This study validates prior findings of the predictive validity of the 3-HC/cotinine ratio as a marker for successful quitting with 21mg transdermal NRT (Lerman et al., 2006
). Across two independent studies we have demonstrated that the rate of nicotine metabolism, measured by the 3-HC/cotinine ratio, predicts cessation after treatment with the standard 21mg dose of nicotine patch. While smokers in the upper (faster) quartiles of nicotine metabolism show the expected approximately 30% 8-week response to transdermal NRT, smokers in the lowest quartile (i.e., the slow metabolizers) show quit rates comparable to varenicline (i.e., 47%), a recently marketed pharmacotherapy for nicotine dependence (Jorenby et al., 2006
). While the binary 3-HC/cotinine ratio variable was not associated with baseline cigarettes per day, number of previous quit attempts, FTND, or gender, smokers in the lowest quartile of the 3-HC/cotinine ratio had significantly lower breath CO levels and plasma nicotine levels at baseline prior to treatment, vs. those in the highest three quartiles, and were more likely to be non-Caucasian (results not shown). Together with our initial study, the present data suggest the potential value of this phenotypic marker of nicotine metabolism to select a treatment modality for nicotine dependence.
There is, however, a difference in the pattern of associations between the present study and our previously published study. In the earlier study (Lerman et al., 2006
), we found a relatively linear reduction in the probability of successful cessation from lowest to highest quartile, with the most robust effect detectable among those in the highest 3-HC/cotinine quartile (i.e., the fastest metabolizers), who showed the lowest end-of-treatment quit rates. In contrast, the present data show that the effect of the 3-HC/cotinine ratio on quit rates following transdermal NRT is largely attributable to a statistically significant and clinically meaningful increase in quit rate among those in the lowest 3-HC/cotinine ratio quartile (i.e., the slowest metabolizers), vs. other quartiles.
There are several potential explanations for this difference. First, the two studies differed in some aspects of methodology. The initial study used the step-down procedure for dosing transdermal nicotine, whereas the validation study maintained participants on 21mg patches for the full 8 weeks. Second, there were potentially meaningful differences in subject characteristics comparing the two samples. In the initial trial, participants had faster rates of nicotine metabolism as indicated by the 3-HC/cotinine ratio (mean = .44, SD = .9), versus the present study (mean = .38, SD = .20), and differences in the distributions of 3-HC/cotinine may influence relationships between nicotine metabolism and smoking behaviors. However, using the 3-HC/cotinine quartile cut-offs from our previous study to determine 3-HC/cotinine quartiles in the present sample did not change the results reported here (results not shown). The subject sample in the initial study was also more ethnically/racially diverse than the sample in the present study, and race affects rate of nicotine metabolism (Benowitz, 2008
). Lastly, the sample in the initial study included a higher proportion of women than the present study, and women metabolize nicotine at a faster rate, especially if they are using oral contraceptives (Benowitz et al., 2006
), and may respond less favorably to transdermal nicotine (Perkins and Scott, in press
). Overall, these methodological and sample differences across the two studies may explain the disparity in the nature of the relationship between 3-HC/cotinine ratio groups and response to transdermal NRT observed across the studies.
The present findings also replicate our previous result showing that slow metabolizers have higher plasma nicotine levels during NRT treatment (Lerman et al., 2006
). This finding provides additional validation of the important role played by 3-HC/cotinine ratio in determining therapeutic response to transdermal nicotine, since the slow metabolizers (with higher nicotine concentrations) were significantly less likely to relapse at 8 weeks, vs. intermediate or fast metabolizers (who showed lower levels of plasma nicotine). Thus, the 3-HC/cotinine ratio is related to both the extent of therapeutic nicotine replacement and quit rates.
The present results should, however, be considered in the context of methodological limitations. First, this study did not include a placebo to allow for assessment of whether or not the 3-HC/cotinine ratio effect on cessation is unique to treatment with transdermal nicotine. In a separate study, which included a placebo arm, among placebo-treated subjects, faster metabolizers of nicotine were more likely to relapse compared to slower metabolizers (Patterson et al., 2008), suggesting that the liability to relapse among fast metabolizers is likely a function of several mechanisms, not only nicotine replacement. Second, the present study did not include a long-term follow-up of cessation rates. We are currently following the present cohort to assess long-term outcomes and have demonstrated previously the effect of the 3-HC/cotinine ratio on long-term abstinence rates (Lerman et al., 2006
; Patterson et al., 2008). Lastly, the present sample was comprised of treatment seeking smokers and it did not have substantial ethnic/racial variability. Therefore, the results are generalizable to treatment seeking smokers and should be extended to other racial/ethnic groups in future studies.
Nevertheless, this study underscores the potential value of assessing pretreatment nicotine metabolism rate from cigarette smoking when considering use of a standard dose of transdermal nicotine. Perhaps only slow metabolizers (i.e., the first quartile for 3-HC/cotinine ratio) should be considered good candidates for this dose of transdermal nicotine. Recent data suggest that faster metabolizers (those in the 4th quartile) achieve significant benefit from bupropion for smoking cessation, while slow metabolizers quit at similar rates on bupropion and placebo (Patterson et al., 2008). Since bupropion is not metabolized by CYP2A6, this finding suggests that the rate of nicotine metabolism influences the level of nicotine dependence and the need for additional pharmacotherapy. Additional randomized clinical trials are needed to determine whether other non-nicotine medications, such as varenicline, are also more efficacious for faster metabolizers, and to identify the optimal treatments for smokers with nicotine metabolism rates in the intermediate range. Further, additional studies are needed to determine if faster nicotine metabolizers are more effectively treated with high doses of transdermal nicotine. While the current and previous findings strongly suggest that physicians could use the nicotine metabolite ratio as a noninvasive biomarker of nicotine metabolism to select slow metabolizers for the transdermal patch, future studies are needed to determine the cost-effectiveness of such tailoring and the optimal cut-points for different populations. Once studies determine the appropriate treatments matched to level of nicotine metabolism, this phenotypic marker may be useful to screen individual smokers for a specific form of smoking cessation therapy, thereby increasing treatment response rates and reducing the overall rate of tobacco use.