Research Article - (2022) Volume 8, Issue 11
Received: 22-Nov-2022, Manuscript No. IPCP-22-12824; Editor assigned: 24-Nov-2022, Pre QC No. IPCP-22-12824 (PQ); Reviewed: 08-Dec-2022, QC No. IPCP-22-12824; Revised: 13-Dec-2022, Manuscript No. IPCP-22-12824 (R); Published: 20-Dec-2022, DOI: 10.35841/2471-9854-8.11.169
Background: Schizophrenia is a significant cause of morbidity, and current biologic treatments often fail to achieve remission. Repeated transcranial magnetic stimulation (rTMS) is a non-invasive neuromodulation therapy approved for major depression. Peer-reviewed literature suggests that rTMS may have efficacy for psychosis as well as negative and cognitive symptoms; however, holistic data regarding the use of rTMS for schizophrenia remains unclear.
Objective: We aim to synthesize published data of rTMS efficacy in treating schizophrenia and evaluate the most efficacious treatment parameters.
Methods: A meta-analysis was performed evaluating mean weighted effect sizes (Cohen’s d) and heterogeneity (Cochran’s I2).
Results: 24 studies were included for analysis (N=4091). rTMS demonstrated greater effect sizes over sham in Positive and Negative Syndrome Scale (PANSS) negative (d=0.40, p=0.007 I2=59), PANSS general (d=0.31, p=0.004, I2=0) and Global Assessment of Functioning (GAF) (d=0.470, p=0.020, I2=58.2) scores. rTMS also demonstrated significant effect sizes over sham in PANSS positive (d=0.207, p=0.017, I2=20.2) and MADRS (d=0.457, p=0.023, I2=54.1) Sub-group analyses indicated that the stimulation location and frequency did not statistically influence efficacy.
Conclusion: rTMS may have benefits for treating schizophrenia, particularly in reducing negative symptoms when targeting the dorsal lateral prefrontal cortex (DLPFC) with high frequency stimulation (≥ 10 Hz). There was no evidence to support the efficacy of rTMS on audiovisual hallucinations. Further large-scale clinical trials are necessary to verify these findings and evaluate the durability of treatment effects, as there is limited long-term outcome data for the use of TMS for schizophrenia.
TMS; Neuromodulation; Schizophrenia; Meta-analysis review
Schizophrenia remains one of the most poorly controlled psychiatric diseases. While pharmacologic treatment has focused on targeting the positive symptoms of schizophrenia, few treatment options exist for the negative symptoms. Severity of negative symptoms has been demonstrated to be closely correlated with quality of life and functional outcomes [1]. Even with an international prevalence of approximately 1%, schizophrenia exerts a large cost on both healthcare and social systems, primarily through the loss of work productivity. Only 10%-20% of individuals are able to hold a job, representing a significant economic loss, estimated at $ 155.7 billion with 38% of that total attributed to unemployment in the U.S.A and subsequent financial burden on the patient’s caregivers [2-4]. Moreover, individuals with schizophrenia end up destitute at a much higher rate than the general population [5]. Thus, any treatment modalities that improve symptoms, more specifically negative and cognitive symptoms, could be of significant benefit and greatly reduce the morbidity of the disease.
Negative symptoms and cognitive impairment from psychosis are major contributors to social and functional debilitation for many patients suffering from schizophrenia. Unfortunately, antipsychotic medications have limited efficacy in these Research Domain Criteria (RDoc) constructs, suggesting a need for more treatment options to supplement a more holistic approach towards therapy [6]. One avenue being explored is the use of non-invasive brain stimulation (NIBS) which has been studied primarily in treatment-resistant depression. NIBS, specifically electroconvulsive therapy (ECT), has been a staple of psychiatric treatment for many decades with more focused modalities, such as direct current stimulation (DCS) and transcranial magnetic stimulation (TMS), gaining clinical favor over recent years due to less adverse effects and greater ease of implementation in the outpatient settings [7,8]. TMS applies a magnetic field that can penetrate the skull and induces changes in neuronal activity as well as plasticity at the cortical level through both immediate and long-term potentiation and/or depression [9]. It has been demonstrated that low frequencies (1 Hz or less) reduce cortical activity, while high frequencies (>10 Hz) increase cortical activity [10].
In schizophrenia, hyperactivity of the auditory cortex occurs during auditory hallucinations [11], while hypo-functioning of the dorsal lateral prefrontal cortex (DLPFC) has been implicated in the pathogenesis of negative symptoms [12]. When targeting positive symptoms of schizophrenia, the auditory cortex has been targeted at low frequencies (1 Hz) to reduce cortical activity and induce long-term depression (LTD). In contrast, 10 Hz (or more) has been used over the DLPFC to increase cortical excitation and induce long-term potentiation (LTP) to attempt to alleviate negative symptoms. These long-term changes in potentiation have been postulated to modulate glutamate receptors, such as N-Methyl-D-aspartic acid (NMDA) and α-amino- 3-hydroxy-5-methyl-4-isoxazoleproprionic acid (AMPA) receptors [13].
Multiple clinical trials have been conducted over the last decade utilizing TMS for the treatment of schizophrenia. While the parameters of location and frequency for each respective category of symptoms (e.g., 10 Hz over the DLPFC for negative symptoms and 1 Hz over the temporoparietal junction (TPJ) for positive symptoms) have remained relatively consistent, the length of treatment, total stimulation, type of sham, assessment measures, and patient characteristics are heterogeneous from study to study, leading to conflicting results at times [14]. Recent reviews of the literature have been unable to definitively support or refute the use of TMS in the treatment of schizophrenia, due in part to this heterogeneity in treatment protocols and outcome measures.
In this study, we aim to provide a comprehensive review that synthesizes the published data regarding the efficacy of TMS for schizophrenia and provide sub-group analyses to determine the ideal stimulation parameters for the individual treatment of negative and positive symptoms.
This meta-analysis was performed in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses statement [15].
Literature Search and Study Selection
An online search was conducted reviewing PubMed, Google Scholars, and clinicaltrials.gov using the terms “transcranial magnetic stimulation,” “TMS,” “schizophrenia,” “psychosis,” and “psychotic disorders”. There were no restrictions placed upon date of publication or language. The last database search was performed in January 2019. Literature was screened by two independent evaluators and selected based on inclusion/ exclusion criteria.
Inclusion Criteria
Randomized, double-blind, sham-controlled trials using TMS as an adjunct to antipsychotic treatment, subjects were diagnosed with schizophrenia based on either the Diagnostic or Statistical Manual of Mental Disorders (DSM-III, DSM-IIIR, DSM-IV, DSM-IV-TR, and DSM-V) or International Classification of Diseases (ICD 9 or 10).
Exclusion Criteria
Data provided in the study were unclear and/or inadequate for proper statistical analysis.
Study data overlapped with at least one other study containing a larger patient sample size.
Data Collection
Study data were independently verified by the investigators and all study inclusion and exclusion criteria were determined by consensus.
Data Items
Outcomes examined include (in alphabetical order): Auditory Hallucination Rating Scale (AHRS), Brief Psychiatric Rating Scale (BPRS), Clinical Global Impression-Improvement (CGI-I), Clinical Global Impression-Severity (CGI-S), Global Assessment of Functioning (GAF), Montgomery-Asnerg Depression Rating Scale (MADRS), Positive and Negative Syndrome Scale (PANSS) negative, Positive and Negative Syndrome Scale (PANSS) positive, Positive and Negative Syndrome Scale (PANSS) total, Scale for the Assessment of Negative Symptoms (SANS), Psychotic Symptom Rating Scale (PSYRATS), Scale for the Assessment of Positive Symptoms (SAPS), and Trail Making Test-Part A (TMT-A). Of these, we included only AHRS, GAF, MADRS, PANSS negative, PANSS positive, PANSS general, PANSS total, and SANS due to their having 3 or more studies available for analysis that met our inclusion criteria.
Statistical Analysis
To derive the outcome variable, we used reported mean differences, F-statistics, and t-statistics to calculate the standardized mean difference (Cohen’s d) for each study that did not directly report it. For each reported outcome, we conducted separate meta-analyses, using a random effects model. We used I2 statistics to evaluate for heterogeneity, or the percentage of variance in a meta-analysis that is attributable to variations in study design outside of random chance. We assessed publication bias using Funnel plots and tested for possible small study effects using Egger’s test. Galbraith plots were generated to further evaluate heterogeneity.
We defined two possible factors, location of stimulation (right/ left/bilateral OR temporoparietal vs. other) and frequency (<10 Hz vs. =>10 Hz OR<=1 Hz vs. >1 Hz) that vary between studies and could lead to heterogeneity. To further describe possible heterogeneity, we conducted stratified random effects meta regressions for a subset of outcomes (PANSS positive, PANSS negative, PANSS total) as these were the most often reported results and thus there were sufficient studies to allow for stratified analysis. When the analyses showed little difference in effect estimates across strata (assessed by overlapping confidence intervals), we presented the un-stratified meta-analyses. We used Stata 15 (StataCorp, College Station, TX) to conduct all data management and analyses. A confidence interval which did not include 0 (the null value) was deemed to be statistically significant.
Literature Search
39 studies were assessed for eligibility, and 24 of the 39 studies then were included in accordance with our inclusion and exclusion criteria [16-40]. 14 of the studies were excluded from the meta-analysis because of missing outcome data that did not allow for calculation of an effect size, and one was excluded because it did not compare TMS treatment to sham TMS (Figure 1).
Figure 1: Modified PRISMA 2020 flow diagram for new systematic reviews and meta-analysis
PANSS Negative
TMS augmentation across 13 studies (N=555) showed statistically significant improvement in reducing PANSS negative symptom scores compared to sham with a medium effect size of d=0.40 (p=0.007). There was also moderate variation in outcomes across studies (I2=59%) and a potential publication bias towards positive treatment outcomes. Egger’s test revealed no small study effects.
PANSS Positive
TMS augmentation across 16 studies (N=760) showed a small but statistically significant effect on PANSS positive symptom scores compared to sham with d=0.21 (p=0.017). The included studies demonstrated some variation (I2=20.2%) in outcomes across studies, and there may also be publication bias towards positive treatment outcomes.
PANSS Total
TMS augmentation across 8 studies (N=476) showed no statistically significant effects on PANSS total symptom scores compared to sham with d=0.14 (p=0.261). The included studies had some variation in outcomes across studies (I2=33.2%) and there may also be publication bias towards positive treatment outcomes.
PANSS General
TMS augmentation across 8 studies (N=373) showed statistically significant effects on PANSS general scores compared to sham with d=0.31 (p=0.004). The included studies had minimal variation in outcomes across studies (I2=0%) and there may be publication bias towards positive treatment outcomes. Egger’s test for small study effects was nonsignificant.
AHRS
TMS augmentation across 10 studies (N=320) did not show significant improvement in reducing AHRS scores compared to sham, with an effect size of d=0.234 (p=0.312). The included studies had significant variation in outcomes across studies (I2=74.5%) and there may also be publication bias towards positive treatment outcomes.
SANS
TMS augmentation across 5 studies (N=221) showed a significant moderate effect on SANS scores compared to sham with d=0.396 (p=0.004). The included studies did not have significant variation in outcomes across studies (I2=0%), however, there was some evidence of publication bias towards positive treatment outcomes.
GAF
TMS augmentation across 4 studies (N=294) showed a statistically significant effect in increasing GAF scores compared to sham with a medium effect size of d=0.470 (p=0.020). The included studies demonstrated moderate variation in outcomes across studies (I2=58.2%) and there may also be publication bias towards positive treatment outcomes.
MADRS
TMS augmentation across 4 studies (N=282) demonstrated a statistically significant effect in improving MADRS scores compared to sham with a medium effect size of d=0.457 (p=0.023). The included studies demonstrated moderate variation in outcomes across studies (I2=54.1%) and there may also be publication bias towards positive treatment outcomes.
rTMS Moderator Effects
Stratified random effects meta regression analysis for factors of:(1) location of stimulation (right/left/bilateral dorsal lateral prefrontal cortex vs. temporoparietal and other) and (2) frequency (<10 Hz vs. =>10 Hz OR<=1 Hz vs. >1 Hz) did not demonstrate a statistically significant impact on PANSS positive, PANSS negative, or PANSS total scores (data not shown). However, the reduction in PANSS score was greater in the higher frequency rates (10 Hz) or when targeting the DLPFC.
Based on our meta-analysis examining the efficacy of TMS augmentation for the treatment of schizophrenia, we found that TMS was favored over sham in multiple treatment outcome domains. The majority of the benefits appears to be weighted towards improving negative symptoms compared to positive symptoms, given the greater effect size seen in reducing PANSS negative scores (d=0.40) compared to reducing PANSS positive scores (d=0.21). There was significant heterogeneity across studies that measured PANSS negative scores (I2=59%) that may compromise our ability to draw conclusions from this piece of evidence, as studies that measured PANSS negative scores varied in their selection of stimulation target and stimulation frequency. Additional outcomes including AHRS, CGI, PANSS total did not demonstrate statistically significant effects between treatment and sham groups.
Currently, it remains unclear how sustainable the treatment effects are, as most TMS for schizophrenia studies have short treatment durations with limited long term follow up data. Studies of TMS treatment in other illnesses like depression have demonstrated sustained benefits for several months, with extension of treatment benefits with maintenance treatment. Moreover, Li et al’s 2016 study of 47 participants with schizophrenia demonstrated a delayed effect of rTMS on negative symptoms, finding that SANS scores had no change at 4 weeks of treatment but were significantly improved compared to sham by 8 weeks of treatment. Like target, frequency and pattern of stimulation, duration of stimulation treatment is yet another variable that needs further investigation to determine optimal standardized treatment protocol. Schizophrenia is a highly heterogenous disorder with symptom profiles that vary between patients; therefore, it may benefit future researchers to focus on standardizing treatment parameters for specific outcome measures relevant to schizophrenia, rather than attempting to find one optimal standardized treatment for schizophrenia as a whole.
The current sub-group analysis sought to elucidate the optimal treatment parameters based on existing data. We analyzed location of stimulation and frequency of stimulation. Current data from clinical studies suggest that higher frequency DLPFC stimulation demonstrated greater effect sizes for the improvement of negative symptoms while low frequency TPJ stimulation demonstrated greater effect sizes for auditory hallucinations, but these were not statistically significant. However, this data is limited by the heterogeneity of treatment parameters with little standardization of protocols. For instance, the duration of treatment ranged from subjects receiving 20 sessions to 100 sessions of TMS (the average study had only 20 sessions). To date, no clinical trial has provided direct comparison of varying treatment parameters on outcome effects in the use of TMS augmentation for schizophrenia. Another significant limitation is the lack of available follow up data. The lack of longitudinal data leads to unclear sustainability of treatment effects and impacts clinical utility, though studies of the impact of neuro- modulation on network dynamics have suggested that TMS could potentially induce sustained neurologic changes for up to 6 months.
Current data suggest there may be clinical utility in using low frequency TPJ TMS for short-term treatment of positive symptoms and high frequency DLPFC TMS for short-term treatment of negative symptoms. It remains to be seen, however, whether TMS can provide sustained improvement in the treatment of schizophrenia beyond the acute to subacute treatment period. Studies of the efficacy of TMS for schizophrenia that follow up for 12 months or more are needed in order to properly evaluate the durability of treatment effects. In addition, our results should be interpreted with caution due to the substantial heterogeneity present in the results of many studies included in this meta-analysis, as well as the possibility that there was a publication bias toward positive outcomes. Future clinical trials should identify informed treatment outcomes and appropriate parameters including targeted location, frequency of stimulation, and total stimulation on impact of outcome changes. The development of a more rigorously standardized protocol for the use of TMS in schizophrenia could aid in clarifying the utility of the procedure for this indication.
This study was made possible through grants provided by the Della Martin Foundation and the APA 2019 Colloquium for Junior Psychiatrist Investigators. Funding from the 2019 Colloquium for Junior Psychiatrist Investigators conference was made possible (in part) by the National Institute on Drug Abuse of the National Institute of Health under Award Number R13DA042568. The views expressed in written conference materials or publications and by speakers and moderators do not necessarily reflect the official policies of the Department of Health and Human Services; nor does mention by trade means, commercial practices, or organizations imply endorsement by the US Government.
Our funding sources had no involvement in the design of the study, collection of data, statistical analysis, writing of the article, or decision to submit the article for publication.
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
[Crossref] [Google Scholar] [PubMed]
Citation: Chen A, McCarron M, Schooner L, Brasfield K, Chesley J, et al. (2022) Transcranial Magnetic Stimulation for Schizophrenia: A Meta-Analysis. Clin Psychiatry. 8.11.169.
Copyright: © 2022 Chen A, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited