Thymosin alpha-1 got the large, blinded trial. FDA's advisers still voted 17 to 4 against it.

Why this note exists

Thymosin alpha-1 sits in an unusual place in the research market. It is not an obscure sequence with a rodent paper behind it: it was isolated from calf thymus in 1977, it has been made synthetically as thymalfasin for decades, it is sold as a prescription medicine in a long list of countries, and it has accumulated a real clinical literature across a dozen indications.[1] That is precisely why it is worth auditing carefully. When a compound has this much human data, the honest question stops being "is there any evidence?" and becomes "what did the evidence say?"

In the index generated on August 15, 2026, 75 of 112 vendors carried a thymosin alpha-1 listing — 103 listings, nineteenth of 83 compounds by vendor coverage.[17] Where a size is stated, 10 mg is the common vial (55 of the 103), followed by 5 mg (22).[17] Not one of the 103 listings uses the word thymalfasin or the brand the trials were run on.[17]

A note on what kind of evidence this is: every efficacy and safety figure below is human data, with its design and participant count attached in the same sentence. One paragraph reports animal and in vitro toxicology and is labeled as such. Where a result comes from a single arm, a retrospective chart review or a case series, it is called that. pepmg does not convert a published study dose into a protocol for a research vial.

The regulatory status

Approved in a lot of places. Not here, and FDA could not verify the list.

Start with the plain facts, because the marketing around this peptide leans on a genuine ambiguity. FDA's own presentation states that "FDA has approved no drug products containing Ta1" — written specifically to correct a website FDA found advertising a compounded subcutaneous injectable as "FDA-approved."[1] A query of Drugs@FDA and FDA's label database on August 17, 2026 returned no matching application and no prescribing information for thymalfasin or thymosin alpha-1.[16]

Abroad, the picture is real but softer than it sounds. FDA records that according to the manufacturer's 2014 annual report the peptide is "approved" for use in countries in the Asia-Pacific region, Latin America, Eastern Europe and the Middle East — and then adds, in the same slide, that "FDA is unable to independently verify these claims of approval in all the specified countries."[1] FDA further states it is "not approved in the United States, Japan, or Europe (except Italy)" and "not recognized in the European or Japanese Pharmacopoeias."[1]

The compounding thread is the one that generated headlines, and it is routinely read backwards. Thymosin alpha-1 was nominated for the 503A Bulks List — the list of bulk substances a pharmacy may compound with under section 503A — with the proposed product being a 3 mg/mL injection for subcutaneous administration.[1][5] That nomination was later withdrawn. FDA's slide is one sentence: "The nomination was withdrawn, and FDA is evaluating the substances at its discretion."[1]

A withdrawn nomination is a procedural event, not a verdict, and FDA's public page reflects that. As of a query on August 17, 2026, thymosin alpha-1 still appears there under bulk drug substances nominated but withdrawn — substances "previously in category 2 of the interim policies" — with FDA's safety language kept intact beside it: compounded drugs containing it "may pose significant risk for immunogenicity for certain routes of administration and may have complexities with regard to peptide-related impurities and API characterization," and "[t]he safety-related information is inadequate for the agency to sufficiently understand the extent of any safety issues raised by the proposed compounded drug."[4]

The vote

Four in favour, seventeen against — twice

On December 4, 2024 the Pharmacy Compounding Advisory Committee took up three peptide families in one sitting: CJC-1295, AOD-9604, and thymosin alpha-1.[2][3] The committee first voted 20 to 1 to handle the two thymosin alpha-1 substances as a group, then split them anyway and voted on each.[2]

Vote results and the roster of evaluated uses as recorded in FDA's final summary minutes and its evaluation presentation.[1][2]

The minutes are worth quoting because they show a committee that was not unanimous and did not pretend to be. On the free base: "The majority of the Committee members voted against placing Thymosin alpha-1 (free base) on the 503A Bulks List. Several Committee members who voted 'No' agreed there was no compelling evidence demonstrating clinical effectiveness and safety in the uses of Thymosin alpha-1 (free base) under review. Some Committee members who voted in favor of placing Thymosin alpha-1 (free base) on the 503A Bulks List commented that this substance should be accessible to patients as an option for use."[2] On the acetate, the minutes record the majority voted against "due to the lack of convincing effectiveness data."[2]

FDA's own recommendation, in its closing slide, was that "a balancing of the four evaluation criteria weighs against Ta1 (free base) and Ta1 acetate being added to the 503A Bulks List" — the four criteria being physicochemical characterization, historical use in compounding, safety, and evidence of effectiveness.[1] As with every PCAC vote, this is advice. It is not an approval, not a ban, and not a final rule.[5]

The pivotal evidence

Three sepsis trials, and the biggest one is the flattest

Sepsis is where thymosin alpha-1 has its most serious human evidence, and it is also where the trajectory is clearest: as the trials got larger and better blinded, the effect got smaller.

Enrollment and results as reported in each publication and, for the 2022 trial, as summarized in FDA's evaluation. Neither of the concomitant therapies in the 2022 trial is standard US practice: FDA notes ulinastatin is not approved in the United States and blood purification is not a standard therapy for sepsis.[1][6][9]

TESTS is the one to read closely. It enrolled 1,106 adults aged 18 to 85 meeting sepsis-3 criteria at 22 centres in China between September 2016 and December 2020, randomised 1:1, stratified by age and centre, with the registry recording quadruple masking — participant, care provider, investigator and outcomes assessor.[6][8] The intervention was a subcutaneous injection of thymosin alpha-1 1.6 mg or matched placebo every 12 hours for up to seven days.[6][8] The primary outcome was 28-day all-cause mortality on a modified intention-to-treat set of 1,089.[6]

The result was 23.4% against 24.1%, hazard ratio 0.99 (0.77 to 1.27), P=0.93 by log-rank, and the authors' stated conclusion is that the trial "found no clear evidence to suggest that thymosin α1 decreases 28 day all cause mortality in adults with sepsis."[6] Two prespecified subgroup interactions were reported and are the part most likely to be quoted out of context: by age, the hazard ratio was 1.67 (1.04 to 2.67) under 60 and 0.81 (0.61 to 1.09) at 60 and over, P for interaction 0.01; by diabetes, 0.58 (0.35 to 0.99) with diabetes and 1.16 (0.87 to 1.53) without, P for interaction 0.04.[6] A subgroup interaction inside a null trial is a hypothesis, not a finding, and the direction of the age interaction points the wrong way for most of the people buying this peptide.

Two disclosures belong with the number. The authors declare support from Sun Yat-sen University's clinical research programme, the Guangdong Clinical Research Center for Critical Care Medicine, and SciClone Pharmaceuticals — the company behind the branded product — with several authors reporting grants from SciClone and consultancy fees from SciClone and others.[6] And the paper carries a published correction.[7] Neither undoes the result; a sponsor-supported trial that reports a null is, if anything, the more credible kind of null.

Read the subgroups before you read the meta-analysis headline

A 2025 systematic review and meta-analysis in Frontiers in Cellular and Infection Microbiology, from the same Sun Yat-sen group, pooled 11 randomized trials — 967 patients on thymosin alpha-1 against 960 controls — and is the source most likely to be cited as vindication.[10] Its headline number is a significant reduction in 28-day mortality, odds ratio 0.73 (0.59 to 0.90), P=0.003.[10]

The same abstract then takes it back, and this is the sentence that matters: the high-quality subgroup gave OR 0.82 (0.65 to 1.03, P=0.09) and the multi-centre subgroup gave OR 0.86 (0.68 to 1.08, P=0.20) — neither statistically significant — while a trial sequential analysis concluded that "the current sample size is inadequate."[10] The heterogeneity analysis flagged possible benefit in cancer patients at moderate credibility and in diabetes and coronary heart disease at low credibility, and the authors' own conclusion is framed as potential rather than demonstrated.[10] In plain terms: the pooled benefit lives in the smaller, single-centre, lower-quality trials and disappears in the better ones.

A separate 2025 meta-analysis, in severe acute pancreatitis, shows what the honest positive version of this literature looks like. Five randomized trials, 706 patients: thymosin alpha-1 raised CD4+ percentages and the CD4+/CD8+ ratio, and reduced the overall incidence of extrapancreatic infections (RR 0.56, 0.40 to 0.78, P=0.0005), while not significantly reducing hospital stay (MD −4.22 days, −11.53 to 3.10, P=0.26).[11] Immune markers moved. An infection endpoint moved. Length of stay did not, and mortality was not the endpoint.[11]

The sweep

What FDA found across the other eleven uses

FDA reviewed twelve indications and reached a "lack of evidence" or "insufficient evidence" conclusion on each one it examined, summarizing that "[n]one of the clinical practice guidelines for U.S. health professionals recommend Ta1 (free base) or Ta1 acetate" and that the studies in the serious conditions considered "were inconclusive and limited by small sample sizes and design deficiencies."[1] Four of those files are worth naming because they are the ones the market cites.

Hepatitis B. Five monotherapy studies from 1998 to 2005 reported mixed results; the one randomized, double-blind, placebo-controlled study among them found no difference in undetectable HBV DNA six months after treatment — 20% on thymosin alpha-1 against 21% on placebo.[1] In a later randomized open-label trial adding it to entecavir, undetectable HBV DNA was 64.6% versus 69.6% at week 52 and 87.7% versus 85.5% at week 104, and the authors concluded the combination had "a similar effect as entecavir monotherapy."[1]

Malignant melanoma. The largest oncology trial FDA identified was an exploratory, multicentre, open-label randomized study in 488 patients with stage IV melanoma, comparing dacarbazine-based regimens with and without thymosin alpha-1 at three doses. FDA's summary is one word long: it "[f]ailed to show a significant difference in the ORR with any Ta1 containing regimens compared to the control group."[1]

COVID-19. FDA identified four meta-analyses and reports that three of the four concluded there was no decrease in mortality in patients treated with thymosin alpha-1; it also notes that IDSA, NIH and WHO treatment guidelines do not discuss the drug for COVID-19, and that it could find no literature on use in children or in adult outpatients.[1]

Stem cell transplant, and the two uses with no data at all. In transplant recipients the picture includes a signal in the wrong direction: FDA cites a 2013 case series of eight recipients in which the infection rate increased in the treated subjects, and concludes the studies "do not provide evidence that Ta1 reduces infections and/or infection related mortality after HSCT."[1] For myalgic encephalomyelitis and chronic fatigue syndrome — a use the peptide is marketed for — FDA's finding is that it "did not identify any clinical studies using Ta1 in subjects with ME/CFS."[1] On use as an influenza vaccine adjuvant, FDA found five studies and noted that four of the five measured immune response by ELISA, "which is not a measure of clinical effectiveness."[1]

Safety

Well tolerated in the studies that have been done, with two specific warnings and one open question

The tolerability record is genuinely unremarkable, and saying so is part of reporting it accurately. FDA's overall conclusion is that "[i]n most clinical studies, Ta1 has not been associated with significant adverse events attributable to Ta1 when administered in doses in the range of 1-16 mg via the SC ROA for up to 12 months," with the most common reactions being local irritation, redness or discomfort at the injection site.[1] FDA identified a single FAERS report — a 46-year-old man on peginterferon alfa-2a and thymosin alpha-1 in a clinical trial, hospitalised with anxiety, atrial fibrillation and a transient TSH decrease — and notes causality is confounded because those are labeled adverse events for the interferon.[1]

Two warnings are specific enough to be worth carrying. First, the immunosuppressed: FDA raises "[p]otential safety concerns with administering Ta1 in patients who are undergoing deliberate immunosuppression," noting it "could cause or worsen acute or chronic graft-vs-host disease and lead to engraftment failure" in transplant recipients, and elsewhere cites a report of fatal immune haemolytic anaemia and engraftment failure in that population.[1] Second, foreign labels: products marketed outside the United States carry warnings and contraindications for use in children, in pregnant and lactating women, in patients with autoimmune diseases, and in immunosuppressed populations.[1]

The open question is immunogenicity, and it is a chemistry question as much as a clinical one. Thymosin alpha-1 is a 28-amino-acid peptide given subcutaneously; FDA writes that it "may pose a significant risk for immunogenicity, potentially amplified by aggregation as well as potential peptide-related impurities," that the nomination "did not include, and FDA is not aware of, information about Ta1 to suggest that this substance does not present these risks," and — the concrete version — that the highest strength identified to date in clinical studies was 2 mg/mL, while the nominated product was 3 mg/mL, with no human data supporting the higher strength.[1]

Animal and in vitro data, labeled as such: summaries in the manufacturer's product monograph describe no drug-related safety signals from single subcutaneous doses up to 20 mg/kg in mice, rats and marmosets, or from repeat dosing up to 6 mg/kg/day for 13 weeks or 1 mg/kg/day for 26 weeks, and no genotoxicity signals in in vivo and in vitro assays.[1] FDA adds two caveats in the same section: the underlying nonclinical data are not included in the monograph for review, and carcinogenicity studies "were not identified."[1] None of that is human data.

The market

A 1.6 mg study dose, a 10 mg vial, and a solubility ceiling

The gap between the studied article and the sold article is smaller here than for most compounds in this index — both are lyophilized peptide for subcutaneous injection — but it is not nothing, and the interesting part is the chemistry.

FDA's characterization section states that thymosin alpha-1 free base is soluble in water up to 2 mg/mL and the acetate salt up to 1 mg/mL, and concludes flatly that "[i]t is difficult to compound Ta1 acetate when reconstituting it using water as the solvent due to its limited water solubility."[1] On the nominated 3 mg/mL injectable it notes the proposed concentration "is greater than the solubility of Ta1 (free base) in water (2 mg/mL). No information was provided regarding how a greater solubility is achieved."[1] Set that against the vials on sale: 55 of the 103 listings in the index are 10 mg.[17] pepmg does not publish reconstitution or administration technique and is not doing so here; the point is only that a published solubility figure is a fact about the substance, and it is one FDA thought was decision-relevant.

Characterization is the other half. FDA concluded that both forms are "not well-characterized," citing a lack of critical data on impurities, aggregates and bioburden or bacterial endotoxin levels, no USP drug substance monograph, no certificate of analysis for the free base in the nomination, and a literature search that "could not find other CoAs nor other information to evaluate potential single impurity and total impurity limits."[1] Storage is unforgiving too: the free base is recommended stored desiccated below −18 °C, and in solution is described as stable for 2 to 7 days at 4 °C.[1]

Finally, the naming. The molecule FDA reviewed has a CAS number (62304-98-7), a molecular weight of 3108.3, and a synthetic name — thymalfasin — that identifies it as identical in sequence to the natural peptide.[1] None of the 103 listings in pepmg's index use that name.[17] That is not an accusation about anyone's vial; pepmg makes no claim about the contents of any vendor's product. It is an observation about search: a buyer looking up "thymosin alpha-1" and a clinician looking up "thymalfasin" are looking at the same peptide and will not find the same literature.

Where the field is going, briefly

This is an active programme, not a closed file. A ClinicalTrials.gov search on August 17, 2026 returned 63 registered studies of thymalfasin, 61 of them interventional, with 19 currently recruiting, not yet recruiting, or active and not recruiting.[13] Nearly all of the recent activity is oncology, and nearly all of it is in China.[13]

Two are worth watching for different reasons. The largest is a phase 3 trial of thymosin alpha-1 as adjuvant treatment after radical resection of high-risk stage II and stage III colorectal cancer, randomized with a blinded outcomes assessor, estimated enrollment 2,500, primary outcome 3-year disease-free survival, primary completion estimated March 2027.[14] If that reads out positive it will be the strongest result this compound has ever had; it is also open-label to patients and investigators, which will matter when it is read.[14] The other is small and American: a randomized, open-label phase 1 study at a US research institute, estimated enrollment 75, testing thymosin alpha-1 before a COVID-19 booster in older adults, currently recruiting, estimated completion December 2026.[15]

The published 2026 work so far is early-phase and uncontrolled. A prospective phase 2 trial published in BMC Medicine in February 2026 added thymalfasin to neoadjuvant anti-PD-1 plus chemotherapy in 30 patients with stage III gastric or gastro-oesophageal junction adenocarcinoma and reported a 30.0% pathological complete response rate and 56.7% major pathological response at a median 14 months' follow-up.[12] Thirty patients, single arm, no control group — which means the trial cannot separate the peptide's contribution from the chemotherapy and the checkpoint inhibitor it was given with, and its authors frame their immune findings as a hypothesis.[12]

Three things this note is not saying

First, it is not saying thymosin alpha-1 does nothing. It has a described mechanism at Toll-like receptor 9 on dendritic and lymphoid progenitor cells, it reproducibly moves immune markers in humans — CD4+ percentages, CD4+/CD8+ ratio, monocyte HLA-DR — and in severe acute pancreatitis a pooled analysis of five randomized trials found fewer extrapancreatic infections.[1][9][11] Moving a marker is a real pharmacological effect. It is not the same claim as changing an outcome.

Second, it is not saying the advisory vote settled the science. Seventeen to four is a clear margin, but four members voted the other way and the minutes record their reasoning — that the substance should be accessible to patients as an option.[2] A PCAC vote is advice about whether pharmacies may compound with a bulk substance under section 503A. It is not a finding of fact about efficacy, and it is not FDA's final action.[5]

Third, it is not saying the trial that matters has not been run. It has, and that is the unusual thing about this file. The most common shape of a peptide argument is an absence of evidence; here there is a 1,106-patient, double-blinded, placebo-controlled, sponsor-supported phase 3 in a serious condition, and it returned a hazard ratio of 0.99 with a confidence interval that comfortably contains no effect.[6]

What this note is saying is narrower: thymosin alpha-1 is the compound in this index with the best claim to a real clinical programme, and its best-designed study is null. Everything else — twelve indications, decades of literature, approvals in countries FDA could not fully verify — has to be read against that.[1][6]

Questions people are asking