Tigris — Polymyxin B Haemoadsorption in Endotoxic Septic Shock

1. Publication

• Title: Polymyxin B haemoadsorption in endotoxic septic shock (Tigris): a multicentre, open-label, Bayesian, randomised, controlled, phase 3 trial
• Acronym: Tigris
• Year & Journal: Lancet Respiratory Medicine, epublished March 23, 2026 (print: 2026;14(5):443-452)
• Citation: Neyra JA, Legrand M, Tidswell MA, et al. Polymyxin B haemoadsorption in endotoxic septic shock (Tigris): a multicentre, open-label, Bayesian, randomised, controlled, phase 3 trial. Lancet Respir Med. 2026;14(5):443-452. doi:10.1016/S2213-2600(26)00047-0

2. Context & Rationale

Background: Endotoxic septic shock (high endotoxin activity + multiorgan failure) carries especially high mortality. Polymyxin B haemoadsorption (removes circulating endotoxin) had mixed prior evidence: EUPHAS suggested benefit, ABDOMIX and EUPHRATES were neutral. A network meta-analysis suggested possible benefit (RR 0.70, 95% CI 0.58–0.86) but with substantial heterogeneity. Post-hoc EUPHRATES analysis identified a biologically plausible subgroup (endotoxin activity assay [EAA] 0.60–0.89) that might benefit — a hypothesis no prior trial had prospectively tested.

Research Question/Hypothesis: In adults with septic shock, multiorgan dysfunction, and EAA 0.60–0.89, does polymyxin B haemoadsorption plus standard of care reduce mortality compared with standard of care alone?

Why This Matters: No specific therapy targeting endotoxin removal is currently approved in the US for this population; if effective, this represents a targeted, biomarker-selected intervention for a subset of septic shock with mortality approaching 40–45%.

3. Design & Methods

• Study Type: Open-label, randomized, controlled, Bayesian, phase 3 trial
• Setting & Centers: 19 US hospitals; enrolled September 30, 2019 – April 10, 2025
• Population:
• Inclusion: Adults ≥18y; septic shock requiring vasopressors; multiple organ dysfunction; EAA between 0.60 and 0.89 units
• Exclusions: Not detailed in available trial text
• Intervention: Two sessions of polymyxin B haemoadsorption via hemodialysis, blood flow 80–120 mL/min for 90–120 min/session, 22h apart, plus standard of care
• Comparator: Standard of care alone
• Randomization: 2:1 (polymyxin B:control), block randomization stratified by site
• Blinding: Open-label for treating clinicians; study personnel assessing certain outcomes were masked to allocation
• Statistical Power & Follow-Up: Bayesian framework with informative prior for treatment effect; primary outcome 28-day mortality in the intention-to-treat cohort; safety cohort included all patients exposed to any study treatment. Follow-up extended to 90 days, with a subsequent 12-month follow-up reported separately.

4. Key Results

Enrollment: 14,890 patients screened; 157 enrolled (106 polymyxin B, 51 control); 66 (42%) women, 91 (58%) men.

Subgroup analyses/signals: Not reported in trial text beyond the EAA-selected population itself, which was the trial's core predefined enrichment strategy.

5. Internal Validity Assessment

• Randomization & Allocation: Block randomization stratified by site; 2:1 allocation ratio (unequal, common in device trials to gather more treatment-arm safety data).
• Protocol Adherence & Separation: Not fully detailed in available text; two-session protocol appears to have been standardized.
• Blinding & Detection Bias: Open-label for treating clinicians introduces performance-bias risk for mortality-adjacent care decisions; masking of some personnel to allocation partially mitigates detection bias.
• Missing Data & Sensitivity Analyses: Not reported in trial text.
• Overall Internal Validity Conclusion: Moderate — small sample size (n=157) despite a large Bayesian posterior probability of benefit; a Bayesian "probability of benefit" is not equivalent to a conventional frequentist significance threshold and CrIs for 28-day mortality (0.39–1.08) cross 1.0.

6. External Validity Assessment

• Population Representativeness: Restricted to a biomarker-defined (EAA 0.60–0.89) subset of septic shock with multiorgan dysfunction — not generalizable to septic shock broadly, only to this specific endotoxin-activity window.
• Practice Context: Requires EAA testing and hemoperfusion capability — not universally available outside larger, resourced US hospitals; single-country (US) trial.
• Overall External Validity Conclusion: Limited-Moderate — applicable only where EAA testing and polymyxin B hemoperfusion are both available, and only to the specific EAA-defined population studied.

7. Strengths & Limitations

Strengths:

• First trial to prospectively test the EAA 0.60–0.89 subgroup hypothesized (post-hoc) from EUPHRATES to benefit
• Bayesian design allows probabilistic interpretation of a modestly sized trial
• Mortality benefit persisted and strengthened from 28 to 90 days

Limitations:

• Small sample (n=157) relative to the number needed to resolve a therapy with a two-decade history of neutral/mixed trials
• Open-label design
• Increased serious adverse events numerically in the treatment arm (30% vs 22%)
• Bayesian "probability of benefit" framing can be less intuitive/more prone to overinterpretation than a frequentist CI/p-value, especially when the 95% credible interval for the primary 28-day outcome crosses 1.0

8. Interpretation & Practice Impact

• Clinical Implications: Supports considering polymyxin B haemoadsorption specifically in septic shock patients with EAA 0.60–0.89 and multiorgan dysfunction, pending EAA availability — not a general septic shock intervention.
• Mechanistic Coherence: Consistent with the biological rationale that a specific "intermediate" endotoxin activity window may represent patients most likely to benefit from endotoxin removal (neither too low to matter, nor too high/refractory).
• Systems-Level Takeaway: Implementation would require EAA testing infrastructure and hemoperfusion capability — a meaningful resource and workflow commitment for a biomarker-selected minority of septic shock patients.

9. Controversies & Subsequent Evidence

• Editorial Commentary/Debates: The trial addresses two decades of mixed polymyxin B evidence (EUPHAS positive, ABDOMIX/EUPHRATES neutral); Tigris represents the first purpose-built prospective test of the EAA-enrichment hypothesis rather than a repeat of prior all-comer designs.
• Guideline Integration: Not yet reflected in major international guidelines at time of this handbook's compilation; the sponsor (Spectral Medical/Vantive) has released topline 12-month follow-up data via press release (June 2026) showing 52.8% vs 66.7% mortality, which has not yet undergone independent peer review.

10. Summary & Executive Takeaway

Summary: Tigris randomized 157 patients with endotoxic septic shock (EAA 0.60–0.89, multiorgan dysfunction) 2:1 to polymyxin B haemoadsorption plus standard of care or standard of care alone. 28-day mortality was 39% vs 45% (adjusted OR 0.67, 95% CrI 0.39–1.08, 95.3% posterior probability of benefit); the mortality signal strengthened by 90 days (OR 0.54, 95% CrI 0.32–0.87, 99.4% posterior probability of benefit). Serious adverse events were numerically higher in the treatment arm.

Overall Takeaway: Tigris provides a biologically plausible, prospectively tested signal for polymyxin B haemoadsorption in a narrowly defined, EAA-selected septic shock subgroup — promising but based on a modest sample and a Bayesian credible interval that still crosses the null for the primary 28-day endpoint; independent replication and peer-reviewed long-term data are needed before broad practice change.

11. Bibliography

• Cruz DN, et al. Early use of polymyxin B hemoperfusion in abdominal septic shock (EUPHAS). JAMA. 2009;301(23):2445-2452.
• Payen DM, et al. Early use of polymyxin B hemoperfusion in patients with septic shock (ABDOMIX). Intensive Care Med. 2015;41(6):975-984.
• Dellinger RP, et al. Effect of targeted polymyxin B hemoperfusion on 28-day mortality (EUPHRATES). JAMA. 2018;320(14):1455-1463.