Psychiatry · Medical Literature
Neuroplastic white matter changes in patients with major depression following lysergic acid diethylamide treatment.
Abstract
The clinical trial NCT03866252 investigates the antidepressant effects of lysergic acid diethylamide (LSD) in 61 patients with major depressive disorder (MDD) randomized to low-dose LSD (LD-LSD; 2 × 25 μg) or moderate-to-high-dose LSD (HD-LSD; 100 μg followed by 200 μg) 4 weeks apart. Although the trial reports positive clinical outcomes, underlying mechanisms remain unclear. Here, we test whether LSD alters white matter (WM) microstructure, potentially reflecting enhanced neuroplasticity. Diffusion tensor imaging data from 35 patients (17 HD-LSD) include pre- and post-intervention scans. Voxel-wise permutation tests reveal group-by-time interactions, with increased fractional anisotropy (FA) in the internal and external capsule, sagittal stratum, and fornix/stria terminalis in the HD-LSD group. In this group, post-intervention FA values correlate with improvements in depressive symptoms at 2, 6, and 12 weeks, measured using the Inventory of Depressive Symptomatology (IDS-clinician rated [C] and IDS- self report [SR]). These findings suggest that LSD-induced WM microstructural changes are associated with antidepressant effects in MDD.
Clinical relevance
Does moderate-to-high-dose LSD (HD-LSD: 100 µg followed by 200 µg) alter white matter microstructure in patients with major depressive disorder, and are any such changes associated with clinical antidepressant outcomes over 12 weeks? __SNAPSHOT__ {"studyDesign":"Neuroimaging sub-study (diffusion tensor imaging; pre/post design with group-by-time interaction analysis using voxel-wise permutation tests) nested within a randomized, parallel, double-blind, low-dose-controlled phase 2 trial (NCT03866252)","setting":"University Hospital Basel, Switzerland; multisite academic psychiatric setting","sampleSize":"n=35 with usable DTI data (HD-LSD: n=17; LD-LSD: n=18), drawn from the parent trial of n=61 randomized patients","population":"Adults >25 years, DSM-5 diagnosis of moderate-to-severe MDD; IDS-C score 24–46 and IDS-SR score 26–48 at baseline (extremes of severity excluded). HD-LSD group: mean age 41.8 ± 12.3 years, 7 female/10 male. LD-LSD group: mean age 38.1 ± 11.7 years, 6 female/12 male. Notably, the HD-LSD neuroimaging subgroup had significantly more severe baseline depression scores than the LD-LSD subgroup.","intervention":"Two sessions of moderate-to-high-dose LSD (100 µg at session 1; 200 µg at session 2) administered 4 weeks apart, combined with supportive psychotherapy","comparator":"Two sessions of low-dose LSD (25 µg × 2, active placebo/control) administered 4 weeks apart, combined with supportive psychotherapy","primaryOutcome":"White matter microstructural change assessed by fractional anisotropy (FA) on diffusion tensor imaging (DTI), using voxel-wise group-by-time permutation tests; secondary: correlation of post-intervention FA with changes in IDS-C and IDS-SR scores at 2, 6, and 12 weeks post-intervention","followUp":"12 weeks post-second administration (neuroimaging: pre- and post-intervention scans; clinical: 2, 6, and 12 weeks)","fundingSource":"Gertrud Thalmann Fund for depression research (non-profit/charitable); knowhow and data licensed to Mind Medicine, Inc., but the sponsor had no role in financing, planning, conducting, or publishing the study"}
Key findings
In the HD-LSD group, voxel-wise permutation tests revealed significant group-by-time interactions characterized by increased fractional anisotropy (FA) in the internal capsule, external capsule, sagittal stratum, and fornix/stria terminalis — white matter tracts implicated in depression-associated structural abnormalities. No such increases were observed in the LD-LSD (active comparator) group. Post-intervention FA values in the HD-LSD group correlated with improvements in depressive symptoms at 2, 6, and 12 weeks on both the IDS-C (clinician-rated) and IDS-SR (self-report) scales. Specific FA effect sizes, correlation coefficients, and p-values for individual tract analyses are not reported in the abstract. From the parent clinical trial (Müller et al., Med 2025), the IDS-SR least-squares mean change from baseline was −3.9 (LD) vs. −11.8 (HD) (difference: −7.9; 95% CI, −16.0 to 0.3; effect size: −0.5; p=0.059); IDS-C least-squares mean change was −3.6 (LD) vs. −12.9 (HD) (difference: −9.2; 95% CI, −17.1 to −1.3; effect size: −0.6; p=0.023), though this did not survive adjustment for baseline depression severity (p=0.086 after correction). NNT is not calculable from the available data. No specific adverse events are reported in this neuroimaging sub-study abstract; the parent trial noted one participant with systolic BP >180 mmHg in the HD group, with vital signs normalized at end-of-study.
Practice implications
This study provides the first human in vivo evidence that moderate-to-high-dose LSD is associated with increased white matter FA in limbic and corticospinal tracts relevant to MDD pathophysiology, and that these changes correlate with sustained symptom improvement over 12 weeks. However, several critical interpretive constraints apply. First, this is an abstract-only summary, and full statistical details (effect sizes per tract, corrected p-values, correlation coefficients) are not available for scrutiny. Second, the neuroimaging subsample is small (n=35 total, n=17 HD-LSD), substantially reducing the power of this sub-study and raising concerns about false positive findings without correction for multiple comparisons across brain voxels (though permutation testing was used, exact thresholds are not reported here). Third, and critically acknowledged by the lead author, it cannot be determined whether white matter changes are a direct neurobiological effect of LSD or a secondary consequence of clinical improvement (i.e., reverse causation). Fourth, the HD-LSD neuroimaging subgroup had significantly more severe baseline depression than the LD-LSD subgroup, introducing a confound that could inflate apparent group differences. Fifth, the parent trial's primary IDS-C result did not survive covariate adjustment for baseline severity. The findings do not apply to patients excluded from the trial (extreme severity, those not consenting to neuroimaging, age ≤25). Adequate blinding was not achieved in a considerable number of cases in the parent trial, which may have biased clinician ratings. These data are mechanistically hypothesis-generating and should not alter clinical prescribing practice. Replication in larger, adequately powered trials with an inactive placebo arm is required. __LIMITATIONS__ Very small neuroimaging subsample (n=35; n=17 in HD-LSD arm), substantially limiting statistical power and generalizability.,Significant baseline imbalance in depression severity between HD-LSD and LD-LSD neuroimaging subgroups, a potential confound not fully addressed in the abstract.,Active low-dose LSD comparator (not inactive placebo), making it impossible to fully disentangle drug-specific neurobiological effects from psychotherapy effects or natural illness trajectory.,Blinding was not effective in a considerable number of participants in the parent trial, potentially biasing clinician-rated outcomes (IDS-C).,Causal directionality of the FA–symptom correlation is unestablished: white matter changes may be a consequence of mood improvement rather than the mechanism of action.,Parent trial IDS-C primary endpoint did not survive correction for baseline severity (p=0.086 adjusted); IDS-SR primary endpoint was not statistically significant (p=0.059).,Only abstract is available for this neuroimaging paper; full statistical details (per-tract effect sizes, exact permutation p-values, correlation coefficients, multiple comparison corrections) cannot be independently evaluated.,DTI fractional anisotropy is a surrogate neuroimaging marker whose relationship to actual axonal structure, myelin integrity, or clinically meaningful plasticity in humans remains incompletely understood.,The study does not report long-term safety data beyond 12 weeks or address repeated-dose, dose-escalation, or rechallenge risks.,No inactive placebo arm; industry knowhow/data licensing to Mind Medicine, Inc. may represent a potential conflict of interest despite stated non-involvement in the study. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ This small neuroimaging sub-study (n=35) provides preliminary, association-level evidence that moderate-to-high-dose LSD-assisted therapy is accompanied by increased white matter FA in limbic and projection tracts in MDD, with FA values correlating with 12-week symptom improvement; however, causality cannot be established and the parent trial's primary clinical outcomes did not consistently survive covariate adjustment. These mechanistic findings are hypothesis-generating and, given the small sample, baseline imbalances, imperfect blinding, and surrogate neuroimaging endpoints, should not be interpreted as sufficient evidence to recommend LSD-assisted therapy in clinical practice. Confirmation in larger, adequately powered phase 3 trials with an inactive placebo control is required before clinical application can be considered.