Endothelial Dysfunction is Associated with Increased Incidence, Worsened Severity, and Prolonged Duration of Acute Kidney Injury after Severe Trauma

Abstract

Introduction:

Nearly half of severely injured patients suffer acute kidney injury (AKI), but little is known about its pathogenesis or optimal management. We hypothesized that endothelial dysfunction, evidenced by elevated systemic soluble thrombomodulin (sTM) and syndecan-1, would be associated with higher incidence, worsened severity, and prolonged duration of AKI after severe trauma.

Methods:

A single-center cohort study of severely injured patients surviving ≥24 hours from 2012-2016 was performed. Arrival plasma sTM and syndecan-1 were measured by ELISA. Outcomes included seven-day AKI incidence, stage, and prolonged AKI ≥2 days. The KDIGO guidelines were used for AKI diagnosis and staging. Univariate and multivariable analyses were performed.

Results:

Of 477 patients, 78% were male. Patients had a median age of 38 (IQR 27-54) and injury severity score of 17 (IQR 10-26). AKI developed in 51% of patients. Those with AKI were older and displayed worse arrival physiology. Patients with AKI had higher plasma levels of syndecan-1 (median 34.9ng/mL vs 20.1) and sTM (6.5ng/mL vs 4.8). After adjustment, sTM and syndecan-1 were both associated with higher AKI incidence, worse AKI severity, and prolonged AKI duration. The strength and precision of the association of sTM and these outcomes were greater than those for syndecan-1. A sensitivity analysis excluding patients with AKI on arrival demonstrated the same relationship.

Conclusions:

Elevated sTM and syndecan-1, indicating endothelial dysfunction, were associated with higher incidence, worsened severity, and prolonged duration of AKI after severe trauma. Treatments that stabilize the endothelium hold promise for AKI treatment in severely injured patients.

Introduction

Acute kidney injury (AKI) occurs in up to half of critically injured patients and is one of the most common manifestations of organ failure after severe trauma. [1, 2] Despite the frequency of AKI development in this population, its clinical pathogenesis is poorly understood. [3] There is a growing understanding that AKI is a heterogeneous clinical syndrome with multiple etiologies which results in a sudden decrease in kidney function. [4]. Inadequate differentiation of the etiologies of AKI may be partially to blame for stymied therapeutic advances. Standard treatment of AKI remains centered on removing the offending agent and minimizing other toxic substances to allow uninhibited recovery. [3]

Hemodynamic compromise has been cited as the most likely global etiology of AKI after trauma. This is supported by studies showing that the number of blood transfusions is associated with incidence and severity of AKI. [3, 4] Recent studies have demonstrated a relationship between dysfunction of the vascular endothelium and outcomes after severe trauma, regardless of hemorrhage. [5–7] This relationship may be particularly important in cases of post-traumatic AKI, because the endothelium plays a pivotal role in kidney function. Experimental models have shown that damage to the kidney’s endothelium results in renal vascular rarefaction, organ fibrosis, and subsequent dysfunction. [8] Clinical studies of patients with chronic diseases such as diabetes or obesity and of patients with sepsis or endotoxemia reveal a relationship between circulating markers of endothelial activation or glycocalyx shedding and poor renal outcomes. [9–13] However, the clinical relationship between endothelial damage and dysfunction secondary to traumatic injury and AKI has yet to be elucidated. We hypothesized that endothelial dysfunction, evidenced by elevated systemic soluble thrombomodulin and syndecan-1 would be associated with higher incidence, worsened severity, and prolonged duration of AKI after severe trauma.

Methods

The McGovern Medical School at UTHealth institutional review board (IRB) approved this study (HSC-GEN-12-0059). This was a post hoc analysis of a prospective, observational study which randomly selected plasma samples to measure markers of endothelial dysfunction from 2012 to 2016. All patients were adult trauma patients (≥16 years) requiring the highest level activation at the Red Duke Trauma Institute. The criteria for highest activation include: Glasgow Coma Score (GCS) ≤10, heart rate >120 beats per minute, systolic blood pressure ≤90 mmHg, respiratory rate <10 or >29 per minute, intubation, penetrating injury to torso, groin, head or neck, amputation proximal to ankle or wrist, paraplegia, quadriplegia, uncontrolled external hemorrhage, fracture to pelvis or two or more long bone fractures, or receiving blood en route. Patients enrolled in alternative studies were not included in this analysis.

Samples were collected when research staff was available to process samples and was not directly involved in the other studies. The IRB approved delayed consent for this study, which was obtained from the patient or their legally authorized representative within 72 hours of admission or as soon as possible. A waiver of consent was obtained for patients who were discharged or died within 24 hours of admission. If consent could not be obtained, the patient was excluded from the study and their blood samples destroyed. Pregnant women and known prisoners were also excluded.

On emergency department (ED) arrival, 20mL of blood was obtained, then transferred into a 3.2% citrate vacutainer tube and inverted to ensure proper anticoagulation. Commercial enzyme-linked immunosorbent assays were performed to quantify levels of syndecan-1 (Diaclone SAS) and soluble thrombomodulin (Nordic Biosite) according to manufacturer’s procedures. Patients were selected for this study based on availability in the database of syndecan-1 and soluble thrombomodulin measurements in arrival plasma. Demographic characteristics, injury details, and outcomes were obtained from the institution’s professionally maintained trauma registry. Laboratory results, need for renal replacement therapy (RRT), and urine output were extracted from the medical record. Patients who died within 24 hours of hospital arrival were excluded to decrease the influence of death as a competing outcome with AKI.

Outcome Measures:

AKI diagnosis was made based on the current guidelines published by the Kidney Disease Improving Global Outcomes (KDIGO) group. [14] The Modified Diet of Renal Diseases (MDRD) uses race, age, and sex to estimate a reference creatinine and has been shown to be superior to alternate reference estimates for AKI diagnosis after traumatic injury. [15–17] The MDRD used in this study was calibrated to an assumed creatinine clearance of 75ml/min. [2, 18] AKI incidence was defined as the proportion of patients who met AKI criteria at any point in the first 7 days to total included patients. This study was limited to the first 7 days to focus on AKI related to initial traumatic injury. Highest AKI stage was defined as the stage met by the highest creatinine during the 7 days following admission, following the KDIGO guidelines. Stage 1 was assigned if there was an increase in serum creatinine by ≥0.3 mg/dl within 48 hours, an increase in serum creatinine to ≥1.5 times reference within 7 days, or a urine output of <0.5 ml/kg/h for 6-12 contiguous hours. Stage 2 was assigned if there is an increase in serum creatinine by ≥2 times reference or if the urine output was <0.5 ml/kg/hr for ≥12 contiguous hours. Finally, stage 3 was assigned if there was an increase in serum creatinine by ≥3 times reference, if the serum creatinine was over 4.0 mg/dl at any time, or if the patient required renal replacement therapy. Stage 3 was also assigned if the urine output was <0.3 ml/kg/h for ≥24 hours or 0 ml/kg/h for ≥12 hours. Short duration AKI was defined as ≤2 days and long duration was defined as >2 days to remain consistent with prior studies. [19]

Statistical Analysis:

Patients with AKI were compared to those without AKI. Demographics and outcomes with continuous variables were presented as medians and interquartile ranges (IQRs). Chi-square and Wilcoxon rank-sum tests were utilized to compare categorical and continuous demographic data and outcomes, respectively. Multivariable log-binomial regression models were utilized to assess the relationship between markers of endothelial dysfunction and outcomes of interest. All associations were reported as risk ratios with 95% confidence intervals. Covariables known or suspected to be confounders between endothelial dysfunction and AKI were chosen a priori: age, arrival systolic blood pressure, and injury severity score (ISS). Full models were also created and included additional variables of arrival Glasgow Coma Scale (GCS) score, mechanism of injury, sex, and 24-hour transfusion requriements. All data analyses were completed using R version 3.53 (R Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria).

Results

There were 520 patients with measured plasma endothelial dysfunction eligible for this study. Forty-three patients were excluded for death earlier than 24 hours after arrival, leaving 477 patients in the final cohort. No patients suffered pre-existing end-stage renal disease. No demographic data were missing. The median age was 38 (IQR 27-54) years. Male sex was predominant, accounting for 78% of the cohort. Participants were severely injured with a median ISS of 17 (IQR 10-26). (Table 1)

Table 1:

Characteristics of Study Participants

	All Patients (n=477)	No AKI (n=235, 49%)	AKI (n=242, 51%)	p
Age, years	38 (27-54)	33 (25-47)	45 (29-60)	<0.001
Male Sex	370 (78%)	178 (74%)	192 (82%)	0.04
Blunt Mechanism of Injury	337 (71%)	141 (60%)	196 (81%)	<0.001
Injury Severity Score	17 (10-26)	14 (5-24)	22 (13-29)	<0.001
Arrival Systolic Blood Pressure, mmHg	123 (103-142)	125 (110-142)	120 (95-141)	0.02
Arrival Pulse, Beats Per Minute	93 (78-110)	92 (79-110)	95 (78-111)	0.31
Arrival Glasgow Coma Scale	12 (3-15)	14 (3-15)	7 (3-15)	0.05
Transfused in First 24h	230 (48%)	72 (31%)	158 (68%)	0.001
Transfusions Required in First 24h, units  Total  Packed Red Blood Cells  Fresh Frozen Plasma  Platelets	0 (0-6) 0 (0-3) 0 (0-3) 0 (0-0)	0 (0-2) 0 (0-2) 0 (0-2) 0 (0-0)	3 (0-12) 2 (0-4) 2 (0-4) 0 (0-6)	<0.001
Arrival Syndecan-1, ng/mL	27.8 (14.7-56.2)	20.1 (12.1-37.0)	34.9 (19.4-77.1)	<0.001
Arrival Soluble Thrombomodulin, ng/mL	5.4 (4.1-7.4)	4.8 (3.6-6.1)	6.5 (4.7-8.5)	<0.001
AKI on Arrival	168 (35%)	0	168 (69%)	N/A
Highest AKI Stage  No AKI  Stage 1  Stage 2  Stage 3	235 (49%) 157 (33%) 65 (14%) 20 (4%)	N/A	0 157 (65%) 65 (27%) 20 (8%)	N/A
Need for Dialysis	7 (2%)	0	7 (3%)	N/A
AKI Duration, Days	1 (0-2)	N/A	1 (1-3)	N/A
Long Duration AKI	85 (18%)	0	85 (35%)	N/A
Hospital Length of Stay, Days	9 (3-19)	4 (1-12)	13 (5-23)	<0.001
ICU Length of Stay, Days	3 (0-8)	1 (0-4)	4 (2-11)	<0.001
In-hospital Mortality	49 (10%)	8 (3%)	41 (17%)	<0.001

Continuous data presented as: median (IQR)

AKI – Acute Kidney Injury; ICU – Intensive Care Unit; N/A – Not Applicable

There were 242 patients (51%) who developed AKI within the first 7 days of hospitalization. Patients who developed AKI were older (median 45 vs 33), more frequently male (82% vs 74%), and suffered blunt mechanism of injury (81% vs 60%). (Table 1) Their arrival systolic blood pressure and GCS were lower and they required more 24-hour transfusions than those who did not develop AKI. Patients with AKI had higher plasma levels of syndecan-1 (median 34.9 vs 20.1 ng/mL) and soluble thrombomodulin (6.5 vs 4.8 ng/mL), suggesting patients who developed AKI suffered more severe endothelial dysfunction on arrival than those who did not develop AKI. (Figure 1) Furthermore, patients with AKI required longer hospitalizations (13 days vs 4 days) and suffered a higher mortality (17% vs 3%) than patients without AKI.

Figure 1:

Markers of Endothelial Dysfunction by AKI Status in Severely Injured Patients

The majority (69%) of the 242 patients who developed AKI met criteria for AKI on hospital arrival. Most of the 242 patients (65%) who suffered AKI met criteria for stage 1 AKI at the highest. Only 7 patients (3% of all patients with AKI) required RRT. The duration of AKI was 2 days or greater in 35% of all patients who developed AKI.

Elevated soluble thrombomodulin on arrival was associated with increased risk of AKI development, prolonged AKI, worse AKI (stage 2 or 3 AKI), and mortality. (Table 2) Similarly, elevated syndecan-1 was also associated with increased risk of AKI development and prolonged AKI, and imprecisely associated with an increased risk of AKI stage 2 or 3. When compared to syndecan-1, soluble thrombomodulin displayed a stronger association and more precise relationship with all outcomes evaluated. After expanding to full models with additional covariables, there were no changes in the associations between both soluble thrombomodulin and worse AKI outcomes, however the association between soluble thrombomodulin and mortality was less precise (Supplemental Table 1). Finally, on sensitivity analysis excluding patients with AKI on arrival, there were again no substantive changes in the findings.

Table 2:

Association between Markers of Endothelial Dysfunction and Outcomes on Multivariable Analyses

	AKI Development	Prolonged AKI	AKI Stage 2 or 3	Mortality
Soluble Thrombomodulin, per ng/ml	1.25 (1.16-1.37), p<0.001	1.20 (1.11-1.30), p<0.001	1.19 (1.10-1.29), p<0.001	1.13 (1.04-1.23), p=0.004
Syndecan-1, per 10 ng/ml	1.06 (1.03-1.10), p<0.001	1.05 (1.02-1.08), p=0.002	1.03 (1.00-1.06), p=0.08	1.02 (0.97-1.05), p=0.41

Findings Reported as Relative Risks (95% Confidence Interval)

AKI – Acute Kidney Injury

Discussion

The present study revealed an association between increased arrival systemic markers of endothelial activation and glycocalyx shedding, soluble thrombomodulin and syndecan-1, with AKI development, AKI severity, and AKI duration. These relationships were incremental, with higher levels of soluble thrombomodulin and syndecan-1 being associated with poorer outcomes. As current therapy for clinical AKI currently is mainly supportive in nature, this finding may have potential implications to guide investigations into optimal treatments. Specifically, stabilization of the endothelium surface and function may benefit patients with AKI after trauma.

In a state of health, the endothelium is the pillar of normal renal function. While the specific phenotype and physiology varies by region of renal microcirculation, it has been postulated that the importance of a well-functioning endothelium may explain why the kidney tends to be a target organ in systemic disease. Severely injured trauma patients suffer a complex, systemic insult that often results in endothelial activation, glycocalyx shedding, and renal dysfunction. [5] As Wohlauer, et al. noted, the kidney is the “canary in the coal mine” for multiple organ failure and death after trauma. [19] Renal endothelial dysfunction may offer insight into other risk factors for AKI after trauma, including increased fluid balance. [20, 21] Increased renal vascular permeability, paired with large-volume resuscitation, may exacerbate organ congestion and tissue edema, and may occasionally result in renal failure due to a compartment syndrome within the renal capsule. [22] While decapsulation has been attempted as a reactive therapy for this problem, understanding and addressing the endothelial dysfunction may be a more prudent and non-invasive approach. [22]

The present study revealed that elevations of soluble thrombomodulin may be a superior indicator of renal dysfunction than syndecan-1. This finding is consistent with a recent study by Inkinen, et al. who observed that soluble thrombomodulin was independently associated with AKI development in patients with sepsis, while syndecan-1 was associated with AKI development on unadjusted analysis only. [12, 23] Importantly, syndecan-1 and soluble thrombomodulin are physiologically distinct, and the conditions under which one may be released into circulation may not apply to the other. Syndecan-1 is a proteoglycan that anchors heparin sulfate to the cell surface, augments mechanotransduction, modulates endothelial interactions with platelets and leukocytes, and regulates the fluid flux. [24, 25] Syndecan-1 is shed in response to inflammatory stimuli, resulting in hemostatic dysregulation, glycocalyx barrier disruption, and increased vascular permeability. In contrast, thrombomodulin is an endothelial bound protein that acts as a receptor for thrombin. [26] Once bound to thrombin, thrombomodulin accelerates protein C activation. [27, 28] It is released into circulation, now denoted a soluble thrombomodulin, after endothelial cellular injury by oxidative stress. [29] The increase in strength and precision of the association between soluble thrombomodulin and AKI may also be due to an intact endothelial anticoagulant system that protects the cells even in the event of glycocalyx shedding. Shedding or cleavage of the cell membrane bound thrombomodulin renders the endothelium unprotected to all injurious stimuli present in the circulating blood and this likely also increases the risk of tight-junction breakage with capillary leakage. [30, 31] Finally, clearance of these soluble markers should be considered. It has previously been suggested that soluble thrombomodulin is cleared by the kidneys, resulting in pronounced plasma elevations in patients with AKI. [32] Further study of the release and clearance of both soluble thrombomodulin and syndecan-1 after severe trauma are needed.

With growing evidence that AKI is closely linked to endothelial dysfunction after ischemia or sepsis, it may be reasonably postulated that stabilization of the endothelium may be beneficial for the treatment of AKI in these settings. Fresh frozen plasma (FFP) is one of the leading clinically available treatments that has been shown to be effective at stabilizing the endothelium after severe trauma and in experimental hemorrhage models. [33, 34] Both fibrinogen and antithrombin III have been recognized for their roles in this observed effect. [35–37] The benefit of these individual factors on AKI specifically has yet to be evaluated. As each potential treatment has its own risk-benefit balance, it may be appropriate to utilize incremental cutoff values to escalate treatment options. Active AKI preventive measures, such as pharmacology modifications and dynamic fluid status assessments may benefit all patients arriving with evidence of endothelial dysfunction after trauma, while antithrombin may benefit only the most severely ill, given the possible detrimental anticoagulant side effects.

Limitations:

Several limitations existed in this study. First, mortality and discharge are competing outcomes with AKI development. Although the majority of AKI related to traumatic injury presents within the first days of hospitalization, it is possible that AKI cases were missed. Second, clinical creatinine data were utilized for AKI diagnoses and therefore the timing of creatinine measurements was irregular. The maximum creatinine for a 24-hour period was utilized to indicate whether a patient met criteria for that clinical day. This may have led to a minor over-estimation of the duration of AKI. Finally, we did not account for pre-existing conditions such as diabetes or chronic kidney disease in this study. While many chronic medical conditions are risk factors for AKI, the reliability of retrospectively collected medical history data in the severely injured is poor and thus these data were not utilized.

Conclusions:

Elevated plasma soluble thrombomodulin and syndecan-1, indicating endothelial damage and dysfunction, were incrementally associated with higher incidence and severity, and prolonged duration of AKI after severe trauma. Treatments aiming to stabilize the endothelium may be promising therapies for AKI treatment in severely injured patients.