Early Identification of a Cytokine-Low Immune Phenotype in Suspected Sepsis Using a Point-of-Care Whole-Blood TNF-α...

Routine biomarkers and clinical scores do not directly assess functional immune responsiveness in suspected sepsis. We evaluated whether the TNF-α Release Assay (TARA), a whole-blood LPS-induced functional assay, is associated with an early cytokine-low immune phenotype in Emergency Department (ED) patients.

In this prospective single-center cohort, adults with suspected sepsis and NEWS2 ≥ 3 were sampled at presentation, at 4 h and, when available, at 24 h. Immune phenotypes were defined independently of TARA by unsupervised k-means clustering of 12 circulating cytokines at 4 h, and a continuous Low-Response Cytokine Score (LRCS) was derived, with higher values indicating lower global cytokine concentrations.

Among 152 patients, three phenotypes were identified: C1 cytokine-low (n = 56), C2 inflammatory/intermediate (n = 63), and C3 IFN-γ/MCP-3/IL-17A-high (n = 33). Indicative TARA low-response status was independently associated with a higher LRCS after adjustment for clinical covariates and procalcitonin (PCT), both at presentation (β = 0.33; 95% CI, 0.14-0.53; p = 0.001) and at 4 h (β = 0.38; 95% CI, 0.19-0.57; p < 0.001). Indicative TARA low-response results at 4 h were most frequent in C1 and least frequent in C3 (75% vs. 45%). Exploratory clinical comparisons showed lower recorded early escalation in C1, whereas Sepsis-3 final diagnosis and in-hospital mortality were similar across phenotypes.

A TARA low-response status was independently associated with an early circulating cytokine-low phenotype defined without reference to TARA or outcomes, providing information complementary to routine biomarkers and clinical severity scores. These exploratory findings support TARA as a complementary functional immune readout but require external validation and clinical-impact evaluation.