Transfusion and Cellular Therapy in Pediatric Sickle Cell Disease

Transfusion management of sickle cell disease

Red blood cell (RBC) transfusion is an essential treatment for sickle cell disease (SCD), with more than 50% of affected children receiving at least one transfusion.¹ RBC transfusion improves blood oxygen delivery by reducing the percentage of hemoglobin S (HbS%) and blood viscosity. Transfusion continues to be a common therapy for prevention and management of acute and chronic complications of SCD.^2–4

RBC transfusion method

RBCs can be infused via simple transfusion or by red blood cell exchange (RBCX).⁵ A comparison of the different transfusion methods is summarized in Table 1. Simple transfusion can be administered via peripheral access and does not require specialized devices or trained personnel. Chronic simple transfusion inevitably leads to iron overload, and thus, initiation of iron chelation therapy is recommended after 1-2 years of transfusions and when the serum ferritin exceeds 1000 ng/mL on two occasions or the liver iron content is >3 mg/g dry weight.⁶ To avoid potential blood hyperviscosity, simple transfusion should only raise the hemoglobin level (Hb) to 10-11 g/dL, particularly when HbS% is >50%.

Table 1.

Comparison of Blood Transfusion Method

	Simple Transfusion	Manual RBCX	Automated RBCX
Specialized instrument	No	No	Yes
Trained personnel	No	Yes	Yes
Vascular access	Peripheral	Central venous catheter might be required	Central venous catheter might be required
RBC exposure	Fewest	Intermediate	Highest
Hyperviscosity	High risk*	Intermediate risk*	Minimal risk
Fluid balance	Potential fluid overload	Minimal volume shifts	Isovolemic
Iron overload	Inevitable	Intermediate risk	Minimal risk

^*Minimized by avoiding post-transfusion hemoglobin >11 g/dL

HbS, hemoglobin S; RBCX, red cell exchange

RBCX replaces RBCs containing HbS with healthy donor cells, thereby reducing HbS% with minimal effects on blood viscosity, fluid balance, and iron burden. RBCX is indicated for patients with high pre-transfusion Hb levels (i.e., Hb > 9-10 g/dL), for specific complications including acute ischemic stroke, for patients with cardiac disease who are unable to tolerate the added volume of a simple transfusion, and for individuals with iron overload and unable to iron chelate. For RBCX, a target HbS% of ≤ 30% is often used to prevent and manage most SCD-associated complications.^7,8 Although RBCX increases donor unit exposures, it does not appear to increase the risk of alloimmunization.⁹ RBCX can be performed by manual RBCX, automated RBCX, and automated RBCX with isovolemic hemodilution (RBCX-IHD). Manual RBCX sequentially removes patient blood and infuses donor RBCs, and typically requires 2 to 8 hours to complete. Automated RBCX can reach target hematocrit (Hct) and HbS% levels in 1–3 hours and can maintain isovolemia to support hemodynamic stability. Automated RBCX requires apheresis devices and trained staff, which may not be available in small community centers. For pediatric patients with low body weight, apheresis devices should be primed with donor RBCs to prevent substantial intra-procedure volume shifts and acute anemia due to the relatively large extracorporeal volume of the apheresis devices for these patients. Central venous access is often required for automated RBCX to provide sufficient venous access for the rapid blood flow rates and high negative pressure in the withdrawal line. For patients requiring chronic transfusion therapy, automated RBCX is recommended over simple transfusion or manual RBCX if possible.² Automated RBCX-IHD includes an initial phase of isovolemic hemodilution or depletion in which patient blood is replaced by saline or 5% albumin instead of donor RBCs. This may slightly decrease the donor unit requirement, compared with conventional automated RBCX, to reach the same target HbS% level. Although RBCX-IHD maintains isovolemia, saline/or albumin replacement acutely lowers the patient’s Hct. Therefore, RBCX-IHD is not recommended for patients with acute complications (i.e., stroke or acute chest syndrome), or for those with severe central nervous system disease, vasculopathy, or cardiopulmonary disease, in which further anemia during IHD phase may be detrimental.² It is recommended not to decrease the Hct to < 21% and/or more than 20% from the patient’s baseline.² For patients experiencing hypotension during depletion phase, 5% albumin can be used as a replacement fluid.

Indications for RBC transfusion

RBC transfusion is a mainstay treatment for various complications of SCD (Table 2). The benefit of chronic transfusion therapy is well established for preventing primary and secondary strokes, and recurrent silent cerebral infarctions (SCIs). Transfusion can correct acute anemia in aplastic crises, acute splenic and hepatic sequestration; reduce HbS% in acute chest syndrome (ACS), acute stroke, acute multiorgan failure, and acute intrahepatic cholestasis. Transfusion also decreases SCD-associated complications in patients undergoing surgery and during pregnancy. Transfusion is not recommended for patients with steady-state anemia, uncomplicated painful vaso-occlusive crisis (VOC), priapism, leg ulcers and nonsurgically managed avascular necrosis. For patients with recurrent ACS, recurrent VOC, pulmonary hypertension, and uncomplicated pregnancy, the benefit of transfusion is not established; therefore, transfusion therapy should be considered on a case-by-case basis.

Table 2.

Indications for Transfusion in Patients with Sickle Cell Disease

Condition	Transfusion indication	Transfusion method
Acute ischemic stroke	Accepted	Exchange transfusion preferred
Primary stroke prevention	Accepted	Chronic simple or exchange transfusion
Secondary stroke prevention	Accepted	Chronic simple or exchange transfusion
Acute chest syndrome, severe*	Accepted	Exchange transfusion preferred
Acute chest syndrome, moderate	Accepted	Simple or exchange transfusion
Acute splenic sequestration	Accepted	Simple transfusion
Acute splenic sequestration, recurrent	Accepted	Chronic simple (before splenectomy)
Preoperative, requiring general anesthesia and surgery lasting > 1 hour)	Accepted	Simple or exchange transfusion**
Transient aplastic crisis	Accepted	Simple transfusion
Acute multi-organ failure	Accepted	Simple or exchange transfusion
Acute hepatic sequestration	Accepted	Simple or exchange transfusion
Acute intrahepatic cholestasis	Accepted	Simple or exchange transfusion
Pregnancy, complicated***	Accepted	Prophylactic simple or exchange transfusion
Acute chest syndrome, recurrent	Controversial	Chronic simple or exchange transfusion
Vaso-occlusive crisis, recurrent	Controversial	Chronic simple or exchange transfusion
Pulmonary hypertension	Controversial	Chronic simple or exchange transfusion
Vaso-occlusive crisis, uncomplicated	Not indicated	NA
Priapism	Not indicated	NA
Leg ulcers	Not indicated	NA
Avascular necrosis, nonsurgically managed	Not indicated	NA
Pregnancy, uncomplicated	Unknown benefit	NA

^*Severe acute chest syndrome, usually considered with rapidly declining hemoglobin levels, severe hypoxia (SpO2 ≤ 94% or considerably below patient baseline), and/or requiring invasive respiratory support

^**Exchange transfusion when pre-transfusion hemoglobin > 9-10 gm/dL that precludes administration of simple transfusion, or for high-risk surgery (e.g. neurosurgery)

^***Pregnant women with a history of severe or frequent sickle cell disease-related complications before the current pregnancy, onset of sickle cell disease-related complications during current pregnancy, or high-risk pregnancy

Neurologic complications

Approximately 11% of individuals with SCD are estimated to experience overt strokes by age 20 years.¹⁰ Children with SCD who are at risk for stroke can be identified by abnormally high blood flow velocity on transcranial doppler ultrasonography (TCD).¹¹ In the Stroke Prevention in Sickle Cell Anemia (STOP 1) trial, chronic transfusion to maintain HbS% < 30% reduced the incidence of initial strokes in children with abnormal TCD findings by 92%.¹² The STOP 2 trial supported the use of chronic transfusion indefinitely, as discontinuation after 30 months resulted in an increased rate of abnormal TCD conversion and stroke.¹³ Most recently, the TCD with Transfusions Changing to Hydroxyurea (TWiTCH) trial demonstrated that hydroxyurea is not inferior to chronic transfusion for primary stoke prevention by lowering TCD velocities in children with abnormal TCDs.¹⁴ Given the burden and potential side effects of chronic transfusion therapy, the ability to prevent stroke with hydroxyurea is a major advance.

Chronic transfusion is the standard of care for secondary stroke prevention. Up to 90% of patients with a stroke may experience a recurrence without therapeutic intervention, and the risk of secondary stroke decreases to ~20% with chronic transfusion therapy that maintains a HbS% < 30%.^15–17 Indefinite transfusion therapy is recommended as the Stroke With Transfusions Changing to Hydroxyurea (SWiTCH) trial was closed due to statistical futility on the composite endpoint of iron overload resolution and stroke prevention. At time of study closure, 10% of patients treated with hydroxyurea and phlebotomy had experienced a second stroke, as compared with no patients receiving chronic transfusion therapy.¹⁸

SCIs are more common than overt stroke, and 25% of children with SCD experience SCIs.^19,20 SCIs are associated with an increased risk of new or enlarged SCIs, overt stroke, low IQ, and poor academic performance.^20–22 The Silent Cerebral Infarct Multi-Center Clinical Trial (SIT) demonstrated that chronic transfusion therapy decreased the incidence of recurrent SCIs by 58% (14% in the observation group compared to 6% in the transfusion group).²³ Given the high incidence of SCIs in children with SCD, the resources needed to provide chronic transfusion therapy to this entire group would be considerable, and thus chronic transfusion therapy is usually considered for those at highest risk (i.e. magnetic resonance angiography-defined vasculopathy, poor school performance). The efficacy of hydroxyurea for preventing recurrent SCIs is yet to be determined.

Acute chest syndrome

ACS is the leading cause of death and second most common cause of hospitalization for patients with SCD.²⁴ The clinical presentation of ACS varies from fever, cough, chest pain and mild dyspnea to severe hypoxia requiring ventilation, and even death.²⁵ RBC transfusion is often used to treat moderate and severe ACS, in addition to antimicrobial and respiratory support.^26–28 No established clinical or laboratory criteria are available to define the severity of ACS or to identify patients with poor prognosis. Severe ACS is usually considered in patients with rapidly declining Hb levels, significant hypoxia (SpO2 ≤ 94% or considerably below patient baseline), and/or requiring invasive respiratory support.² RBCX is recommended over simple transfusion for severe ACS management.² But if a significant delay is expected while mobilizing blood and apheresis personnel, simple transfusion should be administered in the interim to patients with Hb levels < 9 g/dL. There is insufficient evidence to support RBCX over simple transfusion for moderate ACS, so either is appropriate.² However, RBCX should be considered for patients with rapidly progressive ACS, those who do not respond to simple transfusion, or those with pre-transfusion Hb levels ≥ 9-10 g/dL that preclude simple transfusion. Recurrent ACS occurs in 44% of patients with SCD,²⁹ for which the primary treatment is hydroxyurea.³ Chronic transfusion reduces the incidence of ACS and subsequent hospitalizations, so should be considered if hydroxyurea is ineffective or contraindicated.^22–23,30

Preoperative transfusion support

Patients with SCD may experience ACS, painful VOC, and other post-operative complications, particularly for patients with low preoperative Hb levels (< 9 g/dL), severe SCD genotypes (i.e., HbSS or HbSβ°thal) or phenotypes, and those receiving high-risk surgery (cardiac and neurosurgery). The Transfusion Alternatives Preoperatively in Sickle Cell Disease (TAPS) trial showed that preoperative transfusion to increase Hb levels to 10 g/dL reduces the risk of overall perioperative complications, particularly ACS, in patients undergoing low-to-moderate-risk surgeries.³¹ Patients receiving aggressive preoperative transfusion to reach a Hb level of 10 g/dL and HbS% < 30% experience a similar number of perioperative complications as those having conservative preoperative transfusion to only increase Hb levels to 10 g/dL (with no specific target HbS%).³² Furthermore, preoperative transfusion does not improve hospital stay duration, readmission rate, or mortality of patients with SCD.³¹ Hence, preoperative transfusions to increase Hb levels to 10 g/dL (with no specific target HbS%) are recommended for patients with SCD requiring surgery with general anesthesia and lasting greater than one hour.² RBCX should be considered for patients with high baseline Hb levels (>9-10 g/dL) or undergoing high-risk surgery.

Complications of RBC transfusion

Alloimmunization

Non-ABO antigen RBC alloimmunization remains a significant complication of transfusion. It may also lead to difficulty identifying compatible units, transfusion delays for acute complications of SCD, and impact the ability to provide transfusion support for hematopoietic stem cell transplantation and surgeries with a high risk of blood loss. Alloimmunization can lead to life-threatening reactions that reduce the survival of patients with SCD.^33,34

Alloimmunization increases the risk of hemolytic transfusion reactions, both acute and delayed. Delayed hemolytic transfusion reactions (DHTRs) occur in 4-16% of transfused patients with SCD. DHTRs are classically caused by evanescent alloantibodies, which are undetectable at the time of transfusion, but appear upon re-exposure to the offending RBC antigen(s).³⁵ Patients usually present within 7-10 days (and up to 28 days) from the last transfusion with signs and symptoms of anemia and hemolysis, including fatigue, jaundice, hemoglobinuria, fever, and/or pain. Often the direct antiglobulin test (DAT) will be positive, and the evanescent alloantibody(ies) will be identified, but not always. When a new antibody is identified, the patient’s Hb, reticulocyte count and HbS% should be obtained to evaluate for laboratory evidence suggestive of hemolysis. High clinical suspicion and monitoring for DHTRs is warranted, as they may be more common in patients with SCD than appreciated and sometimes clinically subtle.³⁶ Rarely, hyperhemolysis syndrome may occur in which both transfused and endogenous RBCs are hemolyzed, leading to severe and potentially lethal anemia.³⁷ The DAT is often negative in hyperhemolysis syndrome, and no antibody specificity identified. Patients should be managed with supportive treatment (oxygen and erythropoietin) and immunosuppressive therapy. Intravenous immunoglobulin and/or steroids should be initiated promptly, while eculizumab may be considered for those unresponsive to first-line agents.² Transfusion is contraindicated in hyperhemolysis syndrome unless the anemia is life-threatening. If transfusion is warranted, extended-matched (C/c, E/e, K, Jk^a/Jk^b, Fy^a/Fy^b, S/s) RBCs should be considered.² For patients requiring transfusion in the following weeks to months (i.e. chronic transfusion for stroke prophylaxis), rituximab may be considered, particularly if no specific antibody is identified.

The prevalence and rate of alloimmunization are higher in patients with SCD than in patients with other diseases.³⁸ While alloimmunization prevalence can approach 50% in chronically transfused individuals, the rate of alloimmunization is approximately 0.3-0.5 alloantibodies per 100 units transfused with Rh and K matched transfusion regardless of chronic versus episodic transfusion.^23,39–42 Alloimmunization risk is associated with transfusion burden, RBC antigen differences between patients of African descent and Caucasian blood donors, patient immune responses, and inflammatory status when receiving transfusion.^39,43,44

Alloantibodies commonly identified in patients with SCD include antibodies against Rh (D, C/c, E/e), K, Duffy (Fy^a, Fy^b), Kidd (Jk^a, Jk^b), and S, in part due to the discrepant antigen expression in patients of primarily African ethnicity compared to the majority of donors with European background in the US.⁴⁵ To avoid alloimmunization, prophylactic antigen matching with Rh (C, E or C/c, E/e) and K-matched RBCs is recommended.^2–4 Extended RBC antigen matching to also include Duffy, Kidd, and S antigens can provide further protection from alloimmunization,⁴⁶ but identifying adequate matched units would be challenging.^47,48

Prophylactic antigen matching requires blood group antigen information. An extended antigen profile (i.e., C/c, E/e, K, Jka/Jkb, Fya/Fyb, M/N, and S/s at minimum) is preferred and should be obtained at the earliest opportunity, ideally before the first transfusion.² Knowledge of the antigen profile also assists antibody identification and additional antigen matching in alloimmunized patients. A RBC antigen genotype is preferred over a serologic phenotype since it provides additional information and increased accuracy for Fy^b and Rh antigen matching.² Patients with mutations in the transcription factor GATA-1 binding site of the ACKR1/DARC gene encoding Fy antigens do not express Fy^b on RBCs.⁴⁹ These patients are not at risk of forming anti-Fy^b since Fy^b is expressed in other tissues and can thereby safely receive Fy^b+ RBCs. RBC genotyping can also identify variant Rh antigens that are common in African-descended individuals and contribute to Rh alloimmunization despite serologic Rh antigen matching.^40,50,51 Variant Rh antigens, caused by extensive single nucleotide polymorphisms and genetic rearrangements of RH genes, differ from conventional Rh antigens with loss of common Rh epitopes and/or expression of new Rh epitopes. Patients may become Rh-alloimmunized when exposed to Rh antigens that are absent on their own RBCs. Over 50 Rh variant antigens and more than 500 RH variant alleles have been described, with the number of newly identified alleles continuing to grow.^52,53 While serologic Rh antigen matching decreases Rh alloimmunization, RH genotype matching between patients and blood donors is likely needed to eliminate Rh alloantibody formation.⁵⁴

Iron overload

Patients receiving chronic simple transfusion and those who undergo RBCX but require an increase in their end procedure Hct will accumulate iron stores. Free iron accumulates in the liver, heart, and endocrine system, resulting in organ dysfunction. Since SCD-associated inflammation promotes iron retention in the reticuloendothelial system, the liver is the most commonly affected organ.⁵⁵ Cardiac iron overload is uncommon for patients with SCD and typically occurs in a subgroup of patients with exceedingly high iron levels over a prolonged period of time.^56–58

Iron burden can be evaluated by serum ferritin levels, liver and cardiac magnetic resonance imaging (MRI). Serum ferritin levels can be readily and frequently obtained and are a convenient option for monitoring iron burden over time.⁵⁹ However, serum ferritin levels increase with SCD-associated inflammation and may not be a reliable marker for total body iron burden. Liver iron content (LIC) is a good indicator of total body iron burden and was traditionally assessed by liver biopsy.^60,61 LIC measured by MRI (R2, T2* or R2*) correlates well with liver biopsy results and has become the primary screening tool for iron overload.^62,63 LIC should be screened by MRI every 1–2 years for patients on chronic transfusion therapy and/or with a serum iron level ≥ 1000 ng/mL.² Cardiac iron overload is uncommon for patients with SCD, and thus routine cardiac iron content screening by T2* MRI is not required unless a history of poor iron control, cardiac dysfunction or other end organ damage is identified.²

Iron chelation therapy (i.e., deferoxamine and deferasirox) can maintain a negative or neutral iron balance to prevent hemosiderosis. The ferritin trend and liver R2* measurements of iron accumulation are used to titrate iron chelation regimens. For patients with severe iron overload, combination iron chelation can be used, or iron chelation therapy may be combined with RBCX.

Allogeneic hematopoietic stem cell transplantation and transfusion support

HSCT as a curative therapy for SCD was first reported in 1984, in which a patient with SCD received HSCT for leukemia and both disorders were cured.⁶⁴ Since then, various conditioning regimens, such as myeloablative, reduced-intensity, and nonmyeloablative, and different hematopoietic stem cell (HSC) sources, including human leukocyte antigen (HLA)-matched sibling donors, unrelated HLA-matched donors, umbilical cord blood, and haploidentical-related donors, have been examined in multiple clinical trials. Matched sibling donor HSCT, with a high overall survival rate (>90%), low incidence of graft rejection (< 3%), and chronic graft-versus-host disease (GVHD; <15%), has evolved into a curative treatment option for pediatric patients with SCD with fully matched sibling donors.^65,66 Pediatric patients with less SCD-associated end organ damage than adult patients are more likely to tolerate intense conditioning regimens. Transfusion support for patients with SCD undergoing HSCT requires specific considerations because of the unique clinical characteristics of SCD and prevalence of RBC and HLA alloimmunization in this patient population.

Transfusion support

Reducing the HbS% to < 30% by simple transfusion or RBCX for months or just prior to the start of HSCT conditioning has become common practice. In a murine model of SCD, SCD-associated inflammation and hypoxia lead to vascular tortuosity and sinusoidal stasis in bone marrow, which are reversed by chronic transfusion therapy.⁶⁷ This observation suggests that pre-HSCT blood transfusions can be beneficial by altering the marrow environment to improve engraftment. This benefit must be weighed against the risk of RBC and HLA alloimmunization and other transfusion reactions. Therefore, the number and duration of pre-HSCT transfusions should be individualized.

During the peri-HSCT period, patients with SCD are at increased risk of neurovascular complications, including intracranial hemorrhage, seizure, and posterior reversible encephalopathy syndrome (PRES).⁶⁸ The transfusion threshold for platelets is < 50 × 10³/μL to prevent intracranial hemorrhage, and maintenance of platelet counts above this transfusion threshold is particularly important for patients with a history of stroke. Hb should be maintained between 9 and 11 g/dL to prevent both anemia and hyperviscosity. The risk of PRES and seizures can be further decreased by close monitoring and control of electrolytes and blood pressure.

Blood products selected for patients with SCD undergoing HSCT should meet certain criteria. All blood products should be irradiated and leukoreduced. Irradiation prevents transfusion-associated graft-versus-host-disease. Leukoreduction decreases febrile non-hemolytic transfusion reactions, HLA sensitization, and transmission of infectious diseases carried in white blood cells (e.g. cytomegalovirus). Patients with SCD should receive HbS negative RBC units ideally stored for < 14 - 21 days to maximize in vivo circulatory half-life. Units should be prophylactically matched for Rh (C, E or C/c, E/e) and K antigens of the patients, and lacking antigens for which the patient has previously formed alloantibodies against. Some transfusion services extend the matching to Fy^a/Fy^b, Jk^a/Jk^b and S antigens, particularly if a patient is heavily alloimmunized.

ABO incompatibility as a special consideration

Major ABO incompatibility refers to the presence of natural anti-A and/or anti-B antibodies in recipients against donor A and/or B blood group antigens. Major ABO incompatibility can cause 1) acute intravascular hemolysis of donor RBCs present in HSC products upon infusion; and 2) chronic destruction of donor erythrocyte precursors, leading to delayed RBC engraftment (DRE) and even pure RBC aplasia (PRCA).⁷⁰ Acute hemolysis is rare when HSC products are collected by apheresis since these products typically contain 2 to 5% RBCs and a total RBC volume of < 20 mL.⁸ In contrast, HSC products collected by bone marrow harvest have an increased risk due to a high RBC content of 25 to 35%. Acute hemolytic transfusion reaction can be minimized by RBC reduction during HSC product processing by automatic cell processors, hydroxyethyl starch sedimentation, or Ficoll-Paque density gradient separation.⁷⁰ General practice is to not exceed 0.3 to 0.4 mL/kg or a total of 10 to 30 mL of RBCs for pediatric patients receiving an HSC product with ABO incompatibility.⁷¹

Recipient ABO antibodies can persist for many months causing DRE and PRCA, leading to prolonged transfusion-dependence after HSCT. The primary source of ABO antibodies are recipient plasma cells.⁷² Patients undergoing nonmyeloablative and reduced intensity HSCT are at particular risk since their own plasma cells are more likely to persist. PRCA is treated with erythropoietin, rapid tapering of immunosuppressants, and supportive transfusions. While therapeutic plasma exchange may theoretically improve PRCA by removing incompatible ABO antibodies, it is not recommended because of its inability to remove antibody-producing recipient plasma cells.⁶⁹

Minor ABO incompatibility refers to the presence of natural anti-A and/or anti-B antibodies in donors against recipient A and/or B blood group antigens. Minor ABO incompatibility can cause transient and usually self-limited hemolysis of recipient RBCs during infusion, which can be avoided by plasma reduction of HSC products to less than 200 to 300 mL or 5 mL/kg.^70–71 In addition, clinically severe and possibly fatal hemolysis can occur 5 to 16 days after HSCT due to passenger lymphocyte syndrome, in which donor lymphocytes in grafts produce antibodies against recipient RBCs after sensitization to recipient antigens.⁷⁰ In severe cases, RBCX can be used to replace recipient RBCs with donor-compatible RBCs.

To support ABO incompatible HSCT, RBC and plasma-rich components compatible to both donors and recipients are selected.⁷⁰ Specifically, for ABO major incompatible HSCT (e.g., blood type O recipients transplanted with blood type A donor), RBCs compatible to both donors and recipients are transfused (blood type O in this case). For ABO minor incompatible HSCT (e.g., blood type A recipient transplanted with blood type O donor), plasma-rich components compatible to both donors and recipients are chosen (blood types A and AB in this case).

RBC and HLA alloimmunization as a special consideration

Patients with SCD often require more transfusion support in the peri-HSCT period compared to patients with other disorders, and their transfusion requirements vary considerably. In one study, a median of 7 RBC transfusions (range, 3–15) and 13.5 platelet transfusions (range, 4–48) per patient were administered before reaching transfusion independence.⁷³ The increased transfusion demands may be due to the high prophylactic transfusion thresholds and the high prevalence of RBC and/or HLA alloimmunization in this patient population.

RBC alloimmunization

RBC alloimmunization is common in patients with SCD undergoing HSCT, with approximately 31 to 36% of patients having a history of RBC alloimmunization.^74,75 RBC alloimmunization is strongly associated with increased RBC transfusion demands (7 units for RBC alloimmunized patients vs. 4 units for non-alloimmunized patients in the first 45 days after HSCT).⁷⁵ For alloimmunized recipients with antibodies against donor RBC antigens, recipient plasma cells may persist even after conditioning regimens, and continue to produce donor-specific RBC antibodies resulting in DRE or PRCA. Notably, one study found that patients who are alloimmunized but lack antibodies specific to their donor RBC antigens also required more transfusions (6.5 units for RBC alloimmunized patients vs. 4 units for RBC nonalloimmunized patients in the first 45 days after HSCT).^72,75 Since these patients were treated with nonmyeloablative conditioning regimens, their residual immune responses may have removed donor and/or transfused RBCs in an antibody-independent manner. For alloimmunized patients, HSC donor RBC antigen profiles should be considered when selecting from equivalent HLA matched donors, and extended antigen-matched RBCs may be indicated for select patients. Advance communication between the transplant and transfusion services can be helpful for both HSC donor selection and peri-HSCT transfusion support.

HLA alloimmunization

Patients with SCD managed with chronic transfusion therapy are at risk for HLA alloimmunization. Approximately 30% of patients undergoing HSCT have HLA class I antibodies.^74,76 Since HLA class I antigens are present on platelets and crossmatch-compatible or HLA-matched platelets are not routinely administered, HLA class I alloimmunization can lead to poor post-transfusion platelet increments or platelet refractoriness. Specifically, platelet refractoriness is defined as 1-hour post-transfusion corrected count increments < 5000 to 7000/μL {[platelet increment (μL) × body surface area (m²)] / [number of single platelet units × 3]}. In limited studies of pediatric patients with SCD undergoing HLA-matched HSCT, those with HLA class I antibodies required nearly twice as many platelet transfusions (2.5–19 units for HLA alloimmunized vs. 1–7.5 HLA nonalloimmunized in the first 45 days after HSCT).^74,76 High platelet transfusion burden is associated with adverse outcomes in the context of HSCT and various transfusion reactions in particular, pulmonary complications.^77,78 HLA antibody screening prior to HSCT for patients on chronic transfusion is recommended to identify patients at risk for platelet refractoriness, and estimate the potential need for crossmatch-compatible or HLA-matched platelets.

Complications of RBC transfusion

RBC Alloimmunization

Patients with SCD remain at risk for new RBC alloantibody formation during the peri-HSCT period, which is more likely with nonmyeloablative HSCT. Nonmyeloablative conditioning regimens often results in mixed chimerism of donor and recipient-derived RBCs and immune cells. Residual recipient immune cells can form antibodies against donor RBC antigens, causing hemolysis of donor RBCs and DRE or PRCA. Donor-derived immune cells can become immunized to recipient RBC antigens, resulting in hemolysis. In a study of 61 patients undergoing non-myeloablative HLA-matched or haploidentical HSCT, 6 patients formed 11 new RBC alloantibodies and 2 autoantibodies.⁷⁵ Three of the alloantibodies were incompatible with either donor or recipient antigens. Most new antibodies were detected within 30 days of HSCT; the median time to first detection was 17 days. Two patients experienced acute hemolysis and 3 patients had DRE with severe anemia. Pre-transplant extended RBC antigen phenotyping or genotyping of patients and donors can identify potential mismatches, and prophylactic extended RBC matching (beyond Rh and Kell) can be considered but may not be feasible due to donor availability.

Autologous hematopoietic stem cell transplantation and transfusion support

Allogeneic HSCT from HLA-matched sibling donors is an established curative therapy for SCD but many patients lack an unaffected HLA-matched sibling donor. Identifying unrelated HLA-matched donors is also challenging due to the low number of African-descended individuals in bone marrow donor registries.⁷⁹ Clinical trials using alternative donor options including umbilical cord blood or haploidentical-related donors have had encouraging results but are also associated with a high incidence of graft rejection and GVHD.^80–83 Autologous HSC gene therapy provides a novel approach with the potential to be available to all patients and to avoid many complications or toxicities associated with allogeneic HSCT. Many gene editing strategies are currently under investigation, including correction of SCD-causing mutations, overexpression of healthy or modified β^A-like globin gene, and induction of fetal globin expression to outcompete sickle p globin during Hb assembly.^84,85 Multiple phase I/II clinical trials are ongoing and the first patient who received a modified β^A-globin gene encoding an anti-sickling variant (β^A87Thr:Gln [β^A-T87Q]) in 2014 achieved complete clinical remission.⁸⁶

A large number of HSCs, approximately 10–15 ×10⁶ /kg, are required for most gene therapies for SCD.⁸⁴ HSCs are preferentially collected from peripheral blood via leukocytapheresis since cells harvested by bone marrow aspiration require general anesthesia, require at least 2 to 4 procedures to obtain an adequate HSC number, and are associated with an increased risk of complications in patients with SCD. Bone marrow mobilization is needed for adequate HSC collection by leukocytapheresis. Granulocyte colony-stimulating factor is contraindicated in patients with SCD due to severe adverse events, such as VOC, ACS, massive splenomegaly, and even death.^87–89 Plerixafor, a CXCR4 antagonist, can enable rapid and efficient HSC mobilization with minimal toxicity in patients with SCD (Table 3).^90–93 Plerixafor reversibly blocks interaction between CXCR4 on HSCs and stromal-derived factor-1α (SDF-1α) in bone marrow niche.⁹⁴ This induces rapid release of HSCs from bone marrow into the peripheral circulation, with peripheral blood CD34⁺ cell counts peaking as early as 3 to 6 hours after administration. Plerixafor-mobilized HSCs contain a high percentage of long-term repopulating HSCs (CD34^high/CD90⁺/CD45RA⁻), which are amenable to genetic modification.^91,93,95 Adverse events associated with plerixafor mobilization are mild and limited to VOC that resolves with medical therapy.^90–93 Approximately 3 - 6 × 10⁶ CD34⁺ cells/kg can be obtained in a single leukocytapheresis collection; therefore, 2 to 3 collections with plerixafor mobilization are typically required to obtain sufficient HSCs for gene therapy.^90,91,93 Additional agents are currently being investigated,⁸⁴ including GROβ, a CXCR2 agonist that efficiently mobilizes highly engraftable long-term repopulating HSCs alone or synergistically with plerixafor.^96,97

Table 3.

Plerixafor HSC Mobilization in Patients with Sickle Cell Disease

Study	Plerixafor dose (μg/kg)	No. of Patient	Patient Age	Male %	No. of patient on HU	No. of patient with transfusions before plerixafor	Pre-plerixafor HbS (%)	Peak CD34 count (/μL)	Product Hct (%)	CD34+ cell yield (x106/kg)	Adverse Events
Uchida et al., 2020	240	15	29* (20-50)	47	11	15	27 (15.1-37.7)	52 (9-183)	4.5 (2.7-7.5)	6.3 (2.2-12.0)	5†
Lagresle-Peyrou et al., 2018	240	3	20 (19-21)	100	1	3	<30	>80	5.8 (4.8-8.2)	4.6 (4.5-5.8)	0
Esrick et al., 2018	180	3	26 (19-30)	100	0	3	9.9 (5.5-13.3)	36 (31-65)	5.6 (3.7-17)	0.616 (0.069-1.2)	0
	240	3	25 (25-38)	100	0	3	9.2 (6.9-21.4)	156 (27-290)	11.7 (10.5-16.4)	16.38 (2.94-24.53)	0
Boulad et al., 2018	80	6	30.5 (21-34)	66.7	4	0	84.8 (65.6-89.9)	27.5 (7-132)**	NA	NA	1‡
	160	3	32 (25-37)	33.3	2	1	74.9 (41.9-89.9)	43 (7-251)**	NA	NA	0
	240	6	34.5 (23-46)	83.3	4	0	79.6 (71.1-93.1)	30.5 (10-95)**	NA	NA	1‡

^*Median (range)

^**At 12 hours post plerixafor administration

^†Four severe vaso-occlusive crises and one delayed hemolytic transfusion reaction

^‡Severe vaso-occlusive crisis

HSC, hematopoietic stem cell; HU, hydroxyurea; No, number

HSC yields from leukocytapheresis collections vary considerably among patients, which can be attributed to multiple factors including hydroxyurea treatment and the unique characteristics of SCD-derived HSCs. Hydroxyurea suppresses bone marrow hematopoiesis, reducing CD34+ cell numbers in the bone marrow and peripheral blood at steady state and upon mobilization.^98,99 Hydroxyurea is typically discontinued in the months prior to HSC collection and replaced by simple RBC transfusion or RBCX.^91,93 The optimal Hb and HbS% levels and duration of transfusions before marrow mobilization are unknown, but Hb levels of 10 g/dL and/or HbS% < 30% are parameters used by several recent and ongoing clinical trials.^90,91,93

Despite successful bone marrow mobilization, patients with SCD have lower HSC collection efficiency and yield than do other patients with standard leukocytapheresis.^84,100 SCD-associated inflammation is thought to alter the physical properties of HSCs, resulting in adhesion of the HSCs to other cells. Consequently, SCD-derived HSCs may migrate to a deeper layer during centrifugation and thereby fail to be collected with standard leukocytapheresis techniques. Deeper interface collection improves HSC collection efficiency and yield,^90,93 but increases RBC contamination of leukocytapheresis products that may interfere with subsequent HSC purification and recovery.

Conclusion

RBC and platelet transfusions are critical for treating SCD-associated complications and supporting patients during allogeneic and autologous HSCT. A major risk of transfusion therapy is RBC and HLA alloimmunization, for which contributing donor and recipient factors warrant consideration for HSCT donor selection and peri-HSCT transfusion support. HSCT is currently the only cure for SCD, but promising progress in novel gene therapy approaches may broaden accessibility to curative therapies in the near future.

KEY POINTS.

• Red blood cell (RBC) transfusion is a mainstay treatment for acute and chronic complications of sickle cell disease (SCD). Prophylactic transfusion with Rh (C, E or C/c, E/e) and K-matched RBCs is recommended to prevent RBC alloimmunization. Iron status should be closely monitored in patients receiving chronic transfusion.

• Matched sibling donor hematopoietic stem cell transplantation (HSCT) is a treatment option for patients with SCD. Platelets should be maintained ≥ 50 × 10³/μL and hemoglobin (Hb) 9-11 g/dL in the peritransplant period. RBC and HLA alloimmunization are strongly associated with increased transfusion demands.

• Autologous HSCT with gene corrected hematopoietic stem cells (HSCs) for SCD requires a large number of cells. Plerixafor enables rapid and efficient HSC mobilization with minimal toxicity. Hydroxyurea cessation and optimized HSC leukocytapheresis collection increase the HSC yield.

SYNOPSIS.

Red blood cell (RBC) transfusion is critical in managing acute and chronic complications of sickle cell disease (SCD). Alloimmunization and iron overload remain significant complications of transfusion therapy, and are minimized with prophylactic Rh and K antigen RBC matching and iron chelation. Matched sibling donor hematopoietic stem cell transplantation (HSCT) is a curative therapeutic option. Autologous hematopoietic stem cell (HSC)-based gene therapy has recently shown great promise, for which obtaining sufficient HSCs is essential for success. This review is focused on RBC transfusion indications and complications, transfusion support during HSCT, and HSC mobilization and collection for autologous HSCT with gene therapy.