cibinetide

Q: What is cibinetide (ARA-290)?

Cibinetide is an 11-amino-acid peptide, Pyr-Glu-Gln-Leu-Glu-Arg-Ala-Leu-Asn-Ser-Ser, modelled on the aqueous face of helix B of erythropoietin. It was engineered to bind the tissue-protective EPO receptor and beta-common receptor complex without driving the red-cell response of the parent hormone. It is also called ARA 290 and pyroglutamate helix B surface peptide, and it is investigational everywhere.

Q: What does cibinetide do?

In animals it reduces allodynia and protects tissue across kidney, retina, brain, gut, islet and cardiac models without raising haematocrit, and receptor dependency has been tested twice by different methods. In humans the record is thin and conflicting. Across six published trials and roughly 130 people ever dosed, two between-group primary endpoints separated: corneal nerve fibre area at 4 mg in the 64-person phase 2b, p=0.012, with the 1 mg and 8 mg arms failing, and HbA1c in the 48-patient type 2 diabetes trial, p=0.002. The developers' own 2018 reanalysis of the corneal images reported no change in nerve fibre density. No group outside the originating network has reproduced any positive result.

Q: How is cibinetide administered in research?

In the published trials it was given subcutaneously at 1 to 8 mg daily for 28 days, or intravenously at 2 mg three times weekly for 4 weeks. The single longest exposure on record is 4 mg daily subcutaneously for 12 weeks in an uncontrolled 9-patient study. There is no approved dosing protocol, no drug label, and no primary human pharmacokinetic paper with Cmax, AUC or a terminal half-life table has been located.

Q: What are the side effects of cibinetide?

The record is too small to characterise. The phase 2b posted its adverse events by arm, and all four arms belong in the sentence: participants with at least one treatment-emergent event ran 14 of 16 at 1 mg, 11 of 16 at 4 mg, 12 of 14 at 8 mg and 12 of 16 on placebo. Serious treatment-emergent events were two in the 1 mg arm, none at 4 mg, one at 8 mg recorded as suicidal ideation, and none on placebo. The most common non-serious events were gastrointestinal, headache and fatigue. Roughly 130 people have ever been dosed and none for longer than 12 weeks, so no dose-related harm pattern can be shown either way.

Q: Is cibinetide FDA approved?

No. Queries against the FDA approvals and label endpoints return no records for cibinetide, and the EU centralised medicines register returns no authorised product. It holds two EU orphan designations, EU/3/13/1191 for sarcoidosis granted October 2013 and EU/3/16/1721 for prevention of graft loss in pancreatic islet transplantation granted August 2016. An orphan designation is a development incentive, not a marketing authorisation. Cibinetide also sits in Category 3 of the FDA's 503A compounding nomination list, the bucket for substances nominated without adequate supporting information.

Q: How much does cibinetide cost?

There is no clinical retail price because there is no approved product. In US drug-code listings cibinetide appears only as a bulk ingredient offered by API suppliers, three records, all in the bulk ingredient marketing category, with no finished dosage form and no application number. What circulates on the research market is priced by those suppliers and their resellers, and identity, concentration and contamination status are unverified unless a lot-specific certificate of analysis says otherwise.

an erythropoietin fragment engineered to protect tissue without raising red cells. six published human trials, five of them randomised, and one primary endpoint that separated.

tier D · healing · 1 of 3 doses beat placebo

verdict

does it work: mixed. two between-group primary endpoints separated in the whole human record, corneal nerve fibre area at 4 mg (p=0.012) and HbA1c in type 2 diabetes (p=0.002), neither reproduced by anyone outside the originating network. roughly 130 people have ever been dosed.

if the question is simply whether it works — mixed, and the specific failure is replication rather than effect. two between-group primary endpoints separated in the whole human record. corneal nerve fibre area rose 697 µm² over placebo at 4 mg in the 64-person phase 2b (95% CI 159 to 1236, p=0.012), with 1 mg and 8 mg failing. HbA1c separated in the 48-patient type 2 diabetes trial, -0.21% at day 56 against +0.21% on placebo, p=0.002, with PainDetect at p=0.037. no group outside the originating network has reproduced either, arms run 13 to 24 people, and nothing has started since april 2016. what was measured and failed: pain intensity everywhere, skin nerve fibre density, every objective retinal endpoint. what was never measured: anything past 28 randomised days.

if the question is whether cibinetide regrows small nerve fibres — one imaging metric, at one dose, in one trial. the 64-person phase 2b in sarcoidosis reported corneal nerve fibre area rising 697 µm² over placebo at 4 mg (95% CI 159 to 1236, p=0.012), while 1 mg (+109 µm²) and 8 mg (+431 µm²) did not separate. regenerating GAP-43 positive skin fibres also rose in that arm (p=0.035). then the same group reanalysed those same images in 2018 and reported that corneal nerve fibre density did not change and that fibre length did not separate by ANCOVA. no group outside the originators has reproduced it.

if the question is pain — pain intensity has never separated from placebo anywhere in this record. the 2012 pilot found the Brief Pain Inventory and the Fatigue Assessment Scale improving significantly and equally in both arms. the 2013 trial found pain intensity improving in both arms against baseline (p=0.01) with no between-group difference, though the pain interference dimension did separate, 36% against 16%, p<0.02 by the third week of dosing. in the phase 2b, pain fell in every arm including placebo, and the moderate-to-severe subgroup on 4 mg reached only p=0.157.

if the question is whether the programme is still running — it stopped. the last trial to start did so in april 2016, ran 9 patients with diabetic macular oedema for 12 weeks, missed every objective endpoint, and terminated with the registry reason recorded as expiry of study drug with no replacement available. no phase 3 was ever registered under any spelling of the name. cibinetide holds two EU orphan designations and no marketing authorisation anywhere, and the FDA approval and label endpoints return nothing. vials sold under the ARA-290 name are selling a shelved phase 2 asset.

based on published trials, posted registry results and primary regulatory records. cibinetide is investigational and approved in no jurisdiction. not medical advice.

why D-tier

D, at the top of the band. the human data first: six published trials, five of them randomised and placebo controlled, 218 enrolled and roughly 130 ever dosed. that is more than most compounds on this board will ever get, and it is why cibinetide does not sit at C with the mostly-preclinical entries. it sits at D because of what those trials found, and the rationale is about replication rather than about whether anything happened. two between-group primary endpoints separated in the whole record: corneal nerve fibre area at 4 mg in the 64-person phase 2b (p=0.012), with 1 mg and 8 mg failing and 14 to 16 randomised per arm, and HbA1c in the 48-patient type 2 diabetes trial (p=0.002, with PainDetect at p=0.037). neither repeated. the developers' own 2018 reanalysis of the phase 2b images reported no change in nerve fibre density and no ANCOVA separation on fibre length, so the most-repeated claim about this compound is walked back by the authors who made it, and the registry-posted secondaries put placebo numerically ahead on pain interference and on skin nerve fibre density. pain intensity has never separated from placebo anywhere, though the 2013 trial's pain interference dimension did (p<0.02 by week 3). the longest exposure, 12 weeks in 9 macular oedema patients, missed every objective endpoint and moved only a questionnaire. two independent reviews declined to endorse, including the 2021 ERS guideline co-authored by the phase 2b principal investigator. no group outside the originating network has replicated anything, and two supporting preclinical papers have been retracted. nothing has started since april 2016 and no phase 3 was ever registered. F is wrong here: F means harm or nothing, and the 4 mg imaging result plus the GAP-43 rise are real numbers with confidence intervals, the receptor route has been probed twice by different methods, and the posted adverse-event tables show no dose-related harm pattern. D is the honest reading of human evidence that is thin and conflicting rather than absent.

the core tension

the molecule does exactly what it was engineered to do and then does not do the thing people want it for. the receptor route has been probed by two different methods, and animals repeatedly show tissue protection without the red-cell effect of the parent hormone, which was the entire design goal. in 42 rats, five injections bought 20 weeks of reduced allodynia. then in humans, across five randomised trials, cibinetide moved a picture of the corneal nerve at one dose out of three and never moved pain intensity.

what it is

Cibinetide is an 11-residue peptide, Pyr-Glu-Gln-Leu-Glu-Arg-Ala-Leu-Asn-Ser-Ser, mapped onto the aqueous face of helix B of erythropoietin. It also appears in the literature as ARA 290 in the clinical work and as pyroglutamate helix B surface peptide, or pHBSP, in much of the animal work. The WHO proposed the name cibinetide in 2015 and recommended it in 2017. The design goal was separation. Erythropoietin does two things that can be pulled apart: it raises red cell mass through the classical receptor homodimer, and it protects injured tissue through a heterocomplex of the EPO receptor with the beta-common receptor. Cibinetide was built to bind only the second route. It has no FDA approval and no EU marketing authorisation.

what it does

In animal models the compound reduces allodynia, dampens innate immune activation, and protects tissue in kidney, retina, brain, gut, islet and cardiac models, without the haematocrit rise that limits erythropoietin itself. In humans the measured effects are narrow. One corneal imaging metric at one dose, a rise in regenerating GAP-43 positive skin fibres in that same arm, symptom-questionnaire movement in two sarcoidosis trials, and a single HbA1c reading in type 2 diabetes. The objective endpoints in the longest human study went the wrong way.

origin

Michael Brines and Anthony Cerami, who did much of the original erythropoietin tissue-protection work, described the molecule in a 2008 PNAS paper and founded Araim Pharmaceuticals to develop it. The clinical programme ran through Araim with Albert Dahan's group at Leiden University Medical Center and Daniel Culver at Cleveland Clinic. That concentration matters when reading the results. Brines, Cerami or the Leiden group appear on every positive trial, and five of the fourteen authors on the phase 2b carried an Araim affiliation. The last primary company document is a press release dated 8 May 2017.

why researchers are interested

The mechanism story is unusually clean for this market. There is a named receptor complex, a stated design goal, a founding paper that tested the separation directly, and animal work that repeatedly delivers protection without the red-cell effect of the parent hormone. The human trial count is also genuinely high by board standards. Five randomised placebo-controlled trials is more than most compounds here will ever see, and one of them met its primary endpoint.

does it work

Mixed, and the failure is replication rather than effect. Across roughly 130 people ever dosed in six published trials, two between-group primary endpoints separated: corneal nerve fibre area at 4 mg in the phase 2b, p=0.012, and HbA1c in the type 2 diabetes trial, p=0.002 by repeated-measures ANOVA, with PainDetect alongside it at p=0.037. Neither has been reproduced. The corneal dose response was non-monotonic, with 14 to 16 randomised per arm, and the developers' own 2018 reanalysis of those images reported no change in nerve fibre density and no ANCOVA separation on fibre length. Pain intensity, skin nerve fibre density and every objective retinal endpoint were measured and failed. Durable benefit past 28 randomised days, and the HbA1c reading in any second trial, were never measured at all. Those are different states and this page keeps them apart. Two independent reviews looked and declined to endorse, including the 2021 ERS sarcoidosis guideline, which found insufficient evidence to recommend anything for small fibre neuropathy and which the phase 2b principal investigator co-authored. Two supporting preclinical papers have been retracted. Nothing has started since 2016.

claims vs the data

cibinetide regenerates small nerve fibres — partially true — Corneal nerve fibre area rose 697 µm² over placebo at 4 mg in the phase 2b (95% CI 159 to 1236, p=0.012), and regenerating GAP-43 positive intraepidermal fibres rose in the same arm (p=0.035). The 1 mg and 8 mg arms did not separate, and the developers' 2018 reanalysis of the same images found no change in fibre density and no ANCOVA separation on fibre length. One metric, one dose, one trial, never independently reproduced.
it treats neuropathic pain — weak — Pain intensity has not separated from placebo in any trial that measured it. The Brief Pain Inventory and Fatigue Assessment Scale improved equally in both arms of the 2012 pilot. Pain intensity improved in both arms of the 2013 trial against baseline (p=0.01) with no between-group difference. The phase 2b moderate-to-severe subgroup on 4 mg reached p=0.157. The one positive reading is the pain interference dimension in 2013, 36% against 16%, p<0.02 by week 3.
it is a proven treatment for sarcoidosis — contradicted — The 2021 European Respiratory Society sarcoidosis guideline found insufficient evidence to make any recommendation on small fibre neuropathy. Its author list includes Daniel Culver, principal investigator of the cibinetide phase 2b. A 2017 systematic review of fatigue management in sarcoidosis also declined to place ARA 290 among the interventions with supporting evidence.
it protects tissue without the blood effects of erythropoietin — preclinical — That separation is the stated design goal of the 2008 founding paper and is the consistent finding across animal work, including a diabetic retinopathy study that measured haematocrit directly and found it unchanged. It is a demonstrated property of the molecule in animals. It is not a demonstrated clinical benefit in people.
it improves blood sugar control — weak — This is one of only two between-group primary endpoints that separated anywhere in the record, so it should not be filed as noise. One randomised placebo-controlled trial in type 2 diabetes, 49 enrolled and 48 analysed over 28 days of dosing with 28 days of follow-up, listed HbA1c change as a stated primary endpoint and reported a between-group difference by repeated-measures ANOVA (p=0.002): -0.16% at day 28 and -0.21% at day 56 against -0.01% and +0.21% on placebo. PainDetect separated in the same trial at p=0.037. Never replicated, never extended, arms of 24, and the supporting mouse insulin-resistance paper was retracted in April 2024.
it helps diabetic macular oedema — contradicted — The only trial, 9 recruited and 8 completed over 12 weeks with no control arm, missed every objective endpoint. Visual acuity fell 2.9 letters, central retinal thickness rose 10 µm, retinal sensitivity fell 0.53 dB and tear production fell 0.13 mm. Only a patient-reported questionnaire moved.
the programme is still in development — contradicted — The last trial started in April 2016 and terminated with the registry reason recorded as expiry of study drug with no replacement available. No phase 3 was ever registered under any spelling. Several earlier registrations, including a 12-patient rheumatoid arthritis study and a 12-patient pharmacokinetic study, were withdrawn, suspended or never reported.

key facts

molecular formula: C₅₁H₈₄N₁₆O₂₁
molecular weight: 1257.3 g/mol
amino acids: 11
half-life: ~2 min in plasma, from a developer-authored review; no primary human PK paper located
type: erythropoietin helix B surface peptide (EPOR / beta-common receptor agonist)
CAS: 1208243-50-8
6 published human trials, five randomised
~130 people ever dosed with cibinetide
1 of 3 doses that separated in the phase 2b
0 phase 3 trials, approvals, or independent replications

frequently asked questions

What is cibinetide (ARA-290)?

Cibinetide is an 11-amino-acid peptide, Pyr-Glu-Gln-Leu-Glu-Arg-Ala-Leu-Asn-Ser-Ser, modelled on the aqueous face of helix B of erythropoietin. It was engineered to bind the tissue-protective EPO receptor and beta-common receptor complex without driving the red-cell response of the parent hormone. It is also called ARA 290 and pyroglutamate helix B surface peptide, and it is investigational everywhere.

What does cibinetide do?

In animals it reduces allodynia and protects tissue across kidney, retina, brain, gut, islet and cardiac models without raising haematocrit, and receptor dependency has been tested twice by different methods. In humans the record is thin and conflicting. Across six published trials and roughly 130 people ever dosed, two between-group primary endpoints separated: corneal nerve fibre area at 4 mg in the 64-person phase 2b, p=0.012, with the 1 mg and 8 mg arms failing, and HbA1c in the 48-patient type 2 diabetes trial, p=0.002. The developers' own 2018 reanalysis of the corneal images reported no change in nerve fibre density. No group outside the originating network has reproduced any positive result.

How is cibinetide administered in research?

In the published trials it was given subcutaneously at 1 to 8 mg daily for 28 days, or intravenously at 2 mg three times weekly for 4 weeks. The single longest exposure on record is 4 mg daily subcutaneously for 12 weeks in an uncontrolled 9-patient study. There is no approved dosing protocol, no drug label, and no primary human pharmacokinetic paper with Cmax, AUC or a terminal half-life table has been located.

What are the side effects of cibinetide?

The record is too small to characterise. The phase 2b posted its adverse events by arm, and all four arms belong in the sentence: participants with at least one treatment-emergent event ran 14 of 16 at 1 mg, 11 of 16 at 4 mg, 12 of 14 at 8 mg and 12 of 16 on placebo. Serious treatment-emergent events were two in the 1 mg arm, none at 4 mg, one at 8 mg recorded as suicidal ideation, and none on placebo. The most common non-serious events were gastrointestinal, headache and fatigue. Roughly 130 people have ever been dosed and none for longer than 12 weeks, so no dose-related harm pattern can be shown either way.

Is cibinetide FDA approved?

No. Queries against the FDA approvals and label endpoints return no records for cibinetide, and the EU centralised medicines register returns no authorised product. It holds two EU orphan designations, EU/3/13/1191 for sarcoidosis granted October 2013 and EU/3/16/1721 for prevention of graft loss in pancreatic islet transplantation granted August 2016. An orphan designation is a development incentive, not a marketing authorisation. Cibinetide also sits in Category 3 of the FDA's 503A compounding nomination list, the bucket for substances nominated without adequate supporting information.

How much does cibinetide cost?

There is no clinical retail price because there is no approved product. In US drug-code listings cibinetide appears only as a bulk ingredient offered by API suppliers, three records, all in the bulk ingredient marketing category, with no finished dosage form and no application number. What circulates on the research market is priced by those suppliers and their resellers, and identity, concentration and contamination status are unverified unless a lot-specific certificate of analysis says otherwise.

related peptides

ss-31 — the other tissue-protection peptide graded on mechanism versus outcome
bpc-157 — the healing peptide with the opposite problem: huge animal base, almost no trials
kpv — anti-inflammatory peptide studied in overlapping innate-immune models

reptides grades the research record and cites the literature behind every call. research reference only; not medical advice.