The novel biomarker t⁶A accurately identified septic patients at admission but failed to predict outcome

Sepsis is a life-threatening condition caused by the body’s extreme response to infection [1]. Early diagnosis of sepsis is crucial for improving patient outcomes, yet current diagnostic methods including microbiological cultures are delayed and frequently inconclusive. This has driven the search for novel biomarkers and detection systems capable of recognizing sepsis more rapidly and accurately [2]. Procalcitonin (PCT) is the most widely used biomarker for sepsis detection but its use is limited as circulating PCT concentration is influenced by noninfectious inflammation (e.g. trauma, surgery).

Nucleoside modifications are a hallmark of the post-transcriptional processing of transfer ribonucleic acid (tRNA) that generate multiple structurally modified nucleosides [3]. One of such tRNA-modified nucleosides, N6-threonylcarbamoyladenosine (t⁶A), is critical for ensuring efficient protein synthesis in health and disease [4]. We discovered that circulating t⁶A possesses a diagnostic potential in sepsis. Consequently, we evaluated the diagnostic accuracy of t⁶A in differentiating bacterial sepsis and COVID-19 from two different non-septic patient cohorts: i) patients undergoing elective coronary artery bypass graft (CABG) surgery and ii) severe polytrauma patients. In all cohorts, t⁶A diagnostic accuracy was compared to PCT. Additionally, we tested t⁶A potential to predict death/survival in patients with sepsis.

This multicenter retrospective observational study analyzed plasma samples from four cohorts (Supplementary Table 1). The study included 81 patients with bacterial sepsis (cohort 1) and 49 patients with severe COVID-19 infection (cohort 2) diagnosed upon ICU/Emergency admission , 87 patients undergoing elective CABG surgery (cohort 3) and 64 severe (Injury Severity Score > 15) polytrauma patients (cohort 4). Sepsis in cohort 1 was defined according to the Sepsis-3 criteria, and all patients received treatment aligned with the Surviving Sepsis Campaign guidelines. In COVID-19 patients, SARS-CoV-2 infection was confirmed by molecular test. The CABG and polytrauma groups were used as non-septic comparators to assess t⁶A’s diagnostic specificity and accuracy against patients with sepsis and those with COVID-19. We compared blood samples collected at admission for the sepsis and COVID-19 cohorts, to samples collected 24 h post-surgery (cohort 3) and trauma (cohort 4). Plasma t⁶A concentrations were measured using tandem mass spectrometry with stable isotope internal standardization. PCT was measured using standard Brahms PCT luminescence immunoassay. Statistical analyses of the receiver operating characteristic (ROC) curves were carried out with a total of 50,000 stratified bootstrap samples to estimate the 95% confidence interval (CI) of the area under the curve (AUC) of the ROC curve and its optimal threshold. The AUCs were compared via De Long’s two-sided test [5].

t⁶A demonstrated outstanding diagnostic accuracy for sepsis at the ICU/Emergency admission. In separately evaluated exploratory and validation CABG cohorts, t⁶A consistently showed greater accuracy (AUC > 90%) compared to PCT (AUC 80–88%). t⁶A accuracy against the pooled CABG patients (cohort 3) achieved an AUC of 95% (vs. 88% for PCT, p < 0.05) (Fig. 1A). Against the polytrauma group, which represented a more clinically relevant comparator, t⁶A again outperformed PCT; its AUC was 97% compared to 88% for PCT (p < 0.05) (Fig. 1B). In the COVID-19 cohort, t⁶A also displayed excellent diagnostic performance when SARS-CoV-2-infected patients were compared to CABG (AUC 87%, Fig. 1C) and polytrauma (AUC 93%, Fig. 1D) patients. PCT measurement was not differentiating (AUC 50–52%, p > 0.05). Notably, the optimal diagnostic threshold for t⁶A was nearly identical (3% deviation) against CABG (40.0 ng/mL) and polytrauma (38.8 ng/mL) comparators; it varied by 22% for PCT (2.04 and 1.6 ng/mL).

Fig. 1

Diagnostic performance of t⁶A and PCT for detecting bacterial sepsis (A, B) and COVID-19 (C, D) at the ICU/Emergency admission. n: the first number in parentheses defines the number of sepsis or COVID-19 patients; the second number indicates the number of CABG or polytrauma patients. CI: 95% Confidence Intervals

Despite its strong diagnostic performance, t⁶A showed a limited utility for predicting sepsis outcomes. At the ICU/Emergency admission, t⁶A concentrations exhibited a considerable overlap between survivors and non-survivors and failed to show any protracted post-admission separation, resulting in an AUC of only 62%. PCT concentrations demonstrated a similar overlap dynamics and performed only modestly better, with an AUC of 72%, suggesting that while t⁶A is highly effective for identifying sepsis, it lacks prognostic value. This discrepancy highlights the biomarker’s primary utility in diagnosis rather than outcome prediction.

This is the first report regarding the potential utility of t⁶A as a diagnostic biomarker in patients with sepsis. We show t⁶A as a highly accurate and reliable biomarker for early sepsis diagnosis at the ICU/Emergency admission. t⁶A outperformed PCT in differentiating septic patients from non-septic controls including CABG and polytrauma cohorts. Insensitivity of t⁶A to sterile inflammation positions it as a promising diagnostic tool for clinical use. However, the lack of a commercially available assay for rapid t⁶A measurement limits its utility and this deficiency needs to be addressed. Further studies are required to evaluate t⁶A under different ICU conditions, including localized (non-sepsis) infections and extracorporeal therapies, to confirm its broader clinical applicability.