◈ Nootpedia
Compact view
Research pass: thorough Pharmaceutical · Oral WATCH-LIST MEDIUM

KW-6356

Extended Research
◈ Extended Research ◈

Our depth — beyond the mirror

Deeper analysis, verdict reasoning, and per-archetype recommendations from our research team.

▸ Our verdict WATCH-LIST MEDIUM

Mechanism is genuinely novel (true inverse agonist + insurmountable antagonist with 100x istradefylline affinity), Phase 2b adjunct-to-levodopa primary endpoint hit cleanly with favorable safety, AND the original task-brief assumption "not commercially available" is wrong — Umbrella Labs and others sell it as research chem with COA. But Kyowa Kirin discontinued the program in July 2022 (regulatory/timeline calculus, not efficacy), no Phase 3 data exists or will exist, healthy-adult cognitive-enhancement evidence is zero (extrapolated entirely from istradefylline data which itself is mostly null on cold cognition), and Dylan has zero use case beyond curiosity. Verdict shifts to OPTIONAL-ADD only if (a) replicated nootropic anecdote base develops with consistent positive cognitive signal, AND (b) Dylan's modafinil + bromantane V5 stack underdelivers on motivational-drive vector. Verdict shifts to SKIP-PERMANENT if research-chem batches show purity/safety problems or if any A2A-class neuropsychiatric signal (impulse-control, hallucinations) emerges in non-PD users.

Research pass: thorough
▸ Decision matrix by user profile Per-archetype
  • Dylan20-30, brain-priority, high cognitive workload (Dylan-archetype)
    WATCH-LIST

    Mechanism is genuinely interesting but evidence in healthy young adults is zero. Dylan's V5 plan already covers wakefulness (modafinil) and motivation (bromantane) with much better evidence. Adding an unstudied-in-healthy compound with theoretical impulse-control concerns and a long half-life that risks sleep disruption (especially given late chronotype) is poor risk/reward right now. Verdict shifts to OPTIONAL-ADD only if: anecdotal nootropic database develops with consistent positive signal; AND modafinil + bromantane underdeliver on motivation specifically; AND Dylan's 23andMe shows no concerning ADORA2A variant. Sleep risk is the operative blocker for Dylan specifically.

  • 30-50, executive maintenance
    WATCH-LIST

    Same reasoning as Dylan archetype. If clinical PD is ever in their family history, A2A class is one to track for cognitive protection in the long arc. Caffeine + modafinil + lifestyle is the better risk/reward right now.

  • 50+, mild cognitive decline
    WATCH-LIST

    → OPTIONAL-ADD pending more evidence. This is actually the most theoretically interesting use case — A2A blockade has preclinical signal in AD models (autophagy/Nrf2, LTP restoration, neuroinflammation reduction, glutamate excitotoxicity) and the inverse-agonism of KW-6356 is a more thorough A2A intervention than istradefylline. However: istradefylline single-dose in healthy adults was largely null on cold cognition, no long-term cognitive readouts exist, and the A2A class's impulse-control class signal becomes more concerning in older adults with thinner cognitive reserve. Tracking phase, not actionable phase. Donepezil + rivastigmine + lifestyle remain the established options.

  • Anxiety-prone
    SKIP-FOR-NOW

    ADORA2A variants linked to caffeine-induced anxiety mean A2A blockade is a coin flip in this archetype. Until genotype is known, default to caution. There are better-tolerated nootropics for anxiety-prone users (theanine, ashwagandha, low-dose buspirone class).

  • High athletic load, tested status
    SKIP

    Not on WADA prohibited list per current 2024-2026 lists, but its discontinued/research-chem status means no clean documentation of its absence from supplement contamination scenarios. Untested-status (Dylan) makes this irrelevant; for tested athletes, default to known-safe options.

  • Sleep-disordered
    SKIP

    Wrong mechanism + half-life. Modafinil or solriamfetol or pitolisant for narcolepsy/EDS are better-evidenced.

  • Recovery-focused (post-injury, post-illness)
    SKIP-FOR-NOW

    No injury/recovery indication explored. No cognitive recovery data.

  • Strength/anabolic-focused
    SKIP

    No anabolic effect, no athletic-performance data, theoretical impulse-control concern actively unhelpful.

  • DylanParkinson's clinical (not Dylan's archetype but most relevant)
    STRONG-CANDIDATE

    if reapproved/relicensed. The Phase 2b data was meaningfully positive, the dyskinesia profile is better than istradefylline, and the program was killed for commercial reasons not efficacy. If a partner picks up the program (no public signs as of May 2026) and brings it through Phase 3, this would be a real PD adjunct option. Not currently available as treatment.

▸ Subjective experience (deep)

Honest answer: largely unknown for non-PD users.

From PD trials (n=671 cumulative across Phase 2a+2b):

  • Motor improvement on MDS-UPDRS Part III is the dominant signal. Reduced bradykinesia, rigidity, tremor.
  • OFF-time reduction (Phase 2b) — the periods when L-DOPA effect wears off and motor symptoms return. Patients spend more of their day in functional state.
  • No major dyskinesia signal — meaningfully better than istradefylline on this dimension.
  • "Non-motor symptoms" addressed in Phase 2b abstracts but specifics on cognitive/mood domains not detailed in available reporting. No clean cognitive enhancement readout.

From healthy-volunteer Phase 1:

  • Most common reported effect: insomnia, dose-dependent (Tayama 2023). Consistent with wake-promoting mechanism even absent dopamine pathology.
  • No characterized "subjective cognitive lift" signal in formal Phase 1 reports.

What we'd predict if KW-6356 produces a usable healthy-adult subjective effect:

  • Mild wake-promotion (insomnia signal in Phase 1 is the proxy).
  • Possible motivational tilt — A2A blockade in nucleus accumbens has been associated in preclinical work with increased willingness to expend effort for reward (Pardo et al., Soares-Cunha et al.). Theoretically translates to "drive" or "task-engagement lift" — but this is theoretical extrapolation, not demonstrated in humans.
  • Possible cognitive flexibility lift — A2A antagonists improve set-shifting in rodent models; istradefylline's social-information signal in healthy adults is a faint analogue.
  • Probably not a stimulant-feeling experience like caffeine or modafinil — no monoaminergic cascade.

What we'd predict could go wrong:

  • Insomnia if dosed too late.
  • Theoretical impulse-control concerns — A2A blockade indirectly potentiates D2 signaling, and D2-leaning drugs (pramipexole, ropinirole) have notable ICD signals in PD populations. Whether this translates to selective A2A inverse agonists in healthy adults is unknown but plausible. Worth flagging.
  • Theoretical anxiety in genetically susceptible individuals — see Pharmacogenomics on ADORA2A.
▸ Tolerance + cycling deep dive
  • Tolerance: unknown. No long-term human use data outside the 24-week Phase 2b PD trial, in which efficacy was maintained. A2A receptor upregulation in response to chronic blockade is theoretically plausible (analogous to A1 upregulation with chronic caffeine) but not characterized for KW-6356 specifically.
  • Recommended cycle: No empirical basis. If hypothetically pursued, 5-on/2-off weekly + every 8-12 weeks taking a 1-2 week reset would be a sensible default extrapolated from caffeine tolerance literature.
  • Reset protocol: Stop drug → 4-6 weeks washout (5+ half-lives, conservative) → reassess.
▸ Stacking deep dive

Synergistic with

  • Modafinil — Different mechanism (DAT/orexin/histamine cascade vs A2A inverse agonism). In theory complementary wakefulness without overlap. Caveat: no published data on this combination in any context. Both are wake-promoting; sleep risk is additive.
  • L-tyrosine / ALCAR — Provides dopamine precursor + acetyl-coA carrier; A2A blockade enhances D2 signaling, so substrate availability matters. Theoretical synergy, no empirical data.
  • Bromantane — Gentle dopamine release driver in V5 plan. A2A blockade + mild dopamine push could be additive on motivation vector. Same caveat: no empirical data.

Avoid stacking with

  • Other dopamine modulators (high-dose): Pramipexole, ropinirole, amphetamines, high-dose selegiline. Theoretical impulse-control disinhibition risk is class-additive.
  • Other A2A antagonists: Caffeine (especially high-dose), istradefylline, theophylline. Receptor saturation already occurs at clinical KW-6356 doses; adding more A2A blockade is wasted (or could push past target into off-target territory).
  • MAO-B inhibitors at high dose (selegiline >5 mg, rasagiline) — additive dopamine retention plus A2A disinhibition could push toward dyskinesia or impulse-control concerns.

Neutral / safe co-administration (theoretically)

  • All Dylan's V4 supplements appear neutral on paper (omega-3, magnesium, citicoline, NAC, PS, curcumin, rhodiola, theanine, glycine, D3+K2, beta-alanine, vitamin C). None hit A2A or D2 directly.
  • Operative caveat: No human data on KW-6356 + supplement stacks. Theoretical neutrality is not an empirical safety claim.
▸ Drug interactions deep dive

CYP enzymes (from Phase 1 DDI study NCT03970798, results not fully public):

  • Tested as potential perpetrator on CYP3A4 (midazolam probe), CYP1A2 (caffeine probe), BCRP/OATP1B1 (rosuvastatin probe). Specific magnitude of effects not clearly reported in publicly accessible summaries.
  • Conservative assumption: KW-6356 likely has some CYP3A4 effect (given its molecular features and the fact Kyowa specifically tested midazolam) but probably modest. Treat as if it could mildly affect CYP3A4 substrate clearance until data is published.

CYP1A2 + caffeine:

  • The fact that Kyowa tested KW-6356 + caffeine specifically is meaningful — they clearly anticipated potential clinical relevance. Without published results, assume potential modest interaction in either direction.

Hormonal contraceptives: If KW-6356 modestly induces or inhibits CYP3A4, hormonal contraceptive efficacy could be affected. Not Dylan-relevant but worth flagging.

Adenosine itself: KW-6356 is an A2A blocker. Adenosine-based diagnostic stress tests or rare adenosine therapeutic uses (Adenocard for SVT) would have unpredictable interactions.

Other dopamine-modulating drugs: Theoretical class additivity as noted in Stacking.

▸ Pharmacogenomics

Primary gene of interest: ADORA2A (the A2A receptor itself).

  • rs5751876 (1976C>T, synonymous coding variant): Strongly associated with caffeine sensitivity, anxiety response to caffeine, and sleep impact of caffeine. T/T homozygotes report greater anxiogenic response to caffeine and greater sleep disruption. Mechanism likely linkage with the 2592T/− polymorphism that may alter receptor expression.
  • Implication for KW-6356: Individuals genetically more sensitive to A2A blockade (T/T at rs5751876) would theoretically be more susceptible to BOTH the desired effects (motivational, wake-promoting) AND the adverse effects (insomnia, possibly anxiety) at lower doses. Dose escalation should be slower in this genotype.
  • Other ADORA2A variants (rs2298383, rs5996696) have weaker associations with anxiety phenotypes in caffeine challenge studies.
  • Dylan's genotype: unknown until 23andMe results June 2026. This is a meaningful pre-screening point if KW-6356 ever became a serious candidate.

Secondary gene: ADORA1 (A1 receptor):

  • Less directly relevant since KW-6356 is A2A-selective. However, A1-A2A co-localization patterns and downstream signaling overlap means ADORA1 polymorphisms could modulate net response.

CYP enzymes (from limited data):

  • KW-6356 specific metabolic pathway not fully characterized in public literature. The Phase 1 DDI study suggests CYP3A4 is a relevant enzyme. Standard CYP3A4 polymorphism considerations would apply.

Caffeine metabolism (CYP1A2):

  • If KW-6356 + caffeine has a modest interaction (per Kyowa's DDI study premise), CYP1A2 fast vs slow metabolizers would experience different KW-6356 + caffeine combination dynamics. Slow CYP1A2 metabolizers (~20% of population) get prolonged caffeine exposure and could have additive A2A blockade duration if KW-6356 has any CYP1A2 inhibitory effect.
▸ Sourcing deep dive
Path Vendor Cost Reliability Notes
Kyowa Kirin / pharmaceutical Rx N/A N/A N/A Program discontinued July 2022. No commercial product exists. No partner picked up the program publicly as of May 2026.
Research-chem powder/liquid (Umbrella Labs) Umbrella Labs ("Metabolite 6") $49.99 / 30 mL liquid (6 mg/mL = 180 mg total) or $49.99 / 60 caps × 6 mg = 360 mg Medium — COA dated Feb 2026 available on product page; standard research-chem disclaimers ("not for human consumption"). Umbrella Labs has mixed Trustpilot reviews. At 3 mg/day, 180 mg liquid = ~60 day supply, 360 mg capsules = ~120 day supply. Effective monthly cost ~$25-50.
Research-chem solution (Everychem) Everychem KW-6356 solution 6 mg/mL Low — Dylan's profile explicitly notes "Avoid: Everychem (vendor reputation issues)." Skip.
Reference standard (analytical) MedChemExpress, Probechem, Abmole, MedKoo, ChemScene Varies (typically $100-500+ for mg-scale analytical reference quantities) High purity, but expensive and not designed for chronic dosing Useful for COA cross-reference if buying from research-chem vendor. Not practical as supply source.
Compounding pharmacy None N/A N/A No legitimate Rx pathway exists. No compounding pharmacy will compound an investigational discontinued compound without an IND.

Bottom line on sourcing: Available as research chem from Umbrella Labs at low cost (~$0.25-0.40 per 3 mg dose). COA verification is the bottleneck for Dylan's risk-tolerance criteria. The original task brief assumption "not commercially available" is incorrect — it IS available, just as research-chem with no human consumption disclaimer. This is an accuracy flag.

▸ Biomarkers to track (deep)

Not actively recommended for Dylan to pursue. If hypothetically pursued (academic):

  • Baseline: ADORA2A genotype (23andMe rs5751876), CYP3A4 phenotype, sleep VAS + Pittsburgh Sleep Quality Index baseline, anxiety VAS (state + trait), motivation/drive VAS, HR/BP, full lipid + liver baseline.
  • During use (weekly first month, monthly thereafter): Sleep VAS daily log (key biomarker — insomnia is the leading AE), anxiety log, mood log, motivation/task-engagement subjective, screen for impulse-control changes (spending log, gambling/risk-taking log, sexual behavior log, eating patterns), HR/BP weekly.
  • Post-cycle: Sleep VAS recovery curve (how quickly sleep normalizes), motivation/anhedonia rebound assessment.
▸ Controversies / open debates Live debate

1. Is "A2A inverse agonism" clinically distinct from "A2A antagonism" in practice?

  • In vitro: yes, demonstrably — KW-6356 suppresses constitutive cAMP signaling, istradefylline does not.
  • In vivo PD efficacy: KW-6356 hit monotherapy (Phase 2a) where istradefylline mostly failed to develop monotherapy indication. Dyskinesia profile is better. This is suggestive of clinical translation of the in vitro distinction.
  • In vivo healthy cognition: untested. The single istradefylline-in-healthy study was mostly null. Whether KW-6356's deeper A2A suppression produces a healthy-adult cognitive signal is genuinely unknown.

2. Why was the 3 mg dose more effective than 6 mg in BOTH Phase 2a and Phase 2b?

  • Possible explanations: target saturation + off-target effects emerging at higher dose; inverted-U dose response on indirect-pathway D2 disinhibition (similar pattern seen with other A2A antagonists, possibly with D2 agonists); compensatory adenosine or A2A density upregulation kicking in faster at higher exposure; heterodimer A2A-D2 stoichiometry effects.
  • Implication for any consumer use: "more is more" intuition is wrong. 3 mg is the right starting dose, not a fraction-of-clinical-dose default.

3. Why did Kyowa Kirin discontinue despite positive Phase 2b?

  • Per company statement (July 2022): "global regulatory landscape, development hurdles, and timelines for potential market entry."
  • Per industry observers: PD adjunct market is competitive (istradefylline already approved as Nourianz; opicapone, safinamide, amantadine ER all available), Phase 3 capital-intensive, and Kyowa likely calculated the ROI on a follow-up to its own already-approved Nourianz didn't justify the spend.
  • Implication: Discontinuation does not equal "drug doesn't work." Could conceivably revive under licensing partner. As of May 2026, no public partner pickup signals.

4. Is research-chem KW-6356 "Metabolite 6" reliably the actual compound?

  • Umbrella Labs publishes COAs dated Feb 2026 on the product page. Without independent verification, unknown.
  • Dylan's stated risk tolerance for research chems requires COA verification. KW-6356 is a published compound with known SMILES/CAS (858979-50-7) and analytical reference standards available from MedChemExpress etc. for cross-reference.
  • Open question: Has any independent third party verified Umbrella Labs' KW-6356 batches against analytical reference? No public reports. This is the single biggest practical risk.

5. Will the impulse-control class signal show up in healthy users?

  • Theoretical risk (A2A blockade indirectly enhances D2 transmission; D2-leaning drugs have ICD signal in PD).
  • Not characterized in any healthy-adult dataset.
  • Would only emerge with a real anecdotal user base over months of use — which doesn't exist yet.
  • This is a genuinely open question that should temper enthusiasm.

6. Original task-brief framing accuracy check:

  • "A-tier Phase 3 Parkinson's adjunct" — WRONG. Phase 3 was never run. Program discontinued post-Phase 2b. Phase 2b data was good but is not Phase 3 evidence.
  • "non-PD cognitive evidence thin" — accurate (it's actually closer to nonexistent).
  • "research-chem availability minimal" — partially wrong. Available from Umbrella Labs and Everychem with COA. Not as easy to find as modafinil or bromantane but not "minimal."
  • "not commercially available" — accurate for pharmaceutical channel; inaccurate for research-chem channel.
  • These accuracy issues are flagged in the report at the end.
▸ Verdict change log
  • 2026-05-05 — Initial verdict: WATCH-LIST / MEDIUM CONFIDENCE. Mechanism genuinely novel and interesting, Phase 2b PD adjunct data is real and positive, but program was discontinued July 2022, no Phase 3 data exists or will exist, healthy-adult cognitive evidence is essentially zero, anecdotal nootropic database does not yet exist, theoretical sleep + impulse-control risks are operative concerns for Dylan specifically, and his V5 plan (modafinil + bromantane) already covers wakefulness and motivation with much better evidence. Worth tracking for nootropic adoption signal over 12-24 months but not a current candidate. Verdict shifts to OPTIONAL-ADD on (a) anecdotal nootropic database developing with consistent positive cognitive signal, (b) Dylan's modafinil + bromantane V5 stack underdelivering on motivational drive, (c) Dylan's 23andMe ADORA2A status confirmed neutral. Verdict shifts to SKIP-PERMANENT on (a) any neuropsychiatric signal (impulse control, hallucinations) emerging in non-PD users, (b) research-chem batches showing purity/safety problems, (c) better A2A or A2A-adjacent compounds emerging with cleaner profile.
▸ Open questions / gaps Open
  1. Will the program revive under a licensing partner? No public signals as of May 2026. Worth a single-query check at next review (May 2027).
  2. Will an independent cognitive-enhancement study in healthy adults appear? Unlikely — there's no commercial or academic incentive once Kyowa dropped the program. The compound is already publicly synthesizable, so investigator-initiated work is theoretically possible but no public protocols.
  3. Does the inverted-U dose response (3 mg > 6 mg) hold in healthy adults? Unknown. Could be PD-specific.
  4. Does KW-6356 + caffeine have a meaningful PK or PD interaction? Phase 1 DDI study completed (NCT03970798) but specific results not in publicly accessible summaries. Practical concern for Dylan in V4 caffeine ramp phase.
  5. Are research-chem KW-6356 batches reliably the right compound at the right purity? No third-party verification of vendor COAs publicly available.
  6. Does the impulse-control class signal show up in non-PD users? Genuinely open. Won't be answered without consumer-population use base, which doesn't exist yet.
  7. Long-term (>24 weeks) tolerance + receptor adaptation? Unknown. A2A receptor density upregulation in response to chronic blockade is theoretically plausible (cf. caffeine + A1 upregulation) but not characterized for KW-6356.
  8. Does ADORA2A rs5751876 genotype meaningfully predict KW-6356 response? Hypothesized but not empirically tested.
▸ Sources (full, with our context)
Back to compact view