Multilevel Neuroimaging Insights Into Electroconvulsive Therapy Mechanisms in Clozapine-Refractory Schizophrenia: A Comparative Perspective With Depression

A recent study by Roy and colleagues (1) in the previous issue of Biological Psychiatry: Global Open Science explored the putative mechanisms of electroconvulsive therapy (ECT) in clozapine-refractory schizophrenia using both structural and functional magnetic resonance imaging (MRI), identifying changes in amygdala volume and posterior cerebellar connectivity as correlates of positive symptom improvement. In this commentary, we examine the relevance of these findings within the broader context of ECT as a treatment for schizophrenia and consider their implications in light of comparable ECT research in depression. Finally, we highlight the innovative integration of structural and functional neuroimaging in their analysis.

Neurobiological Mechanisms Behind ECT in (Clozapine-Refractory) Schizophrenia

While depression remains the primary indication for ECT in most Western countries, schizophrenia—especially cases resistant to pharmacological treatment—has historically been the leading indication in many Asian clinical contexts. This practice is increasingly supported by contemporary evidence demonstrating that ECT is effective in alleviating positive symptoms in individuals with treatment-resistant schizophrenia, including those who have not adequately responded to clozapine (2).

Although research on ECT-related neuroplasticity in schizophrenia is less extensive than in depression, accumulating evidence highlights both structural and functional brain changes after ECT (3). Structural MRI studies have consistently reported increased gray matter volumes in the hippocampus, parahippocampal gyrus, amygdala, insula, and frontotemporal regions. Notably, these volumetric increases—particularly in the hippocampus and insula—seem specific to ECT rather than antipsychotic treatment alone. Functional MRI findings of populations with schizophrenia demonstrate ECT-induced modulation of large-scale brain networks, including increased connectivity within the default mode network (DMN), especially in the medial prefrontal and posterior cingulate cortices. Simultaneously, ECT seems to reduce aberrant connectivity within the salience network and between limbic structures such as the amygdala and hippocampus. Enhanced prefrontal-hippocampal connectivity and integration within the DMN have been linked to symptomatic improvement, particularly in general psychopathology and positive symptoms. These results support the hypothesis that ECT facilitates a recalibration of dysregulated network dynamics in schizophrenia through neuroplastic processes.

Importantly, the recent study by Roy et al. (1) builds on and extends this literature by implementing a multilevel, within-subject design that combines structural and functional MRI in a cohort of patients with clozapine-refractory schizophrenia. Their findings of concurrent increases in amygdala volume and reductions in posterior cerebellar hyperconnectivity—both of which were significantly associated with clinical improvement—add valuable mechanistic insight. Unlike most earlier studies using one singular imaging modality, Roy et al. (1) uniquely bridge these levels of analysis within a single population and time frame. This approach not only reinforces the role of limbic structures but also introduces the cerebellum as a potential actor in the therapeutic mechanism of ECT.

Comparative Insights From ECT Research in Schizophrenia and Depression

ECT has been shown to induce structural and functional brain changes across psychiatric disorders, although emerging evidence highlights both shared and disorder-specific effects in schizophrenia and depression. One of the key findings in schizophrenia is the volumetric enhancement of the amygdala after ECT, which is significantly associated with clinical improvement. This mirrors earlier observations in depressive cohorts, where increases in gray matter volume in the medial temporal lobe have frequently been reported after ECT (4,5). However, while the amygdala enlargement in schizophrenia seems to correlate more frequently with symptom reduction, the relationship in depression remains less robust, with some studies indicating only transient changes unrelated to sustained clinical benefits (6). Such volumetric increases in emotion-related structures such as the amygdala and hippocampus suggest a shared neuroplastic mechanism, potentially reflecting the remodeling of limbic circuits involved in affective and psychotic symptomatology. Nevertheless, the mechanistic role of the amygdala may be more pronounced in schizophrenia, particularly in patients with clozapine-refractory schizophrenia, in which dysregulation of dopaminergic and limbic pathways is a hallmark feature (7).

Functional neuroimaging findings further underscore this divergence. Roy et al. (1) reported ECT-induced reductions in posterior cerebellar connectivity in schizophrenia, which correlated with improvements in positive symptoms. This pattern contrasts with findings in depression, where cerebellar connectivity changes seem more variable and are often not directly associated with clinical outcomes (8). In addition, the observed post-ECT increase in right frontoparietal connectivity in clozapine-refractory schizophrenia diverges from the typical pattern in depression, where decreases in frontoparietal network connectivity are more commonly reported (8). These contrasting trends emphasize the importance of interpreting functional imaging results within the neurobiological context of each disorder, rather than assuming a uniform model of ECT response.

Taken together, studies investigating ECT across schizophrenia and depression cohorts reveal overlapping patterns—such as gray matter increases in the hippocampus and amygdala—and important differences in connectivity dynamics. While shared structural effects may point to common mechanisms of neural plasticity, differences in functional connectivity, particularly within the cerebellum and frontoparietal networks, suggest disorder-specific pathways of symptom modulation. Most importantly, only in schizophrenia has the amygdala volume increase demonstrated a consistent, linear relationship with clinical response, further supporting the notion of differential neurobiological underpinnings of ECT efficacy across diagnostic categories.

The Added Value of Multimodal Neuroimaging

A major strength of the Roy et al. (1) study lies in its integration of structural and functional MRI within the same individuals undergoing ECT for clozapine-refractory schizophrenia. This multilevel design offers a comprehensive window into how therapeutic effects may unfold across different neurobiological domains simultaneously. While many studies examine either structural or functional outcomes in isolation, the ability to relate volumetric increases in emotion-relevant structures such as the amygdala with concurrent reductions in dysfunctional posterior cerebellar connectivity adds crucial mechanistic insight. In particular, increased amygdala volume may reflect ECT-induced structural neuroplasticity, while decreased cerebellar hyperconnectivity points to a parallel functional recalibration of aberrant network dynamics. This dual-layered evidence reinforces the idea that ECT induces system-wide neuroadaptation, with both local and network-level consequences that are measurable within a shared timescale.

The approach of Roy et al. (1) aligns with the broader perspective put forward in our recent review (9), which advocates for a multilevel conceptualization of ECT-induced plasticity encompassing molecular, structural, and functional domains. Although most studies combine these levels across heterogeneous samples or time points, the within-subject, cross-domain approach by Roy et al. (1) exemplifies this innovative methodological approach. Their use of both inductive (whole-brain, hypothesis-free) and deductive (targeted, hypothesis-driven) frameworks further strengthens the internal validity of their findings and helps differentiate true therapeutic mechanisms from nonspecific seizure effects or imaging artifacts.

Future research should prioritize this type of multilevel, multimodal design, particularly within clearly defined clinical subpopulations, such as those with clozapine-refractory schizophrenia or difficult-to-treat depression. Doing so will not only refine our understanding of the diverse neurobiological effects of ECT but may also help identify predictive markers of treatment response and disentangle the etiopathogenesis across diagnostic categories. Longitudinal follow-up studies are particularly needed to determine the persistence of these multilevel changes and their relevance for sustained clinical improvement.

Limitations and Opportunities for Future Research

While the study by Roy et al. (1) provides valuable mechanistic insights into ECT effects in clozapine-refractory schizophrenia, several limitations merit consideration. First, the sample size—although adequate for an exploratory design—may not fully account for the clinical and neurobiological heterogeneity characteristic of this patient population. Second, the absence of longitudinal follow-up limits conclusions about the temporal stability and clinical relevance of the observed neuroimaging changes. In depression, for instance, multiple studies including a large-scale mega-analysis from the Global ECT-MRI Research Collaboration (GEMRIC) consortium have shown that ECT-induced gray matter volume increases are transient and normalize after the treatment is terminated (10). Another key challenge lies in interpreting the functional connectivity findings. Although the reduction in posterior cerebellar connectivity was associated with clinical improvement, the nature of this relationship remains correlational. To advance our understanding of causality, future studies could explore the use of neuromodulatory interventions such as targeted transcranial magnetic stimulation to the posterior cerebellum. Finally, larger multicenter studies with replication cohorts of patients with schizophrenia undergoing ECT are essential to validate these preliminary results and to generalize findings across clinical settings.

Conclusions

The study by Roy et al. (1) offers important mechanistic insights into how ECT may alleviate symptoms in clozapine-refractory schizophrenia by inducing concurrent structural and functional brain changes. Their multilevel, within-subject design highlights the value of integrating imaging modalities to capture the multilevel complexity of the neurobiological effects of ECT. When viewed alongside research in depression, the findings highlight both shared and disorder-specific neuroplastic responses.