The absence of an association between ROI activation

The absence of an association between ROI activation and treatment outcome during neutral trials suggests the importance of motivating (by reward cue) optimal performance on a relatively challenging cognitive control task, such as antisaccade. Prior work with the reward cue AS task indicated that SUD youth do not generally lack inhibitory control, and that motivational cues (e.g., monetary incentive) can facilitate activation of regions involved in inhibitory control (Chung et al., 2011). For example, right vlPFC activation has been associated with inhibitory control (Aron and Poldrack, 2006), and specifically, inhibition of prepotent oculomotor responses (Massen, 2004). Results in this sample of treated youth indicating that greater activation of right vlPFC sphere during reward (relative to neutral) trials was associated with lower marijuana problem severity at follow-up suggest a potential role for reward in optimizing or motivating activation, particularly in this ROI, which in turn, was associated with treatment outcome. This study\'s finding that greater activation in regions associated with cognitive control (e.g., vlPFC) was associated with better marijuana outcomes in treated adolescents is similar to results obtained using an fMRI Stroop task in cocaine (Brewer et al., 2008) and cannabis dependent adults (Kober et al., 2014). The current study extends prior work with treatment seeking adults by using an AS task as a different assay of cognitive control, examining adolescents in community-based substance use treatment, and investigating the interplay of reward response and cognitive control in one task. An important difference between the prior work with adults and the current study of adolescents is that regional activation in cognitive control regions during reward, but not neutral, trials were associated with adolescent treatment outcomes. The finding in the current study that youth with more marijuana symptoms at follow-up tended to show deactivation in ROIs that were associated with outcome may reflect, for example, decreased sensitivity to reward or less efficient brompheniramine maleate functioning during the reward condition. In addition, prior work with treatment seeking adults examined brain response prior to treatment, whereas in the current study of treated adolescents, brain response was assessed during or shortly after treatment completion. Nevertheless, research with both treated adults and this pilot study of treated adolescents suggests that regional activation associated with cognitive control is associated with subsequent substance use outcomes. Notably, the reward cue antisaccade task leverages an individual\'s sensitivity to reward to motivate or optimize cognitive control. Consistent with proposed brain-based mechanisms by which incentives might increase correct AS (Geier et al., 2010; Geier, 2013), greater activation during reward trials in regions involved in reward processing (e.g., amygdala, nAcc), oculomotor control (e.g., putamen, supplementary eye field sphere), and executive functioning (e.g., vlPFC) were associated with fewer marijuana symptoms at 6-month follow-up in treated adolescents. The association of activation during reward condition in nAcc, but not OFC, with marijuana outcome also is consistent with reliance on bottom-up processing of reward during adolescence, in the context of a relatively immature OFC (Geier, 2013). Adolescent sensitivity to reward cues, may bias action to immediate and salient rewards, such as substance use. However, the use of incentive to support cognitive control over behavior, for example, in contingency management interventions (Stanger and Budney, 2010), capitalizes on adolescent sensitivity to reward to promote positive behavior change and effortful control over behavior. As a comparison to findings with treated adolescents in the current study, research on reward response with cocaine dependent adults in treatment correlated regional activation during reward processing using an fMRI monetary incentive delay task (MIDT) with treatment outcome, and found, in general, that reduced activity in regions involved in reward processing was associated with more positive response to behavioral therapy (Jia et al., 2011). In contrast, results from the current study suggest that youth with more marijuana symptoms at follow-up showed deactivation during the reward condition in ROIs associated with treatment outcome. Differences in results across these two studies may be due, in part, to differences in the role of reward in the two tasks (i.e., use of reward to enhance cognitive control versus response to monetary reward), differences in sample age (and brain development in adolescents versus adults), primary substance of abuse, and severity and duration of substance use. Importantly, with the rewarded antisaccade task, we were able to assess sensitivity to reward, inhibitory control, and their interplay (i.e., the use of reward to optimize inhibitory control) in a single task.