A Low-Cost, High-Throughput Digital Image Analysis of Stain Patterns on Smoked Cigarette Filter Butts to Estimate Mainstream Smoke Exposure

Abstract

Standard machine smoking protocols provide useful information for examining the impact of design parameters, such as filter ventilation, on mainstream smoke delivery. Unfortunately, their results do not accurately reflect human smoke exposure. Clinical research and topography devices in human studies yield insights into how products are used, but a clinical setting or smoking a cigarette attached to such a device may alter smoking behavior. To better understand smokers’ use of filtered cigarette products in a more natural environment, we developed a low-cost, high-throughput approach to estimate mainstream cigarette smoke exposure on a per-cigarette basis. This approach uses an inexpensive flatbed scanner to scan smoked cigarette filter butts and custom software to analyze tar-staining patterns. Total luminosity, or optical staining density, of the scanned images provides quantitative information proportional to mainstream smoke-constituent deliveries on a cigarette-by-cigarette basis. Duplicate sample analysis using this new approach and our laboratory’s gold-standard liquid chromatography/tandem mass spectrometry (LC/MS/MS) solanesol method yielded comparable results (+7% bias) from the analysis of 20 commercial cigarettes brands (menthol and nonmentholated). The brands varied in design parameters such as length, filter ventilation, and diameter. Plots correlating the luminosity to mainstream smoked-nicotine deliveries on a per-cigarette basis for these cigarette brands were linear (average R2 > 0.91 for nicotine and R2 > 0.83 for the tobacco-specific nitrosamine NNK), on a per-brand basis, with linearity ranging from 0.15 to 3.00 mg nicotine/cigarette. Analysis of spent cigarette filters allows exposures to be characterized on a per-cigarette basis or a “daily dose” via summing across results from all filter butts collected over a 24 h period. This scanner method has a 100-fold lower initial capital cost for equipment than the LC/MS/MS solanesol method and provides high-throughput results (~200 samples per day). Thus, this new method is useful for characterizing exposure related to filtered tobacco-product use.