Biochemical estimation of noncompliance with smoking of very low nicotine content cigarettes

Abstract

Background: The reduction of the nicotine content of cigarettes to nonaddicting levels is a potential federal regulatory intervention to reduce the prevalence of cigarette smoking and related disease. Many clinical trials on the effects and safety of nicotine reduction are ongoing. An important methodologic concern is noncompliance with reduced nicotine content cigarettes in the context of freely available conventional cigarettes. We propose two approaches using biomarkers to estimate noncompliance in smokers of very low nicotine content (VLNC) cigarettes in a clinical trial.

Methods: Data from 50 subjects in a study of gradual nicotine reduction were analyzed. Using plasma cotinine concentrations measured at baseline and while smoking VLNC cigarettes, we compared within-subject ratios of plasma cotinine comparing usual brand to VLNC in relation to nicotine content of these cigarettes. In another approach, we used nicotine pharmacokinetic data to estimate absolute plasma cotinine/cigarettes per day (CPD) threshold values for compliance based on the nicotine content of VLNC.

Results: The two approaches showed concordance, indicating at least 60% noncompliance with smoking VLNC. In a sensitivity analysis assuming extreme compensation and extreme values for nicotine metabolic parameters, noncompliance was still at least 40%, much higher than self-reported noncompliance.

Conclusion: Biomarker analysis demonstrates a high degree of noncompliance with smoking VLNC cigarettes, indicating that smokers are supplementing these with conventional cigarettes.

Impact: We propose a practical approach to assessing compliance with smoking VLNC in clinical trials of nicotine reduction.

Figure 1

Observed plasma cotinine concentration vs cigarettes smoked per day in 50 subjects while smoking usual brand (circles) and VLNC cigarettes (boxes). Linear regression lines shown for different cigarette types (solid line = usual brand; dashed line = VLNC cigarettes). Data derived from Benowitz et al 2012 (14).

Figure 2

Frequency distribution of within-subject ratios of the plasma cotinine/CPD ratio while smoking VLNC and plasma cotinine/CPD while smoking usual brand. A ratio of ratios greater than 0.2 suggests noncompliance with VLNC only smoking condition. The insert provides further details on those subjects with ratios ≤ 0.2. Data derived from Benowitz et al 2012. (14)

Figure 3

Frequency distribution of ratios of plasma cotinine/CPD while smoking usual brand (solid portion of the bar) and VLNC cigarettes (striped portion of the bars) while smoking usual cigarettes (baseline) and VLNC (6 months). A ratio value of greater than 2.4 ng/ml/cigarette suggests non-compliance with VLNC cigarettes. Data derived from Benowitz et al 2012. (14)