Circulating endothelial microparticles: a promising biomarker of acute kidney injury after cardiac surgery with cardiopulmonary bypass.

BACKGROUND: Current diagnostic strategies for acute kidney injury (AKI) after cardiac surgery with cardiopulmonary bypass (CPB) are nonspecific and limited. Previously, we demonstrated that circulating microparticles (MPs) in patients with valve heart disease (VHD) and congenital heart diseases (CHD) induce endothelial dysfunction and neutrophil chemotaxis, which may result in kidney injury. We also found that circulating MPs increase after cardiac surgery with CPB and are related to cardiac function. However, the relationship between circulating MPs and AKI after CPB is unknown.
METHODS: Eighty-five patients undergoing cardiac surgery with CPB were enrolled. Patients were divided into AKI and non-AKI groups based on the serum creatinine levels at 12 h and 3 d post-CPB. Circulating MPs were isolated from plasma, and their levels including its subtypes were detected by flow cytometer. Independent risk factors for the CPB-associated AKI (CPB-AKI) were determined by multivariate logistic regression analysis. Receiver operating characteristic (ROC) analysis was used to measure the prognostic potential of CPB-AKI.
RESULTS: The morbidity of AKI at 12 h and 3 d after cardiac surgery with CPB was 40% and 31.76%, respectively. The concentrations of total MPs and platelet-derived MPs (PMP) remained unchanged at 12 h and then increased at 3 d post-CPB, while that of endothelial-derived MPs (EMP) increased at both time points. In patients with AKI, PMP and EMP were elevated compared with the patients without AKI. However, no significant change was detected on monocyte-derived MPs (MMP) at 12 h and 3 d post-CPB. The logistic regression analysis showed that EMP was the independent risk factor for AKI both at 12 h and 3 d post-CPB. The area under ROC for the concentrations of EMP at 12 h and 3 d post-CPB was 0.86 and 0.91, with the specificity up to 0.88 and 0.91, respectively.
CONCLUSIONS: Circulating EMP may serve as a potential biomarker of AKI after cardiac surgery with CPB. 2021 Annals of Translational Medicine. All rights reserved.

Introduction

Acute kidney injury (AKI) is a common complication in patients undergoing cardiac surgery with cardiopulmonary bypass (CPB), with morbidity up to 30% (1). CPB-associated AKI (CPB-AKI) increases the need for mechanical ventilation, days of intensive care, and mortality (2). In CPB, aorta cross-clamping and non-pulsatile blood flow contribute to the inflammatory responses and oxidative stress injury, which may induce CPB-AKI (3). The early diagnosis of CPB-AKI is difficult, which is a challenge for the therapeutic strategy (3). Nowadays, microparticles (MPs), especially its subtype and endothelial-derived microparticles (EMP) from circulation are becoming a promising diagnostic biomarker for AKI (4).

Circulating MPs ranging from 100 to 1,000 nm are derived from circulating cells upon activation (5). Concentrations of MPs are closely associated with cardiovascular disease such as hypercholesterolemia, atherosclerosis, congenital heart diseases (CHD), and valve heart disease (VHD), and increase after cardiac surgery with CPB (5-10). More importantly, a nanoparticle tracking analysis reveals a higher concentration of circulating MPs in patients with acute heart failure at 12 h after cardiac surgery with CPB (10). Circulating MPs are distinguished based on the source cells such as platelet-derived MP (PMP), EMP, and monocyte-derived MP (MMP) (11,12). Recent studies found high concentrations of MPs after CPB (10,13,14). However, the relationship between CPB-AKI and circulating MPs, especially its subpopulations, remains unclear.

More evidence show that microparticles mediate inflammation, coagulation, oxidative stress injury, and endothelial dysfunction in patients with renal disease and become a promising biomarker of various renal diseases such as thrombotic microangiopathies, nephrotic syndrome, IgA nephropathy, and renal failure (4,15). In view of molecular mechanism, MPs, especially EMP, induce endothelial dysfunction by increasing the release of inflammatory cytokines and reducing nitric oxide (5-8,16-18). Recently, we found that circulating MPs from patients with VHD induce chemokines expression and renal neutrophil chemotaxis, which partly contribute to renal dysfunction (18). Recent investigations on the correlation between circulating EMP and renal dysfunction found that EMP are elevated in patients with chronic kidney disease (CKD) or septic patients with AKI (19). However, the relationship between circulating EMP and AKI after CPB also remains unknown.

We hypothesized that the activation of the endothelium during CPB could lead to the release of EMP, and the elevated EMP are correlated with kidney injury and may be a diagnostic biomarker for CPB-AKI. Thus, flow cytometry was performed to analyze circulating MPs and its sub-population including EMP, PMP, and MMP in patients undergoing cardiac surgery with CPB and the correlations between MPs especially EMP and AKI were analyzed.

We present the following article in accordance with the STARD reporting checklist (available at http://dx.doi.org/10.21037/atm-20-7828).

Methods

Study cohort

In this retrospective observational study, 85 adult patients, diagnosed with heart disease undergoing cardiac surgery with CPB were enrolled from the First Affiliated Hospital, Sun Yat-sen University. Patients with trauma, shock, sepsis, chronic pulmonary disease, cancer, and severe hepatic dysfunction were excluded from the study. The patients were assigned into AKI and non-AKI groups based on the serum creatinine levels at 12 h and 3 d post-CPB according to the Kidney Disease Improving Global Guidelines (KDIGO) Clinical Practice Guidelines for AKI (20). Summarily, serum creatinine increased in patients after operation by ≥26.5 μmol/L within 48 h or ≥1.5 times baseline level occurred within 7 days prior to the operation are diagnosed with AKI based on KIDGO. This study complied with the declaration of Helsinki (as revised in 2013) and was also approved by the Ethics Review Board of the First Affiliated Hospital, Sun Yat-sen University [(2011)14]. All subjects recruited in this study signed the written informed consent.

Collection of clinical information

The clinical data of patients enrolled in this study including age, sex, weight, height, diabetes mellitus (DM), hypertension, cardiac disease history (CDH), and personal history (drinking and smoking) were collected. The preoperative data such as ultrasonography of the heart, serum creatinine, uric acid, and blood routine, were collected alongside the intraoperative information: subtypes of cardiac surgery, operation time, CPB time and cross clamp time, and serum creatinine at the 12 h and 3 d post-CPB.

MPs isolation

MPs were isolated from enrolled patients by differential centrifugation as previously reported (5,10,18). According to the methodological guidelines to study MPs (21), intravenous blood was drawn by the vacuum citrate-anticoagulant tubes and the first tube was discarded to avoid the influence of blood injury induced by venipuncture. The blood was then centrifuged at 2,000 g for 20 min at 4 °C twice to discard the deposits of blood cells within 1 h. The plasma was gently absorbed at 0.5 cm above the hemocyte layer to reduce the pollution of hemocytes. Afterwards, the platelet-rich plasma was centrifuged at 11,000 g at 4 °C for 2 min to pellet platelets. The upper platelet-poor plasma was stored at −80 °C before use.

Flow cytometry (FCM)

The platelet-poor plasma was thawed in the water bath at 37 °C and kept at room temperature (RT) before detection. The measurement of MPs and its subpopulations was done as previously done by FCM (18). The plasma was incubated with AnnexinV-FITC with binding buffer for 30 min in the dark at RT to stain the total MPs. EMP, PMP, and MMP were labeled with specific biomarkers of their source cell such as CD144 for endothelial cells, CD41 for platelets and CD14 for monocytes, respectively. The plasma was then divided into three samples and incubated with CD144 conjugated with PE, CD41 conjugated with percpcy5.5 and CD14 conjugated with APC, individually. The plasma was incubated with a final volume of 300 µL at RT for 30 min. To quantity the MPs, the Megmix-Plus SSC beads with four diameters (160, 200, 240, and 500 nm) were performed to set MPs gate according to the instruction. The fluorescent microspheres with 3.8 µm in diameter were added before being detected by the high-sensitivity flow cytometer (Cytoflex S, Beckman coulter).

Statistical analyses

Continuous variables were described as mean ± standard deviation (mean ± SD) unless specially illustrated and the categorical variables were represented as frequency (N). For our risk factor analysis, missing values were stochastically imputed using a multiple imputation method. The comparison of concentrations of MPs and its subpopulations were performed with the Welch’s test. The other continuous variables were compared with unpaired t-test. Categorical variables were compared using χ2 analysis.

Covariant screening was used to screen for possible confounders. The screening criteria included factors that changes > by 10% when covariates are introduced into the underlying model or when covariates are removed from the regression model. Univariate analysis was conducted for the potential risk factors of AKI. A systemic search was conducted to exclude highly correlated risk factors. If two variables had a high Pearson correlation, the one with the largest average absolute correlation was removed from consideration. After univariate analysis, the significant factors were introduced into the multivariate logistic analysis to find out the independent risk factors.

The receiver operating characteristic curve (ROC) and area under curve (AUC) were calculated to evaluate the diagnostic value of the promising biomarkers for AKI. AUC confidence interval and significance test were calculated using nonparametric repeated sampling method (Bootstrap resampling times =500).

The statistical software SPSS 17.0 and R3.6.3 was used for data analysis, and P<0.05 was considered statistically significant.

Results

Demographic and perioperative clinical characteristics

The morbidity of AKI at 12 h (12 h post-CPB) and 3 d (3 d post-CPB) after cardiac surgery with CPB was 40% and 31.76% respectively as described in . The perioperative characteristics including gender, weight, height, BMI, smoking, CDH, hypertension, hyperlipidemia (HLP), ejection fraction (EF)% and the grade of New York Heart Association (NYHA) and the levels of creatinine before surgery were not significantly changed between non-AKI and AKI groups at 12 h post-CPB or 3 d post-CPB (). The percentages of old age and history of drinking were higher in AKI at 12 h post-CPB, but not at 3 d post-CPB (). Compared with non-AKI group, the underlying disease DM was found in the patients with AKI at 3 d post-CPB. The occurrence of AKI at 12 h post-CPB seems to be less related with the types of cardiac surgery, and the degree of surgical complexity was associated with AKI at 3 d post-CPB (). As expected, the CPB time was increased in AKI both at 12 h post-CPB and 3 d post-CPB groups than non-AKI groups, and similar changes were found in the operation time of cardiac surgery and cross clamp time ().

Table 1

Demographic and perioperative characteristics between AKI and Non-AKI group at 12 h and 3 d post-CPB

	AKI-12 h after operation				AKI-3 d after operation
	No	Yes	P		No	Yes	P
Case	51 (60%)	34 (40%)			58 (68.24%)	27 (31.76%)
Age (years)	54.12±11.31	59.94±8.17	0.011		55.45±10.81	58.59±9.69	0.201
Gender (male/female)	26/25	22/12	0.211		33/25	15/12	0.908
Weight (kg)	60.17±10.66	64.44±10.67	0.074		61.4±11.07	62.91±10.33	0.55
Height (cm)	159.42±8.99	160.85±6.67	0.430		160.02±8.85	159.94±6.45	0.969
BMI (kg/m2)	23.60±3.33	24.87±3.64	0.102		23.91±3.55	24.54±3.38	0.444
Drinking (yes/no)	3/48	7/27	0.039		7/51	3/24	0.898
Smoking (yes/no)	10/41	13/21	0.058		16/42	7/20	0.837
DM (yes/no)	3/48	6/28	0.084		2/56	7/20	0.002
CDH (yes/no)	3/48	1/33	0.53		3/55	1/26	0.766
Hypertension (No/I/II/III)	41/2/5/3	18/3/8/5	0.064		45/2/7/4	14/3/6/4	0.108
HLP (yes/no)	10/41	11/23	0.182		11/47	10/17	0.072
EF (%)	64.18±13.18	60.74±14.05	0.253		62.8±14.35	62.81±11.94	0.996
NYHA (I/II/III/IV)	3/30/16/2	3/14/16/1	0.413		5/30/21/2	1/14/11/1	0.865
Creatinine
Before operation	79.02±24.99	84.71±25.29	0.309		79.21±18.88	85.78±35.01	0.264
12 h after operation	85.88±25.97	138.03±32.62	<0.001		92.97±28.03	136.33±41.55	<0.001
3 d after operation	–	–			73.10±22.33	151.33±50.94	<0.001
Operation			0.053				0.010
VR	31	17			38	10
CABG	8	6			7	7
VR+CABG	3	4			3	4
Aortic operation	1	4			1	4
Other	8	3			9	2
Operation time (min)	329.29±119.9	458.32±172.3	<0.001		334.65±115.34	480.26±184.94	<0.001
CPB time (min)	142.92±54.61	229.62±100.64	<0.001		145.84±54.22	245.81±104.69	<0.001
Cross clamp time	82.39±40.6	133.47±58.6	<0.001		85.74±39.78	139.52±63.70	<0.001

Data are presented as mean ± SD or number of patients (%). AKI, acute kidney injury; CPB, cardiopulmonary bypass; BMI, body mass index; EF, ejection fraction; DM, diabetes mellitus; CDH, cardiac disease history; HLP, hyperlipoidemia; NYHA, New York Heart Association; VR, valve replacement /repair; CABG, coronary artery bypass grafting.

The circulating MPs, especially EMP, elevated after CPB

Previously, we demonstrated that the circulating MPs were increased in patients with VHD (5,10). To explore the change of MPs and its subpopulations during the perioperative period of cardiac surgery, the association with CPB and FCM were performed to measure the levels of MPs and its subtypes.

The concentration of circulating MPs did not significantly change at 12 h post-CPB compared with pre-CPB. However, MPs were elevated at 3 d post-CPB (). The PMP, most abundant MPs in circulation, showed similar changes with MPs, and were elevated at 3 d post-CPB (). The CPB significantly reduced the level of MMP at 12 h post-CPB compared with pre-CPB, and the MMP slightly returned at 3 d post-CPB, especially compared with 12 h post-CPB (). The EMP was significantly elevated at 12 h post-CPB, and their release was slightly lower at 3 d post-CPB compared with 12 h post-CPB, but still higher than that of preoperative values ().

Figure 1

Cardiac surgery with CPB changes the levels of circulating MPs and its subpopulation. The plasma was isolated form enrolled patients at different times (per-operation, 12 h post-CPB, and 3 d post-CPB). The thawed plasma was incubated with different biomarkers to identify its source and detected by flow cytometry. Data were presented as mean ± standard deviation (mean ± SD); &, vs. pre-CPB; #, vs. 12 h post-CPB, P<0.05. CPB, cardiopulmonary bypass; MPs, circulating microparticles; PMP, platelet-derived microparticles; MMP, monocyte-derived microparticles; EMP, endothelial-derived microparticles.

EMP elevated in AKI at 12 h post-CPB

To further investigate the correlation of the subtypes of circulating MPs with AKI after cardiac surgery, the subpopulations of MPs were labeled with specific markers of source cells such as CD144 for EMP, CD41 for PMP, and CD14 for MMP. illustrates the changes in the concentration of the subpopulations of circulating MPs at three time points (pre-CPB, 12 h post-CPB, and 3 d post-CPB) in the AKI and non-AKI groups occurred at 12 h and 3 d post-CPB.

Figure 2

Changes in the concentrations of circulating EMP, MMP, and PMP in the AKI and non-AKI group post-CPB. The enrolled patients were divided into the AKI and non-AKI groups according to the KIDGO at 12 h and 3 d post-CPB and the levels of subtypes of circulating MPs (EMP, MMP, and PMP) detected by a flow cytometry were compared. (A) The bar chart shows the levels of circulating PMP in the patients with AKI group occurred at 12 h post-CPB and non-AKI group; (B) the bar chart shows the levels of circulating PMP in the patients with AKI group occurred at 3 d post-CPB and non-AKI group; (C,D) the bar chart described the changes of MMP in the AKI at 12 h (C) and 3 d (D) post-CPB and non-AKI; (E,F) the alteration of EMP in the patients undergoing cardiac surgery with or without AKI at 12 h (E) and 3 d (F) post-CPB. Data were presented as mean ± standard deviation (mean ± SD). &, vs. non-AKI-12 h group; *, vs. non-AKI-3 d group, P<0.05. CPB, cardiopulmonary bypass; PMP, platelet-derived microparticles; MMP, monocyte-derived microparticles; EMP, endothelial-derived microparticles.

In patients with AKI at 12 h post-CPB (AKI-12 h), the concentration of PMP was not altered at pre-CPB but elevated at 12 h post-CPB (). Interestingly, in the patients with AKI it occurred at 3 d post-CPB (AKI-3 d). The concentration of PMP at pre-CPB was not changed, but increased at both 12 h and 3 d post-CPB compared with the non-AKI group ().

Compared with the non-AKI patients, no significant change of levels of MMP at two different time points (pre-CPB, and 12 h post-CPB) were detected in the patients with AKI occurred at 12 h post-CPB (). There was also no difference in the MMP levels in the perioperative period between AKI at 3 d post-CPB and the non-AKI group ().

The EMP at pre-CPB showed no difference between AKI and non-AKI groups at 12 h or 3 d post-CPB (). Compared with the patients without AKI, the EMP levels at 12 h post- CPB showed a remarkable increase in the AKI group at 12 h post-CPB (). shows the concentration of EMP at 12 h and 3 d post-CPB elevated in the patients with AKI at 3 d post-CPB than in the non-AKI ().

Circulating EMP as a potential risk factor in predicting if AKI occurred post CPB

describes the associated factors with AKI at 12 h and 3 d post-CPB. The univariate analysis for AKI at 12 h post-CPB illustrated that the operation time, CPB time, cross clamp time, and the levels of PMP and EMP at 12 h post-CPB led to the elevated risk of AKI at 12 h post-CPB (). Further multivariate logistic analysis indicated that the concentrations of EMP and PMP were independent risk factors for AKI at 12 h post-CPB, and the EMP was the most promising biomarker for AKI at 12 h post-CPB ().

Table 2

Univariate and multivariate analysis of potential risk factors for AKI occurred after operation with CPB

	Univariate			Multivariate§
	OR (95% CI)	P		OR (95% CI)	P
Postoperative 12 h
Age	1.06 (1.01–1.11)	0.015
Gender	0.57 (0.15–1.26)	0.213
Drinking	4.15 (0.99–17.37)	0.051
Operation time	1.01 (1–1.01)	0.001
CPB time	1.02 (1.01–1.02)	<0.001		No select
Cross clamp time	1.02 (1.01–1.03)	0.001
PMP-12h†	1.13 (1.05–1.21)	0.001		1.11 (1.01–1.22)	0.023
EMP-12h‡	2.77 (1.86–4.11)	<0.001		2.49 (1.61–3.85)	<0.001
Postoperative 3 d
Age	1.03 (0.98–1.08)	0.201
Gender	1.06 (0.42–2.65)	0.908
Diabetes mellitus	9.8 (1.88–51.15)	0.007
Operation time	1 (1.00–1.01)	<0.001
CPB time	1.02 (1.01–1.03)	<0.001		No select
Cross clamp time	1.02 (1.01–1.03)	<0.001
PMP-3d†	1.13 (1.07–1.2)	<0.001		1.06 (0.99–1.14)	0.106
EMP-3d‡	4.05 (2.26–7.25)	<0.001		3.18 (1.69–5.97)	<0.001

†, variable units is ×106 particles/mL. ‡, variable units is ×105 particles/mL. §, adjust model adjust for: drinking and cross clamp time at postoperative 12 h, operation time and cross clamp time at postoperative 3 d. AKI, acute kidney injury; CPB, cardiopulmonary bypass. PMP-12h, platelet-derived microparticles at 12 h post-CPB; EMP-12h, endothelial-derived microparticles at 12 h post-CPB; PMP-3d, platelet-derived microparticles at 3 d post-CPB and EMP-3d, endothelial-derived microparticles at 3 d post-CPB. OR, odds ratio; CI, confidence interval.

To further explore the risk factors for AKI which occurred at 3 d post-CPB, the logistic analysis was performed and similar results to AKI at 12 h post-CPB () were found. The univariate analysis found that DM, operation time, CPB time, cross clamp time, and EMP and PMP at 3 d post-CPB were the associated factors with AKI which occurred at 3 d post-CPB (). The next multivariate logistic analysis also showed that the EMP, not PMP at 3 d post-CPB correlated with AKI ().

Diagnostic value of EMP for AKI occurred at post-CPB

The ROC curve was performed to detect the diagnostic value of EMP for AKI at post-CPB. The AUC for the counts of EMP at 12 h post-CPB was 0.86 (). shows that the best cut-off valve of EMP in diagnosing AKI at 12 h post-CPB was 3.99×105 particles/mL, and the sensitivity and specificity were 0.88 and 0.88, respectively ().

Figure 3

Receiver operating characteristic curve analysis of EMP levels. Receiver operating characteristic (ROC) curve analysis of EMP levels. AUC confidence interval and significance test using a nonparametric repeated sampling method (Bootstrap resampling times =500). (A) ROC for AKI occurred at 12 h post-CPB (AKI-12 h); (B) ROC for AKI at 3 d post-CPB (AKI-3d). EMP, endothelial-derived microparticles; AKI, acute kidney injury; CPB, cardiopulmonary bypass.

Table 3

Parameters of ROC curve for AKI happened after operation with CPB

	AUC (95% CI)	Cut-off value (×105 particles/mL)	At cut-off value		Accuracy
			Sensitivity	Specificity
EMP-12h	0.86 (0.77–0.96)	3.99	0.88	0.88	0.89
EMP-3d	0.91 (0.83–0.97)	4.56	0.85	0.91	0.89

ROC, receiver operating characteristic; AKI, acute kidney injury; CPB, cardiopulmonary bypass; AUC, area under the curve; EMP-12h, endothelial-derived microparticles at 12 h post-CPB; EMP-3 d, endothelial-derived microparticles at 3 d post-CPB.

To further evaluate the diagnostic value of EMP in AKI at 3 d post-CPB, the ROC curve analysis was conducted and the value of AUC was 0.91 (). The best cut-off value of EMP in diagnosing AKI at 3 d post-CPB was 4.56×105 particles/mL with a sensitivity of 0.85 and a specificity of 0.91 ().

Discussion

We previously demonstrated that circulating MPs are increased in heart disease such as VHD (5) and CHD (8). It induces neutrophil chemotaxis in the kidneys (18), which may be the underlying cause of kidney injury associated with cardiac disease. Accumulating evidence proved that MPs and its subtypes including EMP are elevated in AKI and is becoming a novel biomarker of AKI (4,15,22). In this study, we found that the circulating EMP were elevated in the patients undergoing cardiac surgery with CPB, which further increased with the occurrence of AKI at both 12 h and 3 d post-CPB and was a promising diagnostic marker for CPB-AKI.

The KDIGO criteria are conducted to define CPB-AKI mainly depending on the increase of serum creatinine (≥26.5 µmol/L within 48 h or 1.5–1.9 times the baseline levels within 7 days) (20). Among the 85 patients enrolled, the incidence of CPB-AKI was 40% (12 h post-CPB) and 31.76% (3 d post-CPB), which are similar to other studies ranging from 5% to 42% (23). We found that the patient-related factors such as old age, drinking, DM, and the operation-related factors including complex cardiac surgery, CPB time, cross clamp time, and operation time are the potential risk factors for AKI after CPB (24). Others also found that old age and female sex are associated with CPB-AKI (25) and the different results from us maybe due to the small sample size in this study.

Although the serum creatinine is the gold standard to diagnose AKI in clinically, its use is limited due to the physiological processing, drugs, and diabetes which could affect the serum creatinine level (26). Some biomarkers for AKI such as kidney injury molecule-1, IL-18, and neutrophil gelatinase-associated lipocalin are investigated. However, large trials show a weaker diagnostic ability with AUC less than 0.77 (27). Therefore, new potential markers are needed to explore. We previously found that circulating MPs detected by FCM are increased in patients with cardiac disease compared with healthy participants. However, surgery did not alter the levels of MPs at 12 h post-CPB (5,10). Based on the results of MPs detected by FCM, we confirmed that cardiac surgery with CPB does not change the concentration of MPs at 12 h post-CPB including the main subtype PMP. Others’ ex vivo experiments also found the fresh blood challenged with the CPB circuit does not increase plasma levels of PMP (28). During CPB, complement activation and inflammatory response are aggregated and contributes to the release of MPs as predicted. The conflict results may be due to adhesion of PMP to CPB circuit, hemodilution or mechanical cell salvage during CPB (29). Another reason may be the detective limit of FCM, which are unable to measure the MPs (mainly ranging 100–300 nm) less than 300 nm (21,30). The MPs levels including PMP were increased at 3 d post-CPB compared with 12 h and the underlying reason may be the recovery of blood volume and the reduction of PMP clearance from CPB with the treatment in the intensive care unit within 3 d after operation (5,10).

Elevated EMP is becoming a promising biomarker in many cardiorenal disease based on endothelial dysfunction such as CHD (8), VHD (7), atherosclerosis (9), DM (31), acute coronary syndrome (31), kidney failure (32), and end-stage renal disease (33). Previously, we demonstrated that EMP isolated from endothelial cells aggravates endothelial dysfunction, leading to the anti-angiogenesis in the heart and acute lung injury (6-8,16). In this study, we found that the level of EMP was elevated both at 12 h and 3 d after surgery with CPB. The endothelium activation and injury increased the release of EMP, and endothelial cells mostly located at the surface of blood vessels will not be cleared during CPB. The endothelial injury induced by CPB mediates essential pathophysiological processes including the regulation of blood flow, inflammation, and vascular permeability in the AKI (34). The characteristic of reflecting endothelial injury makes the elevated EMP a specific biomarker of AKI possible.

Further investigation on the relationship between MPs including its subtypes and CPB-AKI found that the patients with AKI occurred at 12 h post-CPB meanwhile 3 d post-CPB shows high levels of PMP and EMP, not MMP. The next univariate logistic analysis found that both EMP and PMP are the risk factors for AKI at 12 h and 3 d post-CPB. However, the multivariate logistic analysis revealed that EMP was a more powerful diagnostic biomarker for CPB-AKI. The association between MPs and CPB-AKI are rarely reported. To the best of our knowledge, it is the first time to perform an accurate count of EMP measured by FCM to diagnose AKI after cardiac surgery. We found that the best cut-off valves of EMP are 3.99×105 and 4.56×105 particles/mL,with the specificity of 0.88 and 0.91 for 12 h and 3 d after surgery, respectively. Some studies found the EMP are elevated in the cardiorenal dysfunction-related diseases. In the patients with acute myocardial infraction, EMP are further increased with renal dysfunction (35). EMP are also elevated in CKD patients with coronary artery disease compared with non-coronary artery disease (36). These findings may provide a promising biomarker of EMP for the early diagnosis of CPB-AKI. These findings are very important since 12 h post-CPB is an early stage after cardiac surgery, and the doctors may prevent the occurrence of AKI or treat the patients at an earlier stage of renal function failure through detecting this biomarker clinically.

Previously, we found that the circulating MPs, especially EMP, are increased in heart disease (7,8) and the elevated MPs induced the neutrophil accumulated in the kidneys (18). EMP were derived from endothelial cells upon activation and other stimulations. The glomerulus endothelial dysfunction is the initial step of AKI (34). The changes of EMP reflect the conditions of endothelial function and glomerulus endothelial function (8,19,29,37). The elevated EMP after cardiac surgery can induce endothelial dysfunction and aggravate kidney injury. Thus, EMP may be used to early predict AKI occurrence with more specificity and sensitivity after cardiac surgery.

Several limitations exist in this study. First, these results are based on the measurement of EMP performed on FCM; the widely used technique to determine the cellular origin of individual MPs and the concentration of MPs. The electron microscope confirms that the size of MPs ranges from 100 to 1,000 nm (10,21,30). Due to the technical factors of FCM, parts of MPs are missed. To reduce error, we used the Megamix-Plus SSC beads with four diameters (160, 200, 240, and 500 nm) to quantify the MPs. Second, the heart valve replacement surgery is the major surgery in this study, which needs CPB. Only few cases of cardiac surgery were performed without CPB. Therefore, the number of cases was insufficient as a control variable for patients undergoing cardiac surgery without CPB. Future study will use circulating microparticle levels on patients undergoing cardiac surgery without CPB as a control. Other limitations are the small sample size and the heterogeneity with respect to other modality of operation.

Conclusions

In summary, our present study shows the diagnostic value of EMP for AKI occurred after cardiac surgery with CPB. We found that the concentration of EMP detected by FCM could predict CPB-AKI at 12 h and 3 d post-CPB with high specificities.

The article’s supplementary files as