Objective  To examine changes in β₂*-nAChR availability during acute and prolonged abstinence from tobacco smoking and to determine how changes in β₂*-nAChR availability were related to clinical features of tobacco smoking.

Design  Tobacco smokers participated in up to 4 iodide 123–labeled 5-iodo-A-85380 ([¹²³I]5-IA) single-photon emission computed tomography (SPECT) scans during abstinence at 1 day (n = 7) and 1 (n = 17), 2 (n = 7), 4 (n = 11), and 6 to 12 (n = 6) weeks. Age-matched nonsmokers participated in a single [¹²³I]5-IA SPECT scan. All participants completed 1 magnetic resonance imaging study.

Setting  Academic imaging center.

Participants  Tobacco smokers (n = 19) and an age-matched nonsmoker comparison group (n = 20).

Main Outcome Measure  The [¹²³I]5-IA SPECT images were converted to distribution volume and were analyzed using regions of interest.

Results  Compared with nonsmokers, β₂*-nAChR availability in the striatum, cortex, and cerebellum of smokers was not different at 1 day of abstinence, was significantly higher at 1 week of abstinence, and was not different at 4 or at 6 to 12 weeks of abstinence. In smokers, β₂*-nAChR availability was significantly lower in the cortex and cerebellum at 6 to 12 weeks compared with 1 week of abstinence. In addition, cerebellar β₂*-nAChR availability at 4 weeks of abstinence was positively correlated with craving on the day of the SPECT scan.

Conclusions  These data suggest that higher β₂*-nAChR availability persists up to 1 month of abstinence and normalizes to nonsmoker levels by 6 to 12 weeks of abstinence from tobacco smoking. These marked and persistent changes in β₂*-nAChR availability may contribute to difficulties with tobacco cessation.

Preclinical studies have demonstrated that nAChR levels return to control levels after termination of nicotine exposure but with great variability in the time course.^15-17 A postmortem human study indicated that individuals who had quit smoking at least 2 months before their death (range, 2 months to 30 years) had nicotine binding levels similar to those of control nonsmokers.² A recent in vivo study in a small number of male smokers demonstrated a trend for a normalization of the β₂*-nAChR availability by 21 days of smoking cessation.¹⁸ However, these data need to be confirmed in a larger, more heterogeneous population during prolonged abstinence.

The exact subunit combination of nAChR that upregulates in response to nicotine is emerging. The nAChRs that contain α₄ and β₂ subunits^19-20 are the most abundant nAChRs in the brain, and nicotine demonstrates the highest affinity for these receptors.²¹ In recent years, evidence has emerged that the β₂*-nAChRs are a critical neural substrate mediating the effects of nicotine in the brain. Much of this information has been derived from studies in β₂ knockout mice. These studies have demonstrated that the β₂ subunit is critical for self-administration,²² conditioned place preference,²³ discriminative stimulus and taste aversion,²⁴ dopamine release,^{22, 25-26} dopamine-dependent locomotor activation,²⁷ and enhancement of incentive aspects of motivation²⁸ of nicotine. Studies in wild-type animals further confirm that the β₂ subunit is critical to the reinforcing properties of nicotine.²⁹ The β₂ subunit also determines the sensitivity to nicotine^30-32 but does not play a critical role in nicotine withdrawal symptoms in rodents.³³ Nicotine-induced increases in nAChR, termed upregulation, are conferred by a specific microdomain that is in the β₂ subunit,³⁴ and thus this upregulation is confined to nAChRs that contain the β₂ subunit.^{4, 19, 31, 35-38}

Availability of β₂*-nAChR can be measured in vivo with the radiotracer iodide 123–labeled 5-iodo-A-85380([¹²³I]5-IA) and single-photon emission computed tomography (SPECT). The nicotinic agonist [¹²³I]5-IA binds with high affinity to the nicotine-binding site on nAChRs that contain the β₂ subunit.³⁹ This ligand demonstrates low nonspecific binding⁴⁰ and has acceptable dosimetry in humans with high brain uptake^41-42 and good test-retest reproducibility.⁴³ Because [¹²³I]5-IA is administered at a trace dose (<1% occupancy) for SPECT, it does not interfere with receptor and cell function. Imaging with [¹²³I]5-IA SPECT in nonhuman primates and humans results in a binding pattern that is consistent with the established regional distribution of β₂-nAChR and is highest in the thalamus and intermediate throughout the cortex and cerebellum.^43-44

The primary goal of the present study was to evaluate the time course of change in β₂*-nAChR availability during prolonged abstinence by using [¹²³I]5-IA SPECT. A secondary goal was to explore the relationships between β₂*-nAChR availability and behavioral features of tobacco smoking and withdrawal. We hypothesized that during acute abstinence (ie, at 1 day of withdrawal), β₂*-nAChR availability would be lower compared with 1 week of abstinence because of the presence of nicotine in the brain, which would block radiotracer binding. Consistent with our previous study,⁸ we hypothesized that, compared with control nonsmokers, β₂*-nAChR availability would be higher at 1 week of withdrawal and then would progressively decline during the course of prolonged abstinence.

METHODS

Nineteen healthy tobacco smokers (9 men and 10 women; mean [SD] age, 41.1 [9.0] years; 13 white American, 5 African American, and 1 Hispanic American) and 20 age-matched healthy control participants (9 men and 11 women; mean [SD] age, 42.4 [9.8] years; 10 white American, 7 African American, 2 Hispanic, and 1 Asian American) participated in this study. Smokers participated in up to 4 [¹²³I]5-IA SPECT scans and 1 magnetic resonance imaging (MRI) study. Participants were grouped by the following abstinence times (mean [SD]): 1 day (1.0 [0] days; n = 7), 1 week (7.7 [1.4] days; n = 17), 2 weeks (17.9 [3.0] days; n = 7), 4 weeks (30.5 [4.3] days; n = 11), and 6 to 12 weeks (69.0 [23.5] days; n = 6). Participants were grouped as described because of the difficulty in retaining participants who remained abstinent for long periods and because of the challenges in having participants complete time-consuming (eg, >8 h/d) scans on multiple days. Participants were scanned during a range of days at each time point because of scheduling constraints. Nonsmoker controls (n = 20) participated in 1 [¹²³I]5-IA SPECT scan and 1 MRI study.

This study was approved by the Yale University School of Medicine Human Investigation Committee, the West Haven Veterans Administration Human Subjects Subcommittee, and the Radiation Safety Committee. The use of the radiotracer [¹²³I]5-IA was approved by the US Food and Drug Administration. Participants were recruited from the community by word of mouth, posters, and newspaper advertisements. Eligibility was determined as follows. All participants underwent a medical examination by a study physician (E.P.) to exclude any major medical issues or neurological disorders. This included a physical examination, electrocardiography, serum chemistry evaluations, thyroid function studies, complete blood cell count, urinalysis, and urine toxicological screening. Participants were given structured interviews using the Structured Clinical Interview for DSM-IV to rule out any Axis I disorder except for nicotine dependence. All tobacco smokers had to smoke at least 10 cigarettes per day for at least 1 year. Smoking status was confirmed by plasma cotinine levels of more than 150 ng/mL, urine cotinine levels of more than 100 ng/mL (to convert to nanomoles per liter, multiply by 5.675), and carbon monoxide levels of more than 11 ppm on the day of intake. Smokers were helped to quit smoking using clinical practice guidelines and contingency management.^45-47 Briefly, contingency management is a behavioral therapy in which reinforcement is provided contingent on a successful response. In this study, monetary reinforcement was provided contingent on abstinence from smoking measured by urine cotinine and breath carbon monoxide levels. Breath carbon monoxide and urine cotinine levels were monitored daily for the first 8 days of smoking cessation and a minimum of twice weekly thereafter. Participants were instructed that they could not use any form of nicotine replacement therapy or medication throughout the study. All controls were nonsmokers (defined as having smoked <100 cigarettes in their lifetime) and had no history of significant medical illness or major head trauma. Nonsmoking status was confirmed by plasma cotinine levels of less than 15 ng/mL, urine cotinine levels of less than 100 ng/mL, and carbon monoxide levels of less than 11 ppm on the day of intake and the day of the scan. All women were required to have a negative pregnancy test result during the screening process and before radiotracer injection on each study day. Plasma nicotine and cotinine levels were measured as previously described.⁸ Urine cotinine levels were measured using cotinine test strips (Accutest NicoMeter [Jant Pharmacal Corporation, Encino, California] or NicAlert [Nymox Pharmaceutical Corporation, Hasbrouck Heights, New Jersey]).

The severity of nicotine dependence was assessed at intake with the Fagerström Test for Nicotine Dependence,⁴⁸ and craving and nicotine withdrawal symptoms were assessed with the Urge to Smoke Questionnaire (QSU)⁴⁹ and the Minnesota Nicotine Withdrawal Scale,⁵⁰ respectively, at baseline (ie, on the day of their intake before quitting smoking) and on each day they underwent [¹²³I]5-IA SPECT scans. The QSU has 2 main factors, the intention/desire to smoke (QSU-Intent) and relief of negative affect and withdrawal (QSU-Relief).

MRI STUDIES

The MRI studies were obtained on a 1.5-T camera (Siemens AG, Munich, Germany) in a standard orientation (echo time, 5-7 ms; repetition time, 24 ms; 256 x 192 matrix; number of signals acquired, 1; field of view, 30 cm; 124 contiguous sections with 1.2-mm thickness) and were used for coregistration to the SPECT images to provide an anatomical guide for placement of regions of interest.

[¹²³I]5-IA SPECT SCANS

All participants received a 0.6-g saturated solution of potassium iodide (to protect their thyroid from possible exposure to radioactive iodide) in the hour before radiotracer administration. The radiotracer [¹²³I]5-IA was synthesized as previously described⁵¹ and administered as a bolus to constant infusion at a ratio of 7.0 for 8 hours. Participants were injected with mean (SD) equivalent doses of a bolus and constant infusion as follows: control nonsmokers (157.9 [14.6] MBq and 22.8 [2.1] MBq/h, respectively) and smokers abstinent 1 day (134.5 [24.6] MBq and 20.5 [3.8] MBq/h, respectively), 1 week (154.3 [15.9] MBq and 22.2 [2.6] MBq/h, respectively), 2 weeks (159.4 [5.6] MBq and 23.3 [0] MBq/h, respectively), 4 weeks (158.9 [7.4] MBq and 22.6 [1.7] MBq/h, respectively), and 6 to 12 weeks (141.5 [30.5] MBq and 21.9 [3.5] MBq/h, respectively). Three consecutive 30-minute emission scans and one 15-minute simultaneous transmission and emission protocol scan were obtained from hours 6 to 8 of the [¹²³I]5-IA infusion on a SPECT camera (PRISM 3000 XP; Picker, Cleveland, Ohio). The SPECT camera is a 3-headed camera equipped with a low-energy, ultrahigh-resolution fan-beam collimator (photopeak window, 159 keV ± 10%; matrix, 128 x 128) with a uniform sensitivity across the field of view. A cobalt 57–distributed source was measured with each experiment to control for day-to-day variation in camera sensitivity. The axial resolution (full width at half maximum) is 12.2 mm, measured with a [¹²³I] line source in water in a cylindrical phantom. Blood was drawn before injection and at the beginning and end of the emission scans for analysis of plasma total parent and free fraction of parent tracer in plasma (free fraction). The chemical fate of [¹²³I]5-IA after the injection was assessed in plasma as previously described.⁵¹ Briefly, plasma total parent was assessed by acetonitrile protein denaturation, whereas the free fraction was determined by ultrafiltration (Centrifree units; Amicon, Beverly, Massachusetts).

IMAGE ANALYSIS AND OUTCOME MEASURES

Images were reconstructed and analyzed as previously described, including a nonuniform attenuation correction⁴³ with 1 exception. Specifically, in participants who underwent more than 1 SPECT scan, the second and subsequent SPECT scans were coregistered to the same position as the first scan to apply the same region-of-interest template for that subject. The MRIs were coregistered to the SPECT scans to provide an anatomical guide for placement of the regions of interest using commercially available software (MEDx, version 3.4; Medical Numerics, Inc, Maryland). Regions of interest chosen were those known to contain β₂-nAChRs and included frontal, parietal, anterior cingulate, temporal, and occipital cortices and the thalamus, striatum (an average of caudate and putamen), and cerebellum. Regions of interest were corrected to account for differences in size. Two raters (K.P.C. and J.B.) conducted the analysis. Variability between the raters was less than 12% across the regions of interest. The mean of the analysis from the 2 raters is reported.

The outcome measure of regional activity divided by the free plasma parent between 6 and 8 hours (V_T/f_P) was used to correct for possible differences in radiotracer metabolism or plasma protein binding between groups and participants. Specifically, V_T/f_P equals [¹²³I]5-IA uptake in a region of interest divided by the free plasma parent (both measured in kilobecquerels per milliliter).⁵² We refer to V_T/f_P as β₂*-nAChR availability because we are only measuring receptors that are "available" to be bound by the radiotracer. Receptors that are already occupied (ie, by residual nicotine, by a pharmacologically active metabolite [cotinine or nornicotine], or by endogenous neurotransmitter [acetylcholine]) are not available. The outcome measure V_T/f_P is proportional to the binding potential (Bmax/K_D, in milliliters per gram), which is proportional to the receptor number (Bmax) at equilibrium, given the assumptions that there is no change in affinity (K_D) and that nondisplaceable (nonspecific and free) uptake does not differ between participants or comparison groups. As described previously,⁴³ there is no appropriate reference region for this radiotracer, so nondisplaceable [¹²³I]5-IA uptake could not be measured. The measures of total plasma parent, free fraction, and free plasma parent (f_P; which is defined as total parent x free fraction) were compared between groups to determine differences in radiotracer metabolism or protein binding.

STATISTICAL ANALYSES

Data were analyzed using SAS, version 9.1 (SAS Institute Inc, Cary, North Carolina). We assessed differences in blood measures (total and free parent and free fraction) and regional β₂*-nAChR availability, V_T/f_P, between participants in the nonsmoker control group and each of the abstinent smoker groups at 1 day and 1, 2, 4, and 6 to 12 weeks using 2-sample t tests. Differences in regional brain V_T/f_P and behavioral measures of tobacco smoking and withdrawal between participants in the abstinent smoker groups were assessed using repeated-measures mixed-effects regression models with group as a fixed effect and compound symmetry covariance structure across repeated measurements. We evaluated these between-group differences using repeated-measures mixed-effects models to account for the observations between abstinent smoker groups that were not entirely independent, given that some abstinent smokers contribute observations to more than 1 group. For models examining between-group differences for which there was a significant effect of group, we conducted 4 planned post hoc between-group comparisons (ie, between abstinent smokers at 1 week and at 1 day and 2, 4, and 6-12 weeks). For these post hoc tests, P values were Bonferroni corrected for multiple comparisons, and statistical significance was considered at P  .00625. Correlational analyses for the associations between receptor availability and smoking assessments were conducted using SPSS, version 16.0 (SPSS Inc, Chicago, Illinois). Correlations among β₂*-nAChR availability, V_T/f_P, at each point, and clinical variables (smoking, craving, and withdrawal) were assessed with Spearman correlation coefficients (). Different slopes for each abstinence point were estimated to illustrate the relationship between significant clinical variables and β₂*-nAChR availability. Owing to multiple comparisons, statistical significance was considered at P  .01.

RESULTS

Nineteen tobacco smokers and 20 age-matched nonsmokers were included in the study. Smokers who participated in scans at different times since the last cigarette were equivalent in age, level of nicotine dependence (as assessed by the Fagerström Test for Nicotine Dependence at intake), number of cigarettes smoked per day, and years of smoking (Table 1). Plasma cotinine and nicotine and carbon monoxide levels were negligible in nonsmokers and were highest at 1 day of abstinence and decreased over time in smokers, confirming abstinence from smoking (Table 1).

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 1. Demographics and Smoking Characteristics of Nonsmoker Control Subjects and Smokers During Acute and Prolonged Abstinencea

β₂*-nAChR AVAILABILITY DURING ACUTE AND PROLONGED ABSTINENCE

There were no differences between groups in injected dose, bolus to infusion ratio, or time of scan (data not shown). Concentrations of [¹²³I]5-IA activity in the blood were measured to correct for potential differences between groups in radiotracer metabolism or protein binding. There were significant differences in total parent levels between nonsmokers and smokers abstinent 1 day and in free fraction levels between nonsmokers and smokers abstinent 2 weeks (Table 2). There were no differences in total parent or free fraction levels between groups at other time points or between groups at any time point in free parent levels. There was also variability between smokers who participated in multiple scans in changes in β₂*-nAChR availability over time, with some abstinent smokers showing dramatic changes in β₂*-nAChR availability (48% change in the cortex) and others showing barely any difference (eg, <5%) over time (individual data not shown).

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 2. Blood [123I]5-IA Measures of Nonsmoker Control Subjects and Smokers During Acute and Prolonged Abstinencea

Regional β₂*-nAChR availability, reflected by V_T/f_P, was compared between nonsmokers and each of the abstinent smoker groups and between the abstinent smoker groups (Table 3, Figure 1, and Figure 2). In smokers abstinent 1 day compared with nonsmokers, V_T/f_P was significantly reduced in the thalamus. In smokers abstinent 1 week compared with nonsmokers, V_T/f_P was significantly higher in the striatum and cerebellum and throughout the cortex. In smokers abstinent 4 weeks compared with nonsmokers, V_T/f_P was significantly higher in the occipital cortex (Table 3 and Figures 1 and 2).

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 3. β2*-nAChR Availability Throughout the Brain in Nonsmoker Control Subjects and Smokers During Acute and Prolonged Abstinence

View larger version (26K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. Availability of the nicotinic acetylcholine receptor containing the β2 subunit (β2*-nAChR) measured as regional activity divided by the free plasma parent. Availability is shown in individual nonsmokers (NS) and tobacco smokers at 1 day and 1, 2, 4, and 6 to 12 weeks of abstinence in the thalamus (A), striatum (average of caudate and putamen) (B), cortex (average of cortical regions including parietal, frontal, anterior cingulate, temporal, and occipital cortices) (C), and cerebellum (D). The line in each scatterplot represents the mean value of those participants. *Significant difference from control nonsmokers after Bonferroni correction using 2-sample t tests. Significant difference from smokers abstinent 1 week after Bonferroni correction using planned post hoc between-group comparisons subsequent to the analysis of repeated-measures mixed-effects regression models, including the overall effect of abstinent smoker group.

View larger version (27K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Mean parametric images illustrating the availability of the nicotinic acetylcholine receptor containing the β2 subunit in nonsmokers and tobacco smokers. Measurements for the smokers were obtained at 1 day and 1, 2, 4, and 6 to 12 weeks of abstinence at similar transaxial levels of the brain. The color scale is shown with red, yellow, green, and blue corresponding to values for regional activity divided by the free plasma parent.

Among the abstinent smoker groups, there were significant between-group differences in V_T/f_P in the thalamus (F_4,25 = 4.12; P = .01), parietal cortex (F_4,25 = 2.95; P = .04), frontal cortex (F_4,25 = 2.88; P = .04), anterior cingulate (F_4,25 = 3.75; P = .02), occipital cortex (F_4,25 = 3.42; P = .02), and cerebellum (F_4,25 = 4.00; P = .01). Compared with smokers abstinent 1 week, those abstinent 1 day had significantly lower V_T/f_P in the thalamus (t₂₅ = –3.51; corrected P = .007) and cerebellum (t₂₅ = –3.02; corrected P = .02). Compared with smokers abstinent 1 week, those abstinent 6 to 12 weeks had significantly lower V_T/f_P in the parietal cortex (t₂₅ = –2.87; corrected P = .03), frontal cortex (t₂₅ = –2.80; corrected P = .04), anterior cingulate (t₂₅ = –3.21; corrected P = .01), occipital cortex (t₂₅ = –3.10; corrected P = .02), and cerebellum (t₂₅ = –3.17; corrected P = .02). There were no significant differences in V_T/f_P between tobacco smokers abstinent 1 week and smokers abstinent 2 or 4 weeks (Table 3 and Figures 1 and 2).

RELATIONSHIP BETWEEN β₂*-nAChR AVAILABILITY AND CLINICAL FEATURES

There were significant differences in QSU-Intent (F_4,25 = 7.41; P < .001) and QSU-Relief scores (F_4,25 = 5.63; P = .002), but not in Minnesota Nicotine Withdrawal Scale scores (F_4,25 = 0.33; P = .86), between groups of abstinent smokers (Table 4). Specifically, overall group differences in QSU-Intent and QSU-Relief scores were attributable to significantly higher levels of each of these measures on the day of the scan in the smokers abstinent 1 day compared with those abstinent at other points.

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 4. Clinical Characteristics of Smokers During Acute and Prolonged Abstinencea

There were significant correlations between regional β₂*-nAChR availability and clinical features and assessment scores at baseline, eg, at intake before quitting smoking (Table 5). Specifically, baseline QSU-Intent scores correlated negatively with β₂*-nAChR availability at 1 day of abstinence in the thalamus ( = –0.90; P = .006) and parietal cortex ( = –0.88; P = .008). There were also significant correlations between regional β₂*-nAChR availability and assessment scores taken at each abstinence point (Table 6). Specifically, a positive correlation was observed between cerebellar β₂*-nAChR availability and craving on the QSU-Intent ( = 0.74; P = .01) and QSU-Relief ( = 0.74; P = .01) scores at 4 weeks of abstinence. There were no significant correlations between baseline smoking variables (eg, Fagerström Test for Nicotine Dependence, number of years smoked, or number of cigarettes smoked per day) with β₂*-nAChR availability at any point (data not shown).

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 5. Correlations Between β2*-nAChR Availability and Baseline Assessments

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 6. Correlations Between β2*-nAChR Availability and Assessments on Day of Scan

COMMENT

In the smokers abstinent 1 day, the levels of total parent of the radiotracer were significantly lower, but normalized quickly, by 1 week of abstinence. This finding highlights the impact of nicotine or another chemical in tobacco smoke on metabolism, eg, because nicotine was still present, it may have changed the metabolism of the radiotracer, resulting in lower total parent levels at 1 day of abstinence. Cytochrome P450 (CYP2A6) is primarily responsible for the metabolism of nicotine to its main metabolite cotinine.⁵³ There is evidence that nicotine is metabolized faster in smokers than in nonsmokers, and there are genetically mediated differences in the metabolism of nicotine in smokers.⁵⁴ In addition, nicotine can interfere with the metabolism of other drugs.⁵⁵ Consistently, we expect that [¹²³I]5-IA is metabolized in the liver by enzymes in the cytochrome P450 family, such as CYP2A6, which acts on nicotine, and CYP2B6 and CYP2D6, which catalyze the dealkylation of aromatic ethers.^56-57 In radiotracer imaging studies, it is imperative that brain uptake is corrected for radiotracer metabolism because differences in metabolism of the radiotracer determine how much radiotracer is available to the brain. That is, fast metabolizers will have less radiotracer available to the brain, whereas slow metabolizers will have more radiotracer available to the brain for a given dose. By using the outcome measure V_T/f_P, we corrected for individual differences in radiotracer metabolism and protein binding.

In the present study, we report a negative correlation between baseline craving scores before smoking cessation and β₂*-nAChR availability at 1 day of abstinence in the thalamus and parietal cortices. Because receptor availability is defined as receptors that are available to be bound by the radiotracer, at 1 day of abstinence participants with lower receptor availability have more nicotine present in the brain occupying receptors and blocking the radiotracer from binding to the receptor. Thus, we believe that participants who reported high baseline craving likely smoked more cigarettes immediately before their quit day and had lower receptor availability at 1 day of abstinence. However, the experience of craving in the presence of nicotine occupancy of the β₂*-nAChR is not unusual. Smokers experience craving within 2 hours of their last cigarette, despite continued occupancy of the receptor by nicotine.¹³ Dopamine release has been associated with the feeling of craving,⁵⁸ and nicotine⁵⁹ and cotinine⁶⁰ have been shown to facilitate dopamine release; thus, the prolonged partial occupancy of the receptor by nicotine or cotinine may contribute to the feelings of craving that are reported 2 hours after the last cigarette and throughout the first week of abstinence.

We also report a positive correlation between craving on the day of the scan at 4 weeks of abstinence and cerebellar β₂*-nAChR availability at 4 weeks of abstinence. Thus, individuals with higher cerebellar β₂*-nAChR availability at 4 weeks of abstinence report a greater urge to smoke on that day. Associations between nicotine and craving have previously been identified in the thalamus^61-62 and generally in areas associated with emotion and reward and areas with high densities of nAChRs.⁶³ A previous study found associations between craving and regions subserving motor functions, including the primary motor cortex, premotor cortex, and supplementary motor area,⁶⁴ which require input from the cerebellum. In addition, when smokers are told to resist craving actively during cigarette cue exposure, the motor cortex is deactivated.⁶⁵ Our finding of a link between the cerebellum and craving further suggests that craving has a motor component, so that craving for cigarettes may elicit preparation for action or voluntary movement, goal-directed actions (eg, lighting a cigarette or bringing a cigarette up to the mouth), and/or motor imagery that are linked to the motor system.^66-67 It can be estimated that during the course of 20 years, a smoker who consumed 1 pack per day (assuming 11 puffs on average per cigarette⁶⁸) may perform the action of bringing a lit cigarette to the mouth more than 1.5 million times. Thus, the physical action of smoking a cigarette is likely to be critically tied to craving during abstinence.

In addition to the role of the cerebellum in motor functions, there is increasing interest in the involvement of the cerebellum in cognition. Specifically, activation of the cerebellum has been associated with tasks requiring explicit, episodic memory (eg, recall of autobiographical events).^69-70 With regard to drug abuse, imaging studies have determined that the cerebellum is activated in response to smoking-related cues⁷¹ and is associated with cue-induced craving in cocaine abusers^72-73 and recently abstinent alcoholics.⁷⁴ Activity in the cerebellum has also been linked to executive dysfunction in cocaine users.⁷⁵ Together these studies and the present findings highlight the cerebellum as a brain region that is critically linked to craving by both motor and cognitive functions.

We report no significant correlations between nicotine withdrawal and β₂*-nAChR availability. This is consistent with the preclinical literature suggesting that the β₂*-subtype does not play a critical role in the physical symptoms of nicotine withdrawal.^{33, 76} In this study, participants reported a mild-to-moderate level of nicotine withdrawal symptoms at baseline and over the course of the study; thus, these results require replication in a larger sample with a greater range of nicotine withdrawal symptoms. In addition, we did not obtain significant correlations between β₂*-nAChR availability at 1 week of abstinence and clinical features. We previously found a negative correlation between the urge to smoke to relieve withdrawal symptoms and β₂*-nAChR availability in the sensorimotor cortex at approximately 1 week of abstinence.⁸ This discrepancy may be due to differences in correlational analysis methods, that is, voxel-based analyses in the previous study vs Spearman correlations with regions of interest in the present study. Voxel-based analyses may be more sensitive to detecting significance in smaller brain regions, but that was beyond the scope of the present study. The lack of additional correlations at 1 week of abstinence was discussed previously.⁸

Consistent with our previous study,⁸ we report significantly higher β₂*-nAChR availability in smokers at 1 week of abstinence in the cortex, striatum, and cerebellum but not in the thalamus compared with nonsmokers. The difference between recently abstinent smokers and nonsmokers in the previous study⁸ was of a greater magnitude (26%-36% in the cerebral cortex and 27% in the striatum) than in the present study (21%-29% in the cerebral cortex and 22% in the striatum). This may be because of the older average age (approximately 5 years) of participants in the current study, because β₂*-nAChR availability has been shown to decrease with age in nonsmokers.⁷⁷ This study and previous in vivo positron emission tomography⁷⁸ and SPECT^{8, 18} studies report no upregulation of thalamic β₂*-nAChR availability during acute abstinence, which conflicts with the findings of postmortem² and animal^{8, 79} studies. In general, this may be because of differences in methods or the higher relative dose of nicotine in the postmortem and animal studies. However, 2 smokers in the present study exhibited increased β₂*-nAChR availability in the thalamus compared with nonsmokers, highlighting the role of individual differences in receptor regulation.

These findings are also consistent with those of a previous study¹⁸ demonstrating that higher β₂*-nAChR availability in recently abstinent tobacco smokers compared with nonsmokers is temporary. In the previous study in men, β₂*-nAChR availability decreased to nonsmoker levels in some participants by 21 days of abstinence. Specifically, compared with nonsmokers, smokers had significantly lower β₂*-nAChR availability at 4 hours of abstinence, significantly higher β₂*-nAChR availability at 10 days of abstinence, and similar β₂*-nAChR availability at 21 days of abstinence. In addition, they reported significantly lower β₂*-nAChR availability at 21 days of abstinence compared with 10 days of abstinence. One difference is that the previous study had significantly lower β₂*-nAChR availability in all regions at 4 hours of abstinence compared with the nonsmokers, whereas the current study reports lower β₂*-nAChR availability in the thalamus but similar β₂*-nAChR availability in the striatum and cerebellum and throughout the cortex compared with the nonsmoker group at 1 day of abstinence. This finding is interesting and likely attributable to high levels of residual nicotine or metabolites present in the brain at 4 hours of abstinence (vs approximately 24 hours in the present study), resulting in lower β₂*-nAChR availability. Also, although the present study did not find a significant decrease, on average, in β₂*-nAChR availability by 4 weeks of abstinence compared with 1 week of abstinence, in some participants normalization had occurred by this point. The high heterogeneity of the subject population in the present study compared with the previous study in men only¹⁸ likely accounts for the high individual variability with regard to receptor changes during prolonged abstinence. Thus, the present study contributes to the literature with a larger, more heterogeneous subject group, a more prolonged period of abstinence, and additional assessments of behavioral features of tobacco smoking.

Previous preclinical studies demonstrated that, in animals receiving long-term nicotine treatment, nAChR levels returned to levels observed in control animals after termination of nicotine administration, but with variable timing ranging from 1 to 3 weeks.^{15, 80-81} The previous human study¹⁸ indicated a return to control levels within 21 days of abstinence, and the present study indicates that, on average, β₂*-nAChR availability in recently abstinent tobacco smokers does not normalize until 4 to 12 weeks, although this is highly variable between individuals. The differences in the time course changes may result from differences in dosing regimen, chronicity of nicotine administration, route of administration, metabolism between species, specificity of the radioligand, or sex and/or genetic differences in nAChR subunit expression and composition of nicotinic agonist binding sites. However, when taken together, the results consistently highlight a return to control levels or normalization of the β₂*-nAChR availability after termination of long-term nicotine administration. The preclinical studies support a process of prolonged normalization when we consider that 1 to 3 weeks is substantial in the life span of a rodent. In addition, this prolonged normalization of the receptor availability is in line with the protracted withdrawal symptoms reported by tobacco smokers. For example, although withdrawal symptoms such as anger, anxiety, depression, and difficulty concentrating tend to peak within the first week after smoking cessation, they continue up to 4 weeks after the cessation attempt.⁸² We are currently examining variables that may be associated with the rate of normalization.

One limitation of this study is that [¹²³I]5-IA measures the availability of the β₂ subunit of nAChR, primarily the α₄β₂ subunit, but other subunits also contribute to the regulation of nAChR by tobacco smoking. The β₂ subunit has been linked primarily to the reinforcing effects of nicotine^{22, 28, 83} and combines with the α_3-6 subunits. Different β₂* subunit combinations appear to be differentially regulated by nicotine, with α₃/α₆, α₆α₄β₂*, and α5α4β2* not upregulating or decreasing in response to nicotine,^{38, 84-87} whereas α₆β₂ (non-α₄) subunits of nAChR increase in response to nicotine.⁸⁷ The nAChR subunit expression varies regionally, and differential regulation of these distinct subunit combinations that are measured by a single nicotinic agonist ligand such as [¹²³I]5-IA will result in regional differences in the degree of upregulation in smokers and the degree of receptor normalization during abstinence. Thus, we are likely measuring the β₂ subunit in combination with α_3-6 subunits, leaving the in vivo measurement of the β₃ and α₇ subunits for the future, with further radiotracer development. The α₇ and β₃ subunits have been implicated in modulating dopamine release^{14, 88} and thus may also play a role in the rewarding properties of tobacco smoke. Although the α₇ subunit does not upregulate in response to chronic nicotine use,³¹ it is critically involved along with the β₂ subunit in mediating desensitization/inactivation of the neuronal nAChRs in response to chronic nicotine.⁸⁹

In summary, our data extend the findings of previous studies^{8, 18} that showed that the upregulation of the β₂*-nAChR in recently abstinent tobacco smokers was temporary and could be measured with [¹²³I]5-IA SPECT imaging. Specifically, we demonstrate that the normalization of upregulated β₂*-nAChR in tobacco smokers during smoking cessation is prolonged. This is consistent with the clinical course of tobacco smoking in which craving, withdrawal symptoms, and risk for relapse are prolonged. The variation between individuals in the magnitude of upregulation and rates of normalization may ultimately be used to delineate subgroups based on genetics, sex, and comorbid mental illness and thus to target treatment medications.

AUTHOR INFORMATION

Submitted for Publication: June 5, 2008; final revision received October 28, 2008; accepted November 14, 2008.

Financial Disclosure: None reported.

Funding/Support: This study was supported in part by grants RO1 DA015577 and KO2 DA21863 (Dr Staley), KO1 DA20651 (Dr Cosgrove), P50 DA13334 and P50 AA15632 (Dr O’Malley), K25 NS044316 (Dr Maciejewski), and T32 14276 from the National Institute of Health.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Alcohol Abuse and Alcoholism, the National Institute on Drug Abuse, or the National Institutes of Health.

Additional Contributions: Marina Picciotto, PhD, and Robert Makuch, PhD, provided helpful comments on the manuscript. Louis Amici, BS, and the nuclear technologists at the Institute for Neurodegenerative Disorders provided technical assistance. The laboratory of Peter Jatlow, MD, determined plasma nicotine and cotinine levels.