Thromboelastometric analysis of the risk factors for return of spontaneous circulation in adult patients with out-of-hospital cardiac arrest

It is well known that coagulopathy is observed in patients with out-of-hospital cardiac arrest (OHCA). Thrombolytic therapy for those patients has been controversial until now. The purpose of this study was to identify a significant predictor for return of spontaneous circulation (ROSC) of OHCA patients in the emergency department (ED) using whole blood viscoelastic testing. Adult non-trauma OHCA patients transported to our hospital that underwent thromboelastometry (ROTEM) during cardiopulmonary resuscitation between January 2013 and December 2015 were enrolled in this study. We divided patients into two groups based on the presence or absence of ROSC, and performed statistical analysis utilizing patient characteristics, prehospital data, laboratory data, and ROTEM data. Seventy-five patients were enrolled. The ROSC group and non-ROSC group included 23 and 52 patients, respectively. The logistic regression analysis, utilizing significant parameters by univariate analysis, demonstrated that lactate level [odds ratio (OR) 0.880, 95% confidence interval (CI) 0.785–0.986, p = 0.028] and A30 of EXTEM test [OR 1.039, 95% CI 1.010–1.070, p = 0.009] were independent risk factors for ROSC. The cut-off values of lactate and A30 in EXTEM were 12.0 mmol/L and A 48.0 mm, respectively. We defined a positive prediction for ROSC if the patient presented lower lactate level (<12.0 mmol/L) and higher A30 of EXTEM (≥48.0 mm) with high specificity (94.7%) and accuracy (75.0%). The present study showed that lactate level and ROTEM parameter of clot firmness were reliable predictors of ROSC in the ED for adult patients with OHCA.

Rotational thromboelastometry (ROTEM 1 ; TEM International, GmbH, Munich, Germany) is a viscoelastic testing device utilizing citrated whole blood samples. There have been more than one thousand articles about this point-of-care test in various fields of medicine [13][14][15][16]. The features of this device can contribute to prompt pathophysiological evaluations of coagulation and fibrinolytic status compared with the conventional plasma based coagulation test [17,18]. However, there is little evidence about coagulopathy of OHCA patients using viscoelastic devices such as ROTEM [19]. Moreover, no evidence was found for the relationship between the presence of coagulopathy (hypercoagulable? or hyperfibrinolytic?) and a possibility of return of spontaneous circulation (ROSC) until now.
The main goal of this study is to perform univariate and multivariate analyses in order to identify significant predictive indicators for ROSC utilizing patient characteristics, variables in prehospital care, standard laboratory tests, and ROTEM findings.

Patients and study setting
This retrospective study has been certified by the institutional review board of Saga University (20140907). The study population included out-of-hospital cardiac arrest (OHCA) patients who were transported to the emergency department (ED) of Saga University Hospital between January 2013 and December 2015. Blood samples were obtained from patients on admission in order to perform a complete blood count, standard coagulation tests, blood gas, and ROTEM analysis during resuscitation. The indication for ROTEM test was determined by every physician in charge. We retrospectively evaluated the demographics, past medical history, data on prehospital care, and the laboratory tests including ROTEM findings. Exclusion groups included trauma, pregnancy, and patients under 18 years old.
Japanese prehospital care in a case of out-of-hospital cardiac arrest In general, prehospital resuscitation based on basic life support (BLS) tends to be performed by emergency medical services (EMS) on OHCA patients. Sometimes EMS activates the Dr. Car system, which is a prehospital patient care system in Japan that utilizes an ambulance in which an emergency physician and nurse, as well as paramedics, are on board. If Dr. Car is activated, medical personnel perform cardiopulmonary resuscitation according to the advanced cardiac life support (ACLS) at the scene. Our hospital launched this system in 2011 and we perform 300 to 350 missions a year since 2011 to the present.
Variables in prehospital care by Japanese emergency medical service Data on prehospital care performed by Japanese EMS were obtained. The following parameters were used: the presence of bystander CPR (expressed as numbers and percentages), the presence of Dr. Car activation (expressed as numbers and percentages), initial rhythm at scene and ED (categorized as ventricular fibrillation (VF)/ ventricular tachycardia (VT), pulseless electric activity (PEA) and asystole; expressed as numbers and percentages) and time from onset to blood sampling after ED admission (categorized as 30 minutes or less, from 31 to 60 minutes, from 61 to 120 minutes, 121 minutes or more; expressed as numbers and percentages). The definition of return of spontaneous circulation (ROSC) includes the presence of pulse of carotid artery and measurable blood pressure by an automated digital manometer at upper limb.

Thromboelastometry (ROTEM) analysis
Thromboelastometric evaluation was focused on the extrinsic coagulation pathway (EXTEM) and function of fibrinogen pathway in the extrinsic pathway (FIBTEM). Hyperfibrinolysis was defined as below 45 mm of amplitude (A) 10 in EXTEM test, over 600 seconds of clotting time (CT) in FIBTEM test, or more than 15% of maximum lysis (ML) within 60 minutes. [20] All tests were initiated less than one hour after admission and ran more than 60 minutes at 37˚C.

Statistical assessment
All continuous variables between both groups are represented as median (interquartile range (IQR)) and categorical variables as number (percentages). The P values were calculated from the Mann-Whitney U test for continuous variables, and Fisher's exact test and Chi-square tests were used for categorical variables. Values of P<0.05 were considered to be significant. The factors with significant differences in the univariate analysis were analyzed in a multivariate analysis. We selected the logistic regression analysis (Forward: LR method) for multivariate analysis. Cut-off values were calculated using the factors with significant differences in the multivariate analysis. IBM SPSS Statistics version 22 (IBM Corp., Armonk, NY, USA) was utilized for statistical analyses.

Results
A total of 384 adult patients with OHCA were transported to the ED during the study observation period (Fig 1). Out of them, 309 cases (80.5%) met the exclusion criteria. Finally, 75 patients with ROTEM performed during the cardio pulmonary resuscitation in the ED were enrolled in this study. Out of all patients, ROSC was observed in 23 patients (30.7%). Next, included patients were divided into two groups based on the accomplishment of ROSC in the ED; ROSC group (n = 23) and non-ROSC group (n = 52). Table 1 shows results of univariate analysis on patient characteristics and prehospital care by EMS between ROSC and non-ROSC groups. ROSC after admission to the ED was not significantly related to age, gender, medication history, or presence of liver cirrhosis. Bystander CPR was performed on more than half of the patients within each group, however there was no statistical significance (ROSC; 59.1% vs. non-ROSC; 51.0%). In addition, there were no significant differences in the percentages of Dr. Car activation in both groups (21.7% vs. 21.2%, respectively). Although there tended to be more patients in non-ROSC group with asystole (84.0%) as an initial monitor rhythm at scene as compared with ROSC group (60.9%), no significant difference was observed. The same tendency was confirmed for initial rhythm in the ED. About 65% of the patients in ROSC group were transported to our ED within 60 minutes compared with non-ROSC group (36.5%), but this was also not significant.

Complete blood counts, standard coagulation tests and lactate level
The results of standard laboratory tests are shown in Table 2. Median platelet count of the ROSC group was significantly higher than that of the non-ROSC group (13.3x10 4 /μL vs. 10.2x10 4 /μL; p = 0.034). Median APTT of ROSC group was significantly shorter than that of the non-ROSC group (59.6 s vs. 75.4 s; p = 0.014). The ROSC group showed higher fibrinogen level (279 mg/dL vs.208 mg/dL; p = 0.004) and lower FDP (52.7 μg/mL vs. 345.8 μg/mL; p = 0.001) and DD (18.42 μg/mL vs. 105.70 μg/mL; p = 0.001) compared with the non-ROSC group. On the other hand, WBC, Hb and PT-INR were not significantly different between the two groups.
The results of ROTEM findings Table 2 also shows the results of the ROTEM test. The ROSC group indicated significantly shorter median CT of EXTEM than the non-ROSC group (75 s vs. 97 s; p = 0.032). The median alpha angle of the ROSC group was significantly higher than that of the non-ROSC group (74v s. 67˚; p = 0.048). The parameters on clot firmness (A10, A20, A30 and MCF) were  (Table 3). Next, we calculated the cut-off values of lactate and CT of EXTEM by the receiver-operating characteristic (ROC) curve analysis of ROSC. The cut-off value of lactate was 12.0 mmol/L, the area under the curve (AUC) was 0.674, sensitivity was 56.5%, and specificity was 80.0% (Fig 2A and 2C). In addition, the cut-off value of A30 of EXTEM was 48.0 mm, the AUC was 0.715, sensitivity was 72.7%, and specificity was 65.0% (Fig 2B and 2C).

The ROTEM-based predictors of ROSC in ED
We evaluated both parameters in all cases to determine which values were reliable predictors for ROSC in OHCA patients (Fig 3). The population with low lactate value (<12 mmol/L) and high A30 (> = 48 mm) had the highest successful rate for ROSC (81.8%). On the other hand, high lactate value (> = 12 mmol/L) and low A30 (<48 mm) had the lowest successful rate for ROSC (13.6%) and the worst clinical outcome. Next, we created ROTEM-based predictors of ROSC in the ED utilizing these cut-off values of significant parameters by multivariate analysis. We defined the positive criteria to be met  for the blood samples of OHCA patients as follows: lower lactate level (<12.0 mmol/L) and higher A30 of EXTEM (!48.0 mm) (Fig 3). These novel criteria for ROSC show low sensitivity (40.9%), high specificity (94.7%) and accuracy (75.0%).

Discussion
We conducted a retrospective single-center observational study measuring patients' characteristics, complete blood counts, and standard coagulation tests, as well as viscoelastic tests that reflect the extrinsic coagulation cascade. These findings demonstrated that the lactate level and clot firmness of 30 minutes after initiation of clot formation by EXTEM test were independent predictor for ROSC in patients with OHCA. The ROTEM-based predictors we created had high specificity and accuracy in the acute phase of cardiopulmonary resuscitation. To the best of our knowledge, no previous reports ever focused on ROSC and coagulopathy using wholeblood viscoelastic testing such as ROTEM. These results will help medical personnel find the requirement for advanced post resuscitative treatments such as therapeutic hypothermia for patients by referring to two simple parameters within approximately 30 minutes after admission to the ED. Out of 23 ROSC patients after ED admission, 12 patients (52.2%) were survived for more than 6 hours after ROSC, 6 patients (26.1%) were for more than 24 hours, and only 1 patient (4.3%) with cerebral performance categories (CPC) 1 was discharged from our hospital (data not shown). In addition, our results may support to the decision-making for termination of futile cardiopulmonary resuscitation, which may be beneficial for the allocation of limited healthcare resources. OHCA has numerous causes, which makes it difficult for emergency physicians and paramedics to understand its pathophysiology. Myocardial infarction and pulmonary embolism are the most common etiologies for adult OHCA patients, which accounted for over 70% [1,6,11]. Previous reports using thrombolytic agents such as tissue plasminogen activator (t-PA) showed thrombolysis therapy during cardiac resuscitation was related to higher ROSC ratio compared with control [8,10,12]. However, the randomized control trial on the effectiveness of t-PA for 233 cardiac arrest patients with pulseless electric activities demonstrated fibrinolysis therapy did not improve the percentage of ROSC or clinical outcomes [5]. A meta-analysis in 2006 concluded that thrombolytic therapy was related to 2.5 times as many ROSC cases compared to control, although there were significantly more cases of severe bleeding in the thrombolytic group [7]. Unfortunately, the TROICA study in 2008 did not show any advantage of thrombolytic agents for better clinical outcome, including ROSC in OHCA cases with possible cardiac origins. On the contrary, it only indicated that this therapy was correlated with a significantly higher rate of intracranial hemorrhage [6]. There have not been any largescale studies conducted after the TROICA study.
The present study indicated that both lactate level and a parameter of ROTEM (A30 in EXTEM test) are independent predictors of ROSC, and their combination is strongly connected with the possibility of ROSC. Both parameters, known as point-of-care testing, are measured in the ED in our hospital. The lactate level can be measured within 5 minutes. It only takes 1 minute for the ROTEM to start measurement of EXTEM, 1 to 2 minutes for CT of EXTEM, and 30 minutes for the A30. Therefore, the result of A30 should be ready in less than 35 minutes, suggesting that all important parameters are accessible in a matter of half an hour during cardiac resuscitation in the ED. These findings enable us to promptly determine the indication of aggressive treatment for post cardiac arrest syndrome (PCAS).
The present study has some drawbacks. Firstly, this is retrospective study of one single center with a relatively small sample size. In the emergency department setting, we simply do not have medical personnel that operate ROTEM. In fact, ROTEM analysis was missing for 76% of the patients because only a few physicians (not all physicians) could handle the ROTEM, and physicians on shift had no time for operating the ROTEM during CPR. In addition, there are some missing data about prehospital information and laboratory testing. Vasopressors including adrenalin and vasopressin affect the constriction of systemic blood vessels. Unfortunately, we couldn't analyze the doses of the vasopressors. Finally, there is no institutional guideline for discontinuation of aggressive cardiopulmonary resuscitation. In our hospital, every physician on call decides on the resuscitative management for the OHCA patient.
Is an anti-thrombolytic therapy the best for the OHCA patients? To date, thrombolytic therapy, including t-PA, has some evidence for the improvement of the clinical outcome for myocardial infarction and pulmonary embolism [7,8,10,12]. However, recent article using ROTEM in 2013 documented that hyperfibrinolysis is observed in many more patients (35.8%) compared with previous reports [19]. Surprisingly, the current study estimated that 89.3% of OHCA patients showed hyperfibrinolysis on admission although no statistical difference was found between each group. ROTEM can detect multiple aspects of thrombogenicity and thrombolysis in real time. The viscoelastic testing using whole blood shows potential for a reliable predictor of coagulopathy in patients with OHCA.

Conclusions
The new diagnostic criteria including lactate level and the ROTEM parameter of clot firmness predicts a possibility of ROSC in the ED for adult patients with OHCA. In the future, thromboelastometric analysis using larger sample populations will be warranted in order to evaluate the relationships between OHCA induced coagulopathy, including hyperfibrinolysis, and short-term clinical outcome.