Resuscitative endovascular balloon occlusion of the aorta may increase the bleeding of minor thoracic injury in severe multiple trauma patients: a case report

A 50-year-old Japanese woman fell from a height of approximately 10 m and was brought to our hospital by an ambulance. She had no remarkable medical and family history, was a social drinker, and a non-smoker of tobacco. Her marital status was stable with her husband. Her consciousness levels were 14 points (E4, V4, M6) according to the Glasgow Coma Scale at initial arrival. She complained of respiratory discomfort and low back pain, and was in an unrestful state. Her respiratory rate was 24 breaths/minute, blood oxygen saturation (SpO₂) was 95% under 10 L/minute oxygenation, subcutaneous emphysema was recognized in the left side of her chest, and her breathing sounded weak. A hemoperitoneum was not detected on focused assessment with ultrasonography for trauma. Her pulse rate was 90 beats per minute (bpm) and her blood pressure was 180/120 mmHg as measured by an automatic blood pressure monitor. However, her capillary refill time was 3 seconds and radial artery pulsation was feeble with cold sweat in her extremities; hence, she was recognized as being in a state of shock. Her face swelled, and bleeding from her nasal and the oral cavities continued. A subcutaneous hematoma was found in her lumbar and her left femoral region. Her left elbow was deformed and swollen. Because she was in a restless state, accurate neurological assessments were difficult, but coarse paralysis of limbs was not observed.

First, a 28-French chest drain was inserted in her left thoracic cavity for the diagnosis of tension pneumothorax. Hemothorax was not recognized. A pelvis X-ray showed unstable pelvic fracture and a contrast-enhanced computed tomography (CT) pan scan was additionally performed. In the contrast CT, retroperitoneal bleeding with the extravasation of contrast media was recognized (Fig. 1). Other injuries and laboratory data on initial arrival are shown in Tables 1 and 2 respectively.

The injury severity score in this case was 66 and the probability of survival was calculated as 59.3%. Multiple rib fractures were seen, even though active bleeding from intercostal artery injury and pulmonary laceration were not noted at the initial visit. She had deteriorated during CT (blood pressure 67/42 mmHg, pulse rate 75 bpm), and blood transfusion was therefore started and intubation was performed. Transcatheter arterial embolization was performed for hemorrhagic shock with the pelvic fracture as the main bleeding source. While starting angiography, she did not respond to the massive blood transfusion and her systolic blood pressure was maintained at around 60 mmHg. Therefore, first, a 7-French aortic occlusion catheter (RESCUE BALLOON®, Tokai Medical Products, Aichi, Japan) was inserted from her left femoral artery. Her hemodynamics had improved and her systolic blood pressure was 90 mmHg due to REBOA (30 mL saline inflation into the balloon) at the first lumbar vertebra level. Next, transcatheter arterial embolization was started using the sheath inserted in her right femoral artery. Her bilateral internal iliac arteries were embolized from the origin portion with a gelatin sponge. Furthermore, we embolized the extravasation of contrast media of her middle sacral artery and lumbar artery with n-butyl-2-cyanoacrylate. Her blood pressure had been monitored by continuous invasive arterial measurement from her upper right brachial artery during interventional radiology. The systolic blood pressure under REBOA was intended to be controlled under 100 mmHg (so called permissive hypotension); however, it was actually in the range of 90 to 160 mmHg. After 18 minutes of occlusion, the balloon was deflated and the aortic occlusion catheter was removed. However, she was still hemodynamically unstable. The internal iliac artery angiography was performed again, and the site that had just been embolized, as described above, presented recanalization. Since coagulopathy was recognized as a complication, we interpreted that it would be difficult to restrain the hemorrhage with the gelatin sponge. Therefore, it was re-embolized from the origin portion of her bilateral internal iliac arteries by n-butyl-2-cyanoacrylate (Fig. 2). After embolization, retroperitoneal gauze packing and pelvic external skeletal fixation were additionally performed. Further, a repeat CT pan scan was performed, and a massive hemothorax had appeared in her left pleural cavity, which had not been captured at the first arrival (Fig. 3). As the inserted chest drain at the first arrival was detained between lobes, the amount of bleeding in the pleural cavity had not been reflected precisely. Thoracotomy hemostasis was performed and a hemothorax of approximately 2500 ml was aspirated to search for the source of the bleeding. However, clear active bleeding was not captured and only extremely minute bleeding from the rib fracture site, chest wall, and pulmonary contusion area was detected. Therefore, special hemostasis treatment was not needed. After a massive hematoma was removed, a chest drain was placed again and her chest was closed. Continuous blood transfusion was not needed before or after the operation and her postoperative hemodynamics became stable. The timeline of the initial treatment in our emergency department is shown in Table 3.

The retroperitoneal packing gauze was removed on day 3. On day 24, the chest drain was removed without increasing the hemothorax. Complications accompanied by internal iliac artery occlusion, such as gluteal muscle necrosis, did not appear. She regained the ability to walk, and was moved to another hospital on day 154 for rehabilitation. She was discharged from the rehabilitation hospital 6 months after the injury, and her activities of daily living were almost independent.

This case was in a state of threatened cardiac arrest from hemorrhagic shock due to pelvic fracture, but cardiac arrest was avoided by REBOA. However, bleeding of the thoracic injury proximal to the occlusion increased over time, and additional thoracotomy hemostasis was required. The CT at the initial arrival had presented only mild lung contusion. Arterial hemorrhage that resulted in massive hemothorax in a short time was not shown. It is known that delayed massive hemothorax develops in 7.4% of blunt thoracic injuries with rib fracture, even if it is a minor injury in the initial diagnosis [4]. It usually develops from 18 hours to 6 days (3 days on an average) after injury [5]. Moreover, its cause is bleeding from the intercostal artery in most cases, but diaphragm injuries can also be the cause in some cases. In this study, the findings obtained by CT and open thoracotomy after the removal of the aortic occlusion did not show active bleeding, and intercostal aneurysms and diaphragm injuries were not confirmed. In a head and pelvic injury case treated by REBOA, Uchino et al. reported that a complication developed in the site proximal to the occlusion by REBOA in which minor intracranial bleeding was seen in the head CT at the initial diagnosis, wherein the bleeding suddenly increased and a cerebral hernia developed and the patient died [3]. In this case, REBOA excessively raised the perfusion pressure in the chest area proximal to the occlusion, and coagulopathy occurred, induced by trauma shock and trauma itself. Therefore, we considered that bleeding accompanied by extremely minor thoracic injury had increased unexpectedly.

Several methods to avoid complications while performing insertions have been reported, such as the insertion method with ultrasonography [6, 7], insertion training with a cadaver [8], and the use of smaller introducer sheaths for REBOA [9]. In fact, it has also been reported that a good outcome was obtained by REBOA in the course of prehospital care for severe pelvic fracture as it can be inserted safely without fluoroscopy [10].

However, only a few studies have focused on blood flow evaluation at the site proximal to the occlusion in an aortic occlusion, or discussed perfusion pressure in detail. Therefore, a clear standard does not exist for the best occlusion rate, or for bleeding control at a site distal from the occlusion without increasing bleeding at a site proximal to the occlusion. Furthermore, it has been reported that in an analysis of data from the Japan Trauma Data Bank, the use of REBOA increased mortality compared with that of cases without REBOA [11]. REBOA is a revival treatment required in an emergency situation, where there is no other way but to employ REBOA, and it is used at a stage when, it is assumed, detailed information about the lesion and the severity has not yet been grasped. Furthermore, since the frequency of its use is expected to rise in various situations both in and out of hospital for trauma and endogenous diseases in the future, it is necessary to study the evaluation of blood flow at the site proximal to occlusions, the optimal methods of monitoring such blood flow, and the optimal occlusion rates.

A method for the evaluation of blood flow has not been established for sites proximal to occlusions by REBOA. We report this case to call attention to the use of REBOA which may increase bleeding in sites proximal to occlusions unexpectedly in multiple trauma with head or thoracic injuries, even in the case of minor injuries without active bleeding at the initial diagnosis.