Sepsis future or investigational therapies
|
Resident Survival Guide |
|
Sepsis Microchapters |
|
Diagnosis |
|---|
|
Treatment |
|
Case Studies |
|
Sepsis future or investigational therapies On the Web |
|
American Roentgen Ray Society Images of Sepsis future or investigational therapies |
|
Risk calculators and risk factors for Sepsis future or investigational therapies |
Editor-In-Chief: C. Michael Gibson, M.S., M.D. [1]; Associate Editor(s)-In-Chief: Parth Vikram Singh, MBBS[2] Jason Le, B.S.[3]
Synonyms and keywords: sepsis syndrome; septic shock; septicemia
Sepsis future or investigational therapies
More than 100 immune-modulating trials in sepsis have failed to produce a consistently beneficial adjunctive therapy, and mortality reduction has plateaued. Biological heterogeneity is a major barrier: sepsis encompasses patients with hyperinflammation and patients with immunoparalysis who may respond differently, or oppositely, to the same intervention. Current therapeutic development therefore emphasizes precision medicine, including biologically defined endotypes or subphenotypes and predictive enrichment for host-directed therapies.[1][2]
The therapies described below remain investigational unless otherwise stated. Established sepsis management, including antimicrobials, fluids, vasopressors, corticosteroids, and source control, belongs in the Sepsis medical therapy microchapter.
Sepsis heterogeneity, endotypes, and subphenotypes
Multiple independent groups have derived blood-transcriptomic endotypes with prognostic and, in some settings, potential predictive significance:
- Sepsis Response Signatures (SRS1/SRS2): SRS1 is associated with an immunosuppressed phenotype and higher mortality.[3][2]
- MARS1-MARS4: four transcriptomic endotypes, with MARS1 characterized by immunosuppression and higher mortality.[2]
- Inflammopathic, adaptive, and coagulopathic endotypes: a framework described by Sweeney and colleagues and incorporated into subsequent consensus work.[1][2]
- Five emergency-department endotypes: NPS, INF, IHD, IFN, and ADA have been described using transcriptomic approaches.[3][2]
A 2025 consensus framework identified convergence among previously described systems, including three consensus transcriptomic subtypes and four consensus immune-dysregulation endotypes across pediatric and adult, bacterial and viral, and ICU and emergency-department cohorts.[1][4]
The clinical proof-of-concept for treatment-selective biology remains the post-hoc VANISH analysis. In that analysis, the SRS2 endotype was associated with a signal for harm from corticosteroids, whereas this signal was not observed in SRS1, illustrating the potential for treatment effects to differ by immune endotype.[5]
Biomarker-guided precision immunotherapy
The most clinically developed precision strategy is to identify the dominant immune phenotype and select a therapy accordingly.
- Macrophage activation-like syndrome (MALS): hyperferritinemia and other markers of immune dysregulation have been used to identify patients for investigation of IL-1 blockade with anakinra.[6]
- Immunoparalysis: low monocyte HLA-DR expression, impaired ex-vivo LPS-induced TNF-alpha production, and lymphopenia have been investigated as markers for therapies including interferon gamma, granulocyte-macrophage colony-stimulating factor (GM-CSF), and interleukin-7.[2][7]
ImmunoSep randomized clinical trial
The ImmunoSep randomized clinical trial was a double-blind, double-dummy, placebo-controlled trial in 276 analyzed patients across 6 countries. Patients were stratified by immune state: those with macrophage activation-like syndrome, defined by ferritin >4420 ng/mL, received intravenous anakinra, whereas those with sepsis-induced immunoparalysis, defined by ferritin <=4420 ng/mL and <5000 HLA-DR receptors per monocyte, received subcutaneous recombinant human interferon gamma. Precision immunotherapy met the primary endpoint of a decrease of at least 1.4 points in mean SOFA score by day 9 (35.1% [46/131] vs 17.9% [26/145] with placebo; difference 17.2%, 95% CI 6.8%-27.2%; P=.002), but did not significantly reduce 28-day mortality.[8]
The trial provides prospective evidence that biomarker-guided immunotherapy can improve organ dysfunction, but it does not establish a mortality benefit or routine clinical use.
Anakinra
Anakinra is being investigated for patients with hyperinflammatory/MALS phenotypes. Reanalysis of a prior phase III trial identified a mortality benefit signal in a subgroup with features of macrophage activation syndrome, including concurrent hepatobiliary dysfunction and disseminated intravascular coagulation, and benefit was also associated with high baseline IL-1 receptor antagonist concentrations.[9] These findings remain hypothesis-generating outside biomarker-selected clinical trials.
Anakinra is FDA-approved for rheumatoid arthritis, neonatal-onset multisystem inflammatory disease (NOMID), and deficiency of interleukin-1 receptor antagonist (DIRA); its use for sepsis is investigational and off-label.[10]
Complement inhibition
The complement anaphylatoxin C5a is an investigational target because of its role in neutrophil activation, endothelial injury, and coagulation. The anti-C5a monoclonal antibody vilobelimab achieved effective C5a inhibition in early sepsis studies, with signals for increased vasopressor- and ventilator-free days at higher doses.[11] Selective C5a blockade preserves downstream C5b and membrane-attack-complex formation, providing a theoretical advantage over upstream complement inhibition, but sepsis-specific phase III evidence is lacking.[12]
Immunostimulation for sepsis-induced immunoparalysis
Several therapies aim to restore immune function in patients with an immunoparalysis phenotype:
- GM-CSF: GM-CSF can increase monocyte HLA-DR expression and has shown signals for improved infection resolution in selected patients. In the GRID randomized trial, however, an HLA-DR-guided strategy was stopped early after 98 of 166 planned patients and did not reduce ICU-acquired infection or 28-day mortality.[13]
- Interferon-gamma: recombinant interferon-gamma has been investigated to restore monocyte immune function and was the immunoparalysis-directed therapy used in ImmunoSep.[7][8]
- Interleukin-7: IL-7 is antiapoptotic and expands CD4+ and CD8+ T cells. Early studies support biologic activity without clear evidence of improved clinical outcomes.[14][7]
- Anti-PD-1/PD-L1 checkpoint inhibition: checkpoint inhibitors are being investigated to reverse lymphocyte exhaustion, but clinical evidence remains limited to early safety and proof-of-concept studies.[7]
Endothelial and vascular-barrier stabilization
Vascular leak contributes to shock and organ dysfunction. Investigational approaches include Ang-1/Tie2 agonists, sphingosine-1-phosphate receptor agonists, Slit2N, and antibodies directed against HMGB1 or VEGFR2. Most remain preclinical or in early-phase development.[15][16]
Mesenchymal stromal cells
Mesenchymal stromal cells are being investigated as host-directed therapy through paracrine, immunomodulatory, antimicrobial, antiapoptotic, and reparative effects rather than durable engraftment.[17] Early phase studies in septic shock and COVID-19-associated acute respiratory distress syndrome have generally supported short-term safety, but efficacy signals are inconsistent. Multiple phase I/II studies remain investigational.[17][16]
Metabolic-resuscitation therapy and intravenous vitamin C
The vitamin C/thiamine/hydrocortisone metabolic-resuscitation approach is no longer supported as a routine investigational strategy outside research. Early observational enthusiasm was followed by negative randomized trials, including VITAMINS, VICTAS, and LOVIT.[18]
The LOVIT trial enrolled 872 adults with sepsis in the intensive care unit and found a higher risk of death or persistent organ dysfunction at 28 days among patients randomized to high-dose intravenous vitamin C (50 mg/kg every 6 hours for up to 4 days) than among those receiving placebo.[19][18]
The 2021 Surviving Sepsis Campaign issued a weak recommendation against intravenous vitamin C in sepsis.[20] The 2026 Surviving Sepsis Campaign revisited this question and again suggested against intravenous vitamin C in sepsis or septic shock (conditional recommendation, low certainty evidence), finding no 90-day mortality difference in low-risk-of-bias trials (RR 1.06; 95% CI 0.95-1.18).[21]
Artificial intelligence and machine learning
Artificial intelligence and machine-learning applications under investigation include early detection, molecular or clinical subphenotyping, and individualized treatment guidance.
- Early-warning systems: the Targeted Real-time Early Warning System (TREWS) has been associated with improved outcomes when alerts are confirmed promptly, whereas the original Epic Sepsis Model failed external validation in some settings, illustrating the importance of local validation and alert-performance assessment.[22]
- Sepsis ImmunoScore: an FDA-authorized machine-learning tool has demonstrated AUROC values of approximately 0.80-0.85 for prediction of Sepsis-3 sepsis within 24 hours in its validation studies.[23]
- Reinforcement-learning treatment guidance: the AI Clinician approach to fluid and vasopressor dosing has shown feasibility and is moving toward prospective evaluation, including the OVISS program.[24]
Important limitations include algorithmic bias, reliance on proxy labels for an incompletely observable sepsis onset, limited external validation, and insufficient prospective evidence that AI-guided strategies improve patient-centered outcomes.[25][24]
The 2026 Surviving Sepsis Campaign did not recommend any specific host-response diagnostic aid, including sepsis biomarker or immune-response tools, because no study had established that a diagnostic strategy using these tools improves patient-centered outcomes.[21]
Global context and low- and middle-income countries
Most sepsis randomized trials have been conducted in high-income countries, and their findings may not transfer directly to low- and middle-income countries (LMICs).
In Zambia, the SSSP-2 randomized clinical trial found that an early-resuscitation protocol using aggressive fluid and vasopressor therapy increased in-hospital mortality compared with usual care among adults with sepsis and hypotension. The population was largely young, HIV-positive, and at risk for tuberculosis, emphasizing the importance of local disease epidemiology, available organ-support capacity, and context-specific trial evidence.[26]
These findings caution against directly transplanting fluid-liberal resuscitation protocols developed in resource-rich settings to populations with limited ventilatory support or substantially different host and pathogen characteristics.[27]
Research priorities
The 2023 Surviving Sepsis Campaign research priorities identify precision and personalized therapy, real-time predictive screening, individualized fluid and vasopressor strategies, and improved definition of organ injury among important clinical research questions. Basic-science priorities include improved animal models, investigation of the microbiome, and host genetic and epigenetic factors.[28]
Clinically actionable recommendations
- Do not use investigational host-directed agents routinely. Anakinra, interferon-gamma, GM-CSF, IL-7, vilobelimab, mesenchymal stromal cells, and other investigational immunomodulatory therapies should not be considered routine sepsis treatment outside an appropriate clinical trial.
- Do not use intravenous vitamin C for routine sepsis treatment. The 2021 Surviving Sepsis Campaign recommended against intravenous vitamin C, and the 2026 guideline again suggested against it with a conditional recommendation and low-certainty evidence.[20][21]
- Do not extrapolate resource-rich resuscitation protocols automatically to LMIC settings. Local evidence and available organ-support capacity should inform implementation of sepsis interventions.[26]
- Use AI sepsis tools only with appropriate validation and monitoring. Local external validation, prospective performance assessment, and monitoring for false alarms and bias are important before clinical deployment.[22][25]
- Prefer enrollment in precision-immunotherapy and endotype-stratified trials when eligible. The ImmunoSep trial demonstrates improvement in organ dysfunction with biomarker-guided therapy but does not establish a mortality benefit.[8]
Areas of uncertainty
- No endotype has yet been definitively validated prospectively as treatment-selective for a specific sepsis therapy. The relative value of transcriptomic versus immunologic stratification and the temporal stability of endotypes remain uncertain.[2][4]
- Immunostimulation may worsen injury if administered to patients with a predominantly hyperinflammatory phenotype, providing a rationale for accurate immune-state stratification.[2]
- The infection-risk profile and clinical efficacy of complement inhibition in sepsis remain uncertain because sepsis-specific phase III evidence is lacking.[12]
- Mesenchymal stromal-cell efficacy remains unproven despite generally reassuring early safety data.[16]
- Regulatory authorization of an AI tool does not establish mortality benefit or demonstrate that AI-guided care improves patient-centered outcomes.[25][21]
High-yield clinical pearls
- Sepsis can involve opposing immune states; hyperinflammation and immunoparalysis may require fundamentally different therapeutic strategies.[2]
- Hyperferritinemia can identify a MALS phenotype being investigated for IL-1 blockade, whereas low monocyte HLA-DR can identify an immunoparalysis phenotype being investigated for immunostimulation.[6][8]
- Biomarker-guided immunotherapy in ImmunoSep improved a prespecified measure of organ dysfunction but did not demonstrate a significant 28-day mortality benefit.[8]
- For AI-based sepsis systems, timely clinician response and rigorous external validation are at least as important as algorithmic performance metrics.[22]
Common pitfalls
- Treating sepsis as a homogeneous disease and expecting a universal adjunctive immunotherapy.
- Using intravenous vitamin C or HAT based on early observational studies while disregarding subsequent randomized evidence.
- Deploying a commercial AI sepsis model without local validation or prospective monitoring.
- Applying resource-rich resuscitation protocols to LMIC populations without considering local evidence, disease epidemiology, and organ-support capacity.
- Administering immunostimulatory therapies empirically without an appropriate immune-state selection strategy.
References
- ↑ 1.0 1.1 1.2 Moore AR, Zheng H, Ganesan A; et al. (2025). "A consensus immune dysregulation framework for sepsis and critical illnesses". Nature Medicine. 31 (12): 4084–4096. doi:10.1038/s41591-025-03956-5.
- ↑ 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 Giamarellos-Bourboulis EJ, Aschenbrenner AC, Bauer M; et al. (2024). "The pathophysiology of sepsis and precision-medicine-based immunotherapy". Nature Immunology. 25 (1): 19–28. doi:10.1038/s41590-023-01660-5.
- ↑ 3.0 3.1 Pelaia TM, Shojaei M, McLean AS. (2023). "The Role of Transcriptomics in Redefining Critical Illness". Critical Care. 27 (1): 89. doi:10.1186/s13054-023-04364-2.
- ↑ 4.0 4.1 Scicluna BP, Cano-Gamez K, Burnham KL; et al. (2025). "A consensus blood transcriptomic framework for sepsis". Nature Medicine. 31 (12): 4119–4130. doi:10.1038/s41591-025-03964-5.
- ↑ Antcliffe DB, Burnham KL, Al-Beidh F; et al. (2019). "Transcriptomic Signatures in Sepsis and a Differential Response to Steroids. From the VANISH Randomized Trial". American Journal of Respiratory and Critical Care Medicine. 199 (8): 980–986. doi:10.1164/rccm.201807-1419OC. PMID 30365341.
- ↑ 6.0 6.1 Cajander S, Kox M, Scicluna BP; et al. (2024). "Profiling the Dysregulated Immune Response in Sepsis: Overcoming Challenges to Achieve the Goal of Precision Medicine". The Lancet Respiratory Medicine. 12 (4): 305–322. doi:10.1016/S2213-2600(23)00330-2. PMID 38142698 Check
|pmid=value (help). - ↑ 7.0 7.1 7.2 7.3 Leligdowicz A, Harhay MO, Calfee CS. (2022). "Immune Modulation in Sepsis, ARDS, and Covid-19 — The Road Traveled and the Road Ahead". NEJM Evidence. 1 (11): EVIDra2200118. doi:10.1056/EVIDra2200118.
- ↑ 8.0 8.1 8.2 8.3 8.4 Giamarellos-Bourboulis EJ, Kotsaki A, Kotsamidi I; et al. (2026). "Precision Immunotherapy to Improve Sepsis Outcomes: The ImmunoSep Randomized Clinical Trial". JAMA. 335 (9): 775–786. doi:10.1001/jama.2025.24175. PMID 41359996 Check
|pmid=value (help). - ↑ Shakoory B, Carcillo JA, Chatham WW; et al. (2016). "Interleukin-1 Receptor Blockade Is Associated With Reduced Mortality in Sepsis Patients With Features of Macrophage Activation Syndrome: Reanalysis of a Prior Phase III Trial". Critical Care Medicine. 44 (2): 275–281. doi:10.1097/CCM.0000000000001402. PMID 26584195.
- ↑ U.S. Food and Drug Administration. "Approved Drug Products with Therapeutic Equivalence Evaluations (Orange Book)".
- ↑ Bode C, Weis S, Sauer A, Wendel-Garcia P, David S. (2023). "Targeting the host response in sepsis: current approaches and future evidence". Critical Care. 27 (1): 478. doi:10.1186/s13054-023-04762-6.
- ↑ 12.0 12.1 Kox M, Bauer M, Bos LDJ; et al. (2026). "The immunology of sepsis: translating new insights into clinical practice". Nature Reviews Nephrology. 22 (1): 30–49. doi:10.1038/s41581-025-01004-6.
- ↑ Vacheron CH, Lepape A, Venet F; et al. (2023). "Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) in Patients Presenting Sepsis-Induced Immunosuppression: The GRID Randomized Controlled Trial". Journal of Critical Care. 78: 154330. doi:10.1016/j.jcrc.2023.154330. PMID 37267804 Check
|pmid=value (help). - ↑ van der Poll T, van de Veerdonk FL, Scicluna BP, Netea MG. (2017). "The immunopathology of sepsis and potential therapeutic targets". Nature Reviews Immunology. 17 (7): 407–420. doi:10.1038/nri.2017.36.
- ↑ Opal SM, van der Poll T. (2015). "Endothelial barrier dysfunction in septic shock". Journal of Internal Medicine. 277 (3): 277–293. doi:10.1111/joim.12331.
- ↑ 16.0 16.1 16.2 McMullan RR, McAuley DF, O'Kane CM, Silversides JA. (2024). "Vascular leak in sepsis: physiological basis and potential therapeutic advances". Critical Care. 28 (1): 97. doi:10.1186/s13054-024-04875-6.
- ↑ 17.0 17.1 Shaw TD, Krasnodembskaya AD, Schroeder GN; et al. (2021). "Mesenchymal Stromal Cells: An Antimicrobial and Host-Directed Therapy for Complex Infectious Diseases". Clinical Microbiology Reviews. 34 (4): e0006421. doi:10.1128/CMR.00064-21. PMID 34612662 Check
|pmid=value (help). - ↑ 18.0 18.1 Vera-Ponce VJ, Ballena-Caicedo J, Valladolid-Sandoval LAM; et al. (2025). "Monotherapy or combinations? Intravenous vitamin C in sepsis and septic shock: An umbrella review of 31 systematic reviews". PLoS One. doi:10.1371/journal.pone.0351072.
- ↑ Lamontagne F, Masse MH, Menard J; et al. (2022). "Intravenous Vitamin C in Adults with Sepsis in the Intensive Care Unit". The New England Journal of Medicine. 386 (25): 2387–2398. doi:10.1056/NEJMoa2200644.
- ↑ 20.0 20.1 Evans L, Rhodes A, Alhazzani W; et al. (2021). "Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021". Critical Care Medicine. 49 (11): e1063–e1143. doi:10.1097/CCM.0000000000005337.
- ↑ 21.0 21.1 21.2 21.3 Prescott HC, Antonelli M, Alhazzani W; et al. (2026). "Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2026". Critical Care Medicine. 54 (4): 725–812. doi:10.1097/CCM.0000000000007075.
- ↑ 22.0 22.1 22.2 Lafon T, Weingart M, Vaidie J; et al. (2026). "Challenges in Early Detection and Prognostication of Sepsis: New Approaches From the Emergency Department and Intensive Care Unit". EClinicalMedicine. 94: 103864. doi:10.1016/j.eclinm.2026.103864. PMID 42005923 Check
|pmid=value (help). - ↑ Bhargava A, López-Espina C, Schmalz L; et al. (2024). "FDA-Authorized AI/ML Tool for Sepsis Prediction: Development and Validation". NEJM AI. 1 (12). doi:10.1056/aioa2400867.
- ↑ 24.0 24.1 Papareddy P, Lobo TJ, Holub M; et al. (2025). "Transforming sepsis management: AI-driven innovations in early detection and tailored therapies". Critical Care. 29 (1): 366. doi:10.1186/s13054-025-05588-0.
- ↑ 25.0 25.1 25.2 Bignami EG, Berdini M, Panizzi M; et al. (2025). "Artificial Intelligence in Sepsis Management: An Overview for Clinicians". Journal of Clinical Medicine. 14 (1): 286. doi:10.3390/jcm14010286. PMID 39797368 Check
|pmid=value (help). - ↑ 26.0 26.1 Andrews B, Semler MW, Muchemwa L; et al. (2017). "Effect of an Early Resuscitation Protocol on In-hospital Mortality Among Adults With Sepsis and Hypotension: A Randomized Clinical Trial". JAMA. 318 (13): 1233–1240. doi:10.1001/jama.2017.10913.
- ↑ Machado FR, Angus DC. (2017). "Trying to Improve Sepsis Care in Low-Resource Settings". JAMA. 318 (13): 1225–1227. doi:10.1001/jama.2017.10547.
- ↑ De Backer D, Deutschman CS, Hellman J; et al. (2024). "Surviving Sepsis Campaign Research Priorities 2023". Critical Care Medicine. 52 (2): 268–296. doi:10.1097/CCM.0000000000006135. PMID 38240508 Check
|pmid=value (help).