Significant variability in cured leaf tobacco-specific nitrosamine (TSNA) content is commonly observed when sampling within dark air-curing barns. This variability may be due to inconsistency in the curing environment within different areas of the barn. A study was initiated in 2012 through support from a CORESTA Study Grant to evaluate if leaf TSNA content is related to microenvironmental conditions in the barn. Seed screened for low conversion of nicotine to nornicotine (sc) and high converter (HC) selections of TR Madole dark tobacco were cured in barns near Princeton and Lexington, Kentucky in 2012 and 2013. Temperature and relative humidity were measured with data loggers placed at 27 locations within each barn for the duration of curing. TSNA content was determined from 20-leaf samples collected from each selection at each of the 27 locations within each barn. There were no significant effects of individual data logger placement in either variety selection on hours above 24°C temperature, hours above 80% relative humidity, or TSNA; therefore, we investigated these data within 3-dimensional aspects of tier, room, and bent within each barn. There were various effects of tier, room, and bent on temperature, relative humidity, and TSNA; but limited significant relationships between temperature, relative humidity, and TSNA.

Measuring urinary tobacco specific nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) and its glucuronide conjugate may provide the best biomarker of tobacco smoke lung carcinogen metabolism. Existence of differences in the extent of NNAL metabolism rates may be potentially related to an individuals’ lung cancer susceptibility. Low concentrations of NNAL in smokers urine (<1 ng/mL) require sensitive and selective methods for analysis. Traditionally, this involves extensive, time-consuming sample preparation that limits throughput and adds to measurement variability. Molecularly imprinted polymers (MIPs) have been developed for the analysis of urinary NNAL by offline cartridge extraction combined with LC-MS/MS. This method when reproduced demonstrated problems with matrix effects. In the first part of this work, investigation of matrix effects and related problems with sensitivity for the published offline extraction method has been conducted. In order to address the need to improve throughput and other analytical figures of merit for the original method, the second part of this work deals with development of a high-throughput online microfluidic method using capillary-columns packed with MIP beads for the analysis of urinary NNAL. The method was validated as per the FDA guidance, and enabled low volume, rapid analysis of urinary NNAL by direct injection on a microfluidic column packed with NNAL specific MIP beads. The method was used for analysis of urinary NNAL and NNAL-Gluc in smokers. Chemometric methods were used with this data to develop a potential cancer-risk-assessment tool based on pattern recognition in the concentrations of these compounds in urine. In the last part, method comparison approaches for the online and the offline sample extraction techniques were investigated. A ‘fixed’ range acceptance criterion based on combined considerations of method precision and accuracy, and the FDA bioanalytical guidance limits on precision and accuracy was proposed. Data simulations studies to evaluate the probabilities of successful transfers using the proposed criteria were performed. Various experimental designs were evaluated and a design comprised of 3 runs with 3 replicates each with an acceptance range of ±20% was found appropriate. The off-line and the on-line sample extraction methods for NNAL analysis were found comparable using the proposed fixed range acceptance criteria.