Decreased GABA levels in anterior cingulate cortex/medial prefrontal cortex in panic disorder

Abstract

Changes of various brain metabolites including γ-aminobutyric acid (GABA), measured by 1H-magnetic resonance spectroscopy (MRS), have been reported in panic disorder (PD). Deficits in GABA have been implicated in the pathophysiology of PD. Furthermore, it has been suggested that cortical metabolite changes in PD are familial. Eleven PD patients, including five with and six without a PD family history, and eight age- and gender-matched healthy controls without a family history of psychopathology were recruited. Each subject underwent MRS exams and behavioral assessments (resting visual analog anxiety level and the Panic Disorder Severity Scale). GABA was detected with a MEGA-PRESS J-editing sequence and fitted to minimize macromolecule contaminations. A significant decrease in GABA, expressed as the ratio of GABA over total creatine (GABA/tCr), was detected in the anterior cingulate cortex (ACC)/medial prefrontal cortex (mPFC) in PD patients (p < 0.05), which tends to be pronounced in patients with a PD family history. No other patient/control differences in metabolites were noted in the ACC/mPFC or occipital cortex (OCC). Overall, our results indicate that deficits in GABA levels in PD patients vary by brain regions and possibly by family history status.

Introduction

Panic disorder (PD) is one of the most common psychiatric disorders with a lifetime prevalence of up to 5% in the US (Grant et al., 2006). Emerging evidence has implicated dysfunction in the main CNS inhibitory neurotransmitter system, the γ-aminobutyric acid (GABA) system, in the pathophysiology of PD. For example, abnormal reductions in GABA_A-benzodiazepine receptor binding were consistently found in various brain regions by positron emission tomography (PET) and single-photon emission computed tomography (SPECT) (Cameron et al., 2007, Hasler et al., 2008, Kaschka et al., 1995, Malizia et al., 1998, Schlegel et al., 1994), consistent with abnormal inhibitory neuromodulation in PD.

Magnetic resonance spectroscopy (MRS) provides a unique opportunity to noninvasively detect and quantify in vivo brain metabolites such as GABA in humans. In the conventional short-echo-time (short-TE) 1H spectrum, a series of metabolites can be directly measured. These include N-Acetyl-aspartate (NAA, a neuronal marker), creatine and phosphocreatine (often referred to as total creatine, tCr, an energy buffer and shuttle), myo-inositol (mI, a glial cell marker), choline (Cho, involved in cell membrane synthesis and degradation), glutamate (Glu, the major excitatory neurotransmitter) and the sum of glutamate and glutamine (Glx). The GABA signal within these spectra is usually covered by other larger, overlapping peaks due to its small concentration (about 1 mM in vivo) and multiplet structure. Hence special GABA editing techniques are needed in order to achieve signal identification and separation of the relatively small GABA peak.

Both short-TE and GABA-edited MRS have been used to study metabolite changes in PD in the last decade with mixed results. Goddard et al. first showed decreased occipital GABA in unmedicated PD patients using a GABA editing sequence, a decrease which was later suggested to be associated with family history (Goddard et al., 2001, Goddard et al., 2004). Ham et al. observed decreased GABA levels in the anterior cingulate cortex (ACC) and basal ganglia without using a GABA editing sequence. In addition, these authors found increased lactate and Cho in the ACC in medicated patients with PD (Ham et al., 2007). In contrast, no changes of prefrontal cortical GABA, Glx, Cho, and NAA were observed in unmedicated, non-depressed patients with PD by Hasler et al. (2009). Reduced tCr, NAA and Cho were found in the right medial temporal lobe of unmedicated PD patients (Massana et al., 2002). Yet another study reported reduced NAA/Cr and unchanged Cr and Cho/Cr in the left hippocampus in medicated PD patients (Trzesniak et al., 2010). Overall, the literature leads us to hypothesize that metabolite changes, especially GABA deficits, occur in certain brain regions in PD patients.

The ACC appears to regulate both cognitive processing and emotional processing relevant to anxiety (Bush et al., 2000). The occipital cortex and posterior cingulate cortex have been implicated in visual/spatial processing and response preparation to a physical threat (Vogt et al., 1992), functions which could be relevant to anticipatory anxiety in PD. We therefore studied GABA levels, both in the ACC and occipital cortex (OCC), using the MEGA-PRESS J-editing technique (Edden and Barker, 2007, Mescher et al., 1998).

PD is thought to be familial, with genetic factors accounting for 30–40% of the variation in causation (Finn and Smoller, 2001, Smoller et al., 2008, Weissman, 1993, Weissman and Merikangas, 1986). Normal subjects with a family history of anxiety disorder in their first-degree relatives have been found to be more vulnerable to lactate-induced panic attacks than those without a family history of anxiety disorder (Balon et al., 1989). Therefore, we also hypothesize that cortical metabolite changes, especially GABA deficits, depend on family history and may be more profound in PD patients with a family history of PD.