Dear Editor, Lipopolysaccharide (LPS) has a pivotal role in sepsis pathophysiology, but its measurement with limulus amœbocyte lysate (LAL) and other bioassays yielded conflicting results in clinical studies [1]. We developed a mass spectrometry method retaining LAL’s sensitivity in plasma for LPS quantitation, using a specific compound, 3-hydroxy myristate (3HM). When LPS is spiked in plasma, 3HM concentration remains stable, whereas LAL reactivity is decreasing rapidly over time [2]. In this observational study, we used 3HM to describe LPS concentration in a cohort of critically ill patients and in healthy control subjects.

Samples of 507 patients and 49 control subjects were analysed (methods and characteristics of patients are detailed in supplementary material), independently and blindly from clinical data. Patients were classified as septic (sepsis and septic shock), or non-septic (distribution of patients is detailed in Fig. 1 a, b). Septic patients had significantly higher 3HM levels when compared to nonseptic patients (127±75 vs 112±62 pmol/ml, p= 0.02). Moreover, 3HM levels were significantly higher in septic shock patients compared to sepsis patients, although they were similar between non-septic patients and controls (Fig. 1 a). Compared to baseline, 3HM levels decreased by 5 days in septic shock patients, and to a lesser extent in sepsis and non-septic patients (eFigure-1). In patients