A New Thymosin Alpha-1 Study Links It to Lower Sepsis Deaths. The Largest Randomised Trial Found No Difference.

Thymosin alpha-1 is a 28-amino-acid peptide first isolated from the thymus, the gland where immune T cells mature. It has been studied for decades as an immune modulator, and one of the places that research has gone furthest is sepsis: the life-threatening state in which the body's response to an infection starts damaging its own organs.

A study that entered PubMed this autumn reports that thymosin alpha-1 use was associated with substantially lower deaths in one group of sepsis patients. Read on its own, it sounds like a strong result. Read next to the largest randomised trial on the same question, it shows exactly why study design matters.

The new study: hospital records from 327 patients

Writing in BMC Infectious Diseases, critical-care doctors from Chongqing, China, looked back through the records of adults admitted to intensive care with sepsis caused by tuberculosis between 2017 and 2025. Of 327 patients, 160 had received thymosin alpha-1 at some point during their stay as part of their hospital care, and 167 had not.

After adjusting for differences between the groups at admission, thymosin alpha-1 use was associated with a lower risk of death within 28 days, with an adjusted hazard ratio of 0.54. A hazard ratio below 1 means the event happened less often in that group; 0.54 corresponds to roughly half the rate. Sensitivity analyses pointed the same way. An exploratory analysis suggested the association varied with patients' starting lymphocyte counts (a type of white blood cell), which the authors themselves describe as hypothesis-generating.

The authors' conclusion is careful: thymosin alpha-1 use was associated with lower observed mortality, but causality cannot be inferred, and the lymphocyte pattern needs external validation.

Why "associated with" is doing a lot of work

This was a retrospective cohort study. Nobody decided at random who would receive thymosin alpha-1; doctors chose, for reasons that were recorded imperfectly or not at all. Patients judged to be likely candidates, or treated by particular teams, or admitted in particular years, may have differed in ways that statistical adjustment cannot fully remove. This is called confounding, and it is the main reason observational studies can show large effects that disappear in controlled trials.

The study also comes from a single hospital and a specific subgroup, sepsis due to tuberculosis, which limits how far it can be generalised.

The trial: 1,106 patients, blinded, no difference

The question has already been tested the stronger way. The TESTS trial, published in The BMJ in January 2025, enrolled 1,106 adults with sepsis at 22 hospitals in China and randomly assigned half to thymosin alpha-1 and half to a placebo for seven days. Neither patients nor investigators knew who received which. This is the design most able to separate the effect of a compound from everything else going on.

Death within 28 days occurred in 23.4% of the thymosin alpha-1 group and 24.1% of the placebo group, a hazard ratio of 0.99. No secondary or safety outcome differed significantly between the groups. The authors concluded that the trial found no clear evidence that thymosin alpha-1 reduces 28-day mortality in adults with sepsis.

Pre-planned subgroup analyses hinted at differences by age and by diabetes, in opposite directions for younger and older patients. Subgroup findings in a trial that is neutral overall are a standard source of false leads, and the authors present them as a potential differential effect, not a finding.

What the pooled trials show

A systematic review and meta-analysis in Frontiers in Cellular and Infection Microbiology (September 2025) pooled 11 randomised trials with 1,927 patients. Across all of them, thymosin alpha-1 was associated with lower 28-day mortality. But when the reviewers restricted the analysis to the higher-quality trials, or to multi-centre trials, the difference was no longer statistically significant. A trial sequential analysis, a check on whether enough patients have been studied to draw a firm conclusion, found the total sample still inadequate.

That pattern, a signal in the full pool that fades in the most rigorous studies, is common, and it is a reason for caution rather than confidence.

How to weigh the two

	New cohort study	TESTS trial
Design	Retrospective records review	Randomised, double-blind, placebo-controlled
Patients	327, one hospital	1,106, 22 hospitals
Population	Sepsis from tuberculosis	Sepsis from any cause
Result	Association with lower 28-day deaths	No difference in 28-day deaths

The two do not strictly contradict each other: one could argue that tuberculosis-related sepsis is a distinct group. But that is a hypothesis the new study raises, not one it can confirm. The randomised evidence currently carries more weight.

What has not been established

• No randomised trial specifically in tuberculosis-related sepsis appears in these papers.
• Whether any subgroup, by age, diabetes or lymphocyte count, responds differently has not been confirmed in a trial designed to test it.
• These are hospital studies of intensive-care patients under medical supervision. They say nothing about use outside that setting.

What to watch

The meta-analysis authors call for trials that test thymosin alpha-1 in predefined patient groups. A registered trial in tuberculosis-related sepsis, with randomisation and blinding, is the result that would move this question forward.

Sources

1. Wang T, et al. Association between thymosin α1 therapy and 28-day mortality risk in critically ill patients with tuberculosis-related sepsis: a retrospective cohort study. BMC Infectious Diseases (2026)
2. Wu J, et al. The efficacy and safety of thymosin α1 for sepsis (TESTS): multicentre, double blinded, randomised, placebo controlled, phase 3 trial. BMJ (2025)
3. Gu B, et al. Efficacy of thymosin α1 for sepsis: a systematic review and meta-analysis of randomized controlled trials. Frontiers in Cellular and Infection Microbiology (2025)

Compounds in this article

Thymosin Alpha-1