Content

Thymosin Alpha 1 Peptide: The Trial Record

Creator: Julian Reyes, Health Writer
Published: 2026-08-12

Thymalfasin — the drug form of the thymosin alpha 1 peptide — is approved in over 30 countries, and the largest trial ever run on it, in 1,106 sepsis patients, returned a 28-day mortality hazard ratio of 0.97.

Julian Reyes, Health Writer August 12, 2026 14 min read

A woman in a pale sage sweater slicing fresh oranges at a bright kitchen counter in soft morning light

The short version

Thymosin alpha-1 is a 28-amino-acid peptide of about 3,108 daltons, cut from the parent protein prothymosin alpha; its international drug name is thymalfasin.
In the United States it is not FDA-approved for any indication and is dispensed only as a compounded medication. Thymalfasin holds marketing approval in over 30 countries, principally for chronic hepatitis B.
Every adequately powered double-blind placebo-controlled trial has reported a null primary result: sepsis at n=1,106 (HR 0.97, P=0.82), hepatitis C at n=552 (12.7% vs 10.5%, P=0.407), hepatitis B at n=97 (14% vs 4%, P=0.084).
The gap between small studies and large ones is measurable: in an 11-trial sepsis meta-analysis of 1,927 patients, the single-centre stratum gave OR 0.40 (0.25–0.63) and the multicentre stratum was null, OR 0.86 (0.68–1.08).
Safety is the strongest part of the record. In 1,089 sepsis patients, adverse events occurred in 66.4% on thymosin alpha-1 versus 67.6% on placebo (P=0.70), serious adverse events in 26.8% versus 29.3% (P=0.38).
It is a modulator, not a stimulant: no high-affinity receptor has been identified, it induces a tolerance-promoting arm alongside an inflammatory one, and in human lymphocytes it does nothing without a co-stimulus present.

The thymosin alpha 1 peptide is unusual in one respect: it has a drug name, a brand name, and four decades of randomized trials behind it. As thymalfasin, marketed as ZADAXIN, it holds marketing approval in over 30 countries, per the manufacturer's SEC filings, and PubMed indexes 864 records for it — 64 tagged as randomized controlled trials, 21 as meta-analyses, none retracted.

That record has a shape, and it repeats across every indication tested. Small, open-label, single-centre studies report benefit; the large blinded trials do not. A 1,106-patient Phase 3 in sepsis returned a 28-day mortality hazard ratio of 0.97. A 552-patient Phase 3 in hepatitis C missed its primary endpoint at P=0.407. A 97-patient Phase III in chronic hepatitis B came in at P=0.084. Across those same trials the tolerability data are notably clean.

What follows is an inventory of that record, every number reported with the population that produced it — because a response rate from a hepatitis B trial describes hepatitis B patients and no one else.

What the thymosin alpha 1 peptide is

Thymosin alpha-1 is a 28-amino-acid peptide with an average molecular weight of 3,108.3 g/mol, corresponding to residues 2–29 of the parent protein prothymosin alpha and acetylated at its first residue. As a pharmaceutical it is named thymalfasin. It is not the only thymic peptide studied for immune signalling — thymulin is a separate, zinc-dependent molecule from the same gland.

It was isolated and sequenced from calf thymus in 1977, purified out of thymosin fraction 5 — a crude extract it is 10 to 1,000 times more active than in laboratory assays — and today it is made synthetically rather than extracted.

Pharmacokinetics are short. In nine healthy volunteers dosed subcutaneously at 900 µg/m², the half-life measured 1.9 to 3 hours and no accumulation appeared over five days — claims of a multi-day half-life do not match that study. The route is subcutaneous injection in every trial on record, at 1.6 mg twice weekly in the hepatitis indications approved outside the United States and 1.6 mg every 12 hours for seven days in the sepsis trial.

Where thymalfasin is approved, and where it is not

In the United States, thymosin alpha-1 is not FDA-approved for any indication, and it is supplied only as a compounded medication prepared by a licensed U.S. pharmacy. Compounded medications are not reviewed by the FDA for safety, effectiveness, or quality. Outside the country, thymalfasin holds marketing approval in over 30 nations, principally for chronic hepatitis B.

That "over 30" figure comes from the manufacturer's SEC filings, consistent for more than a decade, with approvals concentrated in China, the Pacific Rim, Latin America, Eastern Europe and the Middle East; some countries add hepatitis C or vaccine adjuvant use. The FDA has noted it cannot independently verify approval in every country claimed. Within the European Union, authorization exists in Italy alone, for a narrow indication: enhancing the immune response to influenza vaccination, not hepatitis. Four U.S. orphan-drug designations were granted, and none converted into a marketing approval.

On the compounding side the facts are specific. Thymosin alpha-1 is not listed at 21 CFR 216.23, the 503A bulk drug substances list; there is no USP monograph for it, and it is not a component of an FDA-approved drug — the three routes Section 503A recognizes. In December 2024 an FDA advisory committee voted 4 in favour and 17 against adding it to that list, citing incomplete characterization of the substance, insufficient safety information with immunogenicity singled out, and a lack of evidence of effectiveness across the twelve nominated uses. None of this bars a prescription — a licensed provider may still prescribe, and that decision is between you and your doctor.

The trial record at a glance

Six randomized trials carry most of the weight here. Five reported a null primary endpoint; the sixth missed significance at P=0.062. Sample sizes run from 25 to 1,106, and every population is a hospital population — sepsis, hepatitis, melanoma, liver cancer — rather than healthy adults.

Trial	Population	n	Design	Primary result
TESTS (2025)	ICU patients with sepsis	1,106	Phase 3, multicentre, double-blind, placebo-controlled	28-day mortality 23.4% vs 24.1%; HR 0.97 (0.76–1.24), P=0.82 — null
Ciancio (2012)	Chronic hepatitis C, prior non-responders	552	Phase 3, double-blind, placebo-controlled, 52 sites	Sustained virological response 12.7% vs 10.5%, P=0.407 — failed
Mutchnick (1999)	Chronic hepatitis B	97	Phase III, double-blind, placebo-controlled	Complete response 14% vs 4%, P=0.084 — not significant
Maio (2010)	Metastatic melanoma	488	Randomized Phase 2, open-label, five arms	Median overall survival 9.4 vs 6.6 months; HR 0.80 (0.63–1.02), P=.08 — not significant
ETASS (2013)	Severe sepsis	361	Randomized, single-blind, controlled	28-day mortality 26.0% vs 35.0%; RR 0.74 (0.54–1.02), P=0.062 — not significant
Gish (2009)	Unresectable hepatocellular carcinoma	25	Randomized pilot, chemoembolization ± peptide	No difference in response rate or median overall survival

TESTS: 1,106 sepsis patients, no mortality difference

TESTS is the largest randomized trial of thymosin alpha-1 ever conducted. Across 22 centres in China between September 2016 and December 2020, it randomized 1,106 patients with sepsis, double-blind, to 1.6 mg subcutaneously every 12 hours for seven days or to placebo, and was published in the BMJ in January 2025.

The primary endpoint, 28-day all-cause mortality, was 23.4% versus 24.1% on placebo — hazard ratio 0.97 (95% CI 0.76–1.24), P=0.82. Those are the corrected figures: a May 2025 erratum restored survival data for nine participants recorded as lost to follow-up during the COVID-19 pandemic, moving the originally published 0.99 (0.77–1.27) to 0.97 — the PubMed abstract still carries the original numbers. Ninety-day mortality was 31.0% versus 32.4%, HR 0.95, P=0.61, and no secondary or safety outcome differed statistically significantly. One prespecified subgroup ran the other way: patients under 60 had HR 1.67 (95% CI 1.04–2.67), interaction P=0.01, favouring placebo — which the investigators treat as hypothesis-generating.

The earlier ETASS trial is the one most often called positive, and its own report does not support that reading: in 361 patients with severe sepsis, 28-day mortality was 26.0% versus 35.0%, relative risk 0.74 (95% CI 0.54–1.02), P=0.062 — the primary endpoint was not met. In-hospital mortality, a secondary endpoint, reached P=0.032.

The 552-patient hepatitis C trial

The definitive hepatitis C trial randomized 552 prior non-responders across 52 European sites to peginterferon alfa-2a plus ribavirin with either thymosin alpha-1 at 1.6 mg subcutaneously twice weekly for 48 weeks or placebo, double-blind. Sustained virological response, the primary endpoint, was 12.7% versus 10.5%, P=0.407.

A completer analysis in the same paper reported 41.0% versus 26.3%, P=0.048 — but only 182 of 552 patients, 33%, completed. That is a selected subgroup rather than the randomized comparison, and it is the figure most often quoted without its denominator. The authors concluded the peptide appears to play no role in the primary therapy of the disease. Two further U.S. Phase 3 trials in hepatitis C, roughly 1,000 patients between them, were announced by press release in December 2005 and May 2006 and never published; development in the indication was then discontinued.

Hepatitis B, melanoma and liver cancer

The Phase III chronic hepatitis B trial randomized 97 patients — 49 to 1.6 mg twice weekly for six months, 48 to placebo. Complete response was 7 of 49 (14%) versus 2 of 48 (4%), P=0.084, and the investigators wrote that their results did not confirm the treatment efficacy reported in other clinical studies.

A meta-analysis pooling five hepatitis B trials and 353 patients found no virological difference at end of treatment — odds ratio 0.56 (0.20–1.52), the point estimate favouring control — and a significant difference only 12 months after treatment stopped, OR 2.67 (1.25–5.68). An effect absent while patients are on a drug that appears a year after they stop, across five small trials, is the signature of small-study bias rather than of delayed benefit.

In metastatic melanoma, 488 patients were randomized across five open-label arms combining dacarbazine, interferon alfa and three peptide doses. Objective response ran from 6.1% to 12.1% in the experimental arms versus 4.1% in the control arm, with no dose-response — the 6.4 mg arm did worse than the 3.2 mg arm — and median overall survival was 9.4 versus 6.6 months, HR 0.80 (0.63–1.02), P=.08. It is registered as Phase 2, not the Phase 3 registration trial it is sometimes called. In liver cancer, a randomized pilot adding the peptide to chemoembolization in 14 patients versus 11 controls found no difference in response rate or median overall survival over 24 weeks; it was a safety study, not powered for efficacy.

COVID-19: no randomized treatment trial with results

There is no published randomized controlled trial of thymosin alpha-1 as a COVID-19 treatment with reported results. Every treatment dataset is retrospective or observational, spanning 76 to 2,282 patients, and those datasets disagree with each other in a way that is itself informative.

The most-cited positive report is a retrospective cohort of 76 patients with severe COVID-19 at two Wuhan hospitals: mortality 11.11% versus 30.00%, P=.044 — a causal claim a 76-patient retrospective cohort cannot support. In 771 critically ill patients, unadjusted 28-day mortality was 41.3% versus 60.6%, P<0.001, and after propensity matching 51.0% versus 52.9%, no difference. That single comparison shows how much of the observational signal tracks which patients were selected for treatment. In the largest dataset, 2,282 patients, the adjusted odds of non-recovery were higher with the peptide: OR 1.5 (95% CI 1.1–2.1), P=0.028.

The mechanistic claim attached to the early cohort has not replicated: in 275 patients the peptide made no difference to CD4⁺ counts (P=0.851) or CD8⁺ counts (P=0.842), and viral clearance took longer in the treated group, 14 versus 11 days, P=0.028. IDSA, NIH and WHO guidelines do not discuss the compound, and there is no long-COVID evidence at all; the studies examined acute hospitalised illness.

Why the small trials and the large ones disagree

A 2025 meta-analysis of 11 sepsis trials and 1,927 patients — the only one that includes TESTS — makes the pattern quantitative. Overall it reported 28-day mortality OR 0.73 (0.59–0.90), P=0.003. Split by quality, the high-quality stratum was null at OR 0.82 (0.65–1.03) and the multicentre stratum null at OR 0.86 (0.68–1.08), while the single-centre stratum carried the entire effect, OR 0.40 (0.25–0.63).

Trial sequential analysis in that paper concluded the accumulated sample size remains inadequate. The published record is also skewed relative to the trials actually run. A melanoma Phase 3 with an FDA-agreed Special Protocol Assessment dated 17 November 2008 was never registered or reported; the two U.S. hepatitis C Phase 3 trials were never published; two lung-cancer trials registered in September 2016 never started; and a Phase 4 liver-cancer trial planned for 360 patients has been silent since 2014.

How it works: modulation, not stimulation

Thymosin alpha-1 has no identified high-affinity receptor — a point stated affirmatively in the primary literature, where the absence of a specific receptor is given as the reason researchers turned to studying how the peptide interacts with cell membranes directly.

What is documented is signalling dependency. The FDA's summary describes immunomodulatory properties mediated at least in part by interaction with Toll-like receptor 9 on dendritic and lymphoid progenitor cells. In dendritic-cell work that axis runs through MyD88 to IRF7 and interferon output, and through IDO induction to a tolerance-promoting state involving regulatory T cells and IL-10 — alongside, not instead of, an IL-12-driven inflammatory arm. That evidence comes from knockout and deficiency experiments; direct binding has not been demonstrated.

Direction of effect depends on context: in human monocyte-derived dendritic cells the peptide raised HLA class I and class II expression alongside viral-type stimuli, and sharply lowered the same parameters under bacterial TLR2 and TLR4 stimulation. Human lymphocyte work agrees: it increases the number of high-affinity IL-2 receptors on stimulated cells without altering their affinity — and produces no effect at all without mitogenic stimulation. It behaves as a co-stimulus, not a standalone activator. That is why "immune booster" misdescribes the molecule on the science, before any question of marketing language.

What the safety data show

The strongest safety evidence comes from the same trial that found no mortality benefit. Among 1,089 sepsis patients, any adverse event occurred in 360 of 542 (66.4%) on the peptide versus 370 of 547 (67.6%) on placebo, difference −1.2 percentage points (95% CI −6.8 to 4.4), P=0.70.

Serious adverse events occurred in 145 (26.8%) versus 160 (29.3%), P=0.38 — at the highest sustained dose ever studied, in a critically ill population.

The most common adverse effects across trials are local injection-site reactions: irritation, redness, discomfort. The FDA's review catalogued specific trial signals — ALT flares in chronic hepatitis B, thyroid-stimulating hormone abnormalities in chronic hepatitis C, nipple pain in the liver-cancer pilot — plus one serious population-specific signal: fatal immune haemolytic anaemia and engraftment failure in stem-cell transplant recipients, with the agency stating the peptide could cause or worsen graft-versus-host disease.

Cautions follow. Hypersensitivity rules it out, as does deliberate immunosuppression, including organ- and stem-cell-transplant recipients. Labeling outside the United States warns for children, pregnant and lactating women, autoimmune disease, and immunosuppressed patients. For pregnancy and breastfeeding there is no data at all — no teratogenicity or lactation study has been identified, and with no U.S. label there is no pregnancy category. Treatment courses in the trial record run six to twelve months at most.

What an evaluation looks like

An evaluation starts with why the peptide is being considered at all, because the trial evidence describes hospital populations — sepsis, chronic hepatitis, advanced cancer — and none of those results transfer to a healthy adult. A licensed provider reviews every request, not everyone qualifies, and the provider may decline.

Worth raising at intake: any history of organ or stem-cell transplantation, current immunosuppressive therapy, an autoimmune diagnosis, pregnancy or breastfeeding, thyroid or liver conditions given the ALT and TSH signals from the hepatitis trials, and a full list of current medications. If a prescription is written, the medication is compounded by a licensed U.S. pharmacy and given by subcutaneous injection. Current pricing and the intake questionnaire live on the Thymosin Alpha-1 product page, which sits in the Immune Support category.

It is also where expectations get set against the record above rather than against marketing copy: the tolerability data are reassuring and the efficacy data are unestablished, and both halves of that sentence come from the same trials.

This article is for general education and is not medical advice. It does not describe an outcome any individual should expect. Discuss your own health history, medications and questions with a licensed clinician.

References

Wu J, Pei F, Zhou L, et al. The efficacy and safety of thymosin α1 for sepsis (TESTS): multicentre, double blinded, randomised, placebo controlled, phase 3 trial. BMJ. 2025;388:e082583. PubMed. doi:10.1136/bmj-2024-082583 (correction: PMID 40447307 — survival data updated, HR now 0.97)
Wu J, Zhou L, Liu J, et al. The efficacy of thymosin alpha 1 for severe sepsis (ETASS): a multicenter, single-blind, randomized and controlled trial. Crit Care. 2013;17(1):R8. PubMed. doi:10.1186/cc11932
Gu B, Zhou Y, Nie Y, Wang L, et al. Thymosin alpha 1 for sepsis: a systematic review and meta-analysis of randomised controlled trials. Front Cell Infect Microbiol. 2025;15:1673959. PubMed
Ciancio A, Andreone P, Kaiser S, et al. Thymosin alpha-1 with peginterferon alfa-2a/ribavirin for chronic hepatitis C not responsive to IFN/ribavirin: an adjuvant role? J Viral Hepat. 2012;19 Suppl 1:52–9. PubMed. doi:10.1111/j.1365-2893.2011.01524.x
Mutchnick MG, Lindsay KL, Schiff ER, et al. Thymosin alpha1 treatment of chronic hepatitis B: results of a phase III multicentre, randomized, double-blind and placebo-controlled study. J Viral Hepat. 1999;6(5):397–403. PubMed. doi:10.1046/j.1365-2893.1999.00181.x
Chan HL, Tang JL, Tam W, Sung JJ. The efficacy of thymosin in the treatment of chronic hepatitis B virus infection: a meta-analysis. Aliment Pharmacol Ther. 2001;15(12):1899–905. PubMed
Maio M, Mackiewicz A, Testori A, et al. Large randomized study of thymosin alpha 1, interferon alfa, or both in combination with dacarbazine in patients with metastatic melanoma. J Clin Oncol. 2010;28(10):1780–7. PubMed. doi:10.1200/JCO.2009.25.5208
Gish RG, Gordon SC, Nelson D, Rustgi V, et al. A randomized controlled trial of thymalfasin plus transarterial chemoembolization for unresectable hepatocellular carcinoma. Hepatol Int. 2009;3(3):480–9. PubMed. doi:10.1007/s12072-009-9132-3
Liu Y, Pan Y, Hu Z, et al. Thymosin alpha 1 reduces the mortality of severe coronavirus disease 2019 by restoration of lymphocytopenia and reversion of exhausted T cells. Clin Infect Dis. 2020;71(16):2150–7. PubMed. doi:10.1093/cid/ciaa630
Sun Q, Xie J, Zheng R, et al. The effect of thymosin α1 on mortality of critical COVID-19 patients: a multicenter retrospective study. Int Immunopharmacol. 2021;90:107143. PubMed
Liu J, Shen Y, Wen Z, et al. Efficacy of thymosin alpha 1 in the treatment of COVID-19: a multicenter cohort study. Front Immunol. 2021;12:673693. PubMed
Wang Z, Chen J, Zhu C, et al. Thymosin alpha-1 has no beneficial effect on restoring CD4+ and CD8+ T lymphocyte counts in COVID-19 patients. Front Immunol. 2021;12:568789. PubMed
Rost KL, Wierich W, Masayuki F, Tuthill CW, Horwitz DL, Herrmann WM. Pharmacokinetics of thymosin alpha1 after subcutaneous injection of three different formulations in healthy volunteers. Int J Clin Pharmacol Ther. 1999;37(1):51–7. PubMed
Ancell CD, Phipps J, Young L. Thymosin alpha-1. Am J Health Syst Pharm. 2001;58(10):879–85. PubMed
Leichtling KD, Serrate SA, Sztein MB. Thymosin alpha 1 modulates the expression of high affinity interleukin-2 receptors on normal human lymphocytes. Int J Immunopharmacol. 1990;12(1):19–29. PubMed. doi:10.1016/0192-0561(90)90064-t
Bozza S, Gaziano R, Bonifazi P, et al. Thymosin alpha1 activates the TLR9/MyD88/IRF7-dependent pathway for induction of antiviral responses. Int Immunol. 2007;19(11):1261–70. PubMed. doi:10.1093/intimm/dxm097
Giacomini E, Severa M, Cruciani M, Etna MP, et al. Dual effect of thymosin α1 on human monocyte-derived dendritic cell differentiation and functional activity. Expert Opin Biol Ther. 2015;15 Suppl 1:S59–70. PubMed. doi:10.1517/14712598.2015.1019460
Mandaliti W, Nepravishta R, Pica F, et al. Thymosin α1 interacts with hyaluronic acid electrostatically by its terminal sequence LKEKK. Expert Opin Biol Ther. 2018;18(sup1):33–42. PubMed. doi:10.1080/14712598.2018.1456527

Common questions

Is thymosin alpha-1 FDA-approved?

No. It is not FDA-approved in the United States for any indication, and here it is dispensed only as a compounded medication. Outside the United States the position differs: thymalfasin, the same molecule under its international drug name, holds marketing approval in over 30 countries, principally for chronic hepatitis B.

What is the difference between thymosin alpha-1 and thymalfasin?

There is no chemical difference. Thymalfasin is the international non-proprietary name for the same 28-amino-acid peptide, and ZADAXIN is the brand it has been marketed under. Research papers use "thymosin alpha-1" or "Tα1"; regulatory documents and non-U.S. labeling use "thymalfasin".

What is the difference between thymosin alpha-1 and thymosin beta-4?

They share a name, a thymic origin and a discovery lineage — both were characterized out of thymosin fraction 5 in the 1970s — and little else. Their sequences, structures, mechanisms and trial records are separate, and nothing established about one transfers to the other. TB-500 is a third thing again: a synthetic fragment related to thymosin beta-4, not the molecule itself.

How is thymosin alpha-1 given?

By subcutaneous injection in every clinical trial on record; it is a peptide and is not taken by mouth. The regimen in the hepatitis indications approved outside the United States is 1.6 mg twice weekly, the sepsis trial used 1.6 mg every 12 hours for seven days, and the trial literature spans roughly 1 to 16 mg per dose.

How long does it stay in the body?

Not long. In nine healthy volunteers given 900 µg/m² subcutaneously, the half-life was 1.9 to 3 hours, peak levels arrived at 1 to 2 hours, and no accumulation was detected over five days of dosing. A published pharmacy review corroborates a half-life of about two hours, with levels back at baseline by 24 hours.

What are the most common side effects?

Local injection-site reactions — irritation, redness, discomfort — are the most frequently reported effects across the trial literature. In the 1,089-patient sepsis trial, overall adverse-event rates were 66.4% on the peptide versus 67.6% on placebo (P=0.70), and serious adverse events 26.8% versus 29.3% (P=0.38).

Who should avoid it?

Anyone with known hypersensitivity to the peptide or its components, and anyone who is deliberately immunosuppressed — organ- and stem-cell-transplant recipients in particular, where fatal immune haemolytic anaemia and engraftment failure have been reported. Non-U.S. labeling also warns for children, pregnancy and lactation, autoimmune disease, and immunosuppressed patients generally.

Can thymosin alpha-1 cause immune overactivation?

The laboratory evidence describes a bidirectional molecule rather than a one-way stimulant. It induces a tolerance-promoting arm involving IDO and regulatory T cells alongside an inflammatory arm, and under bacterial-type Toll-like receptor stimulation it lowered measured parameters in human dendritic cells rather than raising them. No trial has reported an overactivation syndrome.

Does it help with COVID-19 or long COVID?

No published randomized trial of the peptide as a COVID-19 treatment has reported results. The observational data conflict: a propensity-matched analysis of 771 critically ill patients found 51.0% versus 52.9% mortality, no difference, and the largest dataset of 2,282 patients found higher adjusted odds of non-recovery, OR 1.5 (1.1–2.1), P=0.028. For long COVID there is no evidence at all.