Secondary Logo

Journal Logo

Basic Science: Original Clinical Research Report

Elevated Presepsin Is Associated With Perioperative Major Adverse Cardiovascular and Cerebrovascular Complications in Elevated-Risk Patients Undergoing Noncardiac Surgery: The Leukocytes and Cardiovascular Perioperative Events Study

Handke, Jessica MSc*; Scholz, Anna S.*; Gillmann, Hans-Jörg MD; Janssen, Henrike MD*; Dehne, Sarah MD*; Arens, Christoph MD*; Kummer, Laura MSc*; Uhle, Florian PhD*; Weigand, Markus A. MD*; Motsch, Johann MD*; Larmann, Jan MD, PhD*

Author Information
doi: 10.1213/ANE.0000000000003738



  • Question: Are leukocyte subpopulation counts, their perioperative change, or their activation during noncardiac surgery associated with perioperative major adverse cardiovascular and cerebrovascular events (MACCEs) in elevated-risk patients?
  • Findings: Noncardiac surgery was associated with changes in monocyte subset counts, and the monocyte activation marker presepsin is associated with MACCE in elevated-risk patients.
  • Meaning: Elevated preoperative presepsin is associated with MACCE after noncardiac surgery and might qualify to improve cardiovascular risk prediction in the future.

Precise measures for preoperative (pre-OP) identification of patients at risk for perioperative major adverse cardiovascular and cerebrovascular events (MACCEs) are essential in clinical routine. Reliable tools to identify patients in need of intensified observation after surgery are scarce, and uncertainty exists regarding the optimal risk stratification model. Scores to predict cardiovascular events or mortality largely rely on the evaluation of clinical risk factors. However, these scores underestimate the risk, particularly in high-risk patients and vascular surgery.1 Guidelines for pre-OP cardiovascular patient evaluation advocate measurement of cardiac troponins and natriuretic peptides in high-risk patients. However, evidence for biomarker measurements is controversial.2 High-sensitive cardiac troponin T (hs-cTnT) and N-terminal probrain natriuretic peptide (NT-proBNP) are established biomarkers for myocardial injury and heart failure, but they are not actively involved in cardiovascular disease progression.

Circulating blood cells are causally involved in atherogenesis. During atherosclerotic lesion development, cells of the innate and adaptive immune system are mobilized from their reservoirs and infiltrate the vessel wall.3 In animal models, increased recruitment and infiltration into atherosclerotic plaques drive perioperative lesion progression and destabilization.4,5 Plaque erosion or rupture often results in occlusion of the vascular bed. Depending on the affected vessel, occlusion might cause life-threatening cardiovascular complications. Leukocyte subsets are ascribed a long-term association with cardiovascular disease and have been shown to predict cardiovascular events in prospective clinical trials independent of surgeries. In detail, high levels of leukocytes,6 natural killer cells,7 and classical8 and intermediate monocytes,9 as well as reduced levels of nonclassical monocytes10 and regulatory T cells,11 positively correlate with adverse cardiovascular outcome (Supplemental Digital Content, Table 1, The monocyte activation marker presepsin (soluble CD14 subtype, sCD14-ST) is an established marker for the early identification of systemic infections12 and has recently been proposed as a biomarker for pre-OP risk prediction in cardiac surgery.13 Presepsin arises from membrane-bound CD14, which is shed from classical monocytes on cellular activation.

Inflammatory cells are causally involved in cardiovascular disease by driving acute plaque destabilization. White blood cells (WBCs) in general expand during the early perioperative period.14 However, it has not been evaluated whether baseline counts of leukocyte subpopulations, their perioperative change, activation, or differentiation during surgery are associated with perioperative MACCE.

Therefore, we analyzed perioperative levels of WBC, monocyte subsets, and lymphoid cell populations for their association with MACCE. In a post hoc analysis, we further evaluated the prognostic performance of the monocyte activation marker presepsin and its additive value for NT-proBNP-based perioperative cardiovascular risk prediction.


Study Design and Population

We performed a single-center, prospective, observational cohort study in patients with coronary artery disease undergoing elective noncardiac surgery at the Department of Anesthesiology, University Hospital Heidelberg, Heidelberg, Germany. The trial was registered prior to patient enrollment at (NCT02874508, Principal investigator: J.M., Date of registration: August 22, 2016). The study protocol conformed to the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Medical Faculty of the Ruprecht-Karls University Heidelberg (S-351/2016, August 4, 2016). We tested the following hypothesis: noncardiac surgery in elevated-risk patients is associated with perioperative changes of atherogenic leukocyte subpopulations. We further tested whether changes of these cell populations are associated with perioperative cardiovascular events. From August to October 2016, consecutive patients ≥18 years of age with coronary heart disease undergoing elective, in-patient, noncardiac surgery were enrolled after providing written informed consent. This article adheres to the applicable Standards for Reporting Diagnostic Accuracy Studies guidelines. Exclusion criteria were pregnancy, breastfeeding, leukemia, baseline leukocytes outside the normal reference of 4–10 cells per nanoliter blood (nL−1), emergency surgery, history of organ transplantation or splenectomy, immunosuppression, chemotherapy, granulocyte-macrophage colony-stimulating factor, or cortisone treatment within the past 14 days or an intraoperative dexamethasone administration. Patients were also excluded if they experienced myocardial infarction (MI), myocardial ischemia, embolic or thrombotic stroke, congestive heart failure, or severe cardiac arrhythmia within the past 28 days before enrollment.

We recruited patients undergoing general, vascular, and urological surgery. General anesthesia was performed according to departmental standard operating procedures. Anesthesia was induced using propofol or etomidate and sufentanil and was maintained as balanced anesthesia using sevoflurane or desflurane. General anesthesia was combined with regional anesthesia in 4 cases.

Data Collection and Conventional Risk Assessment

Preoperatively, demographic data, American Society of Anesthesiologists physical status classification, preexisting diseases including cardiovascular and cerebrovascular events, history of smoking, current medication, and renal status were documented. Estimated glomerular filtration rates (eGFRs) were estimated from pre-OP blood samples by adjusting serum creatinine levels for sex, age, body weight, and height using the Chronic Kidney Disease Epidemiology Collaboration creatinine equation. CKD was defined as eGFR < 60 mL·min−1 (Kidney Disease: Improving Global Outcomes [KDIGO] stage ≥3) as proposed by KDIGO guidelines.15 A 12-lead electrocardiogram was recorded pre-OP. Conventional risk assessment was based on the revised cardiac risk index (RCRI) and pre-OP hs-cTnT and NT-proBNP. hs-cTnT ≥ 14 pg·mL−1 and NT-proBNP ≥ 300 ng·L−1 were considered abnormal.16

Sample Collection, Preparation, and Flow Cytometry Analysis

Blood for flow cytometry and differential blood counts was collected prior to skin incision, 2 and 6 hours after incision, and in the morning of postoperative days (PODs) 1 and 2 in lithium heparin gel or EDTA tubes (Sarstedt, Nümbrecht, Germany). NT-proBNP concentration was measured pre-OP. hs-cTnT was determined in lithium heparinized blood pre-OP and daily on POD1–3. Samples were processed immediately to avoid blood storage effects. Automated differential blood counts, hs-cTnT (Cobas E4111, Roche Diagnostics, Mannheim, Germany), and NT-proBNP (Immulite, Siemens Health care Diagnostics, Erlangen, Germany) measurements were performed in the central laboratory. In an observer-blinded fashion, selected leukocyte subpopulations were identified by flow cytometry (FACSVerse; BD Biosciences, Heidelberg, Germany) based on surface markers: classical (CD14++CD16), intermediate (CD14++CD16+), and nonclassical (CD14+CD16++) monocytes and natural killer cells (CD3CD16+CD56+) and regulatory T cells (CD3+CD4+CD25+CD127). Whole blood was treated with Human TruStain FcX (BioLegend, London, UK) to minimize unspecific binding. Cells were incubated for 30 minutes at 4°C with anti-CD4 APC, anti-CD14 FITC, anti-CD16 PE, anti-CD25 PerCP-Cy5.5 (all BioLegend), anti-CD3 PE-Cy7, and anti-CD127 FITC (both BD Biosciences). Results were analyzed using BD FACSuite Software (version, BD Biosciences). Lymphocytes and monocytes were gated based on forward- and side-scatter features. Leukocyte subpopulations were quantified based on their characteristic expression profiles (Supplemental Digital Content, Figure 1A–F, Absolute cell counts were determined by multiplying relative counts for mononuclear and lymphoid subpopulations obtained by flow cytometry with absolute monocyte and lymphocyte numbers from differential blood counts, respectively. Cell counts are reported as the number of cells per nanoliter blood (nL−1).


Presepsin (sCD14-ST) was measured to determine differentiation of classical to inflammatory nonclassical monocytes prior to surgery and on POD1 using a chemiluminescence-based, noncompetitive immunoassay on the PATHFAST analyzer (LSI Medience, Tokyo, Japan).17 Lithium heparin plasma samples were stored at −80°C until batch processed.

Outcome Analysis

Study participants were followed up until day 30 postsurgery. Patients were screened for the primary composite end point MACCE defined as cardiovascular death, MI, myocardial ischemia, or stroke. Definitions are listed in Supplemental Digital Content, Methods, Predefined secondary end points included individual components of the primary end point MACCE, new-onset atrial fibrillation, peripheral vascular occlusion, acute kidney injury, or myocardial injury after noncardiac surgery (MINS; Supplemental Digital Content, Methods, Post hoc, we assessed presepsin level discrimination for all-cause mortality. Patient visits took place on POD1–3. For outcome analyses, hs-cTnT was monitored pre-OP and daily until POD3. A 12-lead electrocardiogram was recorded at POD3. At the end of follow-up, patient charts were screened, and participants or their family doctors participated in a structured telephone interview.

Statistical Analysis

Statistical analyses were performed using IBM SPSS Statistics 24.0 (SPSS, Chicago, IL), MedCalc 16.8 (MedCalc Software, Ostend, Belgium), and Prism 7.02 (GraphPad Prism Software, Inc, San Diego, CA). Continuous data are presented as median (interquartile range); categorical data are presented as total and relative counts unless otherwise stated. Boxes mark interquartile ranges and whiskers represent 5th to 95th percentiles. Differences in baseline categorical and continuous variables were evaluated using Fisher-Yates and 2-tailed Mann-Whitney U test, respectively (α < .05). Postoperative leukocyte subset counts were individually compared to baseline values using nonparametric Friedman’s test. To account for multiple comparisons of prospectively collected data, Bonferroni correction or Dunn’s post hoc test were used as appropriate. When patients were discharged before completion of post-OP visits, findings and hs-cTnT values from the last visit were carried forward (last observation carried forward analysis). Receiver operating characteristics curve analyses and the maximized Youden index were calculated to define optimal cutoffs and assess discriminatory power for MACCE.18 Data are presented as area under the curve (AUC; 95% CI). Odds ratios (ORs; 95% CI) were calculated using Woolf’s method. Net reclassification improvement (NRI) was calculated to compare reclassification benefits and accuracies of hs-cTnT or presepsin in addition to NT-proBNP.19

Sample Size Calculation

We used G*power 3.1.7 to calculate the sample size necessary to test the null hypothesis stating that mean cell counts are equal before and after surgery. We expected an effect size of 0.3320 and calculated with a Bonferroni-corrected family wise type I error rate of 0.05/6 = 0.008 to correct for 6 different cell types. Analyzing 1 group at 5 time points requires 30 patients to achieve a statistical power of 0.95, assuming a correlation of repeated measures of 0.5 and nonsphericity correction of 0.8. To adjust for nonparametric analyses, 15% individuals were added.21 Expecting 15% dropouts, we estimated a final sample size of 40 patients.


We screened 44 patients with coronary artery disease scheduled for elective, noncardiac surgery. Forty patients were enrolled into the study. Four patients were not enrolled because of not fulfilling inclusion criteria (n = 2) or not meeting exclusion criteria (n = 2). Two patients were discharged home prior to the POD3 visit, and 1 was discharged before the POD2 visit. For those 3 patients, data were imputed (last observation carried forward analysis). For 2 patients, initial POD1 samples for leukocyte analysis were lost due to technical reasons. Another sample was drawn and analyzed in ≤24 hours of the POD1 time point. Two patients withdrew consent and were excluded. The final analysis set consisted of 38 individuals (Supplemental Digital Content, Figure 2,

Patient Characteristics

Table 1.:
Clinical Baseline Characteristics

Main clinical and demographical baseline characteristics are presented in Table 1 and Supplemental Digital Content, Table 2, Mean age was 69 ± 8 years. Eighty-two percent of the participants were male. pre-OP risk factors and cardiovascular diseases were frequent, and most patients received cardiovascular medication. Five patients (13%) experienced MACCE during the 30-day follow-up. Three subjects suffered from MI, 1 experienced myocardial ischemia, and 1 suffered from stroke (Supplemental Digital Content, Figure 2, When stratified for no MACCE versus MACCE, groups were similar in most baseline characteristics. Patients experiencing a MACCE had higher creatinine values and a lower glomerular filtration rate. Accordingly, the proportion of renal impairment (KDIGO stage ≥3; P = .005) was higher in patients undergoing a MACCE (Table 1). Between groups, there was no difference in anesthetics or opioids used for induction and maintenance of anesthesia (Supplemental Digital Content, Table 2, No study-related adverse events were observed.

WBC and Classical and Intermediate Monocyte Subsets Expanded During Noncardiac Surgery

Figure 1.:
Perioperative changes of WBC, monocyte subsets, natural killer cells, and regulatory T cells during noncardiac surgery in all patients. Leukocyte subpopulations, which have been ascribed a crucial role in cardiovascular disease, were quantified pre-OP, 2 and 6 hours after skin incision and at POD1–2 in 38 patients with coronary artery disease scheduled for noncardiac surgery. Perioperative changes of leukocyte subset counts were assessed using nonparametric Friedman test. To account for multiple comparisons, Bonferroni correction was applied, and P < .008 (α < .05/6) was considered statistically significant. (A) WBC (P < .0001), (B) classical (P < .0001), and (C) intermediate monocytes (P < .0001) expanded during surgery, whereas (D) nonclassical monocytes decreased (P = .004). For these populations, postoperative cell counts (blue) were individually compared to pre-OP values (white) using Dunn’s post hoc test. Indicated P values refer to individual comparisons between postoperative and baseline values. Boldface indicates statistical significance. (E) Natural killer cell (P = .045) and (F) regulatory T cell (P = .012) counts did not change during noncardiac surgery (Friedman test). CD indicates cluster of differentiation; nL−1, per nanoliter blood; POD, postoperative day; pre-OP, preoperative; WBC, white blood cell.

WBCs (P < .0001) and classical (P < .0001), intermediate (P < .0001), and nonclassical monocytes (P = .0038) significantly changed in response to surgery (Figure 1A–D). For these cell populations, postoperative values were compared to baseline counts. Total WBC count increased by 59% between baseline and 6 hours after skin incision (5.48 [3.9–6.65] vs 8.73 [6.94–13.1], pre-OP vs 6 hours, P < .0001) and remained elevated until POD2 (Figure 1A). Likewise, classical and intermediate monocytes expanded within the first 24 hours. Classical monocyte expansion was significant 6 hours after incision (0.35 [0.23–0.43] cells·nL−1 vs 0.38 [0.33–0.6] cells· nL−1, pre-OP vs 6 hours, P = .002) and peaked at POD1 (0.45 [0.31–0.66], P < .0001; Figure 1B). Intermediate monocyte levels were found elevated at POD1 (0.017 [0.013–0.021] vs 0.036 [0.022–0.043] cells· nL−1, pre-OP versus POD1, P = .0003; Figure 1C). Inflammatory nonclassical monocytes decreased within the first 6 hours after incision (0.022 [0.012–0.032] vs 0.012 [0.005–0.023] cells·nL−1, pre-OP vs 6 hours, P = .003) and returned back to baseline values at POD1–2 (Figure 1D). We also assessed perioperative changes of natural killer (P = .045) and regulatory T cell levels (P = .012; Figure 1E, F). However, surgery was not associated with changes in either of these lymphocyte subpopulations, suggesting that the observed leukocyte expansion is a specific reaction restricted to certain subsets rather than a general phenomenon.

ΔClassical Monocytes Predict MACCE in Noncardiac Surgery

Figure 2.:
ΔMonocyte subset counts in association with MACCE. Perioperative monocyte subset counts were determined in 38 elevated-risk patients undergoing noncardiac surgery. Delta (Δ) values represent differences between baseline and postoperative day 1 values. Patients were stratified for no MACCE (n = 33) versus MACCE (n = 5). Two-tailed nonparametric Mann-Whitney U test was used to evaluate statistical significance. Bonferroni correction was applied to adjust for multiple comparisons. α Values <0.017 (P < .05/3) were considered statistically significant (boldface). (A) Patients with MACCE lacked classical monocyte expansion and presented with significantly lower ΔClassical monocyte values compared to patients without a MACCE. We did not detect any relevant differences in (B) ΔIntermediate and (C) ΔNonclassical monocyte values between the groups. CD indicates cluster of differentiation; MACCE, major adverse cardiovascular and cerebrovascular event; nL−1: per nanoliter blood.

To further assess perioperative changes of monocyte subsets in association with perioperative cardiovascular complications, we compared delta (Δ) values (POD1 minus pre-OP values) between patients with and without MACCE (Figure 2). Patients suffering MACCE (0.081 [−0.16 to 0.081] cells·nL−1) had significantly lower ΔClassical monocyte values compared to the event-free group (0.179 [0.081–0.292] cells nL−1, P = .016; Figure 2A). Between groups, we did not find any differences in ΔIntermediate monocytes (Figure 2B) or ΔNonclassical monocyte values (Figure 2C).

Presepsin Increase Is Associated With MACCE

Figure 3.:
Comparison of ΔPresepsin and pre-OP presepsin values between patients with and without MACCE and their association with MACCE. Presepsin plasma concentrations were determined before and 1 day after elective noncardiac surgery. Delta (Δ) values represent differences between baseline values and values at POD 1. Patients were stratified for no MACCE (n = 33) versus MACCE (n = 5). Two-tailed nonparametric Mann-Whitney U test was used to evaluate statistical significance. Patients with MACCE presented with significantly higher (A) ΔPresepsin and (B) pre-OP presepsin values. (C) Receiver operating characteristic curve analysis was performed to evaluate the discriminatory power of preoperative presepsin plasma concentrations. (D) To evaluate the association of preoperative presepsin concentrations >184 pg·mL−1 on secondary outcomes, patients were further stratified for the occurrence of secondary end points. ORs (95% CI)were calculated using Woolf’s method. Boldface indicates statistical significance (P < .05). Pre-OP presepsin concentrations >184 pg·mL−1 threshold were associated with an increased risk of perioperative MACCE, MI, and MINS but not with ischemia, stroke, acute kidney injury, arterial occlusion, atrial fibrillation, or all-cause mortality. AUC indicates area under the curve; MACCE, major adverse cardiovascular and cerebrovascular event; MI, myocardial infarction; MINS, myocardial injury after noncardiac surgery; nL−1, per nanoliter blood; OR, odds ratio; pre-OP, preoperative.

Based on differential cell counts in patients with and without MACCE, we aimed to assess whether perioperative activation of leukocyte subsets is associated with MACCE. To further investigate the perioperative shift from circulating classical to nonclassical monocytes between baseline and POD1, we performed post hoc measurements of the monocyte activation marker presepsin (sCD14-ST). Presepsin concentration increased in every patient between pre-OP and POD1, suggesting increased classical monocyte activation associated with the surgical intervention. The increase was more distinct in patients suffering MACCE (ΔPresepsin 147 [101–333] vs 942 [408–1547] pg·mL−1, no MACCE versus MACCE, P = .002; Figure 3A).

Pre-OP Plasma Presepsin Concentrations Predict MACCE

The pre-OP to POD1 change in presepsin levels (ΔPresepsin) will mainly represent the contribution of perioperative factors to classical monocyte activation potentially indicating increased risk. To assess the preexisting risk of the individual patient, we evaluated pre-OP plasma presepsin concentration for association with MACCE. Median pre-OP plasma presepsin levels were higher in patients experiencing MACCE (1528 [406–1897] pg·mL−1) compared to patients not experiencing MACCE (123 [82.2–174] pg·mL−1, P = .0001; Figure 3B). Receiver operating characteristics curve analysis revealed that pre-OP presepsin is associated with MACCE in elevated-risk patients undergoing noncardiac surgery (AUC = 0.96 [0.85–1], P = .001; Figure 3C). By using Youden’s statistics, we calculated the best presepsin cutoff for prediction of MACCE versus no MACCE to be >184 pg·mL−1 (Table 2).

Table 2.:
Comparison of Test Characteristics for Conventional Risk Factors and Preoperative Presepsin for the Identification of Patients Prone to MACCE

We further compared the prognostic value of presepsin to conventional measures for perioperative cardiovascular risk stratification. Test characteristics as a function of established cutoff values are reported in Table 2. Results indicate that the pre-OP plasma presepsin concentration had the highest and most accurate discrimination (accuracy 84%) compared to conventional means of individual patient-related risk assessments, such as RCRI (accuracy 18%), hs-cTnT (accuracy, 66%; AUC = 0.90 [0.76–0.97]; P = .005), NT-proBNP (accuracy, 55%; AUC = 0.91 [0.77–0.97]; P = .004), and eGFR (accuracy, 74%, AUC = 0.88 [0.72–1]; P = .038). Pre-OP presepsin >184 pg·mL−1 threshold was associated with an increased risk of perioperative MACCE (OR, 47 [2.3–952]; P = .0009) and the secondary end points MI (OR, 23 [1.1–484]; P = .02) and MINS (OR, 6.1 [1.3–28]; P = .024). Other secondary end points were not associated with pre-OP presepsin concentrations (Figure 3D).

Net Reclassification Improvement

Given that perioperative guidelines16 suggest additional biomarker measurements in patients with pre-OP NT-proBNP >300 ng·L−1, we compared the reclassification accuracy of NT-proBNP by pre-OP hs-cTnT or presepsin. In our cohort, all 3 biomarker parameterizations (NT-proBNP, ≥300 ng·L−1, hs-cTnT, ≥14 pg·mL−1, and presepsin, >184 pg·mL−1) reached a sensitivity of 100% for the prediction of MACCE (Table 2). The addition of either presepsin (NRI, 0.33 [0.15–0.49]; P = .001; Supplemental Digital Content, Figure 3A, B, or hs-cTnT (NRI, 0.18 [0.04–0.31]; P = .014; Supplemental Digital Content, Figure 3C, D, measurements to NT-proBNP improved predictive accuracy due to correct reclassifications in the nonevent group. In comparison to hs-cTnT, presepsin led to a 16% higher accuracy improvement of risk classification (28 [74%] vs 22 [58%] correctly classified patients for presepsin versus hs-cTnT, P = .014; Supplemental Digital Content, Figure 3,


In this study, we demonstrate that surgery was associated with a perioperative increase of classical and intermediate monocytes in elevated-risk patients undergoing noncardiac surgery. Data resulting from a post hoc analysis further suggest that a lack of classical monocyte expansion and perioperative plasma presepsin (sCD14-ST) elevation was associated with MACCE and that pre-OP presepsin concentrations >184 pg·mL−1 threshold were associated with an increased risk for MACCE. In our cohort, presepsin was superior to hs-cTnT for the improvement of NT-proBNP-based perioperative prediction of MACCE.

Rupture of destabilized atherosclerotic lesions is seen in about 50% of perioperative MI.22 Monocyte subtypes are ascribed diverse roles in exacerbation of atherosclerotic lesions and promotion of plaque destabilization. While classical and intermediate monocytes exert tissue repair function,23 nonclassical monocytes phagocyte oxidized LDL,24 which is essential for differentiation into lipid-ingesting macrophages promoting plaque destabilization.25

We confirmed our hypothesis stating that noncardiac surgery is related to changes of atherogenic leukocyte subpopulations indicating an inflammatory reaction to the surgical procedure. In our cohort, we only detected classical and intermediate monocyte expansion, but no changes occurred in circulating lymphoid populations during the perioperative period. An increase of classical monocyte levels was restricted to patients not suffering MACCE. Vice versa, a lack of ΔClassical monocyte elevation was associated with MACCE (Figure 2A). On activation, classical monocytes differentiate into proatherogenic nonclassical monocytes via the transitional intermediate state.23,26,27 This finding is supported by our observation of the chronological expansion of classical and intermediate monocytes. Lack of detection of classical and intermediate monocyte elevation in patients suffering MACCE can be explained by actual absence of expansion or increased differentiation of classical and intermediate monocytes into the inflammatory nonclassical subtype and their subsequent trafficking from the circulation into tissue (Figure 4A and B). The conversion of classical into nonclassical monocytes is accompanied by shedding of CD14.28,29 Soluble CD14 is further processed to sCD14 subtype (sCD14-ST)30 called presepsin.12 Presepsin thereby largely reflects monocyte activation.28 Elevated ΔPresepsin in patients experiencing a MACCE corroborates our hypothesis by giving indirect evidence for elevated monocyte turnover28 in these patients undergoing a MACCE compared to individuals without MACCE (Figure 3A). Thus, increased turnover and rapid removal of nonclassical monocytes might obscure monocyte expansion.

Figure 4.:
Proposed model for increased classical to nonclassical monocyte turnover resulting in elevated presepsin concentrations in patients prone to MACCE. Steps highlighted in yellow are supported by published evidence. Blue and red circles indicate findings described in this report and conclusions drawn from the reported data, respectively. (A) (1) During homeostasis, classical monocytes are continuously mobilized from the bone marrow and spleen into the circulation.40 (2) On activation, they give rise to inflammatory, proatherogenic nonclassical monocytes via an intermediate state.23 , 26 , 27 (3) During differentiation toward nonclassical monocytes, CD14 is shed from classical monocytes29 and is further processed to presepsin, which reflects monocyte activation.28 Preoperatively, monocyte subpopulation counts did not differ between stable and vulnerable patients. (4) However, we found that vulnerable patients showed higher preoperative presepsin levels, indicating increased classical monocyte activation (boldface arrows). (B) (5) In response to surgery, more classical monocytes were mobilized from their reservoirs. (6) However, only stable patients presented with increased classical and intermediate monocyte counts. (7) In vulnerable patients, postoperative monocyte subset counts remained unchanged. (8) Compared to preoperative values, all patients showed elevated presepsin concentrations at postoperative day 1. Although vulnerable patients did not show any perioperative change in monocyte subpopulations, they had markedly higher postoperative presepsin levels compared to stable patients, pointing toward an increased classical to nonclassical monocyte turnover in vulnerable patients. (9) We hypothesize that nonclassical monocyte expansion in blood was masked by rapid removal from the circulation into tissue. Infiltration of nonclassical monocytes into atherosclerotic lesions contributes to plaque destabilization and rupture,24 , 25 thereby precipitating MACCEs. CD indicates cluster of differentiation; MACCE, major adverse cardiovascular and cerebrovascular event.

Plasma presepsin has originally been proposed as a marker for systemic inflammation and septic shock.12 In recent years, the prognostic and diagnostic value of presepsin has been described for various inflammatory diseases, including pancreatitis,31 skin wound infections,32 and systemic lupus erythematosus.33 Moreover, presepsin was found to be associated with inflammatory cardiovascular disorders such as acute MI34 and mesenteric ischemia.35 Recently, Bomberg et al13 proposed presepsin as a biomarker for the prediction of mortality after cardiac surgery. In a post hoc analysis, we investigated the association of presepsin with major adverse vascular complications after noncardiac surgery and found a significant correlation between elevated pre-OP presepsin concentrations >184 pg·mL−1 and 30-day MACCE. In addition, elevated presepsin was associated with an increased risk for perioperative MI and MINS. This study adds to the literature that pre-OP presepsin is valuable for the prediction of MACCE in patients undergoing noncardiac surgery.

Based on the sample size calculated to study the effect of surgery on circulating leukocytes, we recruited a relatively small number of 38 patients. Only 5 patients reached the primary end point MACCE. Therefore, conclusions from the post hoc analysis should be interpreted with caution and need to be considered as explorative rather than confirmatory.

Renal failure affects presepsin plasma levels.36 In our cohort, patients experiencing MACCE presented with increased pre-OP creatinine and reduced eGFR, and preexisting renal impairment was associated with perioperative MACCE. The analysis set of 38 patients limits the power of our study and renders it insufficient for multivariable analyses that would allow considering additional risk factors such as renal impairment. The test characteristic confirms an association between reduced kidney function and perioperative MACCE. However, this association appeared less strong compared to pre-OP presepsin. Furthermore, hs-cTnT and NT-proBNP are likewise affected by reduced kidney function. In addition, perioperative acute renal failure was not associated with presepsin (Figure 3D). As the majority of our patients are male, generalization of our findings is difficult. To include a broad spectrum of surgeries, we recruited patients scheduled for various types of interventions, including low-, intermediate-, and high-risk surgery. General anesthesia might affect perioperative inflammatory responses.37,38 For the majority of patients, standardized anesthetic plans were used, and no differences between groups regarding the anesthetics used to induce or maintain anesthesia were observed (Supplemental Digital Content, Table 2, Therefore, it is unlikely that potential anesthetic- or opioid-mediated effects on circulating blood cells confound our findings. Endogenous and exogenous catecholamines contribute to the regulation of inflammation.39 Because no patient received catecholamines prior to surgery, catecholamine-mediated effects on pre-OP presepsin levels can be excluded. However, we cannot fully exclude effects on perioperative leukocyte subsets.

This is the first prospective study to provide insights into perioperative changes of proatherogenic leukocyte subpopulations associated with adverse perioperative vascular outcomes. Quantification of pre-OP and postoperative leukocytes combined with presepsin findings suggest that increased classical monocyte differentiation toward inflammatory nonclassical monocytes is associated with MACCE. While other biomarkers, such as hs-cTnT or NT-proBNP, identify patients with preexisting myocardial damage or heart failure, presepsin identifies patients with elevated formation of inflammatory nonclassical monocytes, which might play a causative role for perioperative cardiovascular events.

Identification of elevated-risk patients is crucial for awareness of potential perioperative risks and to improve analysis and communication of benefit-to-risk ratio. Initiation of pre-OP cardiovascular optimization, resource allocation, and adjustment of postoperative monitoring in extent and duration might improve individual patient outcome. In the future, modulation of pre-OP inflammatory cell levels could comprise a promising new target for the prevention of cardiovascular events. Additional studies are needed to elucidate the role of leukocyte subpopulations and their activation in perioperative cardiovascular disease.

Recent guidelines by the Canadian Cardiovascular Society on perioperative cardiac risk assessment suggest measuring NT-proBNP in patients with RCRI ≥ 2 to guide perioperative patients’ pathways.16 We demonstrate that baseline presepsin improved the prediction of perioperative MACCE and reduced false-positive classification based on NT-proBNP. For risk prediction, adding presepsin to the NT-proBNP model was superior to hs-cTnT as a second biomarker. In conclusion, this study suggests that increased pre-OP plasma presepsin is strongly associated with 30-day MACCE in elevated cardiovascular risk patients undergoing elective noncardiac surgery. After validation in an independent data set, a presepsin cutoff of 184 pg·mL−1 might qualify to complement NT-proBNP-based risk prediction, thereby increasing the proportion of correctly identified high-risk patients and, in turn, improving individual patient outcome.


We are deeply indebted to Manuela Schwegler, Department of Anesthesiology, Heidelberg University Hospital, Heidelberg, Germany, for her constant administrative support. Elements of Figure 4 were taken and adjusted from Servier Medical Art at, licensed under a Creative Commons Attribution 3.0 Unported License.


Name: Jessica Handke, MSc.

Contribution: This author helped design the study, conduct the study, analyze the data, and prepare the manuscript.

Conflicts of Interest: None.

Name: Anna S. Scholz.

Contribution: This author helped design and conduct the study.

Conflicts of Interest: None.

Name: Hans-Jörg Gillmann, MD.

Contribution: This author helped analyze the data and prepare the manuscript.

Conflicts of Interest: None.

Name: Henrike Janssen, MD.

Contribution: This author helped conduct the study and prepare the manuscript.

Conflicts of Interest: None.

Name: Sarah Dehne, MD.

Contribution: This author helped conduct the study and prepare the manuscript.

Conflicts of Interest: None.

Name: Christoph Arens, MD.

Contribution: This author helped conduct the study and prepare the manuscript.

Conflicts of Interest: None.

Name: Laura Kummer, MSc.

Contribution: This author helped conduct the study and prepare the manuscript.

Conflicts of Interest: None.

Name: Florian Uhle, PhD.

Contribution: This author helped establish sample analyses and prepare the manuscript.

Conflicts of Interest: None.

Name: Markus A. Weigand, MD.

Contribution: This author helped design the study, conduct the study, and prepare the manuscript.

Conflicts of Interest: None.

Name: Johann Motsch, MD.

Contribution: This author helped design the study, conduct the study, and prepare the manuscript.

Conflicts of Interest: None.

Name: Jan Larmann, MD, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and prepare the manuscript.

Conflicts of Interest: J. Larmann received speaker fees from Mitsubishi Chemical Europe.

This manuscript was handled by: Alexander Zarbock, MD.


1. Cohn SL, Fernandez Ros N. Comparison of 4 cardiac risk calculators in predicting postoperative cardiac complications after noncardiac operations. Am J Cardiol. 2018;121:125–130.
2. Fleisher LA. The value of preoperative assessment before noncardiac surgery in the era of value-based care. Circulation. 2017;136:1769–1771.
3. Swirski FK, Nahrendorf M. Leukocyte behavior in atherosclerosis, myocardial infarction, and heart failure. Science. 2013;339:161–166.
4. Janssen H, Wagner CS, Demmer P, et al. Acute perioperative-stress-induced increase of atherosclerotic plaque volume and vulnerability to rupture in apolipoprotein-E-deficient mice is amenable to statin treatment and IL-6 inhibition. Dis Model Mech. 2015;8:1071–1080.
5. Larmann J, Frenzel T, Schmitz M. In vivo fluorescence-mediated tomography imaging demonstrates atorvastatin-mediated reduction of lesion macrophages in ApoE-/- mice. Anesthesiology. 2013;119:129–141.
6. Li C, Engström G, Hedblad B. Leukocyte count is associated with incidence of coronary events, but not with stroke: a prospective cohort study. Atherosclerosis. 2010;209:545–550.
7. Björkbacka H, Berg KE, Manjer J, et al. CD4+ CD56+ natural killer T-like cells secreting interferon-γ are associated with incident coronary events. J Intern Med. 2016;279:78–88.
8. Berg KE, Ljungcrantz I, Andersson L, et al. Elevated CD14++CD16- monocytes predict cardiovascular events. Circ Cardiovasc Genet. 2012;5:122–131.
9. Rogacev KS, Cremers B, Zawada AM, et al. CD14++CD16+ monocytes independently predict cardiovascular events: a cohort study of 951 patients referred for elective coronary angiography. J Am Coll Cardiol. 2012;60:1512–1520.
10. Urra X, Villamor N, Amaro S, et al. Monocyte subtypes predict clinical course and prognosis in human stroke. J Cereb Blood Flow Metab. 2009;29:994–1002.
11. Wigren M, Björkbacka H, Andersson L, et al. Low levels of circulating CD4+FoxP3+ T cells are associated with an increased risk for development of myocardial infarction but not for stroke. Arterioscler Thromb Vasc Biol. 2012;32:2000–2004.
12. Yaegashi Y, Shirakawa K, Sato N, et al. Evaluation of a newly identified soluble CD14 subtype as a marker for sepsis. J Infect Chemother. 2005;11:234–238.
13. Bomberg H, Klingele M, Wagenpfeil S, et al. Presepsin (sCD14-ST) is a novel marker for risk stratification in cardiac surgery patients. Anesthesiology. 2017;126:631–642.
14. Cohen B, Dery E, Cattan A, Matot I. Is leukocytosis a common finding in the postoperative period? Paper presented at: American Society of Anesthesiologists' 2013 Annual Meeting; October 14, 2013; San Francisco, CA.
15. Levey AS, Eckardt KU, Tsukamoto Y, et al. Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int. 2005;67:2089–2100.
16. Duceppe E, Parlow J, MacDonald P, et al. Canadian cardiovascular society guidelines on perioperative cardiac risk assessment and management for patients who undergo noncardiac surgery. Can J Cardiol. 2017;33:17–32.
17. Kurihara T, Yanagida A, Yokoi H, et al. Evaluation of cardiac assays on a benchtop chemiluminescent enzyme immunoassay analyzer, PATHFAST. Anal Biochem. 2008;375:144–146.
18. Youden WJ. Index for rating diagnostic tests. Cancer. 1950;3:32–35.
19. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157–172.
20. Kim D, Han H-S, Yoon Y-S, et al. Postoperative white blood cell counts change after pancreatoduodenectomy: early sign for pancreatic fistula. Korean J Clin Oncol. 2015;11:95–100.
21. Lehmann EL. Nonparametrics: Statistical Methods Based on Ranks, Revised. 1998:London, UK: Pearson Education; 76–81.
22. Dawood MM, Gutpa DK, Southern J, Walia A, Atkinson JB, Eagle KA. Pathology of fatal perioperative myocardial infarction: implications regarding pathophysiology and prevention. Int J Cardiol. 1996;57:37–44.
23. Wong KL, Tai JJ, Wong WC, et al. Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets. Blood. 2011;118:e16–e31.
24. Mosig S, Rennert K, Krause S, et al. Different functions of monocyte subsets in familial hypercholesterolemia: potential function of CD14+ CD16+ monocytes in detoxification of oxidized LDL. FASEB J. 2009;23:866–874.
25. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352:1685–1695.
26. Zawada AM, Rogacev KS, Rotter B, et al. SuperSAGE evidence for CD14++CD16+ monocytes as a third monocyte subset. Blood. 2011;118:e50–e61.
27. Merino A, Buendia P, Martin-Malo A, Aljama P, Ramirez R, Carracedo J. Senescent CD14+CD16+ monocytes exhibit proinflammatory and proatherosclerotic activity. J Immunol. 2011;186:1809–1815.
28. Shive CL, Jiang W, Anthony DD, Lederman MM. Soluble CD14 is a nonspecific marker of monocyte activation. AIDS. 2015;29:1263–1265.
29. Bazil V, Strominger JL. Shedding as a mechanism of down-modulation of CD14 on stimulated human monocytes. J Immunol. 1991;147:1567–1574.
30. Chenevier-Gobeaux C, Borderie D, Weiss N, Mallet-Coste T, Claessens YE. Presepsin (sCD14-ST), an innate immune response marker in sepsis. Clin Chim Acta. 2015;450:97–103.
31. Lin J, Li Z, Zheng Y, et al. Elevated presepsin levels are associated with severity and prognosis of severe acute pancreatitis. Clin Lab. 2016;62:1699–1708.
32. Shiota J, Tagawa H, Ohura N, Kasahara H. Presepsin is a potent biomarker for diagnosing skin wound infection in hemodialysis patients compared to white blood cell count, high-sensitivity C-reactive protein, procalcitonin, and soluble CD14. Ren Replace Ther. 2017;331.
33. Mukherjee R, Kanti Barman P, Kumar Thatoi P, Tripathy R, Kumar Das B, Ravindran B. Non-classical monocytes display inflammatory features: validation in sepsis and systemic lupus erythematous. Sci Rep. 2015;5:13886.
34. Caglar FNT, Isiksacan N, Biyik I, Opan S, Cebe H, Akturk IF. Presepsin (sCD14-ST): could it be a novel marker for the diagnosis of ST elevation myocardial infarction? Arch Med Sci Atheroscler Dis. 2017;2:e3–e8.
35. Stroeder J, Bomberg H, Wagenpfeil S, et al. Presepsin and inflammatory markers correlate with occurrence and severity of nonocclusive mesenteric ischemia after cardiovascular surgery. Crit Care Med. 2018;46:e575–e583.
36. Nagata T, Yasuda Y, Ando M, et al. Clinical impact of kidney function on presepsin levels. PLoS One. 2015;10:e0129159.
37. Yeager MP, Procopio MA, DeLeo JA, Arruda JL, Hildebrandt L, Howell AL. Intravenous fentanyl increases natural killer cell cytotoxicity and circulating CD16(+) lymphocytes in humans. Anesth Analg. 2002;94:94–99.
38. Rossaint J, Zarbock A. Perioperative inflammation and its modulation by anesthetics. Anesth Analg. 2018;126:1058–1067.
39. Padro CJ, Sanders VM. Neuroendocrine regulation of inflammation. Semin Immunol. 2014;26:357–368.
40. Swirski FK, Nahrendorf M, Etzrodt M, et al. Identification of splenic reservoir monocytes and their deployment to inflammatory sites. Science. 2009;325:612–616.

Supplemental Digital Content

Copyright © 2018 International Anesthesia Research Society