ImmunoSep — Precision Immunotherapy to Improve Sepsis Outcomes

1. Publication

• Title: Precision Immunotherapy to Improve Sepsis Outcomes: The ImmunoSep Randomized Clinical Trial
• Acronym: ImmunoSep
• Year & Journal: JAMA, epublished December 8, 2025 (print: 2026;335(9):775-786)
• Citation: Giamarellos-Bourboulis EJ, Kotsaki A, Kotsamidi I, et al. Precision Immunotherapy to Improve Sepsis Outcomes: The ImmunoSep Randomized Clinical Trial. JAMA. 2026;335(9):775-786. doi:10.1001/jama.2025.24175

2. Context & Rationale

Background: Sepsis is increasingly recognized as a heterogeneous syndrome with divergent immune responses ranging from hyperinflammation to immunosuppression, a variability widely blamed for repeated failures of "one-size-fits-all" immunotherapies. Prior trials assumed a uniform trajectory (early hyperinflammation progressing to later immunoparalysis), but this had not been rigorously tested as a basis for biomarker-guided, phenotype-specific treatment selection.

Research Question/Hypothesis: In patients with sepsis due to pneumonia or bacteremia, does precision immunotherapy — guided by the presence of macrophage activation-like syndrome (MALS) or sepsis-induced immunoparalysis — improve organ dysfunction (SOFA score decrease ≥1.4 points) by day 9, compared with placebo?

Why This Matters: Tests whether biomarker-guided immune-phenotyping, rather than a uniform immunotherapy approach, can finally succeed where decades of unselected sepsis immunotherapy trials have failed.

3. Design & Methods

• Study Type: International, randomized, double-blind, double-dummy, placebo-controlled, phase 2b trial
• Setting & Centers: 33 ICUs, 6 European countries; enrolled 2021–2024
• Population:
• Inclusion: Adults meeting Sepsis-3 criteria with pneumonia or bacteremia as infection source
• Stratification: Ferritin >4,420 ng/mL and <5,000 HLA-DR receptors/monocyte used to classify patients into MALS (20.0% of cohort), immunoparalysis (42.9%), or intermediate (37.1%) — unstratified/intermediate patients received placebo only
• Mortality by phenotype at baseline: MALS 79.1%, immunoparalysis 66.9%, intermediate 41.6% — confirming phenotype itself carries strong prognostic information
• Intervention: Anakinra (IL-1 receptor antagonist) for MALS patients; recombinant human interferon gamma for immunoparalysis patients (double-dummy design so both study drugs/placebos were matched)
• Comparator: Matched placebo
• Randomization: 276 patients total across the phenotype-stratified design
• Blinding: Double-blind, double-dummy
• Statistical Power & Follow-Up: Primary outcome: SOFA score decrease ≥1.4 points from baseline to day 9.

4. Key Results

276 patients enrolled and phenotyped.

Note on data completeness: Full mortality data and the exact effect-size statistic (OR/RR with CI) for the primary outcome were not available in the accessible source text beyond the percentages and p-value above.

5. Internal Validity Assessment

• Randomization & Allocation: International, multicenter randomization; double-dummy design (matched placebos for both active agents) is a genuine methodological strength allowing two different immunotherapies to be tested within one blinded framework.
• Protocol Adherence & Separation: Biomarker-based phenotyping (ferritin, HLA-DR monocyte expression) was used prospectively to allocate treatment, not just post-hoc — a key design strength distinguishing this from prior "one-size-fits-all" immunotherapy trials.
• Blinding & Detection Bias: Full double-blind, double-dummy design minimizes both performance and detection bias.
• Missing Data & Sensitivity Analyses: Not detailed in available trial text.
• Overall Internal Validity Conclusion: Moderate-to-strong — rigorous double-blind, double-dummy, biomarker-stratified design with a statistically significant, clinically meaningful primary result (35.1% vs 17.9%, P=0.002); as a phase 2b trial, definitive confirmation in a larger phase 3 trial would be needed before this changes standard practice.

6. External Validity Assessment

• Population Representativeness: International (6-country) European ICU population with sepsis specifically due to pneumonia or bacteremia; phenotype distribution (20% MALS, 43% immunoparalysis, 37% intermediate) suggests roughly a quarter to two-thirds of sepsis patients could potentially be candidates for this approach.
• Practice Context: Requires biomarker testing (ferritin, HLA-DR monocyte expression via flow cytometry) not universally available in all ICUs — a meaningful implementation barrier for widespread adoption.
• Overall External Validity Conclusion: Moderate — applicable to ICUs with access to the specific immune-phenotyping biomarkers used; the intermediate-phenotype group (37.1% of patients, receiving placebo only in this trial) remains an open question for future targeted therapy.

7. Strengths & Limitations

Strengths:

• First trial to prospectively use biomarker-guided immune phenotyping to select an immunotherapy for each patient, addressing the core theoretical failure point of decades of prior unselected sepsis immunotherapy trials
• Double-blind, double-dummy design across two different active agents
• Statistically significant, clinically meaningful primary result
• No safety concerns raised
• Investigators note the finding overturns the traditional assumption that all patients progress uniformly from hyperinflammation to immunoparalysis — a genuinely novel mechanistic insight

Limitations:

• Phase 2b trial (not yet phase 3 confirmatory)
• Full mortality and complete effect-size statistics not accessible for this summary
• Requires specialized biomarker testing infrastructure
• Intermediate-phenotype patients (37.1%) were not offered an active intervention in this trial design

8. Interpretation & Practice Impact

• Clinical Implications: Supports further development and phase 3 testing of biomarker-guided precision immunotherapy in sepsis — a genuinely novel paradigm shift from unselected immunomodulation toward phenotype-matched treatment.
• Mechanistic Coherence: The finding that immunoparalysis and hyperinflammation can appear as separate entities from early illness onset (rather than a uniform progression) is described by the lead investigator as the trial's most important teaching, and may help explain contradictory results across prior corticosteroid-in-pneumonia trials.
• Systems-Level Takeaway: If confirmed in larger trials, could justify ICU investment in rapid immune-phenotyping biomarker platforms as a companion diagnostic for precision sepsis immunotherapy.

9. Controversies & Subsequent Evidence

• Editorial Commentary: Coverage by Angus DC (MedPage Today) framed the trial as demonstrating that "adding precision immunotherapy improved sepsis outcomes" — a notable departure from the field's historical pattern of negative unselected immunotherapy trials.
• Guideline Integration: As a phase 2b trial, not yet incorporated into major sepsis guidelines; represents an important proof-of-concept requiring phase 3 confirmation before practice change.

10. Summary & Executive Takeaway

Summary: ImmunoSep, an international double-blind, double-dummy, placebo-controlled phase 2b trial, enrolled 276 sepsis patients (pneumonia/bacteremia) across 33 European ICUs, stratifying them by immune phenotype (MALS or immunoparalysis) and administering phenotype-matched immunotherapy (anakinra or interferon gamma) or placebo. The primary outcome (SOFA decrease ≥1.4 points by day 9) was achieved in 35.1% vs 17.9% (P=0.002), without safety concerns.

Overall Takeaway: ImmunoSep provides the first robust evidence that biomarker-guided, phenotype-matched immunotherapy can improve organ dysfunction in sepsis — overturning the traditional assumption of a uniform hyperinflammation-to-immunoparalysis trajectory and offering a genuinely new precision-medicine paradigm for sepsis immunotherapy, pending phase 3 confirmation.

11. Bibliography

• Giamarellos-Bourboulis EJ, et al. Personalized Immunotherapy in Sepsis (ImmunoSep) Protocol. BMJ Open. 2022;12(12):e067251.