Age of transfused red cells had no effect on mortality (TRANSFUSE)

Oct 12, 2017

U.S. medical centers vary widely in the average shelf life of the blood in their blood banks. Trauma and high-volume surgical centers receive the oldest blood from the Red Cross, on the premise that they’ll be likely to transfuse it. All blood banks tend to dispense the oldest units first. This reduces waste of donated blood. In the U.S., the FDA sets the maximum shelf life of stored red cells at 42 days. Other developed countries’ practices are similar.

But could older blood be harmful? Blood cells deteriorate during storage, and in experiments, transfused older blood causes relatively more damage to traversed capillaries. The age of transfused blood has been hypothesized as a source of the numerous harms that can result from blood transfusions. Critically ill patients appear more susceptible to harm from transfused blood.

The large TRANSFUSE randomized trial in the New England Journal of Medicine strongly suggests these concerns are unfounded. Transfused blood was equally safe (or unsafe) in critical illness, regardless of the red cells’ storage time.

Australasian authors randomized 4,919 critically ill patients at 59 centers in 5 countries outside the U.S. to receive newer blood (12 days old on average) or older blood (average 22 days old) between 2012 and 2016. Patients were transfused with a wide variety of diagnoses; their average hemoglobin when transfused was 7.7 g/dL.

There was no difference in 90-day mortality, the primary outcome (24.8% with newer blood, 24.1% with older blood). Nor were there differences in mortality at 28 days; organ dysfunction; need for mechanical ventilation or renal-replacement therapy, or ICU days. At six months of follow-up, the survival between groups was virtually identical.

Interestingly, transfusing newer blood caused significantly more febrile transfusion reactions, which although not usually dangerous, result in stoppage of transfusion and wastage of the blood.

The TRANSFUSE trial is even more convincing in the wake of its largest predecessor, INFORM (in Canada), which also showed no significant difference in survival to hospital discharge among all hospitalized patients, or among critically ill patients, regardless of the age of blood transfused.

Although TRANSFUSE and INFORM were performed outside the U.S., the similar standardization of blood storage and transfusion practices across developed countries mean the age of transfused blood is unlikely to be important in the U.S., either.

Researchers will need to look elsewhere than the age of blood to explain the harms associated with blood transfusions. Immunogenicity is a potential source: even the best-crossmatched red cells inevitably contain numerous mismatches among the more than 300 antigens in 33 blood group systems. Only nine of these 33 systems are believed to be “clinically significant” (i.e., causing apparent transfusion reactions). Whether there could be subtler effects in critically ill patients from minor group mismatches, such as transiently increasing susceptibility to infection, is unknown.

Experts advise not transfusing blood in most anemic critically ill patients with a hemoglobin of at least 7 g/dL who are not actively bleeding (with exceptions considered for patients with cardiac disease, brain injury, and other conditions). Other studies have suggested no benefit of transfusing to higher targets in septic shock, that red cell transfusion may increase the risk of hospital-acquired infection, and that transfusion may be harmful in certain patients with GI bleeding.