Diagnostic and prognostic value of presepsin in non-infectious organ failure, sepsis, and septic shock: a prospective observational study using Sepsis-3

We investigated the diagnostic and prognostic value of presepsin among patients with organ failure in the emergency department. This prospective observational study included 420 patients divided into three groups: non-infectious organ failure (n = 142), sepsis (n = 141), and septic shock (n = 137). Optimal cut-off values of presepsin to discriminate between the three groups were evaluated using receiver operating characteristic curve analysis. We extracted the optimal cut-off value of presepsin to predict mortality associated with sepsis and septic shock and performed Kaplan–Meier survival curve analysis according to the cut-off value. Presepsin levels were signicantly higher in sepsis than in non-infectious organ failure (p < 0.001) and signicantly higher in septic shock than in sepsis (p = 0.002). The optimal cut-off value for presepsin to discriminate between sepsis and non-infectious organ failure was 582 pg/mL (p < 0.001) and to discriminate between sepsis and septic shock was 1285 pg/mL (p < 0.001). The optimal cut-off value for presepsin for predicting 30-day mortality was 821 pg/mL (p = 0.005) in septic patients. Patients with higher presepsin levels ( ≥ 821 pg/mL) had signicantly higher mortality rates than patients with lower presepsin levels (< 821 pg/mL) (log-rank test; p = 0.004). Presepsin levels could effectively differentiate sepsis from non-infectious organ failure and help clinicians identify sepsis patients with poor prognosis.

Introduction that presepsin could effectively discriminate sepsis without shock from non-septic patients with an increase in sepsis-related organ failure assessment (SOFA) score of ≥ 2 17 .
However, to the best of our knowledge, there has been no study on the diagnostic and prognostic value of presepsin, including patients with organ failure in the emergency department (ED), according to the latest Sepsis-3 de nitions. Therefore, we aimed to investigate the diagnostic value of presepsin levels in patients with non-infectious organ failure, sepsis, and septic shock, as well as the prognostic value of presepsin levels in patients with sepsis and septic shock.

Results
Study population and baseline characteristics. During the study period, 517 patients with positive qSOFA scores upon presentation to the ED were screened using the i-SMS (Fig. 1). Among them, 97 patients were excluded due to refusal to participate (n = 54), increase in SOFA score of < 2 (n = 31), admission for trauma care (n = 7), or unknown outcomes (loss to follow-up) (n = 5). The nal study population comprised 420 patients. Of these patients, 142 had non-infectious organ failure, 141 had sepsis, and 137 had septic shock. A owchart of the study population is shown in Fig. 1. The baseline characteristics of the study population are presented in Table 1. Patients with sepsis and septic shock were older than those with non-infectious organ failure. Sex and the Charlson comorbidity index did not differ between the three groups. Acute Physiology and Chronic Health Evaluation (APACHE) II, SOFA, National Early Warning (NEWS), and Modi ed Early Warning (MEWS) scores were signi cantly higher in sepsis and septic shock patients than in non-infectious organ failure patients. The 7-, 14-, 30-, and 90-day mortality rates were higher in patients with septic shock than in the other groups. Table 2 shows the principal clinical diagnoses of patients with non-infectious organ failure according to the affected organ systems: 52, central nervous system disorders; 41, cardiovascular disorders; 21, respiratory disorders; 19, hepatobiliary disorders; 14, renal disorders; and 7, coagulation disorders. The most common diagnoses were hypovolemic shock, metabolic encephalopathy, cerebral hemorrhage, heart failure, chronic obstructive pulmonary disease, asthma, seizure, and liver cirrhosis ( Table 2).   Fig. 2 and Table 1. Presepsin, PCT, and CRP levels were signi cantly higher in patients with sepsis or septic shock than in patients with non-infectious organ failure. Presepsin and PCT levels were signi cantly higher in patients with septic shock than in those with sepsis. In contrast, we observed no signi cant differences in CRP levels between the sepsis and septic shock groups (Fig. 2).

Discussion
To the best of our knowledge, this is the largest prospective observational study on both the diagnostic and prognostic value of presepsin in non-infectious organ failure, sepsis, and septic shock, in accordance with the latest Sepsis-3 de nitions. Presepsin had excellent accuracy in discriminating sepsis from noninfectious organ failure and had fair accuracy in discriminating septic shock from sepsis. The discriminating power of presepsin was comparable to that of PCT among patients with non-infectious organ failure, sepsis, and septic shock. The prognostic value of presepsin was superior to that of PCT and CRP in patients with sepsis and septic shock.
A recent study using Sepsis-3 reported that presepsin and PCT were superior to CRP and lactate in discriminating sepsis, including shock from non-sepsis with SIRS and a SOFA score ≥ 2 16 . The AUC values used to discriminate sepsis from non-sepsis were 0.88 for presepsin, 0.81 for PCT, and 0.65 for CRP 0.65, respectively. The AUC value of presepsin in the study was similar to that in our study (AUC = 0.877), and the sensitivity, speci city, positive predictive value, negative predictive value, and accuracy of presepsin for diagnosing sepsis (including shock) using a cut-off value of 508 pg/mL were 87%, 86%, 93%, 76%, and 87%, respectively. The cut-off value found in the previous study (508 pg/mL) was lower than that in our study (582 pg/mL). Our study is similar to the previous study in that it was performed in the ED according to the latest Sepsis-3 de nitions. However, we included a much larger population and used qSOFA as a screening tool instead of SIRS because it is no longer recommended as a diagnostic criterion for sepsis in the new de nitions 1 . These differences might have caused the difference in the cutoff values between the two studies. Another study using Sepsis-3 in a Spanish population also reported that presepsin can effectively discriminate sepsis from non-infectious SIRS 16 . However, these two studies using Sepsis-3 did not evaluate the prognostic value of presepsin.
A previous study reported that presepsin was superior to PCT and CRP in discriminating sepsis from SIRS in acute abdominal conditions 27 . In contrast, another study showed that the diagnostic capacity of presepsin was not superior to that of PCT 19 , suggesting that its introduction and routine use in clinical practice were not justi ed. Another study also reported that presepsin did not outperform traditional biomarkers in distinguishing sepsis from SIRS and predicting mortality 28 . In fact, results reported about the diagnostic value of presepsin are controversial, probably due to different study designs and settings. Therefore, speci c decision levels are required to determine the clinical roles of presepsin in different settings of non-infectious and infectious diseases 29 .
A multicenter prospective study reported that mean presepsin levels were signi cantly higher in nonsurvivors of sepsis than in survivors 23 . However, in that study, no signi cant correlation was observed between PCT levels and survival outcomes 23 . Similar to the previous study, our results showed that presepsin levels were signi cantly higher in non-survivors than in survivors. No signi cant differences in PCT levels were observed between the non-survivors and survivors. In our study, Kaplan-Meier survival curve analysis according to the optimal cut-off value of presepsin showed that 30-day mortality was signi cantly higher in patients with higher presepsin levels. In accordance with our study, a systematic review and meta-analysis revealed that presepsin levels on the rst day had prognostic value in predicting in-hospital or 30-day mortality in adult patients with sepsis 30 . The combination of presepsin with PCT, Galectin-3, and soluble suppression of tumorigenicity-2 showed better performance in predicting mortality than the single use of presepsin in sepsis patients 10 . The study demonstrated that the combination of presepsin with other biomarkers could help clinicians predict mortality. Further studies with larger cohorts are required to determine the optimal cut-off value of presepsin for predicting mortality associated with sepsis.
The present study had some limitations. First, although the present study included a large sample size compared to that of previous studies, it was a single-center ED-based study. Thus, our results may not be applicable to other EDs or ICUs. Second, only plasma presepsin levels in the ED were measured, and follow-up changes in markers were not determined. Although a previous study reported that dynamic monitoring of presepsin could effectively predict prognosis 31,32 , other trials demonstrated that single measurements of presepsin in the ED also had valuable prognostic value in patients with sepsis 12,23 .
Third, although a previous study reported that presepsin levels were markedly elevated in patients with chronic kidney disease receiving hemodialysis 33 , our study did not consider kidney function. Further studies are needed to investigate the in uence of kidney dysfunction on presepsin levels using repeated marker measurements. Fourth, because the present study included patients with organ dysfunction enrolled in the ED, this might have resulted in selection bias. Nevertheless, we postulate that our study, based on an organ failure cohort, could re ect the clinical characteristics of patients in a real ED setting.
To summarize, the present study, according to the Sepsis-3 de nitions, demonstrated the diagnostic and prognostic value of presepsin levels among patients with non-infectious organ failure, sepsis, and septic shock. Its ability to discriminate sepsis, including shock, from non-infectious organ failure was excellent, and its prognostic ability could help clinicians to prognosticate patients with sepsis and septic shock. Further multicenter prospective studies with larger populations are needed to determine the optimal cutoff value of presepsin for the diagnosis and prognosis of sepsis.

Methods
Study Design and Setting. This single-center prospective observational study was performed at the ED of the Korea University Ansan Hospital, Korea. Our institution is a 910-bed tertiary care teaching hospital with approximately 50,000 annual patient visits to the ED. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of Korea University Ansan Hospital (IRB no. 2020AS0031). Written informed consent was obtained from all patients or their legal representatives.
Study population. From July 2019 to August 2020, adults (≥ 18 years) who had a positive quick sepsisrelated organ failure assessment (qSOFA) score upon ED presentation were screened for participation.
This scoring system uses three criteria: low blood pressure (systolic blood pressure ≤ 100 mmHg), high respiratory rate (≥ 22 breaths/min), and altered mental status (Glasgow coma score < 15). One point was assigned for each criterion, with a nal score ranging from 0 to 3 points. A positive qSOFA score was de ned as a qSOFA score ≥ 2. In the present study, another inclusion criterion was an increase in the SOFA score by ≥ 2 points in the ED, irrespective of the current infection. Since September 2017, our institution has been using the Intelligent Sepsis Management System (i-SMS), a qSOFA alert system, which helps ED clinicians promptly identify sepsis and manage sepsis according to the SSC 2016 guidelines 4,5 . The system automatically enrolled patients who had a positive qSOFA score upon ED arrival and assisted in the decision-making process for sepsis management. If the patients had baseline SOFA scores, we used the standard of an increase in SOFA score of at least 2. If the patients had no previous SOFA score, two infectious disease (ID) experts reviewed the medical records with laboratory data and determined the change in the SOFA score. Exclusion criteria were refusal to consent, an increase in SOFA score < 2, ED visit for trauma care, and unknown outcomes. Therefore, all enrolled patients had a qSOFA score ≥ 2 or an increase in the SOFA score by ≥ 2 points. Eligible patients were divided into the following three groups based on the presence of infection and sepsis severity: non-infectious organ failure, sepsis, and septic shock. All patients were carefully selected and reviewed by two ID experts and an emergency attending physician.

De nitions
According to Sepsis-3 de nitions, sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection 1 .
Septic shock is de ned as a subset of sepsis in which profound circulatory, cellular, and metabolic abnormalities pose a greater risk of mortality than sepsis alone 1,6 . Sepsis-3 recommends the use of the qSOFA score to identify patients with a poor prognosis outside the intensive care unit (ICU). The diagnostic criteria for sepsis include an increase in the SOFA score by ≥ 2 points due to current infection. The criteria for septic shock include the requirement for a vasopressor to maintain a mean arterial pressure of 65 mmHg and serum lactate level > 2 mmol/L despite adequate uid resuscitation. Finally, the criteria for "non-infectious organ failure" included a positive qSOFA score and an increase in SOFA score by ≥ 2 points without current infection. Two independent ID experts reviewed all patients to determine the presence of a current infection.
Assays. We sampled blood for presepsin and PCT from a peripheral vein within 6 hours of ED presentation. Plasma presepsin levels were measured using an automated chemiluminescent enzyme immunoassay (PATHFAST system, LSI Medience Corporation, Tokyo, Japan). This novel system, based on the chemiluminescent enzyme immunoassay principle, was developed to analyze blood samples, providing results within 17 min 34 . During incubation of the sample with alkaline phosphatase (ALP)labeled anti-presepsin polyclonal antibodies and anti-presepsin monoclonal antibody-coated magnetic particles, presepsin binds to anti-presepsin antibodies, assembling an immunocomplex with the ALPlabeled antibodies and mouse monoclonal antibody-coated magnetic particles. The manufacturerclaimed assay range of presepsin was 20-20,000 pg/mL. Plasma presepsin concentrations were measured after the enrolled patients were discharged from the ED. Therefore, the assay results were unavailable to ED physicians and could not in uence the management and disposition of the patients.
PCT levels were measured using the Elecsys BRAHMS procalcitonin automated electrochemiluminescence assay (BRAHMS, Henningsdorf, Germany) on the Roche Cobas e-System (Roche Diagnostics, Basel, Switzerland). The manufacturer-claimed assay range of PCT was 0.02-100 ng/mL.
Outcomes. The primary outcome in the present study was 30-day mortality, and the secondary outcome was 90-day mortality. We excluded patients who were lost to follow-up from the 30-day and 90-day analyses.  Flowchart of the study population.  Kaplan-Meier survival curve analysis and log-rank test according to the optimal cut-off of presepsin for predicting 30-day mortality in patients with sepsis and septic shock.