Daily Sepsis Research Analysis
Analyzed 41 papers and selected 3 impactful papers.
Summary
Three clinically oriented sepsis studies stood out today: a pooled analysis shows weekly rezafungin is non-inferior to daily caspofungin for candidaemia/invasive candidiasis and achieves faster blood-culture clearance; an observational two-center cohort links early deep lymphopenia to higher risk of recurrent sepsis; and a large multicenter database study finds the lactate-to-albumin ratio outperforms lactate alone for 28-day mortality prediction.
Research Themes
- Antifungal therapy optimization in sepsis-related invasive candidiasis
- Host immune trajectories predicting recurrent sepsis
- Data-driven risk stratification using composite biomarkers
Selected Articles
1. Rezafungin Versus Caspofungin in Candidaemia and Invasive Candidiasis: A Post Hoc Pooled Analysis of Phase 2 and Phase 3 Trials.
Across pooled Phase 2/3 randomized, double-blind trials (n=372), weekly rezafungin was non-inferior to daily caspofungin for Day-30 all-cause mortality and achieved faster time to negative blood cultures (23.1 vs 36.7 hours; p=0.0073) with similar safety. Early mycological eradication rates favored rezafungin at Day 5 and were comparable by Day 14.
Impact: Provides robust, randomized evidence supporting a once-weekly echinocandin that maintains efficacy while accelerating bloodstream clearance, potentially simplifying inpatient-to-outpatient transitions.
Clinical Implications: Rezafungin is a viable alternative to caspofungin for candidaemia/invasive candidiasis with the advantage of weekly dosing and faster blood-culture clearance; formulary decisions and OPAT pathways may prioritize rezafungin for suitable patients.
Key Findings
- Day-30 all-cause mortality: 20% with rezafungin vs 22% with caspofungin (weighted difference −2.4%, 95% CI −11.2 to 6.4).
- Time to negative blood cultures was shorter with rezafungin (23.1 h) than with caspofungin (36.7 h), p=0.0073.
- Day-5 mycological eradication 73% (rezafungin) vs 65% (caspofungin); Day-14 70% vs 68%, respectively.
- Adverse event profiles were similar between treatment arms.
Methodological Strengths
- Pooled double-blind randomized Phase 2 and 3 trials with a diverse, international cohort
- Predefined efficacy endpoints including Day-30 mortality and mycological outcomes
- Statistically significant advantage in time to culture clearance with balanced safety
Limitations
- Post hoc pooled analysis may introduce heterogeneity across trials
- Not powered to detect small differences in mortality
- Limited data on specific sepsis subpopulations or source-control timing
Future Directions: Head-to-head pragmatic trials comparing outpatient pathways, PK/PD studies in special populations (renal/hepatic dysfunction), and cost-effectiveness analyses to guide formulary adoption.
BACKGROUND: The efficacy and safety of rezafungin in invasive candidiasis (IC) and candidaemia was demonstrated in the Phase 2 STRIVE and Phase 3 ReSTORE trials, and an additional cohort of patients in ReSTORE from China (ReSTORE China). OBJECTIVES: This post hoc analysis is the first time that combined data from all double-blind, randomised Phase 2 and 3 rezafungin trials have been evaluated. Outcomes were comprehensively assessed across a large, diverse population. METHODS: Data were assessed from adults with candidaemia/IC in STRIVE (NCT02734862) and ReSTORE (NCT03667690) (including the China extension cohort) who received weekly rezafungin 400/200 mg or daily caspofungin 70/50 mg for ≤ 4 weeks. Efficacy was assessed in the modified intent-to-treat population. Endpoints were Day 30 all-cause mortality (ACM, primary), mycological response at Days 5 and 14 (secondary) and time to negative blood cultures (TTNBC, secondary). Treatment-emergent adverse events were assessed in the safety population. RESULTS: Overall, 372 patients were included (rezafungin 179; caspofungin 193). Day-30 ACM was 20% (33/161) for rezafungin versus 22% (39/178) for caspofungin (weighted difference, -2.4% [95% confidence interval -11.2 to 6.4]). Day-5 and Day-14 mycological eradication rates were 73% and 70% for rezafungin versus 65% and 68% for caspofungin, respectively. Median TTNBC was shorter for rezafungin than for caspofungin (23.1 versus 36.7 h, respectively; p = 0.0073). Adverse-event profiles were generally similar across treatments. CONCLUSIONS: This pooled analysis of Phase 2 and Phase 3 data, including the China extension cohort, supports the efficacy and safety of rezafungin. Rezafungin was non-inferior to caspofungin, with earlier mycological eradication.
2. Sepsis-associated lymphopenia: Dynamic evaluation and its association with recurrent sepsis in critically ill patients.
In 4,701 ICU patients with sepsis/septic shock, sustained deep lymphopenia within the first 48 hours (ALC ≤500/µL) showed a dose-response association with recurrent sepsis using competing-risk models (sHR 1.61 for 1 day; 2.22 for 2 days). A low 48-hour time-weighted average ALC (≤500/µL) was similarly predictive, whereas milder lymphopenia had weaker ties.
Impact: Operationalizes early immune suppression into simple, time-aware metrics that stratify risk of recurrent sepsis, informing surveillance and potential immunoadjuvant trial enrollment.
Clinical Implications: Early monitoring of ALC depth and duration can identify patients at high risk of recurrent sepsis, supporting intensified infection surveillance, stewardship review, and consideration of immune-monitoring or immunostimulatory strategies.
Key Findings
- Recurrent sepsis occurred in 9.1% (429/4701) with 20.8% dying without recurrence (competing risk).
- ALC ≤500/µL for 1 and 2 days associated with higher recurrence risk (adjusted sHR 1.61 and 2.22, respectively).
- Lower 48-hour time-weighted average ALC (≤500/µL) independently predicted recurrence.
- Milder lymphopenia (≤1000/µL) showed weaker and inconsistent associations.
Methodological Strengths
- Large two-center cohort with robust event numbers
- Fine-Gray competing-risk models accounting for death without recurrence
- Dose-response assessment using both duration and time-weighted metrics
Limitations
- Retrospective design with potential residual confounding
- Lack of external validation across health systems or countries
- No interventional testing of immunomodulatory strategies guided by these metrics
Future Directions: Prospective validation and biomarker-integrated immune phenotyping to trigger targeted immunostimulatory or infection-prevention strategies; evaluation of thresholds for clinical decision support.
OBJECTIVES: To evaluate whether early lymphocyte trajectories, integrating both the depth and duration of lymphopenia, predict recurrent sepsis in critically ill adults. DESIGN: Retrospective two-center cohort study. SETTING: Intensive care units of two tertiary hospitals. PATIENTS OR PARTICIPANTS: Adult patients admitted with sepsis or septic shock between January 2011 and July 2024. INTERVENTIONS: None. MAIN VARIABLES OF INTEREST: Early lymphopenia during the first 48 h was quantified as: (1) number of days with absolute lymphocyte count (ALC) ≤1000 or ≤500 cells/µL, and (2) 48-h time-weighted average lymphocyte count (TWA-L). Recurrent sepsis, defined as a new septic episode occurring ≥7 days after the index event, was analyzed using Fine-Gray competing-risk models, treating death without recurrence as a competing event. RESULTS: Among 4701 patients, 429 (9.1%) developed recurrent sepsis and 972 (20.8%) died without recurrence. Profound lymphopenia (ALC ≤ 500 cells/µL) showed a dose-response relationship: compared with 0 days, 1 and 2 days were associated with adjusted subdistribution hazard ratios (sHRs) of 1.61 (95% CI 1.22-2.13) and 2.22 (95% CI 1.74-2.84), respectively. Higher TWA-L ≤500/µL was also independently associated with increased risk, whereas milder lymphopenia (≤1000/µL) showed weaker and inconsistent associations. CONCLUSIONS: Early sustained deep lymphopenia (ALC ≤ 500/µL), quantified by duration or cumulative burden, identifies patients at increased risk of recurrent sepsis and may support early immune risk stratification.
3. Machine Learning-Augmented Traditional Analysis of Lactate vs Lactate-to-Albumin Ratio for Predicting Mortality Risk in Patients With Sepsis: Large-Scale Retrospective Study.
In a 3,637-patient multicenter ICU cohort, the lactate-to-albumin ratio (LAR) outperformed lactate alone for predicting 28-day in-hospital mortality using paired AUC comparisons and multiple interpretable ML models. The prognostic advantage was particularly relevant among patients without severe hyperlactatemia.
Impact: Validates a simple composite biomarker that improves mortality discrimination over a widely used standard, backed by robust statistical testing and ML interpretability.
Clinical Implications: Incorporating LAR into early sepsis risk stratification may improve triage and resource allocation, especially when lactate is only modestly elevated; EHR-based calculators could automate deployment.
Key Findings
- Retrospective multicenter cohort (n=3,637) from eICU-CRD (208 US hospitals, 2014–2015).
- LAR showed higher and more stable discrimination for 28-day in-hospital mortality than lactate by DeLong AUC comparison.
- Interpretable ML models (nine algorithms with SHAP) consistently prioritized LAR over lactate across subgroups.
- Findings were particularly relevant for patients without severe hyperlactatemia.
Methodological Strengths
- Large multicenter cohort with paired AUC testing (DeLong) for fair biomarker comparison
- Use of restricted cubic splines and threshold analyses to probe nonlinearity
- Model-agnostic ML validation with SHAP explanations for interpretability
Limitations
- Retrospective design with potential selection bias (excluding patients dying within 48 hours)
- Temporal and assay variability in albumin/lactate measurements across hospitals
- Generalizability outside US 2014–2015 ICU settings remains to be tested
Future Directions: Prospective validation and EHR-integrated implementation trials comparing LAR-informed triage vs standard care; exploration of dynamic LAR trajectories.
BACKGROUND: Effective risk stratification in sepsis remains a critical clinical challenge. Serum lactate is a cornerstone biomarker of metabolic dysfunction, yet its predictive limitations-particularly in patients without severe hyperlactatemia-are well recognized. The lactate-to-albumin ratio (LAR), a composite mixed-unit index integrating markers of acute metabolic dysfunction and systemic inflammation, has emerged as a promising predictor; however, its incremental discriminative advantage over lactate had not been formally tested in a large multicenter cohort using paired statistical methodology. OBJECTIVE: This study aims to determine whether LAR provides statistically significantly higher prediction of 28-day mortality than lactate alone in adult intensive care unit (ICU) patients with sepsis, using threshold effect analysis, restricted cubic splines, DeLong test, and 9 interpretable machine learning models. METHODS: We conducted a retrospective analysis of 3637 adult patients with sepsis from the multicenter eICU Collaborative Research Database (eICU-CRD; 208 hospitals, United States, 2014-2015). The primary outcome was 28-day all-cause in-hospital mortality among patients surviving the initial 48-hour ICU admission period. We used multivariable logistic regression (LR), Cox proportional-hazards regression, threshold effect analysis, restricted cubic spline modeling, DeLong test for area under the receiver operating characteristic curve (AUC) comparison, and machine learning models evaluated with Shapley additive explanations (SHAP) for interpretability. The cohort was divided 70/30 (stratified) into training and held-out test sets; the Synthetic Minority Oversampling Technique was applied exclusively within the training partition to prevent data leakage. RESULTS: LAR consistently demonstrated stronger and more stable associations with mortality than lactate across all subgroups. DeLong test confirmed statistically significantly higher AUC for LAR: 28-day hospital mortality (AUC CONCLUSIONS: LAR demonstrates statistically significantly higher discrimination than lactate alone for 28-day sepsis mortality prediction. LAR may offer greater prognostic utility in patients without severe hyperlactatemia, a population in whom early risk stratification may be particularly relevant.