Comparison of Acute Followed by Maintenance ECT vs Clozapine on Psychopathology and Regional Cerebral Blood Flow in Treatment-Resistant Schizophrenia: A Randomized Controlled Trial

Abstract

Background and Hypothesis

In treatment-resistant schizophrenia (TRS), Clozapine is only approved treatment with undesirable side-effects, warranting better alternatives. Our hypothesis is acute followed by maintenance Electroconvulsive Therapy (M-ECT) will be comparable in efficacy and safety to Clozapine in TRS.

Study Design

In this open-label trial, 60 TRS patients were randomized equally to M-ECT (following an acute-course) or Clozapine. Positive and Negative Syndrome Scale (PANSS), Clinical Global Impression Schizophrenia Scale (CGI-SCH), Montreal Cognitive Assessment (MoCA), and Global assessment of functioning (GAF) were measured and compared within and between the groups at baseline, 6 weeks, 12 weeks, and 24 weeks. SPECT-CT brain was done at baseline and 24 weeks to compare the changes in regional cerebral perfusion between the groups and correlate with the changes in the outcome-measures.

Study Results

The PANSS-T scores changes from baseline over the observation-points were significant in both M-ECT and clozapine groups (P < .001), with comparatively better reduction with M-ECT (P < .001). Similar trends were observed in PANSS subscales, CGI-SCH and GAF in both groups, with significantly better improvement with M-ECT over the study-period. After 24 weeks, there was significantly better perfusion with M-ECT in bilateral prefrontal and temporal cortices (P < .05). With M-ECT, a positive correlation was found between changes in PANSS-P scores and left-lateral Temporal cortical perfusion (r = .465, P = .017).

Conclusions

Acute followed by M-ECT was more effective than clozapine over 6 months in reducing the positive and negative symptoms, general psychopathology, illness-severity, and improving the global functionality in TRS [clinicaltrials.gov: NCT03807882].

Introduction

Schizophrenia is a disabling disorder having an early onset, with a chronic, unremitting course resulting in functional deterioration. Antipsychotics are the first-line treatment, but eventually, 25%–33% of the patients on antipsychotics develop treatment-resistance.¹ Clozapine is the only US-FDA approved drug for treatment-resistant schizophrenia (TRS); however, about 40% of TRS patients show poor response even to clozapine.^2,3 Also, the plethora of adverse-effects with clozapine leading to high discontinuation rates necessitates the search for alternative treatment options.⁴

Electroconvulsive therapy (ECT) is a safe and effective treatment for various psychiatric disorders, including schizophrenia. Regular ECT augmentation with antipsychotics is reported to be more efficacious than antipsychotics alone.^5,6 In a retrospective data-based analysis evaluating the effectiveness and speed of response of ECT in TRS, half of the patients on ECT showed >40% reduction in psychotic symptoms with maximum improvement seen between third and sixth ECT sessions.⁷ ECT-augmentation has also been found to be safe and effective in clozapine-resistant schizophrenia⁸ but, the treatment effects of ECT may last for few weeks to months resulting in resurgence of symptoms.⁹ Therefore, maintenance ECT (M-ECT) are ECT sessions given periodically after an acute course to prevent relapses along the continuous course of schizophrenia.⁹

The literature on the therapeutic effects of M-ECT in TRS is limited to few anecdotal reports and retrospective studies.^10–14 The post-hoc analysis of a 2-year retrospective study¹⁵ showed M-ECT to be superior to clozapine in reducing the clinical severity in TRS. However, there is a paucity of controlled clinical trials on M-ECT in TRS. In this study, we compared the efficacy and safety of M-ECT (following an acute-course) vs the standard clozapine therapy on the various symptom dimensions in patients with TRS.

Functional neuroimaging in schizophrenia, using single-photon emission computed tomography (SPECT-CT) brain, have consistently demonstrated hypoperfusion predominantly in the frontal lobe, temporal lobe and basal ganglion.¹⁶ Previous studies in schizophrenia have demonstrated relative improvement of blood flow in the frontal, temporal, and basal ganglia regions and increased activity in the motor cortex with treatment.^17–20 In the present study, we compared the regional blood flow changes with the ECT treatment vs clozapine in TRS using SPECT-CT brain. We further explored the correlations between the regional blood flow changes and the changes in the various symptom dimensions of schizophrenia to substantiate our findings.

Methods

This study was conducted after approval from the institutional ethics committee. The Indian Council of Medical Research (ICMR) ethical guidelines for biomedical research on human subjects (2017) and Good Clinical Practice guidelines were practiced. Before enrollment, written informed consent was taken from the legally authorized representative (LAR) of the patients after explaining the details of the study, the benefits and harms of participating and the freedom of withdrawing at any moment.

Study Design

This was a randomized, open-label, parallel-design clinical trial conducted in a tertiary-care center over a time frame of 14 months [clinicaltrials.gov ID: NCT03807882].

Study Population

The patients with schizophrenia who had non-response (<20% clinical improvement and significant functional impairment) despite trials with ≥2 different antipsychotics at doses ≥600mg/day chlorpromazine-equivalent for ≥6-weeks were defined as TRS (TRRIP consensus criteria).²¹ The patients were screened based on their clinical history and case-record file. Patients with TRS, between 18 and 60 years, of either gender, were screened for eligibility. Those with a history of clozapine or ECT use, psychoactive-substance abuse, or any comorbid major medical condition, and pregnant or breastfeeding females were excluded.

Randomization and Blinding

The study-population was randomized into two treatment arms in 1:1 allocation-ratio by computer-generated block-randomization with a fixed block-size of 6 (by a co-investigator). All the psychiatric outcome parameters were assessed by a single rater (another co-investigator, a senior faculty member of the Department of Psychiatry). Every case (patient with reliable informants) was interviewed by the rater for approximately 60 min, findings were interpreted carefully and in instances of ambiguity a higher rating was applied based on the available information. The outcome assessor and the nuclear medicine specialist were kept blinded to the recruitment and treatment allocation.

Study Procedure

The socio-demographic and clinical data were collected in a structured case record form (CRF). At baseline, the Positive and Negative Syndrome Scale (PANSS) was administered to determine the severity of positive symptoms [PANSS-P], negative symptoms [PANSS-N], general psychopathology [PANSS-G], and total score [PANSS-T].²² Clinical Global Inventory Schizophrenia Scale (CGI–SCH),²³ was administered to determine the baseline severity of the illness, and Global assessment of functioning (GAF),²⁴ to assess the baseline socio-occupational and psychological functioning. Montreal Cognitive Assessment (MoCA) was administered to measure the baseline cognitive deficits.²⁵ Before commencing treatment, SPECT-CT brain was done to measure the baseline regional cerebral perfusion. The study-population was randomized to receive either M-ECT (following acute-course) or clozapine. PANSS, CGI–SCH, GAF, and MoCA were re-administered after 6 weeks, 3 months, and 6 months to compare the changes in the scores within each group and between the groups. The clinical outcome-measures were scored a day before the scheduled M-ECT sessions in the M-ECT group. Post-treatment SPECT-CT brain was repeated at the end of 6 months to evaluate the changes in regional cerebral perfusion.

Interventions

Electroconvulsive Therapy (ECT). In the M-ECT group, after pre-anesthetic evaluation, the patients were administered with 1.5 times supra-threshold, bilateral, brief pulse electrical stimulus under anesthesia (thiopentone/propofol and succinylcholine). Following acute treatment with ECT of six sessions over 2 weeks, M-ECT was administered in a tapered schedule²⁶ at a frequency of weekly sessions for 1 month, then fortnightly for 2 months and monthly once for the following 3 months. In advent of post-ECT agitation or confusion, injection lorazepam 4mg was given parenterally. The ongoing antipsychotics were continued at the same dose for ethical concerns.

Clozapine Monotherapy. Clozapine was prescribed following Maudsley guideline: 12.5 mg on the first day, followed by 12.5 mg twice daily on the second day, followed by 25 mg twice daily for the next 2 days and then an increment of 25 mg every two days till the target dose of 250–400 mg/day in two divided doses as per tolerability of the patients.²⁷ The median stable dose of clozapine was 350 mg/day. Serum clozapine levels could not be estimated; however, the pill-count method was adopted to ensure compliance. The ongoing antipsychotic was gradually tapered and stopped over 1–2 weeks.

Rescue Treatment. The patients with treatment relapse or treatment non-response to the interventions were given rescue treatment. Treatment relapse was defined as “symptom exacerbation following a period of partial recovery” ²⁸ and treatment non-response as “less than 20% improvement in PANSS-T score after 4 weeks of treatment with ECT or 6 months with clozapine”.²⁹ The rescue treatment consisted of combined treatment with clozapine and M-ECT every fortnightly.^8,30–34

Brain Perfusion SPECT-CT Study Protocol

Freshly prepared ^99mTc-ethyl cysteine dimer (ECD) at a 20 mCi dose was administered intravenously. After 30 min, SPECT-CT brain was acquired using low energy high-resolution collimator, at 20 s/frame, in a 128 × 128 acquisition matrix. To avoid bias, the dose and acquisition parameters were kept constant at baseline and follow-up. The images were interpreted semi-quantitatively and the tracer uptake of the prefrontal, temporal and parietal cortex relative to the cerebellum were studied. The values were represented as the differences in cerebral blood perfusion from the standard reference. The changes in perfusion in the different regions from baseline to follow-up were compared within and between the groups.

Study Outcomes

The primary outcome was the change in PANSS scores after M-ECT (following acute-course) or clozapine treatment over 6 months. The clinical response was defined as >20% reduction in the PANSS-T scores from baseline over the study-period.²⁸ The change in the scores of CGI-SCH [S (Severity), and I (Improvement)], MoCA and GAF, the number of patients requiring rescue medications and the adverse events reported in each group, were the secondary outcome-measures. The changes in regional cerebral perfusion from baseline over the study-period were compared between the groups and correlated with the changes in the PANSS, CGI-SCH, MoCA, and GAF scores.

Safety Methods

The adverse effects were assessed by non-directive questioning of the patients and caregivers at the follow-up visits. All adverse effects, were recorded with their description and opinion was made about their causal relationship to clozapine or ECT.

Statistical Analysis

Continuous variables were represented as mean ± standard deviation (SD)/standard error of the mean (SEM) and categorical variables as percentages. For comparison of PANSS, GAF, MOCA, and CGI at different time-points within the group and between the group, two-way repeated measure ANOVA was done. The Mauchly’s test of sphericity was done to evaluate if the sphericity assumption was followed. In case Mauchly’s test was significant, the Greenhouse–Geisser or the Huynh–Feldt correction was done depending on Greenhouse–Geisser (Ɛ) value. The comparison of the change in regional cerebral blood perfusion (from baseline to 24 weeks) between the groups was done using independent t-test. Missing values were handled using multiple imputations, and intention-to-treat (ITT) analysis was performed for all the outcome-measures, including the SPECT variables. The analysis was done using IBM SPSS 23.0, and P <.05 was considered statistically significant.

Sample Size Calculation

A sample size of 21 per group can achieve 90% power to detect a difference of 10 points in PANSS-T scores between the groups with an SD of 9.9, with a significance-level (alpha) of 0.05 using a two-sided two-sample t-test.³⁵ As the study was in schizophrenia patients and the follow-up period was relatively long, we recruited 30 patients in each group, considering 30% attrition.

Results

Patient Flow

The present study was conducted over 14 months, from February 2019 to March 2020. Initially, 74 patients were screened for eligibility, out of which eight patients did not fulfill the inclusion criteria, and 6 patients declined to participate. Finally, 60 patients were recruited and randomized in 1:1 allocation ratio to receive either ECT or clozapine treatment (figure 1). In the M-ECT group, two patients lost to follow-up for ECT sessions within a month. In M-ECT group, two patients and in clozapine group, three patients were non-responders (section “Rescue Treatment”) requiring rescue treatment.

Fig. 1.

CONSORT flow diagram of the progress through the phases of the parallel randomized trial of the two groups.

Baseline Characteristics

At baseline, the demographic characteristics and various clinical parameters were comparable between the study groups without any significant differences. The SPECT-CT brain measurements were also comparable between the groups, suggesting homogeneity among the study-population (table 1).

Table 1.

Baseline Sample Characteristics of the Various Demographics, Clinical Variables, and Primary, Secondary Outcome Measures

Variables	Groups		P Value
	Group-1 (M-ECT) (n = 30)	Group-2 (Clozapine) (n = 30)
Male:female (%)	60:40	57:43	.99
Age (in years)	34.47 ± 10.29	35.80 ± 10.36	.62
Education (in years)	11.77 ± 3.75	12.23 ± 3.48	.62
Age of onset (in years)	25.5 ± 9.58	22.67 ± 5.76	.17
Duration of illness (DOI) (in years)	9.6 ± 7.36	12.8 ± 7.98	.11
PANSS-P	35.65 ± 4.87	35.78 ± 5.20	.93
PANSS-N	36.96 ± 6.35	37.41 ± 6.63	.79
PANSS-G	70.69 ± 9.69	70.85 ± 10.02	.95
PANSS-T	143.31 ± 14.13	144.04 ± 17.85	.86
CGI-SCH-S	6.192 ± 0.49	6.185 ± 0.60	.96
MOCA	19.08 ± 2.36	19.22 ± 2.19	.81
GAF	24.85 ± 5.97	22.11 ± 8.22	.15
Regional cerebral blood perfusion (SPECT-CT brain)a
RLPFC	−0.108 ± 0.04	−0.118 ± 0.06	.47
LLPFC	−0.103 ± 0.05	−0.110 ± 0.06	.62
RMPFC	−0.053 ± 0.06	−0.06 ± 0.06	.47
LMPFC	−0.057 ± 0.06	−0.088 ± 0.15	.30
RLT	−0.062 ± 0.04	−0.047 ± 0.05	.22
LLT	−0.061 ± 0.05	−0.046 ± 0.06	.28
RMT	−0.030 ± 0.05	−0.022 ± 0.06	.52
LMT	−0.018 ± 0.04	−0.001 ± 0.05	.14
RSP	−0.019 ± 0.05	−0.020 ± 0.05	.92
LSP	−0.002 ± 0.05	−0.021 ± 0.05	.16
RIP	−0.064 ± 0.03	−0.067 ± 0.06	.80
LIP	−0.062 ± 0.05	−0.076 ± 0.06	.28

Data are in mean ± SD, Independent t-test/Fisher’s exact test.

^aValues of SPECT-CT brain reflect the difference of regional cerebral blood perfusion of patient from normal reference. RLPFC, right lateral prefrontal cortex; LLPFC, left lateral prefrontal cortex; RMPFC, right medial prefrontal cortex; LMPFC, left media prefrontal cortex; RLT, right lateral temporal; LLT, left lateral temporal; RMT, right medial temporal; LMT, left medial Temporal; RSP, right superior parietal; LSP, left superior parietal; RIP, right inferior parietal; LIP, left inferior parietal.

Change in Psychopathology (PANSS Score)

There was a mean reduction of >20% in PANSS-T scores from the baseline over the study-period in both groups. The changes in the PANSS-T scores from the baseline over the different time-points were significant in both M-ECT and clozapine groups (P < .001). The post-hoc analysis showed that the differences in the PANSS-T were significant for baseline vs 6 weeks (P < .001), 6 vs 12 weeks (P < .001) and 12 vs 24 weeks (P < .001) in both groups. However, the inter-group comparisons showed that the reduction in the PANSS-T scores across the time-points were significantly better with M-ECT over clozapine treatment (P < .001). Similarly, significant reduction in PANSS-P, -N, and -G scores were observed in both groups over the study-period. Again, compared to the clozapine group, the M-ECT group showed significantly better reduction from baseline in the PANSS-P (P < .001), -N (P < .001), and -G (P < .001) subscales across all time-points (table 2 and figure 2).

Table 2.

Improvement of Different Outcome Parameters Over Time (After Intention-to-Treat Analysis)

Parameter	Time Points, Sphericity, Intragroup Comparisons	Group-1 (M-ECT) (n = 30)	Group-2 (Clozapine) (n = 30)	Between the Group Comparison (P value)b
PANSS-P	Baseline	35.65 ± 0.89	35.78 ± 0.95	.925
	6 week	20.42 ± 0.74	29.93 ± 0.82	<.001
	12 week	15.62 ± 0.79	23.70 ± 1.07	<.001
	24 week	13.39 ± 0.70	19.89 ± 1.03	<.001
	Mauchly’s test of sphericity (P value)	<0.001	<0.001
	Greenhouse−Geisser (Ɛ)	0.515	0.613
	P value (with Greenhouse−Geisser correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	<0.001	<0.001
	12 weeks vs 24 weeks	<0.001	<0.001
PANSS-N	Baseline	36.96 ± 1.16	37.41 ± 1.21	.791
	6 week	26.23 ± 1.07	32.56 ± 1.23	<.001
	12 week	22.12 ± 1.03	28.04 ± 1.19	<.001
	24 week	19.96 ± 1.03	25.15 ± 1.39	<.001
	Mauchly’s test of sphericity (P value)	<0.001	<0.001
	Greenhouse−Geisser (Ɛ)	0.542	0.552
	P value (with Greenhouse−Geisser correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	<0.001	<0.001
	12 weeks vs 24 weeks	<0.001	<0.001
PANSS-G	Baseline	70.69 ± 1.77	70.85 ± 1.83	.950
	6 week	46.15 ± 1.86	60.00 ± 1.90	<.001
	12 week	38.27 ± 1.06	50.74 ± 1.92	<.001
	24 week	34.42 ± 1.23	43.44 ± 2.05	<.001
	Mauchly’s test of sphericity (P value)	<0.001	0.048
	Greenhouse−Geisser (Ɛ)	0.464	0.752
	P value (with Greenhouse−Geisser correction or Huynd−Feldt correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	<0.001	<0.001
	12 weeks vs 24 weeks	<0.001	<0.001
PANSS-T	Baseline	143.31 ± 2.58	144.04 ± 3.26	.862
	6 week	92.81 ± 2.11	122.48 ± 3.24	<.001
	12 week	76.00 ± 2.54	102.48 ± 3.63	<.001
	24 week	67.77 ± 2.67	88.48 ± 4.01	<.001
	Mauchly’s test of sphericity (P value)	<0.001	<0.001
	Greenhouse−Geisser (Ɛ)	0.515	0.609
	P value (with Greenhouse−Geisser correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	<0.001	<0.001
	12 weeks vs 24 weeks	<0.001	<0.001
GAF	Baseline	24.85 ± 1.09	22.11 ± 1.50	.149
	6 week	53.08 ± 1.76	30.59 ± 1.45	<.001
	12 week	63.39 ± 1.92	44.82 ± 2.38	<.001
	24 week	68.12 ± 2.06	51.33 ± 2.94	<.001
	Mauchly’s test of sphericity (P value)	0.001	<0.001
	Greenhouse−Geisser (Ɛ)	0.708	0.572
	P value (with Greenhouse−Geisser correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	<0.001	<0.001
	12 weeks vs 24 weeks	<0.001	<0.001
MOCA	Baseline	19.08 ± 0.43	19.22 ± 0.40	805
	6 week	20.27 ± 0.37	20.07 ± 0.39	722
	12 week	20.35 ± 0.35	20.74 ± 0.34	426
	24 week	20.46 ± 0.42	20.96 ± 0.32	.340
	Mauchly’s test of sphericity (P value)	<0.001	<0.001
	Greenhouse−Geisser (Ɛ)	0.662	0.596
	P value (with Greenhouse−Geisser correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	0.49	<0.001
	12 weeks vs 24 weeks	0.42	0.01
CGI-SCH-S	Baseline	6.192 ± 0.09	6.185 ± 0.11	.961
	6 week	3.69 ± 0.11	5.11 ± 0.13	<.001
	12 week	3.12 ± 0.12	4.15 ± 0.17	<.001
	24 week	3.00 ± 0.12	3.78 ± 0.19	<.001
	Mauchly’s test of sphericity (P value)	<0.001	0.012
	Greenhouse−Geisser (Ɛ)	0.656	0.732
	P value (with Greenhouse−Geisser correction)a	<0.001	<0.001
	Baseline vs 6 weeks	<0.001	<0.001
	6 weeks vs 12 weeks	<0.001	<0.001
	12 weeks vs 24 weeks	0.083	0.022
CGI-SCH-I	6 week	2.27 ± 0.12	3.29 ± 0.10	<.001
	12 week	2.04 ± 0.14	2.67 ± 0.13	.002
	24 week	2.00 ± 0.14	2.56 ± 0.15	.007
	Mauchly’s test of sphericity (P value)	<0.001	<0.001
	Greenhouse−Geisser (Ɛ)	0.582	0.687
	P value (with Greenhouse−Geisser correction)a	0.039	<0.001
	6 weeks vs 12 weeks	0.056	<0.001
	12 weeks vs 24 weeks	0.327	0.185

Data presented as mean ± SEM.

^aIntra-group (repeated measure ANOVA).

^bInter-group (repeated measure ANOVA).

Fig. 2.

Response trajectory of different outcome-measures over time. Group I: M-ECT, Group II: Clozapine treatment.

Change in Severity of Illness, Improvement (CGI-SCH Scores), Cognitions (MoCA Score), and Global Functioning (GAF Score)

The changes from baseline in the severity of illness (CGI-SCH-S) over the observation points were significant in both M-ECT (P < .001) and clozapine group (P = .012) over the study-period. However, the significance of the changes in the CGI-SCH-S scores reduced in both the groups for 12 vs 24 weeks period. The inter-group comparison of the changes in CGI-SCH-S score significantly favored M-ECT over clozapine treatment across all observation points (P < .001). Similarly, the improvement with treatment as measured by CGI-SCH-I was found to be significant in both groups, which was particularly noted for 6 vs 12 weeks period (P < .001). The inter-group comparisons showed significantly better improvement of CGI-SCH-I scores with M-ECT over clozapine treatment across the observation points. However, the significance of the difference in the CGI-SCH-I scores between the groups started weaning over the study-period [6 weeks (P < .001), 12 weeks (P = .002), and 24 weeks (P = .007)].

The effect of treatment on global functioning and cognition was assessed by the change from baseline in GAF and MoCA scores, respectively. There was significant increase in the GAF scores from the baseline across the observation points in both groups (P < .001). Even the changes in the GAF scores between each subsequent assessment were significant (P < .001) in both groups. Again, the change in the GAF scores across all the study-points was significantly better with M-ECT over clozapine treatment (P < .001). In the MoCA scores, significant changes were observed in both groups across all the observation-points (P < .001), with maximal improvement during the initial 6 weeks (P < .001). Although there were significant improvements in the MoCA scores in the clozapine group at 6 vs 12 weeks, and 12 vs 24 weeks, the inter-group assessments did not reveal any significant difference in the change in MoCA scores between the groups across the study-points (table 2 and figure 2).

Changes in Regional Cerebral Blood Perfusion (SPECT-CT Brain)

Table 3 compares the regional cerebral blood perfusion changes between the treatment groups over 24 weeks. At 24 weeks, there was significantly better blood perfusion in bilateral prefrontal and temporal cortices with M-ECT compared to clozapine treatment. However, no significant differences in blood perfusion were noted in the other regions between the groups. Figure 3 shows exemplar-scans of the cerebral blood perfusion changes in both groups.

Table 3.

Comparison of the Change in Regional Cerebral Blood Perfusion (SPECT CT Brain) (From Baseline to the End of 24 Weeks) Between the Treatment Groups

Variables	Groups		P Value
	Group-1 (M-ECT) (n = 30)	Group-2 (Clozapine) (n = 30)
	Mean ± SEM	Mean ± SEM
RLPFC	−0.014 ± 0.004	0.002 ± 0.006	.035
LLPFC	−0.025 ± 0.008	0.017 ± 0.009	.001
RMPFC	−0.009 ± 0.005	0.016 ± 0.006	.001
LMPFC	−0.019 ± 0.007	0.007 ± 0.006	.007
RLT	−0.019 ± 0.006	0.005 ± 0.007	.023
LLT	−0.011 ± 0.004	0.017 ± 0.005	<.001
RMT	−0.035 ± 0.008	−0.024 ± 0.010	.375
LMT	−0.022 ± 0.013	0.021 ± 0.015	.047
RSP	−0.013 ± 0.008	0.005 ± 0.010	.166
LSP	−0.014 ± 0.010	−0.007 ± 0.008	.547
RIP	−0.001 ± 0.008	−0.004 ± 0.009	.773
LIP	0.015 ± 0.009	0.006 ± 0.007	.459

Data are in mean ± SEM. Values reflect the difference of regional cerebral blood perfusion of patient from normal reference, and more negative value suggests better cerebral blood flow.

RLPFC, right lateral prefrontal cortex; LLPFC, left lateral prefrontal cortex; RMPFC, right medial prefrontal cortex; LMPFC, left media prefrontal cortex; RLT, right lateral temporal; LLT, left lateral temporal; RMT, right medial temporal; LMT, left medial temporal; RSP, right superior parietal; LSP, left superior parietal; RIP, right inferior parietal; LIP, left inferior parietal.

Fig. 3.

SPECT-CT 3-D stereotactic surface projection (3D-SSP) maps of the brain of single exemplar cases from the M-ECT and clozapine groups: (A) A patient in the M-ECT group, showing significant improvement in perfusion in the left prefrontal cortex and right temporal cortex in post treatment scan as compared to base-line scan. (B) A patient in the Clozapine group, showing improvement in perfusion in the left prefrontal cortex, partial improvement in right temporal cortex and no improvement in left temporal cortex in post treatment scan compared to base-line scan.

Correlations Between Different Parameters

Correlations Between PANSS Scores vs CGI-SCH-S Score, MoCA Score and GAF Score at Baseline. At baseline, the CGI-SCH-S scores correlated significantly positively with the PANSS-T (r = 0.387, P = 0.002), -N (r = .346, P = .007), and -G (r = 0.314, P = 0.015) scores of the study-population. The GAF scores correlated significantly negatively with the PANSS-T (r = −0.510, P < 0.001), -P (r = −0.342, P = .007), -N (r = −0.306, P = .017), and -G (r = −0.466, P < .001) scores. Similarly, the MoCA score was found to correlate significantly negatively with the various PANSS sub-scores.

Correlations Between the Changes in PANSS Scores and the Changes in CGI-SCH-S Score, MoCA Score and GAF Score (From Baseline to 24 Weeks). In the M-ECT group, the change in PANSS-T, -P, -N, and -G scores correlated significantly positively with the change in CGI-SCH-S score (r = 0.681, P < .001; r = 0.531, P = .005; r = .585, P = .002 and r = 0.605, P < .001, respectively). Similarly, in the clozapine group, CGI-SCH-S scores showed a significantly positive correlation with the PANSS-T (r = 0.637, P < .001), -P (r = 0.584, P < .001), -N (r = 0.543, P = .003), and -G (r = 0.566, P = .002) scores. In the clozapine group, the change in PANSS-N score and -G score correlated significantly negatively with the change in the GAF score (r = −0.606, P < .001 and r = −0.684, P < .001, respectively). The change in the various PANSS scores did not correlate significantly with the change in MoCA scores in either group.

Correlations Between the Changes in PANSS Scores and the Changes in Regional Cerebral Blood Perfusion (SPECT-CT Brain) (From Baseline to 24 Weeks). No significant correlations were found between any variables except for a significant positive correlation between the change in PANSS-P score and the change in blood perfusion in the left lateral temporal cortex in the M-ECT group (r = 0.465, P = .017).

Rescue Treatment

Two patients from the M-ECT group and three from the clozapine group required rescue treatment during the study-period.

Safety Evaluation

In most patients of the M-ECT group, the intervention was uneventful, except for post-ECT headache in 12 patients, lasting for 6−8 h, resolving with tablet paracetamol 650 mg. In two patients, post-ECT cognitive deficits were noted only during the last two acute ECT sessions, without any such reports during the maintenance phase. Weekly leucocyte count and absolute neutrophil count for the patients on clozapine showed no evidence of agranulocytosis during the study-period. Sialorrhoea was noted in most patients on clozapine, and tablet glycopyrrolate 2 mg was given in cases with excessive salivation. Constipation was reported in 18 patients at clozapine dosage >200 mg/day, which was treated with 10−15 ml of liquid paraffin. However, none of the patients discontinued clozapine due to intolerance to the side-effects.

Discussion

In this study, the mean reduction of the PANSS-T and the various subscale scores over the study-period were significant in both the M-ECT and clozapine groups. However, M-ECT produced significantly better reduction over clozapine in the PANSS scores across the different time-points. Similarly, M-ECT compared to clozapine, showed significantly better reduction in severity of illness (CGI-SCH-S scores), resulting in significantly better clinical improvement (CGI-SCH-I scores). Interestingly, the significance of the differences between the treatment effects of the two groups started reducing after 12 weeks (figure 2). The improvement with treatment in global functioning as measured by GAF also significantly favored M-ECT over Clozapine. Both the treatments were comparable in improving the cognitive deficits, safety and the need for rescue treatment. The increased blood perfusion with M-ECT was significantly more than clozapine in the prefrontal and temporal cortices of both cerebral hemispheres, particularly in the left lateral temporal region. In the M-ECT group, the reduction in the PANSS-P scores correlated significantly with the improvement of blood perfusion in the left lateral temporal cortex.

At baseline, the treatment groups were homogenous with regard to the various demographic and clinical parameters. The baseline PANSS-T and various subscale scores were consistently higher than their respective values in the previous studies on TRS,^36–40 which could be due to the higher severity of illness in our study-population. This could be corroborated with similarly higher CGI and lower GAF scores at baseline as compared to earlier studies.^36–40 The higher severity of symptoms in our study-population can be partly explained by our study-venue being a tertiary care referral center with a wide catchment area where patients with high symptom severity usually come for treatment. The cognitive deficits in schizophrenia as measured by MoCA have been described in previous studies in the score range of 18−26.²⁵ The lower mean MoCA score in our study-population can be justified by the negative impact of the higher clinical severity on the cognitions. Previous studies have found hypoperfusion in the frontal^18,19,41 and temporal^17,42 cortices in patients with schizophrenia. Furthermore, hypoperfusion in the prefrontal cortex has been described in TRS.⁴³ Similarly, in our study, reduced cerebral blood perfusion was evident in all the regions of the frontal, temporal and parietal cortices (figure 3).

The better efficacy of ECT over clozapine was apparent from early in the study-period due to the faster onset of therapeutic response with acute ECT. For clozapine, time is required for dose optimization and additional time for the onset of its action.⁴⁴ This acute benefit of the initial ECT sessions were subsequently carried forward by the M-ECT, as reflected by the better scores in the outcome-measures with M-ECT relative to clozapine across the follow-up points. By the end of 6-months, probably both interventions were reaching their maximum therapeutic-efficacy, as evident in the narrowing of the gap between their response-time curves of the various clinical parameters (figure 3).

Thus, M-ECT (following acute-course) was more effective than clozapine in reducing the various symptom dimensions of schizophrenia. The decrease in positive symptoms could have reduced the secondary negative symptoms, which led to a reduction in clinical severity and improvement in global functioning. Also, ECT could have a direct beneficial effect on the negative symptoms.⁴⁵ This can be further corroborated by the significant improvement in blood perfusion in the M-ECT group, particularly in the left lateral temporal cortex (region constantly associated with positive symptoms).⁴⁶ The improvement in the secondary negative symptoms can be validated by the corresponding improvement in the prefrontal cortical perfusion in the M-ECT group.⁴⁷ Previous studies have found both ECT and clozapine to improve the cognitive functioning in schizophrenia.^37,48 In our study, both the treatments had a beneficial effect on the cognition particularly in the initial 6 weeks which continued in the clozapine group. However, inter-group assessments of the MoCA scores showed comparable improvements. The usual cognitive side-effects were not evident in the M-ECT group, possibly because the brain got sufficient time to recover between two stimuli.⁴⁹ Patients needing rescue treatment were given combined clozapine and M-ECT treatment in view of the findings of the previous studies on refractory TRS.^8,30–34

The distinct difference between the two treatment options could be explained by the different neurobiological properties of ECT from clozapine. ECT produces faster response by transiently breaching the blood-brain barrier, enhancing the release/expression of neurotransmitters, neurotrophins and transcription factors.⁵⁰ It also brings changes in the regional cerebral blood flow, glucose metabolism and neuronal electrical milieu.⁵⁰ Moreover, non-dopaminergic and inflammatory-stress underlying TRS may have a comparatively better response to ECT.^50,51 ECT also reportedly increases functional connectivity, neuroplasticity and gene expressions in the brain.⁵²

To the best of our knowledge, this is the first randomized controlled trial comparing the efficacy and safety of acute followed by M-ECT vs clozapine in TRS, involving an adequate sample size over a clinically relevant follow-up period. We followed a robust study design, where the clinical measures were corroborated with the SPECT-CT brain findings. Serum clozapine estimation was not possible in our study setting, which was a major limitation. Hence, the current findings may be carefully extrapolated to settings where clozapine is dosed based on blood levels. The outcome assessor and the nuclear physician were blinded to the treatment allocation. However, double-blinding could not be done because of the ethical concerns of giving anesthesia before sham-ECT. An observation point for comparisons at the end of the acute-course of ECT was not feasible as clozapine dose build-up in the parallel group took more time. Cognitive assessments by MoCA were compared between the groups at baseline and the follow-up points only, but cognitive assessment following every ECT session would have been ideal. A longer follow-up would have been ideal for observing the further response trajectories of the interventions. Moreover, being a single-center study, the generalizability of our findings is limited.

In conclusion, acute followed by M-ECT was more effective than clozapine treatment over 6-months in reducing the positive and negative symptoms, general psychopathology, and the illness severity, along with improving the global functionality in patients with TRS with similar effects on cognition. Possibly, the initial improvements with acute ECT were sustained with M-ECT, ultimately showing a better outcome with M-ECT over the study-period. These clinical findings corroborated with the better blood perfusion changes with M-ECT particularly in the prefrontal and temporal cortices. Both the treatments were comparable in terms of safety, without any profound adverse effects. In view of our findings, M-ECT following an initial acute-course can be a viable option in TRS, particularly in patients with contraindications, intolerance, or poor-response to clozapine. Our findings can be further explored through future multicentric studies involving larger samples.

Funding

The study has been conducted under Intramural research grant received by the first author from All India Institute of Medical Sciences, Bhubaneswar, India vide grant number: IMF/01/2018.