A new paper in Biological Psychiatry by Nielsen and colleagues at Copenhagen University Hospital and University College London helps to clear up a confusing literature on reward processing in patients with psychotic illness. The authors report on data from an impressively large sample of 31 neuroleptic-na´ve schizophrenia patients and 31 demographically matched controls who were administered a Monetary Incentive Delay (MID) task. Variants of this task have been used for over a decade, in conjunction with neuroimaging, to examine the neural substrates underlying the anticipation and integration of rewards and punishments in human subjects. Results of multiple studies using MID tasks point to phasic activation of the human ventral striatum (VS) during the anticipation of rewards, leading some to draw a connection between these neural activity patterns and evidence of nucleus accumbens involvement in what Berridge and Robinson have termed, "incentive salience."
In schizophrenia, in particular, researchers have used MID paradigms to test the idea (Kapur, 2003; Kapur et al., 2005) that supposed dopamine system dysfunction leads to aberrations in the signaling of salience. In fact, initial studies appeared to show that, in unmedicated schizophrenia patients (Juckel et al., 2006b), and in schizophrenia patients on first-generation antipsychotics (Juckel et al., 2006a), VS activity associated with reward anticipation is blunted. However, additional studies also appeared to show that VS activity associated with reward anticipation in schizophrenia patients on second-generation antipsychotics is either normalized (Schlagenhauf et al., 2008; Walter et al., 2009; Gradin et al., 2011), or limited to patients with severe negative symptoms (Waltz et al., 2010; Simon et al., 2010).
These early studies, often underpowered, have left several important questions unanswered, including the following:
1. Is blunted VS activity associated with reward anticipation characteristic of schizophrenia patients prior to treatment with antipsychotic drugs?
2. Is blunted VS activity associated with reward anticipation characteristic of most or all subtypes of patients with schizophrenia, or is it especially characteristic of patients with severe negative symptoms?
3. What is the relationship between VS activity associated with reward anticipation and positive symptoms?
4. Is blunted VS activity in schizophrenia specifically associated with the anticipation of rewards, or does it reflect a disrupted ability to signal salience in general?
Nielsen and colleagues set out to address some of these issues, using a cleverly designed MID paradigm. In their task, Nielsen et al. crossed certainty with valence to enable them to examine whether valence or (unvalenced) salience, or some combination of the two, drives VS activity in this sort of task. Specifically, the authors manipulated whether outcomes were 1) positive, negative, or neutral, and 2) certain or uncertain, presenting 24 trials with each of these kinds of cues for 144 total trials, in 36 minutes of scanning.
The authors performed BOLD signal contrasts meant to isolate VS activity evoked by salient cues (cues of either valence associated with uncertain outcomes), along with VS activity evoked by cues to gains (as opposed by cues to losses). They found that group differences in neural activity were evident in VS and the dopaminergic midbrain evoked by cues to uncertain outcomes, relative to both neutral outcomes and certain outcomes. Furthermore, the strength of salience-evoked responses in medication-na´ve schizophrenia patients was negatively correlated with the severity of psychotic symptoms (as measured by the PANSS) in these patients.
These results appear to indicate that abnormal dopamine system function in schizophrenia affects not only (or even primarily) the ability to anticipate rewarding outcomes, but rather the ability to modulate attention based on the salience of a cue. That is, these findings suggest that the certainty with which a cue predicts a valenced outcome is a stronger modulator of neural activity in ventral striatum than valence is. Additionally, the authors provide important evidence for the relationship between reward processing and psychosis: what are often blunted in psychotic individuals are signals associated with uncertain events. It will now be interesting to find out how antipsychotic medications affect salience-related activity, in dopaminergic nuclei and target areas, in patients never treated with antipsychotic drugs.
The findings are in accord with a number of noteworthy results in the field, such as evidence that the human VS plays a role in signaling the salience of an event, independent of its valence (Zink et al., 2003; Jensen et al., 2007; Guitart-Masip et al., 2010), and plays a particular role in coding for the action made in response to a cue, rather than the valence of the outcome predicted by the cue (Guitart-Masip et al., 2011). Furthermore, the findings of Nielsen et al. add to a body of evidence that abnormal DA system function linked to schizophrenia in general, and to psychotic symptoms in particular, is often associated with aberrant processing of unexpected events, such as prediction errors, regardless of their valence (Murray et al., 2008; Corlett et al., 2007). In general, these results should prompt investigators to consider properties of stimuli and actions beyond their expected value (such as certainty about their value) in examining the possible consequences of DA dysfunction in schizophrenia and other mental disorders.
As is to be expected, these intriguing results prompt a number of questions as well. For example, one would like to know if psychotic symptoms also correlated with behavioral measures (rather than just measures of neural activity). Specifically, did more psychotic patients show less RT differentiation between certain and uncertain cues? More generally, one is led to wonder about design differences between MID paradigms that appear to evoke robust cue-valence signals and those that do not. Whereas most MID paradigms involve variation in the magnitude of the potential gain or loss, Nielsen and colleagues held the magnitude of the potential gain (+7) and loss (-7) constant in their particular implementation (although they manipulated the expected value by varying the probability of the outcome).
In any case, Nielsen and colleagues are to be commended for their solid work in this area, which has potentially important implications.
Corlett PR, Murray GK, Honey GD, Aitken MR, Shanks DR, Robbins TW, Bullmore
ET, Dickinson A, Fletcher PC. Disrupted prediction-error signal in psychosis:
evidence for an associative account of delusions. Brain. 2007 Sep;130(Pt
Gradin VB, Kumar P, Waiter G, Ahearn T, Stickle C, Milders M, Reid I, Hall J,
Steele JD. Expected value and prediction error abnormalities in depression and
schizophrenia. Brain. 2011 Jun;134(Pt 6):1751-64. Abstract
Guitart-Masip M, Bunzeck N, Stephan KE, Dolan RJ, DŘzel E. Contextual novelty
changes reward representations in the striatum. J Neurosci. 2010 Feb
Guitart-Masip M, Fuentemilla L, Bach DR, Huys QJ, Dayan P, Dolan RJ, Duzel E.
Action dominates valence in anticipatory representations in the human striatum
and dopaminergic midbrain. J Neurosci. 2011 May 25;31(21):7867-75.
Jensen J, Smith AJ, Willeit M, Crawley AP, Mikulis DJ, Vitcu I, Kapur S.
Separate brain regions code for salience vs. valence during reward prediction in
humans. Hum Brain Mapp. 2007 Apr;28(4):294-302. Abstract
Juckel G, Schlagenhauf F, Koslowski M, Filonov D, WŘstenberg T, Villringer A,
Knutson B, Kienast T, Gallinat J, Wrase J, Heinz A. Dysfunction of ventral
striatal reward prediction in schizophrenic patients treated with typical, not
atypical, neuroleptics. Psychopharmacology (Berl). 2006a Aug;187(2):222-8. Abstract
Juckel G, Schlagenhauf F, Koslowski M, WŘstenberg T, Villringer A, Knutson B,
Wrase J, Heinz A. Dysfunction of ventral striatal reward prediction in
schizophrenia. Neuroimage. 2006b Jan 15;29(2):409-16. Abstract
Kapur S. Psychosis as a state of aberrant salience: a framework linking
biology, phenomenology, and pharmacology in schizophrenia. Am J Psychiatry. 2003
Jan;160(1):13-23. Review. Abstract
Kapur S, Mizrahi R, Li M. From dopamine to salience to psychosis--linking
biology, pharmacology and phenomenology of psychosis. Schizophr Res. 2005 Nov
1;79(1):59-68. Review. Abstract
Murray GK, Corlett PR, Clark L, Pessiglione M, Blackwell AD, Honey G, Jones
PB, Bullmore ET, Robbins TW, Fletcher PC. Substantia nigra/ventral tegmental
reward prediction error disruption in psychosis. Mol Psychiatry. 2008
Mar;13(3):239, 267-76. Abstract
Nielsen MO, Rostrup E, Wulff S, Bak N, Lublin H, Kapur S, Glenth°j B.
Alterations of the Brain Reward System in Antipsychotic Na´ve Schizophrenia
Patients. Biol Psychiatry. 2012 Mar 12. Abstract
Schlagenhauf F, Juckel G, Koslowski M, Kahnt T, Knutson B, Dembler T, Kienast
T, Gallinat J, Wrase J, Heinz A. Reward system activation in schizophrenic
patients switched from typical neuroleptics to olanzapine. Psychopharmacology
(Berl). 2008 Mar;196(4):673-84. Abstract
Simon JJ, Biller A, Walther S, Roesch-Ely D, Stippich C, Weisbrod M, Kaiser S.
Neural correlates of reward processing in schizophrenia--relationship to apathy
and depression. Schizophr Res. 2010 May;118(1-3):154-61. Abstract
Walter H, Kammerer H, Frasch K, Spitzer M, Abler B. Altered reward functions
in patients on atypical antipsychotic medication in line with the revised
dopamine hypothesis of schizophrenia. Psychopharmacology (Berl). 2009
Waltz JA, Schweitzer JB, Ross TJ, Kurup PK, Salmeron BJ, Rose EJ, Gold JM,
Stein EA. Abnormal responses to monetary outcomes in cortex, but not in the basal
ganglia, in schizophrenia. Neuropsychopharmacology. 2010 Nov;35(12):2427-39. Abstract
Zink CF, Pagnoni G, Martin ME, Dhamala M, Berns GS. Human striatal response to
salient nonrewarding stimuli. J Neurosci. 2003 Sep 3;23(22):8092-7.
View all comments by James A. Waltz
This new paper by Nielsen and colleagues can be considered a true landmark study on reward processing in patients with schizophrenia. It has long been suggested that dopamine dysregulation leads to a reduced response to rewarding or salient cues. A number of previous studies have used functional magnetic resonance imaging (fMRI) with the monetary incentive delay (MID) task to provide empirical evidence for this hypothesis. These studies have suggested that the BOLD response to rewarding cues is indeed reduced in unmedicated patients, but not in patients treated with atypical antipsychotics (Juckel et al., 2006; Schlagenhauf et al., 2008; Simon et al., 2010; Walter et al., 2009; Waltz et al., 2010). However, these studies had small sample sizes and could not account for possible effects of illness chronicity. These problems have now been conclusively resolved by Nielsen and colleagues, who unequivocally show that the ventral striatal response to salient cues is reduced in unmedicated, first-episode patients. Thus, they establish a fact about schizophrenia that will provide a landmark for future research.
One important issue for discussion that arises from the authors' findings is the relationship between the fMRI results and expression of psychopathology. The authors find a negative correlation between striatal response to salient cues and positive symptoms. In contrast, most previous studies using the MID task have suggested a relationship between reward anticipation and negative symptoms (Juckel et al., 2006; Simon et al., 2010; Waltz et al., 2010). While this can possibly be accounted for by the effects of medication and chronicity in at least some of the previous studies, these might not be the only explanations for the divergent findings. In this comment, we want to address two topics that might deserve continued attention in further research.
First, it is very difficult to disentangle the neurophysiological processes underlying the BOLD signal measured in response to salient cues. Recent results from animal research have shown that neighboring or even overlapping dopaminergic projection regions code for very different types of information such as value, uncertainty, and salience (Bromberg-Martin et al., 2010; Schultz, 2010). How these different types of coding relate to each other is still a matter of considerable debate. A huge problem for human research is the fact that fMRI does not provide enough temporal, and probably not even enough spatial, resolution to clearly disentangle these processes in analogy to the animal literature.
This renders careful task design even more important. Nielsen and colleagues take a big step in this direction by using a modification of the MID task to disentangle overall salience from motivational and value signals. The overall salience contrast yields the strongest findings with regard to group differences and relationship with positive symptoms. This contrast is a very complex integration of elements, including value (positive and negative), uncertainty, and behavioral relevance. In contrast, most studies showing a relationship with negative symptoms have focused on positive value coding, although behavioral relevance could not be clearly disentangled. Thus, one could hypothesize that overall salience is more strongly associated with positive symptoms, while positive value coding might be more strongly related to negative symptoms.
Second, when relating brain activation to psychopathology, it is a critical question of which rating scales to use. Some of the disparities between studies might also be related to different methods of psychopathological assessment. Future research might strongly benefit from some agreement on which scales to use to interpret brain activation patterns in relation to psychopathology. Nielsen and colleagues employ the PANSS, which is the most commonly used instrument for the assessment of schizophrenic psychopathology. There are now a number of factor analytic studies of the PANSS, which yield factor solutions that differ from the original distinction of positive, negative, and general psychopathology (Wallwork et al., 2012). In future imaging research, it might be helpful to rely on these factors, since they are also more likely to have common neurobiological correlates. Furthermore, in the domain of negative symptoms, it might be worthwhile to disentangle apathy/avolition from diminished expression, since the former is more likely to be related to dysfunctional reward processing (Foussias and Remington, 2010; Messinger et al., 2011).
Thus, the investigation of reward processing in relation to the symptoms of schizophrenia is likely to remain a challenging field in the future. Due to its unique sample and sophisticated design, the study by Nielsen and colleagues provides a strong reference point that will benefit all research in the field. It should also not be forgotten that Nielsen and colleagues are planning to address the effects of antipsychotic medication. We all look forward to reading their next paper, which will certainly be as valuable to the research community as their present work.
Bromberg-Martin, E.S., Matsumoto, M., and Hikosaka, O. (2010). Dopamine in motivational control: rewarding, aversive, and alerting. Neuron 68, 815-834. Abstract
Foussias, G., and Remington, G. (2010). Negative symptoms in schizophrenia: avolition and Occam's razor. Schizophrenia bulletin 36, 359-369. Abstract
Juckel, G., Schlagenhauf, F., Koslowski, M., Wustenberg, T., Villringer, A., Knutson, B., Wrase, J., and Heinz, A. (2006). Dysfunction of ventral striatal reward prediction in schizophrenia. NeuroImage 29, 409-416. Abstract
Messinger, J.W., Tremeau, F., Antonius, D., Mendelsohn, E., Prudent, V., Stanford, A.D., and Malaspina, D. (2011). Avolition and expressive deficits capture negative symptom phenomenology: implications for DSM-5 and schizophrenia research. Clinical psychology review 31, 161-168. Abstract
Schlagenhauf, F., Juckel, G., Koslowski, M., Kahnt, T., Knutson, B., Dembler, T., Kienast, T., Gallinat, J., Wrase, J., and Heinz, A. (2008). Reward system activation in schizophrenic patients switched from typical neuroleptics to olanzapine. Psychopharmacology 196, 673-684. Abstract
Schultz, W. (2010). Dopamine signals for reward value and risk: basic and recent data. Behav Brain Funct 6, 24. Abstract
Simon, J.J., Biller, A., Walther, S., Roesch-Ely, D., Stippich, C., Weisbrod, M., and Kaiser, S. (2010). Neural correlates of reward processing in schizophrenia--relationship to apathy and depression. Schizophrenia research 118, 154-161. Abstract
Wallwork, R.S., Fortgang, R., Hashimoto, R., Weinberger, D.R., and Dickinson, D. (in press). Searching for a consensus five-factor model of the Positive and Negative Syndrome Scale for schizophrenia. Schizophrenia research. Abstract
Walter, H., Kammerer, H., Frasch, K., Spitzer, M., and Abler, B. (2009). Altered reward functions in patients on atypical antipsychotic medication in line with the revised dopamine hypothesis of schizophrenia. Psychopharmacology 206, 121-132. Abstract
Waltz, J.A., Schweitzer, J.B., Ross, T.J., Kurup, P.K., Salmeron, B.J., Rose, E.J., Gold, J.M., and Stein, E.A. (2010). Abnormal responses to monetary outcomes in cortex, but not in the basal ganglia, in schizophrenia. Neuropsychopharmacology 35, 2427-2439. AbstractView all comments by Stefan KaiserView all comments by Joe J. Simon