Editorial
October 12, 2023
Contemporary Adjuncts to Hemorrhage Control
Samuel A. Tisherman, Megan L. Brenner
JAMA. Published online October 12, 2023. doi:10.1001/jama.2023.16135
Trauma remains the most common cause of death among people 1 to 45 years of age, with hemorrhage being the most common cause of preventable death. For patients with hemorrhagic shock, expeditious control of bleeding is required to maximize chances for survival. Hemostasis requires that patients have sufficient functioning of the coagulation system, and, in most cases, expeditious interventions (open surgical or endovascular) aimed at controlling the hemorrhage.
This issue of JAMA includes 2 randomized clinical trials1,2 of interventions aimed at achieving rapid hemostasis, both carried out in many of the same institutions. Davenport et al1 studied early, empirical, high-dose cryoprecipitate to correct hypofibrinogenemia that is often found in trauma patients. Jansen et al2 studied the use of resuscitative endovascular balloon occlusion of the aorta (REBOA) to decrease bleeding and improve hemodynamics until procedures to achieve definitive hemostasis were performed.
Both interventions were applied in addition to standard care in a randomized fashion. Presumably patients were not enrolled in both studies. These are the first randomized clinical trials of these interventions. Although neither intervention improved outcomes, the studies have provided important information pertinent to trauma care and lessons for future studies.
Cryoprecipitate Intervention
Trauma patients with severe hemorrhagic shock often require a massive transfusion (>10 units of packed red blood cells [PRBCs] within 24 hours). They frequently develop the coagulopathy of trauma caused by dilution of clotting factors, consumption of factors at the site of bleeding, intravascular coagulation, fibrinolysis, hypothermia, acidosis, and inflammation. Although patients bleed whole blood, modern blood banking involves storage of blood components, PRBCs, fresh frozen plasma (FFP), cryoprecipitate, and platelets. Traditionally, PRBCs were administered first, followed by other products based on laboratory data. More recently, a proactive approach called damage control resuscitation has been recommended,3 in which PRBCs, FFP, and platelets are administered empirically in a 1:1:1 ratio. In addition, use of whole blood has become standard in many trauma centers.4
Fibrinogen, as the precursor of fibrin, is critical for stable clot formation. Studies of damage control resuscitation have not addressed management of hypofibrinogenemia frequently found in trauma patients, though it is associated with increased mortality. Davenport et al1 have tried to address this issue with the CRYOSTAT-2 trial. They hypothesized that early, empirical administration of high-dose cryoprecipitate to provide fibrinogen in addition to a standard major hemorrhage protocol would improve survival in trauma patients with evidence of uncontrolled bleeding who required activation of the hemorrhage protocol and had received at least 1 unit of any blood product. Patients were randomized to receive either standard care with a fixed ratio of PRBCs to FFP or standard care plus 3 pools of cryoprecipitate (approximately 6 g of fibrinogen) as soon as possible. Among 1604 patients enrolled, no differences in mortality or any differences in safety outcomes were observed, including thrombotic events.
There are several potential explanations for these negative findings. First, cryoprecipitate may not add benefit for trauma patients already receiving balanced resuscitation with PRBCs and FFP. Fibrinogen levels were measured in the pilot study (CRYOSTAT), but were not measured in the CRYOSTAT-2 trial1; therefore, it is unclear how many patients experienced hypofibrinogenemia in CRYOSTAT-2 or how the patients responded to cryoprecipitate.
Second, the study may not have enrolled enough patients with uncontrolled bleeding for whom cryoprecipitate would most likely benefit. This is evidenced by the fact that the median systolic blood pressure was greater than 100 mm Hg upon arrival at the emergency department and less than one-quarter of the patients required a massive transfusion. Of the patients in the cryoprecipitate group, 15% did not receive cryoprecipitate because they did not have uncontrolled bleeding, the bleeding had been controlled, or they had died.
Third, there was overlap in the timing of administration of cryoprecipitate between groups, decreasing the separation between the treatment strategies. It is possible that there was a Hawthorne effect on the administration of cryoprecipitate in the standard care (control) group. In addition, the delay in administration of cryoprecipitate because of the need for thawing may decrease its benefit when timing is so critical. Future studies with a fibrinogen concentrate, which could be administered outside the hospital, instead of cryoprecipitate may be better positioned to demonstrate benefit.5
REBOA Intervention
The REBOA intervention involves placement of a balloon-tipped catheter through the common femoral artery into the aorta. The balloon is inflated to occlude the aorta proximal to the presumed site of bleeding in the abdomen, pelvis, or both places. Multiple studies have suggested that REBOA does not worsen outcomes6-8 and, in many institutions and settings, improves outcomes with increased institutional experience.9 When REBOA is used in a setting with standardized training, protocol, expertise, and experience, outcomes are improved compared with patients who do not receive REBOA.8 However, no randomized clinical trials had been published on the use of REBOA.
Jansen et al2 studied the use of REBOA in trauma patients with confirmed or suspected life-threatening torso hemorrhage that may benefit from treatment with REBOA. Patients were randomized to REBOA and standard care vs standard care alone. Ninety patients were enrolled in the trial. Among the 46 patients in the REBOA and standard care group, only 19 patients (41%) had the balloon inserted and inflated.2 The procedure required 30 minutes and was not possible in 8 patients. Of the 46 patients in the REBOA and standard care group, 17 (37%) improved hemodynamically and did not need REBOA and 2 (4%) deteriorated before the catheter was inserted.2Even though most patients (n = 27) in the REBOA group did not receive the intervention, the study was stopped early because of increased mortality, specifically from hemorrhage (the main indication for REBOA) in the REBOA and standard care group.
Based on clinical experience and observations from previous studies, there are several variables and details regarding REBOA that affect mortality and are absent from this study. First, the type and size of the balloon occlusion device was not controlled or described. Studies have demonstrated higher mortality with larger devices.10,11
Second, appropriate training, credentialling, and ongoing experience with any new device are necessary for optimal outcomes. The article lacks detailed descriptions of REBOA training for the clinicians, the specialties of the clinicians, and credentialling for the procedure. It is concerning that the highest enrolling institutions in this study may have only performed 1 REBOA procedure every 2.5 months to 5 months during the trial, and that the other trauma centers performed the procedure even less frequently. The fact that the median time from emergency department arrival to REBOA balloon inflation was 32 minutes is concerning because this is significantly longer than the times reported at more experienced trauma centers.
Third, the duration of aortic occlusion, particularly beyond 90 minutes, increases mortality. Without knowing the size of the devices, training requirements, and specific durations of occlusion, the findings must be interpreted with caution. The conclusions may be appropriate for the specific trauma centers included in this trial, but likely do not reflect the effect of REBOA at health systems with significantly greater institutional and individual clinician experience with massive hemorrhage.
A consensus statement12 from the American College of Surgeons Committee on Trauma, the American College of Emergency Physicians, the National Association of Emergency Medical Services Physicians, and the National Association of Emergency Medical Technicians recommends that REBOA should be used only in settings in which expeditious hemorrhage control is available. Furthermore, complete aortic occlusion at the distal thoracic aorta should be limited to 15 minutes, and more distal occlusion limited to less than 30 to 60 minutes.
Despite the significant limitations mentioned above, the trial by Jansen et al2 is another small step in the ultimate goal of determining which patients and in which settings REBOA may improve outcomes. This trial highlights the need for additional studies of REBOA conducted at experienced trauma centers with well-defined indications, training, and credentialling.
Future Research
The fact that both of these randomized clinical trials1,2 had negative outcomes demonstrates the complexities of studying therapies in the fast-paced, complex environment of trauma management. Several factors should be considered in future studies of these therapies. First, participant selection is critical. Judgment by a clinician that a patient has uncontrolled bleeding is very subjective. Both studies enrolled many patients who actually were not actively bleeding. More detailed selection criteria are needed.
Second, the interventions must be implemented as soon as possible. Delay in cryoprecipitate administration or REBOA placement likely decreased potential benefit.
Third, adequate experience and training are critical when conducting research on procedures so that the study is testing the procedure itself and not the ability to train clinicians on the procedure and their performance.
Conclusions
The authors of both of these randomized clinical trials1,2 should be commended for trying to rigorously test relatively new approaches to the management of blood loss. Both studies add to the growing literature related to hemorrhage control and can help guide future work in these areas. Although both trials found no benefit, it is time to conduct more rigorous studies of cryoprecipitate and REBOA.
