There seems to be a lot of really interesting papers on Traumatic Brain Injury (TBI) this year. A welcome relief from COVID 19 perhaps but also a reminder that other pathologies exist and that TBI is a major source of morbidity and mortality for our trauma patients.
CRASH 3 was one of the standout papers of the year. It showed that TXA was likely most effective in the group of patients in whom further bleeding (and thus its prevention) was likely. The principle of managing the secondary brain injury by reducing the amount of bleeding from cerebral contusions and blood collections certainly makes pathophysiological sense, with TXA affecting fibrinolysis, but what of other aspects of the clotting mechanism?
Back in 2018 the PAMPER trial (2) reported on the outcome of a prehospital randomised controlled trial of early plasma use in trauma patients. We reviewed the paper on St Emlyn’s here. In brief this was an RCT of 501 patients in whom a 10% benefit in terms of mortality was found. This is a remarkable effect which certainly raised some skeptical eyebrows, but there are plausible reasons why it might be true.
This month we have a secondary analysis of the PAMPER trial with a particular focus on those patients with traumatic brain injury (1). The abstract is below, but as always we strongly recommend that you read the full paper.
What type of paper is this?
This is a secondary analysis of an RCT. Whilst it’s great that the trial is based on an RCT, we need to be cautious in this being a secondary analysis. It was not the main aim of the study and should therefore be considered as an observational hypothesis generating study.
What did they do?
You can read the full protocol in the original study published in the NEJM (2), but in essence this authors randomised trauma patients in US aeromedical services to either standard care (which was quite variable) or to early use of prehospital plasma. The original trial recruited 501 patients who were pretty sick and showing signs of hypoperfusion. The primary outcome was mortality at 30 days in the original study.
In this secondary analysis the authors have looked to see if there was a significant survival advantage to patients with head injury. This is interesting as in general I think many people think about coagulopathy management when dealing with large volume bleeds. In the brain/cranium bleeds are rarely large volume but even though volumes are small the impact can be devastating. Should we therefore be managing coagulopathy as keenly in head trauma as in large volume bleeds. CRASH-3 would certainly support that notion, but what about plasma?
Anything to note in the methods?
As we said earlier this is a post-hoc analysis and that should make us skeptical. The authors identified traumatic brain injury (TBI) as AIS score for the head greater than 2 in the original PAMPER trial, but in this secondary analysis they changed that to any finding consistent with TBI as defined by a radiologist at initial head CT. This is likely to be a broader definition of what constitutes a TBI.
They also looked at some biomarkers of brain injury, glial fibrillary acidic protein (GFAP) and ubiquitin C-terminal hydrolase (UCH-L1).
What are the main findings?
Of the 501 patients in PAMPER, 166 patients had TBI as defined in this study, with a fair spread between the two treatment arms. 74 TBI patients were allocted to the plasma group, 92 to standard care. In general patients with TBI across the whole PAMPER cohort were sicker and required more interventions.
Like non-TBI patients in the PAMPER cohort, those with TBI in the plasma group had fewer resus requirements, fewer inotropes and fewer subsequent transfusions.
For the main outcome of this study (30 day mortality amongst those with head injury), then there appears to be a difference between those patients with TBI treated with plasma. More interestingly, when non-TBI patients are analysed separately then there is little difference found. This suggests that much of the overall difference found in the PAMPER trials is as a result of benefits seen in TBI patients. This difference can be seen in the Kaplan-Meier curves in Karim Brohi’s tweet below.
It is interesting to note that the impact of plasma appears to be early (as in keeping with CRASH-3), but the exact mechanisms are still unclear as outlined in the accompanying editorial.
So should we give plasma to patients with TBI?
As this is a secondary analysis we cannot be certain on this. Association is not causation and although the effect here could be clinically important there are still unknowns. We still don’t know whether survival translates into positive neurological outcomes for these patients, we don’t know whether timing is important (though that is likely) and we don’t have enough detail here to determine what the mechanism might be. So for now it could not be considered to be routine care, but it is a fascinating finding that deserves further investigation.
- Association of Prehospital Plasma With Survival in Patients With Traumatic Brain Injury https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7563075/
- Sperry JL, Guyette FX, Brown JB, et al. ; PAMPer Study Group . Prehospital plasma during air medical transport in trauma patients at risk for hemorrhagic shock https://pubmed.ncbi.nlm.nih.gov/30044935/
- CRASH3 review on St Emlyn’s https://www.stemlynsblog.org/jc-tranexamic-acid-txa-in-head-injury-the-crash-3-results-st-emlyns/
- CRASH-3. The Lancet. https://www.thelancet.com/. Published October 14, 2019. Accessed October 14, 2019