Personalised transcranial magnetic stimulation for treatment-resistant depression, depression with comorbid anxiety and negative symptoms of schizophrenia: a narrative review

Abstract

Transcranial magnetic stimulation (TMS) is a promising intervention for treatment-resistant psychiatric disorders. However, conventional TMS typically utilises a one-size-fits-all approach when determining stimulation targets. Recent retrospective brain circuit-based analyses using lesion network mapping have suggested that a left dorsal lateral prefrontal cortex target has a higher efficacy for alleviating depression symptoms, a dorsomedial prefrontal cortex target is more effective for anxiety symptoms, and a rostromedial prefrontal cortex target is effective for schizophrenia-associated psychiatric symptoms. Nonetheless, symptom-specific brain circuit targeting has not been tested prospectively. We conducted a narrative review of selected literature to investigate individualised targeting for TMS and discuss potential future directions to elucidate the efficacy of this approach.

INTRODUCTION

Transcranial magnetic stimulation (TMS) is a promising intervention for treatment-resistant psychiatric disorders. Recent retrospective brain circuit-based analyses using lesion network mapping have suggested that symptom-specific brain circuit targeting could be more effective for alleviating depression, anxiety and schizophrenia-associated psychiatric symptoms. Nonetheless, symptom-specific brain circuit targeting has not been tested prospectively. We conducted a narrative review of selected literature published in the past 20 years, with a focus on investigating individualised targeting for TMS and three domains: depressive symptoms, depression with comorbid anxiety symptoms and negative symptoms of patients with schizophrenia. The review also highlights an innovative TMS, individualised therapeutic strategies based on functional magnetic resonance imaging (fMRI) images and accelerated intermittent theta burst stimulation (iTBS) treatment parameters.

CLINICAL CHALLENGES

Treatment-resistant depression

Depression is a common mental disorder that affects more than 350 million people worldwide and confers a large burden of disease because of its early onset, high prevalence, profound disability and suicide death combined.[1] The 2016 Singapore Mental Health Study (SMHS) reported that the lifetime prevalence of depression among Singapore adults was 6.3% with a 12-month prevalence of 2.4%.[2] The economic burden of depression in the Asia Pacific region[3] and Singapore[4] is high, and up to 50% of indirect costs are associated with lost productivity and unemployment. Treatment of depression is estimated to give a benefit-to-cost ratio of 5.7-to-1,[5] while the current mean annual cost of treatment for depression in Singapore is USD 7638.[4] The current standard of care for depression remains suboptimal, with a large proportion of patients not responding satisfactorily to treatment and classified as having treatment-resistant depression (TRD).[6]

Depression and comorbid anxiety

Depression and anxiety are commonly comorbid. The lifetime global population prevalence of generalised anxiety disorder (GAD) is estimated to be 3.7%.[7] In the Singapore population, the estimated prevalence of GAD increased from 0.9% in 2013 to 1.6% in 2016.[8] The 2010 Global Burden of Disease study reported that 45.7% of individuals with lifetime major depressive disorder (MDD) had a lifetime history of one or more anxiety disorders.[9] The SMHS 2016[10] reported that approximately 27% of individuals with a lifetime history of MDD also have a history of anxiety disorder. Some studies with clinical registration data suggested that most outpatients with MDD present with comorbid GAD.[11,12]

Anxiety disorders and depression occur early in a person’s development, with anxiety disorders commonly beginning during childhood and early adolescence and major depression tending to emerge during adolescence and early to mid-adulthood.[13,14,15] In relation to the evolution of comorbidity, studies demonstrate that anxiety disorders generally precede the presentation of MDD.[14] Overlaps between anxiety and depression further complicate diagnosis and render treatment challenging. Compared to depressive patients without significant anxiety symptoms, patients with anxiodepression (comorbid depression and anxiety) often exhibit more severe depression symptoms, including suicidal ideation/attempt, poorer chronicity and more severe daily functional impairment.[12,16,17] Multiple pharmacological studies have demonstrated evidence of poorer MDD treatment outcome in depression with comorbid anxiety as compared to depression without comorbid anxiety.[16,18,19,20,21,22] In addition, greater frequency and intensity of drug side effects, and higher frequency of serious adverse effects have been reported in depressed patients with comorbid anxiety.[20] Comorbid anxiety with depression also predicts poorer outcomes with a higher percentage of treatment resistance, that is, inadequate or failed response to one or several drug trials, than either disorder occurring alone.[23]

Negative symptoms of schizophrenia

Schizophrenia is a severe mental disorder with a profound impact on patients, their families and caregivers, and the society. In Singapore, the prevalence of lifetime schizophrenia or other psychotic disorders was reported to be 2.3% in a national survey study conducted in 2018.[24] The health and economic burden of schizophrenia is massive due to the resources required to provide services to patients and other indirect costs derived from a loss of productivity of both patients and caregivers.[25] Moreover, schizophrenia is a heterogeneous condition with a variety of natural history trajectories[26,27] that generally show a pattern of deterioration and increasing treatment resistance with each trial of antipsychotics and recurrence of symptoms.[28]

While positive symptoms of schizophrenia reflect an excess or distortion of normal function (e.g., delusions, hallucinations, disorganised behaviour), negative symptoms are another core component of schizophrenia that are associated with significant diminution or absence of normal behaviours related to motivation and interest (e.g., avolition, inability to feel pleasure, social withdrawal) or expression (e.g., blunted affect, poverty of speech), which accounts for a large part of the poor functional outcome, cognitive impairment and long-term morbidity in patients with this disorder.[29,30,31] In a retrospective study of the onset of schizophrenia in 4707 patients seeking psychiatric assistance, negative symptoms were observed in 95% of the examined subjects.[32] Naturalistically, in Singapore, many patients with chronic schizophrenia have drug resistance and significant negative symptoms, resulting in poor functioning, unemployment and homelessness. Thus, other than the psychotic symptoms, which remain the focus of therapy for schizophrenia patients, there is also an urgent need for early intervention in schizophrenia to prevent such chronic sequelae of negative symptoms of schizophrenia.

We conducted a selective narrative review of pertinent literature in this field and present a summary of the recent literature in this area and upcoming local studies investigating these clinical challenges.

CURRENT TREATMENTS

The 20-year-old landmark naturalistic National Institute of Mental Health-funded study on sequenced treatment alternatives to relieve depression (STAR*D) revealed that the current standard of care for depression/anxiety was suboptimal, with only 47% of patients responding to standard pharmacotherapy[33] and the remainder being classified as TRD.[6] Real-world outcomes are even worse, with an estimated 15%[34] to 28%[35] of patients recovering from depression. There is limited evidence of augmentation strategies for treatment-resistant anxiety.[36] Furthermore, polypharmacy is common and likely increasing[37] in naturalistic settings, with only half of the drug prescriptions adhering to treatment guidelines[38] despite potential negative patient outcomes.[39]

For patients with schizophrenia, positive symptoms are generally managed effectively with available antipsychotic medications, whereas negative symptoms show little tendency towards spontaneous improvement in the course of the disease. Therefore, negative symptoms in schizophrenia remain a key unmet clinical need with little to no pharmacological or psychotherapeutic treatment approaches that have a significant effect size.[40,41] Moreover, the side effects of antipsychotics, such as sedation and metabolic changes, can further impact the patient’s functional outcomes and quality of life.[42] Cognitive behavioural therapy and psychosocial interventions are other approaches that have been explored for managing the negative symptoms of schizophrenia. While these treatments can provide some benefit in terms of social and occupational functioning, their impact on negative symptoms is often limited and requires prolonged and consistent engagement from patients.[43,44] Although adjunctive medications, such as antidepressants and stimulants, have been explored in the treatment of negative symptoms of schizophrenia, evidence supporting their efficacy is mixed, and these treatments may introduce additional side effects or interactions with other medications.[45,46] Given these limitations, there is a clear need for novel treatment options that can more effectively target negative symptoms without imposing additional burdens on these patients.

Neurostimulation is the use of electrical or magnetic stimulation of the brain to treat patients with mental disorders.[47] Among neurostimulation techniques, electroconvulsive therapy (ECT) is considered the gold standard treatment, yielding 50%–70% response in patients with treatment-resistant affective mental disorders or schizophrenic-related psychotic symptoms.[48] However, ECT is a relatively invasive procedure involving induction of a generalised seizure under general anaesthesia, and although highly effective, the use of ECT is limited by stigma and cognitive side effects.[49] Hence, it is usually considered only as a last resort when other treatment modalities have been tested and failed. Furthermore, ECT is time-consuming, with each treatment session (including the preparation time) lasting 90–120 min, and a total treatment course spanning over 6–12 weeks.

NON-INVASIVE NEUROSTIMULATION THERAPY

Transcranial magnetic stimulation

In recent decades, there is great interest in alternative, non-invasive brain stimulation therapies such as TMS. Transcranial magnetic stimulation involves using focal magnetic stimulation from an electromagnetic coil to induce action potentials of neurons in targeted cortical regions, most commonly the dorsolateral prefrontal cortex (DLPFC) for TRD. Developed in the 1980s, TMS gained attention for its potential in treating psychiatric disorders by targeting and stimulating certain neural circuits.[50] Repetitive TMS (rTMS) is effective in TRD[51,52,53] and is approved for clinical use in multiple countries (e.g., the USA, the UK, Australia, Singapore, Canada and Israel). It has an excellent safety profile.[54] Treatment of depression with rTMS is cost-effective,[55] and a local cost-effectiveness study shows that rTMS is a highly cost-effective option as compared to ECT for non-psychotic depression.[56]

Transcranial magnetic stimulation therapy involves using a coil to generate magnetic fields that induce electric currents in specific brain regions, modulating neuronal activity and promoting neuroplasticity. The TMS-induced electric currents in the brain can alter synaptic transmission and promote neuroplasticity, potentially improving mood and cognitive function.[57] Also, TMS is thought to increase the activity of underactive regions in DLPFC, which may lead to normalisation of the imbalance in prefrontal–limbic connectivity often observed in depressive disorders and negative symptoms of schizophrenia.[58,59] This modulation can lead to increased levels of neurotransmitters such as serotonin, norepinephrine and dopamine, which play critical roles in mood regulation and emotional processing.[60,61] In addition, TMS may enhance synaptic plasticity by promoting long-term potentiation and other mechanisms that strengthen neuronal connections, ultimately contributing to the alleviation of depressive symptoms and negative symptoms of schizophrenia.[62]

The primary benefit of TMS is its ability to promote neuroplasticity, the brain’s inherent capacity to reorganise and adapt through the formation of new neural connections.[63] This adaptive process can lead to improvements in symptoms, particularly those associated with negative symptomatology in schizophrenia and depression.[64] The ability of TMS to target specific brain regions has been a key factor in its effectiveness. Stimulation of this area, for instance, can enhance activity in the prefrontal cortex, which is often dysregulated in patients with negative symptoms.[65,66] This modulation then leads to improved emotional regulation, social functioning and cognitive processes. Furthermore, studies have suggested that TMS may increase the levels of neurotransmitters like dopamine and glutamate, further contributing to its therapeutic effects.[67]

One of the major advantages of TMS is its non-invasive nature. This makes it a preferable option for patients who may not tolerate the side effects associated with traditional pharmacological treatments. Treatments are generally well tolerated, with mild and transient side effects such as scalp discomfort or headaches. These are generally short-lived and can be managed effectively.[68] In addition to its therapeutic effects, TMS offers a patient-friendly treatment experience. Patients undergoing TMS can continue with their daily activities and do not require anaesthesia or hospitalisation, rendering TMS a convenient option for those who may be hesitant about invasive procedures or long recovery periods. Multiple published trials have also demonstrated that TMS can improve the general functioning of patients with schizophrenia, particularly for both positive symptoms like auditory hallucinations[69,70] and negative symptoms[71,72,73] of schizophrenia disorder. A recent meta-analysis that collected data from a large number of studies and patients provided aggregated evidence that TMS has a moderate but significant effect size for patients with schizophrenia suffering from negative symptoms.[74,75]

Theta burst stimulation

Theta burst stimulation, a subtype of rTMS, involves pulses being applied in bursts at high frequency with an interburst interval, matching the range of theta frequency. Existing protocol of TBS showed powerful effects on synaptic plasticity using repeated short bursts (three pulses) of high-frequency stimulation at 50–100 Hz given three to five times per second.[76,77] Continuous TBS is expected to suppress cortical excitability,[78] while iTBS is generally considered as facilitating neuronal activity.[79] The iTBS protocol is an extensively researched, US Food and Drug Administration (US FDA)-approved novel form of magnetic stimulation for the treatment of depression, which can be used to produce the same or greater physiologic effects on depressive symptoms as compared to standard rTMS, but in a markedly reduced period of time (3–10 min vs. 20–38 min of rTMS protocol for a standard session).[80,81,82] There is growing evidence to support the application of accelerated forms of iTBS to treat patients with TRD.[83,84,85] Pooled response rates of various non-invasive TMS treatment modalities (including accelerated iTBS) for TRD have ranged from 25.1%[86] to 46.6%.[87] In a recent study of the Stanford accelerated intelligent neuromodulation therapy (SAINT), highly accelerated (ten sessions per day for five consecutive workdays) iTBS was delivered to patients’ individualised localisation targets. Patients reported a remarkably high response rate (85.7% in the treatment group vs. 26.7% in the sham group) and remission rate (78.6% in the treatment group vs. 13.3% in the sham group) after only 1 week of treatment.[88] The SAINT^™ was approved by the US FDA for the treatment of MDD in adults in September 2022.

Meanwhile, although the data of iTBS treatment for anxiety symptoms remain scarce, recent trial results from functional near-infrared spectroscopy study suggest that iTBS offers promising treatment potential for anxiety symptoms in patients with concurrent depressive and anxiety diagnosis.[89] Thus, iTBS would appear to be an ideal intervention to treat both depression and anxiety with an accelerated protocol, whereby multiple daily sessions could be administered but still kept within a reduced amount of time. Currently, there is still limited evidence regarding the treatment effectiveness of accelerated iTBS for patients with schizophrenia.

Brain circuit-based TMS therapeutic approaches

Functional MRI is a type of imaging that shows activity in specific areas of the brain. It was initially used as a mapping device and has been proven to be essential for patients undergoing brain surgery. It not only allows neurosurgeons to precisely locate the intended surgical sites, considering the slightest individualistic variations, but also minimises the risk of potentially damaging any critical brain region. Recently, resting state fMRI has been used to study non-task-oriented functional connectivity, allowing researchers to investigate the interaction and correlations between various brain circuits.[90]

Emerging evidence suggests that individualised fMRI-based functional connectivity (connectome)-guided localisation, a type of precision medicine for TMS, may yield better response rates than anatomically guided localisation.[91] More specifically, the response rate may depend on exactly where one administers TMS in DLPFC.[92,93] Importantly, psychiatric disorders are now conceptualised as disorders of brain networks, not individual brain regions. Similarly, TMS is now conceptualised as a network therapy; although stimulation is typically applied to a single brain region, its effects are mediated via distributed networks. Advances in mapping brain networks and connectivity now allow us to identify these networks in individuals and potentially refine our therapeutic targets for mental disorders. Under the individualised connectome-guided framework, for any given individual, the region in DLPFC with the strongest negative correlation with the subgenual cingulate cortex (during resting state fMRI) is selected as the stimulation location. Such individualised connectome-guided target locations can potentially improve TMS response rates as compared to anatomic localisation.[94] In a post hoc analysis of TMS treatment for patients with depression, clinical response to TMS was significantly better when patients were serendipitously stimulated in closer proximity to individualised connectome-guided targets, whereas therapeutic outcome was unrelated to proximity to non-personalised group average stimulation targets.[95,96] A version of the individualised connectome-guided TMS approach (i.e., SAINT) was approved by the US FDA in September 2022.

In Siddiqi et al.,[97] the dorsomedial prefrontal cortex region that was proposed for TMS targeting for anxiety symptoms overlapped broadly with the default mode network, which has been extensively implicated in rumination, a core component of trait anxiety (i.e. anxiety that is part of someone’s personality or way of seeing the world).[98,99,100] Utilising the retrospective data analysis of brain functional connectivity map and TMS treatment outcome, the identified anxiosomatic target locations for TMS targeting also outperformed multiple other brain regions that have previously been targeted for anxiety treatment. However, the association of these novel TMS targeting spots with anxiety-correlated brain network needs to be validated with prospective TMS treatment data.

Advances in mapping brain networks and connectivity could also potentially refine our therapeutic targets for specific psychiatric symptoms. The same approach was used by Siddiqi et al.[97] to identify the group-level TMS brain targeting spot with a weighted combination of several TMS- and lesion-derived anxiety circuits. Individualised resting-MRI functional connectivity map could also be utilised to predict the individualised TMS targeting spot for patients with prominent anxiety symptoms and holds great potential to improve individual TMS treatment effectiveness.

Identifying the anatomy of circuits causal of psychosis could inform treatment targets for schizophrenia. Andrew et al.[101] identified a large set of published case reports of brain lesions that caused new-onset psychosis. A location in the rostromedial prefrontal cortex with high connectivity to this psychosis circuit was identified as a potential target for TMS. Considering the shared psychopathology of psychotic symptoms and negative symptoms of schizophrenia, we suspect the rostromedial prefrontal cortex could also be an ideal TMS brain circuit target for negative symptoms.

CURRENT GAPS AND FUTURE RESEARCH

Current individualised connectome-guided TMS protocols have been primarily developed and validated within Western populations, leaving their applicability to Asian populations untested. This raises concerns about the generalisability of these protocols due to potential ethnic and racial differences in resting state fMRI. Therefore, it is essential to investigate the efficacy of such approaches before they can be adopted as clinical services in Singapore. In addition, resting state fMRI at the individual level is known to suffer from poor reliability, which could impact the accuracy of TMS target acquisition. However, advancements have been made in this area with the development of a multisession hierarchical Bayesian model (MS-HBM).[102,103] This model estimates brain networks in individual participants from resting state fMRI data, accounting for both within-subject and between-subject variability. The MS-HBM networks, generated from just 10 min of data, are reportedly as reliable as other methods requiring 50 min of data.

Further research could focus on understanding the reliability of target locations when existing individualised connectome-guided localisation approaches are combined with MS-HBM. The MS-HBM could be used to divide the cerebral cortex of participants into individual-specific networks. In addition, it is important to consider how the dorsal and ventral attention networks are negatively correlated with the subgenual cingulate cortex. This suggests that combining the largest DLPFC components of these networks into a single region of interest (ROI) within DLPFC could be effective. As gyri are closer to the scalp where iTBS stimulation is applied, the MS-HBM ROI could be refined by limiting ROIs to gyral regions. The final stimulation location could then be chosen based on the highest negative correlation to the subgenual cingulate cortex.

Numerous research designs and clinical trials could stem from these findings to better understand treatment outcomes for brain circuit-based individualised precision TMS for psychiatric disorders. We have initiated an open-label clinical trial for investigating the treatment outcomes and cost-efficiency of an individualised connectome-guided accelerated iTBS approach for treating TRD. Furthermore, we are also conducting a randomised, double-blind clinical trial comparing the efficacy of individualised connectome-guided localisation against traditional anatomically guided localisation for accelerated iTBS in Asian patients with TRD. In this study, the treatment arm would undergo accelerated iTBS to individualised connectome-guided target locations based on the MS-HBM approach, while the control arm would use the traditional anatomically guided (Beam F3) approach. Consistent treatment protocols and fMRI scans before and after treatment would help identify any significant differences between the treatment arms.

Further TMS clinical trials could explore its effectiveness in treating negative symptoms of schizophrenia, considering the global dysconnectivity and aberrant networks observed in these patients. In an ongoing local pilot study for patients with persistent negative symptoms, we are investigating stimulation to the rostromedial prefrontal region alongside DLPFC. An accelerated iTBS protocol could be implemented to enhance therapeutic benefits and reduce treatment duration. Should this approach prove effective, it could transform the treatment of psychotic disorders in clinical settings.

In addition, personalised iTBS treatment for comorbid anxiety symptoms in patients with depression warrants investigation. There is currently no established protocol for TMS efficacy in MDD with comorbid anxiety. A study could compare two treatment arms receiving TMS stimulation to either dorsomedial prefrontal cortex or left-dorsal lateral prefrontal cortex, assessing the treatment outcomes for anxious versus depressive symptoms. Functional MRI scans before and after treatment would help determine ROI and examine brain connectivity changes. The primary outcome measure could be improvements in anxiety symptoms, assessed using the Beck Anxiety Scale.

Future studies could also explore whether symptom-specific TMS targeting induces circuit-specific changes in brain connectivity, correlating individualised TMS targets with better symptom improvements. Long-term modulation of network dynamics induced by rTMS trains could be another area of interest. By analysing functional connectivity changes pre- and post-treatment and correlating them with psychiatric symptoms, these studies could provide new insights into the neural mechanisms underlying mental disorders and recovery.

CONCLUSION

Individualised brain circuit-based TMS could be utilised to address pragmatic and scientific considerations with regards to treatment-resistant mental disorders — a major international and local public health challenge, in addition to being a leading cause of years lived with disability. Furthermore, iTBS has the potential to be a paradigm changing treatment for depression, anxiety and schizophrenia. It could even achieve Porter’s triple aim[104] of improving the individual experience of care, improving population health and reducing the per capita cost of care. Porter’s triple aim is achievable due to the significantly improved tolerability of iTBS, as compared against ECT (current gold standard treatment for depression), within a fraction of the time (1 week[105] vs. 4 weeks of standard ECT treatment course[106]), with an equal or superior clinical treatment efficacy and greater cost-effectiveness.

On a more intangible level, precision TMS treatment has the potential to change the population’s mindset of psychiatric illness treatment from a chronic disease model (requiring months to years of treatment) to a rapid procedural approach. This then decreases the stigma towards psychiatric treatment and encourages earlier and more effective population-level treatments for psychiatric and other neurological diseases such as Parkinson’s and Alzheimer’s disease.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.