The Death-Anxiety Trials:
Psilocybin and End-of-Life
Existential Distress

There is a particular terror that comes with a terminal diagnosis — not only the fear of the dying process, but the deeper existential dread of what it means to cease to exist. For most of human history, religion and community ritual provided a container for this terror. For many people today, those containers are absent, and the medical system that manages the body has almost nothing to offer the mind confronting oblivion. Between 25% and 75% of terminal cancer patients experience clinically significant anxiety or depression, with existential distress often proving more refractory than physical pain. Then came 2016, and two clinical trials that changed what science understands about dying.

The Existential Architecture of a Terminal Diagnosis

To understand why end-of-life anxiety is so resistant to conventional treatment, it is necessary to understand what it actually is. Death anxiety in terminal illness is not simply sadness about dying. It is a cluster of overlapping existential disruptions, each of which standard pharmacology is poorly equipped to address.

The first is what existential psychologists call annihilation anxiety — the terror of non-existence, of the permanent cessation of the self. Unlike the fear of pain or functional decline, which conventional anxiolytics can meaningfully attenuate, annihilation anxiety is fundamentally a problem of meaning and identity. The benzodiazepine that reduces the physiological arousal associated with anxiety leaves the existential question completely untouched. When the drug metabolises, the question is still there.

The second is loss of narrative coherence — the disruption of the life story that gives a person their sense of identity, continuity, and meaning. A terminal diagnosis violently interrupts the expected arc of a life. The future collapses. Plans, goals, and the scaffolding of prospective meaning dissolve. Yalom (1980) described this as the confrontation between the human need for cosmic significance and the universe’s fundamental indifference — a confrontation that most people spend their lives successfully avoiding, and that terminal illness makes impossible to defer.

The third is what Irvin Yalom termed “existential isolation” — the recognition that dying is ultimately a solitary act, that however much love surrounds a person at the end, they must pass through the final threshold alone. This awareness, in its full weight, is something that the social transactions of normal life are designed to obscure. Terminal illness removes that obscurity.

Breitbart et al. (2000) documented that will-to-live in cancer patients — the subjective intensity of the desire to remain alive — is far more strongly predicted by existential and psychological variables than by physical symptom burden. Patients with severe pain but intact meaning reported higher will-to-live than patients with well-controlled pain and profound existential despair. This finding reframes the problem: existential distress in terminal illness is not a secondary or epiphenomenal concern. It is the primary clinical challenge for a substantial proportion of patients, and its pharmacological treatment options are primitive.

Why Standard Palliative Care Falls Short

Palliative psychiatry has made significant advances in symptom management for the dying — opioids for pain, antiemetics for nausea, anxiolytics for acute distress, antidepressants for depressive episodes. But the standard pharmacological toolkit for existential distress is almost entirely borrowed from general psychiatry, and its performance in this population reflects that mismatch.

SSRIs require four to six weeks to produce clinically meaningful antidepressant effect. For a patient with weeks or months to live, this timescale is therapeutically irrelevant. Benzodiazepines produce anxiolytic effect acutely but at the cost of cognitive dulling, tolerance, and the rebound anxiety that accompanies discontinuation — in a population for whom cognitive clarity in the final months of life may be a priority. Antipsychotics prescribed for severe agitation in terminal patients carry sedation burdens that further diminish conscious experience at the very time when meaning-making is most urgent.

Psychotherapy approaches — particularly meaning-centred therapy (Breitbart et al.) and dignity therapy (Chochinov et al.) — show genuine promise for existential distress in terminal illness. But these therapies require multiple sessions across weeks, trained therapists willing to work in palliative settings, and patients with sufficient cognitive and functional capacity to engage. The gap between need and available resource remains enormous. In 2014, a landmark analysis estimated that fewer than 1% of terminal cancer patients with clinically significant psychological distress in the United States received specialist palliative psychological care.

What the field needed was a treatment capable of working rapidly — in hours rather than weeks — that addressed the existential dimension of distress directly rather than managing its physiological expression, and that could produce durable benefit from a small number of sessions. In 2016, two research groups reported that they had found it.

The 2016 Double Breakthrough: Griffiths and Ross

In December 2016, two research teams published simultaneous landmark reports in the Journal of Psychopharmacology. The papers arrived together — a deliberate coordination between the Johns Hopkins group and the NYU group — and together they constituted the most significant clinical evidence for psilocybin-assisted therapy since the psychedelic research moratorium of the early 1970s.

Griffiths et al. (2016) at Johns Hopkins enrolled 51 adults with life-threatening cancer diagnoses and clinically significant anxiety or depression. The study used a double-blind crossover design with two conditions: high-dose psilocybin (22 mg or 30 mg per 70 kg body weight) and an active control (a very low psilocybin dose of 1–3 mg/70 kg, effectively a placebo that matched the ritual context without producing significant psychedelic effects). Participants were randomised to the order of sessions, with a five-week crossover period. All sessions were conducted in a carefully designed, home-like environment with two trained facilitators present throughout the 6–8 hour session.

The primary outcomes were anxiety and depression, measured by established instruments including the Hospital Anxiety and Depression Scale (HADS) and the Beck Depression Inventory. At the 6-month follow-up — the primary endpoint — approximately 80% of participants showed clinically significant decreases in depression and anxiety. Approximately 60% achieved full clinical remission of their depressive symptoms. These effect sizes exceeded anything previously documented in this population for pharmacological or psychotherapeutic interventions.

Ross et al. (2016) at NYU enrolled 29 participants with cancer-related anxiety or depression in a crossover RCT comparing a single moderate-to-high dose psilocybin session against an active placebo control using niacin. The NYU trial found rapid and sustained reductions in anxiety and depression, with 60–80% of participants demonstrating clinically significant improvements at the 6.5-month follow-up. Critically, the effect was immediate: improvements were evident at the 1-day and 1-week post-session assessments, distinguishing psilocybin’s mechanism categorically from conventional antidepressants. The brain had changed before any drug concentration remained.

Both studies found that the intensity of the acute mystical experience during the psilocybin session was a significant predictor of long-term therapeutic outcome. Participants who reported higher scores on the Mystical Experience Questionnaire — a validated instrument measuring unity, transcendence, positive mood, sacredness, and noetic quality — showed greater reductions in anxiety and depression at follow-up. This relationship between experience quality and outcome would prove central to understanding how the treatment works.

The Mechanism: How Mystical Experience Changes the Relationship to Death

The finding that mystical experience mediates outcomes raises an immediate question: how does a subjective psychological experience — however profound — produce lasting changes in the neurobiological substrate of existential distress? Two levels of explanation, the phenomenological and the mechanistic, are necessary to answer it.

At the phenomenological level, the mystical experience that psilocybin reliably produces in a sufficient proportion of participants at therapeutic doses appears to provide a direct experiential encounter with a state that transcends the boundaries of ordinary selfhood. Descriptions from participants across multiple trials share a family resemblance: a dissolution of the boundary between self and world, a sense of unity that extends beyond the personal, a conviction that consciousness is not coextensive with the biological body, and a deeply felt positive valence that persists in memory long after the acute state has ended.

For a person facing death who has identified their entire existence with the physical body, this experience — regardless of its ultimate metaphysical status — provides an experiential ground for a different relationship to mortality. The terror of annihilation is premised on the assumption that the self that will cease to exist is the totality of what one is. An experience that temporarily but vividly dissolves that identification provides not an intellectual argument against the fear but an experiential counter-evidence. Participants in psilocybin death-anxiety trials consistently describe not an absence of fear but a shift in perspective — a sense that the dying of the body is not necessarily the end of something essential.

At the mechanistic level, psilocybin’s 5-HT2A agonism produces the neurological conditions for this experience through a specific pattern of network-level brain changes. Siegel et al. (2024) documented that psilocybin produces a profound and lasting desynchronisation of the Default Mode Network — the network most active during self-referential thought, autobiographical memory retrieval, and ego-maintenance. The DMN is, in many ways, the neural substrate of the ordinary sense of self. Its acute disruption under psilocybin is the mechanistic correlate of the phenomenological dissolution of self-boundaries that characterises the mystical experience. The weeks-long persistence of reduced DMN connectivity documented by Siegel et al. suggests that this is not merely an acute drug state but a structural reorganisation of the network that generates ordinary selfhood — a change in the architecture of ego, with lasting consequences for how a person relates to questions of identity and continuity.

Agrawal 2024: The Group Therapy Innovation

One of the significant practical challenges in scaling psilocybin-assisted therapy is the resource intensity of the standard model: each session typically requires two trained therapists present for six to eight hours, making per-patient cost prohibitive and access severely constrained. Agrawal et al. (2024) at Dana-Farber Cancer Institute and Sunstone Therapies addressed this directly with a Phase 2 trial evaluating psilocybin-assisted group therapy in cancer patients with comorbid Major Depressive Disorder.

The trial enrolled 30 participants, each of whom received a single 25 mg psilocybin dose in a group therapeutic setting — a significant departure from the standard individual therapy model used in the 2016 trials. The group format allowed multiple participants to receive the treatment simultaneously under shared therapist supervision, substantially reducing per-patient resource requirements while maintaining the core therapeutic architecture: careful preparation sessions, a carefully designed session environment, and structured integration support.

Primary outcome was assessed using the Montgomery-Åsberg Depression Rating Scale (MADRS), a clinician-rated scale sensitive to depressive symptom change. At 8 weeks post-treatment, participants showed a mean MADRS score reduction of 19.1 points — a reduction that represents clinically significant antidepressant response and that was achieved, again, from a single therapeutic session. The group format did not appear to substantially attenuate the treatment effect compared to the individual therapy model, suggesting that the active ingredient — the quality of the psilocybin experience and the integration framework — is not uniquely dependent on a dyadic therapeutic relationship.

Agin-Liebes 2020: Four and a Half Years Later

The most striking evidence for the durability of psilocybin-assisted treatment for end-of-life distress came from Agin-Liebes et al. (2020), who conducted a long-term follow-up of 15 participants from the original NYU cohort at 4.5 years after their single psilocybin session.

At that four-and-a-half-year mark, 60–80% of participants still met the clinical response criteria that had been established at the 6.5-month follow-up. The therapeutic benefit of a single psilocybin session — delivered years earlier — had persisted through the remainder of many participants’ lives. Some of these individuals had died by the time of the follow-up assessment; their reports came from surviving family members or from assessments conducted before death. This context lends particular weight to the finding: the treatment had altered not just a psychological measurement at a given time point, but the quality of the final years of a human life.

Additionally, 71–100% of participants rated the psilocybin experience as one of the most personally meaningful experiences of their lives — comparable to the birth of a child or the death of a parent in its significance. This is an extraordinary report for a clinical intervention. Effective antidepressants do not typically register as profoundly meaningful in the way that characterises other peak life experiences. Psilocybin-assisted therapy appears to produce not just symptom reduction but an experience of such significance that it restructures the participant’s sense of what their life has contained.

The Agrawal 2025 Two-Year Follow-Up: Group Therapy Durability

Agrawal et al. (2025) extended the Dana-Farber group therapy evidence with a 2-year follow-up assessment of 28 participants from the original 2024 trial. At the 2-year assessment, 53.6% of the cohort showed sustained significant reductions in depressive symptoms compared to their baseline pre-treatment assessment. Fifty percent were in sustained remission — meeting full remission criteria at both the 8-week and 2-year assessments.

These figures — more than half of terminal cancer patients with MDD in sustained remission at two years from a single group psilocybin session — are clinically extraordinary by the standards of any treatment modality in this population. Conventional antidepressants in cancer populations produce response rates of approximately 25–40% over a comparable period, with significantly higher dropout rates, side effect burden, and ongoing daily dosing requirements. The comparison is not perfectly equivalent — the 2025 follow-up is of a selected cohort, without a randomised control arm at the 2-year point — but the magnitude of the effect in absolute terms is striking.

The Agrawal 2025 findings also demonstrated that the group therapy model, initially evaluated primarily for its scalability advantages, appears to produce durability comparable to the individual therapy model documented in the 2020 Agin-Liebes follow-up. Whether this is because the group setting provides additional social and relational resources that support integration, or because the active ingredient — the quality of the acute psilocybin experience — operates largely independently of therapeutic format, remains an important open question for future research.

Why This Treatment Works: The Phenomenology of Dying Differently

The consistent finding across the end-of-life psilocybin literature is that what predicts therapeutic outcome is not simply the pharmacological action of psilocybin — not just the 5-HT2A activation and its downstream neuroplastic effects — but the specific quality of the subjective experience it enables. Across both the 2016 Johns Hopkins and NYU trials, and in the theoretical framework developed by Griffiths, Piff, and Yaden among others, the mystical experience is not incidental to the mechanism. It is the mechanism.

For terminal cancer patients, the specific therapeutic action of the mystical experience appears to operate through what might be called a re-hierarchisation of the self — a temporary but vivid experience of the self as something larger than, or continuous with, the dying body. Participants do not report that psilocybin made them unafraid of death in an intellectual sense. They report that after the experience, death felt less catastrophically final — that whatever they are seemed, experientially, to extend beyond the biological form that would cease. They retain the knowledge of their diagnosis and the expectation of death. What changes is the emotional valence of that knowledge, and the extent to which it colonises every remaining moment of conscious experience.

Piff et al. (2015) documented that awe-inducing experiences — characterised precisely by the sense of encountering something vast and beyond the self — reliably diminish the salience of the individual self and reduce mortality salience. Psilocybin reliably produces this state at a neurobiological level by disrupting the Default Mode Network that generates ordinary self-referential consciousness. The result is not sedation or suppression of existential awareness, but a genuine shift in the experiential relationship between the person and the prospect of their own death.

The Limitations, Stated Honestly

The end-of-life psilocybin literature is among the most compelling in the psychedelic research canon, but intellectual honesty requires clear articulation of what it does and does not show. The 2016 trials — the foundational studies — had sample sizes of 51 (Johns Hopkins) and 29 (NYU). These are sufficient to establish a signal and characterise effect size, but too small to detect rare adverse events or to identify the subset of patients for whom the treatment may be ineffective or harmful.

Blinding is a structural challenge in psilocybin research that is particularly acute in crossover designs like those used in 2016. A participant who receives a high dose of psilocybin and experiences 6–8 hours of profound altered consciousness knows with certainty that they are in the active condition. The very low-dose active control used by Griffiths et al. partially addresses this, but expectancy effects — the therapeutic benefit derived from believing one has received an effective treatment — cannot be fully disentangled from specific drug effects in the absence of a pharmacologically inert placebo that matches the phenomenological experience. No such placebo currently exists.

The long-term follow-up data from Agin-Liebes et al. (2020) is based on 15 participants — a subset of the already small NYU cohort — and involves participants who were willing and able to complete a 4.5-year follow-up, introducing potential survival and selection bias. The Agrawal 2025 2-year data similarly lacks a randomised control comparator at the follow-up point.

These limitations do not invalidate the evidence — they contextualise it. A treatment that produces clinically significant benefits in ~80% of participants from a single session, with effects documented to persist for years, in a population for whom existing treatments fail to achieve these outcomes, represents a signal too strong to dismiss. The appropriate scientific response is not scepticism but rigour: the Phase 3 randomised controlled trials that will confirm, characterise, and contextualise this signal are the work that comes next.

What the Evidence Actually Tells Us

The death-anxiety trials have produced something almost without precedent in modern psychiatry: a treatment that addresses the existential dimension of human suffering directly, rapidly, and durably. The Griffiths 2016 and Ross 2016 trials showed that approximately 80% of people with terminal cancer diagnoses and clinically significant death anxiety experienced meaningful relief after a single psilocybin session — a relief that appeared within hours and persisted for months. Agin-Liebes et al. showed that this relief persisted for years, in a population for whom “years” was often the entire remaining span of their lives. The Agrawal 2024 and 2025 findings extended this evidence to a group therapy model, demonstrating scalability without apparent loss of efficacy.

What ties these results together is not simply psilocybin’s pharmacology, but the specific type of experience it reliably enables: a temporary dissolution of the ordinary self, an encounter with something experienced as vast and continuous beyond the personal, and a lasting re-hierarchisation of what the dying person takes themselves to be. This experience does not resolve death. It changes the relationship to it.

For a species that has known for as long as it has existed that it will die, and that has spent most of that existence constructing containers of meaning and ritual to make that knowledge bearable, the discovery that a pharmacological compound can reliably produce something resembling those containers — in hours, in a clinical setting, in the middle of a hospital system — is among the most consequential things that psychedelic research has established. The trials are proceeding. The question is no longer whether this works. The question is how quickly it can reach the people who need it.