Relief from intractable phantom pain by combining psilocybin and mirror visual-feedback (MVF)

Phantom limb phenomena are extremely common after amputation, with about 95% of amputees experiencing phantom sensations and roughly two-thirds reporting severe phantom limb pain (PLP). Earlier experimental work has linked these phenomena to cortical reorganisation—such as invasion of the deafferented foot area by neighbouring thigh representations in primary somatosensory cortex—and to failures of visuo-motor feedback that can produce a 'learned paralysis' of the phantom. Mirror visual feedback (MVF), in which a mirror superimposes the intact limb's reflection onto the felt location of the phantom, has previously been shown to restore a visuo-motor loop and relieve pain in many patients. The authors also situate their work within the literature on mirror systems, including touch- and pain-related mirror responses in secondary somatosensory and cingulate regions, and suggest these systems may contribute to inter-personal referral of sensation when afferent veto signals are absent. Ramachandran and colleagues present a single-case series of informal experiments in a 35-year-old man (‘‘AL'') with a right lower-leg amputation and intractable PLP. The report focuses on several phenomena: topographic referral of sensation to the phantom, MVF effects including relief of pain and of dysuria-associated pain, referral of warmth from an artificial flame, evocative re-experiencing of implanted nails with a telescoping-nail illusion, and most centrally the effects of psilocybin alone and in combination with MVF. The authors aim to describe these observations, consider mechanisms (notably 5-HT2A-mediated plasticity and cross-modal facilitation), and to raise hypotheses for further controlled work.

This paper reports observational experiments conducted with a single patient, AL, who suffered a traumatic right lower-leg amputation following a motor vehicle accident. At the time of testing he was 35 years old and had previously undergone surgical fixation with three nails; he reported persistent, often severe phantom pain despite analgesics including opioids and pregabalin, and he had tried medicinal cannabis without benefit. All reported interventions and observations were performed clinically and informally; the extracted text does not present a formal protocol, randomisation, blinding, standardised pain scales, or statistical analysis plan. Experimental manipulations centred on mirror visual feedback and person-to-person referral of sensation. A large mirror (approximately 60 × 150 cm) was placed parasagittally between the intact limb and the space of the phantom to visually 'resurrect' the missing leg. A volunteer (V) positioned her right bare leg to overlap or approximate the felt location of AL's phantom; the investigators then applied touch, stroking, tapping or massage to V's leg, or used a plastic foot, while AL observed. The team also used an artificial flame to approach V's foot and a telescoping metal antenna to mimic removal of implanted nails. The authors report that effects were observed on two separate occasions separated by one week. Psilocybin was administered by AL outside the laboratory context as part of his own experimentation; doses are described in the paper as dried weight of whole mushrooms and were varied, but the extracted text does not clearly report specific dose amounts, timing, or administration protocol. The investigators observed responses to psilocybin alone and to psilocybin paired with MVF. Outcomes were recorded descriptively (subjective reports of sensation and pain intensity, reports of paroxysmal episodes, and temporal characteristics such as onset latency and duration); no formal quantitative outcome instruments or objective physiological measures are reported in the extracted text.

Cortical reorganisation and referred sensations: Touching particular points on AL's residual limb and adjacent thigh produced sharply localised sensations in specific locations of the phantom toes, consistent with reorganisation of the Penfield somatosensory map where thigh input contacts the deafferented foot region. Mirror-action and referred touch: When AL watched the volunteer V's right leg being stroked, tapped or massaged while the volunteer's leg was overlapped with the felt location of the phantom, he reported distinct, localised touch sensations on his phantom. With his eyes closed he did not feel these referrals. Viewing the volunteer's foot plantar-flex from a dorsiflexed posture produced an equivalent, felt plantar-flexion of the phantom and immediate analgesia on many trials; MVF relieved cramping and painful sensations on about half of the trials reported. Using a plastic foot in place of the volunteer produced similar referrals. Massage of the intact limb sometimes produced referred massage sensations and partial pain relief of the phantom. Effects on dysuria: Placing a mirror parasagittally beside AL's penis during micturition produced a striking reduction and apparent permanent resolution of intense stinging pain he had experienced with urination and ejaculation. Phantom heat and flame: Bringing an artificial (toy) flame close to the volunteer's foot produced a subjective warmth in AL's phantom after a 7–14 s delay; AL described the sensation as soothing and pain-relieving. This effect was reproducible across two trial sets separated by a week. Phantom pin removal: Touching precise sites corresponding to the surgical nails evoked a distinct sensation of embedded nails. Using a telescoping antenna to mimic nail withdrawal produced a metallic sliding sensation in the phantom and the subjective impression of removal. Psilocybin, alone and combined with MVF: Prior pharmacologic treatments were largely ineffective for AL. He self-administered psilocybin-containing mushrooms in varying doses (exact doses not clearly reported in the extracted text). A single large psilocybin dose produced immediate, substantial pain relief lasting about 3 hours. Psilocybin alone afforded only limited, transient long-term benefit in the weeks of experimentation. However, when psilocybin was paired with MVF the patient reported a striking response: a reported 50% reduction in pain and complete elimination of paroxysmal pain episodes over a 2-week period, with pain reduction sustained for a further three weeks, after which AL was able to discontinue the drug–mirror combination. The authors note that these observations derive from a small number of trials and that detailed dose–response relationships were not established; figures referenced in the paper reportedly show dose comparisons but are not included in the extracted text. The investigators acknowledge that placebo responses are common in analgesic trials but argue that AL's prior refractoriness to multiple treatments and the temporal association with MVF–psilocybin episodes make spontaneous remission less likely.

Ramachandran and colleagues interpret these observations as supporting several mechanistic and clinical ideas. They view inter-subject referral of touch, heat and pain as arising from activation of mirror-like neuronal systems in sensory and insular cortices; in the deafferented state, the usual peripheral 'veto' signal is absent, allowing visual or observed tactile/thermal stimuli to be experienced as somatic sensations in the phantom. The authors propose that psilocybin, via agonism at 5-HT2A receptors, may promote cross-modal functional connectivity and neural plasticity, thereby enhancing the efficacy and durability of MVF in 'unlearning' learned paralysis and reducing PLP. This hypothesis is used to explain the reported synergistic, longer-lasting analgesic effect when psilocybin was combined with MVF compared with either intervention alone. They also suggest that heat referral from an artificial flame implicates neurons with mirror-like properties for thermal sensation—possibly acquired by Hebbian association between the sight of fire and the felt warmth—located in insular cortex. On the clinical side, the persistence of relief from micturition-associated pain after mirror placement, and the evocative pin-removal illusion, are presented as examples of how targeted visual input can access specific phantom memories and sensations with therapeutic potential. The authors acknowledge important limitations: the report concerns a single patient, the interventions and outcomes were informal and observational, dosing and trial numbers are limited and not controlled, and placebo effects are possible. They call explicitly for controlled, placebo-controlled trials to confirm efficacy, to characterise dose–response relationships, and to test whether other serotonergic agonists produce similar facilitation of MVF. Finally, the investigators argue that these findings challenge strictly modular, feedforward models of brain function, favouring a dynamic, interactive view in which perception emerges from interactions among internally and externally driven signals; they propose this perspective as one motivation for further mechanistic and clinical research.

The authors conclude that their informal tests in a lower-limb amputee reproduced known MVF phenomena and revealed additional effects with potential therapeutic relevance. They report replication of inter-subject somatosensory referral, demonstration of visually induced thermal sensation in the phantom, alleviation and apparent permanent resolution of micturition-related pain via mirror placement, and evocative re-experiencing of implanted nails with a telescoping-nail illusion. Most importantly, they describe a marked and long-lasting reduction of phantom pain when psilocybin was combined with MVF—an effect they attribute to serotoninergic promotion of cross-modal plasticity that enhances and prolongs the benefits of mirror therapy. Ramachandran and colleagues emphasise the need for additional controlled trials to confirm these findings and to investigate mechanisms, dose–response relations, and generalisability, and they suggest these observations support a model of the brain as dynamically interacting, plastic modules rather than as an inflexible, strictly hierarchical processor.