INTRODUCTION — Autologous and allogeneic hematopoietic cell transplantation (HCT) are associated with neutropenia, anemia, and thrombocytopenia in the peri-transplant period. The degree of myelosuppression and the time to hematopoietic recovery differ with multiple factors including the preparative regimen and graft source. Blood product transfusions and hematopoietic growth factors are essential components of transplant care.
The term "hematopoietic cell transplantation" will be used throughout this review as a general term to cover transplantation of progenitor cells from any source (eg, bone marrow, peripheral blood, umbilical cord blood). Otherwise, the source of such cells will be specified (eg, autologous peripheral blood progenitor cell transplantation).
Hematopoietic support after HCT will be discussed here. Other supportive care issues surrounding HCT are presented separately, as is quality of life following HCT and acute and chronic graft-versus-host disease (See "Early complications of hematopoietic cell transplantation" and "Treatment of chronic graft-versus-host disease" and "Prevention of graft-versus-host disease".)
The diagnosis and treatment of pulmonary, renal, and infectious complications following HCT are also presented separately. (See "Pulmonary complications after allogeneic hematopoietic cell transplantation: Causes" and "Pulmonary complications after autologous hematopoietic cell transplantation" and "Kidney disease following hematopoietic cell transplantation" and "Overview of infections following hematopoietic cell transplantation".)
BLOOD PRODUCT SUPPORT — Blood product support is usually required before, during, and following HCT [1]. In addition, blood product support, with resulting iron overload, is commonly seen in those patients undergoing HCT for a hematologic disorder. Such iron overload (eg, serum ferritin >1000 ng/mL) may adversely affect overall survival post-HCT, increasing the likelihood of acute graft-versus-host disease, as well as the incidence of blood stream infections and sinusoidal obstruction syndrome of the liver [2-4].
Virtually all patients undergoing transplantation require blood product support in the form of red blood cell and platelet transfusions until the transplanted marrow cells engraft sufficiently to support hematopoiesis. This generally requires 14 to 21 days or more with bone marrow but is accelerated with peripheral blood progenitor cells with engraftment typically requiring 10 to 14 days.
All cytomegalovirus (CMV)-negative patients who receive hematopoietic cells from a CMV-negative donor should receive CMV seronegative or leukoreduced blood products whenever possible. Leukoreduced blood also reduces the risk of febrile nonhemolytic transfusion reactions (FNHTRs). In addition, it is highly recommended to use irradiated blood products to avoid the risk of transfusion-associated graft-versus-host disease. (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'Specialized modifications and products' and "Transfusion-associated graft-versus-host disease".)
Transfusion needs are generally lower in patients undergoing nonmyeloablative preparative regimens. As an example, in a comparative retrospective analysis, platelet transfusions were required in 23 versus 100 percent, and red cell transfusions in 63 versus 96 percent of those undergoing nonmyeloablative or myeloablative transplants, respectively [5].
Red blood cell transfusion — There is no consensus regarding a hemoglobin (Hb) threshold that should trigger transfusion of red blood cells (RBC). Practices vary, but many centers use Hb 7 to 8 g/dL to trigger RBC transfusion. Such a restrictive transfusion strategy is safe, effective, and may reduce transfusion-related adverse events. (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'Whole blood'.)
The transfusion threshold may be influenced by the clinical condition of the patient, number of days since the transplant, and evidence of engraftment of other cell types, including reticulocytes. Patients with graft-versus-host disease (GVHD) or those being treated with immunosuppressive drugs such as cyclosporine or tacrolimus may have continued blood product requirements from bleeding and/or microangiopathic hemolysis. Typically, type O blood is utilized if there is a major ABO mismatch between donor and recipient until engraftment of donor RBCs is confirmed. Cases of delayed hemolysis and/or pure red cell aplasia have been described in patients with pre-existing alloantibodies against Rh or ABO antigens, despite the presence of 100 percent donor chimerism of the circulating lymphocytes. This is discussed in more detail separately. (See "Donor selection for hematopoietic cell transplantation", section on 'ABO and Rh status'.)
In a multicenter trial that included 300 patients undergoing autologous or allogeneic HCT for a hematologic malignancy, there was no difference in outcomes between patients who were randomly assigned to a restrictive transfusion threshold (Hb 7g/dL) versus a liberal transfusion strategy (Hb 9g/dL) [6]. Compared with the liberal threshold, patients treated with a restrictive strategy received fewer transfusions (2.7 versus 5.0, respectively), but there was no difference in health-related quality of life or transplant outcomes, including transplant-related mortality at day 100, length of hospital stay, intensive care unit admissions, hospital re-admissions, acute GVHD, grade ≥3 infections, or bleeding.
Platelet transfusion — Patients who are thrombocytopenic with bleeding require platelet transfusions. The role of prophylactic platelet transfusions remains controversial. We generally transfuse for platelet counts less than 10,000/microL or for higher values if clinical bleeding is present (table 1) [7]. This is discussed in more detail separately. (See "Platelet transfusion: Indications, ordering, and associated risks".)
Many transplant centers have relied upon threshold values of platelet numbers to determine the timing of platelet transfusions. This has been questioned because of the risk of accelerating alloimmunization against future platelet transfusions, the relatively short half-life of transfused platelets, and expense. Several multicenter studies of platelet recovery have addressed this issue in the setting of HCT:
●In a randomized trial, 400 patients with acute myeloid leukemia (AML; patients with acute promyelocytic leukemia were excluded) and patients undergoing autologous HCT for hematologic malignancies were assigned to receive platelet transfusions when morning platelet counts were ≤10,000/microL or only for active bleeding [8]. Patients transfused only for active bleeding received fewer platelet transfusions during the 14-day period after induction or consolidation chemotherapy (1.63 versus 2.44 per patient). Patients undergoing HCT experienced more bleeding episodes when transfused only for active bleeding, but most of these were minor.
●In another randomized trial, 600 patients with hematologic malignancies receiving chemotherapy, autologous, or allogeneic HCT were assigned to receive platelet transfusion for a platelet count ≤10,000/microL or only for active bleeding (the Trial of Prophylactic Platelets [TOPPS]) [9-11]. Compared with those who received prophylactic transfusions, patients transfused only for active bleeding received fewer platelet transfusions during the 30-day period after randomization, but had a higher incidence of major bleeding (50 versus 43 percent) and a shorter time to first bleed (1.2 versus 1.7 days) [12]. There were no differences in the duration of hospitalization, and no deaths due to bleeding. In a predefined subgroup analysis, patients undergoing autologous HCT had similar rates of major bleeding whether they were transfused for a platelet count ≤10,000/microL or only for active bleeding (45 and 47 percent).
●In a multicenter observational study which included 789 patients transplanted in 1995 at 18 centers in the United States and Canada, 11 percent of all patients had a significant hemorrhagic event during the first 60 days following the transplant, contributing to death in 2 percent [13]. Most hemorrhagic events (66 percent) occurred when the morning platelet count was >20,000/microL. Variables associated with accelerated platelet recovery included a higher CD34+ cell count of the infused hematopoietic stem cell product, a higher platelet count at the start of myeloablative therapy, and a graft from an HLA-identical sibling donor. Variables associated with delayed platelet recovery included prior radiation therapy, post-transplant fever, hepatic veno-occlusive disease, the presence of graft-versus-host disease, CMV infection, cell dose, and the use of post-transplant growth factors [14].
These results suggest that prophylactic transfusion for patients with platelets ≤10,000/microL is reasonable. Further enhancing platelet recovery is not likely to have a significant impact on 60-day mortality but could significantly decrease health care costs and potentially improve patient quality of life. Similar conclusions have been reached in studies of patients who became thrombocytopenic after allogeneic HCT or following chemotherapy for acute leukemia; reducing the platelet transfusion threshold from 20,000 or 30,000/microL to 10,000/microL led to decreased use of platelet transfusions without any significant effect on morbidity and only a small adverse effect [15,16] or no effect [17] on bleeding. The choice of stem cell source also impacts platelet recovery with a more rapid recovery following peripheral blood progenitor cells than with bone marrow transplants [18]. (See "Use of blood products in the critically ill" and "Platelet transfusion: Indications, ordering, and associated risks", section on 'Leukemia, chemotherapy, and HSCT'.)
Granulocyte transfusions — Granulocyte transfusions have generally not been used because of the lack of efficacy in clinical trials and the difficulty of obtaining sufficient numbers of cells due to the short half-life of neutrophils. Currently, there are very few if any clinical indications for granulocyte transfusions outside of a clinical trial setting. (See "Granulocyte transfusions".)
GROWTH FACTOR SUPPORT — Cloned hematopoietic growth factors that have been explored in the transplant setting include granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony stimulating factor (GM-CSF), erythropoietin (EPO), and thrombopoietin (TPO). (See "Introduction to recombinant hematopoietic growth factors".)
G-CSF and GM-CSF — The use of cloned recombinant human hematopoietic growth factors such as G-CSF and GM-CSF has resulted in clear clinical benefits in the autologous transplant setting, particularly following autologous transplantation with bone marrow or peripheral blood progenitor cells (PBPCs) and for the mobilization of PBPCs.
The American Society of Clinical Oncology (ASCO) has issued guidelines for the administration of hematopoietic colony-stimulating factors after bone marrow and stem cell transplantation [19]. Available data supported the use of colony-stimulating factors for mobilization of PBPCs for transplantation, as well as following autologous transplantation using either bone marrow or PBPC.
In the post-transplant setting the usual dose is 5 mcg/kg per day for G-CSF (filgrastim), which we prefer, or 250 mcg/m2 per day for GM-CSF (sargramostim) [19]. Therapy is usually begun one to five days after transplantation and continued until the absolute neutrophil count reaches 10,000/microL; however, a shorter duration until clinically adequate neutrophil recovery is achieved is a reasonable alternative. Subcutaneous injection is preferred to intravenous injection when either agent is used due to improved pharmacokinetics and the suggestion of improved efficacy in one small randomized trial [20].
Given the limited general utility of colony-stimulating factors following allogeneic transplantation, and the fact that their use may have been deleterious in some allogeneic settings, there appears to be little reason to routinely treat allogeneic HCT patients with these growth factors as post-treatment prophylaxis [21-23]. Exceptions include those patients with delayed neutrophil engraftment, a reduction in white blood cells due to infection or drug treatment, or those being treated with umbilical cord blood as a stem cell source. (See 'Allogeneic HCT' below.)
Autologous HCT — Randomized phase III trials with both GM-CSF and G-CSF have shown a reduced number of days required for neutrophil engraftment following transplantation with bone marrow as a stem cell source, associated in some studies with a reduction in the days of antibiotic therapy and length of hospitalization [24-28]. These findings led to the US Food and Drug Administration approval of both drugs for this indication. However, the positive effects in these trials were limited to neutrophil engraftment and neutropenia-related complications. In no study was red blood cell or platelet engraftment enhanced, nor was a survival advantage demonstrable.
A study of once daily versus split (twice daily) lenograstim was not associated with superior clinical efficacy [29]. Similarly, three randomized phase III trials of single dose pegfilgrastim (6 mg) versus daily filgrastim (5 mcg/kg per day) in patients undergoing autologous peripheral blood stem cell transplantation revealed no difference in time to neutrophil engraftment or any clinical sequelae [30-32].
Autologous PBPC transplantation — The observation that hematopoietic growth factor administration produces a time-limited enhancement in hematopoietic stem cell mobilization has had a major impact on autologous transplantation. Upon re-infusion of these "mobilized" peripheral blood progenitor cells, hematopoietic engraftment of all lineages is accelerated significantly. This is especially true for platelet engraftment [18,33,34]. (See "Sources of hematopoietic stem cells".)
A multinational study compared the use of filgrastim (G-CSF)-mobilized autologous PBPCs versus autologous bone marrow in 58 patients with relapsed Hodgkin lymphoma and non-Hodgkin lymphoma treated with carmustine, etoposide, cytarabine, and melphalan [18]. Early post-transplant morbidity and mortality and overall survival (median follow-up, 311 days) were similar in both groups, but use of filgrastim-mobilized PBPCs was associated with the following significant benefits:
●A shorter time to platelet recovery above 20,000/microL (16 versus 23 days)
●A shorter time to neutrophil recovery (11 versus 14 days)
●A shorter time in hospital (17 versus 23 days)
●A cost saving of 23 percent (13,521 US dollars) due to lower autograft collection costs and shorter hospitalizations with less supportive care [35]
There was no notable toxicity attributable to filgrastim. Similar results were noted in another report in which patients who received G-CSF-primed PBPCs, as compared with placebo, had more rapid recovery to an absolute neutrophil count above 500/microL (10 versus 20 days) and a shorter time to platelet transfusion independence (16 versus 31 days) [34].
The value of colony-stimulating factors following autologous PBPC transplant is less well defined. The addition of G-CSF may further accelerate neutrophil engraftment, with an absolute neutrophil count >500/microL being achieved one to six days more quickly [36-39]. We begin G-CSF on day six following infusion of autologous PBPCs which results in rapid hematopoietic engraftment by day 10 to 12.
Allogeneic HCT — As with autologous transplantation, accelerated engraftment has been noted in the allogeneic setting with the use of G-CSF mobilized PBPCs [40-42]. The impact of G-CSF-mobilized PBPCs as compared with bone marrow-derived hematopoietic stem cells on the risk of acute and chronic graft-versus-host disease is an area of controversy and is discussed separately. (See "Sources of hematopoietic stem cells", section on 'PBPC versus bone marrow for malignant disease'.)
However, a separate issue is the value of administration of hematopoietic colony-stimulating factors following transplantation.
●In a randomized, controlled trial, 109 patients who underwent allogeneic HCT received either GM-CSF or placebo [43]. GM-CSF therapy was associated with significant reductions in the time to an absolute neutrophil count above 500/microL (13 versus 17 days) and the incidence of infection; no differences in platelet or erythrocyte recovery were noted.
●In a second randomized, placebo-controlled study of 54 patients undergoing HLA-matched allogeneic PBPC transplantation, G-CSF treatment was associated with a significant reduction in the time to an absolute neutrophil count >500 (11 versus 15 days); there were no significant differences in time to a platelet count >20,000/microL (13 versus 15.5 days), red cell transfusion independence, incidence of acute graft-versus-host disease, or 100-day mortality [44]. Similar findings were noted in a third study involving 42 patients, in which there was a trend to earlier hospital discharge in the G-CSF-treated group (day 16 versus day 20) [45].
Similar results have been observed in children undergoing autologous or allogeneic HCT. In one study, 221 children receiving allogeneic or autologous bone marrow (BM) or PBPC transplants were randomly assigned to receive G-CSF or not following transplantation [46]. Myeloid engraftment and neutrophil recovery were significantly faster in transplant recipients receiving G-CSF. A significantly reduced platelet transfusion requirement and earlier patient discharge was noted for patients who received G-CSF in the BM but not the PBPC group.
Graft-versus-host disease — Whether use of colony-stimulating factors following allogeneic HCT increases the risk of developing graft-versus-host disease (GVHD) has not been adequately settled. Two meta-analyses, eight retrospective cohort studies, and one case-control study concluded that there was no significant change in the risk of acute or chronic GVHD after allogeneic HCT when growth factors were used to shorten the initial period of neutropenia [21,47].
This issue was also investigated in a large, retrospective, multi-institutional study that included 1789 patients with acute leukemia [22]. Use of G-CSF in patients receiving transplantation with bone marrow-derived cells was associated with an increased risk of acute GVHD, chronic GVHD, and transplant-related mortality, along with reduced survival. No such effects of G-CSF were noted in patients receiving transplantation with peripheral blood stem cells.
Another large retrospective analysis of 2719 patients from the International Blood and Marrow Transplant Research database evaluated the impact of G-CSF on outcomes for patients transplanted between 1995 and 2000. In this study, which included both patients receiving BM and PBPCs, G-CSF shortened the neutropenic period following transplantation but did not affect treatment-related mortality, acute or chronic GVHD, or leukemia-free survival [48].
Erythropoietin — Erythropoietin (EPO) has been used in an effort to accelerate the recovery of red blood cells. The rationale for this approach was provided in part by the observation that EPO levels were lower than predicted for the degree of anemia following transplantation [49]. However, studies in both allogeneic and autologous transplants have not shown prominent benefits.
Allogeneic transplantation — Most centers do not routinely employ EPO early after allogeneic transplantation. Instead, EPO is generally reserved for those patients with prolonged anemia (eg, transfusion requirements after day 28) and those who develop recurrent anemia after an initial hemoglobin recovery.
Initial small trials suggested that EPO administration produced more rapid recovery of red blood cells and reduced transfusional requirements [50-52]. Randomized trials that used variable doses and schedules of EPO have had mixed results:
●One such trial randomly assigned 215 patients receiving allogeneic HCT recipients to receive placebo or EPO (150 units/kg per day as a continuous infusion) from bone marrow infusion until stable hemoglobin levels were achieved for seven days [53]. The median time to transfusion independence was reduced from 27 to 19 days, but transfusion requirements for the two groups as a whole were similar. There were, however, subsets at high risk for transfusion (eg, GVHD III and IV) in which EPO reduced transfusion requirements.
●Another trial from Australia evaluated 91 patients who underwent allogeneic transplantation followed by randomization to placebo or EPO (300 units/kg three times weekly) [54]. There was no reduction in transfusion requirements, although EPO therapy was associated with increases in the reticulocyte count, hemoglobin concentration, and bone marrow erythropoiesis on day 14.
●In a third trial, 131 patients undergoing allogeneic HCT were randomly assigned to receive placebo or EPO (500 units/kg weekly) as part of one of three cohorts: myeloablative HCT with placebo or EPO starting on day 28; nonmyeloablative HCT with placebo or EPO starting on day 28; or nonmyeloablative HCT with placebo or EPO starting on day 0 [55]. EPO was associated with a lower percentage of patients requiring transfusion (42 verus 77 percent) and fewer units transfused per patient (4.1 versus 6.9 units). Patients who received EPO were also more likely to have attained a hemoglobin ≥13 g/dL by day 126 (63 versus 8 percent, median 90 days versus not reached). Among those undergoing nonmyeloablative HCT, there was no additional benefit to starting the EPO at day 0 when compared with day 28.
Post-transplant anemia may be multifactorial and impacted by transplant complications (eg, GVHD) and medications (eg, ganciclovir, cyclosporine). EPO treatment may decrease the requirements for red blood cell transfusions in patients with GVHD.
Autologous transplantation — Three randomized clinical trials have evaluated the efficacy of EPO after autologous transplantation in patients who were treated with G-CSF; none found a significant reduction in the transfusional requirement for red cells [53,56-58]. The lack of benefit is especially true now that the majority of these transplants are being performed with PBPCs.
Thrombopoietin — Thrombopoietin has the potential to mobilize platelet progenitors, thereby accelerating platelet recovery and treating patients with prolonged platelet recovery and graft failure states. The development of antibodies, resulting in significant thrombocytopenia, has limited the development of these products. (See "Biology and physiology of thrombopoietin".)
With the approval of the thrombopoietic growth factors romiplostim and eltrombopag, additional studies are needed in the HCT setting to determine if these agents can effectively stimulate platelet production, reduce bleeding risks, and reduce transfusional requirements for platelets. (See "Clinical applications of thrombopoietic growth factors".)
CHIMERISM — Allogeneic HCT recipients are assessed periodically with chimerism studies that determine the genotypic origin of post-transplant hematopoiesis. Hematopoiesis can be derived entirely from the donor (ie, complete chimerism), can contain a variable ratio of donor- to recipient-derived cells (ie, mixed chimerism), or be derived entirely from the recipient (ie, engraftment failure). Studies can be performed on the blood or bone marrow and can evaluate total chimerism and chimerism in subsets including granulocytes, and T, B, NK, and CD34+ cells, where clinically appropriate. Chimerism studies help to assess engraftment, potential graft failure, and early relapse, and drive major therapeutic decisions including modifications to immunosuppression post-HCT.
Several methods have been used to assess chimerism after HCT, including determination of short tandem repeats (STR) by quantitative fluorescent polymerase chain reaction (QF-PCR), quantitative real-time PCR (qPCR) of single nucleotide polymorphisms (SNP), and, in the setting of gender-mismatched transplants, fluorescence in situ hybridization (FISH) for the Y chromosome [59].
Chimerism studies rely on the identification of unique DNA fragments in both the donor and the recipient using pre-transplant samples. The post-transplant sample is then assessed for the presence of these unique identifiers. Graft rejection or graft failure is identified when the post-transplant sample lacks the donor-specific DNA fragment. The recipient-specific DNA fragment provides a measurement of residual recipient-derived hematopoiesis.
Chimerism is usually expressed as a percent with 100 percent chimerism reflecting "complete chimerism" with all hematopoiesis being of donor origin. Acceptable levels depend on the underlying condition (inherited/acquired benign condition versus malignant), the type of transplantation (myeloablative versus reduced intensity/nonmyeloablative), timing of the study, the clinical situation (eg, suspected disease relapse), and results from prior studies. Stable mixed chimerism of 10 to 20 percent may be sufficient for certain marrow failure disorders. In contrast, increasing levels of mixed chimerism may precede overt relapse in patients with hematopoietic malignancy. Currently, there are no reliable strategies to enhance donor specific chimerism in the setting of mixed chimerism.
SUMMARY
●Virtually all patients undergoing hematopoietic cell transplantation (HCT) require blood product support in the form of red blood cell and platelet transfusions until the transplanted marrow cells engraft sufficiently to support hematopoiesis. This generally requires 14 to 21 days or more with bone marrow and 10 to 14 days with peripheral blood progenitor cells.
•All cytomegalovirus (CMV)-negative patients who receive bone marrow cells from a CMV-negative donor should receive seronegative blood products. In addition, irradiated blood products with leukodepletion methods should be employed to avoid the risk of transfusion-associated graft-versus-host disease and other transfusion-related complications. (See 'Red blood cell transfusion' above.)
•The indications for transfusion of blood products vary from center to center. With respect to transfusion of red cells, most centers use as a threshold hemoglobin of 7 to 8 g. (See 'Red blood cell transfusion' above.)
•Many centers transfuse for platelet counts less than 10,000/microL or for higher values if clinical bleeding is present (table 1). (See 'Platelet transfusion' above.)
●Cloned hematopoietic growth factors used in the transplant setting include granulocyte colony-stimulating factor (G-CSF) and granulocyte-macrophage colony-stimulating factor (GM-CSF). Randomized phase III trials with both GM-CSF and G-CSF have shown a reduced number of days required for neutrophil engraftment following transplantation with bone marrow as a stem cell source, associated in some studies with a reduction in the days of antibiotic therapy and length of hospitalization. (See 'Growth factor support' above.)
•The usual dose is 5 mcg/kg per day by subcutaneous (SC) injection for G-CSF (filgrastim), which we prefer, or 250 mcg/m2 per day SC for GM-CSF (sargramostim). Therapy is usually begun one to six days after transplantation and continued until the absolute neutrophil count reaches 1,000/microL; a shorter duration until clinically adequate neutrophil recovery is achieved is a reasonable alternative. G-CSF is more commonly used for this purpose. (See 'G-CSF and GM-CSF' above.)
•Given the limited general utility of colony-stimulating factors following allogeneic transplantation, and the fact that their use may have been deleterious in some allogeneic settings, there appears to be little reason to routinely treat allogeneic HCT patients with these growth factors as post-treatment prophylaxis. Exceptions include those patients with delayed neutrophil engraftment or a reduction in white blood cells due to infection or drug treatment.
•Erythropoietin (EPO) does not appear to be of benefit immediately following allogeneic or autologous transplant. (See 'Erythropoietin' above.)
•Additional studies are needed in the transplant setting to determine if thrombopoietic growth factors can effectively stimulate platelet production, reduce bleeding risks, and reduce transfusional requirements for platelets. (See 'Thrombopoietin' above.)
