Thymosin Alpha-1: One Trial, Two Answers

The short answer on thymosin alpha-1

Thymosin alpha-1 is a 28-residue peptide the body produces by cleaving a larger precursor protein. Research describes it as an immune modulator rather than a simple stimulant. Its largest sepsis trial found no mortality benefit, and its hepatitis B evidence sits at very low certainty. It holds no FDA approval in the United States.

*Last updated 16 October 2026. Every figure here was checked against its primary record. Trial and review data came from PubMed and PubMed Central. Regulatory statements came from Federal Register notices, FDA briefing documents, the advisory committee's Final Summary Minutes, the Orphan Drug Product designation database and Drugs@FDA. The Cochrane review was verified at the level of its published abstract; its full text could not be retrieved, and the section on it says what that leaves unconfirmed. Prepared by the PrimeTime BioLabs editorial team. A named medical reviewer has not yet been assigned.*

Why the published thymosin alpha-1 numbers do not agree

Pages covering this peptide quote wildly different figures, and most of them drop the study that produced each one. Below, each figure is set against what the primary record says.

Method note. Every row was checked against the original publication or the official regulatory document on 16 October 2026. Where two renderings of the same paper disagree, both appear. Where a claim could not be substantiated, it is marked and excluded from the rest of this page.

The sepsis result. Circulating sources quote either a hazard ratio near 1.0 or an odds ratio of 0.73, often without naming either source. The trial reported 0.99. The meta-analysis reported 0.73 overall and 0.82 among higher-quality studies, with that second interval crossing the line of no effect. Different studies, different populations, different statistical methods.

The trial's own hazard ratio, in two places. The BMJ article of record gives 0.99, with a confidence interval of 0.77 to 1.27 and a p-value of 0.93, stated consistently in its abstract, results and limitations. The PubMed Central rendering of that same paper gives 0.97, 0.76 to 1.24, and 0.82. This page uses the BMJ figures throughout, because that is the version of record. Anyone checking the PMC copy will see something different, and now knows why.

The Cochrane certainty rating. Secondary coverage describes the 2026 Cochrane review as showing very low certainty evidence, which is accurate but incomplete. Its point estimates favour thymosin alpha-1 across most outcomes. Very low certainty means the review authors have little confidence in those estimates, not that the estimates were unfavourable.

The COVID-19 figures. A risk ratio of 1.03 with heterogeneity of 90% circulates widely and is usually dated 2023. It belongs to a review published online in December 2022. A separate review, genuinely published in 2023, reported a risk ratio of 0.59 in the opposite direction. Both exist, and quoting one without the other misrepresents the field.

The hepatitis C result. Vendor pages sometimes report a 41% response rate. That figure describes only patients who completed all 48 weeks of therapy, a group of 83 out of 275 originally assigned. In the full randomised population the result was 12.7% against 10.5%, which was not statistically significant.

The 1999 hepatitis B trial. This study appears on several pages as supporting evidence. Its authors wrote that their results did not confirm the efficacy reported in other studies. The difference between arms did not reach significance.

The orphan designation count. FDA's own briefing document states that four US orphan designations exist. Three are individually retrievable from FDA's designation database. This page reports both facts rather than picking one.

Two figures this page does not use. A claim that the compound is approved in more than 35 countries traces back to a manufacturer's own filings with securities regulators, which said over 30 in one year and over 34 in another. No regulatory document substantiates 35. A second claim, of more than 30 clinical trials covering more than 11,000 subjects, has no locatable source at all. Neither figure appears anywhere else on this page.

What thymosin alpha-1 actually is

Thymosin alpha-1 is a peptide built from 28 amino acids, with an acetyl group capping its front end. The body makes it by cleaving a larger precursor protein called prothymosin alpha, a relationship described in the published mechanism literature.

Four names circulate for the same thing, and keeping them apart solves most of the confusion readers arrive with.

Thymosin alpha-1 names the molecule itself. Thymalfasin is the international nonproprietary name, which is the name regulators and pharmacopoeias use for the synthetic pharmaceutical version. Zadaxin is a brand name under which that pharmaceutical version has been marketed outside the United States. TA-1 is shorthand that appears in research papers and vendor listings.

Thymosin fraction 5 comes up in the history and deserves one clarification. It was a mixture extracted from thymus tissue, not a single compound, and thymosin alpha-1 was isolated from it. A page describing fraction 5 is describing a historical preparation rather than this peptide.

Precursor is not the same as equivalent. Prothymosin alpha is a larger protein with its own biology, and thymosin alpha-1 is one fragment released from it. Research findings about one do not automatically transfer to the other.

This page does not state a molecular formula, molecular weight or full sequence for the peptide. Those identifiers appear on product listings across the industry, and they were not independently verified against a primary chemical reference for this review. Publishing an unverified identifier would undercut the point of the exercise.

Where thymosin alpha-1 came from

Allan Goldstein characterised this peptide in 1972, working from the thymus extract described above. A synthetic version followed, and clinical development ran across several decades, multiple companies and several countries.

An overview of that development programme was published in 2010 by authors at SciClone Pharmaceuticals, the company that marketed the compound. That paper describes studies ranging from small investigator-led work in primary immune deficiency through to large phase 3 trials. Readers should weigh it as a company account of a company product, which is what it is.

A long development history and an absence of US approval can sit together comfortably. Trials get run, results come back mixed, sponsors change priorities, and no marketing application reaches the finish line. That is roughly what happened here.

How thymosin alpha-1 is thought to work

Vendor pages tend to describe this peptide as boosting immunity. Published mechanism work describes something more specific and considerably stranger.

Researchers at the University of Perugia reported in 2006 that thymosin alpha-1 primes dendritic cells through Toll-like receptor 9 signalling. That same pathway activates indoleamine 2,3-dioxygenase, an enzyme that breaks down tryptophan and dampens immune responses. In their experiments, the peptide induced that enzyme, produced interleukin-10 and generated regulatory T cells.

Regulatory T cells suppress immune activity. A compound that produces them alongside T-helper type 1 responses is not simply switching the immune system up.

The authors described this as establishing a regulatory environment that balances inflammation against tolerance. Their follow-up review in 2007 extended the picture to plasmacytoid dendritic cells and interferon signalling, and described protection against fungal and viral challenge in mice.

Species and model matter enormously here. That 2006 work used human and mouse dendritic cells in culture, with the in vivo experiments done in mice. The 2007 paper is a review of the same research programme. On the grading scale used below, this sits at levels 4 through 6, meaning animal, in-vitro and mechanistic evidence.

What this body of work does not establish is any clinical outcome in any patient. A pathway effect is a reason to run a trial, not a result from one.

Does thymosin alpha-1 raise or lower inflammation

Honestly, the published mechanism work points both ways, and that is the interesting part. The same signalling route that primes a T-helper type 1 response also activates a tolerising enzyme. Researchers have described the net effect as regulatory rather than directional.

Anyone who tells you the answer is simply "lowers" or simply "raises" has picked one half of the literature.

What the thymosin alpha-1 trials actually found

Grades below follow one hierarchy, declared here so the labels mean something.

• Primary human interventional evidence
• Pooled human evidence, carrying its source's own certainty rating
• Human observational evidence
• Animal evidence
• In-vitro evidence
• Mechanistic interpretation
• Insufficient or conflicting evidence

Method note. Each study below was located through PubMed and checked against its primary record. The trial and meta-analysis entries were additionally checked against their full texts in PubMed Central. The Cochrane entry was verified at the level of its published abstract only, because its full text could not be retrieved through any available route. Pooled analyses keep whatever certainty their own authors assigned, rather than receiving a house label.

The trials and what each one tested

• Study: TESTS, Wu and colleagues, 2025 · Population: Adults aged 18 to 85 with sepsis by Sepsis-3 criteria · Design: Multicentre, double blinded, placebo controlled, phase 3, 22 centres in China · Randomised: 1,106 · Period: September 2016 to December 2020 · Primary endpoint: 28 day all cause mortality
• Study: Gu and colleagues, 2025 · Population: Adults with sepsis across 11 randomised trials · Design: Systematic review and meta-analysis with trial sequential analysis · Participants: 1,927 · Period: Searched to 2024 · Primary endpoint: 28 day mortality
• Study: Cochrane review, Naing and colleagues, 2026 · Population: Adults with chronic hepatitis B across 10 randomised trials · Design: Cochrane systematic review with GRADE assessment · Participants: 1,349 · Period: Trials published 1991 to 2018, searched to 10 June 2026 · Primary endpoints: All cause mortality, serious adverse events, quality of life
• Study: Shang and colleagues, 2022 · Population: Hospitalised adults with COVID-19 across 9 studies · Design: Systematic review and meta-analysis · Participants: 5,352 · Primary endpoint: Mortality
• Study: Soeroto and colleagues, 2023 · Population: Adults with moderate to critical COVID-19 across 8 studies · Design: Systematic review, meta-analysis and meta-regression · Primary endpoint: Mortality
• Study: Ciancio and colleagues, 2012 · Population: Chronic hepatitis C not responding to prior therapy · Design: Randomised, placebo controlled, multicentre · Randomised: 552 · Period: 48 weeks of treatment · Primary endpoint: Sustained viral response
• Study: Mutchnick and colleagues, 1999 · Population: Adults positive for hepatitis B e antigen and viral DNA · Design: Multicentre, randomised, double blind, placebo controlled, phase 3 · Randomised: 97 · Period: 6 months of treatment, 6 months follow-up · Primary endpoint: Sustained loss of viral DNA
• Study: ETASS, Wu and colleagues, 2013 · Population: Adults with severe sepsis in six Chinese hospitals · Design: Multicentre, single blind, randomised, controlled · Randomised: 361 · Period: May 2008 to December 2010 · Primary endpoint: 28 day all cause mortality

What each one reported

• Study: TESTS 2025 · Result: Death at 28 days occurred in 127 of 542 (23.4%) against 132 of 547 (24.1%). Hazard ratio 0.99, 95% CI 0.77 to 1.27, P=0.93 · Other findings: No secondary or safety outcome differed significantly. Prespecified subgroup interactions were explored, and none met the trial's own corrected significance threshold of P=0.006 · Authors' conclusion: No clear evidence that the drug reduces 28 day mortality in sepsis
• Study: Gu 2025 · Result: Overall odds ratio 0.73, 95% CI 0.59 to 0.90, P=0.003, from 218 deaths against 274 · Other findings: Higher-quality subgroup OR 0.82, 0.65 to 1.03, P=0.09, across 5 studies and 1,568 patients. Multicentre subgroup OR 0.86, 0.68 to 1.08, P=0.20. Trial sequential analysis found the required information size was not reached · Authors' conclusion: Potential to reduce mortality, with efficacy differing across subgroups
• Study: Cochrane 2026 · Result: All cause mortality risk ratio 0.53, 0.29 to 0.96, no heterogeneity, 3 studies, 907 participants, very low certainty · Other findings: Serious adverse events RR 0.72, 0.53 to 0.99, 5 studies, 1,056 participants, low certainty. Non-serious adverse events RR 0.47, 0.27 to 0.83, 5 studies, 300 participants, very low certainty. Histological improvement RR 0.51, 0.13 to 2.06, heterogeneity 74%, 2 studies, 702 participants, very low certainty. Hepatitis B related morbidity RR 0.86, 0.54 to 1.40, 3 studies, 854 participants, very low certainty · Authors' conclusion: Uncertain whether the drug reduces mortality or serious adverse events
• Study: Shang 2022 · Result: Mortality risk ratio 1.03, 95% CI 0.60 to 1.75, P=0.92, heterogeneity 90% · Other findings: Subgroup results favoured treatment in older patients and in severe or critical disease · Authors' conclusion: The results do not support use in hospitalised adults with COVID-19
• Study: Soeroto 2023 · Result: Mortality risk ratio 0.59, 95% CI 0.37 to 0.93, P=0.02, heterogeneity 84% · Other findings: No difference in need for mechanical ventilation or in hospital length of stay · Authors' conclusion: May reduce mortality, with randomised trials still required
• Study: Ciancio 2012 · Result: Sustained viral response 12.7% against 10.5% in the full randomised population, P=0.407 · Other findings: Among the 182 patients who completed all 48 weeks, response was 41.0% against 26.3%, P=0.048. No significant difference in adverse events · Authors' conclusion: No role in primary therapy, with a possible adjuvant role in preventing relapse
• Study: Mutchnick 1999 · Result: Complete response in 7 of 49 (14%) against 2 of 48 (4%), P=0.084 · Other findings: Sustained viral DNA loss with antigen clearance in 12 (25%) against 6 (13%) · Authors' conclusion: These results do not confirm the efficacy reported in other studies
• Study: ETASS 2013 · Result: Death at 28 days in 26.0% against 35.0%. Relative risk 0.74, 95% CI 0.54 to 1.02. Unstratified P=0.062, log rank P=0.049 · Other findings: Greater improvement in a monocyte surface marker at days 3 and 7 · Authors' conclusion: May be effective in a targeted population with severe sepsis

Evidence status, funding and citations

• Study: TESTS 2025 · Main limitation: Conducted entirely in China, which limits generalisability. The authors concede the trial may have been underpowered · Funding and conflicts: Funders included SciClone Pharmaceuticals, which markets the compound. Twelve authors declared grants from that company and three declared consultancy fees. The company is stated to have had no role in design or interpretation · Grade: 1, primary human interventional · Citation: BMJ. 2025;388:e082583. PMID 39814420
• Study: Gu 2025 · Main limitation: Not independent of TESTS. Three authors were TESTS investigators, including its first, second and senior authors, and TESTS supplies 57% of pooled patients. The higher-quality subgroup used an author-devised numerical score rather than a standard grading instrument · Funding and conflicts: Public research grants. The authors declared no commercial or financial relationship constituting a conflict · Grade: 2, pooled human evidence · Citation: Front Cell Infect Microbiol. 2025;15:1673959. PMID 40969554
• Study: Cochrane 2026 · Main limitation: Certainty very low for every outcome except serious adverse events, which was low. Downgrading reflected risk of bias, imprecision and heterogeneity. Sixteen further studies await assessment because of incomplete reporting. All cause mortality and hepatitis B related mortality share identical statistics across the same three studies, so they are one finding rather than two · Funding and conflicts: The review had no dedicated funding. Four of the ten included trials were industry funded · Grade: 2, pooled human evidence, very low to low certainty · Citation: Cochrane Database Syst Rev. 2026;9(9):CD014610. PMID 42713852
• Study: Shang 2022 · Main limitation: Heterogeneity of 90% means the included studies disagree profoundly with each other · Grade: 2, pooled human evidence · Citation: Int Immunopharmacol. 2022;114:109584. PMID 36527881
• Study: Soeroto 2023 · Main limitation: Heterogeneity of 84%, and the authors themselves call for randomised trials · Grade: 2, pooled human evidence · Citation: Inflammopharmacology. 2023;31(6):3317-3325. PMID 37845598
• Study: Ciancio 2012 · Main limitation: The favourable completer analysis covers 182 of 552 randomised patients, so it describes a self-selected group · Grade: 1, primary human interventional, negative on its primary endpoint · Citation: J Viral Hepat. 2012;19 Suppl 1:52-9. PMID 22233415
• Study: Mutchnick 1999 · Main limitation: Small sample, and the primary comparison did not reach significance · Grade: 1, primary human interventional, negative · Citation: J Viral Hepat. 1999;6(5):397-403. PMID 10607256
• Study: ETASS 2013 · Main limitation: Single blind rather than double blind, and the result depends on which statistical test is read · Grade: 1, primary human interventional, borderline · Citation: Crit Care. 2013;17(1):R8. PMID 23327199

Eight studies, four clinical questions. Sepsis, chronic hepatitis B, chronic hepatitis C and COVID-19 are not interchangeable, and counting positive against negative across them would be meaningless.

Why the big sepsis trial and the meta-analysis disagree

Neither paper got its arithmetic wrong. The disagreement comes almost entirely from which studies you let into the pool.

Pooling all 11 trials gave an odds ratio of 0.73, which favours treatment and reached statistical significance. Restricting the pool to the five studies the authors rated higher quality moved it to 0.82, with a confidence interval from 0.65 to 1.03. That interval crosses 1.0, meaning no benefit remains demonstrated. Restricting instead to multicentre trials gave 0.86, which also crosses.

Smaller single-centre trials tend to report larger effects than bigger multicentre ones. In this analysis the single-centre pool gave an odds ratio of 0.40, against 0.86 for the multicentre pool. Several explanations compete, including publication patterns, looser blinding and chance. The authors reported the subgroup differences as significant, at P=0.02 for the quality split and P=0.004 for the centre split. They did not name the test that produced those values.

One detail about that quality rating deserves stating. The authors built their own numerical score from the Cochrane risk-of-bias domains. They awarded points per domain, and called 11 or more out of 14 relatively high quality. That is a reasonable approach and it is not a standard, externally validated grading instrument.

The authors also ran a trial sequential analysis, which asks whether enough patients have been studied to settle the question. It concluded that the required information size has not been reached. On their own calculation, roughly 7,681 patients would be needed, against 1,927 accumulated.

Now the part that most coverage leaves out. These two papers are not independent sources. The trial's first author, second author and senior author all appear on the meta-analysis. The trial contributes 57% of the pooled patients and supplied individual patient data for the subgroup work. The two papers also share two funders. The meta-analysis declares no commercial or financial relationship constituting a conflict. The trial it draws on declares manufacturer funding and a dozen authors with company grants.

None of that is misconduct, and none of it makes either paper wrong. It does mean a reader should not treat the meta-analysis as outside confirmation of the trial, or the reverse.

Thymosin alpha-1 and thymosin beta-4 are not the same molecule

No, thymosin alpha-1 is not TB-500. The two compounds share a family name and almost nothing else, and a third molecule complicates the picture further.

Three distinct things circulate under thymosin names. Thymosin alpha-1 is the 28-residue immune-signalling peptide this page covers. Thymosin beta-4 is a separate protein of 43 amino acids, studied mainly for actin binding and tissue repair. TB-500 is a seven-residue fragment taken from thymosin beta-4, not the full protein, as our TB-500 research review sets out in detail.

• Property: Molecule length · Thymosin alpha-1: 28 residues · Thymosin beta-4: 43 residues · TB-500: 7 residues
• Property: Relationship · Thymosin alpha-1: Cleaved from prothymosin alpha · Thymosin beta-4: A complete protein · TB-500: A fragment of thymosin beta-4
• Property: Main research focus · Thymosin alpha-1: Immune signalling, dendritic cells, T cells · Thymosin beta-4: Actin binding, cell migration, tissue repair · TB-500: Carries the actin-binding region of the parent
• Property: Human trial evidence · Thymosin alpha-1: Multiple randomised trials across several conditions · Thymosin beta-4: Trials in wound and eye conditions · TB-500: No published controlled human trial of the fragment alone
• Property: US regulatory status · Thymosin alpha-1: Not FDA approved, no 503A category · Thymosin beta-4: Not FDA approved · TB-500: Not FDA approved, not on the 503A list

The family name describes where these molecules were first found, inside thymus tissue, and says nothing about what they do. Evidence generated on one of them does not transfer to another.

Neither compound is a treatment, and nothing here ranks one above the other.

What the trials report about thymosin alpha-1 safety

Published trials have not reported excess harm. That is a genuinely different statement from saying the compound is safe, and the gap between those two sentences matters.

Start with what the trials found. The largest sepsis trial reported that no safety outcome differed significantly between groups. The hepatitis C trial found no significant difference in the incidence of adverse events across 552 patients over 48 weeks. The Cochrane review reported fewer serious adverse events with treatment, a risk ratio of 0.72, at low certainty. It also reported fewer non-serious adverse events, at very low certainty.

Regulators looked at the same field and raised separate concerns. FDA's reviewers assessed the substance for compounding in 2024. They cited reported liver enzyme elevations and thyroid-stimulating hormone abnormalities. One fatal immune complication was also cited, in a transplant patient given the compound alongside another therapy.

Three limits apply to all of it. Trials of this size cannot detect rare harms. A complication occurring once in several thousand patients would not appear in a population of a few hundred. Trials of this length cannot speak to long-term effects, since the longest follow-up here runs to a few years. And every participant in every trial was ill, so none of this describes what happens in healthy people.

Nobody has studied this compound in a healthy population. That absence is the single largest gap in its safety record.

Where thymosin alpha-1 stands with US regulators

Four concepts get mixed together constantly, and separating them answers most regulatory questions about this peptide in one pass.

FDA approval means the agency reviewed a marketing application and authorised a product for a specific use. Thymosin alpha-1 has none. FDA's own briefing document states that it has approved no drug products containing the substance, and it does not appear in the Drugs@FDA database.

Orphan drug designation is granted early in development for rare conditions. It brings incentives such as fee waivers and tax credits, and it involves no finding of safety or effectiveness and no permission to market. FDA's briefing document states that four US designations exist for this compound. Three are individually retrievable from the designation database, all sponsored by SciClone Pharmaceuticals and all recorded as not approved for the designated indication. Those three cover chronic active hepatitis B, designated 3 May 1991; DiGeorge anomaly with immune defects, designated 8 January 1998; and hepatocellular carcinoma, designated 6 March 2000.

An advisory committee recommendation is a vote by outside experts that FDA may accept or decline. It changes nothing on its own.

Research use only describes how a material may lawfully be offered for sale. Our guide to US peptide law covers what that label does and does not do.

What happened at the December 2024 FDA meeting

FDA's Pharmacy Compounding Advisory Committee met on 4 December 2024 under docket FDA-2024-N-4777, with thymosin alpha-1 as its afternoon item. The committee considered two substances, the free base and the acetate salt.

Members first voted on whether to handle both with a single ballot. That procedural vote went 20 to 1, and because one member objected, two separate votes followed. On each substance the committee divided 4 in favour of listing and 17 against.

A version of this has spread describing a single vote of 17 to 4 covering both forms. The official minutes record something different, namely two votes, each with 4 supporting and 17 opposing, after a separate procedural ballot.

FDA's own reviewers had already recommended against listing. Their briefing document cited gaps in how well the substances are characterised. Those gaps included missing impurity and endotoxin data, immunogenicity risk that could not be ruled out, and questionable solubility. Limited nonclinical data and an unclear history of actual compounding use also counted against listing.

Rulemaking would be required for anything to change, with a proposed rule and a public comment period before any final rule. No such rulemaking has followed the December 2024 recommendation.

Where it sits in the compounding category system

Thymosin alpha-1 appears in none of FDA's three 503A categories as of the agency's list dated 14 May 2026. It sits instead on the list of substances nominated but withdrawn, consistent with the original nomination having been pulled by the nominating pharmacy.

It was also absent from the July 2026 advisory committee agenda, which covered a different set of peptides entirely. At least one secondary source currently claims it returned to Category 1, and FDA's own document does not support that.

Status outside the United States

A pharmaceutical version has been marketed abroad under the Zadaxin name. FDA's briefing document names six jurisdictions. Italy registers it as an adjuvant to influenza vaccination in immunocompromised patients. India, Hong Kong, Singapore, Mexico and Indonesia follow, with hepatitis B among the registered uses in several. FDA describes the Singapore and Mexico entries as monographs rather than marketing authorisations.

FDA recorded that it could not independently verify the manufacturer's broader claims. The company's own filings with securities regulators described approval in over 30 countries in one year and over 34 in another. Approval in another country reflects that country's regulator, and it carries no weight in the United States.

What the thymosin alpha-1 evidence does not settle

Whether it helps in sepsis. The largest trial found nothing, the pooled estimate shrinks toward nothing as study quality rises, and the trial sequential analysis says roughly four times more patients would be needed to settle it.

Whether it helps in chronic hepatitis B. The Cochrane review's point estimates look favourable, and its certainty ratings are very low for almost every outcome. Sixteen more studies sit awaiting assessment because of incomplete reporting.

What it does in COVID-19. Two reviews a year apart reached opposite conclusions, both with heterogeneity above 80%.

Anything about healthy people. Every participant in every trial on this page was ill. No published trial has enrolled a healthy population.

Whether the mechanism translates. The dendritic cell and Toll-like receptor work is careful, and it was done in cells and mice. The path from that work to a human outcome has not been demonstrated.

Absence of evidence is not evidence of absence. These are open questions rather than demonstrated failures, and several of them could be closed by trials nobody has yet run.

Research material and the approved medicine are not the same thing

Every clinical result on this page describes a pharmaceutical product, manufactured under defined controls and given to patients in a registered study. None of it describes research-use material sold under the same compound name.

No published evidence establishes that a research-grade preparation matches a pharmaceutical product's formulation, purity, stability or biological behaviour. The studies that would establish it have not been run. That is a statement about what has been tested, not an accusation about any supplier.

Analytical documentation helps within defined limits. A certificate of analysis can support identity and purity by a stated method, on a stated lot, at a stated date. Safety, effectiveness and regulatory standing all fall outside what any such document can show. Chemical purity says nothing about how a material behaves in a living system. Our guide to reading a certificate of analysis explains what each section covers. Our supplier due-diligence checklist sets out what to ask before ordering anything.

PrimeTime BioLabs supplies thymosin alpha-1 as research material for laboratory use only. The trial evidence on this page does not characterise that material.

Nothing on this page is medical advice. This page reviews published research and public regulatory documents for people doing laboratory work.

The bottom line on thymosin alpha-1

Thymosin alpha-1 has a deeper clinical record than almost any compound sold as a research peptide. It also has a record that has grown less convincing as the trials have grown larger and better controlled.

Its biggest sepsis trial found no mortality benefit. Its most favourable pooled estimate weakens toward nothing when restricted to the stronger studies, and the two papers reporting those results share authors. Its hepatitis B evidence points in a favourable direction at very low certainty. Its COVID-19 evidence contradicts itself across two reviews.

US regulators have approved nothing containing it, an advisory committee voted against making it available for compounding, and no rulemaking has followed either way.

What remains genuinely interesting is the mechanism work, which describes something more subtle than immune stimulation. Whether any of that reaches patients is a question the published trials have not answered.