The Use of ECT and MST in treating depression

Abstract

Electroconvulsive therapy (ECT) has been used clinically since 1938. Its most common use is in the treatment of depression: first line treatment where rapid recovery is a priority, but more frequently as an effective treatment for patients who do not respond to pharmacological and psychological approaches. Whilst it is widely hailed as an effective treatment, concerns about its effect on cognition remain. The development of magnetic seizure therapy (MST) over the past decade has attempted to devise a therapy with comparable efficacy to ECT, but without the associated cognitive side effects. The rationale for this is that MST uses magnetic fields to induce seizures in the cortex, without electrical stimulation of brain structures involved with memory. MST has been used successfully in the treatment of depression, yet there is a dearth of literature in comparison with ECT. We present a systematic review of the literature on ECT (from 2009–2011) and MST (from 2001–2011).

Two take home points

1. ECT is a safe and effective treatment for depression.

2. MST shows potential as an alternative to ECT, with fewer cognitive side effects.

Two key future directions

1. Greater understanding of the mechanism of seizure therapy.

2. Larger, randomized studies to determine efficacy and safety of MST.

Introduction

Major depression is a leading cause of morbidity and mortality, with a lifetime prevalence of 6–12% (Brunoni et al., 2010). Psychological and pharmacological treatments are effective for some, but not all patients, and there is also the issue that delay in response to treatment can leave patients at serious risk of suicide or self-neglect, such as dehydration and starvation. For those patients who do not respond to medication initially, the chance of successful remission is reduced when a second or third drug is tried (Rush et al., 2006; Trivedi et al., 2006a, 2006b). There are many reasons for treatment resistance apart from ineffective treatment, including current social or environmental stressors and incorrect diagnosis. Recent papers have even questioned the efficacy of anti-depressant medication, intriguingly arguing that in more severe illness it is placebos that are less effective, rather than antidepressants that are more so (Kirsch et al., 2008). For those who do not respond to psychological and pharmacological approaches, or who need treatment more rapidly, alternative therapeutic options are needed. Treatment resistance among depressed patients was associated with 40% higher medical care costs compared with those who responded to treatment (Gibson et al., 2010). Early identification and alternative therapies such as ECT or MST may therefore not only benefit individuals, but also the wider society (Gibson et al., 2010).

ECT was first introduced in the 1930s with the aim of inducing seizures to treat severe mental illness (Bolwig & Fink, 2009; Merkl et al., 2009). In the 1950s, concerns regarding side effects together with the availability of new drug treatments led to waning popularity and even abolition of ECT in some countries and US states. Despite its controversial history, ECT has been used safely and successfully in clinical practice, and it remains a rapid and highly effective treatment for depressive disorders, particularly when other treatments have failed (Andrade et al., 2010; Carney et al., 2003; Scott, 2004).

In spite of limited evidence, cognitive side effects remain a real and understandable concern for patients (Fraser et al., 2008). They continue to be a crucial driver for attempts to find alternative physical treatments which may provide equivalent (or superior) efficacy, with lower cognitive side effects (Fitzgerald, 2008). In magnetic seizure therapy (MST) a coil is used to generate high frequency (100 Hz), repetitive trans-cranial magnetic stimulation (rTMS) (Kayser et al., 2011; Morales et al., 2004; Rowny et al., 2009). This leads to depolarization of neurons, causing focal induction of a generalized seizure. The development of this therapy promises the possibility of an effective treatment for depression, without the cognitive side effects seen in ECT. We review the literature of the last 10 years and consider whether this is a feasible and realistic possibility.

Methods

ECT

We used a systematic approach to review recent literature on ECT. Three databases (MEDLINE, Embase and PsycINFO) were searched from January 2009–April 2011. Reference lists were searched for additional studies. We used the following Boolean search terms: [ECT OR electroconvulsive therapy] AND [depression OR depressive]. The search was limited to English language papers on human subjects. The initial search identified 643 papers, reducing to 250 when duplicates were removed (Figure 1). Following a review of titles and abstracts, 147 were excluded, leaving 103 papers. Papers were excluded if they did not focus on ECT and depression; single case reports, brief letters and papers on anaesthesia in ECT were also excluded. Thirteen papers were excluded after reading the full text. This review includes a summary of information from the remaining 90 papers. As this is a synthetic review, where relevant we included major papers preceding this time period in order to present a broad overview of recent literature.

Figure 1. Search results: ECT.

MST

Three databases (MEDLINE, Embase and PsycINFO) were searched from January 2001–June 2011. Reference lists were searched for additional studies. We used the following Boolean search terms: [MST OR Magnetic seizure therapy] AND [depression OR depressive]. The search was limited to English language papers on human subjects. Unpublished studies and repeat populations were not included. We placed no limitation on age of subjects. The initial search identified 137 papers, reducing to 78 when duplicates were removed (Figure 2). Following a review of titles and abstracts, 36 were excluded leaving 42 papers. Papers were commonly excluded, as ‘MST’ also referred to alternative acronyms, e.g. multi systemic therapy. Following review of full text papers a further 20 papers were excluded, these were most frequently review papers which duplicated citations found in the initial search, without adding novel information. Five further papers were included following hand search of reference lists. This review includes a summary of information from the remaining 27 papers.

Figure 2. Search results: MST.

Results

ECT

Is it effective?

ECT is a safe, effective treatment for depression (Carney et al., 2003; Keltner & Boschini, 2009; NICE, 2003; Scott, 2004). Response rates are greater than 80%, with even higher rates observed in those with psychotic depression (Petrides et al., 2001). Patients recover quickly using ECT, with a remission rate of greater than 60% within 3 weeks (Bailine et al., 2010). There is somewhat reduced efficacy for those who have failed to respond to pharmacotherapy; when used in patients who have failed to respond to one or more adequate antidepressant medication trials, ECT response rates are between 50–60% (Heijnen et al., 2010; Kennedy et al., 2009). ECT not only leads to improvement in mood, but also in functional status and self-rated anxiety (Berg & Kononova, 2009; Langius-Eklof & Samuelsson, 2009). Quality of life improves following a course of ECT when rated by both patients and their physicians (Antunes & Fleck, 2009).

Despite an evidence base in its favour (Carney et al., 2003), there is still some opposition to the use of ECT (Read & Bentall, 2010). The methodology in these papers differs considerably. Carney et al. (2003) designed a systematic overview and meta-analysis of randomized controlled trials and observational studies, obtaining data from a wide range of sources including the Cochrane Collaboration Depressive Anxiety and Neurosis and Schizophrenia Groups, Controlled trial registers, Cochrane Controlled Trials register, Biological Abstracts, CINAHL, Embase, LILACS, MEDLINE, PsycINFO, and SIGLE, reference lists, and specialist textbooks. Their main outcome measures were quantitative, based on depressive symptoms, measures of cognitive function, and mortality. In contrast, Read and Bentall (2010) provide only a qualitative review, based on a search of two databases (MEDLINE and PsycINFO), one independent review and one textbook. In their literature review, Read and Bentall (2010) cite placebo controlled studies which showed minimal effectiveness for ECT in depression, with no evidence of benefits beyond the treatment period. They conclude that there is no scientific justification for the use of ECT when side effects and cost-benefit analysis are considered; whilst these are realistic concerns there is limited evidence to support this claim. Additionally, they raise the somewhat unrealistic concern that there have been no placebo-controlled studies to evaluate whether ECT prevents suicide (Read & Bentall, 2010). However, given that suicide is such a rare event, epidemiological rather than randomized control trial (RCT) evidence is usually adduced.

One difficulty in ECT research has indeed been the absence of a control with sham ECT, or a higher than expected sham response rate in some studies (Rasmussen, 2009). Whilst some patients may agree with an overall negative analysis, a qualitative review of ECT demonstrated a generally positive attitude towards ECT, with satisfaction independent of treatment outcome (Malekian et al., 2009).

What is the mechanism of action?

ECT increases post-synaptic 5-HT(1A) receptor signalling (Savitz et al., 2009). Whilst serotonergic and noradrenergic mechanisms have been suggested as mechanisms of anti-depressive action for ECT, studies using serotonin and catecholamine depletion following ECT do not induce relapse of depressive symptoms (Cassidy et al., 2009). This and related studies suggest that catecholamine and serotonin availability may not be necessary to mediate the anti-depressant response and benefit of ECT (Cassidy et al., 2010). Furthermore, investigation of 5-HT1A receptor binding using positron emission tomography (PET) imaging, found no difference in patients before or after ECT (Saijo et al., 2010a) suggesting the involvement of other neurotransmission mechanisms (Saijo et al., 2010a). Down-regulation of 5-HT2 receptors following ECT may, for example, explain efficacy in those with antidepressant refractory depression (Yatham et al., 2010).

ECT increases the level of key molecules which may stimulate brain plasticity, for example brain-derived neurotrophic factor (BDNF) which is involved in the neurogenesis and numbers of synapses in several brain areas (Castren & Rantamaki, 2010). Some studies have found no association between improvement in depressive psycho-pathology after ECT, and changes in serum BDNF or other neurotrophic factors (Fernandes et al., 2009; Gronli et al., 2009). However, there have also been reports of positive association between serum BDNF and therapeutic efficacy of ECT for depression (Piccinni et al., 2009; Viikki et al., 2010a). Activation of glial cells may play a part in mediating the antidepressant effects of ECT (Palmio et al., 2010), and successful ECT may also be associated with elevated serum glial cell-line derived neurotrophic factor (GDNF) levels (Zhang et al., 2009). Other molecules have also been investigated; ECT induces erythropoietin, which independently exhibits antidepressant-like efficacy and regulates the expression of genes implicated in antidepressant action (Girgenti et al., 2009). Yet related studies have not provided evidence for the mechanism of action (Valevski et al., 2010).

ECT increases the metabolism and blood flow in the anterior cingulate cortex (ACC), and electroencephalography (EEG) data have shown a normalization of theta activity following ECT (McCormick et al., 2009b). Given that the ACC generates theta rhythms, which are hypoactive in patients with depression and psychotic disorders, this may explain ECT's antipsychotic effect (McCormick et al., 2009b).

Neuroimaging studies have been employed to investigate mechanisms of action. An MRI study of 12 patients found that hippocampal volumes increased significantly after ECT, supporting the hypothesis that hippocampus may play a central role in the treatment of depression (Nordanskog et al., 2010). A positron emission tomography (PET) study showed that electroconvulsive therapy decreased D(2) receptor binding in the rostral anterior cingulate in depressed patients responding to ECT (Saijo et al., 2010b).

Whilst explanations of the mechanism of action of ECT have focused on studies of brain metabolism and neurochemistry, alternative hypotheses have been suggested; for example, ECT may be effective because it facilitates and restores the function of specific memory systems that are deficient in the course of a severe depressive episode (Frais, 2010). ECT also increases vagal activity which may be associated with the beneficial effect seen following this treatment (Bar et al., 2010).

In summary, multiple different mechanisms are likely to be responsible for the mechanism of ECT action. This may include regulation of cortico-limbic circuits, inhibition of neurotransmitter systems and monoamine neurotransmitters (norepinephrine, serotonin and dopamine), endocrinological pathways and neurogenesis (Merkl et al., 2009).

When is it indicated?

The major indication for ECT is in treatment resistant depression or for those who require rapid amelioration of symptoms due to high risk of neglect (NICE, 2003; Shelton et al., 2010). The risks associated with general anaesthesia and its poor public image explain why ECT is not generally a first line treatment for depression; however, it has been recommended for first-line treatment of depression where there is rapidly deteriorating physical status, repeated medication intolerance and acute suicidal ideation (Brunoni et al., 2010; Kennedy et al., 2009). It is effective for treatment of depression across all ages (Birkenhager et al., 2010), but may be particularly useful for elderly patients with psychotic depression (Moksnes & Ilner, 2010). It is effective and well tolerated in geriatric depressed inpatients regardless of pre-existing cognitive impairment (Hausner et al., 2011).

ECT can be used to treat both unipolar and bipolar depression without precipitation of mania (Bailine et al., 2010; Grunze et al., 2010; Medda et al., 2009; Popeo, 2009; Sienaert et al., 2009b; Yatham et al., 2009) and can be used in those presenting with mixed affective state (Medda et al., 2010). It is particularly useful for those who are severely depressed and have failed to respond to combined treatment, those who are severely suicidal (Fountoulakis, 2010), and those with severe psychomotor retardation (Grunze et al., 2010). ECT may be less effective in bipolar I disorder: Medda et al. found remission rates (defined as having a final Hamilton-depression score < 8) were significantly higher in the unipolar group (70.6%), than in bipolar II (43.3%) (OR = 2.0, 95%CI = 1.0–9.9) and in bipolar I disorder (34.8%) (OR = 2.4, 95%CI = 1.3–15.0) (Medda et al., 2009). One further study reported poorer subjective response to ECT than in unipolar disorder (Hallam et al., 2009). Further research is planned to investigate whether ECT is better than medication in treatment resistant bipolar depression (Kessler et al., 2010).

A review of ECT in pregnancy focusing on 339 cases showed good efficacy for treating depression (Anderson & Reti, 2009), with low rates of fetal and maternal complications (11 developed fetal or neonatal complications related to ECT and 18 developed maternal complications). This confirms that ECT may be useful in pregnancy when rapid reduction of symptoms is a priority, or when pharmacotherapy may have adverse implications for the fetus or for lactation. However, whilst the absolute risks are low, use of ECT should follow a careful risk–benefit assessment.

How should treatment be delivered?

ECT involves either unilateral, bitemporal or bifrontal stimulation of the brain with electric current in order to induce a seizure. Unilateral and bitemporal stimulation is most common, and bifrontal stimulation is under-used in clinical practice. This is performed under brief general anaesthesia, with use of muscle relaxants to minimize muscle contractions (and associated injury). Treatment typically takes place 2–3 times per week, comprising 6–12 sessions in total (NICE, 2003).

A multi-centre randomized, double-blind, controlled trial investigated bifrontal, bitemporal and right unilateral electrode placement in ECT (Kellner et al., 2010a). All three electrode placements resulted in clinically and statistically significant antidepressant outcomes: remission rates were 55% (95%CI 43–66%) with right unilateral, 61% with bifrontal (95%CI 50–71%) and 64% (95%CI 53–75%) with bitemporal electrode positioning. These results concurred with earlier studies that bilateral electrode placement is faster and more effective than unilateral (Carney et al., 2003; Kennedy et al., 2009). However, previous studies found that though bilateral ECT is faster and more effective than other electrode placements, it is also associated with greater cognitive side effects (Carney et al., 2003; Kennedy et al., 2009). Kellner et al. (2010a) did not detect differences in the cognitive profile of bifrontal ECT compared to bitemporal ECT. When used with optimal doses, different electrode placements show modest differences in efficacy, clinicians need to consider whether a modest benefit in favour of rapid treatment or in favour of minimizing possible cognitive side effects is preferred, depending on individual patient needs (Kellner et al., 2010b).

Identifying the dose which will deliver timely and effective treatment, causing the minimum of cognitive side effects has been a key area for research. Guidelines from the Royal College of Psychiatrists recommend the dose of electrical stimulus to be 50–100% above seizure threshold for bilateral ECT where rapid treatment is required in emergencies, or 200% above seizure threshold for unilateral ECT when treatment is less urgent (Scott, 2004). The seizure threshold is the minimum charge needed to induce a seizure, and is determined in individual patients by titration of electric charge. Many factors can influence seizure threshold, for example, it correlates strongly with age (Petrides et al., 2009). More recently, one study has shown that a dose of one and a half times seizure threshold for bifrontal and bitemporal ECT is highly effective (Bailine et al., 2010) whereas another comparing bilateral ECT at threshold stimulus and at 1.5 times threshold stimulus found no difference in outcome for depression in terms of the number of ECT treatments needed for clinical improvement and length of hospital stay (Thirthalli et al., 2009). A dose of six times seizure threshold for right unilateral ECT is highly efficacious (Bailine et al., 2010) and right unilateral ECT at high dose is as effective as bilateral ECT, but with fewer cognitive side effects (Sackeim et al., 2000). Dosing is particularly important for unilateral ECT, and given that seizure thresholds can rise steeply during ECT, ‘high’ (e.g. six times) or at least ‘moderate’ (e.g. three times) seizure thresholds should be used (Plakiotis et al., 2010).

Ultra-brief stimulation uses very short pulses (0.3 ms compared to standard ‘brief’ ECT with pulses of 0.5–2ms) at high doses (between 4–10 times seizure threshold), with the aim of using the minimum electrical energy necessary to stimulate a neuron, leading to more efficient seizure induction and potentially minimizing cognitive side effects (Sienaert et al., 2010a). Using ultra-brief stimulation at four times the seizure threshold appears to be effective, has good tolerability (Quante et al., 2011; Sienaert et al., 2009a) and is acceptable to patients (Sienaert et al., 2010b). There is some evidence of fewer cognitive side effects using ultra-brief stimulation compared to standard, brief stimulation (Loo et al., 2008). However, patients treated with ultra-brief right unilateral ECT require significantly more treatments than patients starting on bilateral ECT, and overall ultra-brief ECT appears to be less effective than bilateral ECT (McCormick et al., 2009a), with the potential for few cognitive advantages if more treatment sessions are needed (Quante et al., 2011).

ECT is an effective treatment and after delivering an average of 12 ECT sessions, one study showed remission rates of 56% (Cinar et al., 2010). However, this study showed that patients can continue to show definite improvement beyond 12 sessions, and for a small group of patients remission may only be achieved by a prolonged course of ECT (see section 3.17).

Medication may influence the effectiveness of ECT by interfering with seizure parameters; compared with psychotropic medication, benzodiazepines can have a significant influence and should be used cautiously when trying to achieve remission (Bundy et al., 2010). Guidelines regarding concomitant antidepressants during electroconvulsive therapy (ECT) are inconsistent, yet combination treatment is considered routine by many in order to promote efficacy, especially given the high relapse rate after acute ECT (Haskett & Loo, 2010). Combined ECT and anti-psychotic treatment for depression can be well tolerated with minimal adverse effects (Masdrakis et al., 2010).

A key aspect of good practice in the use of ECT is adequate explanation and consent before treatment (Malekian et al., 2009). Whilst many patients have a favourable experience of ECT, a review showed that a sizeable proportion report poor information about the treatment itself and possible side effects (Chakrabarti et al., 2010). Good practice in ECT requires close attention to safety and monitoring (Scott, 2004). Use of clinical databases can help to ensure treatment is delivered safely, and to high standards, as well as providing a tool for audit and research (Rai et al., 2010).

What are the side effects?

Major depressive disorder itself has an impact on cognitive function (McClintock et al., 2010) and cognitive effects from ECT are three-fold (Kennedy et al., 2009): first, cognition may improve as depression is treated (Bayless et al., 2010). Second, there may be a brief period of confusion, including anterograde amnesia. Third, a minority of patients develop retrograde amnesia, particularly for autobiographical memory. Whilst the total numbers affected are small, this is a highly distressing side effect for those patients affected, and current evidence does not provide an estimate of the degree of retrograde amnesia that is often described (Semkovska & McLoughlin, 2010).

A systematic review with meta-analysis found that the majority of cognitive abnormalities associated with ECT were limited to the first three days after treatment (Semkovska & McLoughlin, 2010). After 15 days, pre-treatment functioning recovered, and processing speed, working memory, anterograde memory, and some aspects of executive function improved beyond baseline levels (Semkovska & McLoughlin, 2010). Brief confusion following ECT or post-ictal delirium may occur following ECT. Catatonic features are predictors of post-ictal delirium and use of propofol may be useful as an alternative anaesthetic (Kikuchi et al., 2009) in this situation. The addition of manual hyperventilation during the early phase of ECT may also help to lessen the impact on immediate orientation following ECT, without impeding clinical response (Mayur et al., 2010).

Using right unilateral ECT at high dosage has been shown to be as effective as bilateral ECT, but with less severe and persistent cognitive effects (Sackeim et al., 2000, 2009). Whilst there are studies which have not found adverse effects on cognitive function (Sienaert et al., 2010a), for patients where minimizing retrograde amnesia is a priority, right unilateral electrode placement should be the preferred initial choice, given the accumulated evidence suggesting fewer cognitive side effects (Kellner et al., 2010a; O'Connor et al., 2010a). Some patient-specific factors increase the likelihood of cognitive side effects: increasing age, lower education level, lower pre-morbid IQ and psychotropic medication (Moreines et al., 2011). Where there are concerns about cognition, use of dose titration to determine the minimum effective dose and limiting number and frequency of sessions can help to reduce the frequency and intensity of cognitive side effects.

Careful and routine monitoring of cognitive function is important. Falconer et al. (2010) used computerized neuropsychological testing and found significant impairments in visual and visuospatial memory during, and in the week following ECT (Falconer et al., 2010). Most impairments resolve within one month following ECT; however, impairment remained in some domains (e.g. spatial recognition). They concluded that those with pre-existing memory concerns may benefit from more wide spread use of neuropsychological instruments such as the Cambridge Neuropsychological Test Automated Battery (CANTAB).

Cardiac side effects from ECT may be related to general anaesthesia; therefore, the process of seizure induction is of particular concern for those with a history of cardiac disease, as well as older people. Despite this, ECT is well tolerated by elderly patients, even those in poor physical condition (Moksnes & Ilner, 2010). ECT has also been used in those with serious cardiovascular disease, for example, a retrospective case series of eight patients showed that ECT was used safely in patients with abdominal aortic aneurysm (Mueller et al., 2009). It is important that physical health should be closely monitored prior to ECT, and ECT should be considered even in those with physical comorbidities if the risks to their mental health are high.

Does ECT prevent relapse?

In patients with major depression whose symptoms had remitted successfully with ECT, over 80% relapsed within six months when left without active therapy (O'Connor et al., 2010b; Sackeim et al., 2001). Non-psychotic patients who had at least one adequate antidepressant medication trial before ECT may be especially prone to early relapse after successful acute remission with ECT (Rasmussen et al., 2009).

One way to prevent relapse following successful ECT may be to continue pharmacotherapy. A double-blind, placebo, controlled trial found that early pharmacotherapy with sertraline provided protection against relapse, with a relapse rate of only 14% (Yildiz et al., 2010). Nortriptyline plus lithium has also been suggested as a useful combination therapy to prevent relapse following ECT (Sackeim et al., 2001), though ultimately the optimal therapy is likely to be determined at an individual level. The efficacy of ECT can be substantially increased by the addition of antidepressant medication, but clinicians should bear in mind that medications may differ in whether they reduce (e.g. nortriptyline) or increase (e.g. venlafaxine) cognitive adverse effects (Sackeim et al., 2009). It has been suggested that patients with treatment resistant depression who switch antidepressants after successful ECT are more likely to maintain clinical remission than those who are maintained on the same antidepressant treatment (Nakajima et al., 2009).

An alternative to pharmacotherapy is continued use of ECT in carefully selected patients at high risk of relapse. This may involve reducing the frequency of ECT after an acute course, lengthening the intervals between treatments from weekly to fortnightly then monthly before stopping (so called ‘continuation’ ECT). Continuation ECT can lead to sustained remission of depression in those with treatment resistant mood disorder (van Waarde et al., 2010), and can reduce and shorten admission for older people with severe affective disorder (O'Connor et al., 2010b). There are no differences in memory outcomes when comparing those receiving continuation ECT compared to those with continuing pharmacotherapy (Smith et al., 2010). Patients who remain at very high risk can also be offered ‘maintenance’ ECT, typically monthly treatments for an indefinite period, extending to years in some cases. The CORE study reported that continuation pharmacotherapy and maintenance ECT were equally effective for relapse prevention during the first 6 months after responding to ECT (Kennedy et al., 2009). The literature on continuation–maintenance ECT is limited; while it may be beneficial for some individuals, pharmacotherapy (and psychological therapy) is likely to have a much wider appeal for relapse prevention.

Can we predict who will benefit from ECT?

Individual responses to ECT vary, from those who improve rapidly, through to slow improvers, to those who show no benefit from the treatment. There may be reasons why people do not improve at all (e.g. incorrect diagnosis, comorbid personality disorder for example); however, it remains difficult to predict an individual's response to ECT (Cinar et al., 2010).

Clinical symptoms may predict efficacy, with one study suggesting that those with treatment resistant depression who score highly for dysphoria on the MADRS scale before treatment were more likely to have a good response to ECT (Okazaki et al., 2010). Physiological markers of treatment adequacy such as, EEG seizure duration, EEG regularity, peak ictal heart rate and post-ictal suppression are modestly predictive of antidepressant response when giving right unilateral ECT (Azuma et al., 2011; Kimball et al., 2009). High levels of parasympathetic modulation at baseline might be associated with a beneficial effect upon ECT treatment (Ebert et al., 2010).

Much emphasis has been placed on the possibilities for ‘personalized medicine’. Recent papers have considered whether specific genetic profiles may increase the likelihood of successful ECT. The catechol-O-methyltransferase (COMT) val158met polymorphism has been associated with better treatment response to ECT, but only in a female subgroup of patients (Domschke et al., 2010). A plausible mechanism for this may be the role of COMT degradation of norepinephrine and dopamine. Treatment response to ECT was not associated with two tryptophan-hydroxylase-2 (TPH2) polymorphisms (Anttila et al., 2009), vascular endothelial growth factor polymorphism (VEGF 2578 C/A) (Viikki, et al., 2010a), angiotensin I-converting enzyme gene (Stewart et al., 2009), or carrier status for the long allele of the serotonin transporter gene polymorphism (Rasmussen and Black, 2009). The CC genotype of one tryptophan hydroxylase polymorphism (TPH1 218A/C) was associated with increased risk of MDD and lower probability of responding treatment (Viikki et al., 2010b).

Magnetic seizure therapy

What is the difference between MST and ECT?

Magnetic seizure therapy, in common with ECT, is a non-invasive technique where a generalized, tonic–clonic seizure is induced under general anaesthesia (Carpenter et al., 2011). In contrast to ECT, it does not involve an electrical current passing through deep brain structures. Instead, a coil placed on the head is used to generate high frequency, high intensity transcranial magnetic stimulation. Magnetic fields pass through the scalp and skull without the impedance affecting electric currents leading to depolarization of neurons in specific regions of the superficial cortex, which causes focal induction of a generalized seizure (Hoy & Fitzgerald, 2010a, 2010b; Lisanby et al., 2003b). The aim is to target specific regions and circuits in the brain which are implicated in the pathogenesis of depression, without impinging on deeper, e.g. hippocampal, structures involved in memory storage and acquisition, thereby reducing (or eliminating) cognitive side effects associated with ECT (Carpenter et al., 2011; Lisanby et al., 2003a).

What progress has been made towards clinical use of MST?

Repetitive trans-cranial magnetic stimulation (rTMS) can inadvertently lead to seizures: in 2000 Lisanby et al. used rTMS to deliberately induce seizures in non-human primates (Lisanby et al., 2001a). Studies in non-human primates provided early safety data and demonstrated no histological lesions after MST and ECT (Dwork et al., 2004; Lisanby et al., 2003c). In contrast to subconvulsive clinical rTMS, which is generally administered at frequencies of 1 Hz or 10 Hz, early MST devices stimulated at 40 Hz (Fitzsimons et al., 2009). However, these early studies found that magnetic stimulators were underpowered to reliably induce seizures in anaesthetized animals (Lisanby, 2002); custom-modified stimulators capable of reaching 100 Hz led the way for technological development of commercially available stimulators which could reach this frequency.

In 2001 ‘magnetic seizure therapy’ was used for the first time on a human subject in a course of four sessions which were well tolerated and led to improvement on the Hamilton Rating Scale for Depression, without any change to cognition (Lisanby et al., 2001b). A randomized study by Lisanby et al. investigated the safety and efficacy of MST compared to ECT in 10 patients (Lisanby et al., 2003a) using a pulse frequency of 40–60 Hz, with train duration of 0.5–8.0 s at maximal stimulator output (400 pulses). The authors found that MST seizures were of shorter duration, had lower ictal EEG amplitude and less post-ictal suppression. Furthermore, patients had fewer subjective side effects, recovered orientation more quickly following therapy and had fewer cognitive side effects. Use of non-focal coils and vertex placement were the most efficient way of administering treatment.

Following the study by Lisanby et al. (2003a), several case reports have been published. One reported successful treatment of a patient with severe treatment-resistant major depression following 12 sessions of MST (Kosel et al., 2003). No adverse events were reported and cognitive assessment demonstrated less severe cognitive side effects than previously reported following ECT; verbal and non-verbal learning tasks improved significantly from baseline compared to one week after completion of the MST trial (Kosel, et al., 2003). Hoy et al. used single stimulation trains of 1000 pulses at 100 Hz stimulation to the vertex in a course of eight sessions delivered three times per week for a woman with treatment-resistant depression whose symptoms had failed to respond adequately to treatment with a range of antidepressants, anti-psychotics, mood stabilizers and ECT (Hoy & Fitzgerald, 2010b). Improvement in depressive symptoms was seen after six sessions and was maintained at one month post treatment. Following individual sessions there was no disorientation or cognitive side effects, and mild improvement was seen across a range of neuropsychological domains. Kayser et al. reported a case study where a patient with treatment resistant depression in the context of bipolar disorder had a partial clinical response with MST, without precipitation of mania (Kayser et al., 2009).

Most recently was a randomized study of 20 patients with treatment resistant depression who received either ECT or MST (stimulation frequency 100 Hz; train duration up to 6 s) (Kayser et al., 2011). Using 100 Hz, there was seizure induction in all patients; in the MST group, six out of 10 were classified as responders (50% reduction of the MADRS score) and three of these reached remission status (MADRS score of less than 10). There was no significant difference in anti-depressant response between the two treatment groups, and cognitive side effects were not observed in either group. Whilst this was a small study, the positive results in favour of MST suggest that it may be a useful alternative to ECT if efficacy and safety data can be confirmed (Kayser et al., 2011).

How is treatment delivered?

MST is performed under general anaesthesia, most frequently within an ECT suite (Lisanby, 2002). During anaesthesia, succinylcholine-induced muscular paralysis is needed in addition to medical and physiological monitoring, including electrocardiogram, pulse oximetry, blood pressure and electroencephalogram (EEG) (Stanford et al., 2005). Early development of MST identified potential adverse effects and associated solutions: earplug were used to protect hearing (rTMS produces a loud clicking caused by the vibrating of the copper wires within the stimulating coil) and electroencephalographic electrodes were modified (using MRI compatible EEG electrodes) to prevent heating and potential for scalp burns (Lisanby, 2002). Use of 100 Hz trans-cranial magnetic stimulation over the vertex, closest to the motor cortex, appears to be the most reliable method of seizure induction (seizures induced in 10 out of 11 patients in one study) (Kirov et al., 2008) and is now the most frequently used stimulation (Kayser et al., 2011).

What are the effects on cognition?

A randomized control trial using non-human primate models showed that MST at 50 Hz had a more benign side effect profile compared with electroconvulsive shock therapy; this effect was most marked for long-term memory of a constant target, short-term memory of a variable target and recall of previously learned three-item lists (Moscrip et al., 2006). In line with these results, a comparison of electroconvulsive shock therapy, MST delivered at 100 Hz and sham therapy found that non-human primates were significantly slower (p < 0.0001) to complete criterion tasks after electroconvulsive shock therapy compared to MST and sham situations (Spellman et al., 2008). This study provided further support for reduced cognitive side effects in MST, compared to other electroconvulsive therapies.

The neuropsychological effects of MST have recently been reviewed (Moreines et al., 2011). Though studies in humans are based on individual case reports or small (less than 20 participants) groups there is growing evidence that MST is associated with few subjective and objective cognitive side effects (Hoy & Fitzgerald, 2010b; Kayser et al., 2011; Kosel et al., 2003; Lisanby et al., 2003a; Lisanby et al., 2001b), as well as rapid recovery of orientation after treatment compared with ECT (Kirov et al., 2008; White et al., 2006). Though more studies are needed, the neuropsychological profile is likely to be superior to ECT. Since MST may be associated with similar efficacy to ECT with fewer cognitive side effects, it may be of particular use in treating childhood depression where patients are unresponsive to medications and the only available option is ECT (Sporn & Lisanby, 2006). It should be noted that whilst many scientists are optimistic about the benefits of MST, concerns have also been raised. For example, use of ECT with an appropriate selection of seizure parameters may diminish cognitive side effects sufficiently such that MST provides fewer additional benefits (Zyss et al., 2010).

What further research is needed before it can be used in practice?

Over the past decade, MST has been systematically developed through device development, feasibility studies in animals and humans, safety testing in non-human primates (with neuroanatomical, neurophysiological and cognitive endpoints), safety testing in patients, initial efficacy testing in patients, dosage optimization, and randomized comparison with ECT (Rowny et al., 2009). However, much existing evidence in favour of MST is based on individual case reports or small-scale studies. Larger studies using randomization and blinding are needed to establish safety and efficacy before MST can be used in routine clinical practice. Of particular interest will be studies that offer a direct comparison with ECT, which can confirm optimum seizure parameters (including coil position), provide evidence of the absence of cognitive side effects and confirm that there is lasting improvement in symptoms. In addition to use in clinical practice, research determining how MST (and other neuro-stimulatory therapies) works, is likely to provide new insights into the neurophysiology of mood regulation (George et al., 2002; Rau et al., 2007).

Discussion

ECT is widely used in clinical practice and, though not without its critics, is an effective treatment for depression (Carney et al., 2003). There is particular concern about the cognitive side effects of ECT which range from mild and temporary, to more severe problems with autobiographical memory (Kennedy et al., 2009). However, optimizing the treatment schedule (e.g. using right unilateral rather bilateral therapy) minimizes side effects – and for some the antidepressant response to ECT leads to improvement in cognition (Sackeim et al., 2009; Sackeim et al., 2000). Thus, despite concerns, ECT remains a valuable treatment; for many patients the benefits of using it far outweigh the risks. There remains a frustrating lack of evidence explaining how ECT works. Whilst this may not matter clinically, greater understanding of the underlying mechanisms of action would be of benefit when developing other neurostimulatory therapies.

MST has been developed over approximately 10 years. Research has been concentrated in a few centres enabling a careful development process from device development though animal studies, to use on human subjects (Rowny et al., 2009). The literature base for MST is small, with many reviews citing the same early papers. Whilst individual case reports have been useful in the development of this therapy, there are few randomized studies. Existing papers provide ample justification for further research in this area, but do not constitute sufficient evidence for widespread clinical use. Larger studies are underway, and are much needed to confirm the efficacy and safety of MST before it can be fully translated into the clinical setting. Thus, whilst the therapy is still in its infancy, the combination of promising research and a sound theoretical basis means that MST has potential as an alternative therapy to ECT, with fewer (cognitive) side effects.

Acknowledgements

Many thanks to Mark Bryant and librarians at Oxford Health NHS Foundation Trust and Oxfordshire Learning Disability NHS Trust (Ridgeway Partnership) for their help in obtaining papers.

Declaration of interest: C.L.A. received support from Oxford University Clinical Academic Graduate School ([email protected]). K.P.E. has previously received support in kind and travel expenses from Dantec and Magstim. The authors alone are responsible for the content and writing of the paper.