Invited Commentary
April 26, 2023
Optimizing Hemostatic Resuscitation in Trauma—Will Endotyping Be the Key to the “Proper” Ratio?
Purvi P. Patel, Matthew J. Martin
JAMA Surg. 2023;158(7):736-737. doi:10.1001/jamasurg.2023.0825
“Measure what is measurable, and make measurable what is not so.”1(p301) Galileo Galilei
Trauma endotypes (TEs) characterize patients based on dynamic changes in immune and inflammatory biomarkers after injury.2,3 Most frequently cited endotypes are hyperinflammatory and hypoinflammatory subtypes. Previous research in trauma, sepsis, and acute respiratory distress syndrome (ARDS) has demonstrated that different endotypes have distinct responses to therapies impacting outcomes.2-4 In this interesting analysis, Thau and colleagues5 have applied the concept of TEs to the Pragmatic, Randomized Optimal Platelet and Plasma Ratio (PROPPR) trial study population. The original PROPPR trial evaluated the impact of 2 different transfusion strategies of plasma, platelets, and packed red blood cells (PRBCs) in trauma patients with hemorrhaging. The groups were randomly assigned to the administration of PRBCs in a 1:1:1 vs 1:1:2 ratio. The key finding of PROPPR was no 24-hour or 30-day mortality difference between groups, although the 1:1:1 group had increased hemostasis and fewer early deaths from bleeding.6 The study by Thau et al5 performs a secondary analysis of the PROPPR trial using results from a panel of 36 plasma biomarkers obtained at admission to categorize patients into endotype groups and assess any differential response to the assigned transfusion ratio arm.5
With this study, the authors identified 2 endotypes, TE-1 and TE-2, with a notable variance in expression of proinflammatory markers. By applying endotypes to assess for a differential response to transfusion, the authors identified that the TE-2 group, which had lower concentrations of all biomarkers, had improved survival with the 1:1:1 transfusion ratio vs the 1:1:2 ratio. This is a comparable result to other studies evaluating ARDS that initially demonstrated similar outcomes between groups based on conventional study characteristics but were able to establish differences once assessing outcomes based on endotypes.2,7 These findings prompt us to ask what other features outside of traditional phenotypes should be considered when evaluating groups in large studies. Taking a systems biology approach and considering patients on proteomic and transcriptomic levels may offer better methods to prognosticate outcomes. Applying the concept of endotypes to clinical practice could allow clinicians to better target therapies and avoid unnecessary trial and error wasting limited and costly resources.
Although these results are novel, interesting, and provocative, they must be considered in light of several limitations, both with the original PROPPR trial and this secondary analysis. It is often overlooked that the PROPPR trial was essentially a negative study—showing no mortality benefit at 24 hours or 30 days with a 1:1:1 transfusion ratio. The 1 positive secondary outcome finding of reduced early death from exsanguination was based on a small number of events, did not translate to any measurable outcome benefit, and has been shown to be an extremely fragile finding based on calculation of the study’s fragility index.8 Attempting to then take these baseline data and further subdivide each randomized treatment arm retrospectively into TEs to identify any interactions on mortality introduces significant additional bias, underpowering, and the inability to make any causal inferences. The manipulation of the various cytokines and other biomarkers by log transformation and z scaling enables better data handling but removes the large amount of variability, skew, assay deviation, and outlier effect that are commonly encountered when measuring inflammatory markers after trauma. Finally, there is currently little in the way of plausible underlying mechanistic explanations for why the less injured/inflammatory endotype would exhibit a benefit of a 1:1:1 transfusion ratio whereas the more severely altered endotype exhibits no such benefit. Although the authors speculate that the answer to who benefits from a 1:1:1 transfusion ratio may lie in TE characterization, it is also entirely plausible that there is little clinical difference between the 1:1:1 vs 1:1:2 ratio strategies. Both of these approaches arguably represent a balanced hemostatic resuscitation, particularly in comparison with the prior paradigm of early administration of large volumes of PRBCs and crystalloid as the primary resuscitative agents. We look forward to further study of this interesting concept of TEs and whether it truly can discern clinically relevant cohorts and guide the approach to early resuscitation ratios or other lifesaving therapies.
