Non-compressible torso hemorrhage (NCTH) remains a preventable cause of death from trauma.^[1] Currently, early systemic fluid resuscitation or blood transfusion combined with effective hemostatic measures is often recommended to stabilize the physiological state of NCTH patients.^[2] Resuscitative endovascular balloon occlusion of the aorta (REBOA) is a readily attainable and effective early in-hospital and pre-hospital intervention for provisionally controlling hemorrhage from NCTH.^[3,4]

REBOA is inherently associated with the burden of ischemia-reperfusion injury (IRI).^[5] Existing evidence indicates that prolonged complete REBOA (cREBOA) leads to distal organ inflammation and injury.^[6] Partial REBOA (pREBOA) can maintain the normal physiological state, reduce the impact of ischemia in the distal organs of the body, and avoid the occurrence of organ failure and prolongation of hospital stay after prolonged blocking time compared with cREBOA.^[7,8] However, the overall survival of patients who underwent prolonged pREBOA did not significantly improve. Moreover, previous animal studies did not involve real injury models, present clinical scenarios of human trauma and blood loss, or perform long-term observation and evaluation after injury. Therefore, it is unclear whether the data from the macrovascular injury model can be extrapolated to hemorrhages from solid organ injury caused by an experimental sliding-chamber ballistic gun. Moreover, whether pREBOA improves the survival rate of animals with hemorrhage due to solid organ injury remains unknown.

The main purpose of this research was to compare the feasibility and effectiveness of pREBOA and cREBOA in a swine hemorrhagic shock model of liver penetrating injury. We also aimed to determine the effectiveness of pREBOA in a solid organ injury model and confirm the best pREBOA pattern for solid organ injury hemorrhage.

This study used healthy adult, evirate male or nonpregnant female Landrace pigs weighing 55±10 kg. This study complied with the principles of the laboratory animal welfare and the guidelines for the protection and use of experimental animals of the National Research Council. Approval was obtained from the ethics committee of Tianjin Bairong Experimental Animal Breeding Technology Co., Ltd., before the start of the experiment (authorization number 2022001).

Animals were randomized into control (no aortic occlusion) (n=5), intervention with complete REBOA (cREBOA) (n=5), continuous pREBOA (C-pREBOA) (n=5), and sequential pREBOA (S-pREBOA) (n=5) groups. The animals were subjected to identical solid organ injury with uncontrolled hemorrhage and resuscitation.

All animals were fasted for 12 h before the initiation of the experiment with free access to water. The animals were premedicated with 0.02 mg/kg atropine sulfate via intramuscular injection. After 30 min, anesthesia was induced via an intramuscular injection of Zoletil 50 (5 mg/kg; Virbac, France). Following endotracheal intubation, general anesthesia was maintained via intravenous infusion of Zoletil 50 (50 mg/h). The core temperature was maintained throughout the experiment (36.5-37.5°C) with a warming blanket. A gastric tube was inserted prior to the procedure.

The animals were placed in the supine position after the induction of anesthesia. The bilateral neck and inguinal areas and pectoral, liver, and renal regions were cleaned and depilated. Under ultrasound guidance (Philips, CX50, Netherlands), an artery catheter (Arrow, ES-04701, USA) was inserted at the left internal carotid artery for proximal blood pressure monitoring and proximal arterial blood sample collection. A jugular 8.5-F sheath (Edwards, 1300F85, USA) was inserted in the right internal jugular to allow access for Swan-Ganz pulmonary artery catheter (Edwards, 744F75, USA) placement, resuscitation, and proximal venous blood sample collection; the right femoral artery was intubated with a 14-F sheath (Cook, RCF-14.0-38-J, USA) to allow placement of a REBOA catheter (Cook, CODA-2-10.0-35-120-32, USA) and monitor distal blood pressure. After external measurement (from the femoral artery to the subxiphoid), the REBOA catheter was pushed to the level of the proximal descending branch of the thoracic aorta. Then, the monitor (Mindray, BeneVision N17, China) was connected.

The experiment was performed according to the defined procedure (supplementary Figure 1). After preparations were complete, a baseline blood sample was collected after the animals had stabilized for 30 min. The process of injury in animals has been described.^[9] An ultrasound scan was performed to locate the lower edge of the liver at the end of exhalation and the lower border of the right rib cage at the end of inhalation; then, a line was drawn between them. The shooting point lied in this arc. The relative height between the shooting point and the highest point of the body surface on the arc (relative to the ground) was recorded as H (H=6 cm) (supplementary Figure 2). An experimental sliding-chamber ballistic gun was used for injury. The diameter and weight of the fragment simulation projectile were 6 mm and 1.1 g, respectively, and the initial velocity of the fragment was 785 m/s. The distance between the muzzle and the shooting point was 1.2 m. After shooting, the wound surface was covered with sterile dressing without pressure dressing. If the mean arterial pressure (MAP) decreased by 30% or more than the baseline value, the swine model of hemorrhagic shock caused by penetrating liver injury was successfully established.

The animals were subsequently randomized into a control group and three REBOA groups. The control group did not inflate the balloon. The MAP difference between proximal and distal was calculated in real time in the REBOA groups, which can judge the degree of aortic occlusion. cREBOA was attained using an aortic occlusion catheter inflated until the right femoral artery waveform disappeared. C-pREBOA was achieved using a standard aortic occlusion catheter inflated until the distal blood pressure gradually decreased by 60%-70% of the pressure gradient (water injection speed, 0.5 mL/min). In the cREBOA and C-pREBOA groups, the balloon was inflated for 60 min. In the S-pREBOA group, the balloon was then maximally inflated for 15 min to achieve full aortic occlusion; afterward, the balloon was slowly deflated until the distal blood pressure gradually decreased by 60%-70% of the pressure gradient and continued to decrease for 45 min. Then, the balloons were slowly deflated until distal arterial waveforms appeared. When the balloon was inflated, 1 mL of normal saline was pushed into the balloon every minute, and when the balloon was deflated, 0.5 mL of normal saline was discharged every minute.

Resuscitation was performed only after the completion of liver injury and shock. When the MAP decreased by 40% or more than the baseline value, fluid replacement therapy of 20 mL/kg was carried out. This scheme involved 250 mL colloids of hydroxyethyl starch 130/0.4 sodium chloride that were initially infused, and the rest was filled with 0.9% sodium chloride injection. The experiment was concluded when MAP < 30 mmHg (1 mmHg=0.133 kPa) or 180-min survival was achieved.

Physiological indices, including heart rate, proximal and distal aortic pressure, and core temperature, and electrocardiography (ECG) were continuously monitored and recorded. Laboratory assessment was performed at baseline, shock after injury, and every 60 min after shock, including measurements of arterial blood gas, hematocrit (Hct), biochemical indicators, and serum cytokines. Immediately after animal euthanasia, tissue samples from the myocardium, liver, kidneys, and duodenum were collected for histological assessment using hematoxylin and eosin staining.

All data were statistically analyzed using SPSS version 22 (IBM Corp., USA). The distribution of continuous data was assessed using the Shapiro-Wilk test. Continuous variables with normal distribution were analyzed using analysis of variance or Student’s t-test. Variables with a nonnormal distribution were analyzed using the Mann-Whitney U-test or Kruskal-Wallis test as appropriate. The log-rank test was used for survival analysis. A P-value of less than 0.05 was considered significant.

The baseline characteristics of the animals were not significantly different (Table 1), similar to the baseline hemodynamic and metabolic indices (all P>0.05) (supplementary Table 1). All animals had penetrating liver injury and a similar degree of shock (P=0.77) (Table 1), and no significant differences were observed in the hemodynamic and laboratory values at various shock time points (all P >0.05) (supplementary Table 1). Without intervention, the condition of the animals deteriorated rapidly after liver injury, and the mean survival time was 57.60 min.

Table 1 also shows the features of the animals and responses to damage and intervention in terms of total volume of blood loss and time to death. The total blood loss volume was statistically lower in the cREBOA group compared with the other three groups (P<0.01). A marked difference was observed in the time of death among the four treatment groups (P=0.03). The average survival time after shock in the cREBOA group was 161.00 min, which was significantly longer than that in the control group.

The Kaplan-Meier curves indicated overall survival (Figure 1). Compared with the control group, the survival rate of the REBOA groups was prominently improved (all P<0.05). Most of the animals in the control group died within 90 min after liver injury. No significant differences were found in the survival time among the REBOA groups.

Table 2 shows the hemodynamic and laboratory indices at various time points. At 60 min before the balloon was deflated, the REBOA groups had higher proximal blood pressure, lower pH, and higher Lac than the control group (all P<0.05). The Hct count was lower at 60 min in the control group and S-pREBOA group compared to that in the cREBOA group (all P<0.05). The alanine aminotransferase (ALT) at 60 min was significantly lower in the cREBOA group than in the control group (P<0.05). The creatinine concentration was lower at 60 min in the control group compared with the cREBOA or S-pREBOA group (all P<0.05). Compared with the C-pREBOA group, the final pH was significantly lower after 180 min in the cREBOA and S-pREBOA groups (P<0.05). The Hct count was higher in the cREBOA group than in the C-pREBOA and S-pREBOA groups at 120 min (all P<0.05). Compared to the cREBOA group, the ALT was higher at 120 min in the S-pREBOA group (P=0.039). At 120 min, the blood urea nitrogen was lower in the C-pREBOA group than in the cREBOA group (P<0.05).

Under the light microscope, the kidney and duodenum showed ischemic injury with tissue loss in all REBOA groups. Representative kidney and duodenum samples harvested from all four groups are displayed in Figure 2. The duodenum tissue of the cREBOA group exhibited significant injury, as evidenced by villus hemorrhage and denudation and lamina propria inflammatory cell infiltration, compared with that in the pREBOA groups. The control group only showed patch disruption of the epithelial cells in the villi. The kidney tissue in the cREBOA group showed more necrotic areas with increased interstitial inflammatory cell infiltration compared to the specimen obtained from the pREBOA groups.

With the advancement of REBOA techniques for NCTH treatment, the complications of aortic occlusion are increasingly well known.^[10] To overcome these complications, novel strategies are needed to maintain MAP while minimizing distal ischemia. However, the anatomical categories of NCTH are diverse;^[2] hence, the specific causes of NCTH should be considered when studying the dispositions and adjunctive treatments, such as REBOA. Most animal studies related to REBOA have used vascular injury models, which may manifest differently when compared with models using other familiar causes of NCTH, such as solid organ injury or serious pelvic fractures. Therefore, this study used a swine model of liver injury and hemorrhage shock caused by an experimental sliding-chamber ballistic gun to analyze the effect of pREBOA and evaluate the situation 3 h after injury. The results suggested that pREBOA was able to improve the survival of subjects with solid organ injury while minimizing IRI after undergoing cREBOA. The C-pREBOA group showed a reduction in the severity of acidosis and end-organ injury.

NCTH is defined by high-grade injury present in one or more of the following anatomic domains: pulmonary, solid abdominal organ, major vascular or pelvic trauma; plus hemodynamic instability or the need for immediate hemorrhage control.^[2] A limited number of hemostatic materials or facilities have been used in the field and hospital settings for the provisional therapy of NCTH, such as multifarious foaming agents and equipment for blocking the vessel^[11] or fibrin sealants for diverse procedures.^[12] Among them, REBOA is a temporary aortic hemostatic method for traumatic hemorrhage;^[13] patients who underwent REBOA showed improved overall survival and a lower risk of early death compared with patients who underwent resuscitation.^[14,15]

However, sustained use of REBOA is associated with IRI; in particular, prolonged zone I inflation to 45 or 60 min was fatal.^[16,17] Thus, reducing the complications of zone I REBOA has become a hot topic around REBOA. An animal study found that pREBOA could better maintain the normal physiological status compared with cREBOA, reduce the ischemic load in those with serious hemorrhagic shock, and lessen the risks of metabolic and inflammatory impacts or organ dysfunction after resuscitation, allowing a longer intervention time.^[7,18] A clinical study in Japan showed that undelayed deployment of REBOA without presenting impending cardiac arrest with shorter balloon occlusion (<30 min at zone I with partial occlusion) might be related to successful hemodynamic stabilization and improved survival.^[19] Clinically, pREBOA might mitigate ischemic-organ injury but is not currently associated with improved survival in prolonged occlusion.^[8] Furthermore, it is difficult to determine the specific advantages of pREBOA that can be used to develop more standardized methods. In addition, the ability to achieve renewable and reliable partial blocking using currently available devices is still unknown, and existing devices cannot strictly control flow. In order to accurately achieve pREBOA, we must optimize the strategy to minimize IRI and extend the tolerance time for aortic occlusion in Zone I.

In our study, we explored the practicality of utilizing simple balloon volume adjustment methods to effectively control NCTH. A study found that monitoring distal pressure and titrating percent balloon volume is feasible to manage pREBOA and that 60% balloon volume was enough inflation to elevate central pressure allowing distal perfusion.^[20] Our study adopted two different partial methods to control NCTH and utilized proximal and distal MAPs for intervention, monitoring, and guidance. Complete and partial obstructed schedules were obviously beneficial in improving the survival of animals with liver injury at 180 min and the proximal blood pressure compared with the control group. However, no prominent difference was observed in the survival time of the partial groups. Moreover, among the above-mentioned hemostatic strategies, complete strategies can better control bleeding. Although early complete blocking for 15 min in the S-pREBOA group may be beneficial for the formation of clots, the total volume of blood loss did not significantly change compared with the control group. This result may be attributed to the fact that venous and portal bleeding caused by liver injury persisted even after REBOA inflation during the experiment. Alternatively, the sudden changes in blood pressure may have caused the blood clots to dislodge. Finally, marked differences were observed in the severity of organ injury and ischemic changes, with less injury observed in the C-pREBOA group than in the other groups. The severity of acidosis in the C-pREBOA group also showed significant improvement over time. This phenomenon may be due to the persistence of distal blood flow; moreover, the influence of proximal organ ischemia eventually reduced over time.

The results demonstrated here are restricted to an animal model and may not be directly suitable for human patients. The study animals were frequently under anesthesia; hence, we were unable to evaluate their long-term outcomes or survival. The currently used devices are insufficient to achieve precise control of aortic occlusion. Finally, considering that long-term REBOA probably requires military-related scenarios, the resuscitation and schedule used were designed to mimic these conditions and are not suitable for dense civilian trauma centers.

In this trauma model with liver injury, a 60-minute pREBOA resulted in improved survival rate and was effective in maintaining reliable aortic pressure, despite persistent hemorrhage. Extended tolerance time for aortic occlusion in Zone I for NCTH is feasible with both continuous partial and sequential partial measures, and the significant improvement in the severity of acidosis and distal organ injury was observed in the sequential pREBOA. Although this study has limitations, it provides data that can serve as a reference for more comprehensive REBOA research.

Funding: The study was supported by military logistics scientific research project (AHJ16J004).

Ethical approval: This research was approved by ethics committee of Tianjin Bairong Experimental Animal Breeding Technology Co., Ltd., before the start of the experiment.

Conflicts of interest: The authors declare no conflicts of interest.

Contributors: TSL: study design, critical revision; YS: literature search, study design, data collection, data analysis, data interpretation, writing; YZ: literature search, study design, data collection, data analysis, data interpretation, writing; CCL: literature search, data collection, study conduct; JXG: literature search, data collection, study conduct; GGS: literature search, data collection.

All the supplementary files in this paper are available at http://wjem.com.cn.