ECT versus Sham for clozapine-resistant schizophrenia: A secondary analysis of a pilot study based on PANSS-30 individual items

Approximately 60% of patients with treatment-resistant schizophrenia (TRS) do not respond to a trial of clozapine [[1]Siskind D. Siskind V. Kisely S. Clozapine response rates among people with treatment-resistant schizophrenia: data from a systematic review and meta-analysis.Can J Psychiatr. 2017; 62: 772-777Crossref PubMed Scopus (174) Google Scholar]. This is termed clozapine-resistant schizophrenia (CRS). A recent metanalysis showed that clozapine augmentation with ECT is efficacious for CRS. However, the authors highlighted the lack of studies with Sham-ECT as a control [[2]Wang G. Zheng W. Li X Bin Wang S Bin Cai D Bin Yang X.H. et al.ECT augmentation of clozapine for clozapine-resistant schizophrenia: a meta-analysis of randomized controlled trials.J Psychiatr Res. 2018; 105: 23-32Crossref PubMed Scopus (59) Google Scholar]. We performed a pilot, randomized, placebo-controlled, single-blinded trial to compare the efficacy of ECT as an augmentation strategy in patients with CRS (or “super-refractory” schizophrenia) in comparison to Sham-ECT, published elsewhere [[3]Melzer-Ribeiro D.L. Rigonatti S.P. Kayo M. Avrichir B.S. Ribeiro R.B. Dos Santos B. et al.Efficacy of electroconvulsive therapy augmentation for partial response to clozapine: a pilot randomized ECT – sham controlled trial.Rev Psiquiatr Clínica. 2017; 44: 45-50Crossref Scopus (7) Google Scholar]. Briefly, patients were recruited at the Institute of Psychiatry of University of São Paulo Medical School, Schizophrenia Research Program (Projesq). The protocol (NCT02049021 – clinicaltrials.gov) was in accordance to the Helsinki Declaration and was approved by the University of Sao Paulo General Hospital ethics committee (protocol 0364/09). All subjects or a legal guardian signed an informed consent form. Patients of both genders aged between 18 and 55 years were included. Patients met the DSM IV-TR diagnostic criteria for schizophrenia or schizoaffective disorder. The symptom and disease severity were evaluated using the Positive and Negative Syndrome Scale (PANSS) [[4]Kay S.R. Fiszbein A. Opler L.A. The positive and negative syndrome scale (PANSS) for schizophrenia.Schizophr Bull. 1987; 13: 261-276Crossref PubMed Scopus (16159) Google Scholar] and the Clinical Global Impression scale (CGI) [[5]Guy W. ECDEU assessment manual for psychopharmacology. US Department of Heath, Education, and Welfare Public Health Service Alcohol, Drug Abuse, and Mental Health Administration, Rockville, MD1976Google Scholar]. Patients were required to have been taking clozapine for at least six months with plasma levels ≥350 ng/mL at baseline and to have a total PANSS ≥60 and CGI ≥4. Patients were defined as having CRS (“super-refractory” schizophrenia in the original publication) based on the Mouaffak et al. [[6]Mouaffak F. Tranulis C. Gourevitch R. Poirier M.F. Douki S. Olié J.P. et al.Augmentation strategies of clozapine with antipsychotics in the treatment of ultraresistant schizophrenia.Clin Neuropharmacol. 2006; 29: 28-33Crossref PubMed Scopus (76) Google Scholar] modified criteria: at least eight weeks’ of clozapine treatment, plasma clozapine levels of >350 ng/mL and moderate severity of psychotic symptoms (≥4) on at least two to four positive symptoms on the PANSS, as well as the CGI. In the ECT group, anesthesia consisted of hypnotic induction with etomidate (0.15–0.3 mg/kg) or propofol (1–2 mg/kg); suxamethonium (0.5 mg/kg) for muscle relaxation with atropine 0.5 mg intravenously. The Sham group received the same anesthetics without the muscle relaxants or electrical stimulus. ECT was administered with bitemporal electrode placement and standard brief pulse stimulus threshold titration and dosing. Twenty-three participated were randomly assigned to the ECT (n = 13) or Sham (n = 10) groups. The groups had similar baseline demographics characteristics, clozapine plasma levels, and psychopathology. Only the total PANSS score differed between groups at baseline (ECT = 81.23[14.46]; Sham = 98.80[19.86] p = 0.023). There were three dropouts in the Sham group and one in the ECT group. Pre- and post-treatment comparisons between groups were carried out with a linear mixed-effects model to accommodate the dropouts [[7]Cnaan A. Laird N.M. Slasor P. Using the general linear mixed model to analyse unbalanced repeated measures and longitudinal data.Stat Med. 1997; 16: 2349-2380Crossref PubMed Scopus (951) Google Scholar] There were no statistically significant differences between groups for the total PANSS (p = 0.668) or the PANSS subscales: Positive (p = 0.646), Negative (p = 0.610) and General Psychopathology (p = 0.501). We used Cohen’s d effect size measure to reanalyze the 30 individual PANSS items, focusing on hallucinations (P1), conceptual disorganization (P2), hallucinatory behavior (P3), blunted affect (N1), passive apathetic withdrawal (N4) and lack of spontaneity & flow of conversation (N6). These items are components of the PANSS-6, a reduced version of the PANSS which has been validated for measurement of symptom severity in patients with TRS [[8]Østergaard S. Foldager L. Mors O. Bech P. Correll C. The validity and sensitivity of PANSS-6 in treatment-resistant schizophrenia.Acta Psychiatr Scand. 2018; 138: 420-431Crossref PubMed Scopus (14) Google Scholar]. For effect sizes calculations we used the usual Cohen’s d formula, subtracting ECT minus Sham means in the numerator, thus negative values represent worsening of psychopathology in the Sham group, and vice-versa. The magnitude of Cohen’s d was interpreted in the usual manner [[9]Borenstein M. Hedges L.J. Higgins J. Rothstein H. Introduction to meta-analysis. John Wiley & Sons, Ltd, Chichester, UK2009Crossref Scopus (11682) Google Scholar]. The results are displayed in Fig. 1 and the supplementary table. At baseline, most items showed a non-significant difference between groups, with the exception of difficulty in abstract thinking (N5), guilt feelings (G3), and uncooperativeness (G8). After treatment, among the PANSS-6 items only conceptual disorganization (P2) showed an improvement, with a reduction of the magnitude of size of the effect. Delusions (P1) and blunted affect (N1) did not change, while other worsened in the Sham group, as can be observed by their medium and large and significant effects sizes: apathetic withdrawal (N4) (−0, 67 CI: -1.64-0.30), hallucinatory behavior (P3) (d = −0.99, 95% CI = −1.97 to −0.00) and lack of spontaneity and flow of conversation (N6) (d = −1.04, 95% CI = −2.02 to −0.04). Other symptoms, not related to the PANSS-6, did improve in the Sham group, for example guilt feelings (G3), with high reduction of effect size magnitude as compared with baseline, while others remained unchanged, such as uncooperativeness (G8), or even worsened in the Sham group, as was the case for disorientation (G10) (d = −1.44, 95% CI = −2.48 to −0.04). However, these findings should be interpreted with caution, as they do not show directly an improvement in patients that received ECT, but rather that patients that received the Sham intervention worsened in some symptoms. Additionally, the sample size was small and a power calculation [[9]Borenstein M. Hedges L.J. Higgins J. Rothstein H. Introduction to meta-analysis. John Wiley & Sons, Ltd, Chichester, UK2009Crossref Scopus (11682) Google Scholar] showed that the study did not have 80% power to detect significant differences between groups in any of the core symptoms such as Hallucinations (50.72% power) and Lack of Spontaneity (53.81% power), thus limiting the precision and the external validity of the results. Another consideration is that some symptoms remained unchanged such as uncooperativeness, regardless the type of intervention. Furthermore, disorientation – a symptom that belongs to the PANSS cognitive domain – worsened with Sham, but not with ECT, which is known to produce transient cognitive impairments. In conclusion, different from the original study, the present reanalysis showed some significant effects when each symptom was analyzed individually. After 12 ECT or Sham sessions, patients in the Sham group showed significant worsening in hallucinatory behavior and lack of spontaneity, which are core symptoms for determining severity in patients with TRS [[8]Østergaard S. Foldager L. Mors O. Bech P. Correll C. The validity and sensitivity of PANSS-6 in treatment-resistant schizophrenia.Acta Psychiatr Scand. 2018; 138: 420-431Crossref PubMed Scopus (14) Google Scholar]. This analytical approach could help us to differentiate the clinical impact range of the ECT-intervention on CRS and may show a path to precision psychiatry, where the intervention should be customized to each patient with CRS. Despite the limitation previously pointed out, this is the first published randomized controlled trial of ECT with a Sham control group in patients with CRS [[10]Wagner E. Kane J.M. Correll C.U. Howes O. Siskind D. Honer W.G. et al.Clozapine combination and augmentation strategies in patients with schizophrenia —recommendations from an international expert survey among the treatment response and resistance in psychosis (TRRIP) working group.Schizophr Bull. 2020 May; 46: S309-S310Crossref Google Scholar]. To overcome the limitations of our pilot study we are presently conducting a new trial, with a larger sample size, to adequately evaluate the efficacy and tolerability of ECT vs Sham-ECT for patients with CRS. Melzer-Ribeiro D.L., and Grilli-Tissot declares have no conflicts of interest to declare and Elkis H declare that between 2018 and 2019 received research grants from São Paulo Research Support Foundation (FAPESP) and honoraria for participation as a member of advisory boards speaker or travel support from the following pharmaceutical companies: Aché, Cristália, Daiichi-Sankyo, Janssen and Mantecorp-Hypera. The following is the supplementary data to this article: Download .doc (.09 MB) Help with doc files Multimedia component 1