Clinical Chemistry, Vol 37, 1276-1280, Copyright © 1991 by American Association for Clinical Chemistry

Liquid-chromatographic determination of nicotine and cotinine in urine from passive smokers: comparison with gas chromatography with a nitrogen-specific detector

M Hariharan and T VanNoord
University of Michigan Medical School, Department of Psychiatry, Ann Arbor 48109-0656.

We describe a simple, sensitive, and specific high-performance liquid- chromatographic method with ultraviolet detection (256 nm) for the simultaneous analysis of nicotine and cotinine in urine of passive smokers. The analytes are extracted and purified from the complex and impure matrix in two stages; first, by liquid-liquid extraction and followed by solid-phase extraction (C2 column). We used a "DB" C8 5- microns-particle column (25 x 0.46 cm) and a mobile phase of phosphate- citrate buffer and acetonitrile (91:9 by vol) containing 5 mL of triethylamine and 600 mg of heptanesulfonate per liter, adjusted to pH 4.4, to separate the compounds. Two internal standards (2- phenylimidazole and N-ethylnorcotinine) were used. The detection limit of the HPLC assay was less than 1 microgram/L for both analytes. The average interassay CV for nicotine was 7.6%, for cotinine 6.5%, in the concentration range 0-60 micrograms/L. The mean analytical recovery of nicotine with respect to the internal standard N-ethylnorcotinine was 102% and that for cotinine was 99%; the mean absolute recoveries of the compounds were as follows: nicotine 85%, cotinine 87%, and N- ethylnorcotinine 87%. To compare the HPLC assay results of 20 samples with the more-sensitive in-laboratory gas-chromatographic method with a nitrogen-phosphorus detector, we used both N-ethylnornicotine and N- ethylnorcotinine as internal standards. The mean correlation coefficient for nicotine values between the two methods was 0.934; for cotinine, 0.987.