Vasoactive Intestinal Peptide (VIP): An Evidence Ledger

Two database records for one molecule

Said and Mutt pulled a vasodilator out of hog small intestine and published it in Science in September 1970. Their paper described a 28-residue polypeptide with systemic vasodilation, hypotension, increased cardiac output, respiratory stimulation and hyperglycaemia among its actions, and stated that it was chemically distinct from the kinins, substance P, glucagon and secretin. The amino-acid sequence followed in 1974 in the European Journal of Biochemistry, determined with the help of kallikrein cleavage. UniProt P01282 places the mature peptide at residues 125 to 152 of a 170-residue precursor that also carries PHM-27 at residues 81 to 107, and records Asn152 as amidated.

That terminal amide splits the identifier record in two. FDA GSRS holds a substance called VASOACTIVE INTESTINAL PEPTIDE, UNII 6J2WVD66KR, CAS 37221-79-7, classed as a chemical, formula C147H237N43O43S, molecular weight 3326.7889. It holds a second substance called AVIPTADIL, UNII A67JUW790C, CAS 40077-57-4, classed as a protein, with the same 28 residues and a complete C-terminal asparaginamide modification at position 28. PubChem CID 53314964 carries the first CAS and the free-acid formula, while the GSRS record for VIP itself cross-references PubChem 16132300, the amidated compound, rather than the free-acid entry its own CAS number resolves to. The identifier that circulates for VIP therefore describes a molecule lacking the amide that the native peptide and the synthetic drug substance both have, and the databases do not agree with each other about which entry the name points at.

Harmar and colleagues, writing for the IUPHAR receptor nomenclature subcommittee in 2012, recorded three class B G-protein-coupled receptors: PAC1, VPAC1 and VPAC2 (PMID 22289055). That assignment constrains every experiment below. PAC1 is selective for PACAP. VPAC1 and VPAC2 respond to both VIP and PACAP with high affinity. No study administering unmodified VIP can therefore attribute an effect to a VIP-specific receptor, because the two receptors it occupies are shared with a structurally related peptide that is present endogenously.

One minute in plasma

The pharmacokinetics were measured in humans in 1978. Domschke and colleagues, writing in Gut, infused graded intravenous doses of 0.6, 1.3 and 3.3 pmol/kg/min over 30-minute periods into four healthy volunteers and measured plasma VIP by radioimmunoassay. After each infusion stopped, plasma concentrations fell by first-order kinetics with an average disappearance half-time of one minute. Apparent metabolic clearance was about 9 mL/kg/min and apparent volume of distribution about 14 mL/kg. The sample was four volunteers, and no later human pharmacokinetic study was located that would independently confirm the figure.

The same experiment recorded what the higher infusion rate did. Plateau concentrations reached the range found in Verner-Morrison syndrome. Glucose, free fatty acids and calcium rose slightly but significantly, pulse rate and blood-pressure amplitude increased, and cutaneous flushing occurred. The authors read their own data as arguing against VIP acting as a circulating hormone under physiological conditions. Every human study in the sections below works around that half-life: continuous infusion, aerosol delivered to the tissue of interest, or injection directly into the target tissue. Nothing here is oral.

The most direct attempt to engineer past the problem produced no published data. Pemziviptadil, development code PB1046, was a long-acting VIP analogue carried on an elastin-like polypeptide. Its phase 2 trial in pulmonary arterial hypertension, NCT03556020, enrolled 35 subjects and was terminated with the reason given as delayed study-drug resupply during COVID-19; it completed in January 2022 and has posted no results. A second trial in hospitalised COVID-19 patients, NCT04433546, enrolled 54 and was terminated because the company was no longer pursuing the indication. PubMed searches for both names returned no trial report.

Pulmonary hypertension and sarcoidosis: small, open, short

Petkov and colleagues published the founding clinical claim in the Journal of Clinical Investigation in 2003, under a title calling VIP a new drug for treatment of primary pulmonary hypertension. The paper reported a deficiency of the peptide in serum and lung tissue of patients by radioimmunoassay and immunohistochemistry, upregulation of corresponding receptor sites by Northern blot, Western blot and immunological techniques, and administration to eight study patients, in whom mean pulmonary artery pressure decreased and cardiac output and mixed venous oxygen saturation increased. There was no control group and no randomisation. The paper's own closing sentence asked for further investigation rather than claiming an established treatment.

Leuchte and colleagues measured the acute effect more carefully five years later in the European Respiratory Journal. Twenty patients with pulmonary hypertension of three different aetiologies — nine with pulmonary arterial hypertension, eight with hypertension in lung disease, three with chronic thromboembolic disease — inhaled a single 100 microgram dose during right-heart catheterisation. Six of the twenty showed a pulmonary vascular resistance reduction greater than 20 per cent. Stroke volume and mixed venous oxygen saturation improved, systemic blood pressure was unaffected, and the authors described the pulmonary vasodilating effect in their own words as modest and short-lived.

Sarcoidosis produced the one immunological human dataset. Prasse and colleagues, in an open phase II study published in the American Journal of Respiratory and Critical Care Medicine in 2010, gave nebulised VIP for four weeks to 20 patients with histologically proven sarcoidosis and active disease. Tumour necrosis factor alpha production by cells isolated from bronchoalveolar lavage fell significantly, and lavage CD4+CD127-CD25+ T cells increased. The study was open-label with no placebo arm, so the design cannot separate the peptide from four weeks of nebulisation and clinical attention. The abstract is silent on lung function and the full text is not open access, so what the paper does or does not report on that endpoint is not established here.

Two COVID-19 trials that do not agree

Youssef and colleagues reported a multicentre randomised trial in Critical Care Medicine in November 2022: 196 patients with COVID-19 respiratory failure across ten United States hospitals, randomised 2:1 to three days of intravenous aviptadil or placebo. The primary endpoint, alive and free from respiratory failure at day 60, returned an odds ratio of 1.6 with a 95 per cent confidence interval of 0.86 to 3.11 — it did not reach significance. A secondary survival comparison at day 60 gave an odds ratio of 2.0, confidence interval 1.1 to 3.9, p equal to 0.035. The authors concluded that the benefit-versus-risk balance was favourable.

A larger trial reached the opposite conclusion. TESICO, run within the NIH ACTIV-3b programme across 28 United States sites and reported by Brown and colleagues in Lancet Respiratory Medicine in September 2023, enrolled 473 participants; 461 formed the modified intention-to-treat population, 231 on aviptadil and 230 on matched placebo. Dosing targeted 600, 1200 and 1800 pmol/kg across three daily 12-hour infusions. The data and safety monitoring board recommended stopping the aviptadil comparison for futility on 25 May 2022. The day-90 ordinal outcome odds ratio was 1.11, confidence interval 0.80 to 1.55, p equal to 0.54.

Mortality in TESICO went the same way. Eighty-six aviptadil participants and 83 placebo participants had died by day 90, a cumulative 38 per cent against 36 per cent, hazard ratio 1.04, confidence interval 0.77 to 1.41, p equal to 0.78. The day-5 composite safety outcome — death, serious adverse events, organ failure, serious infection, or grade 3 or 4 adverse events — occurred in 63 per cent of the aviptadil group and 56 per cent of placebo, odds ratio 1.40, confidence interval 0.94 to 2.08. The accompanying comment by Lee and Slutsky was titled simply as a negative trial.

One smaller inhaled trial reports differently again. Esendagli and colleagues, in Medical Principles and Practice in 2025, randomised 80 hospitalised COVID-19 patients across nine centres 1:1 to inhaled aviptadil or placebo alongside standard care. Mean time to discharge was 7.8 plus or minus 4.0 days against 10.0 plus or minus 5.0, p equal to 0.049. Modified Borg scores did not differ at day 3 and were lower at day 7. CT lung damage score improvement at day 28 was greater. Deaths were 5.1 per cent against 12.2 per cent, in a trial with 80 participants and no power to resolve mortality.

What an infusion does to a person

Two headache-research groups established the acute human profile more precisely than any therapeutic programme did. Hansen and colleagues, in Cephalalgia in 2006, infused 8 pmol/kg/min or saline for 25 minutes into 12 healthy volunteers in a double-blind crossover. Headache was very mild, with a maximum verbal rating of 2 out of 10; five participants developed headache on VIP against one on placebo. The superficial temporal artery dilated markedly in the first 30 minutes, mean flow velocity in the middle cerebral artery fell, regional cerebral blood flow did not change significantly, and heart rate rose significantly.

Lengthening the infusion changed the outcome. Pellesi and colleagues repeated the design in 12 healthy volunteers over two hours, reported in Cephalalgia in 2020, and found superficial temporal artery dilation lasting more than two hours, higher headache incidence than placebo at p equal to 0.003, and increased tear production. The following year, in JAMA Network Open, the same group gave a two-hour infusion to 21 patients with migraine without aura in a randomised double-blind crossover. Fifteen of 21, or 71 per cent with a confidence interval of 48 to 89, developed migraine attacks, against one patient on placebo, p below 0.001. The attacks mimicked the patients' spontaneous ones.

Endogenous excess supplies the other end of the range. VIP-secreting neuroendocrine tumours, reviewed by Siddappa and Vege in Pancreas in 2019, present with a characteristic syndrome of watery diarrhoea; most arise in the pancreas and most are found late. Domschke's flushing, tachycardia and metabolic shifts at the top infusion rate sit on the same continuum. In Pellesi and colleagues' 2021 crossover trial in 21 patients with migraine without aura, 15 of 21 developed migraine attacks after a two-hour VIP infusion against one on placebo (PMID 34357396). Descriptions of this peptide as uniformly calming do not account for that result.

The intranasal record, and what is not in it

One pharmacodynamic study of the nasal route was located. Cui and colleagues, in Die Pharmazie in 2013, injected aggregated amyloid-beta 25-35 intracerebroventricularly, treated animals intranasally over one week, and assessed spatial learning by Morris water maze. At 40 micrograms per millilitre, escape latency did not differ significantly from the model group. At 200 micrograms per millilitre, treated animals did not differ significantly from normal controls. Nasal mucosa showed minor irritation that resolved spontaneously within a week of treatment ending. The abstract is internally inconsistent about species: it describes the intracerebroventricular injection in mice but attributes both the escape-latency result and the mucosal assessment to rats, so the species for the behavioural endpoint cannot be determined from the abstract. Group sizes are not stated. The work is rodent.

The registry contains nothing corresponding. A ClinicalTrials.gov search for intranasal vasoactive intestinal peptide returns one study, on allergen immunotherapy, unrelated to this compound. Registered interventions involving VIP or aviptadil use intravenous infusion at a headache centre, intravenous aviptadil in COVID-19, inhaled or nebulised aviptadil, intracavernosal injection of a fixed-dose combination, and subcutaneous injection of a long-acting analogue. No registered trial gives this peptide to a human being by the nasal route, for any indication, at any phase.

The nasal claims that circulate most widely attach to a mould-illness protocol, and their author's indexed publications do not contain them. Shoemaker's PubMed record includes a 2014 volumetric MRI study in Neurotoxicology and Teratology comparing 17 patients with 18 controls, which reported caudate atrophy and pallidum enlargement, and a 2015 whole-blood transcriptomics study of a ciguatoxin-exposed CIRS cohort (Ryan, Wu and Shoemaker, 11 patients against 11 controls, PMID 25889530) — a different exposure from the water-damaged-building population. Neither administers VIP. Neither has a treatment arm. The protocol material lives in venues PubMed does not index and in a United States patent, which is a document type that establishes a claim of invention rather than a measured result.

What is not known

No regulator has approved this peptide as a medicine in the United States, and the endpoint that would support one has not been met: TESICO, the largest randomised trial in the file, which enrolled 473 participants of whom 461 formed the modified intention-to-treat population, was stopped for futility with a day-90 odds ratio of 1.11 and near-identical mortality in both arms. Outside the United States a fixed-dose intracavernosal combination of aviptadil and phentolamine holds national marketing authorisations, which is a narrow local approval for a route that deposits the peptide directly in the target tissue and says nothing about systemic administration. Chronic dosing is uncharacterised in every indication: the pulmonary hypertension data consist of one uncontrolled series of eight and one single-dose acute study of twenty, the sarcoidosis data of one open four-week series of twenty, and no controlled trial has run beyond the days-long windows used in COVID-19. The long-acting analogue that would have answered the durability question was tested in 35 patients with pulmonary arterial hypertension and 54 with COVID-19, both trials terminated for non-clinical reasons, neither posting results nor producing a publication. Nothing has been given by the nasal route to a human being in any registered study, so no nasal pharmacokinetics, bioavailability, dose-response or adverse-event profile exists in people. Receptor attribution is unresolved at the level of the experiment, since VPAC1 and VPAC2 bind both VIP and PACAP with high affinity. And the identifier record itself is split, with the widely quoted CAS number and PubChem entry describing the free acid rather than the amidated peptide that circulates in the body.

Questions