Adrenomedullin

A native 52-amino-acid vasodilator peptide of the calcitonin family, isolated from human pheochromocytoma in 1993 and tested by Japanese investigators as an intravenous infusion for ulcerative colitis, Crohn's disease, stroke and COVID-19 pneumonia without any approved product.

Also referenced as: AM, ADM, huAM, adrenomedullin (1-52), adrenomedullin acetate

Also appears in: Immune

COA Public product evidence No certificate records are bound to a tracked Adrenomedullin product page A record counts here only when a vendor publishes a certificate against one of its own tracked Adrenomedullin product pages. Other certificates naming this compound may still exist — search the wider certificate database. →

---

What is Adrenomedullin?

Adrenomedullin is a native human peptide of 52 amino acids. It was discovered in 1993 by Kazuo Kitamura, Kenji Kangawa and colleagues in Japan, who purified it from human pheochromocytoma tissue by following its ability to raise platelet cyclic AMP and named it for the adrenal medulla, where it is abundant in normal tissue and in the tumors arising from it (Kitamura et al., Biochemical and Biophysical Research Communications, 1993;192(2):553–560). The mature peptide carries a six-residue ring closed by a disulfide bond between cysteine 16 and cysteine 21 and an amidated C-terminal tyrosine, features that place it in the calcitonin family alongside calcitonin, CGRP, amylin and adrenomedullin 2 (Kita and Kitamura, Hypertension Research, 2022;45(3):389–400). It is cut from a 185-residue precursor, preproadrenomedullin, which also contains the 20-residue proadrenomedullin N-terminal peptide (PAMP) (Kitamura et al., Biochemical and Biophysical Research Communications, 1993;194(2):720–725).

Adrenomedullin circulates in healthy people, rises sharply in sepsis and heart failure, and acts as a potent vasodilator and a protector of the vascular endothelium. The mid-regional precursor fragment (MR-proADM) is a sepsis biomarker, while the peptide itself has been infused as an experimental drug, almost entirely in Japan, for heart failure, myocardial infarction, inflammatory bowel disease, stroke and COVID-19 pneumonia. No adrenomedullin product is approved in any country. Synthetic human adrenomedullin (1-52) is available as a laboratory reagent under its own name or as AM or ADM.

How it works

• A CLR–RAMP2/RAMP3 receptor agonist. The calcitonin receptor-like receptor (CLR) becomes a CGRP receptor when chaperoned to the cell surface by receptor activity-modifying protein 1 (RAMP1) and an adrenomedullin receptor when paired with RAMP2, the discovery that defined the RAMP family; the RAMP3 complex is the second adrenomedullin receptor (McLatchie et al., Nature, 1998;393(6683):333–339). Both complexes couple to Gs and raise cyclic AMP, and a polyethylene glycol chain attached at the N-terminus leaves the maximal cyclic AMP response intact, which is the basis of the long-acting derivatives (Kubo et al., Peptides, 2014;57:118–121).
• Vasodilation with a reflex sympathetic cost. In 13 patients with precapillary pulmonary hypertension, a 0.05 µg/kg/min infusion raised cardiac index by 44%, cut pulmonary vascular resistance from 19.7 to 13.4 units, lowered mean systemic pressure from 81 to 72 mm Hg and raised heart rate from 73 to 79 beats per minute, while placebo changed nothing (Nagaya et al., Heart, 2000;84(6):653–658). Intravenous adrenomedullin stimulates sympathetic nerve activity through its short-acting vasodilation, raising heart rate and renin secretion, and lessening these reactions was the stated reason for PEGylating the peptide (Kubo et al., Peptides, 2014;57:118–121).
• Essential for vascular integrity. Deleting the adrenomedullin gene in mice is lethal in utero: more than 80% of homozygous embryos were dead by embryonic day 13.5 with severe hemorrhage from endothelial cells detaching from their basement membranes, and heterozygotes lived to adulthood with raised blood pressure and reduced nitric oxide production (Shindo et al., Circulation, 2001;104(16):1964–1971). This barrier-protective role underlies the interest in sepsis and inflamed gut mucosa.
• Mucosal healing in the gut. In rodent colitis the peptide suppresses inflammatory cytokines and accelerates vascular and epithelial repair; a 40 kDa PEGylated form given as a single intravenous dose as low as 25 nmol/kg treated dextran sulfate sodium colitis in mice without downregulating receptor expression despite sustained high plasma concentrations (Kitamura et al., International Journal of Molecular Sciences, 2025;26(19):9373).

Research status

Early cardiovascular infusion studies. The first human trials were hemodynamic studies at Japan’s National Cardiovascular Center. In 7 patients with congestive heart failure and 7 healthy volunteers, a 0.05 µg/kg/min infusion raised cardiac index by 49% and 39% respectively, lowered pulmonary capillary wedge pressure and increased urine volume and sodium excretion (Nagaya et al., Circulation, 2000;101(5):498–503). In 10 patients with a first acute myocardial infarction, a 12-hour infusion begun before reperfusion kept hemodynamics stable in all but 2 and was followed by a smaller infarct on cardiac magnetic resonance at 3 months, in an uncontrolled design (Kataoka et al., Journal of Cardiovascular Pharmacology, 2010;56(4):413–419). None of these cardiovascular programs progressed to registration trials.

Inflammatory bowel disease. The University of Miyazaki program began with a case series: 7 patients with refractory ulcerative colitis received 8-hour daily infusions for 14 days, all improved on the Disease Activity Index, endoscopy showed mucosal healing, and 4 remained in remission a year later (Ashizuka et al., Digestive Diseases and Sciences, 2016;61(3):872–880). A double-blind phase 1 study in 36 healthy men found no serious adverse events, one withdrawal for headache and a half-life under 60 minutes (Kita et al., Drug Design, Development and Therapy, 2020;14:1–11). The multicenter phase 2a trial in steroid-resistant ulcerative colitis (Japic CTI-205255) randomized 28 patients to placebo or 5, 10 or 15 ng/kg/min for 8 hours daily for 14 days. The primary endpoint, change in Mayo score at 2 weeks, showed no difference among the four groups; at 8 weeks, with incomplete follow-up, the 15 ng/kg/min group had a larger Mayo score decrease than placebo (−9.3 versus −3.0, P = 0.035) and clinical remission in 3 of 3 evaluable patients versus 0 of 2 (Kita et al., Journal of Gastroenterology, 2021;56(2):147–157). The companion phase 2a trial in biologic-resistant Crohn’s disease randomized 24 patients to placebo or 10 or 15 ng/kg/min for 8 hours daily for 7 days; the primary endpoint, change in Crohn’s Disease Activity Index at 8 weeks, was not met, and the only signal was a sustained reduction in the treated groups that separated from placebo at 24 weeks in a mixed-effects model (P = 0.043) (Kita et al., Journal of Gastroenterology and Hepatology, 2022;37(11):2051–2059). As of November 2026 no later-phase trial of native adrenomedullin in inflammatory bowel disease has been published or registered on ClinicalTrials.gov.

Stroke, CADASIL and COVID-19. The investigator-initiated AMFIS trial (jRCT2051190092) enrolled 60 patients with acute non-embolic ischemic stroke between January 16, 2020 and November 14, 2021 and randomized them to placebo or 9 ng/kg/min given as daily pulses or as a continuous 72-hour infusion followed by daily pulses. There were no treatment-related serious adverse events, but the change in NIHSS score to day 7 was −0.76 with adrenomedullin and −1.08 with placebo, so no benefit was shown (Yoshimoto et al., EClinicalMedicine, 2024;77:102901). A single-arm phase 2 study in 60 patients with CADASIL (AMCAD) used the change in frontal cerebral blood flow at day 28 as its primary endpoint (NCT06072118, completed June 12, 2023); the protocol is published (Washida et al., Cerebral Circulation - Cognition and Behavior, 2024;6:100211) and no results have appeared in PubMed. Two phase 2a trials in COVID-19 pneumonia randomized 33 patients with severe and 31 with moderate pneumonia to 15 ng/kg/min or placebo; neither the duration of mechanical ventilation or oxygen support nor clinical status at 30 days differed, with 3 deaths on placebo and 1 on adrenomedullin in the severe group (Kita et al., Viruses, 2025;17(7):982).

PEGylated adrenomedullin. To escape the short half-life and the hypotensive surge, the Miyazaki group attached polyethylene glycol to the N-terminus; the conjugate kept the same maximal cyclic AMP response as the native peptide with a smaller acute blood-pressure fall and a longer plasma half-life in rats (Kubo et al., Peptides, 2014;57:118–121), and a 40 kDa conjugate with greater plasma stability followed (Kitamura et al., International Journal of Molecular Sciences, 2025;26(19):9373). The derivative reached the clinic as HM201 in a first-in-human, placebo-controlled phase 1 study of single and multiple intravenous doses in 53 healthy adults in Australia, run from November 11, 2021 to December 19, 2022 with ulcerative colitis and Crohn’s disease as target conditions; the registry lists Himuka AM Pharma as a collaborator (NCT05088369). No patient trial of HM201 is registered on ClinicalTrials.gov as of November 2026, and no clinical results have been published. A separate inhaled conjugate, Bayer’s BAY 1097761 (PEG-ADM), was tested in the SEAL phase 2a/b trial in acute respiratory distress syndrome: 90 ventilated patients were randomized to 960 µg, 1,920 µg or placebo. Both doses were tolerated, but there was no effect on the clinical utility index and ventilator-free survival at day 28 was 52% in the 960 µg group against 67% with 1,920 µg and 65% with placebo, so Part B was not conducted (Karagiannidis et al., Critical Care, 2025;29(1):448); the registry records that the decision rested on those results rather than on any safety finding (NCT04417036).

Mechanistic human studies. The Danish Headache Center uses the peptide as a provocation agent: in a crossover study of 20 patients with migraine without aura, a 20-minute infusion triggered migraine-like attacks in 11 patients (55%) versus 3 (15%) after saline (P = 0.039) (Ghanizada et al., Neurology, 2021;96(20):e2488–e2499; NCT04111484). Follow-on crossover studies in post-traumatic headache (NCT07059546) and migraine (NCT07451769, started March 12, 2026) are recruiting and investigate pathophysiology rather than treatment.

Common dosage forms

• No approved presentation. No adrenomedullin product has been authorized by the FDA, the EMA, the PMDA or any other regulator.
• Investigational lyophilized vial. The Japanese trials used a sterile lyophilized formulation for intravenous infusion by hospital staff, supplied only to trial sites.
• PEGylated derivatives. HM201 was an intravenous investigational product in one phase 1 study and BAY 1097761 an inhaled investigational product in the SEAL trial; neither has been marketed.
• Research-market material. Synthetic adrenomedullin (1-52) is supplied as a lyophilized powder for laboratory use.

Key considerations

• Regulatory status. The Research Market status reflects the absence of any FDA-approved or foreign-approved adrenomedullin product and of any registered approval-directed trial of the peptide as of November 2026. HM201 finished a healthy-volunteer phase 1 study in December 2022 without a registered successor, the inhaled conjugate was not taken past Part A of the SEAL trial, and the Japanese investigator-initiated trials have all completed.
• The efficacy record is thin and mostly negative. No randomized trial of the peptide has met its primary endpoint; the positive signals are an 8-week secondary analysis in five evaluable colitis patients and a 24-week modeled difference in Crohn’s disease.
• Vasodilation is the safety signal. Headache, flushing, palpitations, tachycardia and hypotension recur in every human study, and the peptide reliably provokes migraine-like attacks in susceptible people; no serious treatment-related adverse events have been reported in the controlled trials.
• A hospital infusion, not a self-administered peptide. The native peptide has a half-life under 60 minutes, required 8- to 72-hour intravenous infusions in all of its trials, and has no subcutaneous or oral clinical data.
• Naming. Adrenomedullin, AM, ADM, huAM and adrenomedullin (1-52) denote the same 52-residue peptide. HM201, PEG-ADM and BAY 1097761 are PEGylated derivatives; MR-proADM and bio-ADM are diagnostic assays; adrecizumab (enibarcimab) is an antibody against the peptide, not a form of it.