TESTS — Thymosin α1 for Sepsis

1. Publication

• Title: The efficacy and safety of thymosin α1 for sepsis (TESTS): multicentre, double blinded, randomised, placebo controlled, phase 3 trial
• Acronym: TESTS
• Year & Journal: BMJ, published January 15, 2025 (2025;388:e082583)
• Citation: Wu J, Pei F, Zhou L, et al; TESTS study collaborator group. The efficacy and safety of thymosin α1 for sepsis (TESTS): multicentre, double blinded, randomised, placebo controlled, phase 3 trial. BMJ. 2025;388:e082583. doi:10.1136/bmj-2024-082583

2. Context & Rationale

Background: Thymosin α1 (Tα1) is an immunomodulatory drug with bidirectional effects — promoting naive T cell maturation, reversing T cell exhaustion, alleviating cytokine storms, and enhancing antifungal immunity. A prior meta-analysis of 19 smaller studies (1354 patients) suggested Tα1 might benefit sepsis patients, but this evidence was limited by small trial sizes and methodological heterogeneity, warranting a large, definitive phase 3 test.

Research Question/Hypothesis: In adults with sepsis (Sepsis-3), does subcutaneous thymosin α1 reduce 28-day all-cause mortality compared with placebo?

Why This Matters: TESTS was designed as the largest, most rigorous test to date of an immunomodulatory approach that had shown promise in smaller trials but lacked definitive phase 3 confirmation — directly relevant given the broader field's repeated struggles to find effective sepsis immunotherapies (see also ImmunoSep, this handbook, for a phenotype-guided alternative approach).

3. Design & Methods

• Study Type: Multicenter, double-blind, randomized, placebo-controlled, phase 3 trial
• Setting & Centers: 22 centers, China; enrolled September 2016 – December 2020
• Population:
• Inclusion: Adults 18-85y with sepsis per Sepsis-3 criteria
• Stratification: By age (<60 vs ≥60y) and center
• Intervention: Subcutaneous thymosin α1 injection every 12 hours for 7 days (unless discontinued due to ICU discharge) (n=552)
• Comparator: Matched placebo (n=554)
• Randomization: 1:1, stratified block method; 1106 total enrolled
• Blinding: Double-blind
• Statistical Power & Follow-Up: Primary outcome: 28-day all-cause mortality. Modified ITT population: 1089 patients (542 Tα1, 547 placebo). Follow-up included 90-day SF-36 quality-of-life assessment (74.1% of survivors completed).

4. Key Results

1089 patients in modified ITT analysis (542 Tα1, 547 placebo). Median age 65y (IQR 52-73) overall.

Context from a subsequent meta-analysis (Gu et al., Frontiers 2025) pooling TESTS with other trials found a pooled HR of 0.88 (95% CI 0.72-1.08) across ~1450 patients, with heterogeneity-of-treatment-effect analysis suggesting possible subgroup benefit in septic patients with cancer (HR 0.59, 95% CI 0.37-0.94), diabetes (HR 0.64, 95% CI 0.41-0.98), or coronary heart disease (HR 0.56, 95% CI 0.31-0.99) — though these subgroup credibility ratings were graded low-to-moderate.

5. Internal Validity Assessment

• Randomization & Allocation: Stratified block randomization by age and center — sound methodology.
• Protocol Adherence & Separation: Standardized subcutaneous dosing protocol (every 12h for 7 days) with clear discontinuation criteria (ICU discharge).
• Blinding & Detection Bias: Full double-blind design minimizes both performance and detection bias — a key strength given the large sample and hard mortality endpoint.
• Missing Data & Sensitivity Analyses: Modified ITT population (1089/1106, 98.5%) reflects high analytic completeness; 90-day SF-36 completion was lower (74.1% of eligible survivors) but this was a secondary quality-of-life outcome.
• Overall Internal Validity Conclusion: Strong — large (1106-patient), double-blind, adequately powered phase 3 trial with a hard, objective primary endpoint and high analytic completeness; the clean null result (HR 0.99, tight CI) provides high-confidence evidence against a clinically important mortality benefit in the unselected sepsis population studied.

6. External Validity Assessment

• Population Representativeness: Chinese, 22-center, broad sepsis population (Sepsis-3 criteria, age 18-85y) — reasonably representative of general ICU/hospital sepsis populations in similar health systems.
• Practice Context: Subcutaneous injection is simple and broadly deployable; the drug itself (thymosin α1 / thymalfasin) has established manufacturing and distribution in China and other markets.
• Overall External Validity Conclusion: Good for general sepsis populations in similar health systems; single-country (China) design and the drug's variable international availability/regulatory status may limit direct practice transfer elsewhere.

7. Strengths & Limitations

Strengths:

• Largest, most rigorous (double-blind, phase 3) thymosin α1 sepsis trial to date, resolving prior uncertainty from smaller, more heterogeneous studies
• High analytic completeness (98.5% modified ITT)
• Broad, representative unselected sepsis population
• Clean, statistically precise null result (tight confidence interval)

Limitations:

• Long enrollment period (2016-2020) may introduce temporal sepsis-management practice changes
• As an unselected (non-biomarker-guided) immunomodulatory trial, does not address whether a phenotype-selected subgroup (as tested in ImmunoSep, this handbook) might still benefit
• Post-hoc subgroup signals (cancer, diabetes, coronary heart disease) from subsequent meta-analysis are hypothesis-generating only, not confirmatory

8. Interpretation & Practice Impact

• Clinical Implications: Does not support routine thymosin α1 use for unselected sepsis patients — the largest, most rigorous trial to date found no mortality benefit, resolving prior uncertainty from smaller studies in favor of a definitive null.
• Mechanistic Coherence: Fits within the broader pattern (also seen with ImmunoSep, this handbook) that unselected immunomodulation trials in heterogeneous sepsis populations tend toward null results, while biomarker/phenotype-guided approaches may hold more promise — the subsequent meta-analysis's subgroup signals (cancer, diabetes, coronary heart disease) hint that patient selection may matter even for Tα1, though this requires prospective confirmation.
• Systems-Level Takeaway: Reinforces that future sepsis immunomodulator trials likely need better patient selection strategies (biomarker- or comorbidity-based) rather than testing agents in unselected, heterogeneous sepsis populations.

9. Controversies & Subsequent Evidence

• Editorial Commentary/Debates: A subsequent systematic review and meta-analysis (Gu et al., Frontiers in Cellular and Infection Microbiology, 2025) pooling TESTS with 10 other smaller RCTs found an overall pooled HR of 0.88 (95% CI 0.72-1.08, still not statistically significant) but identified potential subgroup benefit in patients with cancer, diabetes, or coronary heart disease — though rated low-to-moderate credibility per ICEMAN criteria, meaning these should be considered hypothesis-generating rather than practice-changing.
• Guideline Integration: As a large, definitive null phase 3 trial, unlikely to support inclusion of thymosin α1 in major international sepsis guidelines for unselected patients.

10. Summary & Executive Takeaway

Summary: TESTS, a large double-blind phase 3 trial, randomized 1089 (modified ITT) Chinese sepsis patients across 22 centers to subcutaneous thymosin α1 or placebo for 7 days. 28-day mortality was nearly identical (23.4% vs 24.1%; HR 0.99, 95% CI 0.77-1.27, P=0.93) — a clean, statistically precise null result.

Overall Takeaway: The largest, most rigorous trial of thymosin α1 in unselected sepsis found no mortality benefit, providing high-confidence evidence against routine use — though intriguing (low-confidence) subgroup signals from subsequent meta-analysis in patients with cancer, diabetes, or coronary heart disease suggest patient selection, rather than the drug itself, may be the more important unresolved question, echoing the broader field's shift toward precision/phenotype-guided immunomodulation approaches.

11. Bibliography

• Liu F, et al. Efficacy of thymosin alpha 1 in critically ill patients with sepsis: a meta-analysis. J Crit Care. 2016.
• Gu B, Zhou Y, Nie Y, et al. Efficacy of thymosin α1 for sepsis: a systematic review and meta-analysis of randomized controlled trials. Front Cell Infect Microbiol. 2025;15:1673959.
• Wu J, Zhou L, Liu J, et al. The efficacy of thymosin alpha 1 for severe sepsis (ETASS). Crit Care. 2013;17(1):R8.