Editorial 

October 12, 2023

Precision in Transfusion Medicine

Matthew D. Neal, Beverley J. Hunt

JAMA. Published online October 12, 2023. doi:10.1001/jama.2023.16134

Blood transfusion management has come a long way since the introduction of blood banks in the 1930s, leading to the widespread use of plasma and blood in the Second World War. The gradual introduction of blood components, such as platelets and cryoprecipitate, occurred in the following decades. However, the utility of blood components in various settings was never assessed using modern clinical trial methodology. Some would argue this has led to the overuse of blood components in many clinical settings, based solely on the assumption that using components is better than not using them.

Today, increasing evidence is guiding how to use blood components; however, much of the data, especially involving patients who are massively bleeding, come from blood loss due to trauma. Severely injured trauma patients experience a unique trauma-induced coagulopathy that is acutely managed with either whole blood or a high ratio of plasma and platelets to red blood cells; early use of tranexamic acid; and, if bleeding continues, targeted therapies based on repeated laboratory coagulation testing and/or viscoelastic hemostatic assays.¹ Even for traumatic bleeding, which has been the subject of numerous randomized clinical trials, the notion of one strategy benefits all has been convincingly disproved. Recent translational science using multiomic approaches has characterized the differential response of various patient endotypes to transfusion, which demonstrates that specific patient characteristics may affect the response to transfusion therapy.^2,3

Moreover, a present challenge in transfusion medicine is that these data, which had been obtained from injured patients with unique coagulopathy, have been extrapolated to resuscitate bleeding in multiple different clinical settings. In the treatment of infectious diseases, it would be rudimentary and wrong to characterize all cases of pneumonia as similar and accept that a particular antibiotic, as an example, would be acceptable to treat any infection, regardless of source (viral, bacterial, fungal), much less the specific organism and microbial resistance pattern. This overly simplified example unfortunately highlights the lack of precision medicine in transfusion—presently a nonspecific, limited repertoire of blood products are available to treat bleeding of all types. Attention to the unique pathophysiology of the cause of bleeding and the likely variation among individual products derived from different donors almost certainly contributes to outcomes in resuscitation in a manner that modern medicine has yet to recognize. Precision diagnostics to accurately identify the specific component requirements of the bleeding patient are lacking. Herein lies the call for an increased focus on precision medicine for blood transfusion.

Design of Precision Transfusion Approaches

The call for precision in transfusion medicine can be applied to both recipients of blood products and to the products themselves. A pertinent example is postpartum hemorrhage, which unacceptably remains a leading global cause of death in women in low- and middle-income countries. Although there are some similarities between postpartum hemorrhage and trauma, including the lifesaving application of tranexamic acid,⁴ understanding of the hemostatic pathophysiology of postpartum hemorrhage is limited. Indeed, apart from the World Maternal Antifibrinolytic (WOMAN)⁴ study, few clinical trials have examined transfusion and hemostatic management. The contrast between the lack of research attention given to a major cause of death among young women compared with the global efforts to improve clinical outcome in trauma, the major cause of death of young men, is stark. Transfusion management in postpartum hemorrhage is currently derived from the application of trauma-derived massive transfusion protocols. These may be of benefit, but this is largely unstudied, and it leaves the clinician with only a blunt and unproven instrument in the treatment of a life-threatening condition that could end tragically in a death with major social and financial implications for the partner and the surviving children.⁵

In addition to conducting trials studying common bleeding scenarios beyond trauma, modern and sophisticated clinical trial design needs to be applied in transfusion medicine. The future of transfusion clinical trials should achieve precision medicine, ideally through adaptive platform trials that allow comparisons among multiple interventions and account for patient-level characteristics, including biomarker-driven studies to more rapidly identify patient cohorts that benefit and save valuable blood product resources by identifying promptly those who do not benefit.⁶

The next level of precision extends to the intervention, ie, understanding the nuances and complexities of the actual blood products themselves. Incompatibilities between donor and recipient have been well recognized since 1818 when James Blundell showed that blood from animals was not suitable for humans: later shown to be due to preexisting xeno antibodies in human recipients.⁷ The discovery of the ABO blood groups by Karl Landsteiner in 1901 demonstrated that mismatch in antigen expression existed between individual humans and that the presence of preformed antibodies against another human’s ABO antigens can also cause a life-threatening reaction.⁸ Today, extended red blood cell phenotype matching is automatically performed for those with a need for regular transfusions because the development of alloantibodies between those with different antigens can hinder further treatment. This means, however, that there can be shortages of donors who are well matched with a minority group requiring regular transfusions. This is highlighted in a UK campaign to increase blood and organ donation from Asian and Black individuals to support the 15 000 UK residents with sickle cell disorders.⁹

Ideally, the precision of donor-recipient matching needs to be increased. A recipient should receive blood from a fully matched donor or a donor whose allogenic antigens have been removed. It is therefore exciting to consider in the Recovery and Survival of Stem Cell Originated Red Cell (RESTORE) trial¹⁰ that red blood cells were grown from stem cells from donors and small amounts of red cells were transfused into recipients without adverse effects. Could this be scaled so that fully compatible red blood cells are obtained from a biological science facility? Similar studies to grow supplies of platelets from stem cells are also in progress.¹¹

Another consideration of donor blood is how it functions within a recipient. Increasing attention has been directed to the variation of function of blood constituents between donors. A recent study¹² has shown that platelet and cryoprecipitate from female donors improved coagulopathy in vitro more than from male blood donors. It begs the question: Do the hemostatic differences between male and female donors affect clinical outcomes? Even more broadly, might we identify ideal donor characteristics (age, sex, and cellular and molecular profiles) that might be best matched to a particular pathophysiology or cause of bleeding? A futuristic outlook includes the potential to design bioinspired products that overcome barriers of blood product storage and can be optimized for specific bleeding conditions.¹³

Implementation of Precision Medicine

Considerable variation in transfusion practice exists both between and within institutions. This has been well recognized since the Safe and Good Use of Blood in Surgery (SANGUIS)¹⁴ study in 1993, which showed that the transfusion rate depends more on the individual clinician than on the type of procedure, patient population, or hospital. This variation still exists today as evidenced by the International Point Prevalence Study of Intensive Care Unit Transfusion Practices (InPUT) study.¹⁵ Why does such variation persist? The reasons for the large variability of transfusion practice remain elusive, and clinicians’ attitudes appear slow to change. Patient blood management guidelines exist but are not well implemented. Barriers include access to knowledge, beliefs about the intervention, and tension for change.

In conclusion, blood transfusion practice has come a long way, but further efforts toward precision medicine are required to ensure that patients receive the most effective components. These products should be matched to patients as individuals who have unique antigens and a variable host response, and how to use the appropriate blood components in different clinical settings must be understood.