Psychopharmacology (2007) 191:823–833 DOI 10.1007/s00213-006-0690-y
ORIGINAL INVESTIGATION
Human reward system activation is modulated by a single dose of olanzapine in healthy subjects in an event-related, double-blind, placebo-controlled fMRI study Birgit Abler & Susanne Erk & Henrik Walter
Received: 27 August 2006 / Accepted: 18 December 2006 / Published online: 31 January 2007 # Springer-Verlag 2007
Abstract Rationale Olanzapine is a neuroleptic drug widely prescribed to treat schizophrenia and bipolar disorder. Although it is long known that modulation of the dopamine system is a basic mechanism of action of neuroleptics, their impact on reward functions mediated by dopamine is still poorly understood. Objective Using functional magnetic resonance imaging (fMRI), we intended to reveal the effects of a single dose of olanzapine on reward-related brain activation. Methods Eight healthy subjects were each scanned twice, once 5 h after intake of 5 mg of olanzapine and once after intake of placebo in a double-blind cross-over design. Subjects performed a delayed incentive paradigm with monetary reward to investigate reward functions and a breath-holding task as a hypercapnic challenge to reveal unspecific drug effects on the fMRI signal. Results Reward-related brain activation in the ventral striatum, anterior cingulate and inferior frontal cortex was reduced on olanzapine compared to placebo. Only the differential effects (high>no reward) in the ventral striatum were independent of overall drug effects as measured with the breath-holding task. Parallel to the differential effects in the ventral striatum, the acceleration of reaction times in the trials with higher rewards was diminished in the olanzapine sessions. B. Abler (*) Department of Psychiatry, University of Ulm, Leimgrubenweg 12-14, 89075 Ulm, Germany e-mail:
[email protected] S. Erk : H. Walter Department of Psychiatry, Division of Medical Psychology, University of Bonn, Bonn, Germany
Conclusions Our behavioural and fMRI results can be interpreted as first evidence from neuroimaging that olanzapine affects the assignment of incentive salience represented by differential activation in dopaminergic brain areas and acceleration of reaction times. This can help to better understand neuroleptic effects in psychiatric diseases. Furthermore, we demonstrate the value of a hypercapnic challenge in functional pharmaco-MRI. Keywords Olanzapine . Pharmaco-fMRI . Incentive salience . Dopaminergic reward system . Hypercapnia . Healthy subjects
Introduction The neuroscientific investigation of reward-related processes in humans and animals in recent times has led to a better understanding of the functioning of dopaminergic brain areas. The involvement of dopamine signals in operant conditioning and learning (Schultz 2000, 2001) as well as addiction (Hyman 2005) has been demonstrated in animal studies. Neuroimaging studies in humans confirmed the involvement of dopaminergic brain regions in reward processes (reviews in: Knutson and Cooper 2005; McClure et al. 2004). Pharmacological neuroimaging studies showed that the human dopaminergic reward system comprising the ventral tegmental area, ventral striatum and orbitofrontal cortex can be stimulated by drugs of abuse like cocaine (Breiter et al. 1997) and modulated by dopamine agonists like amphetamine (Knutson et al. 2004). However, although widely prescribed to treat schizophrenia, bipolar disorder and other psychiatric conditions, the effects of neuroleptic drugs modulating the dopaminergic system on reward functions are still poorly understood. For schizophrenic
824
patients, some evidence points towards an impairment of reward-related but not associative learning correlating with neuroleptic drug dose (Cutmore and Beninger 1990). Furthermore, reduced sensitivity of dopamine receptors in the course of neuroleptic treatment has been associated with increased negative symptoms that may be related to a dysfunctional response to rewarding stimuli in patients (Schmidt et al. 2001). Animal studies suggest a reduction in reward-related behaviour under neuroleptic medication as well. Varvel et al. (2002) showed dose-dependant decreases of responses for food rewards in rats for various typical as well as atypical neuroleptics. Olanzapine is currently one of the most commonly prescribed neuroleptic agents. It is a dopamine receptor antagonist with high affinity to D1-, D2- and D4receptors and, in line with other atypical neuroleptics, seems to selectively affect the mesolimbic and mesocortical but not the dorsal striatal dopamine system (Bymaster et al. 1997; Moore et al. 1997). It induces an elevation of extracellular dopamine levels in vivo in the nucleus accumbens and prefrontal cortex in rats (Koch et al. 2004; Zhang et al. 2000). The antipsychotic effects of olanzapine have been attributed to the dampening of increased mesolimbic dopaminergic stimulation, which reduces the assignment of salience to important stimuli (Kapur 2003; Spitzer 1997). Functional magnetic resonance imaging (fMRI) revealed effects of the chronic administration of olanzapine in schizophrenic patients on functional connectivity (Stephan et al. 2001). The efficacy of only a single dose of olanzapine to increase resting state ventral but not dorsal striatal cerebral blood flow (CBF) in schizophrenic patients has been demonstrated using PET with 15O water (Lahti et al. 2005). In healthy subjects, a single dose of olanzapine resulted in similar peak occupancy rates of striatal dopamine receptors as a comparable dose in chronically treated schizophrenic patients. Peak plasma levels at 6 h after intake occurred concurrently with peak occupancy rates (Tauscher et al. 2002). Furthermore, it has been shown in monkeys that the acute administration of a single dose of olanzapine interferes with reward functions in the context of drug rewards (Howell et al. 2006). Medication with olanzapine reduced the incentive properties of addictive drugs in rats as well (Mechanic et al. 2002) and was suggested a valuable treatment of addiction in schizophrenic patients (Potvin et al. 2003). On the grounds of these previous findings, we expected that the administration of a single dose of olanzapine in the context of a rewarding task would have a dampening effect on differential rewardrelated brain activation as measured by fMRI. We expected that the activation differences indicating the incentive properties of a reward (highly, somewhat or not rewarding) would decrease, although dopamine levels in general are
Psychopharmacology (2007) 191:823–833
supposed to increase with neuroleptics irrespective of the incentive properties of a stimulus. Functional MRI measurements rely on basic mechanisms of CBF. Olanzapine may influence regional CBF (rCBF), for example, due to its interaction with cerebral vascular dopamine receptors (Amenta et al. 1991). To distinguish potentially task-related effects from more general drug effects on CBF, we applied a breath-holding task. The increase in CBF during hypercapnia, as for example induced by breath-holding, is thought to be similar to the coupling of CBF to neuronal activation, as both processes are mediated by an increase in carbon dioxide in cerebral arterioles (Kastrup et al. 1999). Previous studies have shown that breath-holding is a feasible task to investigate the influences of age (Riecker et al. 2003; Thomason et al. 2005) on the sensitivity of the blood oxygenation level dependent (BOLD) contrast to detect neural activation. The induction of hypercapnia can be thought of as a means to induce BOLD contrast in all vascularized tissues. The effects of anaesthetics on the BOLD signal in rat brains have been investigated successfully using a hypercapnic challenge (Brevard et al. 2003). We assumed that hypercapnia would be a means to reveal unspecific effects of other pharmacologic agents, such as olanzapine, on the BOLD signal in humans as well. To elicit reward-related brain activity, we used a validated delayed incentive task with monetary reward (Abler et al. 2005), investigating brain activation upon expectation and receipt of reward. Each participant was scanned twice: once after ingestion of oral olanzapine and once after placebo in a randomized order. We intended to investigate the effects of a single dose of an atypical neuroleptic on brain activation in healthy subjects. Olanzapine was chosen, as it is widely prescribed for psychiatric conditions, for its demonstrated effects on reward functions, specific effects on the ventral striatum and little interference with motor functions as required by our task (Bymaster et al. 1997). Based on the previous findings described above of a dampening effect on reward functions of olanzapine, we expected to find that olanzapine would decrease differential reward-related brain activation within dopaminergic brain areas previously shown to be involved in our task, i.e. the ventral striatum and prefrontal regions. Furthermore, we planned to differentiate specific effects of olanzapine on the reward task from more general effects on CBF as measured with a breath-holding task.
Materials and methods Eight right-handed healthy subjects (4 female) aged 19–43 (mean±SD, 31.3±8.8) were recruited from students and
Psychopharmacology (2007) 191:823–833
staff. The participants’ medical and psychiatric history were assessed in a screening interview by a physician trained in psychiatry. Exclusion criteria were any psychiatric or neurological illnesses currently or in the past, any serious general medical condition, use of illegal drugs and excessive consumption of caffeine, alcohol (>14 units/ week) and cigarettes (>20/day). Five participants were moderate smokers (usual maximum about 12 cigarettes per day). They were asked to refrain from smoking 1.5 h before scanning. One subject took a low dose of levothyroxine for mild hypothyroidism, one a contraceptive pill. Other than that, none of the subjects took any regular medication, and none had taken any other medication within the week before the study. Subjects were asked to refrain from alcohol 24 h before participation and from coffee on the day of the scans. Written informed consent after explanation of the procedures and before the inclusion in the study was obtained from each volunteer. The study was approved by the local ethics committee of the University of Ulm and conducted in accordance with the 1964 Declaration of Helsinki. Procedures Each subject was tested on two occasions, once with placebo (PLAC, gelatine capsule filled with mannitol powder) and once with oral olanzapine (OLAZ, 5 mg, identical capsule with pulverized olanzapine) in a randomized, balanced order double-blind design. Ten days to 8 weeks (average 22 days) passed in-between the two appointments. All participants arrived at the Department of Psychiatry at 8 A.M. for a light standardized breakfast and took the capsule. The scans were conducted between 1 and 2 P.M. Blood was drawn directly after scanning, i.e. 6 h after intake of drug or placebo to assess olanzapine blood levels. Subsequently, subjects took part in a short psychological testing session including a simple reaction times test, a memory task and questions on symptoms and side effects. In the scanner, subjects were tested on the monetary reward task and the breath-holding task. Before scanning, all subjects completed 10 min of a practice version of the reward task to minimize learning effects during scanning at each of the two appointments. Subjects were not paid for the practice task. Reward task Subjects were presented with a validated paradigm (Abler et al. 2005), a monetary incentive task with a parametric variation of possible wins (1 euro, 20 cent, no win). Each of the two sessions consisted of 60 trials (6,250 ms each; ten no-win-trials, 25 trials with potential gain of 1 euro and 25 trials with potential gain of 20 cent). Each trial started with one of three symbols (cue, 750 ms) indicating the possible amount of money to win. After an expectation period (delay, 3,000 ms) subjects had to correctly react with a left or right button press to one of
825
two symbols (a square or a triangle; target) within a fixed interval of 1 s with index or middle finger of their right hand. Subjects were informed that they did not need to react faster and that their chances to win were independent of their reaction times. In reacting correctly, they preserved themselves a 60% chance to win the announced amount of money (1 euro or 20 cent: win trial). In 40% of the trials, subjects were not rewarded despite pressing the correct button (omission trial). Incorrect button presses resulted in a feedback of zero euro at any rate. Win and omission trials as well as the three trial types (1 euro, 20 cent, no win) appeared in random order. In the control trials (no win) no money was announced; subjects only had to press an arbitrary button and could not win any money. To make sure that all trials included a button press of any kind, subjects were informed that they would lose 1 euro if no button press occurred. Feedback (outcome, 1,500 ms) followed the targets’ disappearance and notified subjects about the amount of money they won in the trial. The task was constructed following the task used by Knutson et al. (2001a,b). The main differences are: first, in our task, outcomes (reward/no reward) were independent of the velocity of the button press. This allows for using reaction times as a measure of motivation. Second, we avoided working with punishments but used omission of rewards, compared to receipt of rewards, to allow for investigating prediction error signals similar to animal experiments (Schultz et al. 2000). Reaction times and errors were registered. Median reaction times were calculated across trials for each single subject; means were calculated to average over subjects. Breath-holding task As a hypercapnic challenge, all subjects performed alternating periods of breath holding and self-paced breathing at 30 s intervals after normal expiration (Kastrup et al. 1999; Riecker et al. 2003). The time to stop and start breathing was indicated by a touch to the lower leg by a person standing next to the subject in the scanner room, observing performance and taking the time. Each subject performed two sessions of 2 min each on each of the two appointments. fMRI acquisition A 3 Tesla Siemens ALLEGRA Scanner (Siemens, Erlangen, Germany) equipped with a head coil was used to acquire T1 anatomical volume images (1×1× 1 mm voxels) and functional MR images. Twenty-three axial slices were acquired, with an image size of 64×64 pixels and a FoV of 192 mm. Slice thickness was 3 mm with 0.75 mm gap, resulting in a voxel size of 3×3× 3.75 mm. Images were angled along a line connecting basal forebrain and basal cerebellum. For the reward and breathholding task, images were centred on basal structures of the brain including subcortical regions of interest (basal
826
ganglia, orbitofrontal and ventral frontal regions). Functional images were recorded using a T2*-sensitive gradient echo planar imaging sequence measuring changes in BOLD contrast. Four hundred and one volumes were obtained during each of the two reward sessions at a time of repetition (TR) of 1,500 ms (time of echo (TE) 40 ms, flip 90°). The breath-holding task comprised two sessions of 87 volumes each. Apart from the number of volumes, parameters were identical to the reward task. fMRI analysis Image processing and statistical analysis were carried out using Statistical Parametric Mapping (SPM2, Wellcome Department, London, UK). Images were pre-processed including slice timing, realignment to correct for motion artifacts and spatial normalization to a standard template (Montreal Neurological Institute, MNI) with a resampled voxel size of 3×3×3 mm. Movement was less than the size of 1 voxel or one degree of rotation in any direction over the whole scans in all of the subjects. Smoothing was applied with an 8-mm Gaussian kernel. Intrinsic autocorrelations were accounted for by autoregression, 1st order (AR(1)), and low frequency drifts were removed via high pass filter. After preprocessing, first-level analysis was performed on each subject, estimating the variance of voxels according to a general linear model. Reward task The three expectation periods (including presentation of the cue) as well as the button press and the three different outcome events (win, omission, control trial) were each modeled as a boxcar function and convolved with the hemodynamic response function, resulting in seven orthogonal regressors. The six realignment parameters were included in the model, hence resulting in 13 regressors. The contrast images of parameter estimates were then included in a second-level group analysis (random effects model), treating inter-subject variability as a random effect to account for interindividual variance. We investigated (a) simple effects: effects in each, the OLAZ or PLAC group separately; and (b) interaction effects: comparisons of PLAC and OLAZ scans. Effects were calculated for four contrasts of interest that yielded meaningful results in our previous study (Abler et al. 2005), including: (1) expectation of high vs no reward; (2) expectation of high vs low reward; (3) outcome, win vs omission of reward; (4) outcome, omission of reward vs win. One-sample t tests were used to first investigate simple (within group) effects, separately for the PLAC and OLAZ scans. Paired t tests were calculated to then explore interaction effects comparing OLAZ and PLAC scans. Whole brain statistical maps were thresholded at p
