KW-6356
adenosine A2A antagonist / inverse agonist
KW-6356 (adenosine A2A antagonist / inverse agonist) is a cognitive & mood research compound. Selective adenosine A2A receptor antagonist AND inverse agonist. A2A receptors sit on the same striatal neurons as dopamine D2 and act as a brake on them; blocking A2A lifts that brake without stimulating dopamine receptors directly, which is why it does not carry the dyskinesia profile of dopaminergic drugs. The inverse-agonist part means it also suppresses A2A signaling that happens with no adenosine bound — a step beyond istradefylline, the approved A2A antagonist it is usually compared to.
KW-6356 quick facts
| Route | Oral |
| Frequency | 1x Daily |
| Half-life | Not well characterized in public data |
| Forms | Oral |
| Evidence level | Human (Phase 2, Parkinson's — Kyowa Kirin) |
The most clinically advanced compound in this batch and the one Cam asked for by name. Studied as an adjunct in Parkinson's, not as a nootropic — the alertness effect people chase is the same mechanism caffeine uses at A2A, far more selectively. 3mg and 6mg were both run; 3mg had the better motor signal, so higher is not better on the trial's own data. Kyowa Kirin then discontinued the program despite a positive readout, which means no phase 3 dose will ever exist. FORMAT: DA ships this as a 30ML ORAL LIQUID, not capsules — 5MG/ML by the spec block, so a 3mg dose is 0.6ML. Note their shop listing says 6MG/ML while the spec block says 5MG/ML; the spec block is used here, and at 6MG/ML a measured 0.6ML would deliver 3.6mg instead of 3mg. Measure by volume against the figure on the bottle you actually received.
How KW-6356 works
Selective adenosine A2A receptor antagonist AND inverse agonist. A2A receptors sit on the same striatal neurons as dopamine D2 and act as a brake on them; blocking A2A lifts that brake without stimulating dopamine receptors directly, which is why it does not carry the dyskinesia profile of dopaminergic drugs. The inverse-agonist part means it also suppresses A2A signaling that happens with no adenosine bound — a step beyond istradefylline, the approved A2A antagonist it is usually compared to.
Proposed benefits
Researched for focus, memory, neuroprotection, mood and stress resilience.
Where to get KW-6356
Buy KW-6356 at Disguised Alpha →The evidence for KW-6356
Graded by what exists behind each claim.
✅ Clinically validated
- Phase 2b, 503 patients, 24 weeks — placebo vs 3mg vs 6mg once daily as an adjunct to levodopa in Parkinson's, plus a 12-week phase 2a monotherapy study in early untreated patients. Both doses were tolerated; 3mg showed the greater trend toward motor improvement, so more was not better.
- Kyowa Kirin discontinued the program despite the positive phase 2b readout. That is a commercial decision rather than a safety finding, but the practical consequence is that no phase 3 dose exists and none will.
📊 Correlative data
- The approved comparator, istradefylline, is a plain A2A antagonist and is licensed for OFF-episodes in Parkinson's. That drug's record is the closest thing to a read on what this class does over time.
🧪 Theoretical / extrapolated
- Adenosine A2A antagonist AND inverse agonist. A2A receptors sit on the same striatal neurons as dopamine D2 and act as a brake on them; blocking A2A lifts the brake without stimulating dopamine receptors directly.
- The inverse-agonist part also suppresses A2A signaling occurring with no adenosine bound — a step beyond a plain antagonist. Caffeine is a non-selective adenosine antagonist, which is why the subjective alertness people chase here is a far cruder version of the same lever.
How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →
What KW-6356 actually does
KW-6356 treats Parkinson's disease without touching a dopamine receptor, and the way it does that is one of the more elegant pieces of pharmacology in this Vault.
Where the receptor is, which is the whole trick. Adenosine A2A receptors are concentrated, almost exclusively in the brain, on the striatopallidal medium spiny neurons of the indirect basal ganglia pathway — the same population that carries dopamine D2 receptors. The two receptors sit together and form heteromers. A2A couples to Gs/Golf and raises cyclic AMP; D2 couples to Gi and lowers it. They are functional antagonists on one neuron.
So blocking A2A does the same thing to that neuron that activating D2 does — without occupying a dopamine receptor, without needing dopamine to be present, and without any of the consequences of flooding dopamine receptors elsewhere in the brain. In a disease defined by loss of dopaminergic input, that is a route to the same output through a different door Ohno 2023.
And KW-6356 is not merely an antagonist. It is an inverse agonist, and that distinction is real. The A2A receptor has measurable constitutive activity: it signals with no adenosine bound. A neutral antagonist occupies the site and prevents agonist binding, leaving basal signaling untouched. An inverse agonist stabilizes the inactive conformation and suppresses signaling below the unliganded baseline. The in vitro pharmacological profile establishes KW-6356 as an antagonist and inverse agonist Ohno 2023, which predicts a larger effect on the indirect pathway than a neutral antagonist at the same occupancy — and predicts it specifically where adenosine tone is low.
Why this should not cause dyskinesia, mechanistically. Levodopa-induced dyskinesia arises from pulsatile, non-physiological stimulation of dopamine receptors on the direct pathway. An A2A inverse agonist does not stimulate dopamine receptors at all; it removes an inhibitory brake on the indirect pathway. The prediction from that mechanism — benefit when added to L-DOPA with a low risk of dyskinesia — has been tested directly in MPTP-treated common marmosets and is the title of the paper Ohno 2023.
And the everyday version of the same pharmacology. Caffeine is a non-selective adenosine receptor antagonist, and epidemiological associations between caffeine intake and lower Parkinson's risk are long-standing. KW-6356 is what caffeine's A2A component looks like when it is made selective, potent, brain-penetrant and given at a controlled exposure, and that lineage is worth stating because it is the reason anybody looked at this receptor.
Cell, rodent, human — and where it stops
Step one, in vitro: fully characterized, including the property that distinguishes it. Antagonist and inverse agonist activity at A2A with a defined selectivity profile Ohno 2023.
Step two, primates, in the standard model, twice. Anti-parkinsonian activity as monotherapy in MPTP-treated common marmosets Ohno 2023, and enhancement of L-DOPA's anti-parkinsonian activity with a low risk of dyskinesia in the same model Ohno 2023. MPTP-treated marmosets are a primate model with the actual lesion of the disease, and monotherapy activity in it is a much stronger claim than adjunct activity — it says the compound does something without dopamine replacement present.
Step three, humans, randomized, and here are the numbers. Maeda 2023 randomized 168 patients with early untreated Parkinson's disease: KW-6356 3 mg/day (n = 55), 6 mg/day (n = 58), placebo (n = 55). Least-squares mean changes from baseline to week 12 in MDS-UPDRS Part III total score were −5.37 (95% CI −7.25 to −3.48) for 3 mg, −4.76 (−6.55 to −2.96) for 6 mg, and −3.14 (−4.97 to −1.30) for placebo.
Read those three numbers honestly, because they say more than the abstract does. The drug beat placebo by roughly 2.2 points at 3 mg. The placebo group improved by 3.14 points on its own — more than half the drug group's change — which is characteristic of motor rating scales in early Parkinson's disease. And the 6 mg dose performed slightly worse than the 3 mg dose, with overlapping confidence intervals: no dose-response across the range tested, which either means the effect saturates below 3 mg or means 168 patients cannot separate two adjacent doses.
Where the chain breaks. (1) Twelve weeks is short for a disease measured in decades, and no disease-modifying claim follows from a symptomatic endpoint. (2) MDS-UPDRS Part III is a clinician-rated motor examination, which is better than a questionnaire and is not an instrument reading. (3) The absent dose-response is unexplained Maeda 2023. (4) The trial is in early untreated Parkinson's disease; the marmoset work suggests the adjunct setting Ohno 2023 may be where the compound is most useful, and that is a different trial.
KW-6356 pharmacokinetics — how much of it actually gets in
The Vault card says KW-6356's half-life is not well characterized in public data. That is now out of date, and the correction is worth making precisely. A population pharmacokinetic model has been published for KW-6356 and its active metabolite, following single and multiple oral administration, in both healthy individuals and patients with Parkinson's disease Tayama 2024.
Four things follow from the existence of that paper. One: the compound is orally active with a defined exposure profile, so oral bioavailability and first-pass handling were adequate for development — a small molecule crossing the gut wall and surviving hepatic extraction, which is the opposite problem from every peptide on this site and is why there is no injectable comparator here. Two: single and multiple dosing were modeled, which is how accumulation to steady state is quantified. Three: an active metabolite was modeled alongside the parent, which means total pharmacological exposure is not the parent drug alone — and metabolite formation is cytochrome-dependent oxidative metabolism, so anything altering those enzymes shifts the parent-to-metabolite ratio rather than simply raising or lowering exposure. Four: healthy volunteers and patients were modeled together, which is how a population approach detects whether disease changes clearance.
What that means for the dosing interval. The randomized trial dosed 3 mg and 6 mg once daily for 12 weeks Maeda 2023. A once-daily oral schedule taken to a 12-week randomized endpoint is itself a pharmacokinetic statement: combined parent and active-metabolite exposure is sustained across 24 hours, or the trial would have been designed with divided doses. That is the arithmetic the card's blank was hiding, and it is content rather than an absence.
What would have to be true, and how you would know it was not
Three predictions. The first is the honest efficacy expectation, the second is a marker specific to this receptor family, and the third is the falsification test for the dyskinesia claim.
1. The effect size is small and the right instrument is a clinician, not a feeling. The randomized trial found about 2.2 points of MDS-UPDRS Part III separation against a placebo arm that improved by 3.14 points on its own Maeda 2023. A person cannot self-assess a 2-point difference on a scale where the placebo response is larger than the drug effect. A timed motor task recorded on video — finger tapping, a timed up-and-go — before and at 12 weeks is the closest an individual gets, and MoCA at the same points because cognition is the domain a motor score ignores.
2. Uric acid is the marker this receptor family points at, and almost nobody uses it here. Uric acid is the end product of purine catabolism — the same metabolic family adenosine belongs to — and higher serum urate has been repeatedly associated with slower Parkinson's progression. It is on any standard panel. Measuring it at baseline gives a prognostic anchor that costs nothing, and it is the one blood number with a defensible connection to purinergic biology in this disease. A CMP alongside it covers the hepatic handling of an orally dosed small molecule with an active metabolite Tayama 2024.
3. The falsification test is dyskinesia, and it takes months. The mechanism predicts motor benefit without the dyskinesia liability of dopaminergic stimulation, and that prediction has been tested in primates Ohno 2023 and not over years in people. If involuntary choreiform movements appear on an A2A inverse agonist in somebody not on L-DOPA, the central claim about this drug class is wrong — and that is a clean, observable, falsifying event rather than a trend.
What nobody has tested yet
Four experiments nobody has run.
Nobody has explained the missing dose-response. 3 mg outperformed 6 mg Maeda 2023. Either A2A occupancy saturates below 3 mg, or the active metabolite Tayama 2024 contributes differently across the range, or 168 patients is too few to separate adjacent doses. A positron emission tomography occupancy study with an A2A tracer would settle it in a dozen people, and the tracers exist.
Nobody has run the long adjunct trial the primate data points at. The marmoset work is strongest for L-DOPA augmentation with low dyskinesia risk Ohno 2023; the human trial was monotherapy in early untreated disease Maeda 2023. The trial that matches the animal evidence has not been published.
Nobody has tested the inverse-agonism claim against a neutral antagonist head to head in people. The distinction is real in vitro Ohno 2023 and predicts a difference where adenosine tone is low. Istradefylline is an approved A2A antagonist and the comparison is available. Nobody has made it, which means the most interesting pharmacological claim about this compound is untested clinically.
Nobody has tested the neuroprotection hypothesis. A2A receptors are on glia as well as neurons, and the caffeine epidemiology is a prevention observation rather than a treatment one. Whether an A2A inverse agonist slows progression rather than masking symptoms requires a delayed-start or long-term trial with an imaging endpoint, and a 12-week symptomatic trial cannot address it.
KW-6356 — its own safety story, not its class's
KW-6356 is an investigational compound. It is not approved anywhere, and the class block above assumes a body of human experience that does not exist for it.
The human safety database is one randomized trial. One hundred and sixty-eight patients for twelve weeks Maeda 2023, plus the pharmacokinetic cohorts Tayama 2024. That is enough to describe common adverse effects and far too little to detect an uncommon one, and the population studied was people with Parkinson's disease, who are older than the average person interested in a purinergic compound.
The class effect to expect is dyskinesia in the adjunct setting. For approved A2A antagonists used with L-DOPA, worsening of dyskinesia is the characteristic adverse effect. The primate work specifically reports a low risk for KW-6356 Ohno 2023, which is a promising animal finding and not a human safety result.
The metabolite is a safety consideration in its own right. An active metabolite was modeled alongside the parent Tayama 2024, so total pharmacological exposure depends on metabolic capacity. Anything that inhibits or induces the enzymes forming it changes the parent-to-metabolite ratio rather than simply scaling the dose, and in an investigational compound those interactions are not publicly mapped.
The neuropsychiatric question is open. Adenosine A2A receptors modulate striatal circuits involved in motivation and reward, and the caffeine analogy runs to sleep disruption and anxiety as well as to alertness. Impulse-control problems are a documented hazard of dopaminergic therapy in this disease; whether a non-dopaminergic route to the same circuit carries the same risk is unmeasured.
What this page will not do. Recommend a dose. The trial doses are 3 mg and 6 mg once daily in people with Parkinson's disease under randomization Maeda 2023; the compound is unapproved; and the honest summary is a well-designed molecule with one positive twelve-week trial, an unexplained flat dose-response, and no long-term human data of any kind.
Sources read for this page
- Maeda T, et al. Randomized controlled trial of KW-6356 monotherapy in patients with early untreated Parkinson's disease. Parkinsonism and Related Disorders 2023 · PMID 37948832
- Tayama T, et al. Population Pharmacokinetics of the Novel Adenosine A2A Antagonist/Inverse Agonist KW-6356 and Its Active Metabolite Following Single and Multiple Oral Administration in Healthy Individuals and Patients with Parkinson's Disease. Clinical Pharmacology in Drug Development 2024 · PMID 38178727
- Ohno Y, et al. In vitro pharmacological profile of KW-6356, a novel adenosine A2A receptor antagonist/inverse agonist. Molecular Pharmacology 2023 · PMID 36894319
- Ohno Y, et al. Anti-parkinsonian activity of the adenosine A2A receptor antagonist/inverse agonist KW-6356 as monotherapy in MPTP-treated common marmosets. European Journal of Pharmacology 2023 · PMID 37146707
- Ohno Y, et al. The adenosine A2A receptor antagonist/inverse agonist, KW-6356 enhances the anti-parkinsonian activity of L-DOPA with a low risk of dyskinesia in MPTP-treated common marmosets. Journal of Pharmacological Sciences 2023 · PMID 37257947
KW-6356 — safety, predicted from mechanism
Predicted from mechanism, not from a human safety trial. How that reasoning works →
What the mechanism predicts
Derived from the molecule, not a trial.
- It blocks adenosine A2A receptors, and adenosine is the molecule that accumulates while you are awake and creates sleep pressure. Blocking A2A is the same axis caffeine acts on, far more selectively and for far longer. The first predicted problem is therefore the caffeine problem in a more durable form: insomnia, and a sleep debt you have removed your own ability to notice.
- A2A receptors sit on the same striatal neurons as dopamine D2 and act as a brake on them. Lifting that brake without stimulating D2 directly is the design intent — but the class's real-world experience says the brake still matters. Dyskinesia is the most common adverse effect of istradefylline, the approved A2A antagonist, at roughly 16–18% against 10% on placebo, in patients already taking levodopa.
- The rest of istradefylline's profile is the honest read-across for a compound in the same class: nausea, dizziness, constipation, insomnia, and hallucinations less commonly.
- Being an inverse agonist as well as an antagonist is a step beyond blockade — it suppresses A2A signaling that occurs with no adenosine bound at all. Nobody has characterized what removing that tonic signal does over years, because nobody has run it for years.
What has actually been reported
- Phase 2 in Parkinson's disease with a positive readout, and then a discontinued program. So real human safety data does exist for this molecule — at the doses and durations of a Parkinson's adjunct trial, in Parkinson's patients taking levodopa. Not in healthy people using it for alertness, which is who is reading this.
- For the approved comparator, ALT elevations occurred in 4–11% against 5–6% on placebo — a small signal, but the baseline test is cheap enough that there is no reason not to have it.
How to reduce the risk
Same mechanism as the prediction.
- Treat it as a long-acting stimulant for timing purposes — morning only, and count backwards from your bedtime rather than forwards from waking.
- Remember that it removes the feeling of sleep debt, not the debt. Adenosine keeps accumulating behind a blocked receptor. This is the mechanism by which an alertness compound quietly costs more than it delivers.
- If you smoke or vape nicotine, CYP1A induction is real and your exposure is not the same as a non-smoker's on the same amount.
- Track sleep objectively for the first fortnight. Self-report is the least reliable instrument available in precisely this situation, because the compound acts on the sense being used to report.
What it does to your bloodwork
A fact about the assay.
- A liver panel at baseline, for the modest transaminase signal seen with the approved compound in this class.
Don't run this if
- You take a strong CYP3A4 inhibitor or inducer. The class is cleared through CYP3A4 and CYP1A1 and is sensitive to both directions.
- You have a psychotic disorder or any history of hallucinations — hallucination is a recognized effect of the approved compound in this class.
- You already have a sleep problem, which describes most people shopping for an alertness compound in the first place.
The honest unknown
- The program was discontinued for commercial reasons after a positive readout, so no phase 3 dose for this molecule will ever exist. That is a permanent gap rather than a temporary one, and it is the single most important thing to understand about the evidence base here — the missing data is not on its way.
Not medical advice. If you take prescription medication or have a diagnosed condition, check this with a pharmacist or doctor.
KW-6356 — interference & stacking
Predicted from mechanism, not from an interaction study. How mechanism-predicted claims are made →
What KW-6356 moves on your bloodwork
Expected direction, not a measured one.
- Comprehensive Metabolic Panel (CMP) — ◆ worth watching
Most of this class has no predicted marker movement at all, and saying so is more useful than listing markers that will not move.
What to do: A baseline liver panel is reasonable for anything taken daily and long-term. Beyond that there is nothing specific to chase.
- Dose range and how to work up to it
- When to take it, and why that window
- Cycle length
- Time off between cycles
- Fasted or fed, and when in the day
- Coach Cam's personal notes
- Which compounds push the same lever, and why the dose adds up faster than people count
- What blunts it — the stacks that waste your money
- What compounds the risk, so a side effect arrives sooner than any one of them suggests
- Coach Cam's read on running it alongside the rest of your protocol
Everything above is free and stays free. Skool is where it becomes a plan — KW-6356 in an order, with the rest of what you're running.
Unlock in Skool — $10/mo →Bloodwork to run alongside KW-6356
Baseline first, then again at 8–12 weeks.
| Marker | What it’s watching for |
|---|---|
| TSH (Thyroid-Stimulating Hormone) | Thyroid disease imitates every cognitive complaint there is |
| Vitamin B12 | Deficiency causes fog long before it causes anemia |
| Methylmalonic Acid (MMA) | Catches the deficiency a normal B12 hides |
| Ferritin | Low iron flattens cognition at levels most labs call fine |
| Vitamin D (25-Hydroxy) | Commonly low, cheap to correct, associated with mood |
The Brain Fog & Cognition panel covers these in one order — 12 markers, $233.06 with the discount applied.
Check results you already have → · All 103 markers A–Z
KW-6356 — frequently asked questions
What is KW-6356?
KW-6356 (adenosine A2A antagonist / inverse agonist) is a cognitive & mood research compound. Selective adenosine A2A receptor antagonist AND inverse agonist. A2A receptors sit on the same striatal neurons as dopamine D2 and act as a brake on them; blocking A2A lifts that brake without stimulating dopamine receptors directly, which is why it does not carry the dyskinesia profile of dopaminergic drugs. The inverse-agonist part means it also suppresses A2A signaling that happens with no adenosine bound — a step beyond istradefylline, the approved A2A antagonist it is usually compared to.
Where can I find KW-6356 dosing and protocols?
Dosing, the reconstitution calculator and Coach Cam's full KW-6356 protocol are available to members inside Skool. This public page covers what KW-6356 is, how it works and the evidence.
What is the half-life of KW-6356?
KW-6356 has an approximate half-life of Not well characterized in public data, which is part of what determines how often it's dosed.
What's the evidence behind KW-6356?
Current evidence level: Human (Phase 2, Parkinson's — Kyowa Kirin). KW-6356 is offered for research purposes only and is not an approved medicine.
What KW-6356 is used for
KW-6356 appears under 1 goal in the goal router.
Related Cognitive & Mood compounds
Where this goes next
The pages here are the frameworks. The protocols — the dosing, the order to correct things in, the week-by-week schedule and what to retest — are inside Skool.