According to the recent regulations (Circulaire DGS/DH du 3 avril 2000), tobacco dependence must be determined by the measurement of urine nicotine metabolites. Various assay methods are presently available. They were tested in order to evaluate their analytical performances and to determine how they can be used for the clinical management of smoking cessation.
Material and methods: Urine samples from a single void (n = 97) were obtained from active and abstinent smokers (with or without nicotine substitutive therapy). They were all analyzed by the various methods. Cotinine concentration was measured in six laboratories, using HPLC combined with UV detection according to a standardized procedure (Ann Biol Clin 2002 : 60 : 263-72). Immunoassay methods were also tested and the values obtained from urine samples were compared to urine cotinine measured by HPLC-UV.
Results: HPLC-UV: Urinary cotinine varied in a range from undetectable to 4 mg/L. An interlaboratory comparison was performed according to the Valtec procedure (calculation of equation of Deming, chart of differences). There was a good accordance between laboratories. Cotinine concentration was only slightly influenced by fluid intake, as shown by a poorly significant correlation between cotinine and creatinine (r = 0.23, p = 0.05). Homogeneous immunoassays: The two homogeneous immunoassays (Cotinine) from Thermo Electron and Cotinine Enzyme Immunoassay commercialized by Microgenics were highly correlated (r = 0.97). The correlation was not so strong with HPLC-UV (r = 0.86). Firstly, values were found higher with immunoassays because antibodies crossreact with 3-hydroxycotinine. Secondly, the ratio of immunoassays values to HPLC-UV values varied according to urine specimens. Finally, there was a highly significant correlation with urine creatinine (r = 0.40, p = 0.0001), thus indicating the influence of fluid intake. Heterogeneous immunoassay: The kit Metabolites of Nicotine commercialized by DPC France was tested on the analyzer Immulite, using a procedure specifically established for urine. Antibodies revealed a large spectrum of nicotine metabolites. Therefore, the values were much higher than those observed for the same urine samples with homogeneous immunoassays.
Conclusion: HPLC-UV can be recommended for the measurement of urinary cotinine, as it was shown a good accordance between laboratories. The low detection limit is of interest for the diagnosis of Environmental Tobacco Smoking. Homogeneous immunoassays can be easily used for routine analysis as they can be performed directly on urine specimen. The results must be interpreted according to cut-off values specifically established according to homogeneous or heterogeneous immunoassays. Variability induced by fluid intake must be taken into account. The interest of the heterogeneous immunoassay needs to be confirmed for the diagnosis of Environmental Tobacco Smoking.