Presepsin and Procalcitonin as Potential Biomarkers for Early Diagnosis and Prognosis of Sepsis in Critically Ill Patients ()
1. Introduction
Sepsis is a life-threatening systemic reaction to infection characterized by hyper-inflammatory response followed by immunosuppression during which multiple organ dysfunctions are present [1]. Sepsis is the most common cause of death in critically ill patients. It develops in about 25% of intensive care unit (ICU) patients [2]. The mortality rates range from 20% to 50% in patients with sepsis and are >50% in patients with septic shock [3]. Early diagnosis and timely intervention of those patients is important to decrease mortality and improve sepsis-related survival [4].
Clinical scores have been introduced to predict hospital outcomes for critically ill patients e.g., Acute Physiology and Chronic Health Evaluation II (APACHEII), Sequential Organ Failure Assessment (SOFA) and quick (q) SOFA scores. However, the complicated methods for using these clinical scoring systems, proposing blood biomarkers as promising alternatives [5].
Blood culture is considered the gold standard for the diagnosis of blood infections, but it is time-consuming, even in the most advanced systems, and has poor sensitivity [6]. In addition, C-reactive protein (CRP) is considered as an early marker of infection or inflammation and could help with the monitoring of the progress of inflammation, but it is non-specific and cannot differentiate infectious from non-infectious inflammation as its level can be increased in autoimmune diseases, tumors, myocardial infarction, severe trauma, invasive surgical procedure and burns [7].Therefore, hundreds of the circulating biomarkers have been investigated for early diagnosis, risk stratification and prognosis of sepsis [8].
The soluble cluster of differentiation 14 subtype (sCD14-ST) or presepsin (PS) is a high-affinity receptor in monocytes, macrophages and granulocyte cells and their cell membranes that bind to lipopolysaccharide (LPS)/LPS-binding protein complexes [9]. Following infection, PS is produced and released into circulation either by secretion following phagocytosis of the CD14-pathogen complex or through shedding of CD14 from the cell membrane yielding soluble CD14 [10]. Many studies found that PS levels were elevated in patients with cardiac surgery and renal or liver disorders such as severe chronic kidney disease or liver cirrhosis without infection [2].
Procalcitonin (PCT) is the precursor of calcitonin that is normally synthesized in the C cells of the thyroid gland from pre-procalcitonin [11]. However, during infection, PCT is ectopically secreted into the circulation by liver, kidneys, spleen, lungs, pancreas, small intestine and leukocytes due to inflammatory stimuli mediated by interleukin 6 (IL-6) and tumor necrosis factor-alpha (TNF-α) [12]. PCT levels can be increased in conditions other than infection such as severe trauma, invasive surgical operations, major burns, cancers, autoimmune diseases and tissue necrosis after ischemic heart diseases [13].
An increasing number of studies have shown the ability of PS and PCT to differentiate between sepsis and systemic inflammatory response syndrome (SIRS) of non-infectious origin [14].
We aimed to investigate the role of the serum PS and PCT as early diagnostic and prognostic biomarkers for sepsis in critically ill patients.
2. Subjects and Methods
2.1. Subjects
This case-control study was conducted on adult critically ill patients who were admitted to ICU of Internal Medicine Department, Al-Zahraa University hospital, Cairo, Egypt, during the period from January 2020 to November 2020. The patients who were diagnosed with sepsis were enrolled in the study. The severity of sepsis was assessed according to APACHE II and qSOFA scoring systems at the time of admission and we selected 20 patients with sepsis, 20 patients with severe sepsis and 20 patients with septic shock to be included in the study. In addition, 20 age and sex matched apparently healthy individuals were included as a control group.
Informed consent was taken from all participants before enrollment in the study. The study was done after approval from Research Ethics Committee of Faculty of Medicine for Girls Al-Azhar University, and was conducted in accordance with the previsions of the Declaration of Helsinki (ethical approval number 20202151).
2.1.1. Inclusion Criteria
Adult critically ill patients (>18 years) diagnosed with sepsis as they fulfilled at least 2 criteria for SIRS which were: 1) temperature > 38˚C or <36˚C, 2) heart rate > 90 beats/minute, 3) respiratory rate > 20 breaths/minute or PaCO2 < 32 mmHg when on mechanical ventilation, 4) white blood cell count > 12,000/µL or <4000/µL, or an increase in the number of immature band forms (>10%) [15].
2.1.2. Exclusion Criteria
Pregnancy, immunosuppressive drugs, invasive surgical operations, cancers, acquired immunodeficiency syndrome, autoimmune diseases, ischemic heart diseases and end-stage liver and renal diseases.
2.1.3. Assessment of Subjects
1) Medical history taking and thorough clinical examination for all patients at the time of ICU admission, with special attention for the patient’s vital signs and sepsis parameters. The patients were observed for 28 days, and the mortality rate was recorded.
2) Radiological investigations: based on the patients’ signs and symptoms, to determine the source of infection.
- Abdominal and urogenital imaging: x-ray, ultrasound and/or tomography.
- Thoracic imaging: x-ray and/or tomography.
- Electrocardiogram.
3) Laboratory investigations:
- Urine analysis.
- Urine, stool, sputum and wound swap cultures: (according to the patients’ signs and symptoms).
- Complete blood count (CBC).
- Liver and kidney functions tests.
- CRP.
- Serum PS level.
- Serum PCT level.
2.2. Samples Collection and Preparation
About 7 ml of venous blood were drawn from each participant (within 1 hour after the diagnosis of sepsis) and divided into four aliquots: the first aliquot was 2 ml blood transferred to an EDTA tube for CBC using automated hematology cell counter (Cell dyne Ruby, Germany), the second aliquot was 2 ml blood transferred to plain tube for liver and kidney functions tests using fully automated chemistry analyzer (Cobas c 311, Roche Diagnostics kits, Germany) and the third aliquot was 3 ml blood transferred to plain tube, centrifuged at 3000 rpm for 20 minutes, then serum was separated and stored at −20˚C for measurement of PS, PCT and CRP by enzyme-linked immunosorbent assay (ELISA) in one assay to avoid repeated freeze-thaw cycles.
2.3. Measurement of PS and PCT Serum Levels
The concentrations of PS and PCT in serum were analyzed according to the manufacturer’s instructions by quantitative sandwich ELISA technique, with a complete set of ELISA reader (das 1851), using human Presepsin ELISA kit (Catalog number: 11141), supplied by Glory Science, China, and human Procalcitonin ELISA kit (Catalog number: E0977Hu), supplied by Bioassay Technology Laboratory, China.
2.4. Statistical Analysis
Data were coded and entered to the Statistical Package for Social Science (SPSS) version 23. Data were summarized using numbers and percentages for qualitative data; mean ± standard deviations (SD) for quantitative parametric data and median with inter-quartile range (IQR) for quantitative non parametric data. The comparison between qualitative data was done using Chi-square test, while the comparison between quantitative data was done using independent t-test for parametric and Mann-Whitney test for non-parametric data. The comparison between more than two groups was done by using One Way ANOVA followed by post hoc analysis by least significant difference (LSD) for quantitative parametric data. The correlations between quantitative data were done using Spearman correlation coefficients. Receiver operating characteristic (ROC) curve was performed with area under curve (AUC) analysis to detect the best cut off value, sensitivity and specificity of sUPAR, IL-34 and FIB-4 score for differentiation between patients and controls, and also between patients with severe and mild to moderate hepatic fibrosis. P-values less than 0.05 were considered statistically significant.
3. Results
Our study included 60 critically ill patients with sepsis and 20 controls. The patients were divided into three groups according to the severity of sepsis: group 1 included patients with sepsis (n = 20), group 2 patients with severe sepsis (n = 20) and group 3 patients with septic shock (n = 20). Comparison between patients and controls demonstrated a significant increase in PS, PCT, CRP, APACHE II and qSOFA scores (P < 0.001) in patients than controls. No significant difference was found in age (P = 0.247) and sex (P = 0.796) between patients and controls. The source of infection and number of survivors and non-survivors in the patient group are presented in (Table 1).
The median PS and PCT levels in each patient group and controls are shown in (Table 2) that demonstrated a significant increase in serum PS and PCT levels in each patient group than controls (P < 0.001), in severe sepsis than sepsis (P < 0.001 and 0.004 respectively), in septic shock than severe sepsis (P < 0.001) and in septic shock than sepsis group (P < 0.001).
It was found that PS and PCT levels were significantly higher in non-survival patients than survivors (P = 0.007 and 0.019 respectively) (Table 3).
The correlation of PS with other parameters revealed a significant positive correlations with APACHE II score (r = 0.792, P < 0.001), qSOFA score (r = 0.796, P < 0.001) and CRP level (r = 0.814, P < 0.001). In addition, PCT showed a significant positive correlations with APACHE II score (r = 0.728, P < 0.001), qSOFA score (r = 0.761, P < 0.001) and CRP level (r = 0.781, P < 0.001). There was a significant positive correlation between PS and PCT levels (r = 0.855, P < 0.001) (Table 4).
Furthermore, by using the ROC curve, the ability of PS for predicting sepsis in critically ill patients revealed the highest AUC (0.967) with 97.5% sensitivity, 85% specificity, 92.9% PPV, 94.4% NPV and cut-off of >635.5 pg/mL, followed by PCT that has AUC (0.946) with 97.5% sensitivity, 95% specificity, 97.5% PPV, 95% NPV and cut-off of >319.7 pg/mL. CRP showed the lowest AUC (0.902) with 75% sensitivity, 100% specificity, 100% PPV, 66.7% NPV and cut-off of >7 mg/L (Figure 1, Table 5).
The AUC of a combination of PS and PCT (0.989) or PS & CRP (0.989) or PCT & CRP (0.986) in predicting sepsis was significantly higher than that of PS or PCT or CRP alone (Figure 2, Table 5).
![]()
Table 1. Demographic and clinical characteristics of critically ill patients with sepsis and the controls.
•Independent t-test; *Chi-square test; ≠Mann Whitney test. P-value <0.05: Significant. SD, standard deviation; IQR, inter-quartile range; GE, gastro enteritis; UTI, urinary tract infection; CRP, C-reactive protein; APACHE, Acute Physiology and Chronic Health Evaluation; qSOFA, quick Sequential Organ Failure Assessment.
![]()
Table 2. Comparison between sepsis, severe sepsis and septic shock groups as regards PS and PCT levels.
•One Way ANOVA test; P-value < 0.05: Significant; IQR, inter-quartile range.
LSD, least significant difference; P-value < 0.05: Significant. P1: sepsis vs controls; P2: severe sepsis vs controls; P3: septic shock vs controls; P4: severe sepsis Vs sepsis; P5: septic shock Vs severe sepsis, P6: septic shock Vs sepsis.
![]()
Table 3. Comparison between survivors and non survivors as regards PS and PCT levels.
≠ Mann Whitney test; P-value < 0.05: Significant; IQR, inter-quartile range.
![]()
Table 4. Correlation of PS and PCT levels with other parameters.
P-value < 0.05: Significant; APACHE, Acute Physiology and Chronic Health Evaluation; qSOFA, quick Sequential Organ Failure Assessment; CRP, C-reactive protein.
![]()
Table 5. Diagnostic accuracy of various parameters to predict sepsis.
AUC, area under curve; PPV, positive predictive value; NPV, negative predictive value; CRP, C-reactive protein.
Finally, the ROC curve of PS and PCT for predicting septic shock in septic patients showed that PS has the highest AUC (0.969), with 90% sensitivity, 97.5% specificity, 94.7% PPV, 95.1% NPV and cut-off of >5500.6 pg/mL, followed by CRP that has AUC (0.945) with 90% sensitivity, 87.5% specificity, 78.3% PPV, 94.6% NPV and cut-off of >63 mg/L. PCT showed the lowest AUC (0.889) with 90% sensitivity, 97.5% specificity, 94.7% PPV, 95.1% NPV and cut-off of >822.1 pg/mL (Figure 3, Table 6).
The AUC of a combination of PS and PCT (0.971) or PS & CRP (0.980) or PCT & CRP (0.983) in predicting septic shock was significantly higher than that of PS or PCT or CRP alone (Figure 4, Table 6).
![]()
Figure 1. Roc curve of various parameters to predict sepsis.
![]()
Figure 2. Roc curve of combination of parameters to predict sepsis.
![]()
Table 6. Diagnostic accuracy of various parameters to predict septic shock.
AUC, area under curve; PPV, positive predictive value; NPV, negative predictive value; CRP, C-reactive protein.
![]()
Figure 3. Roc curve of various parameters to predict septic shock.
![]()
Figure 4. Roc curve of combination of parameters to predict septic shock.
4. Discussion
Despite new advances in critical care management, the incidence of sepsis is still increasing among hospitalized patients and early diagnosis still constitutes a challenge [16]. Since the use of the currently available tests were found to be limited, this justifies our aim to investigate the utility of serum PS and PCT as early diagnostic biomarkers of sepsis in critically ill patients and predictors of patients prognosis.
The current study showed a significant increase in PS level in patients with early sepsis than in healthy controls. This was in accordance with Kahveci et al., who found that the plasma PS level in the sepsis group was higher than that in the healthy control group [3].
Liu et al., reported that PS was a valuable biomarker for early diagnosis of sepsis and it is highly specific for diagnosing bacterial infections because it is produced in association with bacterial phagocytosis [17].
In addition, we found that with progression of sepsis severity, the serum PS level increased accordingly, and reached the highest level in septic shock. Similarly to our results, a great number of studies reported significantly higher levels of PS in septic shock patients compared to septic patients without shock [18] [19].
Masson et al., stated that PS was a robust circulating biomarker for sepsis, early stratification of its severity, as well as for patient prognosis. They explained their results by PS that appears to be released in the plasma as a consequence of cellular phagocytosis after bacterial infection and are therefore an indirect marker of sepsis. In addition, PS was a good marker of the host response, and its higher levels, independently of the type of infection, may indicate a state of immune paralysis, leading to a spreading of the related inflammatory reaction, which may lead to multiple organ failure and death [20].
In contrast, Kahveci et al., did not detect a significant difference in plasma PS levels between patients with septic shock and those with severe sepsis [3]. This could be explained by that their study included a less number of patients with acute kidney injury that increased the plasma PS levels. Therefore, a significant difference may not have appeared between the sepsis and septic shock groups. Nakamura et al., reported that not only sepsis but also accompanying kidney injuries contributed to an increase in PS levels [21].
Similarly, our study showed a significant increase in PCT level in patients with early sepsis compared with controls. This was in line with Kibe et al., who concluded that PCT was a biomarker currently used for diagnosis of sepsis [22].
We also found that PCT level increased with increase of sepsis severity till reached the highest level in septic shock patients. This was in agreement with De Oliveiria-Netto et al., who detected a significant increase in PCT in septic shock patients compared with those with sepsis and explained that by the hemodynamic changes presented in septic shock [23].
In contrast, Aliu-Bejtaa et al., did not find association of PCT levels with severity of sepsis [24]. Similarly, Kweon et al., did not report a difference in PCT levels among sepsis, severe sepsis and septic shock groups [25].
We found that the first-day PS and PCT levels were significantly higher in non-survivor septic patients than survivors. Our results confirmed the prognostic value of PS and PCT in adult patients with sepsis by being able to differentiate between patients with sepsis and those with septic shock, as well as between survivors and non-survivors.
In consistence to our results, Zhu et al., found that there is no statistically significant difference between both PS and PCT in predicting mortality of septic patients implying that both were promising prognostic biomarkers of sepsis [2].
In addition, Brodska et al., reported significantly higher PS levels in deceased patients [26]. In another study, Kim et al., reported higher mortality rates in patients with high PS levels [27]. Hassan et al., reported that increasing PS levels within the first week of hospitalization predicted ICU and 90-day mortality [15]. Jedynak et al., revealed that a high PCT level was associated with 28-day mortality in patients with sepsis [28].
While in opposite to our results, Kahveci et al., did not observe a significant difference in PS and PCT levels between survivor and non-survivor patients [3]. Yu et al., found that PCT levels had no statistical difference between survival and non-survival groups and within 12 days, PCT levels in both groups decreased synchronously. Comparatively, they found that PS levels in the survival group decreased persistently, while they rose gradually in the non-survival group [29].
The correlation studies revealed that the PS and PCT levels were positively correlated with APACHE II score, qSOFA score and CRP level, supporting the view that PS and PCT are potential markers for diagnosing sepsis and differentiating sepsis severity that significantly correlates with the activation of the systemic inflammatory state induced by sepsis. We also found that PS and PCT were positively correlated with each other.
Similar to our results, Behnes et al., found significant correlations between PS and PCT, as well as SOFA and APACHE scores, and subsequently severity of the disease [30]. In accordance with these findings, Drăgoescu et al., identified significant correlations between PS and the SOFA score and CRP [31]. Aliu-Bejtaa et al., revealed a strong correlation of PS with SOFA score, thus letting them believe that PS might be a specific sepsis biomarker [24]. However, Brodska et al., reported that initial PS concentrations do not correlate with SOFA score [26]. The fact that the results of the reported studies are incompatible necessitates further research.
We evaluated the diagnostic accuracy of PS and PCT to predict sepsis in comparison to CRP, by using ROC curve analysis. Our results revealed that PS has the highest AUC value for discrimination between control group and patients with early sepsis followed by PCT, while CRP has the lowest AUC value for its discrimination ability in comparison to PS and PCT. We also found that the combinations of PS and PCT, PS and CRP or PCT and CRP improved the accuracy of early diagnosis of sepsis.
Our results could be explained by the rapid pharmacokinetics of PS molecule that makes its level rises within 2 hours of any infection, with a maximum concentration after 3 hours, which is earlier than that reported for PCT and CRP that have significantly longer kinetics in bacterial or fungal infections. This specific feature of PS makes it a superior biomarker to PCT and CRP for early diagnosis of sepsis, while, both PCT and CRP might be still not reliable enough as early indicators for sepsis [32].
This was in agreement with Yamamoto et al., who found that PS had the highest diagnostic accuracy for discriminating non-sepsis from sepsis compared to PCT and CRP [33]. Tan et al., evaluated the diagnostic performance of PCT for the diagnosis of sepsis and found that PCT has a significantly higher diagnostic accuracy than CRP [34].
Different to our findings, Wu et al., suggested that there was no obvious better performance of PS than PCT in the diagnosis of sepsis. In addition, they observed that there was non-significant difference between CRP and PS as biomarkers for the early stage of sepsis [35].
Moreover, Enguix-Armada et al., showed that PCT had the highest AUC in the first 24 h after ICU admission. Also PCT and CRP showed similar diagnostic values [36]. While, Venkataraman et al., found that the diagnostic performance of PCT as a biomarker of sepsis was low and concluded that PCT therefore could not reliably differentiate sepsis from other conditions in critically ill adult patients [37].
Finally, we tested whether PS and PCT could predict septic shock in comparison to CRP, through ROC curve analysis. PS was found to yield higher AUC value than PCT and CRP, while PCT showed the lowest value. The combinations of any two biomarkers increased the accuracy of diagnosis of septic shock.
Liu et al., showed that PS had a higher predictive value than PCT for septic shock in patients with sepsis and the combination with other severity scores proved to be better, which was in line with our results [17]. However, different to our results, Behnes et al., observed that the AUC of PS to diagnose septic shock at day 1 of ICU treatment, was comparable to that of PCT [30].
There were some limitations for our study. First, the relatively low number of patients because it was a single-center study performed with no external funding. Second, a control patient group (non-sepsis ICU patient) was not included in the study. Third, the lack of repeated measurements for the study parameters over time as this was a single time-point measurement study only.
5. Conclusion
Our study demonstrated that the serum PS and PCT levels were found to be promising biomarkers for the early diagnosis of sepsis in critically ill patients, but PS has more diagnostic accuracy than PCT and CRP. Moreover, PS served as a prognostic index towards the progression of sepsis and severe sepsis to septic shock with higher diagnostic accuracy than PCT and CRP. While CRP level was more successful in predicting septic shock compared to the PCT level. The combinations of biomarkers improved the accuracy of early diagnosis of sepsis and septic shock. More comprehensive studies are necessary for further confirmation of our results.