Role of the rostral anterior cingulate cortex in emotion processing in Treatment Resistant Depression

doi:10.21203/rs.3.rs-4907107/v1

Download PDF

Article

Role of the rostral anterior cingulate cortex in emotion processing in Treatment Resistant Depression

https://doi.org/10.21203/rs.3.rs-4907107/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The rostral anterior cingulate cortex (rACC) has been identified as a key region in treatment-resistant depression (TRD), potentially influencing the adaptive interplay between the default mode network and other critical neural networks. This study aims to further uncover the role of the rACC in TRD by investigating its differential connectivity during tasks that evoke conscious and non-conscious emotional responses. Thirty-nine TRD patients, 35 treatment-sensitive depression (TSD) patients, and 38 healthy controls (HC) underwent 3T fMRI while performing tasks designed to elicit emotional responses to facial expressions under supraliminal and subliminal conditions. Connectivity patterns of the rACC were analyzed using seed-based and region-of-interest (ROI) approaches. During the processing of positive emotions in the subliminal task, TRD patients showed increased rACC connectivity to the cerebellum and middle temporal gyrus, compared to HC and TSD. Furthermore, significant hypoconnectivity between the rACC and hippocampus was found in the TRD, during the supraliminal processing of positive emotions, compared to TSD and HC. Altered neural connectivity to both subliminal and supraliminal processing of positive facial emotion distinguishes TRD from healthy individuals and patients who respond to depression treatments. This observation fits with anhedonia symptoms that persist in TRD and our findings identify that the underlying mechanism for this inability to process positive emotions is via altered connectivity of the rACC to regions involved in regulating or contextualising emotions.

Treatment-resistant depression (TRD) represents a significant clinical challenge, characterized by a persistent lack of response to standard antidepressant treatments. This condition profoundly affects patients' quality of life and stresses the need for a deeper exploration into the complex brain dynamics that contribute to its persistence.

The investigation of emotion processing in TRD is particularly crucial given its central role in depressive symptomatology [1]. Emotion processing deficits, characterized by impairments in recognizing, interpreting, and responding to emotional stimuli, are a hallmark of major depressive disorder (MDD) [2, 3, 4, 5], including its treatment-resistant forms [1]. Studies have shown that patients with TRD often exhibit altered neural responses to emotional stimuli, reflecting dysregulation in key brain areas and networks involved in emotion processing [6, 7, 8].

The understanding of the pathophysiology of depression has shifted to a model based on dysregulation of neural networks, rather than a single neuroanatomical location [9]. Central to these investigations are neural networks like the default-mode network (DMN), which is known for its role in self-referential thought processes and mind-wandering. The DMN exhibits peak activity during periods of restful introspection and is diminished when engaging in tasks that require external goal-oriented focus [10]. This reduction in activity, or 'deactivation,' is believed to indicate a shift from internal processing to increased attention on external tasks [10, 11].

The rostral anterior cingulate cortex (rACC), considered a hub of the DMN [12, 13], is intricately linked with regions and networks vital for emotional, cognitive, somatic, and social functions. It acts as a central processor for diverse information streams, enabling fluid transitions between introspective and task-focused cognitive states [14, 15]. Even during specific task engagement, the rACC plays a key role in modulating connectivity with networks that are inherently linked, potentially influencing the adaptive interplay between the DMN and other critical neural networks [16, 17].

Abnormal functioning of the rACC is closely associated with detrimental features of depression such as persistent negative thinking, biased negative perception, attentional disruptions, and impaired regulation of emotions [18, 19, 20]. These dysfunctions are particularly relevant in the context of TRD, where the rACC exhibits reduced activity at rest in patients who do not respond to treatments, compared to those who do, across various therapeutic modalities including pharmacotherapy [21], sleep intervention [22], and rTMS [13]. The atypical activity of the rACC at rest is implicated in the development, persistence, and relapse of depressive episodes [23, 15] with heightened rACC activity as a central neural mechanism underlying rumination and persistent thought patterns [24, 25, 26].

Studies indicate that elevated rACC activity might serve as a dynamic marker for antidepressant treatment responsiveness, influenced by the degree of treatment resistance [27]. This relationship is nuanced, showing variability with the levels of treatment resistance. Our previous work extends these findings, demonstrating that patients who were unresponsive to more intensive treatments like rTMS and ECT display higher rACC activity at rest relative to those treated with first-line therapies such as psychotherapy and antidepressants [28], which suggests rACC activity could be mediated by level of treatment resistance. However, a significant gap remains in our understanding of how these neural networks function during active emotional processing tasks. While resting-state fMRI has revealed much about the role of rACC in TRD, task-based fMRI can provide additional insights into the interplay of the rACC and the key brain regions that are activated and modulated in response to emotional stimuli. This approach is particularly relevant for understanding the neural basis of emotional dysregulation in TRD. Therefore, the goal of our investigation was to extend our understanding of the neural underpinnings of TRD, particularly the role of the rACC beyond resting-state conditions and provide a more comprehensive view of the involvement of this region in the disorder's complex neurobiology. Specifically, this study evaluated the functional connectivity of the rACC in treatment-resistant depression, during supraliminal and subliminal level of processing of both negative and positive emotions. Our study included both healthy individuals and treatment responsive patients with depression for comparison.

Participants

Initially, 39 individuals with TRD and 35 individuals with TSD were recruited through a local network of specialist psychiatrists and clinics. TRD and TSD individuals met DSM-5 criteria for primary diagnosis of MDD, assessed through the Structured Clinical Interview for the DSM-5 (SCID-5) [29]. The inclusion criteria for TRD were: no remission of symptoms with at least two adequate trials (in terms of dosage, duration – 6 weeks for each trial) of antidepressant of different pharmacologic classes, as well as the presence of moderate to severe depressive symptoms (assessed by a rating greater or equal to 16 in the 21-item Hamilton Depression Rating Scale – HAMD-21 [30]. Participants in the TSD group were identified as patients who had achieved complete remission of symptoms for at least two weeks (characterized by a HAMD-21 score of less than or equal to 7). The control group (HC) comprised of 38 healthy individuals recruited through community advertisements with no psychiatric illnesses, assessed using the SCID-5. All participants were aged between 18 and 65 years old.

For both patient groups, indices of illness severity and chronicity were assessed. These indices included age of onset, number of inpatient hospitalizations, length of remission period since last episode, number of previous depressive episodes, history of suicidal ideation and behaviour, and history of suicide attempt. Information on past and current medication and other forms of treatment (e.g. ECT, or TMS) was also collected. Level of functioning was assessed by the Social and Occupational Functioning Assessment Scale (SOFAS) [31].

HC and patient groups (TRD and TSD) were matched for age, sex and education status. Exclusion criteria for all participants included a) inability to provide consent, b) insufficient English proficiency, c) current primary diagnosis of eating disorder, psychosis, personality disorder or primary PTSD, d) substance dependence for the past 3 months, e) pregnancy, f) history or current neurological disorder or prior brain injury, g) ECT or TMS in the last 6 months, h) contraindication to MRI.

Data collection was conducted at Westmead Hospital, Department of Radiology and at the Brain Dynamics Centre, The Westmead Institute for Medical Research, in Sydney, Australia. The research protocol was approved by the Western Sydney Local Health District Human Research Ethics Committee and all participants provided written consent.

MRI data collection

Functional imaging was performed on a Siemens Prisma 3T MRI system running VE11C software and utilising a 2D GRE EPI sequence with a 64-channel head/neck array coil. Whole brain data for both faces tasks was acquired in the axial oblique plane with 47 interleaved slices of 3mm thickness/0mm gap parallel to the AC-PC line. Other parameters were as follows: TR/TE = 2500ms/27ms, Flip angle = 90°, FOV = 240mm, matrix = 86 x 86, voxel size = 3mm x 2.8mm x 2.8mm, parallel imaging r = 2, number of volumes = 120. Total acquisition time was 5 minutes 13 seconds.

Task description

The emotion processing tasks utilized in this study were based on established paradigms previously described in [32] and [33]. These tasks involve the presentation of facial expressions categorized into various emotional states: threat-related emotions (fear and anger), disgust, loss-related emotions (sadness), reward-related emotions (happiness), and neutral expressions. The stimuli are derived from a standardized series of facial expressions and are modified for central positioning at eye level.

There were two tasks: supraliminal emotional face processing, and subliminal emotional face processing, each analysed separately.

In the subliminal task, each emotional face is briefly presented for 16.7 ms, followed by a perceptual mask of a neutral face for 150 ms and an interstimulus interval of 1233.3 ms. This presentation is crafted to be below the level of conscious awareness. Behavioral psychophysical testing, as detailed in Williams et al. [33], confirms that this duration is sufficient to meet the criteria for subliminal threshold detection.

The supraliminal task differs in presentation duration, with each emotional face displayed for 500 ms, based on evidence that this duration ensures conscious processing of the emotion stimulus [33]. The interstimulus interval in this condition is 750 ms. For both subliminal and supraliminal task presentations, faces are grouped in blocks by emotion, with each block containing eight faces and repeated five times, totalling 240 stimuli.

Throughout both paradigms, participants are instructed to focus on each face, with no specific behavioral responses required during scanning. This approach is consistent with findings that passive processing of emotional stimuli can elicit significant brain activation [34].

MRI data processing

MRI data was processed and analysed using Matlab R2018b (The Mathworks inc, Natick, Massachusetts), SPM12 (Wellcome Trust Centre for Neuroimaging, London, UK), CONN functional connectivity toolbox v16b. Details of pre-processing steps are presented in our previous study [35]. Four TRD participant datasets were excluded for excessive head motion during the scan. Five participants who did not complete the Faces tasks were identified and subsequently excluded from the analysis. This resulted in 31 TRD, 35 TSD and 37 healthy controls for final analyses.

Demographic and clinical characteristics

All three groups were compared for age and gender (demographic variables), using a one-way ANOVA and chi square test, respectively. The TRD and TSD groups were compared for age of onset, age of first episode, depression severity (HAMD-21 score), functionality (SOFAS score), severity of worst depressive episode, number of previous depressive episodes, using student t-tests. The groups were also compared for history of hospitalizations, suicidal attempts and suicidal ideation, using chi square tests of independence.

Functional connectivity (FC) analyses

Generalized PsychoPhysiological Interaction (gPPI) analyses, as employed in this study, are designed to explore task-specific changes in the BOLD signal across different brain regions over time. The gPPI approach is particularly suited for studies involving multiple task conditions, such as the varied emotional stimuli in our tasks.

In this analysis, the interaction term is created by convolving each task condition's onset times, representing different emotions, with the hemodynamic response function. This term is then multiplied by the seed region's estimated neural activity, derived from the deconvolved BOLD signal. The gPPI model is distinct from standard PPI in that it includes the interaction factors from all conditions simultaneously in the estimation model in order to better account for between-condition overlaps. In this analysis, PPI effects (interaction terms in PPI model) are relative to the baseline state (the baseline state is defined by the zero values of the interaction term), so they provide a relative measure of connectivity characterizing differential task-specific effects.

One critical contrast employed was "Happy" versus "Neutral," where "Neutral" represents emotionally neutral faces. This contrast was chosen to elucidate the neural mechanisms underlying positive emotion processing. By comparing brain activity associated with viewing happy faces against neutral faces, we aimed to identify the specific brain regions and their interconnections that are selectively engaged in the perception and processing of positive emotional stimuli.

Additionally, we analyzed the aggregated negative valenced conditions (“Sadness”, “Anger”, “Fear”, and “Disgust”) versus "Neutral" to investigate the networks involved in negative emotion processing. By averaging together these negative conditions and contrasting them with neutral emotion processing, we aimed to capture a broad perspective of the brain's response to negative stimuli.

In this study, the primary gPPI analysis employed a seed-to-voxel whole brain approach to comprehensively examine the role of the rACC in emotional processing in this population.

The selection of rACC ROI anatomic landmark derived from the voxels reported by Pizzagalli and colleagues [36], and also used in our prior studies on rACC [28]. Each functional rACC ROI was a 10 × 10 × 10 mm cube placed around the central MNI coordinate in this region ((-10 45 − 5). FC values were calculated from the rACC voxel-wise to the rest of the brain for each participant in the dataset as bivariate Fisher’s z-transformed correlation coefficients. This measures the association between the rACC seed BOLD timeseries and each voxel of the whole brain BOLD timeseries using generalized linear model (GLM).

The statistical parametric maps threshold was set p < 0.05 at the cluster-level family-wise error (FWE)-corrected for multiple comparisons, using an initial whole-brain voxel-wise p < 0.001. We used an arbitrary threshold of a minimum cluster size of 20 as necessary to be considered relevant.

Clinical factors

In order to analyse the associations between the FC measures and treatment resistance in the TRD group, we computed a variable to reflect chronicity of the disorder, by calculating the number of years from age of onset to current age, which is referred to as “number of years since onset”. The relationships between the mean FC values for the significant ROIs and different clinical factors (depression severity, functionality, severity of worst depressive episode, number of previous depressive episodes, and number of years since onset) in the two patient groups separately were analysed using Pearson correlation coefficients.

All effects considered significant at the p < 0.05 significance level corrected for multiple comparisons, using a Bonferroni adjustment. Statistical analyses were performed using SPSS software version 21 (IBM Corp, 2012).

Demographic and clinical characteristics

Demographic and clinical data for the final sample are summarized in Table 1. All three groups were comparable for age and sex. As expected, groups were significantly different in some clinical variables, reflecting a more severe clinical profile for TRD patients. Although the depression groups weren’t different for the number of past depressive episodes, the TRD group had higher rates of history of hospitalizations, history of ECT, and history of suicidal attempts.

Table 1

Summary of demographic and clinical characteristics of the sample.
	TRD (32)	TSD (34)	HC (37)	F/t/X^2	sig
Demographics
Age, Mean ± SD [Min-Max]	42.4 ± 13.8 [18.1–64.3]	37.4 ± 11.1 [20.0-57.4]	42.29 ± 14.1 [18.9–65.9]	n.s.	n.s.
Gender (M), N (%)	12 (38)	17 (50)	17 (44.7)	n.s.	n.s.
Clinical Profile
Age of onset, Mean ± SD [Min-Max]	29.97 ± 13.52 [8–53]	21.68 ± 9.40 [8–50]	n.a	n.s.	n.s.
Number of previous MDE, Mean ± SD [Min-Max]	8 ± 11 [1–40]	6.11 ± 6.16 [1–30]	n.a	n.s.	n.s.
Severity of worst MDE, Mean ± SD [Min-Max]	6.56 ± 1.13 [5–7]	5.52 ± 1.28 [3–7]	n.a	3.318	0.002
HAM-D-17 score, Mean ± SD [Min-Max]	23.66 ± 6.62 [15–39]	3.9 ± 2.92 [0–9]	n.a	15.611	0.000
SOFAS score, Mean ± SD [Min-Max]	74.45 ± 15.98 [40–100]	89.59 ± 5.56 [78–95]	n.a	-5.052	0.000
History of Hospitalizations, N (%)	22 (68)	6 (17.6)	n.a	17.625	0.000
History of ECT, N (%)	8 (25)	0 (0)	n.a	9.672	0.002
History of TMS, N (%)	2 (6.3)	0 (0)	n.a	n.a.	n.a.
History of Suicidal Ideation, N (%)	25 (78)	26 (76.5)	n.a	n.s.	n.s.
History of Suicidal Attempt, N (%)	17 (53)	5 (14.7)	n.a	12.439	0.000
n.a. – not applicable, n.s. – not significant; SD – Standard Deviation; M – male; MDE – Major Depressive Episode; HAMD-21 – Hamilton Depression Rating Scale, 21 items; SOFAS – Social and Occupational Functioning Assessment Scale; CGI-S – Clinical Global Impression, Severity; ECT – Electroconvulsive Therapy; TMS – Transcranial Magnetic Stimulation; N – total number.

There were no significant differences between groups for motion during the scan, after the exclusion of the four TRD participants for excessive motion.

Whole brain seed-based connectivity analyses – rACC

Supraliminal processing:

Functional connectivity differences among the three groups were significant in the context of processing happy versus neutral faces. Specifically, in the whole-brain analysis, a significant difference was observed for connectivity of the rACC with the hippocampus (p = 0.036, FDR corrected, see Fig. 1). Post-hoc cluster analyses performed based on extracted beta connectivity values from the significant rACC-hippocampus connectivity found significant differences between the TRD relative to both TSD and HC groups. TRD patients had reduced connectivity compared to TSD (t = -5.24, p = 0.000) and HC (t = -3.47, p = 0.001) (see Table 2). There were no significant differences in connectivity between TSD and HC (p = 0.061) (see Table 2).

Table 2

Whole-brain seed-based functional connectivity of the rostral anterior cingulate cortex during supraliminal and subliminal emotion processing tasks.
Task	Contrast	Brain region (aal)	P (FDR corrected)	Cluster size (k)	Peak voxel coordinates (x y z)	Post-hoc
Supraliminal task	Positive > Neutral	Hippocampus	0.036	98	40 − 24 -12	HC, TSD > TRD
Supraliminal task	Negative > Neutral	-	-	-	-	-
Subliminal task	Positive > Neutral	Cerebellum	0.001	106	-50 -46 -50	TRD > HC > TSD
		Middle temporal gyrus	0.001	99	-36 12–42	TRD > HC > TSD
		Temporal fusiform cortex	0.04	54	40–40 -24	HC > TRD,TSD
	Negative > Neutral	Dorsal anterior cingulate cortex	< 0.001	80	06 24 24	n.s. (averaged) Disgust: TRD, TSD > HC

There were no significant differences in the FC whole-brain analyses of the rACC for the supraliminal processing of negative emotions versus neutral faces.

Subliminal processing:

Significant differences between the three groups in the functional connectivity of the rACC for the subliminal task were found for the processing of happy versus neutral faces. Significant differences were identified for connectivity of the rACC and three clusters in the brain: the cerebellum (p = 0.001, FDR corrected), the middle temporal gyrus (p = 0.001, FDR corrected), and the temporal fusiform gyrus (p = 0.04, FDR corrected) (see Table 2, Fig. 2).

Post-hoc tests revealed significant differences between all three groups for the connectivity between rACC and the cerebellum and between the rACC and the middle temporal gyrus. TRDs had hyperconnectivity compared to TSD and HC groups for both the rACC-cerebellum connectivity (TRD > TSD: t = 4.816, p < 0.001; TRD > HC: t = 2.415, p = 0.018) and the rACC-middle temporal gyrus connectivity (TRD > TSD: t = 5.154, p < 0.001; TRD > HC: t = 3.260, p = 0.002). TSD on the other hand had lower connectivity compared to HC for the cerebellum cluster (t=-2.610, p = 0.011) and the middle temporal gyrus cluster (t=-2.338, 0.022).

For the third cluster i.e. connectivity between the rACC and the temporal fusiform gyrus, both TRD and TSD groups had lower connectivity than HC but were not different from each other (see Table 2).

For negative versus neutral emotional faces, there was a significant group effect for connectivity between the rACC and the dorsal ACC. Posthoc analyses revealed that group differences were significant only during the processing of disgust with both the TRD and TSD groups with greater connectivity compared to healthy controls (p < 0.001, FDR corrected) but not different from each other (See Table 2).

Due to known biases of neutral facial cues to be interpreted as negative due to psychopathology, we also conducted all analyses contrasting each emotion and implicit baseline i.e. rest presented in Supplementary Material. We also evaluated group differences in connectivity beyond the rACC for other key neural regions involved in emotion processing (amygdala, hippocampus, insula, pregenual and subgenual ACC) in supplementary.

Clinical factors

There were no significant correlations between the clinical variables and the functional connectivity measures.

This study aimed to elucidate the role of the rostral anterior cingulate cortex in the neural underpinnings of emotion processing in TRD. We wanted to evaluate connectivity related to this region during emotion processing, extending from existing knowledge on altered functional connectivity of the rACC – a key region of the default mode network - in TRD during rest [13].

Neuroimaging studies have revealed that the perception of emotion in faces is a complex process, engaging a network of brain regions working together dynamically across time and space, rather than being confined to isolated neural activities in singular areas [37]. Furthermore, these studies highlight the involvement of both supraliminal and subliminal pathways in emotional processing, where supraliminal processing involves the conscious perception of emotions, and subliminal processing occurs without conscious awareness, both of which play crucial roles in how we perceive and respond to emotional facial expressions [38].

The processing of positive and negative emotional faces in the brain engages overlapping yet distinct neural circuits, reflecting the nuanced ways our brains interpret and respond to different emotional stimuli [39]. Understanding these differences is crucial for insights into various psychological conditions. For example, positive emotional faces such as happy facial expressions engage the brain's reward system [40]. This system is less activated by negative emotional faces, which instead may activate brain areas associated with threat detection and stress response [40]. The rACC, a crucial region for emotional regulation and cognitive processing, appears to exhibit atypical activity suppression during tasks, which Leonards and collaborators [17] associated with disturbances in adaptive neural communication and the dynamic balance between internal and external cognitive modes. These disruptions may form the basis of the maladaptive cognitions and biased emotional processing characteristic of depression [17]. This region is involved in emotion processing and regulation for both positive and negative emotions. However, its engagement can vary depending on the complexity of the emotion processing task and the specific emotional context [41].

In our study, we investigated the functional connectivity underpinnings of the processing of positive and negative emotional expressions, contrasting with the processing of emotionally neutral faces. Additionally, we used both a supraliminal emotion perception task and a subliminal emotion perception task to allow us to measure both explicit and implicit level of emotion processing [42].

For negative versus neutral emotional faces, our analyses revealed a significant group effect only for the supraliminal task and this was for connectivity between the rACC and the dorsal ACC. Posthoc analyses revealed that group differences were significant only during the processing of disgust facial expressions. Both the TRD and TSD groups exhibited greater connectivity between the rACC and the dorsal anterior cingulate cortex compared to healthy controls but not different from each other which suggests a shared neural mechanism underlying the processing of negative emotions, particularly disgust, in both TRD and TSD. The enhanced connectivity between the rACC and dACC might indicate an increased demand for cognitive control and emotional regulation in these groups, reflecting the heightened negative emotional reactivity that is a hallmark of treatment-resistant and treatment-sensitive depression [43]. The lack of significant difference between TRD and TSD groups suggests that this connectivity alteration might be a trait feature in depressive disorders and irrespective of response to treatment.

In contrast for positive vs neutral emotional faces, we saw group differences for both the supraliminal and subliminal processing tasks. For the supraliminal processing task, our analysis revealed a pattern of hypoconnectivity between the rACC and the hippocampus when individuals with TRD processed happy emotions compared to neutral faces, that distinguished them from both the TSD and healthy control groups. This observation aligns with existing literature highlighting hypoconnectivity in reaction to emotional faces in patients with major depression who were non-responsive to antidepressant medication treatments [9]. The observed hypoconnectivity in TRD when processing happy relative to neutral faces could reflect a diminished capacity to modulate neural responses to positive stimuli, a characteristic that might contribute to the anhedonia often seen in TRD, as an expression of impaired positive affect regulation [44]. Existing functional neuroimaging literature corroborates these valence-specific alternations in emotion processing in TRD. Patients with TRD compared with healthy volunteers showed reduced responses to positive emotion within the caudate and insula [42].

Anhedonia, or the diminished ability to experience pleasure, is a core feature of major depressive disorder, and it is particularly pronounced in TRD [44]. The rACC has been implicated in the regulation of emotions and in modulating responses to reward-related stimuli. The link between such neural connectivity patterns and anhedonia is supported by findings that suggest alterations in reward processing circuits are associated with anhedonia in depression [44]. Specifically, reduced activation in the ventral striatum, an area closely connected to the rACC and involved in reward processing, has been correlated with the severity of anhedonia in MDD [44]. However, we did not observe connectivity differences between the rACC and the ventral striatum for our tasks which might suggest this neural mechanism is better evident in tasks that directly tap reward processing functions.

The connectivity of the rACC with the hippocampus, an area which is crucial for memory formation and retrieval, suggests a pathway through which emotional experiences are integrated and stored. This connectivity is believed to play a role in how emotional memories influence current emotional states and decision-making [45]. The hippocampus is involved in the cognitive processing of emotional memory, so the decreased functional connectivity between the rACC and hippocampus may be related to abnormal regulation of emotional memory in TRD [46]. Interestingly, our findings did not indicate significant differences in the FC of the rACC with the hippocampus between TSD and HC, suggesting that this altered connectivity might be specific to TRD. Additionally, literature on the connectivity between the DMN and hippocampus indicates hyperconnectivity in MDD patients [9]. In our study, we not only found that this hyperconnectivity is absent in patients who are currently depressed (the TRD group), but it is also not present in the TSD group (patients with depression who recovered), highlighting a unique connectivity pattern in TRD.

Previous literature also has shown increased hippocampal response to positive stimuli associated with treatment response [47], suggesting that high levels of activity in the hippocampal region at baseline may indicate a resilience in neural responsivity in those patients who subsequently showed a clinical improvement [47]. Therefore, hypoconnectivity between the rACC and the hippocampus, a region involved in memory and emotion processing [48, 49], may signal a disrupted neural circuit that contributes to the diminished sensitivity to positive stimuli seen in TRD. This disruption could undermine the ability of individuals with TRD to engage with and retain positive emotional experiences, further perpetuating the cycle of depression.

One of the most notable findings on the subliminal task was the hyperconnectivity between the rACC and the cerebellum also during happy face processing, where connectivity was greatest in TRD compared to HC and TSD. This result is particularly interesting as it introduces a novel aspect of cerebellar involvement in TRD. Traditionally associated with motor function and coordination, the cerebellum has increasingly been recognized for its role in cognitive and emotional processing [50]. The heightened connectivity observed in TRD could suggest an adaptive or maladaptive mechanism where the cerebellum compensates or exacerbates emotional regulation difficulties, particularly in the non-conscious processing of positive emotions. This finding prompts further investigation into how the DMN might interact abnormally with the cerebellum in TRD, potentially disrupting normal emotional regulation.

Additionally, we also observed hyperconnectivity of the rACC with the middle temporal gyrus in the subliminal processing of positive emotions, that distinguished TRD from both HC and TSD. While previous literature commonly associates major depression with decreased activity in the middle temporal gyrus [51], the observed rACC hyperconnectivity to this region in TRD suggests a distinct neural pattern that differentiates it from TSD. The middle temporal gyrus is implicated in the semantic processing of emotions and social cognition [52]. The increased connectivity could reflect an over-engagement or faulty modulation of this region when processing happiness subliminally, pointing to a possible neural basis for the altered perception and integration of positive social cues in TRD.

The third significant finding during the subliminal processing of positive emotions involved the right temporal fusiform gyrus, where HC exhibited greater connectivity than both TRD and TSD. This aligns with existing research indicating reduced fusiform gyrus activity in depression during the processing of facial emotions [51]. The fusiform gyrus plays a crucial role in face recognition and the interpretation of facial expressions. Its reduced involvement in both TRD and TSD groups suggests a shared impairment in processing emotional faces at a non-conscious level, potentially contributing to the social cognition deficits observed in depressive disorders in general rather than specifically to TRD.

The distinct patterns of connectivity observed during supra versus subliminal processing tasks imply a disconnection in how emotions are processed at different levels of awareness in TRD. Non-conscious processing tasks revealed significant connectivity between the rACC and areas typically associated with automatic emotional responses, such as the cerebellum and middle temporal gyrus. This suggests that individuals with TRD may have heightened automatic emotional reactivity, particularly in response to positive stimuli, which is not accessible to conscious awareness and could contribute to the persistent negative emotional bias seen in this disorder. In contrast, supraliminal emotion processing did not show these patterns, indicating that the rACC plays a role in both levels of processing in TRD but engages different regions. Specifically, during subliminal tasks, the rACC connects with automatic processing regions, while during conscious tasks, this region has altered hyperconnectivity with the hippocampus, which is involved in deliberate emotional processing. The lack of observed differences in automatic processing regions during the supraliminal task could be due to the fact that the task is not sensitive enough to capture these mechanisms, rather than these processes being normal in TRD. The differences seen in the subliminal task highlight the complexity of emotional processing in TRD. Collectively these findings underscore a complex network dysfunction in TRD, where the rACC's impaired modulation of neural responses to positive stimuli and its altered connectivity with key regions of the emotion processing network, namely the hippocampus and the middle temporal gyrus.

We acknowledge several limitations in this study. First, the cross-sectional design constrains our ability to infer causality between the observed alterations in functional connectivity and treatment resistance. To better understand the dynamic changes in connectivity and their relationship with treatment outcomes, future research should adopt longitudinal designs. This would also help untangle trait vs state markers of treatment resistance. Additionally, despite careful matching, our sample size is relatively modest, urging caution in the extrapolation of our findings. Studies with larger and more diverse cohorts are essential for corroborating our results. This study's focus on emotion processing through facial expression tasks also presents a limitation. Face processing is a multifaceted function, engaging a network of brain regions often termed the "core face network," including the fusiform face area, occipital face area [53], and superior temporal sulcus [39]. Not all relevant regions within this network were examined in our analysis as our focus was primarily on emotion processing. Future research should distinguish between functional connectivity in face processing and broader emotion processing by employing diverse emotional processing paradigms. The same applies for emotion processing versus emotion regulation. The sample's heterogeneity is another notable limitation, along with the study's focus on HAMD-17 symptoms, excluding a thorough investigation of anhedonia. This aspect is critical for a clinically meaningful interpretation of our findings especially considering that differences were for processing of happy emotions.

In conclusion, our findings, along with the broader literature, suggest a critical role for the rACC in the disrupted emotional processing observed in TRD. Abnormalities in the connectivity of the rACC with the hippocampus and the middle temporal gyrus might reflect an impaired ability to regulate negative emotions and to properly integrate emotional memories, which are essential for adaptive responses to environmental stimuli and for the efficacy of antidepressant treatments. The findings suggest that TRD involves distinct neural pathways for processing emotional stimuli under conscious and non-conscious conditions. Specifically, non-conscious processing seems to involve heightened connectivity between the rACC and brain regions like the cerebellum and the middle temporal gyrus in TRD compared to TSD and controls. This may indicate that while these regions more responsive or exhibit altered connectivity in TRD, these changes occur outside of conscious awareness. Such disruptions in automatic emotional processing could contribute to persistent symptoms, such as pervasive sadness or anhedonia, despite treatment. In other words, although the rACC may be involved in automatic emotional responses, it appears to struggle with integrating these responses into conscious emotional regulation effectively. This suggests that while the rACC may be involved in heightened automatic emotional responses, its hypoconnectivity with the hippocampus suggests that it fails to integrate these responses into conscious emotional regulation effectively.

Addressing these neural circuitry disruptions holds promise for advancing our understanding of the role of the rACC in positive emotion processing in TRD, paving the way for interventions that more precisely target the neural bases of the disorder's core symptoms.

Conflict of Interest

The authors declare no conflicts of interest.

Fan X, Mocchi M, Pascuzzi B, Xiao J, Metzger BA, Mathura RK, et al. Brain mechanisms underlying the emotion processing bias in treatment-resistant depression. bioRxiv [Preprint]. 2023 Aug 28:2023.08.26.554837.
Fitzgerald PB, Laird AR, Maller J, Daskalakis ZJ. A meta-analytic study of changes in brain activation in depression. Hum Brain Mapp. 2008;29(6):683-695.
Fu CH, Williams SC, Cleare AJ, Brammer MJ, Walsh ND, Kim J, et al. Attenuation of the neural response to sad faces in major depression by antidepressant treatment: a prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry. 2004;61(9):877-889.
Vanderlind WM, Millgram Y, Baskin-Sommers AR, Clark MS, Joormann J. Understanding positive emotion deficits in depression: From emotion preferences to emotion regulation. Clin Psychol Rev. 2020 Mar;76:101826.
Anderson IM, Shippen C, Juhasz G, Chase D, Thomas E, Downey D, et al. State-dependent alteration in face emotion recognition in depression. Br J Psychiatry. 2011 Apr;198(4):302-8.
Roseman L, Demetriou L, Wall MB, Nutt DJ, Carhart-Harris RL. Increased amygdala responses to emotional faces after psilocybin for treatment-resistant depression. Neuropharmacology. 2018 Nov;142:263-269.
Murrough JW, Collins KA, Fields J, DeWilde KE, Phillips ML, Mathew SJ, et al. Regulation of neural responses to emotion perception by ketamine in individuals with treatment-resistant major depressive disorder. Transl Psychiatry. 2015 Feb 17;5(2).
Loureiro JRA, Sahib AK, Vasavada M, Leaver A, Kubicki A, Wade B, et al. Ketamine's modulation of cerebro-cerebellar circuitry during response inhibition in major depression. Neuroimage Clin. 2021;32:102792.
Kaiser RH, Andrews-Hanna JR, Wager TD, Pizzagalli DA. Large-scale network dysfunction in major depressive disorder: a meta-analysis of resting-state functional connectivity. JAMA Psychiatry. 2015;72(6):603-611.
Buckner RL, Andrews-Hanna JR, Schacter DL. The brain’s default network: Anatomy, function, and relevance to disease. Ann N Y Acad Sci. 2008;1124:1–38.
Harrison BJ, Pujol J, Contreras-Rodríguez O, Soriano-Mas C, López-Solà M, Deus J, et al. Task-Induced deactivation from rest extends beyond the default mode brain network. PLoS One. 2011;6(7).
Fair DA, Cohen AL, Power JD, Dosenbach NU, Church JA, Miezin FM, et al. Functional brain networks develop from a “local to distributed” organization. PLoS Comput Biol. 2009;5(5).
Ge R, Downar J, Blumberger DM, Daskalakis ZJ, Vila-Rodriguez F. Functional connectivity of the anterior cingulate cortex predicts treatment outcome for rTMS in treatment-resistant depression at 3-month follow-up. Brain Stimul. 2020 Jan-Feb;13(1):206-214.
Harrison BJ, Davey CG, Savage HS, Jamieson AJ, Leonards CA, Moffat BA, et al. Dynamic subcortical modulators of human default mode network function. Cereb Cortex. 2022;32(19):4345–4355.
Lichenstein SD, Verstynen T, Forbes EE. Adolescent brain development and depression: a case for the importance of connectivity of the anterior cingulate cortex. Neurosci Biobehav Rev. 2016;70:271–287.
Jamieson AJ, Leonards CA, Davey CG, Harrison BJ. Major depressive disorder associated alterations in the effective connectivity of the face processing network: a systematic review. Transl Psychiatry. 2024;14(1):62.
Leonards CA, Harrison BJ, Jamieson AJ, Agathos J, Steward T, Davey CG. Altered task-related decoupling of the rostral anterior cingulate cortex in depression. Neuroimage Clin. 2024;41:103564.
Gotlib IH, Joormann J. Cognition and depression: Current status and future directions. Annu Rev Clin Psychol. 2010;6:285–312.
Pizzagalli DA. Frontocingulate dysfunction in depression: toward biomarkers of treatment response. Neuropsychopharmacology. 2011;36(1):183–206.
Price JL, Drevets WC. Neurocircuitry of mood disorders. Neuropsychopharmacology. 2010;35(1):192–216.
Pizzagalli DA, Pascual-Marqui RD, Nitschke JB, Oakes TR, Larson CL, Abercrombie HC, et al. Anterior cingulate activity as a predictor of degree of treatment response in major depression: evidence from brain electrical tomography analysis. Am J Psychiatry. 2001;158(3):405-15.
Kito S, Fujita K, Koga Y. Regional cerebral blood flow changes after low frequency transcranial magnetic stimulation of the right dorsolateral prefrontal cortex in treatment-resistant depression. Neuropsychobiology. 2008;58(1):29-36.
Gotlib IH, Hamilton JP. Neuroimaging and depression: Current status and unresolved issues. Curr Dir Psychol Sci. 2008;17(2):159–163.
Zhou HX, Chen X, Shen YQ, Li L, Chen NX, Zhu ZC, et al. Rumination and the default mode network: meta-analysis of brain imaging studies and implications for depression. Neuroimage. 2020;206:116287.
Greicius MD, Flores BH, Menon V, Glover GH, Solvason HB, Kenna H, et al. Resting-state functional connectivity in major depression: abnormally increased contributions from subgenual cingulate cortex and thalamus. Biol Psychiatry. 2007;62(5):429-437.
Sheline YI, Price JL, Yan Z, Mintun MA. Resting-state functional MRI in depression unmasks increased connectivity between networks via the dorsal nexus. Proc Natl Acad Sci USA. 2010;107(24):11020-11025.
Hunter AM, Korb AS, Cook IA, Leuchter AF. Rostral anterior cingulate activity in major depressive disorder: state or trait marker of responsiveness to medication? J Neuropsychiatry Clin Neurosci. 2013 Spring;25(2):126-33.
Prentice A, Barreiros AR, van der Vinne N, Stuiver S, van Dijk H, van Waarde JA, et al. Rostral Anterior Cingulate Cortex Oscillatory Power Indexes Treatment-Resistance to Multiple Therapies in Major Depressive Disorder. Neuropsychobiology. 2023;82(6):373-383.
Shabani A, Masoumian S, Zamirinejad S, Hejri M, Pirmorad T, Yaghmaeezadeh H. Psychometric properties of Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) in the Iranian context. Psychol Test Assess Model. 2020;62(1):1-12.
Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;23:56–62.
Morosini PL, Magliano L, Brambilla L, Ugolini S, Pioli R. Development, reliability and acceptability of a new version of the DSM-IV Social and Occupational Functioning Assessment Scale (SOFAS) to assess routine social functioning. Acta Psychiatr Scand. 2000 Apr;101(4):323-9.
Korgaonkar MS, Erlinger M, Breukelaar IA, et al. Amygdala activation and connectivity to emotional processing distinguishes asymptomatic patients with bipolar disorders and unipolar depression. Biol Psychiatry Cogn Neurosci Neuroimaging. 2019;4(4):361-370.
Williams LM, Korgaonkar MS, Song YC, et al. Amygdala reactivity to emotional faces in the prediction of general and medication-specific responses to antidepressant treatment in the randomized iSPOT-D trial. Neuropsychopharmacology. 2015;40(10):2398-2408.
Costafreda SG, Brammer MJ, David AS, Fu CH. Predictors of amygdala activation during the processing of emotional stimuli: a meta-analysis of 385 PET and fMRI studies. Brain Res Rev. 2008;58(1):57-70.
Barreiros AR, Breukelaar IA, Mayur P, Andepalli J, Tomimatsu Y, Funayama K, et al. Abnormal habenula functional connectivity characterizes treatment-resistant depression. Neuroimage Clin. 2022;34:102990.
Pizzagalli DA, Webb CA, Dillon DG, Tenke CE, Kayser J, Goer F. Pretreatment rostral anterior cingulate cortex theta activity in relation to symptom improvement in depression: a randomized clinical trial. JAMA Psychiatry. 2018;75(6):547-554.
Vuilleumier P, Pourtois G. Distributed and interactive brain mechanisms during emotion face perception: evidence from functional neuroimaging. Neuropsychologia. 2007 Jan 7;45(1):174-94.
Williams LM, Liddell BJ, Rathjen J, Brown KJ, Gray J, Phillips M, et al. Mapping the time course of nonconscious and conscious perception of fear: an integration of central and peripheral measures. Hum Brain Mapp. 2004 Feb;21(2):64-74.
Haxby JV, Hoffman EA, Gobbini MI. The distributed human neural system for face perception. Trends Cogn Sci. 2000;4(6):223-233.
Haber SN, Knutson B. The reward circuit: linking primate anatomy and human imaging. Neuropsychopharmacology. 2010;35(1):4-26.
Bush G, Luu P, Posner MI. Cognitive and emotional influences in anterior cingulate cortex. Trends Cogn Sci. 2000;4(6):215-222.
Murrough JW, Collins KA, Fields J, DeWilde KE, Phillips ML, Mathew SJ, Wong E, Tang CY, Charney DS, Iosifescu DV. Regulation of neural responses to emotion perception by ketamine in individuals with treatment-resistant major depressive disorder. Transl Psychiatry. 2015 Feb 17;5(2)
Hamilton JP, Etkin A, Furman DJ, Lemus MG, Johnson RF, Gotlib IH. Functional neuroimaging of major depressive disorder: a meta-analysis and new integration of baseline activation and neural response data. Neuroimage. 2012;61(3):677-685.
Pizzagalli DA. Anhedonia and major depression: Insights from affective neuroscience. Biol Psychiatry. 2014;76(5):333-344.
Admon R, Pizzagalli DA. Dysfunctional reward processing in depression. Curr Opin Psychol. 2015;4:114-118.
Zhang X, Liang Y, Gao Y, et al. Altered functional connectivity of the hippocampus in patients with major depressive disorder. J Affect Disord. 2022;298:210-218.
Fu CH, Williams SC, Brammer MJ, Suckling J, Kim J, Cleare AJ, et al. Neural responses to happy facial expressions in major depression following antidepressant treatment. Am J Psychiatry. 2007;164:599-607.
Mayberg HS. Limbic-cortical dysregulation: a proposed model of depression. J Neuropsychiatry Clin Neurosci. 1997;9:471-481.
Phillips ML, Drevets WC, Rauch SL, Lane RD. Neurobiology of emotion perception, II: implications for major psychiatric disorders. Biol Psychiatry. 2003;54:515-528.
Schmahmann JD. The cerebellum and cognition. Neurosci Lett. 2019;688:62-75.
Stuhrmann A, Suslow T, Dannlowski U. Facial emotion processing in major depression: A systematic review of neuroimaging findings. Biol Mood Anxiety Disord. 2011;1(1):10.
Olson IR, Plotzker A, Ezzyat Y. The enigmatic temporal pole: A review of findings on social and emotional processing. Brain. 2007;130(7):1718-1731.
Gauthier I, Skudlarski P, Gore JC, Anderson AW. Expertise for cars and birds recruits brain areas involved in face recognition. Nat Neurosci. 2000;3(2):191-197.

The authors have declared there is NO conflict of interest to disclose

SupplementaryMaterial.docx

Download PDF

Reviewers invited by journal
02 Sep, 2024
Submission checks completed at journal
15 Aug, 2024
First submitted to journal
15 Aug, 2024
Unknown event
14 Aug, 2024
Editor assigned by journal
13 Aug, 2024

You are reading this latest preprint version

Role of the rostral anterior cingulate cortex in emotion processing in Treatment Resistant Depression

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Participants

MRI data collection

Task description

MRI data processing

Demographic and clinical characteristics

Functional connectivity (FC) analyses

Clinical factors

Results

Demographic and clinical characteristics

Whole brain seed-based connectivity analyses – rACC

Supraliminal processing:

Subliminal processing:

Clinical factors

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1