Principles of Red Blood Cell Transfusion

Whole Blood

A unit of whole blood is collected in citrate phosphate dextrose adenine (CPDA)-1 anticoagulant, giving it a shelf-life of 35 days and a volume of approximately 510 mL (450 mL of blood plus 63 mL of CPDA-1). Within 24 hours of collection, the granulocytes are dysfunctional, and several plasma coagulation factor levels including factors V and VIII have fallen. Although clinicians have been taught that refrigerated whole blood stored has no functional platelets, evolving studies have demonstrated in vitro that platelet quantities and function are maintained for 10 to 14 days at 4°C after collection, leading to renewed interest in the use of whole blood for resuscitation.

Whole blood has the advantage of correcting simultaneous deficits in oxygen-carrying capacity, coagulation factor deficiencies, and blood volume. Therefore, whole blood is useful in the management of trauma or in surgical cases involving extensive blood loss. In this setting, whole blood has two distinct advantages: (a) it provides colloid osmotic pressure and coagulation factors not supplied by crystalloid solutions and (b) it does not expose the recipient to RBCs and plasma from different donors.

There has been recent renewed interest in fresh whole blood for patients with severe coagulopathy and shock. Few prospective trials have compared fresh whole blood with component therapy. The potential advantages for fresh whole blood are a relative increase in hemoglobin (Hb) concentration, coagulation factors, and platelets compared with component therapy. In addition, fresh products avoid some of the negative effects of storage and processing. A Food and Drug Administration (FDA) approved platelet-sparing leukocyte reduction filter is available for fresh whole blood. However, the use of fresh whole blood is limited due to the lack of published randomized controlled studies on specific indications and demonstration of significant clinical benefits, as well as logistical considerations such as wastage. The methods to achieve adequate pathogen removal and the determination of optimal storage conditions for whole blood products are the subjects of active investigations.

There has also been recent interest in the use of stored, low-titer group O whole blood (LTOWB) for civilian trauma building on successful military experience. “Low-titer” refers to low titers of anti-A (and sometimes anti-B as well) which are found in the plasma component of whole blood and may cause hemolysis of the recipient’s RBCs, but a universal testing process and titer threshold to define low-titer has not been established. LTOWB can be used if a recipient’s ABO group is unknown at the time of transfusion, but every local institution must adopt policies to define specific patient populations eligible to receive LTOWB and establish monitoring for potential hemolysis. Monitoring may involve regular post-transfusion hemolysis labs as well as clinical assessment of hemolysis symptoms, which are often non-specific (i.e., fever, back pain).

Increasingly studies in trauma patients have demonstrated a benefit to the early use of increased ratios of plasma and platelets to RBCs. The PROPPR study found there was a significant reduction in the time to hemostasis and a reduction in death by exsanguination within 24 hours for patients receiving a 1:1:1 ratio of RBC:plasma:platelets versus a 2:1:1 ratio. In addition, the PAMPer trial demonstrated a significant 30-day survival benefit in patients administered pre-hospital thawed plasma relative to standard-care resuscitation.

Red Blood Cells

RBCs (also referred to as packed RBCs or RBC concentrates ) are obtained from anticoagulated whole blood after removal of most of the platelet-rich plasma for the production of frozen plasma or platelets, or both. At most blood centers, the RBCs are then mixed with 100 mL of an additive nutrient solution that extends the storage period to 42 days and results in flow properties similar to those of whole blood.

RBCs are the product of choice for the correction of an isolated defect in oxygen-carrying capacity, as in cases of chronic anemia. In addition, RBCs rather than whole blood are most commonly used for the emergent transfusion of patients of unknown ABO type. The use of concentrated group O RBCs without plasma containing anti-A and anti-B can minimize potential hemolysis of the recipient’s red cells.

Leukocyte-Reduced Red Blood Cells

Leukocyte-reduced RBCs (LRRCs) can be prepared by a variety of methods, resulting in differing degrees of white blood cell (WBC) removal. Currently, the most widely used method of leukoreduction is pre-storage filtration that is performed at blood collection facilities after collection using blood bags with in-line filters. The various filters on the market result in greater than 99% leukocyte reduction while depleting less than 10% of the red cells.

The major indication for the use of LRRCs is the prevention of the febrile nonhemolytic transfusion reaction, a common adverse effect of transfusion, particularly in multiply transfused patients or multiparous females. These reactions are believed to be mediated by antibodies directed against leukocyte antigens (human leukocyte antigen, HLA, or human neutrophil antigen, HNA). Depletion of leukocytes to less than 5 × 10 ⁶ has been shown to prevent, or at least ameliorate, such reactions in most patients. Increasing evidence suggests that cytokines play a role in causing these reactions. Because cytokines may be released from leukocytes during storage, pre-storage leukoreduction is preferred.

A second important indication for LRRCs is the mitigation of alloimmunization to HLA antigens that can adversely affect post-transfusion platelet increments and the search for HLA-compatible donors for stem cell and solid organ transplantations. This approach will be effective only if leukoreduced platelets are also used. According to current AABB standards, the total leukocyte number must be less than 5 × 10 ⁶ when intended for this purpose, an achievable goal with third-generation leukoreduction filters. A multicenter study known as TRAP (Trial to Reduce Alloimmunization to Platelets) showed that the use of leukoreduction filters for platelet products significantly decreased the rate of alloimmunization but did not completely eliminate the problem.

A third indication for the use of LRRCs is to prevent transfusion-transmitted cytomegalovirus (CMV). CMV is found in low copy numbers outside of cells, and leukoreduced blood components are considered CMV safe. RBCs that are either leukoreduced or from CMV-seronegative donors are associated with only 1% to 1.5% CMV-transmission rates; thus, many institutions provide leukoreduced RBCs to mitigate CMV transmission.

A final indication for the use of LRRCs is to prevent transfusion-related immunomodulation (TRIM). A large meta-analysis previously demonstrated that patients who receive a blood transfusion are more likely to experience a postoperative infection than patients not transfused. This effect has been found to be dose-dependent and is thought to be mediated by suppressing the patient’s immune function. In addition, some studies have suggested that TRIM may be associated with cancer recurrence, but other studies have refuted this notion. The mechanism of TRIM has yet to be defined, but multiple theories have been proposed. It has been suggested that immunologically active WBCs or soluble biologic response modifiers released from WBCs during storage downregulate the recipient's immune function. Alternative theories suggest that soluble mediators circulating in allogeneic plasma may have an immune-modulatory effect. Universal leukoreduction has been found by some studies, but not by others, to mitigate the immunomodulatory effect of allogeneic transfusions.

Washed Red Blood Cells

RBCs are washed using isotonic saline solutions mainly by automated techniques. There is always some degree of RBC loss with each wash cycle. Because most automated methods utilize open systems, the resulting product must be transfused within 24 hours to prevent bacterial growth.

The primary aim of washing is to remove plasma proteins, although some leukocytes and platelets are removed simultaneously. The major indication for washed RBCs is the prevention of severe allergic transfusion reactions, most likely mediated by recipient IgE antibodies against donor plasma proteins. Washing is recommended when reactions are severe and refractory to pre-transfusion steroid and antihistamine administration. In IgA-deficient patients who have preformed antibodies to IgA, the exposure to IgA-containing plasma can cause anaphylaxis. Therefore, high volume cell washes may be required to prepare cellular components for transfusion in IgA-deficient patients.

Irradiated Red Blood Cells

RBCs are irradiated with either gamma-ray or X-ray technology with a minimum dose of 25 Gy not exceeding 50 Gy. RBCs expire 28 days after irradiation or on the original expiration date, whichever is sooner. A method is used to ensure that irradiation has occurred with each batch. The primary aim of irradiation is to prevent the rare, but often fatal risk of transfusion-associated graft-versus-host disease (TA-GVHD) by the abrogation of the proliferative potential of donor T lymphocytes. GVHD can occur after the transfusion of immunologically competent donor lymphocytes, usually to an immuno-incompetent recipient. Some patient populations with indications for irradiated products include neonates, patients with hematologic malignancies, stem cell transplant recipients, and patients with congenital immune deficiencies. There is still much debate among experts regarding which additional patient populations may be at risk for TA-GVHD. It has been suggested that a policy of universal blood component irradiation could prevent TA-GVHD in patients with currently unsuspected risks, including advanced age, unrecognized immune deficiencies, or unsuspected donor-recipient immune similarities. If efficacious systems to apply pathogen reduction to red cell components by induction of DNA/RNA cross-linking become approved, another advantage would be the inactivation of white cells, which would eliminate the need for gamma irradiation to prevent TA-GVHD.

Frozen Red Blood Cells

RBCs can be frozen (with glycerol used as a cryoprotective agent) and stored at −65°C or colder. The required concentration of glycerol depends on the rate and the temperature of freezing. The freezing process destroys other blood constituents, except for a small percentage of immunocompetent lymphocytes. RBCs are prepared for transfusion by thawing and washing away the glycerol using a series of progressively less hypertonic crystalloid solutions, allowing glycerol to diffuse gradually from the cells to prevent hemolysis. The cells are resuspended in an isotonic saline solution containing glucose. The extensive washing removes approximately 99.9% of the plasma as well as cellular debris.

RBCs can be stored in the frozen state for at least 10 years with good viability. After thawing and washing, storage is typically limited to 24 hours because of the open system. Frozen cells have been shown to maintain prefreezing adenosine triphosphate (ATP) and 2,3-diphosphoglycerate (DPG; also known as 2,3-bisphosphoglycerate or 2,3-BPG) levels. To maintain these factors at high levels, the standard is to freeze within 6 days of collection. When it is necessary to freeze older units, rejuvenation with a solution containing pyruvate, glucose, phosphate, and adenine has provided excellent results. The major indication for frozen RBCs is the stockpiling of rare donor units for patients who have developed alloantibodies. Some patients with rare phenotypes can make autologous donations that can be frozen for later use. Cells from autologous donors can be frozen if more units are required than can be collected in the 42-day liquid storage period or if surgery is postponed. Because of the high cost, lower product quality, and cumbersome nature of freeze-thaw procedures, other uses of frozen RBCs are somewhat difficult to justify.

Chronic Anemia

As a rule, signs and symptoms attributable to anemia are unlikely to develop at a Hb level of greater than 7 g/dL. When the anemia is of gradual onset, the body’s compensatory mechanisms for maintaining oxygen delivery to the tissues come into play. Both cardiac output and intracellular 2,3-DPG increase, and thus, oxygen unloads at a lower oxygen saturation of Hb. When chronic anemia is due to red cell destruction, the healthy bone marrow responds by increasing erythropoiesis up to sixfold.

RBC transfusion provides symptomatic support rather than definitive therapy for anemia. Transfusion should be used only when there is no definitive treatment for the underlying cause (i.e., iron, vitamin B12, or folate supplementation) or when the severity of the anemia and the clinical manifestations in the patient make it impossible to wait for the effects of treatment to be realized.

Generalizations about RBC transfusion indications and practices are difficult to make and are usually inappropriate. The clinical impact of anemia varies depending on its pathogenesis, rate of onset, the presence or absence of accompanying hypovolemia, and, most importantly, the individual patient. The Hb level at which a given individual manifests the signs and symptoms of anemia relates, in part, to underlying health status, cardiorespiratory reserve, and tissue oxygen demand. Studies have shown that the risk of death increases with the severity of chronic anemia; it is prudent to maintain the Hb level above four for otherwise healthy patients and six for patients with cardiac disease to avoid sudden cardiac arrhythmias that are often fatal.

Perioperative Period

Many generalizations have been made about the appropriate transfusion management of acute blood loss, often with little evidence to support the arguments. One rule of thumb is that blood loss of 10% or less of total blood volume requires no replacement therapy at all; loss of up to 20% can be replaced exclusively with crystalloid solutions; and loss of greater than 25% generally requires RBC transfusion to restore oxygen-carrying capacity, along with crystalloid and sometimes colloid solutions to restore intravascular oncotic pressure to achieve adequate perfusion. A RBC transfusion trigger of 7 g/dL of Hb is now commonly used during the perioperative period. Each case must be evaluated individually on the basis of clinical signs and symptoms, rather than laboratory values alone. If the cardiovascular system is healthy and the degree of hypoperfusion is not significant, good tissue oxygenation can be maintained at much lower Hb levels. Many anesthesiology and surgery clinical guidelines suggest that most surgical patients do not need transfusion unless the Hb level falls to less than 7 g/dL. Given that RBC transfusion should be tailored to individual needs, the question arises as to whether there is any readily available, objective measurement that can be used to determine how low the Hb level can safely be allowed to fall before RBC transfusion is initiated.

Global hemodynamic parameters do not always correlate with microvascular perfusion. Assessment of tissue oxygenation at the microvascular level would help evaluate the effectiveness of a red cell transfusion, evaluate the effect of red cell storage on end-organ perfusion, and provide data about when to transfuse. Several general methods are available to evaluate the microcirculation and include direct assessment using image techniques and indirect methods of assessment, such as measures of microvascular oxygen availability and function. Direct assessment can be performed using laser Doppler flowmetry, imaging of the microcirculation, intravital microscopy, orthogonal polarization spectral imaging, and sidestream dark-field imaging. Assessments of oxygen availability include oxygen electrodes, reflectance spectrophotometry, and near-infrared spectroscopy. The techniques described are currently considered research tools and are not available in routine clinical practice. None has yet to prove reliable and reproducible in the clinical setting. Further, some are only useful in specific organ systems and do not reflect the global oxygenation of the patient. To be useful at the bedside, a technique must be technically simple, rapid, and noninvasive without large interoperator variation. Such a device has yet to become available.

In 2019, the Society of Cardiovascular Anesthesiologists published clinical practice guidelines focusing on blood conservation interventions, such as erythropoietin or antifibrinolytic administration, intraoperative blood salvage, or normovolemic hemodilution, and provided practical intraoperative transfusion algorithms guided by point-of-care viscoelastic methods and standard coagulation laboratory values.

Randomized clinical trials have evaluated the effects of different transfusion thresholds in distinct clinical settings; however, the thresholds used in the studies differ widely. Many of the studies found no difference in outcome. Most studies were not powered to adequately evaluate clinically important outcomes. Few studies included more than 100 patients. The TRICC trial included 838 intensive care unit (ICU) patients who were randomized to a restrictive transfusion strategy (transfused at Hb 7 g/dL) or liberal strategy (transfused at 10 g/dL). The 30-day mortality was slightly lower in the restrictive group (18.7% vs. 23.3%), but not significantly lower. The FOCUS trial, a 2600 patient, multicenter randomized trial designed to determine whether patients with cardiovascular disease or risk factors undergoing surgical repair of the hip benefit from a lower (<8 g/dL) or liberal transfusion trigger (transfused at 10 g/dL) showed that liberal transfusion did not reduce mortality or in-hospital morbidity in this patient cohort. Other studies have now been published in sepsis (TRISS), upper gastrointestinal (GI) bleeding (Villanueva et al.), pediatric intensive care (TRIPICU), and cardiac surgery (TRACS, TITRe2) that support the use of conservative transfusion triggers.

Red Blood Cell Transfusion in Neonates

In neonates, it is convenient to consider periodic, small-volume transfusion separately from massive transfusion situations. The trigger for transfusion and the optimal type of component are very different in these two settings. The potential adverse effects may be quite distinct.

Low-volume RBC transfusion is rarely indicated in full-term infants unless acute blood loss has occurred at birth or an intrauterine situation has led to prenatal anemia. In contrast, premature infants are frequent recipients of transfusions. In the intensive care setting, the premature infant is subjected to frequent blood sampling, and iatrogenic anemia may necessitate transfusion. Anemia of prematurity is also a well-recognized entity; premature infants have a slightly lower Hb value at birth. In addition, the postnatal decline in Hb occurs earlier and is more pronounced in premature infants. The mechanism for anemia of prematurity appears to involve a relatively lower output of erythropoietin in response to a given degree of anemia. This phenomenon is attributed in part to the fact that the liver, rather than the kidney, is the major site of erythropoietin production in these infants. Although some practitioners have considered this degree of anemia to be physiologic, the benign nature of this condition remains controversial.

Another debate among neonatologists concerns the triggers for RBC transfusion, as in what clinical signs and symptoms are valid reflections of poor tissue oxygenation. Congestive heart failure and severe pulmonary disease are generally accepted indications for transfusion, but recurrent apnea, tachypnea, tachycardia, and failure to thrive are also used as transfusion triggers. In recent times, the rate of transfusion and the donor exposure rate of premature infants have consistently declined. These changes, however, reflect improvements in patient care (e.g., microtesting methods resulting in less iatrogenic blood loss; the use of surfactant resulting in decreased respiratory distress and the use of a single unit to supply one infant over a longer period) rather than being attributable to changes in the transfusion trigger. The prolonged use of single units has become possible with the advent of the sterile docking technology that preserves the full shelf-life of the unit of RBCs, as well as with the accumulating evidence that fresh blood is not necessary for low-volume transfusions in neonates because supernatant potassium and decreased pH are not of concern in this setting. Recent studies have provided additional information about the relative risks and benefits of using restrictive rather than more liberal criteria for very low birth weight infants. Two studies showed that a liberal transfusion practice resulted in more infants receiving transfusion but conferred little evidence of clinical benefit. One study showed a lower risk of apnea and major brain injury for the liberal transfusion arm of the study. Therefore, the safest transfusion trigger in the preterm infant still remains unclear, and further studies are indicated. Most institutions assess the clinical situation and consider the postnatal age and whether a neonate has oxygen requirements when determining the need for a RBC transfusion.

The dose of a RBC transfusion in a neonate can vary by institution between 10 and 20 mL/kg. Few studies have assessed the optimal dose in this patient population, and further studies are needed. Paul et al. compared 10 and 20 mL/kg and found that the larger volume did not cause impaired pulmonary function. Wong et al. demonstrated extra transfusion episodes could be avoided with 20 mL/kg versus 15 mL/kg, without any additional risk to the patient. Many transfusion services now routinely use RBCs stored in additive solutions for low-volume RBC transfusion and thus prefer a dose of 20 mL/kg to account for the lower hematocrit of an additive unit.

Finally, several trials of erythropoietin therapy in premature infants have been undertaken. The administration of relatively high-dose erythropoietin has been shown to raise Hb levels and reticulocyte counts in healthy premature infants, but the effect in sicker neonates is unclear. Although transfusion exposure was decreased, the significance of this observation is diminished, given the promise of new strategies for limiting transfusions and donor exposure. The high cost and the increased risk of retinopathy associated with erythropoietin treatment does not justify its use in this patient population.

In the case of massive transfusion, the situation differs. There have been marked increases in massive transfusion in recent years in full-term as well as premature infants. Hemolytic disease of the newborn remains a prominent indication for exchange transfusion; however, the recent use of intravenous immune globulin to decrease red cell antibody levels in newborns has decreased the necessity of this procedure. The two triggers for exchange are (a) rapidly rising levels of unconjugated bilirubin that may lead to kernicterus and permanent central nervous system damage and (b) congestive heart failure secondary to severe anemia. Whole blood exchange transfusion is especially beneficial in cases of hemolytic disease of the newborn because it clears the bilirubin, the offending antibody, and the antibody-coated red cells before lysis while providing a source of red cells lacking the offending antigen. A two-blood-volume exchange is commonly performed by using a fresh unit of blood concentrated to a final hematocrit level of approximately 50%. In cases of hyperbilirubinemia resulting from other causes (e.g., liver immaturity in premature infants), phototherapy is the treatment of choice because its effects are usually more sustained, and exchange transfusion is used only for cases of marked elevations. Extracorporeal membrane oxygenation and open-heart surgery are two other situations when the neonate may be exposed to large volumes of allogeneic RBCs. The extracorporeal membrane oxygenation circuit requires a prime with RBCs, as do many of the types of extracorporeal circuits used for cardiopulmonary bypass.

Although accumulating evidence supports the safety of using RBC units of any age and with any preservative solution for low-volume transfusions in neonates, the same transfusion policies may not apply to massive transfusion. Newborn physiology is unique in several ways that may have implications for massive transfusion therapy. The newborn does not handle metabolites in a mature fashion. Renal immaturity may lead to problems in clearing potassium or acid from stored RBCs, and the immature liver may not catabolize citrate efficiently. These problems are accentuated and protracted in the premature infant. To address the concern about potassium load, fresh (<7 days old) or washed RBCs are often used, although the necessity of this practice is actively debated. Fresh blood may also be preferred because of its higher 2,3-DPG levels and better red cell integrity. The citrate problem is probably best handled by using slow infusion rates because the use of bicarbonate or calcium replacement to counteract the acid load or calcium-chelating effects of citrate is controversial. AS-3 may be preferentially chosen over AS-1 because of the theoretical concern over mannitol content that is absent in AS-3 but present in AS-1. However, in the context of multiple, small-volume transfusions, either AS is considered safe.

The humoral and cellular immune systems of the neonate are immature, especially in the premature infant. There is a small but real risk of transfusion-induced GVHD in premature infants receiving RBC transfusions and in the fetus undergoing intrauterine transfusion. Irradiation of RBCs should be performed in both settings. Another risk of transfusion in low birth weight (<1500 g) premature infants is the development of clinical CMV infection in infants of CMV-seronegative mothers. CMV-safe blood, either CMV seronegative or leukoreduced, should be provided to these infants. Some retrospective observational studies have suggested that other rare RBC transfusion-associated complications in this particularly vulnerable patient group may include necrotizing enterocolitis, intraventricular hemorrhage, bronchopulmonary dysplasia, and retinopathy of prematurity. However, a causal relationship has yet to be established through clinical studies. In multivariable analyses, some studies have suggested that severe anemia, and not transfusion itself, may be associated with these outcomes.

Ideas to prevent anemia and decrease donor exposure in premature infants and other neonates include delayed cord clamping, cord milking, umbilical vein blood sampling from the delivered placenta, and autologous cord blood transfusion. One review of 10 delayed cord clamping studies demonstrated lower transfusion requirements in the delayed versus early clamped group. However, a randomized controlled trial by Strauss et al. found no difference in transfusion needs between delayed and early clamped groups. A few studies have looked at autologous cord blood transfusions in neonates. One study found that the amount of blood harvested was insufficient to cover all transfusions in low-birth-weight infants. In addition, studies have demonstrated that blood processing problems, bacterial contamination, and costs are all barriers to the routine collection and autotransfusion of cord blood.

Red Blood Cell Transfusion in the Allogeneic Hematopoietic Stem Cell Transplantation Recipient

Red cell engraftment is usually the last phase of hematopoietic recovery after stem cell transplantation; therefore, RBC transfusion is common during the posttransplantation period. As hematopoietic stem cells and progenitor cells lack ABO antigens, the transplantation outcome is not significantly affected by the red cell antigen/antibody incompatibility between the donor and the recipient. However, RBC transfusion requirements and blood product selection can vary significantly depending on the type of ABO incompatibility. The phenotyping of non-ABO/Rh red cell antigens from the donor and recipient is not required in the absence of a positive red cell antibody screen in the recipient. Patients with an autologous stem cell transplant have fewer RBC transfusion requirements when compared with patients receiving allogeneic stem cell transplantation.

In the setting of allogeneic stem cell transplantation, there are four major categories of ABO antigen matching: full compatibility, minor incompatibility, major incompatibility, and bidirectional incompatibility. In addition, Rh type must be taken into consideration. Rh-positive recipients with Rh-negative donors should receive Rh-negative RBCs, but Rh-negative recipients with Rh-positive donors may receive Rh-positive RBCs. Apheresis platelets and plasma may be given without regard to Rh. The blood product selection algorithm is shown in ( Table 112.2 ). Minor incompatibility is defined as the presence of blood group antibodies in donor plasma (e.g., group O donor to group A recipient). In minor incompatible stem cell transplantation, the incompatible plasma in the donor stem cell product may result in some hemolysis of the recipient's endogenous RBCs during the early phase of the posttransplantation period. Minor incompatibility is occasionally complicated by passenger lymphocyte syndrome, where transient hemolysis may occur if donor-derived lymphocytes in the stem cell graft remain viable and form blood group–specific antibodies that are incompatible with the recipient's red cells. Typically, passenger lymphocyte syndrome involves ABO incompatibility, but hemolysis resulting from serologic incompatibility in other blood group systems has been reported. If hemolysis increases a few days after transplantation, passenger lymphocyte syndrome should be considered. Once the donor is engrafted and incompatible recipient red cells are removed, donor red cells will have normal survival in the recipient. Major incompatibility is defined by the presence of blood group antibodies in recipient plasma (e.g., group A donor to a group O recipient). Major ABO-incompatible transplantation may result in pure red cell aplasia (PRCA) due to persisting incompatible ABO antibodies targeting the donor's engrafting erythropoietic precursors expressing ABO antigens. Although there may be no evident hemolysis, the donor red cell engraftment could be further delayed and result in prolonged RBC transfusion support. Finally, bidirectional incompatibility is defined as the presence of incompatible ABO antigens and antibodies contributed by both donor and recipient (e.g., group A donor to group B recipient). Bidirectional incompatible transplants may cause the problems associated with both minor and major ABO-incompatible transplants. To predict the severity of these complications and provide management, an ABO antibody titer can be performed on either the stem cell product or the recipient. However, titers do not correlate perfectly with the clinical outcomes. When a high titer incompatible ABO antibody is discovered in a minor incompatibility, the stem cell product can be plasma-reduced to avoid an immediate hemolytic reaction. Occasionally, in major incompatibilities, plasma exchange can be considered if the recipient has a high level of incompatible ABO antibodies to prevent hemolysis at the time of transplant or PRCA.