Depressive disorders are among the leading causes of disability and global health burden (GBD 2019 Diseases and Injuries Collaborators, 2020), with depression affecting 280 million individuals worldwide (World Health Organization, 2023). Major depressive disorder (MDD) poses a significant challenge for affected individuals in terms of its social, occupational, and economic impacts (Greenberg et al., 2003; Murray & Lopez, 1997). Due to the heterogeneity in both the clinical presentation of the disorder (Drysdale et al., 2017) and how affected individuals respond to treatment (Fava & Davidson, 1996), efforts have been directed toward elucidating the neural underpinnings behind MDD etiology, symptoms, progression, and antidepressant response.

Neuroimaging studies suggest that MDD is associated with altered brain activity and functional connectivity (FC), reflecting disrupted intra-brain communication (Kaiser et al., 2015; Menon, 2011). For instance, the affective network (AN) displays altered intra- and inter-network dynamics in MDD (Demchenko et al., 2022; Dunlop et al., 2019; Wang et al., 2012). One of the neural circuits with the most consistently identified functional abnormalities in MDD is the serotonergic emotion regulation circuitry, centred on the amygdala and regions of the medial prefrontal cortex (PFC) (Phillips et al., 2015). Given its involvement in emotion regulation (Phillips et al., 2015), the amygdala has been consistently reported as one of the principal cortico-limbic structures with abnormal activity in MDD (Perlman et al., 2019).

Research suggests that MDD may be characterized by an overactive emotion processing circuit, which involves the extended amygdala (Matthews et al., 2008). For instance, the amygdala displays hyperactivity during emotional face processing and in response to negatively valenced pictures, words, and performance feedback in participants with MDD (Thomas & Elliot, 2009). In turn, amygdala hyperactivity contributes to core features of depression, such as the inability to cope with stress, emotional reactivity, and rumination on emotionally unpleasant memories (Drevets et al., 2002; Drevets, 2001; Hamilton & Gotlib, 2008). Simultaneously, there is evidence suggesting that participants with MDD demonstrate hypoactivity of the amygdala in response to positive stimuli (Groenewold et al., 2013). This opposing effect based on the emotional valence of stimuli might contribute to the negativity bias in MDD, such that negative information is sent for further processing, whereas positive information is inhibited (Groenewold et al., 2013). Abnormalities in the connectivity between the amygdala and cortical regions are also related to cognitive biases and neuropsychological alterations in MDD (Thomas & Elliot, 2009). For instance, studies have reported decreased amygdala FC with the dorsolateral PFC and anterior cingulate cortex (ACC) (Matthews et al., 2008; Siegle et al., 2007). Thus, the amygdala has a critical role in different aspects of affect processing and mood regulation, and might be implicated in antidepressant response (Chen et al., 2008).

Currently, monoaminergic antidepressants, including selective serotonin reuptake inhibitors (SSRIs), are the first-line treatment option in MDD (Kennedy et al., 2016) and are hypothesized to target neurotransmission in malfunctioning brain circuits (Ceskova, 2016). Serotonin, in particular, influences how the amygdala regulates emotion (Smith & Porrino, 2008) by modulating circuits such as those involved in impulse control (Liu et al., 2018). Single doses of SSRI antidepressants given to healthy participants appear to reduce cerebral blood flow in regions implicated in emotion regulation, including the amygdala, at rest (Chen et al., 2011). Moreover, antidepressants targeting serotonin have been found to enhance positive biases in healthy individuals (Harmer et al., 2004), perhaps by reducing amygdala response to negative stimuli and increasing its response to positive stimuli (Anderson et al., 2007; Norbury et al., 2009). Targeting the amygdala with SSRIs may therefore potentially normalize aberrant activity and FC in emotion regulation circuits among participants with MDD.

Our group's recent systematic review revealed that, compared to healthy control subjects, antidepressant-free participants with MDD commonly demonstrated hyperactivity of the amygdala in response to negatively valenced emotional stimuli; there was less evidence supporting amygdala hypoactivity in response to positive stimuli (Tassone et al., 2022). Our review also identified abnormal patterns of event-related amygdala-ACC FC and resting-state amygdala FC with the insula, PFC, temporal, and parietal cortices at baseline (Tassone et al., 2022). As a result, the current systematic review aimed to examine whether these alterations may be normalized by a course of treatment with SSRI antidepressants. Here, we sought to identify the reported changes in the amygdala across individuals with MDD following SSRI pharmacotherapy, as measured by functional magnetic resonance imaging (fMRI). Amidst the vast number of studies that have proposed neuroimaging correlates of treatment response following pharmacotherapy, inconsistencies in findings arise as differential impacts on the amygdala have been reported across SSRI antidepressants, including altered amygdala FC with a variety of regions following treatment (e.g., sertraline versus fluoxetine) (Anand et al., 2007; Chen et al., 2008).

Previous attempts to synthesize the literature on antidepressant response in MDD have investigated other neuroimaging techniques, such as structural magnetic resonance imaging, magnetic resonance spectroscopy, and positron emission tomography/single photon emission computed tomography (Bellani et al., 2011; Delaveau et al., 2011; Gerlach et al., 2022). One systematic review also assessed resting-state fMRI findings to elucidate MDD antidepressant treatment response (Dichter et al., 2015), overlooking studies that used event-related metrics. In addition, an fMRI review examining pharmacotherapy studies suggested that it was not possible to identify distinctive patterns relating to treatment (Arnone, 2019). Each of these studies has focused on the overall brain effects of different classes of antidepressants, with no predefined region of interest (ROI) or explicit focus on SSRIs. Moreover, meta-analyses assessing amygdala activity and FC have examined both medicated and unmedicated samples of participants with MDD (Li & Wang, 2021; Tang et al., 2018). Since the functioning of the amygdala may be altered through treatment, combining these types of samples introduces confounds and precludes an understanding of SSRI-related effects. Given the crucial role of the amygdala in affect regulation, a core component of MDD, a systematic review analyzing the effects of SSRI antidepressant treatment on this particular region via resting-state and event-related fMRI is warranted as the amygdala has the potential to act as a personalized predictor of treatment response.