This study provides new insights into the relationships between E-field strength, functional neuroplasticity, and clinical response to ECT. The key findings include changes in cerebro-cerebellar functional connectivity in association with the E-field. Specifically, two cerebro-cerebellar FNC decrease and seven cerebro-cerebellar FNC increase as E-field strength increases. Further, the Ebrain-related cerebro-cerebellar FNC is associated with both antidepressant outcomes and cognitive side-effects following ECT. Functional neuroplasticity through cerebro-cerebellar FNC mediates the effects of Ebrain on both antidepressant outcomes and cognitive side-effects. Our results demonstrate that E-field strength is associated with ECT-mediated cognitive impairment, specifically verbal dysfluency, through increased FNC between the cerebellum and MOG and decreased FNC between the cerebellum and PCC. E-field strength is also associated with antidepressant outcomes through increased FNC between the cerebellum and IPL.
Mechanisms of ECT-related neuroplasticity, including neurogenesis, angiogenesis, synaptogenesis, and gliogenesis may be associated with the changing electric field (Bouckaert et al., 2014b). The neurogenic hypothesis posits that the depressive brain has an impairment of producing new neurons for proper mood control, and generating new neurons is beneficial for antidepressant efficacy (Scott et al., 2000; Petrik et al., 2012). However, it remains heavily debated (Boldrini et al., 2018; Nogueira et al., 2018) that neurogenesis cannot be supported by existing ECT-imaging literature given the short time frame of the ECT series and widespread structural neuroplasticity (Nordanskog et al., 2010; Abbott et al., 2014; Gbyl and Videbech, 2018; Takamiya et al., 2018; Van Den Bossche et al., 2019). Our results for the first time demonstrated the relationship between E-field strength and functional neuroplasticity. We speculate that the increased brain volumes might change the capabilities and behaviors of neurons to communicate, as reflected by the alterations of functional connectivity between brain regions (Argyelan et al., 2019). Synaptogenesis and functional remodeling are possible mechanisms that may be compatible with both structural and functional neuroplasticity (Chen et al., 2009).
Using individualized connectomes, previous studies have found a set of functional connectivity patterns affected by ECT, whose changes directly or indirectly impact cognitive performance (Wang et al., 2020; Wei et al., 2021). These changes in connectivity involve a variety of functional networks, including the cognitive-control network, the default-mode network, and the cerebellum network (Perrin et al., 2012; Wang et al., 2020). Interestingly, both increased and decreased cerebro-cerebellar connectivity have been observed after ECT, indicating the heterogeneous patterns of ECT modulated functional connectivity changes. The changed temporal coherence of functional connectivity results may be suggestive of synaptic remodeling, in which E-field affects the neurons to modulate the information flow by adopting polarized morphologies (Jonckheere et al., n.d.; Small et al., 2011).
In our study, cognitive outcomes were measured by the DKEFS Verbal Fluency Test, with a primary focus on letter fluency (e.g., phonemic fluency) (Mueller et al., 2015). The cerebellum shows a robust connection to many cognitive and affective cerebral structures (Sang et al., 2012), especially the default-mode network, a system responsible for self-referential information processing and memory (Wang et al., 2020). Consistent with these findings, our results on the association between cognitive performance and decreased PCC-cerebellum connectivity provide further evidence that the default-mode network to cerebellum connectivity might play a direct role in regulating cognitive function (Leech and Sharp, 2014; Habas, 2021). The biological underpinnings of ECT-induced cognitive impairment could be related to the disruption of the cerebro-ponto/reticulo-cerebellar-thalamocortical loops caused by the broken communication between the default-mode network and cerebellum (Habas, 2021).
Another interesting observation in our study is the laterality of ECT-modulated FNC neuroplasticity. Changes in FNC between the right cerebellum and PCC and between the cerebellum and right IPL are associated with cognitive and antidepressant outcomes, respectively. The lateralization of brain changes has been documented in ECT literature (Abbott et al., 2014; Argyelan et al., 2019; Cano et al., 2019; Sartorius et al., 2019), although its underlying mechanisms are still unclear. Considering the right laterality of the electric field (Argyelan et al., 2019; Deng et al., 2022), our findings might suggest a potential linkage between electric field laterality and functional connectivity laterality. Numerous studies believe that the laterality of brain changes might be due to the predominant side of stimulation (Tendolkar et al., 2013; Abbott et al., 2014). We speculate that the E-field-related FNC lateralization might also be associated with the major depressive episodes themselves since functional brain abnormalities in the right hemisphere have been widely reported in the mood- and stress-related disorders (Sartorius et al., 2019).
Our investigation further demonstrated that cerebro-cerebellar FNC is a potential mediator between the E-field and clinical outcomes. Increased E-field strength is associated with decreased FNC between the cerebellum and PCC, which results in cognitive impairment after ECT. Meanwhile, increased E-field strength is associated with decreased connectivity between the cerebellum and right IPL, which results in improved antidepressant outcomes. These mediation effects show similar and unique patterns compared with previous investigations that focused on structural neuroplasticity (Argyelan et al., 2019; Deng et al., 2022). On one hand, both functional and structural neuroplasticity show mediation effects on the association between E-field strength and %ΔHDRS. On the other hand, while the link between structural neuroplasticity and cognitive impairment is less robust, functional neuroplasticity significantly mediates the effect of E-field strength on the DKEFS Letter Fluency Test score. These findings may suggest complementary but differential processes of functional and structural neuroplasticity (Abbott et al., 2014). Additional pre-clinical studies with advanced analytical methods such as data fusion are needed to elucidate the mechanistic link between electric field, functional neuroplasticity, structural neuroplasticity, and clinical outcomes.
There are several limitations in the present study that might influence result interpretation. First, as an important therapeutic component of ECT, seizure activity might impact the effects of the electric field on functional neuroplasticity and clinical outcomes (Segi-Nishida, 2011; Miller et al., 2022). Our investigation did not assess the potential relationships between seizure activity on functional connectivity changes and clinical outcomes. The possible role of seizures on functional neuroplasticity is still unclear. Future studies can employ similar mediation models with topographical ictal power to unveil the impact of E-field and ictal power on ECT’s therapeutic and iatrogenic effects. Second, our study included adult patients with MDD with or without psychotic features and two electrode placements (RUL or BT). Our sample size was underpowered for the investigation of the group- or electrode-specific changes in functional neuroplasticity. In future studies with larger sample sizes, we can assess the effects of psychosis and electrode placement on functional neuroplasticity. We can also confirm and probe the relationships between electric field, functional neuroplasticity, and clinical outcomes within each sub-group. Third, subjects recruited in our study discontinued antidepressant medications, but did receive as-needed medications for anxiety and sleep. Larger samples with different medications will be needed to assess these complex relationships with both antidepressant and cognitive outcomes. Fourth, our present study only focused on the functional neuroplasticity that is related to the electric field and ECT responses. Existing work has suggested potential relationships between structural neuroplasticity, electric field, and clinical outcomes (Argyelan et al., 2019; Deng et al., 2022). Despite these findings, the relationship between structural and functional neuroplasticity in the context of ECT response remains unknown. Structural and functional neuroplasticity might play overlapping and complementary roles in antidepression-related circuitry. In future studies, advanced fusion analysis tools can be combined with mediation analysis to elucidate more details of the mechanisms underlying structural and functional neuroplasticity induced by ECT.
In conclusion, this investigation provides further support for the construct of optimal ECT dosing in relation to the E-field. Increased E-field influences cerebellar-cerebral FNC and improves antidepressant outcomes. However, increased E-field also changes other cerebellar-cerebral FNC and adversely impacts cognitive performance. The relationship between E-field and cognitive performance may serve as an upper limit for determining ECT dosing. With pre-ECT imaging, an individual’s E-field may be calculated prior to the ECT series based on the given electrode placement. E-field informed ECT has the potential to eliminate the trial-and-error dosing strategy (RUL with BT contingency) that was utilized in this investigation. An individualized amplitude may be then calculated to optimize antidepressant outcomes and minimize cognitive risk.