Biobehavioral correlates of an fMRI index of striatal tissue iron in depressed patients

Abstract

Dopaminergic function is a critical transdiagnostic neurophysiological dimension with broad relevance in psychiatry. Normalized T2*-weighted (nT2*w) imaging has been previously investigated as a method to quantify biological properties of tissue in the striatum (e.g., tissue iron), providing a widely available, in vivo marker with potential relevance to dopaminergic function; but no prior study to our knowledge has examined this neuroimaging marker in clinical depression. In a treatment-seeking, clinically depressed sample (n = 110), we quantified tissue iron (nT2*w) in striatal regions. We assessed test-retest reliability and correlated values with dimensional features across levels of analysis, including demographic/biological (sex, age, Body Mass Index), neuroanatomical (hippocampal atrophy, which was quantified using a recently validated machine-learning algorithm), and performance-based (Affective Go/NoGo task performance) indices with relevance to depressive neurocognition. Across patients, decreased tissue iron concentration (as indexed by higher nT2*w) in striatal regions correlated with indices of decreased cognitive-affective function on the Affective Go/NoGo task. Greater caudate nT2*w also correlated with greater hippocampal atrophy. Striatal tissue iron concentrations were robustly lower in female patients than males but gender differences did not explain relations with other neurocognitive variables. A widely available fMRI index of striatal tissue properties, which exhibited strong psychometric properties and can be readily quantified from most fMRI datasets irrespective of study-specific features such as task design, showed relevance to multiple biobehavioral markers of pathophysiology in the context of moderate-to-severe, treatment-resistant depression. Striatal tissue iron may play a role in dimensional and subgroup-specific features of depression, with implications for future research on depression heterogeneity.

Contributions of dopamine-related basal ganglia neurophysiology to the developmental effects of incentives on inhibitory control

“…nT2 *w values were extracted separately across each basal ganglia region of interest (ROI; Fig. 4 A), and are consistent with distributions and developmental trajectories found in prior studies of tissue iron ( Larsen et al, 2020a , Peterson et al, 2019 ; Price et al, 2021 ) using more direct measures obtained using R2’ ( Larsen et al, 2020b , Parr et al, 2021 ), R2* ( Larsen et al, 2020a ) and quantitative susceptibility mapping (QSM; Haacke et al, 2010 , Haacke et al, 2005 ; Peterson et al, 2019 ). nT2*w values were extracted separately across each ROI in order to examine the specificity the relationship with AS performance in individual subregions.…”

“…However, the relationship between the BOLD response and DA physiology remains unclear ( Attwell and Iadecola, 2002 , Brocka et al, 2018 , Logothetis, 2003 , Logothetis and Wandell, 2004 ). Tissue iron has influence on the T2* signal, and has been quantified using a variety of relaxometry-based MR measures including R2* ( Haacke et al, 2010 , Larsen et al, 2020a ), R2’ ( Larsen et al, 2020b , Sedlacik et al, 2014 ), susceptibility-based approaches ( Haacke et al, 2004 , Peterson et al, 2019 ), and nT2*w imaging ( Larsen and Luna, 2015 , Peterson et al, 2019 , Price et al, 2021 ). nT2*w imaging in particular represents a promising, non-invasive, indirect measure of basal ganglia DA-related neurophysiology, as it can be readily obtained using existing T2* -weighted echo planar imaging (EPI) scans acquired during most fMRI protocols.…”

“…Averaging across time enhances the signal to noise ratio ( Larsen and Luna, 2015 ), while the normalization step gives the T2* decay in the basal ganglia relative to the whole brain, and allows for comparison of nT2*w values across participants. Here, we averaged across all functional and resting state runs, in line with prior research showing excellent reliability of this measure within sessions (ICCs ranging from.93 to.91 for within-session estimates of nT2*w values within basal ganglia ROIs; Price et al, 2021 ).…”

“…Distinct from typical fMRI BOLD studies that are interested in fluctuations in the T2* -weighted signal that change across time, here we were interested in the time-invariant aspects of the T2* -weighted signal that have been shown to reflect tissue-iron properties ( Larsen and Luna, 2015 ). Preprocessing procedures for T2* -weighted images have been described in ( Larsen and Luna, 2015 , Peterson et al, 2019 , Price et al, 2021 , Vo et al, 2011 ) and include the following steps. First, each volume was normalized to the whole-brain mean (z-score normalization Larsen and Luna, 2015 ; Peterson et al, 2019 ), using a coverage map created from all inputs for each participant that included only non-zero values.…”

“…In order to determine the role of basal ganglia DA-related neurobiology in developmental changes in the modulation of cognitive control by exogenous rewards, we designed a longitudinal developmental neuroimaging study that characterized age-related changes in AS performance with in both rewarded and neutral contexts (n = 193 sessions, 12–33 years of age, 1–3 visits). We assessed the role of basal ganglia neurophysiology using longitudinal MRI-based assessments of tissue iron obtained via normalizing and time-averaging T2* -weighted images of task and resting state scans (nT2*w; n = 177 sessions, 12–33 years of age, 1–3 visits), which quantifies the relative T2* relaxation across the brain and is sensitive to magnetic field inhomogeneities induced by iron ( Brown et al, 2014 , Langkammer et al, 2010 , Larsen and Luna, 2015 , Peterson et al, 2019 , Price et al, 2021 , Schenck and Zimmerman, 2004 ). Given recent findings showing a specific role for NAcc tissue iron in the development of frontostriatal networks ( Parr et al, 2021 ) and putamen tissue iron in the development of general cognitive performance ( Larsen et al, 2020a ), we explored the contribution of nT2*w extracted separately across basal ganglia subregions.…”