Ketamine plus propofol-electroconvulsive therapy (ECT) transiently improves the antidepressant effects and the associated brain functional alterations in patients with propofol-ECT-resistant depression

Depression remains highly prevalent and often refractory to treatment, with many patients classified as treatment-resistant after failing adequate courses of antidepressants from at least two pharmacological classes and psychotherapeutic interventions. Electroconvulsive therapy (ECT) and other neuromodulation approaches provide options for treatment-resistant depression (TRD), but a substantial subset of patients do not respond to ECT administered under propofol anaesthesia; the investigators label these individuals as having ECT-resistant depression (ECT-RD). Prior studies have reported short-lived antidepressant effects of ketamine and ketamine-related alterations in brain structure and function, but the utility of combining ketamine with ECT (or as an anaesthetic during ECT) remains uncertain and may carry safety concerns including cardiovascular and addiction risks. Zhang and colleagues conducted a pilot study to test whether alternating intravenous low-dose ketamine and propofol-anaesthetised bilateral ECT over two weeks improves clinical outcomes and induces measurable changes in brain functional connectivity in patients with propofol-ECT-resistant depression. The study aimed to characterise both the time course of clinical response, assessed by the 17-item Hamilton Depression Rating Scale (HAMD-17), and treatment-associated changes in global functional connectivity density (gFCD) and functional connectivity strength (FCS) within the default mode network (DMN) using repeated resting-state fMRI assessments.

This was a single-arm, within-subject pilot study approved by the Tianjin Anding Hospital ethics committee. Patients admitted between December 2014 and December 2018 who had failed to respond to at least 12 sessions of propofol-ECT in the prior two months and met TRD criteria were recruited. The extraction reports that 36 propofol-ECT non-responders (6 males, 30 females) consented; eight were excluded (three for poor image quality and five for failing to complete the second MRI), yielding a final sample of 28 patients. All participants were right-handed and underwent structured clinical interviews (SCID) and baseline HAMD-17 assessment. Exclusion criteria included other psychotic or affective disorders (for example bipolar disorder), substance use disorders, organic brain lesions, major medical or neurological disease, MRI contraindications, and history of prolonged unconsciousness. Intervention: participants received alternating treatments over two weeks (six ketamine sessions and six propofol-ECT sessions). Ketamine was administered intravenously at 0.5 mg/kg over 40–50 minutes at 22:00 the night before each ECT session; vital signs were monitored after infusion. Bilateral ECT was delivered the following morning between 08:30 and 09:30 using a MECTA device with propofol (1–2 mg/kg) for anaesthesia and succinylcholine as muscle relaxant; atropine (0.25–1 mg) was used as needed. Treatments were given three times per week for two weeks. Imaging and assessments: resting-state fMRI and structural MRI were acquired on a 3T GE scanner. Six MRI scans were scheduled at baseline and at days 1, 3, 7, 10 and 14 after the first ketamine plus ECT treatment. HAMD-17 ratings were obtained at baseline, day 7 and day 14. Functional metrics calculated were global functional connectivity density (gFCD), computed voxel-wise with a Pearson correlation threshold R>0.6 limited to grey matter and normalised by the brain mean, and functional connectivity strength (FCS) within the DMN, derived from group-independent component analysis seeds (top 100 voxels per DMN region) and pairwise region mean time-course correlations. Preprocessing and statistics: resting-state data were processed with SPM8, discarding the first 10 volumes, correcting slice timing and motion (head motion <2 mm and <2°), regressing covariates including Friston 24-parameter head motion, white matter and CSF signals, censoring volumes with frame-wise displacement >0.5, band-pass filtering (0.01–0.08 Hz), normalising to MNI space and resampling to 3 mm voxels. gFCD maps were smoothed with a 6 mm Gaussian kernel. Paired t-tests compared pre- and post-treatment HAMD scores and imaging metrics; family-wise error (FWE) correction was applied to gFCD and DMN FCS comparisons, with significance set at p<0.05. In one analysis a double threshold (cluster size ≥30 and FWE-corrected p<0.001) is reported for gFCD changes.

According to the extracted text, 28 patients completed the protocol and were analysed. Clinically, mean HAMD-17 scores fell from 32.49 at baseline to 14.57 at day 7 after the treatment sequence, corresponding to a reduction rate of approximately 56.2% (statistically significant, p<0.05). However, symptom improvement was transient: depressive symptoms began to relapse from day 8 onwards, and by day 14 the mean HAMD-17 score had returned to 26.71 ± 5.7, a level described as severe depression. Item-level improvements on day 7 included reductions in depressed mood, work and activities, and psychomotor retardation. Imaging findings indicated dynamic changes in functional connectivity following the combined ketamine plus propofol-ECT regimen. At day 7, gFCD values decreased in several regions including the medial prefrontal cortex, subgenual anterior cingulate cortex (sgACC), posterior cingulate cortex, thalamus, hippocampus and orbitofrontal cortex when applying a double threshold (cluster size ≥30 and FWE-corrected p<0.001); these decreases were reported to have disappeared by day 10. Intra-network analysis of the DMN showed that FCS within the DMN decreased beginning on day 1, reached its nadir at day 3, and then gradually returned toward baseline. The authors report no significant correlation between the magnitude of HAMD score reduction and the observed changes in gFCD or DMN FCS. An increase in gFCD in the amygdala was observed in uncorrected analyses, but this change did not survive FWE correction. The investigators note this trend because of its potential relevance to addiction risk and to hypotheses about tolerance, although it was not statistically significant after correction. The extracted text does not provide detailed adverse event data beyond routine monitoring during ketamine infusions and ECT sessions.

Zhang and colleagues interpret their findings as evidence that alternating low-dose ketamine infusions with propofol-anaesthetised bilateral ECT can produce a clinically meaningful reduction in depressive symptoms in ECT- resistant patients, but that this benefit is short-lived, lasting approximately seven days. They emphasise the mismatch between the temporal courses of clinical improvement and imaging changes: brain functional connectivity alterations (gFCD and DMN FCS) were detectable beyond the point at which clinical effects had waned, a dissociation the authors describe as “disjointed.” Several possible explanations are proposed. One hypothesis is that the haemodynamic basis of BOLD-fMRI (blood oxygen signals) may outlast changes in neural electrical activity, producing an apparent persistence of connectivity changes after clinical symptoms return; the authors acknowledge that a one-week lag exceeds typical vascular–neural timing differences and therefore remains speculative. A second possibility is rapid neurochemical desensitisation or tolerance in this ECT-RD sample, which could blunt the clinical response despite persistent connectivity changes; the authors concede that tolerance developing within a week seems rapid and requires further study. A third consideration is that gFCD and FCS metrics may not directly index the microphysiological mechanisms that drive clinical improvement, since they are derived from correlational calculations rather than direct measures of neuronal firing. The authors relate their results to prior literature indicating ketamine can transiently reduce depressive symptoms and alter sgACC and DMN connectivity, and they suggest that combined ketamine plus ECT may modulate brain metabolism and connectivity in a manner partly analogous to ECT alone. They highlight the non-significant trend toward increased amygdala gFCD and raise concerns about addiction risk and tolerance with repeated low-dose ketamine, recommending long-term follow-up. Key limitations acknowledged include the single-group, self-comparison design which limits causal inference about neural mechanisms, the relatively short observation period that may have missed longer-term dynamics, and the need for controlled and longer-term studies to better characterise efficacy, durability and safety, including addiction potential.

In this pilot study, alternating low-dose ketamine infusions with propofol-anaesthetised bilateral ECT produced a measurable antidepressant effect in ECT-resistant patients that persisted for approximately one week. Treatment-related functional connectivity changes were observed in medial prefrontal regions, sgACC, posterior cingulate, thalamus, hippocampus, orbitofrontal cortex and within the DMN, but these imaging alterations did not align temporally with the clinical response. The authors conclude that while the combined approach yields short-term benefit, its limited duration and unresolved safety questions, including tolerance and addiction risk, warrant further controlled and longer-term investigation.