Pre-Hospital Zone 1 Partial REBOA

Background – What do we already know about REBOA?

In 2023, Dr Zaf Qasim reviewed the UK REBOA trial for us on St Emlyn’s (The UK-REBOA trial – Has the balloon popped?). You’d be forgiven for thinking that there was nothing more to be said on the matter, and many would agree with you. I often hear a casual resus room discussion saying “obviously REBOA is dead.” Why would they think otherwise, when there was an RCT which was stopped early due to showing harm? Perhaps a reasonable assumption, but let’s remind ourselves of some of the critical appraisal points:

The patients weren’t sick enough to need the intervention
The time between injury and intervention was too long
The intervention caused delay to definitive haemorrhage control
Too few patients needed a haemorrhage control procedure
It was an MTC only trial, can it be generalised to pre hospital care?

This week, a team from London’s Air Ambulance (LAA) have published their cohort study using zone 1 REBOA in the pre hospital environment, specifically looking at partial occlusion (P-REBOA). To their knowledge (and ours) this is the first studied use of zone 1 P-REBOA in the pre hospital environment. See the full article here.

The Abstract

Importance – Hemorrhage is the most common cause of preventable death after injury. Most deaths occur early, in the prehospital phase of care.

Objective – To establish whether prehospital zone 1 (supraceliac) partial resuscitative endovascular balloon occlusion of the aorta (Z1 P-REBOA) can be achieved in the resuscitation of adult trauma patients at risk of cardiac arrest and death due to exsanguination.

Methods – This was a prospective observational cohort study (Idea, Development, Exploration, Assessment and Long-term follow-up [IDEAL] 2A design) with recruitment from June 2020 to March 2022 and follow-up until discharge from hospital, death, or 90 days evaluating a physician-led and physician-delivered, urban prehospital trauma service in the Greater London area. Trauma patients aged 16 years and older with suspected exsanguinating subdiaphragmatic hemorrhage, recent or imminent hypovolemic traumatic cardiac arrest (TCA) were included. Those with unsurvivable injuries or who were pregnant were excluded. Of 2960 individuals attended by the service during the study period, 16 were included in the study.

Results – Femoral arterial access for Z1 REBOA was attempted in 16 patients (median [range] age, 30 [17-76] years; 14 [81%] male; median [IQR] Injury Severity Score, 50 [39-57]). In
2 patients with successful arterial access, REBOA was not attempted due to improvement in clinical condition. In the other 14 patients (8 [57%] of whom were in traumatic cardiac arrest [TCA]), 11 successfully underwent cannulation and had aortic balloons inflated in Z1. The 3 individuals in whom cannulation was unsuccessful were in TCA (failure rate = 3/14 [21%]). Median (IQR) pre-REBOA SBP in the 11 individuals for whom cannulation was successful (5 [46%] in TCA) was 47 (33-52) mm Hg. Z1 REBOA plus P-REBOA was associated with a significant improvement in BP (median [IQR] SBP at emergency department arrival, 101 [77-107] mm Hg; 0 of 10 patients were in TCA at arrival). The median group-level improvement in SBP from the pre-REBOA value was 52 (95% CI, 42-77) mm Hg (P < .004). P-REBOA was feasible in 8 individuals (8/11 [73%]) and occurred spontaneously in 4 of these. The 1- and 3-hour postinjury mortality rate was 9% (1/11), 24-hour mortality was 27% (3/11), and 30-day mortality was 82% (9/11). Survival to hospital discharge was 18% (2/11). Both survivors underwent early Z1 P-REBOA.

Conclusions – In this study, prehospital Z1 P-REBOA is feasible and may enable early survival, but with a significant incidence of late death.
Lendrum RA, Perkins Z, Marsden M, et al. Prehospital Partial Resuscitative Endovascular Balloon Occlusion of the Aorta for Exsanguinating Subdiaphragmatic Hemorrhage. JAMA Surg. Published online July 10, 2024.

How is this different?

There is much to compare between the methodology and the results of these 2 landmark REBOA publications. It is important to remember, though, that in so doing, we are comparing one arm of an RCT, with a cohort study.

The first thing to consider is that in this paper, all of the patients received zone 1 aortic occlusion, whereas this was only 53% of the 19 patients who received REBOA in the UK-REBOA trial. Unlike zone 3 occlusion (between the renal arteries and the aortic bifurcation), zone 1 occlusion has the potential to decrease thoracic and abdominal as well as pelvic haemorrhage. However, the significance of this difference is as much to do with preventing cardiogenic shock / cardiac arrest than it is with stopping haemorrhage. We already know about the importance of coronary perfusion pressure in maintaining myocardial perfusion and gaining ROSC from cardiac arrest. It has also been shown that Z1 REBOA improves coronary and cerebral perfusion and decreases haemorrhage. Perhaps therefore, it has a role in rapidly preventing irreversible cardiac and cerebral ischaemic injuries, rather than simply stopping bleeding alone.

The second physiological consideration is that of partial occlusion versus complete. Complete supracoeliac (zone 1) aortic occlusion comes at the consequence of substantial visceral ischaemia. P-REBOA is titratable to allow a limited flow of blood past the aortic balloon. Therefore it could provide a balance between protecting the heart/stopping haemorrhage and distal ischaemia, by facilitating distal permissive hypotension which may decrease ischaemic injury and the later consequences of reperfusion. 8 patients (42% of those who received REBOA) in UK-REBOA had P-REBOA. Interestingly it was also 8 in the London paper, which represented 73% of their cohort.

Finally, the inclusion criteria:
UK-REBOA: Trauma patients, aged 16 years or older, presenting to a participating major trauma centre, with confirmed or suspected life-threatening torso haemorrhage, which was deemed to be amendable to adjunctive treatment with REBOA.

Z1 P-REBOA: Trauma patients, aged 16 years or older, attended by LAA, with a clinical diagnosis of exsanguinating sub diaphragmatic haemorrhage, with recent or imminent risk of hypovolaemic cardiac arrest thought to be amenable to treatment with Z1 REBOA.

Before we even look at the results, these criteria show us that we are probably looking at a substantially sicker cohort, substantially earlier in their timeline since injury, which the treating team felt was not going to reach hospital alive.

What kind of study is this?

This study is a prospective cohort study using data from one pre-hospital critical care service (LAA), which provides a 24-hour service to Greater London, covering a population of 9 million people. The study aimed to investigate the feasibility of delivering pre-hospital zone 1 REBOA, the feasibility of titrating this to P-REBOA, and examine the impact of these interventions on clinical outcomes and sequelae. Data was collected between June 2020 and March 2022, during which LAA attended 2960 cases and recruited 16 patients.

Tell me about the patients

Of the 16 patients recruited, there were 5 who did not receive REBOA. 2 improved with conventional resuscitation and REBOA was not attempted, and percutaneous femoral arterial access was unable to be gained in the remaining 3 (all of whom were in hypovolaemic TCA – this is really difficult!). 11 patients therefore received Z1 REBOA. Patient characteristics were:

Median age: 30 years (IQR: 17-76)
81% were male
81% had injuries caused by high-energy blunt trauma
All had exsanguination pathophysiology – the penetrating injuries were groin/thigh/buttock
Median ISS: 50 (IQR: 39-57)
3 were in TCA initially, with 8 patients in TCA (57%) by the time of intervention

All of the patients received pre-hospital blood product transfusion and had continued to deteriorate despite this. At the time of recruitment, the patients had a median systolic BP of 58 (IQR 47 – 82) and diastolic BP of 35 (IQR 27-49), which had deteriorated to median SBP 47 and DBP 28 pre REBOA. In many conventional settings, without invasive monitoring, this pressure would be unlikely to generate palpable pulses or recordable non-invasive blood pressure readings, and many would declare TCA.

At this stage, let’s compare some features of the patients between UK-REBOA and Z1 P-REBOA.

	UK-REBOA	Z1 P-REBOA
Median ISS	41	50
Initial SBP (mmHg)	84	58
Pre hospital time (999 – hospital arrival [minutes])	90	89
Time from injury to REBOA (minutes)	135	57
Proportion of patients which had a haemorrhage control surgical procedure (of those who reached hospital)	30%	100%
Time from injury to haemorrhage control surgical procedure	4 hours	1.5 hours

What this shows us is that this cohort of patients is a very different group to the ones in the RCT. As suspected they were sicker, seen earlier, all required haemorrhage control surgical procedures and many if not all would not have survived to have been eligible for recruitment into the UK-REBOA trial. What we can also see is that whilst the pre-hospital time is the same in both studies, this included REBOA in the P-REBOA trial, and also caused no delay to the time to haemorrhage control procedures.

What were the measured outcomes in this study?

The primary outcome was to determine the feasibility of deploying pre-hospital Z1 REBOA in critically bleeding trauma patients. Secondary outcomes included:

Feasibility of delivering controlled P-REBOA
The impact on pre-hospital and hospital clinical outcomes and safety

What are the main results?

The main findings of the study are as follows:

Z1 REBOA was deployed in 11 patients, and was deemed feasible
Partial REBOA was achieved in 8 of those 11 and was deemed feasible for patients who demonstrated a response to complete REBOA. The 3 patients where it was not achieved were all too unstable to tolerate partial REBOA
Systolic and diastolic blood pressures were seen to improve by 52 and 42 mmHg respectively (group medians). This was a statistically significant increase between pre-REBOA and ED arrival blood pressures. “This cohort had DBP levels below that required to allow effective coronary perfusion (median DBP preintervention, 28 mm Hg) and were therefore at high risk of myocardial ischemia, asystole, and TCA. Z1 REBOA was associated with an immediate increase in DBP to levels adequate to restore coronary perfusion.”
Patients with occlusion times substantially longer than previously estimated ‘hard limits’ (30 minutes in zone 1), were able to survive (189 and 67 minutes respectively)
10/11 patients (all of those who reached hospital) experienced multiple organ dysfunction syndrome. 2 patients had bowel ischaemia requiring resection. 2 patients had distal arterial thrombus requiring thrombectomy. There were no limb amputations due to ischaemia, but there were 2 patients who had bilateral traumatic amputations which were completed surgically.

How about patient survival?

Two things are important here: that survival wasn’t a primary outcome and to interpret the survival figures in the context of a cohort with already extremely poor survival. It’s reasonable to compare these patients to patients in TCA, given that the majority of patients in this cohort experienced hypovolaemic TCA, and the remainder were felt to be in imminent danger of it by the treating team. Mortality figures for ‘TCA’ in its broadest sense vary substantially. Overall 30 day survival from TCA in a TARN analysis in 2017 was 7.5%. More recent work, some of which is pre-publication, suggests a mortality of 90% from witnessed hypovolaemic TCA in the LAA system, and this mortality rapidly increases over the subsequent 10 minutes to almost 100%.

3 hour mortality: 9%
24 hour mortality: 27%
30 day mortality: 82% (survival to hospital discharge 18%)

The 3 and 24 hour mortality rates for these patients are substantially lower than I would expect for a cohort this severely injured. Clearly there is a very substantial proportion of late deaths with reported high mortality, but even this is lower than the previously reported mortality for hypovolaemic TCA. In terms of complications (the 2 bowel ischaemia cases and the 2 arterial thrombus cases), these are not atypical for patients with this severity of haemorrhage and metabolic insult, even without aortic occlusion.

If we go back to the detail of the UK-REBOA trial patients, they also reported 3 and 24 hour mortality both for the intention to treat REBOA arm (of whom only 19 of the 46 patients received REBOA) and for the 19 patients who received REBOA in a PS/CACE supplementary analysis. If we compare those figures with the Z1 P-REBOA patients we see something potentially very interesting:

	UK-REBOA (Supplementary Analysis)	Z1 P-REBOA
3 hour mortality (%)	26	9
24 hour mortality (%)	42	27
Discharge mortality (%)	68	82

We can see that the 3 and 24-hour mortality for the substantially sicker Z1 P-REBOA patients, is substantially lower. Could this indicate to us that the technique can generate early unexpected survivors, and that the next phase of learning needs to be about managing their ischaemia, cardiogenic shock, and ischaemic reperfusion injuries? The authors comment that early VA-ECMO may have led to survival in some cases, and this is something which is being investigated in London currently.

What about the limitations?

This is outstanding work, and really does shift the boundaries on how much we understand about exsanguinating trauma patients and the potential role of aortic occlusion. That said, we must always consider the relevant limitations:

Study Design: As an observational cohort study without a control group, it is difficult to attribute the outcomes solely to the intervention as there might be other confounding factors influencing the results.
Data Limitations: The level of data presented in the paper and supplementary material is incredible, down to individual patient variables. Unfortunately, no data were collected about blood transfusion volume to go along with this.
Technical Failures: Issues with setting up the required transducer sets mean there are some missing data.
Selection Bias: The procedure was only attempted 16 times out of 2960 cases (0.5%). This suggests that there were elements of selection both by the recruiting clinicians, and by the availability of clinicians who were trained and authorised to recruit to the trial.
External Validity: The study was conducted in a specific urban trauma setting with a highly experienced physician-led prehospital trauma service, which might not be reflective of other regions or services.

Should we change practice based on this study?

Whilst the usual caveats of single-centre, small sample, non-randomised trial apply, it is difficult to think of achievable ways this work could be bettered. Thus I think it is worth considering the substantial detail within. The very high 30 day mortality will be a concern for most, but rather than being off-putting, I think it should be a reminder of how sick this cohort is, and that resuscitating them from these extremes of physiology may require extreme measures. The very low 3 and 24 hour mortality are encouraging and I think shows that this intervention does have promise for patients at the extremes of potential survival. Picking up a REBOA catheter and adding it to your bag tomorrow will not work though; it’s vital to remember that this can only be delivered with substantial institutional and individual practice of endovascular access and techniques and massive system-wide learning to support these incredibly sick and complex patients.

Summary

This study analysed the feasibility, safety and impact of pre hospital zone 1 REBOA on exsanguinating peri-arrest trauma patients. My key take aways from it are:

Pre-hospital zone 1 complete and partial REBOA are both feasible with suitably experienced and trained teams.
It causes a significant increase in systolic and diastolic BP, which may prevent ischaemic brain injury and cardiogenic shock.
With partial occlusion, the ischaemic consequences may be no worse than those that the patient would experience as a result of profound hypotension/low flow/TCA (i.e. without the balloon).
Zone 1 REBOA may be able to be performed for substantially longer than 30 minutes by using partial occlusion – this vastly increases the number of settings and systems it may be applicable to.
Z1 P-REBOA appears to prevent early deaths from exsanguinating haemorrhage. Further work is required to develop the follow on care to similarly prevent later deaths.

The bottom line: Has the balloon burst for REBOA? I don’t think so. Please read the full paper and supplementary information to make your own decisions. If nothing else, it is packed with fascinating data and physiology at the extreme limits of human survival from trauma. Enjoy – I certainly did.