Ara-290

What it is

ARA-290 is the cleanest piece of peptide engineering in the endogenous-protection cluster. The story starts with a discovery that rewrote what EPO actually does.

EPO (erythropoietin) is famous as the blood-doping hormone — what Lance Armstrong used, what raises red blood cell counts, what creates hematocrit, thrombosis, and stroke risk. But Brines and Cerami at Araim Pharmaceuticals found that EPO actually operates through two entirely different receptor systems with entirely different jobs:

• The blood-building receptor — the classical EPOR homodimer (two identical EPO receptor subunits, symmetrical). This is the one everybody knew about. Drives erythropoiesis. Raises hematocrit.
• The tissue-repair receptor — the Innate Repair Receptor (IRR): a heterodimer of two different subunits, EPOR + CD131 (the β common receptor, βcR). A mismatched pair. When EPO binds this one, it tells the cell: survive, repair yourself, calm the inflammation. No red blood cell production at all.

The engineering insight was to identify exactly which surface of the EPO molecule activates the IRR — a specific patch called the “helix-B face.” Brines and Cerami carved that surface out and rebuilt it as a standalone 11-amino-acid peptide. The repair signal, isolated and concentrated. That’s ARA-290.

Why this matters: ARA-290 only fits the IRR — its shape specifically matches the gap between the two different subunits of the heterodimer. It physically cannot dock at the classical EPO homodimer. The result is a peptide that is non-erythropoietic — it does not and cannot raise red blood cell count. EPO’s entire hematological risk profile is engineered out. What remains is a clean tissue-protective signal with a published Phase 2 RCT and FDA Orphan Drug Designation behind it.

Why a separate peptide was needed at all (added 2026-09-22). The repair receptor binds EPO less tightly than the blood-building receptor does. So using EPO itself for tissue protection takes a higher dose than it takes to max out red-cell production, and at that dose EPO also makes platelets more reactive and pushes blood vessels toward clotting. Some popular explanations say the reverse: that EPO protects tissue at doses too low to affect red cells. That is backwards. The mismatch is the reason a receptor-selective molecule was worth building. The key papers:

• Brines 2004 (PNAS, PMID 15456912) identified the EPOR + CD131 heteroreceptor.
• Leist 2004 (Science, PMID 15247477) showed that carbamylated EPO keeps the tissue protection but loses the red-cell effect.
• Brines 2008 (PNAS, PMID 18676614) located the protective surface on helix B and built the 11-amino-acid peptide from it.

---

How it works — mechanism

Receptor targeting

After SubQ injection, ARA-290 circulates passively through the bloodstream. It reaches injured tissue and docks at the Innate Repair Receptor (IRR) — the EPOR + CD131 heterodimer — which is upregulated at sites of injury, inflammation, and metabolic stress (see “The targeting trick” below). At healthy tissue, the molecule largely passes by.

Intracellular cascade

Docking at the IRR fires a downstream signaling cascade: STAT3 → PI3K → Akt

• STAT3: anti-apoptotic gene expression (“keep this cell alive”)
• Akt: specifically disables the cell’s apoptosis program. A metabolically stressed cell hovering near programmed self-destruction — the Akt signal essentially pulls the trigger finger off the button. ARA-290 doesn’t heal; it stops the dying long enough for real repair to catch up.
• Amplification: one docking event activates hundreds of downstream proteins. The signal is amplified and runs long after the peptide itself has cleared.
• Upstream kinase and cytokine profile (added 2026-09-22): the receptor signals through JAK2, which feeds both PI3K/Akt and STAT3. Reported downstream effects are lower TNF-α, IL-6, IL-1β and MCP-1, higher anti-inflammatory IL-10, and suppressed NF-κB. In LPS-activated macrophages, cibinetide’s anti-inflammatory effect required both CD131 and JAK2 and ran through inhibiting NF-κB p65 (mouse colitis model).

Macrophage polarization

ARA-290 shifts macrophages from M1 mode (pro-inflammatory, demolition) to M2 mode (pro-repair, construction). This is why repeated dosing over days changes the whole tissue environment — not one big event, but a sustained shift in the immune tone of the injury site.

The targeting trick — how a “blind” molecule finds the wound

ARA-290 does not navigate. It has no homing machinery. It passively follows the bloodstream, which goes everywhere. The selectivity comes from two filters that stack at the injury site:

Filter 1 — Receptor upregulation. Inflammation, metabolic stress, and tissue damage cause cells to put more IRRs on their surface. Healthy tissue has few IRRs — the peptide floats past with almost nothing to grab. An injured cell has thrown up hundreds of IRRs. The peptide docks, the cascade fires.

Filter 2 — Downstream priming. Even if a healthy cell has some IRRs, the intracellular wiring isn’t primed for the cascade. The signal lands at low volume. At the injury site, the cell is already in metabolic stress and receptive — same molecular event, far louder response. Striking a match in an empty room vs. a room stacked with dry kindling. Same spark, completely different outcome.

Stack both: the molecule is blind, but the wound puts up the locks — and when the molecule floats through, it finds them.

---

Signaling pharmacology (why the short half-life is not a problem)

ARA-290 is cleared from the blood in under an hour. People who see that and conclude “it can’t work” are thinking in the wrong pharmacological framework.

Occupancy pharmacology (ibuprofen, most common drugs): the drug must physically sit on the target the whole time it’s working. Blood levels fall, effect falls. The effect tracks the drug concentration minute by minute.

Signaling pharmacology (ARA-290): the peptide is the trigger, not the effect. Its only job is to reach the receptor, dock, and flip the switch. Once it fires the STAT3/PI3K/Akt cascade, the biology carries the ball. Gene expression changes. Protein synthesis shifts. The machinery runs for days after the peptide is gone.

Think of it as a light switch: your finger touches it for half a second, the lights stay on for hours. ARA-290 is the finger. The cascade is the lights.

This is why the trial outcomes — meaningful pain reduction at weeks 4–8, objective nerve fiber improvement, sustained metabolic changes past the treatment window — are biologically expected, not surprising.

---

Pharmacokinetics

• Route: Subcutaneous injection — subcutaneous depot acts as a slow reservoir, avoiding the sharp spike of an IV push
• Peak blood levels: ~15–30 minutes post-injection
• Blood half-life: under 1 hour — 11 amino acids with no large protective structure; kidneys filter it rapidly
• Compare to: antibody drugs that linger for weeks; large proteins with half-lives of days. ARA-290 is here and gone — the effect, as described above, is not.
• Measured human PK (added 2026-09-22): in healthy volunteers, 4 mg SubQ gave a peak plasma level of ~3 ng/mL (~2.4 nmol/L) with a terminal half-life of ~20 minutes. Preclinical data put the activation threshold for the repair receptor at ≥~1 nmol/L (~1.3 ng/mL). An IV half-life of ~2 minutes is sometimes quoted.
• Why 4 mg SubQ became the trial dose: a crossover PK study in 10 volunteers measured how long plasma stayed above the ~1.3 ng/mL threshold (area under the curve). The results in ng/mL×min: 2 mg IV = 65, 2 mg SubQ = 23, 4 mg SubQ = 59, 6 mg SubQ = 249. 4 mg SubQ was chosen because it matched the 2 mg IV exposure used in the first sarcoidosis trial.

---

What the research shows

Brines 2015 — the cornerstone published RCT

PMID 25387363, PMC4365069. ARA 290, a Nonerythropoietic Peptide Engineered from Erythropoietin, Improves Metabolic Control and Neuropathic Symptoms in Patients with Type 2 Diabetes. Brines M, Dunne AN, van Velzen M, et al. Mol Med 2015.

• Design: double-blind, placebo-controlled Phase 2; Netherlands Trial Register NTR3858; investigator-initiated
• Population: n=48 (24 ARA-290 / 24 placebo); Type 2 diabetics with neuropathic symptoms
• Dose: 4 mg SubQ self-injection daily × 28 days; 28-day observation period after

Endpoint	ARA-290	Placebo	Result
HbA1c at day 28	−0.16%	−0.01%	p=0.002
HbA1c at day 56	−0.21%	+0.21%	continued improvement vs. reversal
Cholesterol/HDL ratio	reduced	—	p=0.039
Triglycerides	decreased	—	p=0.043
PainDetect score at day 28	−3.3 pts	−1.1 pts	significant
PainDetect score at day 56	−4.2 pts	−0.74 pts	p=0.037
PainDetect tingling	improved	—	p=0.01
PainDetect allodynia	improved	—	p=0.04
Corneal nerve fiber density (subgroup)	+2.6 ± 1.0 fibers/mm²	+0.7 ± 1.3 (ns)	p=0.02
RAND-36 vitality	improved	—	p=0.02
RAND-36 physical role functioning	improved	—	p=0.05

Two results that stand out mechanistically:

• PainDetect improvement tracked with CNFD increase in the ARA-290 group but not placebo. Mechanistic + clinical congruence: the nerve regeneration explains the symptom improvement, not just a placebo response.
• No anti-drug antibodies at day 28. No immunogenicity signal — important safety-of-repeated-dosing data.

Conflict-of-interest disclosure (from paper): Brines, Dunne, and Cerami are officers of Araim Pharmaceuticals and hold equity. Funding included Dutch government grant NIRM FES0908, Swedish Research Council ALF, and Swedish Diabetes Association — mixed industry + government-funding profile, disclosed transparently in the paper.

Sarcoidosis SFN trials

Double-blind, 4 mg ARA-290 SubQ daily × 28 days in sarcoidosis-associated small-fiber neuropathy. SFNSL (Small Fiber Neuropathy Screening List) score decreased 9.1 ± 8.5 points at day 28 (figure as reported in Brines 2015’s power calculation). The sarcoidosis programme is the basis for ARA-290’s FDA Orphan Drug and Fast Track designations for neuropathic pain in sarcoidosis.

Resolved 2026-09-22 — there are three sarcoidosis RCTs, not one. All three are placebo-controlled and developer-sponsored; the first two ran at the same Leiden centre, so read them as one programme rather than independent replications.

Trial	Design	Key result
Heij 2012 — Mol Med, PMID 23168581	n=22; IV 2 mg three times weekly × 4 weeks	SFNSL −11.5 vs −2.9 placebo at week 4 (p<0.05); SF-36 pain and physical functioning improved; no safety concerns
Dahan 2013 — Mol Med, PMID 24136731	n=38; SubQ 4 mg daily × 28 days; 16-week follow-up	SFNSL −12.2 vs −3.8 placebo one week after dosing (p=0.005); corneal nerve fibre density rose significantly vs placebo; 6-minute walk +~19 m vs −~15 m; cold/heat pain thresholds improved. Skin-biopsy nerve density did not change significantly (distal-leg trend only)
Culver 2017 — Invest Ophthalmol Vis Sci, PMID 28475703; NCT02039687	Phase 2b; n=64; 1, 4 or 8 mg/day vs placebo × 28 days; two centres (Cleveland Clinic, Leiden)	See dose-ranging below

Correction: this section previously said the 4 mg sarcoidosis trial showed increased nerve density on skin biopsy. In Dahan 2013 the significant nerve gain was in the cornea; skin-biopsy regrowth (GAP-43+ fibres) only reached significance in Culver 2017 at 4 mg.

Culver 2017 — the dose-ranging trial

The only trial that compared doses directly. The primary endpoint was the change in corneal nerve fibre area at day 28 (placebo-corrected, µm²):

Dose	Change in corneal nerve fibre area	Significance
1 mg/day	+109	ns
4 mg/day	+697 (≈23% over baseline)	P = 0.012
8 mg/day	+431	ns

• Skin regrowth: GAP-43+ intraepidermal fibres (a marker for regrowing nerve endings) increased in the 4 mg group (P = 0.035).
• Structure tracks function: corneal gains correlated with skin gains (ρ = 0.575) and with 6-minute walk improvement (ρ = 0.645).
• Pain is the softer result. Pain improved in every group, placebo included. In patients with moderate-to-severe pain, the placebo-corrected drop at 4 mg was described as clinically meaningful but was not statistically significant (P = 0.157). The trial’s strength is the structural nerve data, not the pain score.
• Reading the dose curve honestly: 8 mg gave no advantage over 4 mg in this 28-day trial. Some practitioners describe this as an inverted-U curve with 4 mg as the ceiling. One trial with overlapping confidence intervals shows “no benefit above 4 mg”, not a proven peak.

Diabetic macular edema — Lois 2020

PMID 32674280, J Clin Med 2020; ClinicalTrials.gov NCT06626971; EudraCT 2015-001940-12.

• Design: open-label Phase 2 in Belfast; 4 mg/day SubQ self-injected for 12 weeks; 9 enrolled, 8 completed. This is the longest human exposure on record.
• Findings (descriptive statistics only): no mean improvement in visual acuity, central retinal thickness, retinal sensitivity or tear production.
• What did improve: vision-related quality of life (NEI VFQ-25 +2.7 ± 3.1). Some individuals improved on retinal thickness, tear production, diabetic control and albuminuria.
• Safety: no serious adverse events and no anti-cibinetide antibodies over 12 weeks.
• This trial studied diabetic macular edema. There are no data in macular degeneration, a different disease.

Other registered human studies

• NCT01933529 — prediabetes / type 2 diabetes, Phase 2, n=24 (status unknown on ClinicalTrials.gov).
• NCT02070783 — cognitive and neural effects in healthy volunteers, Phase 1/2, n=36, completed. Related: Cerit 2015, PMID 26431906.

Timeline of clinical effect (from trials + mechanism)

Phase	Timeframe	What’s happening
Trigger	Minutes	IRR docking → STAT3/PI3K/Akt cascade fires
Clearance	<1 hour	Peptide gone; cascade runs independently
Cytokine shift	Hours	Early anti-inflammatory markers shift
Macrophage polarization	Days (repeated dosing)	M1→M2 consolidation; background fatigue may ease
Structural nerve repair	Weeks	Small nerve fiber regrowth; real pain score separation from placebo at weeks 4–8; laser evoked potentials trend the right way

This is the shape of ARA-290’s clinical signature: minimal detectable change in week 1, meaningful separation from placebo at weeks 4–8. Small nerve fibers don’t regrow in a day. The trials’ primary endpoints were designed around this slow-repair biology.

Other studied contexts (preclinical)

• Diet-induced insulin resistance
• Diabetic retinopathy
• Diabetic autonomic neuropathy (Akita mice)
• Myocardial infarction and chronic heart failure
• Traumatic brain injury
• Burns and shock-induced multi-organ failure

Retraction note (added 2026-09-22): a 2012 preclinical paper on pHBSP (ARA-290) in acute kidney injury (Patel et al., Mol Med 2012) was retracted in February 2026 (retraction note PMID 41749083). Preclinical kidney-protection claims should not rest on it.

---

Safety — the honest read

Brines 2015 adverse events (n=48 trial):

• ARA-290 arm (n=24): 54 mild + 9 moderate + 1 severe + 4 serious adverse events (SAEs)
• Placebo arm (n=24): 61 mild + 5 moderate

The four SAEs in the ARA-290 arm (transparent reporting):

1. Renal insufficiency — patient on furosemide with borderline baseline renal function; creatinine rose from 119 → 159 μmol/L after week 2; adjudicated “possibly related”; renal function did not improve after ARA-290 was stopped while furosemide continued. Clinical note: patients with borderline renal function on loop diuretics merit monitoring.
2. Cellulitis + fatal MI — 70-year-old male developed severe lower-extremity cellulitis ~2 weeks after his last dose and suffered a fatal myocardial infarction; adjudicated “unrelated” to ARA-290. 3–4. Two additional SAEs adjudicated “unlikely associated.”

The fatal MI in a 70-year-old — even adjudicated unrelated — is reported here because OHM’s editorial credibility is its transparency. The molecule did Phase 2 trials, the trial had four SAEs in a 24-person arm, and the data is published in full with explicit investigator adjudication. That’s the honest record.

Other safety findings:

• No clinically significant hematology or chemistry changes vs. baseline (confirming non-erythropoietic profile)
• No immunogenicity signal at 28 days
• No long-term multi-year safety data exists

Across the other trials (added 2026-09-22):

• Dahan 2013 (n=38, 28 days SubQ): no serious adverse events through 12 weeks of follow-up, and no pain or irritation at injection sites. One ARA-290 patient had ongoing weight loss that began before enrolment and persisted after dosing stopped.
• Culver 2017 (n=64): no significant safety issues in any dose group.
• Lois 2020 (n=9, 12 weeks): no serious adverse events and no anti-drug antibodies.
• Scale of the evidence: the five published patient trials enrolled about 180 people combined, placebo arms included. Every patient was dosed for 12 weeks or less. The record supports calling ARA-290 well tolerated in short trials. It cannot yet support a claim that there is no downside.
• Kidney context: in the one renal event (Brines 2015), the patient’s furosemide (a diuretic) dose had just been raised, and kidney function did not recover after ARA-290 was stopped. Practitioners commonly list kidney-function changes as a stop signal.
• Theoretical concern: PI3K/Akt activation suppresses programmed cell death. No trial has shown a cancer signal, but the trials were too short and small to rule one out. Practitioners typically list active cancer or a new cancer diagnosis as a reason to avoid or stop. [MECHANISM + ANEC]
• No data: pregnancy, breastfeeding, children, or drug interactions.

---

Real-world protocol

The doses and schedules here are for educational and informational purposes only. These peptides are sold for research use only and are not FDA-approved drugs. This is not medical advice. Consult a qualified physician before beginning any protocol.

Trial-validated dose:

• 4 mg SubQ injection daily × 28 days — the dose from both Brines 2015 (T2D neuropathy) and the sarcoidosis SFN trial
• Reconstitution example: 4 mg vial + 1 mL bacteriostatic water = 4 mg/mL; a 4 mg dose = 1 mL = 100 units on a U-100 insulin syringe (a full syringe)

Community / anti-aging dosing (not formally validated):

• 0.5–1 mg SubQ daily, or 1–2 mg every other day
• Well below the trial dose; used in the biohacker community as a “longevity / nerve maintenance” pattern
• No published data validates this range — it’s extrapolation from the trial dose and general peptide community practice

Storage: refrigerated lyophilized; reconstitute with bacteriostatic water; ~30 days post-reconstitution

Cycle: 28-day blocks are the trial-validated pattern. Extended dosing past 28 days lacks published data. The slow-repair timeline (weeks 4–8 for meaningful symptom change) means a 28-day course starts showing benefits near its own end — a longer block or repeated cycles is the logical extension for chronic neuropathic conditions, but it’s not formally studied.

Practitioner protocol range (added 2026-09-22)

Different sources report different protocols. They are listed here side by side as a range, not a ranking.

Source	Dose	Duration / cycle	Anchoring rationale	Tier
Trial protocol (Dahan 2013, Brines 2015, Culver 2017)	4 mg SubQ daily	28 days	Dose selected by PK matching; only significant arm in the dose-ranging trial
Longest human exposure (Lois 2020)	4 mg SubQ daily	12 weeks	Safety shown to 12 weeks, n=9; eye endpoints not met
Practitioner three-tier ladder (Williams 2026)	Tier 1: 1–2 mg/day for mild symptoms · Tier 2: 4 mg/day × 28 days · Tier 3: 4 mg/day (never higher) for 8–12 weeks in more severe cases	Continued while symptoms are still falling (target ~0–1 out of 10), not stopped at a fixed week. Reported cycles: 4 on / 8 off, 8 on / 4 off, or 12 on / 12 off. No continuous low-dose maintenance	Dose-ranging trial for the ceiling; clinical experience for duration. Reports that most users recover 70–80% within 8–12 weeks (no trial covers that window in neuropathy)
Community longevity pattern (existing entry above)	0.5–1 mg/day, or 1–2 mg every other day	Continuous	Extrapolation

Other reported practice

• Timing: once daily. Time of day and fed or fasted state are reported not to matter.
• Injection site: the trials injected SubQ into the abdomen (Dahan 2013 also used the upper leg). Some users inject near the painful area, such as the foot, and report faster relief. This is untested; the effect is systemic either way.
• Expected timeline: usually little in week 1. Less burning and fewer shock-like sensations around weeks 2–4. Structural change was measured at day 28. Benefits commonly continue 4–8 weeks after stopping. Brines 2015 showed continued improvement to day 56; the longer tail is.
• What gets tracked: burning intensity, electric sensations, temperature sensitivity, night-time symptoms and walking tolerance. Objective options are skin punch biopsy, corneal confocal microscopy and quantitative sensory testing. Routine blood work usually shows nothing.
• Reasons for non-response: not enough time; a dose below 4 mg; a dose above 4 mg; or pain that isn’t driven by small-fibre loss or macrophage inflammation.
• Reported stop signals: signs of infection or allergic reaction, escalating symptoms, a new cancer diagnosis, or any change in kidney function. Starting during an acute illness or within two weeks of surgery is commonly avoided.

Reconstitution chemistry — why some vials gel (added 2026-09-22)

• The chemistry. ARA-290 is strongly acidic: two glutamates, one arginine, a free C-terminus and a blocked N-terminus. Its estimated isoelectric point is about 3.5–4. That is the pH where the molecule carries no net charge, stops repelling itself, and is least soluble. [CALCULATED from sequence]
• The water. Bacteriostatic water is unbuffered, with a labelled pH range of 4.5–7.0 (typical 5.7). The label says it must be made approximately isotonic before use. [DailyMed label]
  • Most vials sit well above the peptide’s least-soluble pH and dissolve fine.
  • A vial at the acidic end sits much closer to it, which is the proposed reason some ARA-290 preparations gel or stay hazy. [MECHANISM — plausible; no published gelling data for ARA-290]
• Phosphate-buffered saline (PBS) as the alternative. PBS (pH 7.4, isotonic) keeps the peptide at a net charge of about −2, which holds it in solution.
  • PBS has no preservative, so the practice reported for PBS-mixed vials is refrigeration and use within about 5–7 days.
• Acetic-acid rescue for a gelled vial. Adding a small share of dilute acetic acid, roughly 10% of the diluent volume, is a reported fallback. It stings on injection, leaves the final pH unbuffered, and may degrade the peptide.
• Syringes. ARA-290 is commonly kept out of shared syringes with other peptides because of its pH.
• Worked arithmetic. A 10 mg vial in 2 mL is 5 mg/mL, so 4 mg = 0.8 mL = 80 units on a U-100 insulin syringe.
• Handling. Diluent run down the side of the vial, no shaking, refrigerated storage.
• Routes. No oral or intranasal data exist.

Stacking landscape (added 2026-09-22)

No combination of ARA-290 with any other compound has been tested in a trial. Everything below is.

• Sequencing logic. ARA-290 is run first, or overlapped, to shift macrophages out of inflammatory mode. BPC-157 / TB-500 are layered in a few weeks later for regeneration, on the reasoning that small nerve fibres are among the slowest tissues to regrow and regrow poorly into an inflamed field.
• Other reported partners:
  • SS-31 (Elamipretide), described as a general amplifier
  • KPV for inflammation
  • a GLP-1 (Tirzepatide / Retatrutide) where diabetes drives the neuropathy
  • Thymosin Alpha-1 for the adaptive immune side, while ARA-290 works on the innate side
  • Low-Dose Naltrexone (LDN) — the immune-modulation bridge for autoimmune patients on GLP-1s, a low-cost adjunct
• Hypothesis only. Whether ARA-290 could ease the skin-burning sensation (dysaesthesia) some Retatrutide users report is untested either way.
• Commonly avoided. Combining with EPO or other erythropoiesis-stimulating agents (never studied), and starting several unfamiliar compounds at once.

Who it is unlikely to help (added 2026-09-22)

• Compression pain. A herniated disc pressing a nerve root is a structural problem, and regrowing small fibres doesn’t decompress anything. [MECHANISM + ANEC]
• Athletic performance. ARA-290 has no red-cell effect, so there is no EPO-like benefit, and it is banned in tested sport (see Regulatory). [VERIFIED mechanism]
• General wellness with no nerve damage. The repair receptor is sparse in healthy tissue, so there is little for the peptide to act on. [MECHANISM]
• Metabolic or cardiovascular optimisation on its own. The HbA1c and lipid effects in Brines 2015 were real but small next to what other metabolic tools deliver.
• Uncontrolled diabetes. The underlying driver of the neuropathy still needs to be dealt with; ARA-290 does not replace that.
• Fibromyalgia, MCAS, long COVID. No human trials. Benefit is plausible only where small-fibre damage is the actual driver.

---

Regulatory status

• FDA Orphan Drug Designation for neuropathic pain in sarcoidosis — granted; meaningful regulatory advance. The FDA reviewed the data and acknowledged real potential benefit for a serious rare disease. Opens fast-track development pathways.
• Full FDA approval: never submitted. Development program never reached Phase 3.
• Araim Pharmaceuticals: closed as of 2026. Not a clinical failure — a financing failure. A small biotech couldn’t fund Phase 3 for niche indications; peptides are expensive to manufacture. Economics ended the company; the biology didn’t fail. The Brines/Cerami dual-receptor discovery and the published Phase 2 data remain in the scientific record regardless of the company’s fate.
• Post-closure status: patents may be expired; the molecule could be picked up by another developer. Orphan Drug precedents for molecule recovery exist.
• Designations in full (added 2026-09-22): US Orphan Drug + Fast Track for neuropathic pain in sarcoidosis; US + EU orphan designations for sarcoidosis; orphan designations relating to pancreatic-islet transplantation. Araim reported a successful end-of-Phase-2 meeting with FDA after the Culver 2017 results (Araim press release, 8 May 2017). No Phase 3 followed. ``
• FDA compounding list (added 2026-09-22): on FDA’s 503A bulk-substance nominations list (updated 2026-05-14), cibinetide (ARA-290) is in Category 3, “nominated without adequate support.” It is not in Category 2 (the significant-safety-risk category). It is also not eligible for 503A pharmacy compounding.
• WADA — CORRECTED 2026-09-22: prohibited. Being non-erythropoietic does not get ARA-290 around the ban. The 2026 WADA Prohibited List names “Innate repair receptor agonists” under S2.1.5 (examples given: asialo EPO, carbamylated EPO), prohibited at all times, in and out of competition. ARA-290 is an innate repair receptor agonist. It is also caught by S0 (non-approved substances). This entry previously said ARA-290 was not banned; that was wrong. [REGULATORY — WADA 2026 Prohibited List]

---

Where experts disagree

Is more than 4 mg better? (added 2026-09-22): one view reads Culver 2017 as an inverted-U curve with 4 mg as the ceiling. The strict reading is that 8 mg showed no advantage over 4 mg in one 28-day trial. Both lead to the same practical number. They differ in how confidently it can be said that higher doses are worse.

How long to run it (added 2026-09-22): the trials ran 28 days, and the eye trial 12 weeks. Some practitioners extend to 8–12 weeks or longer, as long as symptoms keep improving. No neuropathy trial has tested beyond 28 days, so extended courses rest on clinical experience.

Dosing at the community low end: the trial-validated dose is 4 mg/day × 28 days. The biohacker community uses 0.5–1 mg/day continuous as a maintenance or longevity pattern. There’s no published data on the low end; it’s extrapolation. The trial dose is what was actually tested. OHM’s position: present both honestly — the validated trial dose with its outcomes, and the community practice with its label.

“Did it fail?” The molecule produced statistically significant results on hard endpoints (HbA1c, PainDetect, corneal nerve fiber density). The company that held it ran out of money before full approval. These are two different things and the distinction matters: “failed Phase 2” vs. “passed Phase 2, company folded” carry very different weight when assessing the molecule’s potential.

Duration of benefit after the 28-day course: the trial has a 28-day observation period post-treatment; HbA1c continued improving in that window (−0.21% at day 56 vs. +0.21% in placebo). What happens at day 90, day 180 is unknown. The cascade-based signaling model predicts durable benefit, but the human data doesn’t reach that far.

---

Where this sits in the OHM endogenous-protection map

Peptide	Origin	Primary role
ARA-290	Engineered from EPO helix-B face	Peripheral nerve repair + anti-inflammation via IRR
Humanin	Mitochondrial-derived (MT-RNR2 gene)	CNS neuroprotection + longevity + IGF-1 axis
MOTS-c	Mitochondrial-derived (MT-RNR2 gene)	Metabolic / insulin sensitivity
SS-31 (Elamipretide)	Mitochondrial-targeted synthetic	Mitochondrial ROS scavenging; FDA-approved for Barth syndrome

---

Sources

• original research synthesis; source of the Brines 2015 RCT data, safety detail, protocol
• “Fable LLM from Anthropic” (Free Knowledge channel); July 2026; source of the signaling pharmacology framework, two-filter targeting mechanism, timeline systematization, Araim closure 2026 confirmation
• Brines et al. 2015 Mol Med, PMID 25387363, PMC4365069 — the cornerstone ARA-290 Phase 2 trial; fully verified
• “The ARA-290 Masterclass” (one practitioner, 2026). Source of the practitioner protocol range, reconstitution chemistry, stacking landscape, who-it-does-not-help, and the 31-row verification table behind every 2026-09-22 addition.
• Heij 2012 Mol Med, PMID 23168581: sarcoidosis SFN, IV.
• Dahan 2013 Mol Med, PMID 24136731: sarcoidosis SFN, 4 mg SubQ × 28 days; PK dose-selection data.
• Culver 2017 Invest Ophthalmol Vis Sci, PMID 28475703; NCT02039687: dose-ranging (1/4/8 mg).
• Lois 2020 J Clin Med, PMID 32674280; NCT06626971: diabetic macular edema, 12 weeks.
• Brines 2004 PNAS, PMID 15456912; Leist 2004 Science, PMID 15247477; Brines 2008 PNAS, PMID 18676614: receptor discovery and peptide engineering.
• Brines 2010 Blood Purif, PMID 20093809; Hand & Brines 2011 J Investig Med, PMID 20683348: the tissue-protective receptor has lower EPO affinity, so protection with EPO needs a higher dose.
• Nairz 2017 Sci Rep, PMID 29026145: CD131/JAK2 dependence, NF-κB p65.
• Retraction note Mol Med 2026, PMID 41749083: pHBSP acute kidney injury paper (2012) retracted.
• WADA 2026 Prohibited List: S2.1.5 Innate repair receptor agonists; S0 non-approved substances.
• FDA 503A bulk drug substances nominations list (updated 2026-05-14): cibinetide in Category 3.
• Araim Pharmaceuticals press release, 8 May 2017: orphan / Fast Track designations; end-of-Phase-2 meeting.
• PubChem CID 91810664 (cibinetide) · DailyMed Bacteriostatic Water for Injection USP label.

queue (updated 2026-09-22):

• Resolved: mechanism papers (PMIDs 15456912, 18676614); sarcoidosis trial PMIDs (23168581, 24136731, 28475703); WADA status (prohibited); molecular formula and sequence (PubChem CID 91810664).
• Still open:
  1. Current active or lapsed status of the orphan designations after Araim’s closure, and the closure date.
  2. IV half-life (~2 min).
  3. Duration of benefit beyond ~4 weeks after the last dose.
  4. Full-text details of the Nairz 2017 JAK2 experiments.

Related: Humanin & ARA-290 — the endogenous protection cluster · MOTS-c · SS-31 (Elamipretide) · BPC-157 · Retatrutide · atherosclerosis-cardiovascular.