REBEL Cast Ep109: The RePHILL Trial

Background: Currently, trauma resuscitation focuses on immediate hemorrhage control, resuscitation with blood products (i.e. PRBCs, plasma) and early administration of tranexamic acid.  The evidence for the use of blood-based resuscitation in the pre-hospital system has been mixed (PAMPer and COMBAT trials). While early administration of blood products makes physiologic sense, proof of benefit is important due to financial and resource waste burdens.

REBEL Cast Ep109: The RePHILL Trial

Paper: Crombie N et al. Resuscitation with Blood Products in Patients with Trauma-Related Haemorrhagic Shock Receiving Prehospital Care (RePHILL): A Multicentre, Open-Label, Randomised, Controlled, Phase 3 Trial. Lancet Haematol 2022. PMID: 35271808  [Access on Read by QxMD]

Clinical Question: Is prehospital packed red blood cells (PRBCs) and lyophilized plasma (LyoPlas) resuscitation superior to 0.9% saline solution for trauma related hemorrhagic shock?

What They Did:

• Resuscitation with Pre-Hospital Blood Products (RePHILL)
• Multicenter, allocation concealed, open-label, parallel group, randomized, controlled, phase 3 trial
• Performed in 4 civilian prehospital critical care services in the UK
• Randomized in 1:1 fashion to:
  • PRBC-LyoPlas: Up to 2 units of PRBCs + 2 units of LyoPlas
    • Mean volume of 1 unit of PRBCs = 282mL (Range 220 – 340cc)
    • Mean volume of 1 unit of LyoPlas = 213mL (Reconstituted in 200mL of water)
    • Protocol recommended alternating 1U PRBCs with 1U LyoPlas
  • 9% NaCl: Up to 1L of 0.9% sodium chloride
    • Given in 250mL bags
  • Over 4 boluses, similar volumes of fluid should have been administered in each group
  • Both intervention and control fluids were delivered through a fluid warmer
  • For both groups, interventions were administered until either hospital arrival, or return of a SBP ≥90mmHg
  • If all 4 units of intervention were given and further fluids were deemed necessary, non-trial sodium chloride was given (standard UK prehospital practice)
• Trial stopped early before achieving intended sample size of 490 participants due to COVID-19 pandemic

Outcomes:

• Primary: Composite of mortality (Death at any time between injury and discharge) or impaired lactate clearance (<20% per hour in the first 2hrs after randomization) or both
• Key Secondary:
  • Individual components of the primary outcome
  • All-cause mortality within 3 hours and 30 days of randomization
  • Prehospital timing, type, and volume of fluid administered before and after the intervention
  • Serious Adverse Events (Organ failure, multiorgan failure, ARDS, infection, VTE, and transfusion reactions)

Inclusion:

• Adults (Age ≥16 years)
• Trauma-related hemorrhagic shock and hypotension (Defined as SBP <90mmHg or absence of palpable radial pulse)
• On Jan 25^th, 2017, protocol was amended to allow treatment to be administered through IO route

Exclusion:

• Transfusion of prehospital blood products before assessment for eligibility
• Known refusal of prehospital blood products
• Pregnancy (known or apparent)
• Isolated head injury without evidence of major hemorrhage
• Prisoners (For regulatory reasons)
• On Jan 25^th, 2017, protocol was amended to exclude participants with traumatic cardiac arrest where the cardiac arrest occurred before arrival of the prehospital emergency team or the primary cause was not hypovolemia

Results:

• 432 participants randomized
  • Male = 82%
  • Median age = 38 years
  • Cause of trauma mostly road traffic collisions = 62%
  • Median injury severity score = 36 (Range 25 to 50)
  • Non-compressible hemorrhage = 82%
  • Blunt trauma = 79%
  • Concomitant head injury = 48%
  • Average time from call to arrival on scene = 30min
  • Average time from arrival on-scene to administration of 1^st intervention = 25min
  • Average time from call to arrival at hospital = 83min (Range 65 to 101min)
  • Mode of transportation via ground = 62%
  • Average blood pressure at arrival = 73/46mmHg
  • Before randomization most patients had received on average:
    • 430mL of crystalloid fluids
    • 90% received tranexamic acid (TXA)
    • Participants in the PRBC-LyoPlas grorup received on average 1.57U (443mL of PRBCs and 1.25U (266mL of LyoPlas)
    • Participants in the 0.9% saline group received on average 2.55U (638mL)
  • Mortality and/or Lactate Clearrance (Composite Primary Outcome):
    • PRBC-LyoPlas: 64%
    • 9% NaCl: 65%
    • Risk Ratio: 1.01; 95% CI 0.88 to 1.17
    • Adjusted Risk Difference: -0.025%; 95% CI -9.0 to 9.0; p = 0.996
    • Treatment effect was consistent across all predefined subgroups (including injury severity score and scent to hospital transport time)
    • Event rates for the individual components of the primary outcome were also not statistically different between groups
    • Although the point estimate suggested benefit, the confidence intervals were wide and crossed zero making it difficult to determine potential benefit vs harm
  • Rates of Transfusion-Related Complications in First 24hrs After ED Arrival:
    • PRBC-LyoPlas: 7%
    • 9% NaCl: 7%
    • Adjusted RR 1.05; 95% CI 0.46 to 2.42
  • Serious Adverse Events:
    • PRBC-LyoPlas: 6%
    • 9%NaCl: 2%
    • No treatment related deaths

Strengths:

• Treatment packs were identical in external appearance
• Used an intention-to-treat analysis (analysis according to randomization irrespective of treatment received) which mimics real world practice
• Funder of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report
• Groups had similar vital signs and lactate concentrations on arrival to hospital through 24 hours
• Near complete follow up of both groups

Limitations:

• Composite outcome with unequal outcomes…lactate clearance is not equivalent to mortality. One is a patient-oriented outcome, the other is a lab-oriented outcome
• Large amount of 0.9% saline received prior to interventions could potentially cloud the picture
• Prehospital critical care teams in the UK, typically compromise a physician and critical care paramedic which is different than US prehospital systems which may affect generalizability of the results
• Prehospital teams were unaware of the treatment allocation before enrolling participants (Once randomly assigned, health-care professionals administering the trial intervention were aware of the group assignment). This could bias the results, although unclear in which direction
• Wide confidence intervals make the possibility of both benefits and harms
• Recruited only 93% of planned sample size due to COVID-19, however due to the similarity of the primary and secondary outcomes between groups, iit is unlikely that completing recruitment would have changed the outcome
• Open-label study which could have led to performance bias or detection bias, especially for more subjective outcomes such as adverse events

Discussion:

• This trial did not demonstrate that prehospital PRBC-LyoPlas resuscitation was superior to 0.9% saline from trauma related hemorrhagic shock in a civilian population. But the point estimates for the individual components of the primary outcome and some other secondary outcomes (i.e. survival within 3hrs) showed a benefit from allocation to PRBC-LyoPlas.
• Prior to the study authors did a systematic review of studies which compared prehospital blood components (PRBCs, plasma, or whole blood) in both civilian and military settings, with other resuscitative fluids in patients. They found 27 observational studies (16 case series and 11 comparative studies), with no definitive evidence of benefit being found for prehospital blood components and survival (OR for mortality 1.29; 95% CI 0.84 to 1.96), physiology, or in-hospital transfusion requirements
• The investigators needed 438 participants (219 per group) to achieve an 80% power to detect a 10% difference between groups in the primary outcome which they fell just short of (i.e. the study was not powered correctly)
• The total number of blood products up to 24hrs after ED arrival was 6.34U and for plasma was 5.04U in the PRBC-LyoPlas group compared to 4.42U of PRBCs and 3.37U of plasma in the 0l9% saline group (Adjusted Risk Ratio 0.91; 95% CI 0.44 to 1.90; p = 0.80). The logistical implication here is the financial costs and waste of blood products (i.e. more blood products without an apparent improvement in physiological outcomes or survival)
• Why did the RePHILL Study not show benefit of blood products over 0.9% saline?
  • Study took place in a civilian setting within an established major trauma network (i.e. prehospital critical care is provided at the scene of the incident by critical care physicians)
  • Proportion of patients had transport times of <20 minutes (i.e. might not benefit from prehospital transfusion)
  • Use of lyophilized plasma, due to logistical challenges created by a short post-thaw shelf-life of fresh frozen plasma. Lypophilized plasma (LyoPlas) is freeze dried plasma from a single donor that can be stored for years. Requires water to be added to reconstitute the plasma
  • Used plasma to PRBC ratio of 1:1, however unclear if different ratios, the addition of coagulation factors, platelets or use of whole blood would make a bigger difference
  • Majority of patients had blunt traumatic injuries and not penetrating injuries which may not benefit as much from prehospital blood products (i.e. only 9 patients (6%) in the 0.9% saline group had a hemoglobin concentration of <8.0mg/L)

Author Conclusion: “The trial did not show that prehospital PRBC-LyoPlas resuscitation was superior to 0.9% sodium chloride for adult patients with trauma related haemorrhagic shock.  Further research is required to identify the characteristics of patients who might benefit from prehospital transfusion to identify the optimal outcomes for transfusion trials in major trauma.  The decision to commit to routine prehospital transfusion will require careful consideration by all stakeholders.”

Clinical Take Home Point: The RePHILL randomized clinical trial did not show improved mortality or lactate clearance when using PRBC-LyoPlas compared to 0.9% saline in the prehospital setting of trauma-related hemorrhagic shock. However, there may be more specific remote settings, (longer transport times) penetrating trauma patients, etc that may still benefit from prehospital blood products.

References:

1. Crombie N et al. Resuscitation with Blood Products in Patients with Trauma-Related Haemorrhagic Shock Receiving Prehospital Care (RePHILL): A Multicentre, Open-Label, Randomised, Controlled, Phase 3 Trial. Lancet Haematol 2022. PMID: 35271808 [Access on Read by QxMD]

For More Thoughts on This Topic Checkout:

• Critical Care Reviews: RePHILL Trial Results Livestream
• St. Emlyn’s Blog: The RePHILL Trial
• REBEL Cast: Episode 60 – COMBAT and PAMPer – Prehospital Plasma in Trauma
• The Bottom Line: RePHILL
• REBEL EM: The PROPPR Randomized Clinical Trial
• First10EM: The RePHILL Study – Is Saline the Fluid of Choice in Trauma?
• The SGEM: SGEM #369 – Romeo is Bleeding – Does He Need a RePHILL?

Post Peer Reviewed By: Anand Swaminathan, MD (Twitter: @EMSwami)