This project will use dynamic causal modeling (DCM) to investigate neuronal connectivity underlying treatment response to dopaminergic medications in cocaine dependence. Since reduced dopamine (DA) function is likely to be involved in cocaine dependence, improved understanding of DA neuronal systems in cocaine dependence would facilitate the development of new medications to treat cocaine dependence. Previous studies have shown significant correlations between striatal DA and prefrontal cortical activity in cocaine dependent subjects (CDs). However, it is unclear whether the prefrontal cortex (PFC) causally affects the striatum or vice versa in CDs. This is relevant because the prefrontal-striata pathway may be involved in glutamatergic regulation of striatal DA release leading to low DA in CDs. The hypodopaminergic state in CDs may disrupt frontal inhibition, resulting in poor inhibitory control over compulsive drug intake. In addition, low DA in CDs may affect cognitive functions via striatum's gating of connections between PFC and other cortical regions. There is evidence that preexisting differences between CDs in DA circuits may underlie variability in responsiveness to treatments, and that for CDs who have a preexisting hypodopaminergic state, DA enhancement may be a useful target for pharmacotherapy. This project will use stochastic DCM, which is a recent DCM extension that takes into account hidden fluctuations in neuronal and vascular responses, and thus is especially suited for investigating effects of disease or drugs. In addition, this project will use nonlinear DCM, a DCM extension that can measure gating effects by striatum on cortico-cortical pathways. The overall aims of this project are: (1) To conduct functional magnetic resonance imaging-based DCM studies of working memory and impulsivity in order to determine the effective (directional) connectivity between PFC and striatum in treatment-seeking CDs compared to non-drug using controls. We hypothesize that DLPFC causally affects ventral striatum in CDs, and that the strength of this connection is lower in CDs compared to controls. (2) To determine whether the pretreatment gating effect by the dorsal striatum, as a reflection of pretreatment hypodopaminergic state associated with chronic compulsive drug use, predicts the treatment response to dopaminergic pharmacotherapy in CDs. We hypothesize that lower pretreatment gating by the dorsal striatum on prefrontal-parietal effective connectivity predicts greater 12-week improvement from treatment of CDs with DA enhancing medications (combined with cognitive behavioral therapy [CBT]), but not from treatment with placebo (combined with CBT). Establishing the directional relationship between the PFC and striatum and the role of striatal gating in cocaine treatment response will significantly improve our understanding of dopaminergic pharmacotherapeutic mechanisms in CDs and may ultimately lead to useful biomarkers for cocaine dependence and its treatment.
This study investigates the causal relationship between prefrontal cortex and striatal connectivity in cocaine dependence, and whether the gating effect of striatum on the prefrontal-parietal effective connectivity is predictive of the therapeutic response to dopaminergic pharmacotherapy in cocaine-dependent subjects. The evidence supporting the hypotheses will have significant implication in cocaine addiction treatment, and ultimately lead to potential clinically useful biomarkers for cocaine dependence, and thus improve public health through improved treatment for cocaine dependence.