sample="quota" bates="2063595051" isource="pm" decade="1990" class="ui" date="19980900" VALIDATION OF AN ASSAY FOR THE DETERMINATION OF COTININE AND 3-HYRDOXYCOTININE IN HUMAN SALIVA USING AUTOMATED SOLID PHASE EXTRACTION AND LIQUID CHROMATOGRAPHY WITH TANDEM MASS-SPECTROMETRIC DETECTION Mark C Bentley*, Mohammed Abrar, Mark Kelk, Jeremy Cook, Keith Phillips Covance Laboratories Ltd., Otley Road, Harrogate, North Yorkshire, HG3 1PY, England DRAFT Tel. +44(0) 1423 500011 Fax +44(0) 1423 508745 Email mark.bentley@covance.com * Corresponding author Abstract The validation of a high performance liquid chromatographic method for the simultaneous determination of low level cotinine and 3-hydroxycotinine in human saliva is reported. Analytes and deuterated internal standards were extracted from saliva samples using automated solid phase extraction, the columns containing a hyper cross-linked styrene-divinylbenzene copolymer sorbent, and analysed by reversed phase liquid chromatography with tandem mass spectrometric detection (LC-MS-MS). Lower limits of quantitation of 0.05 and 0.10 ng/ml for cotinine and 3-hydroxycotinine respectively were achieved. Intra- and inter-batch precision and accuracy values fell within ±17% for all quality control samples, with the exception of quality control samples prepared at 0.30 ng/ml for 3-hydroxycotinine (inter-day precision 21.1%). Results from the analysis of saliva samples using this assay were consistent with subjects' self-reported environmental tobacco smoke (ETS) exposures, enhancing the applicability of cotinine as a biomarker for the assessment of low level ETS exposure. 1. Introduction Nicotine is the principal alkaloid in tobacco and is present as a major component of tobacco smoke. It is absorbed in measurable quantities by both active and passive smokers, the latter shown to inhale quantities of nicotine proportionally to the product of concentration, duration of exposure and respiration rate [ ]. However, the relatively short half-life for nicotine (t½ 1-2 hours) precludes its use as an accurate marker for environmental tobacco smoke (ETS) exposure, or passive smoking, since assessments of low level exposure over protracted time periods are often desired. Cotinine, a primary metabolite of nicotine formed after C-oxidation via the enzyme cytochrome P450, has a much longer half-life (t½ 18-20 hours) than nicotine resulting in higher and more stable plasma concentrations and is therefore considered a more appropriate biomarker for evaluating ETS exposure [ ]. Following exposure to nicotine, cotinine can be found in most body fluids and methods for its determination in blood (serum/plasma), urine and saliva are generally considered acceptable for estimating nicotine exposure [ ]. In a review by Etzel [ ] evaluating the relationship between saliva cotinine concentrations and ETS exposure, concentrations less than 10 ng/ml would usually result from ETS exposure without active smoking, although heavy passive exposure to tobacco smoke may produce levels in excess of this value. In recent times the specificity of cotinine as a biomarker has been questioned since dietary sources of nicotine have been identified (eg tomato, potato, cauliflower, tea) [ ], although the contribution of dietary nicotine to serum cotinine levels is estimated to be small in comparison to ETS exposure [ ]. In a recent publication by Pirkle et al. [ ], reporting the findings of the Third National Health and Nutrition Examination Survey (NHANES III) in the United States, the estimated geometric mean contribution of dietary intake to serum cotinine levels was less than 0.02 ng/ml. Also reported was a median serum cotinine level of 0.526 ng/ml for adults who had reported some degrees of ETS exposure either at home or at work. Hence, in order to adequately assess very low level exposure to ETS using cotinine as a biological marker, or indeed to determine the degree of any dietary contribution to cotinine levels, assay sensitivity sufficient to quantitate levels significantly lower than 0.5 ng/ml is required. To date, there are very few published methods available for the quantification of cotinine at very low level concentrations. Bernert et al. [ ] have recently reported the development and 'validation' of an assay for determining cotinine in human serum with a limit of detection of 0.05 ng/ml, the method utilising protein precipitation followed by liquid/liquid extraction prior to liquid chromatography with tandem mass-spectrometric detection. From the data presented, this method was estimated to have a lower limit of quantification (LLOQ) in the region of 0.17 ng/ml. The method described here utilises solid phase extraction followed by liquid chromatography with tandem mass-spectrometric detection and has been validated for the simultaneous determination of cotinine and 3-hydroxycotinine in human saliva with LLOQs of 0.05 and 0.10 ng/ml respectively. The automation of sample extraction, which is not readily applicable to liquid/liquid methodologies, reduces the opportunity for sample contamination and contributes significantly to the achievement of low level quantitiation. Also of critical importance for the achievement of these LLOQs was the use of water as a surrogate matrix for the preparation of calibration standards, since analyte free control human saliva was not available. Subsequent evaluation of the influence of any ion suppression or background noise differences between water and human saliva was performed. At present, there are no published methods available for the simultaneous quantification of both cotinine and 3-hydroxycotinine in any matrix at the LLOQs reported here. This validation was designed to fulfil the requirements outlined in the Washington consensus meeting (1990) reported by Shah et al. [ ] as well as incorporating current regulatory opinion. A combined working stock solution containing both cotinine and 3-hydroxycotinine at a concentration of 10 µg/ml was prepared by diluting 1 ml of each primary stock solution to 10 ml with methanol. Dilutions of this combined calibration working stock solution and subsequent diluted working stock solutions, using a minimum number of serial dilutions, were performed in order to provide working standard solutions containing cotinine and 3-hydroxycotinine at concentrations of 400, 360, 200, 40, 20, 8, 4, 2, 1.2 and 0.8 ng/ml in methanol. Working standard solutions were stored refrigerated (nominal 4ºC) in amber glass vessels for up to two months. Calibration standards were prepared fresh on each analysis occasion by the addition of 25 µl of each working standard solution to 1 ml of water giving a calibration range of 0.020 to 10.0 ng/ml. 2.3.2. Preparation of quality/control samples Separate dilutions of the primary stock solutions with methanol were performed in order to provide working stock solutions at concentrations of 1000, 100, 10 and 1 ng/ml for both cotinine and 3-hydroxycotinine. These solutions were stored refrigerated (nominal 4ºC) in amber glass vessels for up to two months. Aliquots of these working stock solutions were spiked into control human saliva (50 ml) to produce quality control samples containing both cotinine and 3-hydroxycotinine at nominal concentrations of 0.060, 0.150, 0.300, 4.00, 8.00 and 20.0 ng/ml, taking accounts of determined endogenerous levels of cotinine and 3-hydroxycotinine within the control saliva pool used. Quality control samples at concentrations below determined endogenous levels in the control matrix pool were prepared by dilution of the control matrix using water to achieve the required concentration of the most abundant analyte and subsequently spiked with the least abundant analyte to the required concentration. Aliquots (2.5 ml) were stored frozen (nominal -20ºC) in polypropylene tubes prior to analysis. 2.3.3. Preparation of internal standard solutions A working stock internal standard solution containing both cotinine-d3 (10 ng/ml) and 3-hydroxycotinine-d3 (20 ng/ml) was prepared by diluting 100 µl and 200 µl of the respective primary stock solutions to 1000 ml with methanol. This solution was stored refrigerated (nominal 4ºC) in an amber glass vessel for up to two months. 2.4 Sample extraction An aliquot of each saliva sample (1 ml) quality control sample or calibration standard (1 ml water containing 25 µl of an appropriate working standard solution) was transferred into a borosilicate glass culture tube (10x75 mm LIP (Equipment and Services) Ltd, Shipley, UK) and 100 µl of internal standard working solution (10 ng/ml cotinine-d3; 20 ng/ml 3-hydroxycotinine-d3) added. Buffer (1/15/M potassium dihydrogen orthophosphate-1/15 M disodium hydrogen orthophosphate 41:59 v/v/) was added (1 ml) and the tube capped, briefly vortex mixed and centrifuged at 1500 g for 5 min. Following centrifugation, the tube cap was removed and the tube was transferred to a Gilson ASPEC™ or ASPEC XL™ instrument (Anachem Ltd, Luton, UK) for automated solid phase extraction. The solid phase extraction column, Isolute ENV+ 100 mg/1 ml (International Sorbent Technology; Jones Chromatography Ltd, Mid Glamorgan, UK), was conditioned with methanol (1 ml) followed by water (1 ml) and the sample solution subsequently applied to the column under low positive pressure. Following sample application, the column was sequentially washed with water (1 ml) and water-methanol (70:30 v/v, 1 ml) and the analytes eluted into ASPEC™ collection tubes (12x45 mm borosilicate glass tubes; purchased from Anachem Ltd, Luton UK) with methanol (2.5 ml). The sample extract was then evaporated to dryness at 40ºC under a gentle stream of nitrogen using a Techne Dri-Block® SC-3 sample concentrator (BDH (Merck) Ltd, Lutterworth, UK) and dissolved in 200 µl of 12.5 mM ammonium formate-methanol-formic acid (80:20:0.5 v/v/v). The tube was briefly vortex mixed and centrifuged at 1500 g for 5 min. Following centrifugation, the sample extract was transferred into a tapered microvial and 100 µl was taken for injection onto the column. Inter-batch (between-run) precision and accuracy were similarly calculated by analysing QC samples (n=6; excluding LLOQ and ULOQ QCs) on nine separate occasions. Mean concentrations (± standard deviation) and the coefficient of variation for each analyte at each concentration are presented in Table 1. These data indicate that the method is precise and accurate for the determination of cotinine over the concentration range 0.050 to 10.0 ng/ml. For 3-hydroxycotinine, assay precision at the low QC concentration (21.1% at 0.300 ng/ml) fell marginally outside the generally accepted limit of 20%. However, this variability was considered to be acceptable since poorer assay sensitivity and generally lower endogenous concentrations for 3-hydroxycotinine make this analyte less important as a marker for ETS exposure than cotinine. 3.3. Analyte recoveries Due to the presence of endogenous analytes in control saliva, recoveries for cotinine and 3-hydroxycotinine were determined using the deuterated internal standards. Analyte peak heights determined from the extraction of control saliva fortified with cotinine-d3 and 3-hydroxycotinine-d3 at low, mid and high QC concentrations were compared with peak heights determined for control saliva extracts fortified with equivalent amounts of cotinine-d3 and 3-hydroxycotinine, representative of 100% recovery at these concentrations. Analyte recoveries expressed as a percent of analyte added are presented in Table 2. The detector response for deuterated analytes was found to be less sensitive than for cotinine and 3-hydroxycotinine and the poor reproducibility of peak heights and exaggerated recoveries calculated for low QC concentrations were attributed to this fact. 3.4. Analyte stability The stability of cotinine and 3-hydroxycotinine in human saliva to three additional freeze-thaw cycles (four in total) and after storage for 24 h under ambient conditions of temperature and lighting was instigated at low, mid and high QC sample concentrations. The stability of analytes in sample extracts stored refrigerated (nominal 4°C) for up to 48 h was also investigated at the same concentrations. These data are presented in Tables 3 and 4 and demonstrate that the analytes were stable under the storage conditions investigated. Deviations from theoretical of 19.0 and 24.7% were observed for 3-hydroxycotinine low QC samples (0.300 ng/ml) following storage for 24 h at room temperature and after 3 additional freeze-thaw cycles respectively. However, these deviations reflected an increase in analyte concentration and determined levels were within 15% of the reported concentrations for 'untreated' low QC samples, included for the assessment of inter-batch assay variability. As such, these deviations were considered to be due to inherent assay variability for 3-hydroxycotinine at low concentrations, rather than a reflection of any analyte instability. 3.5. Application The method as described was used to measure concentrations of cotinine and 3-hydroxycotinine saliva samples collected from subjects with a range of self-reported recent ETS exposure histories. Determined levels for subjects reporting no recent exposure, some recent exposure, living with a smoker, being an occasional smoker and being a smoker are reported in Table 5. Subjects reporting no recent ETS exposure had lower mean saliva cotinine concentrations (0.145 ng/ml) than those reporting some exposure (0.689 ng/ml) or living with a smoker (1.28 ng/ml). However, there was an overlap in the range of cotinine values reported for all of these ETS exposure groups. This was possibly due to nicotine intake derived from sources other than ETS (eg dietary intake) or, more likely, a reflection of subjects' abilities to accurately assess any recent ETS exposure. A less noticeable trend was apparent for 3-hydroxycotinine, with a far greater overlap in the ranges of determined values, which was attributed to the poorer assay sensitivity for this analyte and the fact that saliva concentrations were found to be in the region of 40% of the corresponding cotinine levels. The results from this preliminary investigation indicate that saliva cotinine levels, and to a lesser extent 3-hydroxycotinine levels, determined over these validated concentration ranges will have increased significance as a biomarker for ETS exposure. Acknowledgement The funding for this investigation was made available to Covance Laboratories Ltd. by the Center for Indoor Air Research (CIAR), Linthicum, MD, USA. DRAFT Page 12 of 13 References Figure Captions Fig. 1. Structures for cotinine and 3-hydroxycotinine. Asteriks denote the position of the deuterated label for internal standards. Fig. 2. Representative LC/MS/MS chromatograms from the analysis of cotinine in human saliva for (A) a water blank, (B) a calibration standard at the lower limit of quantitation; 0.05 ng/ml cotinine in water and (C) a sample of control human saliva. Fig. 3. Representative LC/MS/MS chromatograms from the analysis of 3-hydroxycotinine in human saliva for (A) a water blank, (B) a calibration standard at the lower limit of quantitation; 0.10 ng/ml 3-hydroxycotinine in water and (C) a sample of control human saliva.