Prediction of arterial extravasation in pelvic fracture patients with stable hemodynamics using coagulation biomarkers

A double-center, retrospective, observational study was performed. The study protocol was approved by the institutional review board of Gunma University Hospital and Japan Red Cross Maebashi Hospital. The medical records of patients with a pelvic fracture, transferred to the emergency department (ED) of these hospitals between December 2009 and December 2016, were reviewed.

Patients with a pelvic fracture who received prehospital treatment comprising only crystalloids and/or packed red blood cell infusions and who were tested for coagulation biomarkers on ED arrival were included. We defined exclusion criteria according to that of our previous study [10]: (1) abbreviated injury scale (AIS) scores in another region that was higher than the pelvis AIS score, (2) arterial extravasation in regions other than the pelvis, and (3) unknown time of trauma occurrence. In addition, we excluded pelvic fracture patients with unstable hemodynamics (systolic blood pressure below 90 mmHg on ED arrival and/or lactate level above 5.0 mmol/L).

The following parameters were obtained on ED arrival: patient demographics, use of anticoagulation, and/or antiplatelet drugs, trauma mechanism, vital signs at ED arrival (the Glasgow Coma Scale, systolic blood pressure, heart rate, and respiratory rate), results of blood tests (hemoglobin and lactate levels, levels of fibrin degradation products [FDP], D-dimer and fibrinogen, prothrombin time–international normalized ratio [PT–INR]), AIS for each body region, injury severity score (ISS), patterns of pelvic fracture classified by Young and Burgess classification [12], World Society of Emergency Surgery (WSES) classification [13], angiography for the pelvis, treatment for pelvic fracture (TAE, external fixation, preperitoneal packing, and amount of blood transfusion within 24 h (in Japan, 1 U of packed red blood cells is approximately 140 mL), time course (door to CT time and door to angiography time), and mortalities (24 h and 30 days). With regard to analytic variables, the ratio of FDP to fibrinogen (FDP/fibrinogen) was also calculated, as with our previous study [10]. FDP and D-dimer were measured using an immunoturbidimetric method and using Cs-2000i and Cs-5100 systems (Sysmex Corporation, Kobe, Japan). It took about 15 min to gain the results of coagulation biomarkers by the immunoturbidimetric method.

During the study periods, strategies, diagnostic devices, and treatment options as to pelvic fracture were not changed. All patients who were suspected of having a pelvic fracture with stable hemodynamics underwent a contrast-enhanced CT in hospital. Contrast-enhanced CT was performed in arterial and portal venous phases. If arterial extravasation was detected on contrast-enhanced CT, we basically performed angiography. In addition, angiography was performed on patients who progressed to unstable hemodynamics and/or had a progressive retroperitoneal hematoma without obvious arterial extravasation on contrast-enhanced CT. If necessary in the acute phase after TAE, external pelvic fixation (damage control orthopedics) was performed. Basically, pelvic fracture patients were treated by transfusion and resuscitated by the use of resuscitation of endovascular occlusion of the aorta if necessary until TAE is completed.

Arterial extravasation was defined as extravascular high attenuating regions with attenuation similar to or greater than that of the aorta on arterial phase images, and the increase of extravasation during the portal phase compared the arterial phase. Arterial extravasation on contrast-enhanced CT and angiography was analyzed by at least one radiologist.

We divided the studied patients into two groups: The extravasation (+) group was defined as having arterial extravasation on contrast-enhanced CT and/or angiography, and the extravasation (−) group was defined as not having arterial extravasation on contrast-enhanced CT and/or angiography. The emergency physicians and/or the interventional radiologist made the decision to proceed to angioembolization.

Data are expressed as the median (interquartile reference; IQR). Comparisons of each parameter between extravasation (+) and extravasation (−) groups were performed using Mann–Whitney U, and chi-squared tests or Fisher’s exact test. The efficacy of predicting arterial extravasation was evaluated using the area under the receiver-operating characteristic curves (AUROC), with low, medium, and high accuracy defined as < 0.7, ≥ 0.7 to < 0.9, and ≥ 0.9, respectively [14]. The optimal cutoff point was defined by the maximum of the sum of sensitivity and specificity using the Youden index approach. An ordinal 2-by-2 cross-tabulation analysis for diagnostic purposes estimated sensitivity, specificity, positive and negative predictive value, positive and negative likelihood ratio, and diagnostic odds ratio (DOR). Statistical analysis was performed with IBM SPSS Statistics version 25.0 (Armonk, NY, USA). Statistical significance was defined as a two-sided p value < 0.05.

We performed a subgroup analysis. Subgroup analysis included pelvic fracture patients with stable hemodynamics that had AIS scores in another region that was higher than the pelvis AIS score.

This study demonstrated the association of arterial extravasation in pelvic fracture patients with stable hemodynamics and coagulation biomarkers. Coagulation biomarkers could predict of arterial extravasation in pelvic fracture patients with stable hemodynamics.

Since the usefulness of TAE for pelvic fracture was first reported [13‐15], investigators have debated which patients with pelvic fracture should undergo this procedure [15‐18]. Pelvic fracture patients with unstable hemodynamics were thought to require a prompt hemostatic procedure. However, it is clinically important to predict arterial bleeding in pelvic fracture patients with stable hemodynamics and to judge the necessity of hemostatic treatment before the collapse of their hemodynamics. Arterial extravasation is an important indication for needing TAE [19]. The most important finding of this study was that arterial extravasation in pelvic fracture patients with stable hemodynamics could be predicted by coagulation biomarkers. We could predict of arterial extravasation by coagulation biomarkers before CT scanning, and this perception alerted the later collapse of hemodynamics. Actually, the 15 patients of extra (+) became unstable hemodynamics before completion of hemostatic treatment. The diagnostic abilities of coagulation biomarkers were medium accuracy [14] and not so high (Table 2 and Fig. 2); therefore, a combination of contrast-enhanced CT and coagulation biomarkers was thought to be realistic and practical. For instance, we could judge whether the pelvic fracture patient with stable hemodynamics should undergo enhanced CT or not by coagulation biomarker when we doubted that the patient has arterial extravasation from initial physical examination and/or initial pelvis X-ray.

We performed a subgroup analysis to maintain the clinical generality of the study results. For the subgroup analysis, we evaluated whether coagulation biomarkers could be applied to patients with a pelvic fracture who had a more severe injury of another body part. Coagulation biomarkers were obviously affected by the presence of another anatomical injury, except for the pelvis. As a result, the predictive nature of coagulation biomarkers for arterial extravasation in pelvic fracture was decreased; however, this was of a medium level of accuracy [10]. In a clinical situation, patients with severe pelvic fracture often have another more severely injured body region. Subgroup analysis revealed that coagulation biomarkers may be useful in predicting arterial extravasation among multiple trauma pelvic fracture patients.

Whether arterial bleeding in pelvic fracture patients with stable hemodynamics needs TAE is contentious. It was previously reported that only 23% of arterial extravasation cases observed on contrast-enhanced CT needed TAE [20], and we agree that all arterial extravasation cases do not need TAE. The strength of this study was that we could repeatedly predict arterial bleeding in pelvic fracture patients with stable hemodynamics and the necessity of TAE based on coagulopathy. Since the concept of trauma-induced coagulopathy has been advocated, coagulopathy accompanying trauma is associated with mortality [21‐23]. Hemorrhage and coagulopathy affect each other. Therefore, hemorrhage induces more severe coagulopathy and coagulopathy induces more active hemorrhage in the absence of prompt treatment, to the point where a patient’s hemodynamics will eventually collapse. In this study, the extravasation (+) group showed arterial hemorrhage and coagulopathy; both of these are thought to deteriorate together. Of the extravasation (+) group with enhanced CT, 94% of patients (32/34) needed TAE, which indicated clinicians judged the arterial extravasation with enhanced CT could not conservatively be arrested. The high positive rate of arterial extravasation with angiography among arterial extravasation with enhanced CT may be because of coagulopathy [24]. Coagulopathy was more deteriorated in five patients of the extravasation (+) due to anticoagulation and/or antiplatelet drugs (Table 1). Finally, 15 patients of the extravasation (+) group became unstable hemodynamics; however, there was no hemorrhage-related 24-h mortality.

This study has several limitations. First, the values measured for coagulation biomarkers could have been affected by the timing of the test. In this study, we excluded a patient for whom the time of trauma was unknown; however, we did not take into account the time from trauma occurrence to testing. Second, we did not account for the amount of prehospital intravenous fluids given to patients. The administration of fluids can dilute the blood and affect the value of coagulation biomarkers. Third, we had no information about using tranexamic acid in this retrospective study. Tranexamic acid was thought to be affecting the coagulopathic state; therefore, this was an important limitation [25]. Fourth, the number of patients in this study was not large and the validation of study’s result would be necessary.