and the catalytic microreactor feeding the FID. Sample processing consisted of a simple dilution procedure; 50 µL of each sample was diluted into 1 mL of a solution containing 157.3 mg of n-propanol per liter of distilled water. Samples were mixed and then vialed prior to injection. The fortified n-propanol in each sample served as the sole calibration reference and as the internal standard normalizing for any sample and/or instrumental variability encountered during the analysis of each individual sample. No additional sample processing was required. Finally, 0.5 µL of the processed samples was injected into the GC configured with a Jennings cup injector liner.
Solutions and samples containing known concentrations of ethanol were used to evaluate accuracy of the method in a variety of commonly encountered sample types. Water-based samples were obtained from Cerilliant Corp. (Round Rock, TX); whole blood, serum, and urine controls were from UTAK Laboratories, Inc. (Valencia, CA). Materials used had ethanol concentrations ranging from 50 to 300 mg/dL (0.05% to 0.3% ethanol). The water and serum solutions also contained acetone, 2-propanol, and methanol. Fresh ethanol-free urine samples from 10 volunteers were analyzed as blanks and following fortification to contain 100 mg of ethanol per dL of urine. Each sample was processed, analyzed, and quantified exclusively using n-propanol fortified into each sample as an internal calibrator following in-line catalytic conversion to methane. The following equation was used to quantify ethanol in the test samples:
Where:
CA = concentration of the analyte
CC = concentration of the internal calibrator (n-propanol)
AreaA = integrated peak area of the analyte
AreaC = integrated peak area of the internal calibrator (n-propanol)
MwA = molecular weight of the analyte
MwC = molecular weight of the internal calibrator (n-propanol)
#CC = number of carbon atoms per molecule of internal calibrator (3 for n-propanol)
#CA = number of carbon atoms per molecule of analyte (2 for ethanol)
vC = volume of internal calibration solution
vA = volume of tested sample aliquot
Chromatograms generated using the Polyarc coupled to an FID were comparable in sensitivity and resolution to those generated in the mass spectrometer (Figure 2a and b). Ethanol levels quantified against the internal calibrator n-propanol were in agreement with the theoretic ethanol levels of each sample (Figure 3), generating an overall slope of comparison of 1.0062 with a correlation coefficient of 0.9949. Using the catalytic microreactor, solutions containing ethanol in blood, urine, and serum quantified on average to within 4.5% of their theoretical concentrations. Ten fortified and quantified urine samples quantified to within 7.0% of the theoretical concentrations. The sample matrix type had no impact on the accuracy of the results.
Figure 2 – a) Chromatogram of ethanol and other volatiles in serum generated using the Polyarc catalytic microreactor coupled to an FID. b) Total ion chromatogram of the same serum sample analyzed via electron ionization mass spectrometry.
Figure 3 – Parity plot of measured ethanol levels with theoretical concentrations.Conclusion
Incorporation of a catalytic microreactor in series with an FID enabled accurate quantitation without the analytical overhead or potential inaccuracy associated with use of a calibration curve. It can be installed on a standard GC-FID running a capillary column. The reactor was shown to effective for samples such as directly injected blood, plasma, and urine. Ethanol analysis was described here, but the reactor is compatible with any GC-compatible analyte containing carbon atoms. One of the most intriguing potential applications is the quantitative analysis of novel drug substances or metabolites where reference materials are not available. For these uses, the ability to split the column effluent between a mass spectrometer and the catalytic microreactor would provide information-rich quantitative analysis, which has not been previously attainable. Even in routine bioanalytical applications, use of the reactor can simplify quantitative procedures and increase the sample capacity of an instrument by reducing the need for external calibrators. By shifting the basis of quantitation to an internal calibrator fortified within each sample, multiple quantitative variables and assay variables are eliminated.
Gregory C. Janis is an R&D director for Laboratory Corporation of America and scientific director of MedTox Laboratories, 402 West County Rd. D, Saint Paul, MN 55112, U.S.A.; tel.: 651-628-6190; e-mail: [email protected]. Charles S. Spanjers is a product innovation engineer for Activated Research Company, Eden Prairie, MN; www.activatedresearch.com.