Pregabalin effects on neural response to emotional faces

Abstract

Pregabalin has shown promise in the treatment of anxiety disorders. Previous functional magnetic resonance imaging (fMRI) studies indicate agents used to treat anxiety, e.g., SSRIs and benzodiazepines, attenuate amygdala, insula, and medial prefrontal cortex (mPFC) activation during emotional processing. Our prior study has shown that during anticipation of an emotional stimulus, pregabalin attenuates amygdala and insula activation but increases medial PFC activation. In this study, we examined whether, similar to SSRIs and benzodiazepines, pregabalin attenuates amygdala, insula, and medial PFC during emotional face processing. Sixteen healthy volunteers underwent a double-blind within-subjects fMRI study investigating effects of placebo, 50 mg, and 200 mg pregabalin on neural activation during an emotional face-matching task. Linear mixed model analysis revealed that pregabalin dose-dependently attenuated left amygdala activation during fearful face-matching and left anterior insula activation during angry face-matching. The 50 mg dose exhibited more robust effects than the 200 mg dose in the right anterior insula and ventral ACC. Thus, pregabalin shares some similarity to SSRIs and benzodiazepines in attenuating anger and fear-related insula and amygdala activation during emotional face processing. However, there is evidence that a subclinical 50 mg dose of pregabalin produced more robust and widespread effects on neural responses in this paradigm than the more clinically relevant 200 mg dose. Taken together, pregabalin has a slightly different effect on brain activation as it relates to anticipation and emotional face processing, which may account for its unique characteristic as an agent for the treatment of anxiety disorders.

Keywords: anxiety; anxiolytic; benzodiazepine; emotion; fMRI; pharmaco-imaging; pregabalin.

Figure 1

fMRI emotional face matching task. This task was adapted from the block-design task described by Hariri et al. (2002) to an event-related design using the emotional face images from Tottenham et al. (2009) and Matsumoto and Ekman (1988). This task includes conditions requiring the subject to match emotional faces or shapes. For the faces conditions, a target face (on the top of the computer screen) and two probe faces (on the bottom of the screen) are presented, and subjects are to match the probe with the same emotional expression to the target by pressing the corresponding button. For the faces conditions, the subjects are matching an equal number of angry, fearful, and happy emotional faces. During the control shape condition, subjects are to match the target shape to one of two differently oriented probe shapes.

Figure 2

Regions of interest masks used for fMRI analyses, shown from axial views.(A) bilateral amygdala (green) and fusiform gyrus (yellow) (shown at z = −20), (B) bilateral anterior (green) and posterior (yellow) insula (shown at z = 2), (C) ventral medial prefrontal cortex (PFC; shown at z = −6), and (D) dorsal medial PFC (shown at z = 4). These anatomical regions were defined using Talairach Atlas [Lancaster et al. (2000)].

Figure 3

Pregabalin dose effect on reaction time during fMRI task. Pregabalin was associated with slower reaction time across face- and shape-matching conditions [F(1, 171) = 14.25, p < 0.001; (200 = 50) > 0]. There was also a significant difference in reaction time across conditions, with responses being slowest when matching negative emotional faces than happy faces and slower when matching faces than shapes [F(2, 171) = 183.11, p < 0.001]. There was no significant dose by condition effect on reaction time [F(2, 171) = 0.422, p = 0.656].

Figure 4

Task valence effect on fMRI BOLD activation. Regions within bilateral fusiform gyrus (A) and bilateral amygdala (B) showed greater activation for both fearful and angry faces compared to happy. Bilateral anterior insula also showed greater activation for angry compared to happy faces (D2) but showed greater activation to happy compared to fearful faces (D1). Regions of the mPFC exhibited greater activation for happy compared to both angry and fearful faces (C). Graphs depicted below each image represent average PSC for negative (fearful or angry) and happy conditions.

Figure 5

Pregabalin effect on fMRI BOLD activation during processing of fear and angry emotional faces. For fear faces, pregabalin was associated with a consistent decrease in left amygdala [0 > (50 = 200); identified via region of interest (ROI) analyses] activation while 50 mg pregabalin increased ventral ACC activation (identified via whole-brain analyses). For angry faces, pregabalin was associated with a consistent decrease in a region of the left anterior insula/inferior frontal gyrus [B2; 0 > (50 = 200); identified via whole-brain analyses] while 50 mg pregabalin decreased right anterior insula activation [B1; (0 = 200) > 50; identified via ROI analyses] and increased ventral ACC activation [B3; 50 > (0 = 200); identified via ROI analyses]. Graphs depicted with each image represent average PSC for negative emotional faces for each dose (placebo, 50 mg, and 200 mg pregabalin).