Cigarette smokers who completed treatment with either bupropion HCl or PGC showed similar regional changes in relative cerebral glucose metabolism. In the active-treatments group, we found a significant cluster of decreased glucose metabolism in the posterior cingulate gyrus. To determine their relative contributions to the ‘active-treatments' analysis, we considered the bupropion HCl and PGC groups separately, and (compared with placebo-treated individuals) smokers treated with PGC had a greater reduction in glucose metabolism with treatment in the posterior cingulate gyrus, whereas those treated with bupropion HCl showed a trend towards significance in the same region. Furthermore, these cerebral metabolic changes were not because of the change in number of cigarettes smoked per day (which was controlled for in this statistical analysis). As a direction for future study, we conducted an exploratory analysis to reveal more subtle differences between the active treatments. Our results suggest that effective treatments for cigarette smoking affect regional brain function through similar circuitry.
The posterior cingulate gyrus is a region that shows reduced activity in response to goal-directed tasks, regardless of task type (Gusnard and Raichle, 2001
), and is thought to mediate the intrinsic activity of the resting brain as part of the ‘default mode network' (Fransson and Marrelec, 2008
; Raichle et al, 2001
). The reduced posterior cingulate gyrus activity found here indicates that treatment for tobacco dependence with bupropion HCl or PGC results in lower resting glucose metabolism in the default mode network. Consistent with our findings, previous studies of bupropion HCl treatment of subjects with MDD showed decreases in activation of the posterior cingulate gyrus and precuneus in response to emotional stimuli (Robertson et al, 2007
), perhaps indicating a common mechanism of action when bupropion HCl is used for the treatment of MDD or tobacco dependence.
In the exploratory analysis, comparison between the bupropion HCl and PGC groups revealed differences that were in agreement with the previously mentioned changes, in particular a cluster of greater decrease in the PGC group in the posterior cingulate gyrus. In contrast, the bupropion HCl-treated group showed areas of greater metabolic decreases than the PGC-treated group in several areas including the left amygdala and anterior cingulate cortex, a region we have previously shown to exhibit attenuated cigarette cue-induced activation following bupropion HCl treatment in heavy smokers (Brody et al, 2004
). The difference between these active treatments, combined with increased effects at the shared sites (such as the posterior cingulate), may account for the increased efficacy of combined treatment with bupropion HCl and PGC (Reus and Smith, 2008
). Owing to the relaxed threshold we used in this comparison (p
0.001, uncorrected) these results should be considered preliminary and as providing possible direction for future study.
We found a large cluster representing an area of positive correlation between glucose metabolism and CPD in the posterior occipital gyrus, primarily on the left side but extending to the midline. Increased regional activity in the visual system is often observed in response to acute nicotine administration (Domino et al, 2000a
; London et al, 1988b
). Because number of cigarettes per day is correlated with blood nicotine concentration (Hill et al, 1983
; Russell et al, 1976
) and subjects in our study reduced their cigarette usage (and presumably nicotine intake), our results are consistent with past studies showing a relationship between increased plasma nicotine concentration and increased occipital activity. An additional cluster of positive correlation between daily smoking and regional metabolism was found in the left parietal–temporal junction, extending to the superior temporal gyrus. The superior temporal gyrus is linked to both visual and auditory alertness (Thiel and Fink, 2007
), consistent with previous research showing that smoking increases alertness in those who are tobacco dependent (Bates et al, 1995
; Thiel and Fink, 2007
There were several limitations of note in this study. The spatial resolution of the PET scanner may have limited our ability to locate small regions of significant group differences. Before scanning, subjects were allowed to smoke ad libitum, and subjects were scanned in the resting state, which may have resulted in factors outside of the study design, including recent cigarette use, influencing the results. Though we used a specific control for the bupropion group (matching pill placebo), there were no specific control conditions for the many different components of the PGC treatment. Finally, the use of patient report, rather than plasma bupropion or nicotine levels to determine medication compliance and smoking status, respectively, may have limited our ability to assess compliance; however, the use of exhaled carbon monoxide readings and repeated visits with study staff may have minimized this limitation.
The strengths of this study included a relatively large sample size for a study of this type (with 108 PET scans being used for analysis) and the use of standard imaging and image analysis methodology. We used carefully controlled treatment to isolate the effects of bupropion HCl and PGC. It is interesting to note that subjects in the placebo-treated group reduced their reported daily cigarette use, but did not show a corresponding reduction in exhaled CO. This discrepancy may reflect effects of smoking immediately before the post-treatment CO test.
Our results suggest that bupropion HCl and PGC treatment for tobacco dependence have similar effects on regional brain metabolism. Active treatments for tobacco dependence appear to work by lowering resting metabolism in part of the brain default mode network. Thus, this study supports the theory that active treatments for tobacco dependence move the brain into a more goal-oriented state. Future work is needed to assess metabolic changes that occur over a longer-term treatment program, as well as changes that occur when bupropion HCl and PGC are used concurrently (as is commonly done in clinical practice).