| |
HIV infection [slows brain activity] results in ventral-striatal reward system hypo-activation during cue processing
|
| |
| |
Download PDF here
AIDS 2015
"HIV caused a decrease in activity during cue processing in the ventral striatum, with normal cortical functioning during reward outcome processing. Our results therefore suggest that HIV not only has an impact on fronto-striatal systems involved in executive functioning, but also has a direct impact on the function of the ventral-striatal reward system.
...In healthy controls, there is an increase in the ventral-striatal activity during the presentation of reward cues relative to the activity present during neutral cues [14,18]. HIV infection has been associated with hypoactive functional striatal responses, as well as a decreased baseline cerebral blood flow on past neuroimaging studies [21,22, unpublished data].. We therefore predicted that HIV infection is more likely to result in a general decrease in activity in the ventral striatum during reward neutral and reward cues, rather than a specific effect on any given cue.
HIV enters the brain in the early stage of infection [1], and even in the era of combined antiretroviral therapy (cART), HIV causes neurocognitive impairment in up to half of the individuals [2]. Functional imaging studies have demonstrated that HIV affects the functioning of the fronto-striatal network, a key system involved in executive functioning [3,4]. The effect of HIV on the fronto-striatal network could, at least in part, be due to high concentrations of viral RNA accumulating in the striatum and the surrounding areas early on during infection [5]. More specifically, striatal dopamine neurons may be particularly sensitive to viral effects [6,7]. This could either be caused directly by toxic effects of viral proteins such as GP120 [8] or tat [9], or indirectly by means of the de-regulation of the host immune system [10].
As striatal function plays an important role in reward processing [11-14], HIV could potentially interfere with the function of the ventral-striatal reward system. Behavioural evidence thus far has indicated that HIV may affect reward processing [15,16]. For example, HIV-positive (HIV+) substance-dependent individuals exhibit greater risky decision-making behaviour on a gambling task [16]. HIV+ participants favoured relatively larger payoffs, which incurred infrequent large penalties. While controls also selected these payoffs, they quickly learned to avoid them. HIV infection was thought to play a role in this behavioural deficit, as although half of the participants were classified as having past drug dependence, few met criteria for current drug dependence. Furthermore, controlling for these factors did not affect the final results. Although such behavioural studies provide evidence for an affect of HIV on reward processing, the effect of HIV in the absence of drug abuse on the function of the ventral-striatal reward system remains to be demonstrated."
ABSTRACT
Objective: Functional MRI has thus far demonstrated that HIV has an impact on frontal-striatal systems involved in executive functioning. The potential impact of HIV on frontal-striatal systems involved in reward processing has yet to be examined by functional MRI. This study therefore aims to investigate the effects of HIV infection on reward processing by examining the function of the ventral-striatal reward system during a monetary incentive delay task.
Design: This is a cross-sectional case-control study.
Methods: Eighteen combined antiretroviral therapy-naive HIV-positive (HIV+) participants, as well as 16 matched healthy controls, performed a monetary incentive delay task. This paradigm assesses behaviour as well as functional brain activity-associated reward anticipation and reward outcome.
Results: HIV+ participants showed a general decrease in activation associated with both neutral as well as potentially rewarding cues in their ventral striatum. We found normal activity related to reward outcome in the orbito-frontal cortex. Despite HIV+ participants' reaction times being significantly slower when independently measured from the reward paradigm, this performance deficit normalized during the performance of the reward task.
Conclusion: HIV caused a decrease in activity during cue processing in the ventral striatum, with normal cortical functioning during reward outcome processing. Our results therefore suggest that HIV not only has an impact on fronto-striatal systems involved in executive functioning, but also has a direct impact on the function of the ventral-striatal reward system.
Discussion
Here, we investigated brain activity during reward processing in 18 HIV+ substance and cART-naive participants and 16 matched controls using fMRI. To our knowledge, this is the first study to investigate the effects of HIV on fronto-striatal functioning during reward processing. Importantly, both groups were substance and cART-naive. As predicted, HIV+ participants showed a general decrease in ventral-striatal activity during anticipation for both neutral and potentially rewarding trials, when compared to controls. The increase in activity related to reward outcome in the OFC did not differ between the two groups. Despite HIV+ participants' reaction times being significantly slower at baseline, this performance deficit normalized during the performance of the reward task. These findings suggest that HIV has an impact on the function of the ventral striatum during cue processing, with a relative sparing of the cortical function. Striatal dysfunction during cue processing was present despite HIV+ participants still being able to speed up their responses in anticipation of a potential reward, indicating that the impact of HIV is not limited to striatally mediated executive function, but extends to reward processing.
We found a decrease in ventral-striatal activity during cue processing in HIV+ participants. It is well known that HIV's impact on brain function often results in a clinical fronto-striatal dementia [37]. Past functional studies have shown general striatal dysfunction, with hypo-activity in the putamen during reactive inhibition [unpublished data], caudate hypo-activity during semantic event sequencing [22], as well as a general decrease in resting cerebral blood flow in the striatum on arterial spin labeling [21]. Our data extend these findings by showing an HIV-related decrease in activity in the ventral-striatal reward system during cue processing. Given the limitations of BOLD fMRI [38], a potential explanation for a general decrease in regional BOLD signal is that the consequences of HIV infection, such as neuro-inflammation [10], does not affect neural activity per se, but rather causes a general change in haemodynamics in the striatum [39]. This is unlikely to be the case, given that we find no decrease in the ventral-striatal activity during reward feedback in HIV. We therefore have shown that HIV does not only affect striatal functions associated with executive functioning such as the inhibition of voluntary movement [unpublished data] or semantic event sequencing [22], but also reward processing. Previous studies in the behavioural impact of HIV have been potentially confounded by past drug dependence [16]. Our findings further support behavioural studies reporting abnormal behaviour on gambling tasks [16] in HIV infection in the absence of illicit drug use/abuse, as cue processing is vital in predicting future rewards. For example, hypo-activation during ventral-striatal anticipatory activity has been associated with impulsive decision-making in alcohol dependence, as well as pathological gambling [40,41].
Our data show that HIV does not result in a specific deficit during the anticipation of reward cues, but also during activity related to neutral cues. Nevertheless, it is likely that a disruption in general cue processing could potentially result in abnormal reward-based decision-making. This could include risky decision-making with respect to sexual behaviour, aggression, substance abuse and potentially a decreased capacity to appreciate the long-term benefits of using and staying on treatment. Importantly, as our sample has no reported drug use, abnormal reward-related behavior in the HIV+ population is unlikely to be explained by the effects of drug abuse alone [15,42].
We found normal activity in the cortical regions known to be active during reward outcome [17]. This finding is consistent with our previous work which suggested that sub-cortical function is primarily affected in the present sample population [unpublished data]. This negative finding is seemingly at odds with other functional studies, which generally demonstrate HIV-induced hyper-activation in the cortex during working memory, as well as during visual attention tasks [4,43-45]. A possible explanation for why some studies found cortical dysfunction when we have not is that these studies differ from our present sample in terms of patient age and cART use. The additional effect of cART, as well as the effect of aging, has been associated with increases in cortical activation in HIV infection in their own right [46,47]. More importantly, participants from our sample population were all newly diagnosed, just prior to the initiation of treatment. This would suggest that HIV could have a different impact on cortical function after treatment [46]. This will have to be confirmed with further prospective studies utilizing fMRI tasks that reliably engage the cortex and the striatum.
Both controls and HIV+ participants showed significantly faster response times as well as increased response accuracy between neutral trials and potentially rewarding trials. This is in keeping with behavioural responses previously reported in healthy control populations [18]. This indicates task comprehension in both the groups. Furthermore, the HIV+ participants are still able to anticipate potentially rewarding trials. The fact that HIV+ participants did not show generally slower responses than controls is surprising, given that we did indeed find slower simple response times acquired independent of the fMRI task. Although it is well known that HIV infection is associated with reaction time slowing [48], here we show that this response slowing could normalize during situations with a positive motivational valence. This implicates at least, in part, deficient reward processing in response time slowing in HIV.
As we have used a relatively simple reward task that utilizes only one level of reward, we cannot rule out the possibility that HIV+ participants will start showing slower behavioural responses with increased task complexity with multiple levels of reward, as well as the inclusion of punishment trials. We chose to utilize a simpler task to ensure that reward anticipation would not be influenced by a lack of task comprehension. As the present task was successfully utilized in children as young as 10 years, who demonstrated no measurable differences in response accuracy from those of adults, we believe that the task was simple enough not to confound performance in the present study [18]. As our task involved on an average a 50% failure rate, factors such as participant frustration could potentially have differed between the groups. Task accuracy was, however, not significantly different in the present sample, and therefore differences in levels of group frustration should have been minimal.
As our sample largely consisted of female participants, special mention should be made on the possible influence of sex on our findings. It has been postulated that women are differentially at risk for the development of HAND, due to potentially different prevalence of risk factors such as poverty, differences in substance abuse and prevalence of mental health disorders [49-52]. Studies performed thus far have indeed found a small HIV serostatus-by-sex interaction on cognition, with HIV-related comorbidities having a much larger impact [50,52]. The only functional study performed thus far in a largely female cohort did indeed report hippocampal dysfunction in women, but could potentially have been confounded by substance abuse [51]. In the present study, we confirm an impact of HIV on brain function in HIV+ women, and extend these results by demonstrating a functional impact of HIV in the absence of a history of drug use, severe comorbid psychopathology (i.e. major depression), as well as differences in demographic variables, which we controlled for with a strict sample selection.
HIV caused a decrease in activity during cue processing in the ventral striatum, with normal cortical functioning during reward outcome processing. Our results therefore suggest that HIV not only has an impact on the fronto-striatal systems involved in executive functioning, but also has a direct impact on the function of the ventral-striatal reward system.
|
|
| |
| |
|
|
|
