Banisteriopsis caapi, a Forgotten Potential Therapy for Parkinson’s Disease?

The paper reviews the history, pharmacology, and clinical evidence for harmine-containing preparations derived from Banisteriopsis caapi and Peganum harmala as potential therapies for parkinsonism. Earlier research from the 1920s and 1930s identified harmine (also called banisterine, telepathine or yag eine) as the active alkaloid in both plants, and early investigators reported rapid but variable improvements in motor function after parenteral administration. Interest waned with the advent of other agents considered more predictable or effective, and because oral preparations were thought to have limited efficacy. Djamshidian and colleagues set out to summarise the early clinical trials and case series, to outline the known pharmacology of harmine, and to reconsider whether harmine or B. caapi extracts merit renewed investigation for Parkinson's disease (PD). The review also situates harmine among other monoamine oxidase (MAO) inhibitors and discusses safety signals that influenced its historical abandonment, while noting more recent preclinical and small clinical data that suggest mechanisms relevant to PD and a rationale for further controlled trials.

The extracted text presents a narrative, historical review rather than a formal systematic meta-analysis. It collates early 20th-century clinical observations and case series, company records (notably Merck's marketing and internal reports), later experimental animal data, and at least one modern double-blind clinical trial. The paper traces pharmacological studies of harmine and related beta-carbolines and summarises clinical reports of therapeutic use in postencephalitic parkinsonism, paralysis agitans, and other extrapyramidal conditions. The extracted text does not describe a reproducible search strategy, inclusion/exclusion criteria, databases searched, date range, or risk-of-bias assessment. Instead, the study synthesises published clinical anecdotes, early investigator reports (including dosing and routes of administration), experimental toxicology and pharmacology studies in animals and in vitro, and a 2001 double-blind trial of B. caapi extract in de novo PD patients. Where numerical results or dosing regimens are reported in the original sources, the review reproduces these details; where methodological detail from the original trials is absent in the extraction, the authors' reporting is summarised conservatively.

Historical clinical reports from the late 1920s and early 1930s described rapid-onset, sometimes dramatic improvements in motor function after parenteral administration of banisterine/harmine extracts. Lewin and colleagues reported subjective and objective gains in gait and motor control following subcutaneous (SC) injections (reported doses 25–70 mg in some early experiments). Beringer treated 15 postencephalitic patients with P. harmala extracts and described improvements in akinesia, rigidity and oculogyric crisis in some cases, with onset around 30 minutes and duration ranging from several hours up to a few days. Merck marketed "Merck's harmine" in injectable, capsule and suppository forms in 1928; average daily injection doses reported in contemporaneous work were around 20–40 mg/day, with higher-dose side effects appearing at about 60 mg. Several centres reported variable outcomes. Schuster noted rigidity improvement within 15 minutes after SC injections in 18 patients, with effects lasting 2–6 hours and up to 7 days in some. Rustige treated 18 patients with 3–50 mg injections and observed objective improvement within 20–30 minutes lasting up to 1 hour in many; 13 of 18 showed increased speed of movement, 6 had rigidity improvement and 3 had complete resolution of pronounced rigidity. Mueller reported that, after combined SC and oral regimens over 3–4 weeks, 26% of patients had "excellent" and 37% "good" improvement. However, other contemporaneous reports were negative or sceptical: Hill and Worster-Drought treated 38 postencephalitic patients (after withdrawing hyoscine) and reported no improvement with harmine; almost all patients worsened, improving only after hyoscine was restarted. Gausebeck found scopolamine superior to harmine in 9 of 10 cases. Beringer and others emphasised that benefits were often short-lived, variable between patients, and sometimes accompanied by worsening of tremor. Adverse effects reported historically included yawning, nausea, vertigo, headache, agitation, tinnitus, bradycardia, orthostatic dysregulation, hypersalivation, occasional vomiting and, in at least one high-dose intramuscular report, severe intoxication with cramps, tremor, delirium and faintness. Some investigators suspected suggestion/placebo effects in early positive reports. More recent and preclinical data summarised in the review show that harmine and related beta-carbolines are MAO inhibitors (nonselective MAO inhibition demonstrated), can induce striatal dopamine release in vivo, act as NMDA receptor antagonists, inhibit the DYRK1A kinase, and in animal models support dopaminergic neuron survival (for example in MPTP-treated mice) and reduce glutamate-mediated excitotoxicity via upregulation of glutamate transporter 1. On the question of tremor, harmine produces an acute, dose-dependent 8–14 Hz action and postural tremor in rodents, cats and monkeys; a transient coarse tremor has also been reported in humans given high doses of B. caapi. The animal tremor is attributed to activation of inferior olivary nuclei and cerebellar circuits; pharmacological similarities between harmaline-induced tremor and essential tremor are noted. Repeated high-dose harmaline in animals induces tolerance and diminution of tremor, whereas acute intoxication can be severe. A double-blind, placebo-controlled study cited in the extraction enrolled 30 drug-naïve, de novo PD patients in 2001: 15 received 200 mL of B. caapi extract and 15 received taste- and colour-matched placebo. Side effects were reported in all active participants (diarrhoea, nausea, one case of transient visual hallucinations). Motor scores (UPDRS part III) in the active group fell from a baseline mean of 54.4 to 41.4 at 60 minutes, to 22.4 at 120 minutes, and to 25.6 at 4 hours. However, tremor (rest, action and postural) worsened in all patients. The extract used in that trial may have contained multiple alkaloids and is not directly comparable to purified harmine or the historical "Merck's harmine" SC preparations. The review also discusses MAO pharmacology: harmine is a nonselective MAO inhibitor with prominent MAO-A activity. MAO-A inhibition can potentiate levodopa responses but raises safety concerns (hypertensive "cheese effect" with irreversible MAO-A inhibitors, serotonin syndrome risk when combined with certain antidepressants). By contrast, MAO-B inhibitors (selegiline, rasagiline) have been more thoroughly investigated in PD and are considered safer in combination with L-dopa, though they also show variable patient responses and tremor worsening at higher doses.

Djamshidian and colleagues interpret the assembled historical and modern evidence as indicating that harmine and B. caapi extracts possess a pharmacological profile relevant to PD: MAO-A inhibition, NMDA antagonism, potential to release striatal dopamine, DYRK1A inhibition and possible neuroprotective effects in preclinical models. They suggest that these mechanisms could explain the rapid symptomatic improvements reported in some early and one recent trial, particularly for rigidity and bradykinesia. The authors note substantial heterogeneity and inconsistency in the clinical literature. Early positive reports were often based on parenteral (SC) administration of relatively purified harmine, with rapid onset but short duration of effect and substantial interpatient variability. Other contemporaneous investigators failed to reproduce benefits and raised concerns about placebo effects, confounding from withdrawal of established anticholinergic treatments, and dose-related adverse events including tremor and intoxication. Modern controlled data are limited to small trials of herbal extracts, which complicates comparison with historical studies of purified alkaloids. Key limitations acknowledged include the historical nature of many reports, lack of standardised preparations or dosing, heterogeneous patient populations (postencephalitic parkinsonism versus idiopathic PD), short duration of observed benefit in many cases, and safety considerations inherent to MAO-A inhibition (dietary tyramine interactions, hypertensive crises, and risk of serotonin syndrome when combined with other drugs). The authors also emphasise that harmine can acutely provoke tremor in animal models and, at high doses, in humans, which may limit its applicability. Implications discussed by the authors are cautious: given the mixed historical clinical results but supportive preclinical mechanisms, they argue that harmine or B. caapi extracts could warrant carefully controlled modern trials to determine efficacy, safety, optimal formulation and route of administration. They note that harmine's rapid onset could complement existing PD therapies if safety and tolerability can be managed. The review also draws a parallel to other therapies (for example apomorphine) that were initially abandoned but later rehabilitated with improved delivery and adjunctive measures, suggesting that revisiting discarded agents with contemporary methodology can be productive.

The authors conclude that harmine has a multifaceted pharmacological profile—MAO-A inhibition, serotonin receptor affinity, NMDA antagonism, potential antioxidative and neuroprotective properties, and capacity for striatal dopamine release—which supports reconsideration of harmine or Banisteriopsis caapi extracts for PD. They state that although harmine is unlikely to be a highly potent antiparkinsonian agent, it could be as efficacious as currently available selective MAO-B inhibitors in some patients, particularly given its rapid action. The authors recommend further controlled trials, noting that B. caapi extracts may be superior to P. harmala extracts and that modern evaluation should address formulation, dosing, efficacy, and safety.