Reduced functional connectivity between ventromedial prefrontal cortex and insula relates to longer corrected QT interval in HIV+ and HIV− individuals

Highlights

• A seed-based resting state functional connectivity (rsFC) of the ventromedial prefrontal cortex (VMPFC) was compared to QTc length.

• Increased rsFC between VMPFC and anterior insula was linked to shorter QTc intervals.

• Greater frontal-striatal connectivity was linked to higher CD4 and shorter QTc length in HIV+ patients.

Abstract

Objective

Prolongation of the QT interval, i.e., measure of the time between the start of the Q wave and the end of the T wave, is a precursor to fatal cardiac arrhythmias commonly observed in individuals infected with the Human Immunodeficiency Virus (HIV), and is related to dysregulation of the autonomic nervous system. We investigated the relationship between QT interval length and resting state functional connectivity (rsFC) of the ventromedial prefrontal cortex (VMPFC), a core region of the brain that is involved with cardio-autonomic regulation.

Method

Eighteen HIV+ men on antiretroviral therapy and with no history of heart disease were compared with 26 HIV-negative control subjects who had similar demographic and cardio-metabolic characteristics. A seed-based rsFC analysis of the right and left VMPFC was performed at the individual subject level, and 2nd-level analyses were conducted to identify the following: group differences in connectivity, brain regions correlating with corrected (QTc) interval length before and after controlling for those group differences, and regions where seed-based rsFC correlates with CD4 count and QTc interval within HIV+ individuals.

Results

HIV-negative adults showed greater rsFC between the VMPFC seed regions and several default mode network structures. Across groups greater rsFC with the left anterior insula was associated with shorter QTc intervals, whereas right posterior insula connectivity with the VMPFC correlated with greater QTc intervals. HIV patients with lower CD4 counts and higher QTc intervals showed greater rsFC between the right VMPFC and the right posterior insula and dorsal cingulate gyrus.

Conclusions

This study demonstrates that QTc interval lengths are associated with distinct patterns of VMPFC rsFC with posterior and anterior insula. In HIV patients, longer QTc interval and lower CD4 count corresponded to weaker VMPFC connectivity with the dorsal striatrum.

Significance

A forebrain control mechanism may be implicated in the suppression of cardiovagal influence that confers risk for ventricular arrhythmias and sudden cardiac death in HIV+ individuals.

Introduction

Despite the widespread availability of antiretroviral therapy, mortality from cardiovascular disease (CVD) has increased for persons infected with human immunodeficiency virus (HIV) (Enakpene et al., 2015). A large proportion of this CVD comorbidity has been attributed to sudden cardiac death (Tseng et al., 2012). The QT interval reflects the latency between the initiation of the QRS-complex and end of the T-wave and is used as a general marker of ventricular repolarization (Drew et al., 2010). Clinical presentation of ventricular arrhythmias (i.e., torsade de pointes) is partially attributed to prolongation of the QT interval (Viskin, 1999). The corrected QT interval (QTc) length is used clinically for predicting the risk of fatal arrhythmias and CVD mortality (Schouten et al., 1991). Increased sympathetic nervous activity and decreased parasympathetic activity may be linked to the clinical presentation of QT interval prolongation. Decades of research show that adrenergic stimulation of the left and right stellate ganglion causes prolongation followed by transient shortening of the QT interval (Abildskov, 1976, Zipes, 1991). This adrenergic influence is arrhythmogenic and can result in QT lengthening by contributing to the abnormal repolarization currents commensurate with ventricular arrhythmias (Jackman et al., 1988, Zipes, 1991, Shen and Zipes, 2014).

Evidence of significantly longer QT intervals in HIV patients, when compared to uninfected individuals, precedes the advent of combination antiretroviral regimens (Cohen et al., 1991, Shahmanesh et al., 1991, Villa et al., 1995, Kocheril et al., 1997, Enakpene et al., 2015). Current literature The syndrome continues to be present despite greater effectiveness of newer antiretroviral classes (Villa et al., 1995, Ogunmola et al., 2015). Prevalence of QT prolongation is estimated to be 3 to 5 times higher in HIV and AIDS patients than in the HIV-negative population (Sani and Okeahialam, 2005). Moreover, regimens that include protease inhibitors may predispose HIV-infected individuals to QT interval lengthening due to cardiometabolic stress (Sani and Okeahialam, 2005, Chinello et al., 2007, Charbit et al., 2009, Fiorentini et al., 2012, Shavadia et al., 2012, Ogunmola et al., 2015). The persistence of cardiac manifestations associated with QT interval lengthening has therefore become an emergent concern in HIV-CVD treatment as prolongation of this interval is present across the age continuum and at nearly every stage of infection (Wongcharoen et al., 2014, McIntosh et al., 2017).

Cortical and subcortical brain regions that participate in modulation of heart rhythm exhibit the same susceptibility to the deleterious effects of aging and neurodegenerative disease as cognitive or emotional brain regions (Goldstein, 2003, Kaufmann and Biaggioni, 2003, Kaufmann et al., 2004). The central autonomic network (CAN) involves structures that regulate heart rate and blood pressure; e.g., the medial prefrontal cortex (MPFC), insula, anterior cingulate cortex (ACC), hypothalamus, periaqueductal gray, and brainstem regions including the nucleus of the solitary tract (NTS) and the rostral and caudal ventrolateral medulla (Matthews et al., 2004, Wong et al., 2007, Thayer and Lane, 2009). The ventral region of the medial prefrontal cortex (VMPFC) is of particular interest given its role in top-down coordination of sympathetic and parasympathetic response to autonomic maneuvers (Beissner et al., 2013). A neurovisceral integration model describes a network of structures including the anterior and posterior insula, dorsal and subgenual cingulate, and amygdala that are functionally connected to the ventral and orbitofrontal cortex and are thought to regulate adjustment of heart rate and blood pressure to psychological and physiological challenge (Thayer and Lane, 2000, Thayer and Siegle, 2002). Whether VMPFC activity leads or lags behind alterations in cardioautonomic activity has been debated. Granger causality analysis has been used to show that elevations in VMPFC activity precede decreases in skin conductance, a marker of sympathetic arousal (Zhang et al., 2013). Neuronal tract-tracing studies in animals confirm inhibitory and excitatory connections from VMPFC to cortical areas (e.g., insula) and to subcortical cardiorespiratory structures (Verberne and Owens, 1998). Parasympathetic and sympathetic effects of the VMPFC have been lateralized to left and right hemispheres, respectively, in both human (Wittling et al., 1998, Hilz et al., 2001) and animal (Terreberry and Neafsey, 1983, Terreberry and Neafsey, 1987, Neafsey et al., 1986) studies. Central control of cardioautonomic function may be of particular relevance to HIV+ individuals as thinning of bilateral medial frontal and insular brain regions has been reported in this population (Thompson et al., 2005, Kallianpur et al., 2012).

Resting-state functional connectivity (rsFC) is a technique that assesses the brain’s intrinsic function by correlating the time course of spontaneous blood-oxygen-level-dependent (BOLD) activity in a particular seed region of interest (ROI) to signals from other areas. RsFC is a technique well-suited for exploration of functional neural networks (Biswal et al., 1995) and is useful for categorizing patient populations by neural phenotypes. This measure of intrinsic BOLD signal fluctuations is thought to index a property of functional brain organization indicative of behavioral or physiological characteristics (Hahn et al., 2011, Kim et al., 2011, Zhu et al., 2012, Baur et al., 2013). Resting state connectivity of the VMPFC has also been implicated in cardio-autonomic regulation. For example, greater resting-state functional connectivity of the left VMPFC has been related to higher vagal tone indexed by heart rate variability in the temporal domain (Ziegler et al., 2009). There are several reasons why lower rsFC of the VMPFC might relate to QTc interval length in persons with HIV. Compared to healthy controls, individuals infected with HIV show deficits in intrinsic brain connectivity across the HIV disease spectrum (Thomas et al., 2013, Herting et al., 2015). Moreover, marked HIV-related reductions in intra-network connectivity has been reported for the default mode network (DMN), of which the VMPFC is a major hub, as well as reduced inter-network connectivity with the salience network which is anchored by the anterior insula (Wang et al., 2011, Thomas et al., 2013). If these structures are indeed involved in the regulation of sympathetic and parasympathetic tone then it is conceivable QTc interval length might reflect their functional connectivity. In addition, prolongation of the QTc interval in HIV+ individuals has been most consistently linked to cardio-metabolic disease comorbidity (Reinsch et al., 2009), antiretroviral therapy regimen (Chinello et al., 2007, Charbit et al., 2009, Shavadia et al., 2012), and CD4 decline (QaQa et al., 2010, Ige et al., 2014, Wongcharoen et al., 2014, Gaharwar et al., 2017). The aim of the current study was to determine the brain regions whose rsFC with the VMPFC corresponds to QTc interval length; whether these regions differ between HIV patients on stable antiretroviral therapy (ART) and HIV-negative comparison subjects; and whether these rsFC patterns vary as a function of CD4 count in the HIV+ patients. We hypothesized that (1) greater rsFC of the VMPFC with CAN structures, such as the anterior insula, cingulate cortex, and brainstem medulla, would correspond with shorter QTc interval length; (2) VMPFC connectivity with the CAN, as a function of QTc length, would be lower in HIV patients compared to healthy controls; and (3) connectivity between the VMPFC and CAN structures would be lower in HIV patients who had greater QTc interval length and lower CD4 count.