Activation of Orbital and Medial Prefrontal Cortex by Methylphenidate in Cocaine-Addicted Subjects But Not in Controls: Relevance to Addiction

Abstract

Introduction

The ability of cocaine to increase dopamine (DA) in the nucleus accumbens (NAc) seems to be crucial for the reinforcing effects of cocaine (Koob and Bloom, 1988). However, cocaine-induced DA increases in the NAc do not explain addiction, because this effect occurs in nonaddicted as well as addicted subjects (Volkow et al., 1997a). A reasonable hypothesis is that addiction results from adaptations triggered by intermittent cocaine-induced supraphysiological perturbations of the DA system, which could disrupt the circuits regulated by DA. Indeed, positron emission tomography (PET) reveals that DA brain function (assessed by DA D₂ receptors and DA release in the striatum), as well as activity in the orbitofrontal cortex (OFC) and anterior cingulate gyrus (CG) (assessed by brain glucose metabolism), is markedly decreased in cocaine-addicted subjects tested during withdrawal (Volkow et al., 1993). Reductions in OFC and CG metabolism were associated with decreases in striatal DA D₂ receptors, which suggested that these cortical changes reflect the consequences of disrupted DA activity. From these studies, a hypothesis emerged that abnormalities in the OFC and CG, which are regions involved with salience attribution and inhibitory control and implicated in compulsive behaviors, underlie craving and compulsive drug intake in addiction (Volkow and Fowler, 2000). This hypothesis was evaluated in this study by comparing the regional brain metabolic responses of cocaine-addicted subjects with controls after administration of the stimulant drug methylphenidate (MP). In parallel, we also measured the association between the metabolic changes and the DA changes induced by MP to assess whether the brain responses were linked to its dopaminergic effects.

For this purpose, detoxified cocaine-addicted subjects and controls were administered two sequential doses of intravenous MP. PET imaging of [¹⁸F]deoxyglucose (FDG) was used to measure the effects of MP in brain glucose metabolism (marker of brain function) (Sokoloff et al., 1977), and imaging of [¹¹C]raclopride was used to measure DA transmission (Volkow et al., 1994). We selected MP because it is similar pharmacologically to cocaine; both drugs increase DA by blocking DA transporters (DATs) with similar potencies (Volkow et al., 1995) and induce similar changes in regional brain activation (Dirckx et al., 2004). Moreover, cocaine-addicted subjects report that intravenous MP has similar effects to cocaine (Wang et al., 1997). We chose two sequential doses rather than a single dose because this mimics better the pattern of drug intake in cocaine abuse (multiple administrations). Previous publications have documented the effects of MP on brain metabolism in controls (Volkow et al., 1997b) and in cocaine-addicted subjects (Volkow et al., 1999) and have compared DA changes between controls and addicted subjects (Volkow et al., 1997a). Here, we compare for the first time the brain metabolic responses to MP between cocaine-addicted subjects and controls and evaluate the association between MP-induced metabolic and DA changes. We hypothesized that responses to MP of addicted subjects and controls would differ in the OFC and CG and that these differences would be associated with the “desire for the drug” and would be linked to the dopaminergic effects of MP.

Materials and Methods

Subjects. The cocaine-addicted subjects were 21 male right-handed subjects [age, 36 ± 5 years; education, 12 ± 2 years; intelligence quotient (IQ)-verbal, 103 ± 11] who were inpatients at Veterans Administration Hospital (Northport, NY). Subjects fulfilled Diagnostic and Statistical Manual of Mental Disorders, fourth revision (DSM-IV) criteria for cocaine dependence and had used cocaine (free-base or as crack) for at least the previous 6 months (at least “3 g” per week). Note that subjects were excluded who met DSM-IV criteria for cocaine abuse rather than dependence. The exclusion criteria were current or past psychiatric disease other than cocaine dependence; past or present history of neurological, cardiovascular, or endocrinological disease; history of head trauma leading to loss of consciousness >30 min; and current medical illness and dependence or abuse on any substance other than cocaine, nicotine, or caffeine. Care was taken to exclude subjects that abused or were addicted to alcohol or drugs other than cocaine and nicotine. Subjects were tested within 1 month of their last cocaine use (14 ± 7 d) and had an average history of 13 ± 5 years of cocaine use. Controls were 15 healthy right-handed male subjects (age, 35 ± 7 years; education, 15 ± 2 years; IQ-verbal, 115 ± 12), and although they had previous experience with recreational drug use, they were excluded if they met DSM-IV criteria for alcohol or drug abuse or dependence. The exclusion criteria were as for the addicted subjects, except for cocaine dependence. Cocaine-addicted subjects remained in the hospital after completion of the study and were not discharged unless craving for cocaine had subsided. The protocol was approved by the Institutional Review Boards at Veterans Administration Hospital and at Brookhaven National Laboratory. Written informed consent was obtained after complete description of the study to the subjects.

Scans. Each subject underwent two PET FDG and two PET [¹¹C]raclopride scans done on two separate days. On a given day, each subject was given an injection first of [¹¹C]raclopride and then 90 min later an injection of FDG. Before each radiotracer injection, the subjects were given injections of a placebo (3 cc of saline intravenously before [¹¹C]raclopride and 3 cc of saline intravenously before FDG) on the first day and injections of MP (0.5 mg/kg MP intravenously before [¹¹C]raclopride and 0.25 mg/kg MP intravenously before FDG) on the second day (Fig. 1). The order of administration was fixed (placebo was given on the first day, and MP was given on the second day) to avoid possible long-lasting effects of MP on brain metabolism. The study was a single blind design (subjects were blind to the drugs received). For [¹¹C]raclopride, dynamic scans were started immediately after injection of 4-10 mCi of [¹¹C]raclopride (specific activity, 0.5-1.5 Ci/μm at end of bombardment) for a total of 60 min using procedures published previously (Volkow et al., 1994). For FDG, an emission scan was obtained for 20 min and was started 35 min after injection of 4-6 mCi of FDG on a CTI 931 tomograph (6.5 × 5.9 × 5.9 mm full-width at half-maximum; 15 slices) using procedures described previously (Wang et al., 1993). FDG measures were completed in all subjects, but the [¹¹C]raclopride scans in one of the controls and in two of the cocaine-addicted subjects were lost because of technical errors.

Open in a separate window

Figure 1.

Diagram of experimental design. Subjects were tested on two separate days. On each day, they were scanned first with [¹¹C]raclopride, and this was followed by an FDG scan. On the first day of the study, subjects received intravenous placebo before each radiotracer injection, and on the second day, they received intravenous MP before each radiotracer injection.

Behavioral measures. Before and at 27 min after the first and the second MP (or placebo) dose, subjects rated drug effects (high, desire for MP, desire for cocaine, restlessness, and mood) on a 0-10 analog scale (Wang et al., 1997). When asked about the “desire for MP” subjects were asked to rate the desire to have the drug that they had just received (placebo or MP). MP concentration in plasma was measured using capillary gas chromatography/mass spectroscopy (Srinivas et al., 1991). A factorial repeated ANOVA was used to assess the effects of MP in the behavioral measures (drug main effect) and to assess whether the differences differed between controls and addicted subjects (drug by group interaction effect).

Analysis. The data for the metabolic images were analyzed using Statistical Parametric Mapping (SPM) (Friston et al., 1995), and the results were corroborated with region of interest (ROI) analysis. For the SPM analyses, the images were spatially normalized using the template provided in the SPM 99 package, then normalized to the mean metabolic activity for the entire brain (mean of all voxels within the brain) and subsequently smoothed with a 16 mm isotropic Gaussian kernel. Paired-samples t tests were performed to compare the images obtained in the placebo and MP conditions separately for the addicted subjects and the controls. Independent (unpaired) t tests were used to compare the difference images (placebo - MP) for the cocaine-addicted and control groups. Significance was set at p < 0.005 for clusters ≥100 voxels. The statistical maps were overlaid on a magnetic resonance imaging (MRI) structural image

For the ROI analyses, an automated ROI extraction method was used based on the standard brain template of the Talairach atlas (Talairach and Tournoux, 1988). First, to eliminate variations across the brain of individuals, the PET FDG images were mapped into the Talairach brain using the spatial normalization package provided in SPM. The inverse mapping procedure was used to extract the Talairach coordinates of all voxels for a given anatomical region using the stereotaxic coordinates provided in the Talairach Daemon database (Collins et al., 1995; Lancaster et al., 1997, 2000). These anatomically defined ROI were overlapped voxel by voxel onto the normalized PET image (Volkow et al., 2002). This automated ROI method was validated by comparing the metabolic measures obtained with the manually drawn ROI and those obtained with the automated ROI from FDG images of 35 subjects (Ma et al., 2004). The correlations between the two methods were very high and ranged between r = 0.86 for head of caudate and r = 0.99 for the middle temporal gyrus. A factorial repeated ANOVA was used to assess the effects of MP in the metabolic measures (drug effect) and to assess whether the differences differed between controls and addicted subjects (drug by group interaction effect). Because controls and cocaine-addicted subjects differed in years of education and on their IQ scores, we performed a covariance analysis to assess whether these differences influenced the results.

For the [¹¹C]raclopride images, we obtained ROI in the striatum, thalamus, and cerebellum, from which we generated time-activity curves as described previously (Volkow et al., 1994). The time-activity curves for tissue concentration and unchanged tracer in plasma were used to calculate the distribution volumes (DVs), using a graphical analysis technique for reversible systems (Logan et al., 1990), to estimate the equilibrium ratio of tissue concentration to plasma concentration in the striatum, thalamus, and cerebellum. The ratios of DVs in the striatum and in the thalamus to those in the cerebellum correspond to (B_max/K_d) + 1 and are insensitive to changes in cerebral blood flow (Logan et al., 1994). The response to MP was quantified as the percentage of change in B_max/K_d from placebo (B_max/K_dPlacebo - B_max/K_dMP/B_max/K_dPlacebo × 100).

Pearson product moment correlations were used to assess the association between the regional metabolic changes induced by MP (expressed as the percentage of change from the metabolic measures obtained after placebo), which differed between controls and addicted subjects, and the changes in behavioral ratings induced by MP (difference between an individual's highest scores obtained during MP administration and the scores before MP administration). Correlations between the metabolic and behavioral measures were based on data from all subjects, except for the behavioral measure of cocaine craving when correlations were from just the cocaine abusers, because they were the only ones that experienced significant drug craving with MP. We also report the correlations with desire for MP and for “mood” separately for the cocaine abusers. In addition, correlations were obtained between the changes in DA (percentage of change in B_max/K_d) and changes in metabolic activity for regions where the response differed between controls and addicted subjects.

Results

Effects of MP on regional brain metabolism

The SPM analysis conducted separately for each group revealed significant relative metabolic increases with MP in the cerebellum and occipital cortex and relative metabolic decreases in the caudate and temporal insula, both in controls and in addicted subjects (Fig. 2). In addition, in controls, MP increased metabolism in the precuneus and left thalamus and decreased metabolism in the NAc and in the right orbital and medial prefrontal cortices (including BA 25 and medial BA 11). Table 2 provides the Talairach coordinates for the regions that were significantly affected by MP in controls. In contrast, in cocaine-addicted subjects, MP increased metabolism in the right orbital medial prefrontal cortex (OMPFC) and decreased metabolism in the right frontal cortex (BA 8, BA 6, and BA 4) and parietal cortex (BA 40) (Fig. 1). Table 3 provides the Talairach coordinates of the regions that were significantly affected by MP in cocaine-addicted subjects.

Open in a separate window

Figure 2.

Brain regions where MP significantly (p < 0.005) increased (red) or decreased (blue) metabolism in controls and in cocaine-addicted subjects are shown.

Table 2.

Location of regions where metabolism was increased or decreased significantly by MP in controls subjects (p <0.005; volume, ≥100 voxels)

Region	Hemisphere	BA	x	y	z
Increases
Occipital	Right	18	1	−93	4
	Left		−1	−93	5
Cerebellum	Right		1	−65	−28
	Left		−1	−65	−21
Precuneus	Left	31	−5	−63	27
Thalamus	Left		−1	−2	8
Decreases
Caudate head	Right		10	15	5
Temporal insula	Left		−4	−1	8
NAc	Right		5	10	−8
Frontal medial	Right	11, 25	7	27	11

Open in a separate window

Table 3.

Location of regions where metabolism was increased or decreased significantly by MP in cocaine-addicted subjects (p <0.005; volume ≥100 voxels)

Region	Hemisphere	BA	x	y	z
Increases
Occipital	Left	18	−3	−95	7
Cerebellum	Right		1	−77	−25
	Left		−1	−77	−25
Frontal medial	Right	11	11	37	−25
Decreases
Caudate head	Right		5	10	8
Temporal superior	Left	22	−47	−30	−12
Frontal paracentralis	Right	4, 6	45	−15	44
	Left		−47	−17	40
Frontal superior	Right	8	30	35	48
Parietal inferior	Right	40	55	−30	36
	Left		−59	−33	37
Occipital gyrus	Right	19	20	−85	28
	Left		−25	−85	28

Open in a separate window

Regional brain metabolic differences in response to MP between controls and addicted subjects

The SPM analysis of the difference image (placebo - MP) revealed that only two brain regions differed significantly at p < 0.005 between controls and addicted subjects: an area in the right OMPFC [as defined by Ongur and Price (2000)] that included BA 25 and middle BA 11 and an area in the left occipital gyrus (BA 19) (Fig. 3A, Table 4).

Open in a separate window

Figure 3.

A, Brain regions where the response to MP differed between controls and cocaine-addicted subjects. The results are presented with respect to the controls. B, Relative metabolic measures in the right OMPFC in controls and in cocaine-addicted subjects for the placebo and for MP conditions. Metabolism was significantly lower in addicted subjects than in controls for the placebo condition. MP decreased metabolism in controls but increased metabolism in addicted subjects (values correspond to means and SDs). R OMPFC, Right OMPFC. C, Relationship between the changes in metabolism induced by MP and the self-reports for desire for MP (r = 0.55; p < 0.0005). R Bd 25, Right Brodmann area 25.

Table 4.

Location of regions where increases in metabolism induced by MP were increased significantly in cocaine-addicted subjects compared with controls (p < 0.005; volume ≥100 voxels)

Region	Hemisphere	BA	x	y	z
Frontal medial	Right	11,25	17	33	−16
Occipital gyrus	Right	19	−37	−47	−5

Open in a separate window

The ROI comparison of the response to MP corroborated that the response to MP in the OMPFC differed between the groups (ANOVA; interaction effect; F = 31; p < 0.0001) but did not corroborate the differences in BA 19. The difference in the right OMPFC remained significant after covarying for years of education (ANCOVA; p < 0.0001) or for IQ (p < 0.0001). MP significantly increased metabolism in the right OMPFC (p < 0.05), whereas in controls it decreased metabolism (p < 0.0005) (Fig. 3B). Separate analysis for subregions revealed that the differences were significant both for right BA 25 (p < 0.0003) and for right BA 11 (p < 0.01). At baseline (placebo scan), metabolism in the right OMPFC was also significantly lower in cocaine-addicted subjects than in controls (F = 22; 1.1 ±.09 vs 1.24 ±.09; p < 0.0001).

Effects of MP on [¹¹C]raclopride binding and correlations with MP-induced change in metabolism

MP significantly decreased B_max/K_d (measure of receptor availability) in the striatum in both controls (placebo, 2.86 ± 0.28; MP, 2.22 ± 0.38) and cocaine-addicted subjects (placebo, 2.61 ± 0.39; MP, 2.34 ± 0.41; ANOVA; drug effect; F = 27; p < 0.0001). Baseline measures of B_max/K_d in the striatum were smaller in addicted subjects than in controls in the striatum (control subjects, 2.90 ± 0.34; addicted subjects, 2.59 ± 0.34; F = 4.1; p < 0.05) but did not differ in the thalamus (control subjects, 0.42 ± 0.10; addicted subjects, 0.43 ± 0.15; F = 1.9; p = 0.17). The responses in the striatum differed between the groups (ANOVA; interaction effect; F = 5.1; p < 0.05), indicating that the decrements were significantly smaller in addicted subjects (-10 ± 14% change) than in controls (-21 ± 16% change). In the thalamus, MP significantly changed B_max/K_d (ANOVA; drug effect; F = 5.2; p < 0.05), and the response also differed between groups (ANOVA; interaction effect; F = 10; p < 0.004). In addicted subjects but not in controls, MP significantly decreased B_max/K_d in the thalamus (control subjects, 16 ± 34% change; cocaine-addicted subjects, -35 ± 36% change). The correlation between MP-induced changes in B_max/K_d and the changes in metabolism in the right OMPFC were significant in the thalamus (r = 0.49; p < 0.005) (Fig. 4). The correlations with MP-induced changes in DA in the striatum were not significant (data not shown).

Open in a separate window

Figure 4.

Correlation between MP-induced changes in OMPFC metabolism and the changes in thalamic DA. R OMPFC, Right OMPFC.

Discussion

The main difference in the brain metabolic response to MP between the control and cocaine-addicted groups was in an area in the right OMPFC that included BA 25 (ventral region in the anterior CG) and medial BA 11 (medial region in the OFC). The qualitative difference in the response (in controls it decreased metabolism, and in cocaine-addicted subjects it increased metabolism) provides additional evidence of the involvement of the anterior CG and the OFC in cocaine addiction. Abnormalities in the OFC and anterior CG have been reported previously in cocaine-addicted (Modell et al., 1990; Volkow et al., 1991, 1993; Franklin et al., 2002; Bolla et al., 2003; Matochik et al., 2003; Myrick et al., 2004), methamphetamine-addicted (Volkow et al., 2001; Sekine et al., 2003), heroin-addicted (Sell et al., 2000), and marijuana-addicted (Volkow et al., 1996) subjects and in alcoholics (Modell et al., 1990; Volkow et al., 1997c; Dao-Castellana et al., 1998; Myrick et al., 2004), suggesting that they reflect a common abnormality across various types of drug addictions.

Footnotes

References