998 SHORT PAPERS Analyst, V o l , 105 Rapid Gas-Liquid Chromatographic Determination of Cotinine in Biological Fluids C. Feyerabend and M. A, H. Russell Poisons Unit, New Cross Hospital, London, SE14 5ER Addiction Research Unit, Institute of Psychiatry, Maudsley Hospital, London, SE5 8A F Keywords: Cotinine determination; gas chromatography; plasma; saliva; urine Radioactive tracer techniques have been described for the measurement of cotinine in bio- logical fluids,l but they are cumbersome and inappropriate when large numbers of subjects are being investigated. It is preferable to use a gas - liquid chromatographic technique, especially when this can be achieved by simple modification of a routine method for nicotine measurement. We report here a rapid extraction procedure that allows the sensitive deter- mination of cotinine in biological samples without interference or contamination.October, 1980 SHORT PAPERS Experimental 999 Apparatus detector and a Model 3380A integrator was used, with an external time delay relay.A Hewlett-Packard, Model 5730A, chromatograph fitted with an alkali flame-ionisation Reagents All reagents were of analytical-reagent grade. Acetone. Dichloromethane. Pheniramine maleate. Hoechst Pharmaceuticals, Hounslow, Middlesex. Sodium hydroxide solutiora, 5 M. Gas - Liquid Chromatography A glass column (2.5 m x 2 mm i d . ) packed with 10% (m/m) Apiezon L and 10% potassium hydroxide on 80-100 mesh Chromosorb W was used. The temperatures of the oven, detector and injection port were 230, 300 and 250 "C, respectively. The retention times for cotinine and pheniramine free base were 2.66 and 3.17 min, respectively.The time delay relay device was used to overcome difficulties in the integration of peak areas, as described in detail elsewhere.2 Procedure To 1.0 ml of sample in a 12.5-ml centrifuge tube were added 2.0 ml of 5 M sodium hydroxide solution, 100 pl of an aqueous solution of phenirarnine rnaleate (2.2 pg ml-l) as internal standard and 3.0 ml of dichloromethane. The solution was vortex mixed for 2 rnin and then centrifuged for 5 min. Any emulsions were removed by discarding the aqueous layer, vortex mixing the centrifuge tube for a few seconds and centrifuging for 2 min. The organic layer was transferred into a second tube and evaporated to dryness under a stream of nitrogen a t room temperature.Acetone (50 pl) was added and the tube vortexed for 1 min and centri- fuged for 1 min. A 3-pl volume of the acetone solution was injected on to the chromato- graphic column. Calibration A calibration graph was constructed by adding cotinine and the internal standard to blank solutions of the sample type to be analysed to give concentrations of 25,50, 100,200,400,800 and 1000 ng ml-I. Although the calibration graph was linear from 0 to 1000 ng ml-1 and passed through the origin, care was taken to ensure that all cotinine concentrates were stored away from the analytical laboratory. Blanks Many of the problems of positive blanks associated with nicotine analysis3 do not occur in the cotinine method as this compound is present in cigarette smoke in much lower concentra- tions than nicotine.Reproducibility efficient of variation over this range was 1.8% (n = 10). The gas flow-rates were helium (carrier gas) 60, air 50 and hydrogen 3 ml min-l. These solutions were then carried through the extraction procedure. As a result, the risk of contamination is subsequently reduced. The reproducibility over a range of concentrations is shown in Table I. The average co- TABLE I REPRODUCIBILITY OF RESULTS OF TEN DETERMINATIONS AT VARIOUS COTININE CONCENTRATIONS Cotinine added/ng ml-l . . 25 50 100 200 400 600 800 1000 Cotinine found (mean)/ng ml-l 25.0 50.1 100.1 201.6 399.2 598.8 800.4 1000.3 Standard deviation/ng ml-1 . . 0.4 1.0 1.4 3.9 8.6 11.6 13.1 19.81000 SHORT PAPERS Analyst, Vol.105 Recovery The absolute recovery of cotinine (90%) was determined by injecting a mixture in acetone of cotinine (2000 pg 1-l) and pheniramine maleate (2000 +g 1-l). The peak-area ratio was compared with that obtained by injecting an extract of an aqueous solution of cotinine (1 ml; 100 pg 1-l) taken through the extraction procedure. The internal standard was introduced by reconstituting the cotinine extract in 50 p1 of acetone containing pheniramine maleate (2000 pg 1-l). (This procedure resulted in a final extract 20 times more concentrated than the original solution.) Results and Discussion A typical chromatogram of an extract from human plasma is shown in Fig. 1. The repro- ducible lower limit of determination of cotinine was 1 ng ml -l. No interference was found from the following common drugs: atropine, amphetamine, amitl-iptyline, chlorpheniramine, diphenhydramine, diethylpropion, fenfluramine, imipramine, lignocaine, methylamphetamine, nortriptyline, procaine, phentermine, sodium cromoglycate, salbutaniol and terbutaline. - 0 2 4 Timehin Fig.1. Gas chromato- gram of extract from human plasma: 1, cotinine and 2, pheniramine maleate. This method €or the determination of cotinine can be used in coiijunction with a previously reported extraction technique for n i c ~ t i n e . ~ . ~ Nicotine is extracted into diethyl ether and the sample is then re-extracted with dichloromethane to retrieve coiinine. Unlike Hengen and Her~gen,~ who reported no loss of cotinine during an extraction with diethyl ether, we found that a loss of cotinine occurred that was inversely proportional to the ratio of the volume of sample to that of the organic phase.Thus, with sample volumes of 3, 2 and 1 ml, the cotinine losses were 10,22 and 44%, respectively. It is therefore important to standardise the volume of sample if necessary, by the addition of tap water, and t o use the equivalent volume of plasma, saliva or urine standards to construct the calibration graph. In a preliminary experiment, a subject who had abstained from smoking for 10 days smoked a cigarette over a short period (3 min) and blood samples were withdrawn a t frequent intervals via an indwelling venous cannule. Samples of saliva and urine were also collected throughout the course of the experiment (Fig. 2). An increase in plasma cotiriine concentration occurred only 2 min after discarding the cigarette and the concentration rose steeply during the first hour and reached a plateau a t 4 h that persisted throughout the experiment. As expected, plasma nictoine concentrations reached a maximum within 2 min after smoking had ceased, fell sharply during the next 10-15 min and then more slowly over the remaining 6.5 h to approach base-line levels.Although cotinine (pK, 4.56) is essentially unionised in blood a t pH 7.4, the free base is poorly soluble in lipids and therefore its rate of distribution intoOctober, 1980 SHORT PAPERS 1001 60 50 40 30 20 10 0 0 1 0 2 0 3 0 4 0 5 0 6 0 2 3 4 5 6 7 Time/ rn i n Tirneih Fig. 2. Graph showing A, plasma nicotine; B, plasma cotinine and C, saliva cotinine, after smoking one 1.3-mg nicotine cigarette. Smoking period = 3 min.tissues may be slow. This would partially explain the prolonged existence of the compound in blood. Another contributing factor to this is the low rate of renal excretion of cotinine relative to nicotine (Fig. 3). This may account for the high concentrations of this metabolite (about 800 ng ml-l) often found in the plasma of habitual smokers. 0 1 2 3 4 5 6 7 Time/h Fig. 3. Urinary excretion of A , nicotine and B, cotinine over 7 h after smoking a 1.3-mg nicotine cigarette. The pII of the urine a t the different times is also shown. Nicotine is concentrated in saliva to give levels approximately 10 times higher than those measured in plasma,6 whereas salivary and plasma cotinine concentrations were essentially the same. We thank A. E. Bryant for technical assistance and the Medical Research Council for financial support. References 1. 2. 3. 4. 6. 6. Armitage, A. K., Dollery, C. T., George, C. F., Houseman, T. H., Lewis, P. J., and Turner, D. M., Feyerabend, C., and Russell, M. A. H., J . Pharm. Pharmacol., 1979, 31, 73. Feyerabend, C., and Russell, M. A. H., J . Pharm. Pharmacol., 1980, 32, 178. Hengen, N., and Hengen, M., Clin. Chem., 1978, 24, 50. Yamamoto, I., Adv. Pest Control. Res., 1966, 6, 231. Russell, hl. A. H., and Feyerabend, C., Drug Metab. Rev., 1978, 8, 29. Brit. Med. J . , 1975, 4, 313. Received March 19th, 1980 Accepted April 24th, 1980