The Impact of Antidepressant Discontinuation Prior to Treatment with Psilocybin for Treatment-Resistant Depression

Marwood and colleagues situate this study within ongoing efforts to develop psilocybin as a monotherapy for treatment-resistant depression (TRD). Psilocin, psilocybin's active metabolite, is a partial agonist at the serotonin 5-HT2A receptor, a mechanism implicated in acute psychedelic effects that have been associated with therapeutic benefit. The introduction highlights prior concerns that recent or concurrent use of serotonergic antidepressants might down‑regulate receptors or otherwise attenuate psychedelic effects, citing mixed and largely uncontrolled evidence from case reports, online surveys and small open-label studies. A smaller post hoc analysis previously suggested that requiring antidepressant discontinuation could alter comparative outcomes between psilocybin and an SSRI, but evidence is limited and inconsistent. This post hoc analysis uses data from the largest Phase II randomised controlled trial of psilocybin to date to examine whether supervised discontinuation of antidepressant drugs during the screening period affected: depressive symptom change during discontinuation, baseline suicidality, the acute subjective psychedelic experience, or the trial's primary efficacy endpoint (change in MADRS from Baseline to Week 3). The authors aim to clarify whether recent antidepressant discontinuation compromises psilocybin treatment efficacy or safety in people with TRD and to assess the practical feasibility of supervised withdrawal in this setting.

The analysis draws on a Phase II, international, multi‑centre, randomised, fixed‑dose, parallel‑group, double‑blind, dose‑finding trial of investigational COMP360 psilocybin administered with non‑directive psychological support. The investigational compound was given as a single administration at one of three fixed doses: 25 mg, 10 mg or 1 mg. The protocol received independent ethics approval at each site and participants provided written informed consent; full methodological details are reported elsewhere. Participants were 233 adults with a current major depressive episode meeting criteria for treatment‑resistant depression, defined here as non‑response to two to four adequate pharmacological treatments of ≥8 weeks' duration. During a six‑week screening period, participants who were taking antidepressants were required to taper within four weeks and be off all antidepressant medications for ≥2 weeks before Baseline; fluoxetine required a ≥4‑week washout owing to its longer half‑life. Study physicians met weekly with participants during tapering to monitor safety. Unless clinically necessary, participants were expected to remain off antidepressants for at least three weeks following psilocybin administration. Clinical assessments included the clinician‑rated HAM‑D‑17 at Screening and Baseline to examine change during discontinuation, and the clinician‑rated MADRS at Baseline and follow‑up visits to evaluate the primary endpoint (change from Baseline to Week 3). Suicidality was assessed using MADRS item 10 and, after a protocol update, the C‑SSRS; earlier suicidality data captured with the S‑STS were mapped to the FDA‑CASA classification to harmonise measures. The Five‑Dimensional Altered States of Consciousness Questionnaire (5D‑ASC) measured acute subjective psychedelic effects. For these post hoc analyses, the investigators split the sample into two subgroups: those who discontinued at least one antidepressant or antipsychotic medication between Screening and Baseline (ATC level 3 codes) and those who entered the trial antidepressant drug free. Analyses were descriptive subgroup comparisons; the extracted text reports use of descriptive summaries rather than formal inferential modelling for these post hoc comparisons.

Of 233 randomised participants who received psilocybin, 156 (67%) discontinued at least one antidepressant drug during screening and 77 (33%) entered the trial antidepressant drug free. Baseline demographic and clinical characteristics were reported as similar between these two groups. Four participants initially failed to complete discontinuation within the planned four‑week taper period; two were re‑screened and subsequently completed withdrawal successfully. Measures of suicidality at Baseline, including MADRS item 10 and the FDA‑CASA mapping, were comparable and low across both groups. HAM‑D‑17 scores did not show evidence of worsening between Screening and Baseline regardless of the number of antidepressant drugs discontinued. Baseline MADRS scores were not considered to meaningfully differ by number of discontinued antidepressant drugs within the 10 mg and 25 mg dose groups. The primary endpoint—change in MADRS total score from Baseline to Week 3—showed no meaningful differences attributable to antidepressant discontinuation. The authors note a slightly lower mean change in the 10 mg group among participants who discontinued more than one antidepressant and a slightly higher mean change in the 1 mg group for those entering drug free, but these patterns were not interpreted as meaningful. A subgroup comparison by prior antidepressant treatment duration (>6 months versus ≤6 months) did not materially affect results for the active dose groups. Acute subjective experiences measured by the 5D‑ASC did not meaningfully differ by discontinuation status across doses; small differences were somewhat more pronounced in the 10 mg group across four of five dimensions. Post‑treatment behaviour differed by subgroup: 81 participants who had discontinued an antidepressant during screening (51.9%) initiated an antidepressant within 12 weeks after receiving COMP360, compared with 18 participants who had entered antidepressant drug free (23.4%). The extracted text references figures and tables for detailed distributions but those numerical displays are not reproduced here.

Marwood and colleagues interpret their post hoc analyses as showing that supervised antidepressant discontinuation in the screening period was feasible and not associated with increased depressive symptom severity or greater baseline suicidality in this large TRD trial. Psilocybin treatment efficacy (MADRS change to Week 3) and the intensity of acute subjective psychedelic effects did not appear to be attenuated by recent antidepressant use and discontinuation, supporting the view that recent antidepressant discontinuation is unlikely to compromise psilocybin's mechanism or clinical signal in this dataset. The authors emphasise several strengths that increase confidence in these findings: the controlled trial environment, standardised pharmaceutical psilocybin formulation and doses, thorough screening and diagnostic validation, regular clinical monitoring during tapering, and a larger sample than earlier reports. They also note dose‑related variability observed at lower doses (10 mg and 1 mg) where small inconsistencies were seen; the investigators suggest these patterns could indicate greater vulnerability of lower doses to any residual effects of prior medications, but they state that replication is required before drawing firm conclusions. Limitations acknowledged in the extracted text include the post hoc nature of the analysis and the absence of systematic capture of discontinuation withdrawal effects during the tapering window, which the authors recommend for future studies. The impact of the discontinuation requirement on recruitment could not be determined from the available data, since individuals unwilling to attempt withdrawal may not have completed Screening. The authors also note insufficient data to explain why participants who had discontinued prior antidepressants were more likely to restart antidepressant treatment within 12 weeks after psilocybin; several possible reasons are suggested (withdrawal physiology, patient preference, comorbidity, provider or access issues), but no firm data from this trial address these possibilities. Overall, the investigators conclude that supervised antidepressant discontinuation prior to psilocybin administration in this Phase II trial was practicable and did not produce meaningful worsening of depression or suicidality, nor did it reduce psilocybin's acute subjective effects or its short‑term antidepressant efficacy at the doses tested. They recommend further research to systematically capture withdrawal phenomena and to explore the effects of discontinuing specific antidepressant classes across indications.