AICAR (Acadesine)

A purine nucleoside analog and AMP-activated protein kinase (AMPK) activator, developed clinically as acadesine for bypass-surgery ischemia and leukemia without success, known in research circles for a 2008 mouse 'exercise mimetic' study and banned by WADA.

Also referenced as: Acadesine, AICA riboside, AICA-riboside, 5-aminoimidazole-4-carboxamide ribonucleoside, 5-aminoimidazole-4-carboxamide ribonucleotide, 5-aminoimidazole-4-carboxamide 1-beta-D-ribofuranoside, Z-riboside

Also appears in: Other

What is AICAR (Acadesine)?

AICAR is not a peptide. It is a purine nucleoside analog, 5-aminoimidazole-4-carboxamide riboside: the nucleoside form of ZMP, an intermediate of de novo purine synthesis. Cells take the riboside up and phosphorylate it into ZMP, which behaves as an AMP mimic. Under the nonproprietary name acadesine it was studied for two decades as an intravenous drug, first as an “adenosine-regulating agent” for ischemia during coronary artery bypass grafting and later as an apoptosis-inducing agent in B-cell malignancies; neither program produced an approved product anywhere.

It appears in the research-compound market for a different reason: a 2008 Salk Institute study in which four weeks of AICAR alone, with no training, increased treadmill endurance in sedentary mice. That result made it a reference “exercise mimetic,” put it on the World Anti-Doping Agency Prohibited List, and placed it alongside peptides in vendor catalogs despite its unrelated chemistry.

How it works

AICAR is a prodrug. Its effects depend on cellular uptake through nucleoside transporters and phosphorylation by adenosine kinase to ZMP, which then signals as though the cell’s AMP level had risen:

• ZMP activates AMPK without changing ATP, ADP, or AMP — in rat hepatocytes, intracellular ZMP accumulation mimicked both effects of AMP on AMP-activated protein kinase (direct allosteric activation and increased phosphorylation by the upstream kinase), inactivated HMG-CoA reductase, and nearly shut off fatty acid and sterol synthesis; the authors proposed AICAR as a specific tool for activating AMPK in intact cells (Corton et al., European Journal of Biochemistry, 1995, 229(2):558–565)
• Endurance and oxidative-muscle gene expression in mice — AICAR given to sedentary mice for 4 weeks induced metabolic genes in skeletal muscle and increased running endurance by 44%, the finding that coined the “exercise mimetic” label; the same paper showed the PPARδ agonist GW1516 synergizing with training rather than replacing it (Narkar et al., Cell, 2008, 134(3):405–415)
• Adenosine regulation in ischemic tissue — the cardiac-surgery program rested on a separate rationale: in animals, acadesine selectively increases the availability of adenosine in ischemic tissue, which was the basis for perioperative infusion in CABG trials (Leung et al., Anesthesia & Analgesia, 1994, 78(3):420–434)
• Selective apoptosis in B-cell CLL cells — acadesine induced caspase-dependent apoptosis in B-CLL cells from all 70 patient samples tested (EC50 about 380 µM) while T cells were only slightly affected at concentrations up to 4 mM; the effect required uptake and conversion to ZMP, and AMPK phosphorylation accompanied apoptosis in B cells but not in T cells (Campàs et al., Blood, 2003, 101(9):3674–3680). The same group reported AICAR-induced apoptosis in Jurkat T cells that depended on neither ZMP formation nor AMPK (López et al., Biochemical Journal, 2003, 370(Pt 3):1027–1032), a reminder that the compound has actions beyond the ZMP–AMPK axis

In humans, acute AICAR infusion increased skeletal-muscle 2-deoxyglucose uptake about 2-fold in 29 healthy men with only a 7% rise in whole-body glucose disposal, and notably without a measurable increase in muscle AMPK activity or phosphorylation (Cuthbertson et al., Diabetes, 2007, 56(8):2078–2084). In ten men with type 2 diabetes, intravenous AICAR reduced hepatic glucose output and plasma free fatty acids, again without increased muscle AMPK phosphorylation (Boon et al., Diabetologia, 2008, 51(10):1893–1900).

Research status

AICAR has an unusually long human record for a research-market compound, and almost all of it ends in failure or abandonment.

Cardiac surgery (the main program). Acadesine was tested as a 7-hour perioperative infusion (0.1 mg/kg per minute, also added to cardioplegia solution) in patients undergoing on-pump CABG. An initial 116-patient trial found the drug safe, with mild uric acid increases at the high dose and no significant reduction in ischemia (Leung et al., Anesthesia & Analgesia, 1994). The original developer, Gensia, filed a New Drug Application in December 1992 under the trade name Arasine; the FDA would not act on it without a further Phase 3 trial, and when that trial missed its endpoint in 1994 the application went nowhere. An individual-patient meta-analysis of the five randomized trials (4,043 patients) nonetheless reported a 27% reduction in perioperative myocardial infarction, a 50% reduction in cardiac death through postoperative day 4, and a 26% reduction in the combined outcome of MI, stroke, or cardiac death (Mangano, JAMA, 1997, 277(4):325–332). On the strength of that analysis, PeriCor Therapeutics revived the compound and licensed it to Schering-Plough in 2007 for a definitive Phase 3 trial.

RED-CABG (NCT00872001) — stopped for futility. The Reduction in Cardiovascular Events by Acadesine in Patients Undergoing CABG trial randomized intermediate- to high-risk patients at 300 sites in 7 countries between May 2009 and July 2010. A prespecified futility analysis found a very low likelihood of a significant result, and the trial was halted after 3,080 of a planned 7,500 participants. The primary composite of all-cause death, nonfatal stroke, or need for mechanical support for severe left ventricular dysfunction through day 28 occurred in 5.0% of the placebo group and 5.1% of the acadesine group (odds ratio 1.01; 95% CI 0.73–1.41), with no differences in key secondary endpoints (Newman et al., JAMA, 2012, 308(2):157–164). Merck, which had absorbed Schering-Plough, terminated its acadesine development program in the third quarter of 2010. The meta-analytic signal did not survive a single adequately sized trial.

B-cell chronic lymphocytic leukemia (NCT00559624). A multicenter open-label Phase I/II study gave acadesine as a 4-hour intravenous infusion to 24 patients with relapsed or refractory CLL. The maximum tolerated dose was 210 mg/kg; at 315 mg/kg, two of three patients had dose-limiting toxicities, one tumor lysis syndrome and one renal impairment. Grade 2 or higher hyperuricemia was common and managed with prophylactic allopurinol (a single grade 4 case occurred before prophylaxis was introduced), and the authors also recorded transient anemia and thrombocytopenia, renal impairment, and clinically significant infusion-related hypotension. There were no formal complete or partial responses; “anti-leukemic activity” (a greater than 20% fall in B-cell counts or palpable lymph node size) was seen in 16 of 24 patients (Van Den Neste et al., Cancer Chemotherapy and Pharmacology, 2013, 71(3):581–591). No later-phase CLL trial followed.

Myelodysplastic syndromes (NCT01813838). A French cooperative-group Phase I/II trial of acadesine at 140 to 315 mg/kg/day in MDS enrolled 5 patients between 2013 and 2015 before being terminated for renal toxicity, per the registry record.

Other human work. A Phase 2 pilot of AICAR in Lesch-Nyhan disease (NCT00004314), begun in 1996 under the National Center for Research Resources, lists 2 participants and is marked completed without posted results. The two small infusion studies in healthy men and in type 2 diabetes (above) are the principal human data on the metabolic actions that drive research-market interest; a few other acute physiology studies exist, none longer than a single infusion.

Current state (as of October 2026). Acadesine has never been approved anywhere: the 1992 US application was abandoned after the 1994 trial failure, no filing has followed, and no marketing authorization could be identified in any jurisdiction. ClinicalTrials.gov lists no recruiting or active interventional study of acadesine or AICAR as a drug; the only recruiting record returned by an acadesine search is a purine-rich-diet study in AICA-ribosiduria, a rare ATIC-deficiency disorder, in which the compound is a disease biomarker rather than the intervention. No human trial has tested AICAR for endurance, body composition, or performance.

Common dosage forms

• Clinical formulation (historical) — a sterile solution for intravenous infusion used in the cardiac-surgery and CLL trials, also added to cardioplegia solution during bypass. It was never commercialized.
• Research supply — lyophilized powder in vials, most often labeled 50 mg, intended for reconstitution; some suppliers list 56 mg or 100 mg vials, bulk powder, or 10 mg capsules. These are sold as research chemicals with no approved product behind them.
• Reference-grade reagent — AICAR has been a catalog biochemical for activating AMPK in cell culture since the 1990s, which is the format most of the mechanistic literature actually used.

Key considerations

• Regulatory status. Acadesine has no FDA approval, no approval abroad, and no active clinical development; the 1990s application stalled, the Phase 3 cardiac program was stopped for futility in 2010, and the leukemia programs did not advance past Phase I/II. Anything sold today is research-use material. The Research Market status reflects the absence of US approval or any development program aimed at approval. It is also prohibited at all times in sport: the 2026 WADA Prohibited List names AICAR as an example under S4.4, activators of AMPK. It has been on the List since 2009, first under gene doping and, since 2012, under metabolic modulators.
• The endurance claim is a mouse result. The 44% endurance gain comes from four weeks of daily treatment in sedentary mice. The two human infusion studies measured glucose and fatty-acid kinetics for hours, not weeks, and neither detected the increase in muscle AMPK activity the mouse model would predict. There is no human evidence that AICAR improves endurance or body composition.
• Documented human safety signals. Hyperuricemia recurs across the clinical trials because AICAR is metabolized to uric acid; at oncology doses the program recorded a grade 4 hyperuricemia event, tumor lysis syndrome, renal impairment, and infusion-related hypotension, and a separate MDS trial was terminated for renal toxicity. Those doses were far above anything the mechanistic literature uses, but renal and urate effects are the first thing a reader should look for in any human data.
• ZMP is not a clean AMPK switch. AICAR must be taken up and phosphorylated to act on AMPK, yet in Jurkat cells it killed cells without either ZMP formation or AMPK activation, and ZMP itself is the metabolite proposed to be toxic in Lesch-Nyhan disease. Effects attributed to “AMPK activation” in AICAR studies may partly belong to purine metabolism or other targets instead.
• Naming confusion. “AICAR” is used for both the riboside (acadesine, the compound that is dosed) and the ribotide (ZMP, the intracellular active form); some sources, including automated indexes, expand it as “ribonucleotide” rather than “ribonucleoside.” Acadesine, AICA riboside, and Z-riboside all refer to the dosed compound. It is unrelated to the peptide AMPK activator MOTS-c, which sits next to it in vendor catalogs.