Whole Blood versus Components for Transfusion in Trauma and Life-Threatening Hemorrhage-Unresolved Issues

COHORT AND CLINICAL TRIALS OF WHOLE BLOOD TRANSFUSION

Torres et al performed a retrospective cohort study (N = 2795) to determine whether WB as an adjunct to component therapy-based MTP is associated with a survival advantage over component therapy alone in adult civilian trauma patients with severe hemorrhage. The analysis found a mortality benefit at 24 hours (hazard ratio, 0.63, 95% confidence interval, 0.41–0.96, P = .03) and 30 days (hazard ratio 0.53, 95% confidence interval, 0.31–0.93, P =.02) associated with WB and components compared to component therapy alone.³ Cotton et al conducted a pilot randomized clinical trial of (N = 107) comparing WB MTP with component therapy MTP. This study found no difference in secondary outcomes of 24 hour (P = .62) or 30-day mortality rates (P = P = .62).⁴ Hanna et al conducted a retrospective analysis comparing WB MTP with component therapy MTP in trauma patients, and found a reduction in 24-hour mortality (17% vs 25%, P = .002) and hospital mortality (29% vs 40%, P < .0001) associated with WB MTP.⁵ Hazelton et al published an abstract for a prospective multicenter study in patients (N = 299) of all ages (including pediatric patients) which found that WB MTP was associated with a 48% reduction in in-hospital mortality.⁶

The data in most of these studies are potentially confounded by multiple factors that could affect survival. First, selection bias and time-to-treatment are major concerns. Patients who received WB earlier may represent a subset of patients with more immediate access to treatment compared to those receiving component therapy, potentially skewing survival results. For instance, a cohort study by Torres et al (N = 1394) of patients who presented with severe traumatic hemorrhage found an improvement in survival benefit as early as 1 hour after emergency department arrival, with improved 24-hour survival (adjusted hazard ratio, 0.40; 95% confidence interval, 0.22–0.73, P =.003) and 30 day survival (adjusted hazard ratio, 0.32; 95% confidence interval, 0.22–0.45, P <.001) associated with earlier WB administration.⁷

Second, data on concomitant medications that reduce bleeding, such as tranexamic acid and/or arginine vasopressin, are often lacking in these studies, which could confound the observed data on WB versus component therapy.³ Third, differences between fresh and cold-stored WB introduce another variable. Military studies often transfused fresh WB, which maintains minimal loss of labile clotting factors and platelet activity. In contrast, civilian settings primarily use cold-stored WB. Although recent randomized trials have not found increased mortality based on red cell storage lesions, despite abundant cohort data.,^8,9 these existing RCTs did not include WB. Furthermore, these RCTs were not designed to compare patients transfused with very long-stored red cells in the comparative arm, nor did they account for the administration of other potentially toxic blood components, such as prothrombotic and proinflammatory platelets.

May et al compared the biologic profile of component therapy versus WB (N = 5 samples). The investigators found that component therapy resulted in a reduction in hematocrit concentration (29% vs 40%–50% for WB), coagulation factor activity (65% vs 100%), platelet count (88,000 per microliter vs 150,000–400,000 per microliter), and fibrinogen (750 vs 1000 mg/dL), all supporting the possible superiority of WB.¹⁰ However, these in vitro measures (hematocrit, prothrombin time (PT)/partial thromboplastin time (PTT)/international normalized ratio (INR), platelet count, and fibrinogen) do not necessarily correlate with clinical outcomes after hemorrhage. Marginal abnormalities in these metrics are usually of little clinical significance. Conclusions about the “trauma-induced coagulopathy” that rely on minimal changes in PT and aPTT could be misleading. For instance, the PT and aPTT are approximately 1.5 times the midpoint of the normal reference range when 30% residual clotting factors remain, which clinically may represent sufficient clotting factor activity.^11,12 Approximately 30% clotting factor levels typically persist after one blood volume has been lost and replaced with red blood cells. Dilutional severe thrombocytopenia, which occurs later than coagulation factor deficiency, manifests after approximately 2 blood volume replacement.^,12

ADVERSE EFFECTS OF WHOLE BLOOD AND BLOOD COMPONENT TRANSFUSION

Like any blood product, WB is not without risk, carrying all the same risks of component transfusion. Transfusion-transmitted infections and transfusion reactions, including allergic, febrile, transfusion-associated circulatory overload/congestive heart failure (TACO), and transfusion-related acute lung injury/acute respiratory distress syndrome (TRALI) are dose-dependent adverse events.

Specifically, WB has been associated with increased severe complications in some observational data. A quality improvement study by Gallagher et al (N = 83,924) found WB receipt associated with severe sepsis, even after adjusting for age, sex/gender, mechanism of injury, and injury severity score.¹³ Similarly, Moreno et al performed a retrospective cohort study, finding an association between WB and the development of acute respiratory distress syndrome (ARDS) among trauma patients who received blood transfusions, which persisted after adjustment for confounders and sensitivity testing.¹⁴

The association of WB with increased sepsis and ARDS risk, if true, remains of uncertain pathophysiology. Possible mediators include the known proinflammatory and immunomodulatory effects of all allogeneic blood transfusions. Mediators such as supernatant cytokines, microparticles, DNA, histones, platelets, and white cells are thought to be involved. This risk is potentially relevant when comparing WB (containing stored supernatant and platelets), with red cells that have minimal supernatant and few platelets.

If the WB is not leukoreduced, this could theoretically contribute to new or worsening organ dysfunction and increased hospital-acquired infections. Furthermore, platelets and platelet-derived factors, which are numerous in WB (but not leukoreduced red cells), likely contribute to these immunomodulatory effects. For example, platelet-derived microparticles are capable of inducing both immune suppression and activation, which aligns with the known important role platelets play in modulating immune response.¹⁵

POTENTIAL BENEFITS AND TOXICITIES OF ABO IDENTICAL/MISMATCHED TRANSFUSIONS VERSUS LOW-TITER GROUP O WHOLE BLOOD TRANSFUSIONS

Biocompatibility and immunologic compatibility are important factors for transfusion efficacy and safety in all patients. Some studies suggest that low-titer group O WB (low levels of anti-A and anti-B antibodies) has superior efficacy compared to component therapy.^16,17 However, one key factor that has not been investigated at all is the ABO blood group of the patient and the number and volume of ABO-nonidentical components transfused.

ABO-nonidentical transfusions contain incompatible, toxic antibody/antibodies and incompatible, toxic soluble and cellular antigens. Multiple observational studies and in vitro modeling demonstrate that ABO-nonidentical transfusions are associated with increased bleeding, inflammation, impaired platelet function and thrombin generation, endothelial cell damage, and increased bleeding and mortality compared to ABO identical transfusions.^18–21

Although the conventional wisdom has been that group O red blood cells (RBCs), group AB plasma, and any ABO group of platelets and cryoprecipitate are equally effective and safe as ABO identical products, there is no evidence for this dogma. Infusing incompatible soluble A and B antigens to patients with the corresponding antibody is associated with increased bleeding and impaired hemostasis.²² An example is transfusing group AB or A plasma to O recipients. Future studies comparing WB versus component therapy should collect data on ABO identical and ABO-nonidentical blood transfusions and assess what volume of incompatible blood components were transfused. Comparisons of WB with component therapy are not reassuring, as component therapy as currently practiced involves ABO-nonidentical transfusions to almost all recipients.

A potential reason why low-titer group O WB might be more effective or less toxic than component therapy alone is that it is ABO identical with 45% of patients. Component therapy is ABO identical with close to 0% of patients because it typically involves group O red cells, group AB or A plasma, and any ABO type platelets and cryoprecipitate. Low-titer group O WB is ABO identical with approximately half of all patients, and entails infusion of smaller amounts of incompatible ABO antibody compared with untested group O red cells. ABO major incompatible platelet transfusions, for example, are significantly less effective and significantly more toxic than group O platelets, being associated with both increased bleeding and mortality.^18–22

Another nuance in this literature is that the ABO blood group of patients may independently influence bleeding and mortality in trauma patients. Group O patients tend to bleed more than patients of other ABO groups. This is postulated to be related to von Willebrand factor (vWf) levels. Group O individuals have the lowest vWf levels, and group AB have the highest levels.²³ One might expect earlier mortality in group O patients due to bleeding. In contrast, group A, B, and AB patients might be less likely to have early hemorrhagic deaths. Group B and AB patients, who are only 10% and 5% of patients, are also most likely to receive ABO-nonidentical transfusions during resuscitation. Therefore, nongroup O patients might potentially be more likely to experience later increased inflammation, bleeding, and organ injury due to ABO mismatched transfusions.²⁴

Seheult et al performed a retrospective analysis of trauma patients receiving low-titer group O WB during the first 24 hours of admission.²⁵ Patients receiving WB were matched with those who received at least one red blood cell unit during the first 24 hours of admission. The analysis found no significant differences between the groups in overall and 24-hour mortality, hospital and intensive care unit length of stays, and number of RBC units transfused. As with all previous studies, no data were collected on the ABO blood group of recipients, nor on the volume and type of ABO mismatched components transfused. Although it has been demonstrated that transfusion of up to 4 units of low-titer group O WB to civilian trauma patients does not cause significant changes in laboratory or clinical outcomes, these studies were focused on hemolysis, as opposed to other key clinical outcomes such as length of stay, hospital-acquired infections, and mortality.^26–28

Finally, issues of clinical outcomes need to be characterized similarly to those found in recent publications on transfusion in hemorrhagic resuscitation.^29–34

SUMMARY AND CONCLUSIONS

There are as yet no published RCTs that prove that WB MTP is superior to component therapy MTP. There are no comparisons of routine MTP with viscoelastic testing-driven component therapy. The existing data, although suggestive of improved survival, are thus limited to association rather than causation. Data on medication administration (such as anti-fibrinolytics and platelet function-enhancing drugs) are also frequently lacking.

Crucially, there is a complete absence of data on the ABO type of the recipients and the use and quantity of ABO mismatched transfusions. WB is subject to similar risks as separate blood component transfusions. The observational data suggesting increased risks of sepsis and ARDS in WB recipients, compared with component recipients, is consonant with known immunomodulatory effects. This finding raises the question of whether WB may cause additional adverse clinical outcomes in some patients or circumstances. It will be important to assess whether patients who only receive a small number of red cells (perhaps <5) or WB do better or worse with WB. WB would be expected to be more toxic than red cells alone, in terms of thrombosis, risk of nosocomial infection, inflammation, and organ injury. This issue has yet to be explored in existing studies.

Accounting for patient ABO group and quantity of infused ABO-nonidentical WB and other transfusions will significantly improve the likelihood of studies providing clinically useful and accurate information for resuscitation. If improvement in clinical outcomes is, at least in part, due to transfusing ABO identical low-titer group O WB to 45% of patients, further studies of ABO identical transfusions (both WB and component therapy) will be needed. These studies will hopefully clarify the true efficacy, safety, and clinical utility of low-titer group O WB versus component therapy in life-threatening hemorrhage.

Finally, there may be a need for a more individualized approach, given the heterogeneity of trauma and similar patients. There is, at best, modest evidence for balanced transfusions overall, and for WB substitution for component therapy in specific. The likelihood that any single approach will fully address the complexity and variability of life-threatening hemorrhage in trauma and other clinical settings remains uncertain.

DISCLOSURES

Conflicts of Interest: None. Funding: None. This manuscript was handled by: Shannon L. Farmer, DHSc.