Clinical status
Description
Hemoglobin level for transfusion (g/dL)
Level of evidencea
Stable
Asymptomatic, imminent marrow recovery
<7
1C
Vital sign changes
Tachycardia, tachypnea, hypotension
<8
1C
Thrombocytopenia
Recent or active hemorrhage
8–10
1C
Procedure
Potential blood loss
8–10
1C
Anesthesia requirement
<7
1C
Oxygen requirement
Pulmonary or cardiac comorbidities
8–10
1C
Fatigue
Decreased quality of life, especially in adolescents
8–10
1B
Chronic anemia, infancy
Impact on growth or development
8–10
1C
Radiation therapy
Radiosensitizer
See text
2C
2.2.2 Red Blood Cell Administration
Informed consent must be obtained prior to PRBC transfusion with a discussion of the risks, benefits and treatment alternatives to transfusion therapy. PRBCs are often stored in adsol (adenine saline, AS) secondary to a longer storage life (42 days) with a resultant hematocrit of 55–60 % due to the addition of 100 mL of AS per unit blood. All pediatric oncology patients should receive leukoreduced, irradiated PRBCs. Leukocytes are implicated as contributory to the majority of transfusion reactions; leukoreduction has been shown to significantly reduce febrile nonhemolytic transfusion reactions (FNHTRs) as well as infection with viral, bacterial, and protozoal pathogens, and specifically CMV transmission (van Marwijk Kooy et al. 1991; Chu 1999; Vamvakas and Blajchman 2001; Dzik 2002; Heddle 2004; King et al. 2004; Paglino et al. 2004; Yazer et al. 2004; Blumberg et al. 2005). Although leukoreduction decreases CMV transmission, CMV-seronegative units are thought safer in at-risk populations; specifically for pediatric oncology, consensus guidelines recommend CMV-seronegative units in patients receiving hematopoietic stem cell transplantation (HSCT) although this recommendation is controversial (Hillyer et al. 1994; Blajchman et al. 2001; Nichols et al. 2003; Gibson et al. 2004; Ljungman 2004; Vamvakas 2005). Additional studies have shown no difference in CMV transmission rates in HSCT patients receiving leukoreduced versus CMV-seronegative PRBCs (Bowden et al. 1995; Thiele et al. 2011; Nash et al. 2012). PRBC irradiation is recommended in all immunocompromised patients to prevent transfusion-associated graft-versus-host disease (TA-GVHD) by inactivating donor T-cell replication and engraftment in the host (Anderson et al. 1991; Moroff and Luban 1992; Dwyre and Holland 2008; Rühl et al. 2009). Similar to HSCT-related GVHD, TA-GVHD can present with fever, anorexia, vomiting, diarrhea, skin rash, as well as pancytopenia and hepatic dysfunction. Directed donation of PRBCs from family members is generally not recommended due to the cost and time required in addition to not being shown more safe in the prevention of transfusion-associated infection (Strauss et al. 1990).
Determination of the goal hgb should direct the volume of PRBCs transfused and is dependent on the patient’s clinical status, potential for ongoing blood loss and time to recovery from myelosuppressive chemotherapy. Generally 10–20 mL/kg of PRBCs are transfused, rounding to the nearest unit to avoid blood product wastage and historically given over 4 h although this practice is not well studied in hemodynamically stable children. In the patient without ongoing blood loss or alloimmunization, the expected rise in hgb is dependent on the hematocrit concentration of the PRBC product; for an AS preserved unit, the hgb is expected to increase by approximately 2 g/dL for each 10 mL/kg of PRBCs transfused (Davies et al. 2007). Anecdotal teaching that repeat hgb measurement must wait a certain period of time for reequilibration is poorly studied and likely unnecessary (Glatstein et al. 2005; Davies et al. 2007). Patients with severe chronic anemia (i.e., hgb <5 g/dL) are potentially at risk for transfusion-associated circulatory overload (TACO) due to the theoretical concern for cardiogenic pulmonary edema with transfusion in the patient with existing compensatory increase in plasma blood volume to near-normal levels. Variable practice exists, including slow transfusion of 5 mL/kg over 4 h, sometimes with the addition of a diuretic agent such as furosemide. Limited evidence suggests more liberal transfusion rates such as 2 mL/kg/h can be safely used in those patients without underlying evidence of hemodynamic instability or cardiopulmonary compromise (Jayabose et al. 1993; Agrawal et al. 2012).
2.3 Thrombocytopenia
Thrombocytopenia is a common side effect of intensive pediatric oncology therapy with potential risks for morbidity, dependent on the rate of platelet drop and seen more commonly when the platelet count is <20 × 109/L (Belt et al. 1978; Rintels et al. 1994). Petechiae, spontaneous hemorrhage and mucosal bleeding are common with platelet count <20 × 109/L while the risk for severe spontaneous or life-threatening hemorrhage is rare (Slichter and Harker 1978; Consensus Conference 1987; Gmür et al. 1991; Contreras 1998; Norfolk et al. 1998; Schiffer et al. 2001; Athale and Chan 2007). Thrombocytopenia occurs most commonly secondary to suppressed thrombopoiesis from chemotherapy, radiation therapy or infection. Thrombocytopenia at diagnosis in leukemia patients can be secondary to marrow infiltration as well as splenic sequestration in those with splenomegaly. Increased platelet consumption can occur due to hemorrhage and sepsis with secondary disseminated intravascular coagulation (DIC). Frequent platelet transfusion increases the risk of developing platelet antibodies and subsequent refractoriness to further transfusion; therefore, as with PRBC transfusion, the risks and benefits of platelet transfusion must be considered in each individual case prior to the decision to transfuse.
2.3.1 Platelet Transfusion Guidelines
Platelet transfusion guidelines generally recommend prophylactic transfusion at threshold levels depending on the underlying risks of bleeding (Table 2.2). Although no specific pediatric oncology guidelines have been published, the American Society of Clinical Oncology (ASCO) practice guidelines incorporate clinical trials in pediatric oncology (Schiffer et al. 2001). Factors that must be considered when deciding to transfuse platelets include: (1) cause of thrombocytopenia; (2) time to expected resolution; (3) rapidity of platelet count drop; (4) clinical condition of the patient including fever, infection, mucositis, coagulopathy or bleeding; (5) history of severe hemorrhage; (6) recent surgical procedure or planned surgery; and (7) concomitant medications such as amphotericin, enoxaparin and tyrosine kinase inhibitors.
Table 2.2
Platelet transfusion guidelines and level of evidence
Clinical scenario | Description | Platelet count for transfusion (×109/L) | Level of evidencea |
---|---|---|---|
Stable | Asymptomatic, imminent marrow recovery | <10 | 1B |
Procedures | Diagnostic LP | 50–100 (see text) | 1B |
Subsequent LP | <20 | 1C | |
Bone marrow aspiration | Not indicated | 1C | |
Minor surgery: central line placement, bronchoscopy with lavage, sinus aspiration, endoscopy with biopsy | <50 | 1C | |
Major surgery: CNS or solid tumor resection or biopsy | <100 | 1C | |
Signs/symptoms or underlying diagnosis | Minor bleeding: epistaxis, mild mucosal bleeding | <20 | 1C |
Major bleeding: hemoptysis, hemorrhagic cystitis, GI, CNS, tumor necrosis | <100 | 1C | |
Fever | <20 | 2C | |
APL induction | <50 | 1C | |
Newborns | <20–50 | 1C | |
Radiation | <20–50 | 2C | |
DIC | <50 | 1C | |
Coagulopathy | <50 | 1C |
Prospective, randomized trials in adolescents and adults with acute leukemia have reported that prophylactic transfusion can be safely given for a platelet threshold of 10 × 109/L in clinically stable patients (Gmür et al. 1991; Heckman et al. 1997; Rebulla et al. 1997; Wandt et al. 1998).
What platelet threshold is appropriate for patients undergoing procedures, for those with a history of hemorrhage and in those with potential concurrent bleeding risk factors such as fever, infection, and coagulopathy has not been well studied. Similarly, social factors such as distance to clinic and ease of accessing care must also be considered when determining the need for platelet transfusion (Benjamin and Anderson 2002). Although definitive evidence is lacking, consensus panels have concluded that platelets >50 × 109/L and >20 × 109/L are sufficient for major and minor surgical procedures, respectively (Norfolk et al. 1998; Rebulla 2001; Schiffer et al. 2001; BCSH 2003). Risk of bleeding has been found to correlate most with a history of severe hemorrhage rather than platelet count (Friedmann et al. 2002).
What platelet threshold should be utilized for lumbar puncture in pediatric oncology patients is poorly studied. One large retrospective study concluded that a platelet threshold of >10 × 109/L was sufficient for lumbar puncture (LP) although only 3.8 % of patients had platelet count <20 × 109/L at the time of LP (Howard et al. 2000). A follow-up study analyzing risk factors for traumatic LP concluded that African American race, age <1 year, prior traumatic tap within 2 weeks, prior LP with platelets <50 × 109/L, lack of general anesthesia, platelet count <100 × 109/L, interval of <15 days between LPs and a less experienced practitioner were all significant (Howard et al. 2002). What effect this analysis has on practice is unclear. The study was also confounded by a high rate of traumatic (29.3 %; ≥10 RBC/μL) and bloody (10.4 %; ≥500 RBC/μL) LPs (Howard et al. 2002). Based on this analysis the authors conclude that a platelet count >100 × 109/L should be the threshold for diagnostic LP and the procedure be performed by the most experienced practitioner (Howard et al. 2002). Data on the prognostic significance of traumatic LP and theoretical potential of introduction of leukemic blasts into the cerebrospinal fluid are controversial although multiple studies have shown it to be a risk factor for poor outcome (Gajjar et al. 2000; Bürger et al. 2003; te Loo et al. 2006). Whether a lower platelet count in the hands of an experienced practitioner remains a risk factor for traumatic LP is unknown.
Unlike LP procedures, bone marrow aspiration and biopsy can be performed without regard to platelet count as long as pressure is applied to the area after the procedure (BCSH 2003). For patients undergoing central line placement, adult studies have shown that platelet counts >30–50 × 109/L are safe (Stellato et al. 1985; Coit and Turnbull 1988; Lowell and Bothe 1991; Barrera et al. 1996; Doerfler et al. 1996; Ray and Shenoy 1997; Loh and Chui 2007). No similar studies have been reported in pediatric patients. Additionally, the need for a particular platelet count for some post-procedure time period to prevent development of bleeding has not been reported.
2.3.2 Platelet Administration
Platelets for transfusion come as either pooled platelet concentrates (PPCs) or as an apheresis unit. PPCs are aggregated from red blood cell donations and contain ≥5.5 × 1010 platelets. Four to six platelet units are combined to make a PPC. On the other hand, an apheresis platelet unit is obtained from a single donor and is the equivalent of 6–10 PC units (i.e., ≥3 × 1011 platelets). Transfusion with apheresis platelets minimizes exposure to multiple blood donors although whether this decreases the risk of alloimmunization and therefore is of benefit in patient populations requiring frequent transfusion such as pediatric oncology is controversial (NEJM 1997). Apheresis platelet units undergo leukodepletion during the collection procedure, whereas PPCs must be subsequently filtered. Leukodepletion has been shown to decrease the risk of alloimmunization in both PPCs and apheresis units (NEJM 1997). Risk of bacterial contamination is low with both PPCs and apheresis units and has not been found to be significantly different (Schrezenmeier et al. 2007). Storage of platelet units at 20–24 °C with gentle horizontal agitation has been found safe up to 5 days after collection; longer storage times increase the risk for bacterial proliferation and cytokine-mediated reactions (Schiffer et al. 1986; Klein et al. 1997).
As with PRBCs, platelets should be dosed by weight with 10 mL/kg of either PPCs or an apheresis product resulting in an increase of 50–100 × 109/L (Roseff et al. 2002; Fasano and Luban 2008). Determination of response as well as clinical status of the patient can guide future transfusions; generally patients are not given more than one apheresis unit although those with a poor response to transfusion or active bleeding may require higher doses. Platelets are transfused over 30–60 min although more rapid infusion rates have been found safe and effective (Norville et al. 1997). Although considered a risk factor, a direct relationship between number of transfused platelet units and incidence of platelet refractoriness has not been consistently shown in the literature (Howard and Perkins 1978; Dutcher et al. 1981; Schiffer et al. 2001). Platelet refractoriness is defined as an insufficient platelet increment after transfusion on at least two occasions and is the most significant long-term complication of platelet transfusion (Schiffer 1991). ABO incompatibility has been shown to be a risk factor in development of platelet refractoriness (Carr et al. 1990). HLA-matched platelets and crossmatching have been shown effective in improving platelet increment in patients found refractory (Duquesnoy et al. 1977; Heal et al. 1987; Kickler et al. 1988; Welch et al. 1989; O’Connell et al. 1992; Friedberg et al. 1993, 1994; Gelb and Leavitt 1997). As with PRBCs, platelets should be irradiated to prevent TA-GVHD in immunocompromised patients, while CMV seronegativity may be unnecessary in the apheresis product for the CMV-seronegative patient (Luban et al. 2000; Nichols et al. 2003; Dwyre and Holland 2008).
2.4 Granulocyte Transfusion
Patients with prolonged neutropenia are at increasing risk of infection and secondarily lack neutrophils to eradicate infection. Therefore it has been theorized that frequent granulocyte transfusion may be an effective method to fight serious infection in the severely neutropenic patient without imminent count recovery. Although theoretically promising and potentially shown beneficial in small observational studies, consistent data are lacking and some meta-analyses have failed to show a significant benefit (Vamvakas and Pineda 1997; Bishton and Chopra 2004; Robinson and Marks 2004; Grigull et al. 2006; Sachs et al. 2006; van de Wetering et al. 2007; Seidel et al. 2008; Massey et al. 2009; Peters 2009).
In the patient with severe, refractory or progressive bacterial or fungal infection and severe neutropenia likely to continue >1 week, granulocyte transfusion can be considered (Bishton and Chopra 2004). Donors should be mobilized with G-CSF and dexamethasone with a goal transfusion dose of >1.0 × 109 cells/kg in the pediatric patient transfused daily for a minimum of 4–7 days (Chanock and Gorlin 1996; Klein et al. 1996; Massey et al. 2009). Granulocyte transfusion should be ABO compatible and crossmatched as well as irradiated to prevent TA-GVHD (Bishton and Chopra 2004). CMV-seronegative products should be used in CMV-seronegative patients to prevent CMV transmission (Bishton and Chopra 2004).
2.5 Risks of Blood Product Therapy and Their Management
In the past, clerical error leading to ABO-incompatible blood transfusion and secondary acute hemolysis, as well as infectious complications, were the most common transfusion reactions (Williamson et al. 1999; Linden et al. 2000; Myhre and McRuer 2000; Stainsby et al. 2006). Now, due to improved blood safety, transfusion-related acute lung injury (TRALI) has become most common in the adult literature due to increased recognition, and generally noninfectious causes are much more common (Bolton-Maggs and Murphy 2004; Stainsby et al. 2006). Reactions to transfusion may range from mild to life-threatening and the clinician must be able to promptly recognize the potential severe reaction in the face of common transfusion-related symptoms such as fever. Acute reactions are defined as occurring within 24 h while delayed reactions occur beyond the acute period. Potential reactions include acute and delayed hemolysis, FNHTR, allergic reactions, TRALI, TACO, TA-GVHD and infectious complications. In the chronically transfused, alloimmunization to PRBC transfusion must be considered in addition to iron overload. Management of transfusion reactions is generally based on best practice and consensus guidelines rather than an evidence basis (Table 2.3).
Table 2.3
Management guidelines for transfusion reactions and level of evidence
Type | Clinical features | Lab/imaging findings | Management | Level of evidencea |
---|---|---|---|---|
Acute hemolytic transfusion reaction (AHTR) | Immediate onset; fever, anxiety, hypotension, DIC, renal failure | ↑Indirect bilirubin Hematuria | Stop transfusion | 1C |
↑LDH/AST | ICU support | |||
↓Haptoglobin | Fluid resuscitation | |||
↑Plasma-free hgb | Vasopressor support | |||
+DAT | FFP, platelets for DIC | |||
Avoid PRBC transfusion | ||||
Diuretics to maintain urine output once BP stabilized | ||||
Febrile nonhemolytic transfusion reaction (FNHTR) | During or within 4 h of transfusion; fever, chills, rigors, nausea/vomiting, headache | None | Stop transfusion | 1C |
Rule out AHTR and bacterial contamination/sepsis | ||||
Antipyretics | ||||
Restart transfusion if serious adverse reactions ruled out | ||||
Allergic transfusion reaction | Immediate for severe reaction with anaphylaxis (i.e., bronchospasm, hypotension); during or following transfusion for mild reaction (i.e., urticaria, pruritus) | None | Stop transfusion | 1C |
Severe reaction: epinephrine, diphenhydramine, H2 blocker, consider steroid | ||||
Mild reaction: diphenhydramine | ||||
Delayed hemolytic transfusion reaction (DHTR) | >24 h from transfusion and within 2 weeks; fever, chills, jaundice, malaise; can be asymptomatic | ↓Hgb vs expected posttfn increment | Usually no treatment required | 1C |
↑Bilirubin | Potential repeat tfn | |||
↑LDH | Screen for new red cell antibodies | |||
+DAT | ||||
+Red cell alloantibodies | ||||
Transfusion-related acute lung injury (TRALI) | Within 6 h of transfusion; dyspnea, hypoxemia, fever, hypotension, noncardiogenic pulmonary edema | +CXR with diffuse infiltrates | Oxygen | 1C |
Vasopressor support | ||||
Mechanical ventilation | ||||
Unclear benefit for corticosteroids | ||||
Bacterial sepsis | Usually of immediate onset if severe GNR; fever, chills, rigors, hypotension, DIC, oliguria, shock | +Bcx from patient and/or transfusion bag | Stop transfusion | 1C |
Fluid resuscitation | ||||
ICU support | ||||
Vasopressors | ||||
Empiric antibiotics with ceftaz/tobra |
2.5.1 Hemolytic Transfusion Reactions
Acute hemolytic transfusion reactions (AHTRs) present with fever, chills, nausea and vomiting as well as anxiety and discomfort. Additional signs and symptoms include dyspnea, hypotension or shock, oliguria, hemoglobinuria, hemoglobinemia and disseminated intravascular coagulation (DIC). AHTRs generally occur secondary to IgM antibodies to anti-A and anti-B isohemagglutinins but may additionally develop due to other IgM and IgG antibodies. Bystander hemolysis can secondarily lyse recipient red blood cells. AHTR is generally due to ABO incompatibility but may also occur with other immune and nonimmune causes including RBCs damaged by blood warmers, incorrectly prepared frozen PRBCs, bacterial contamination, as well as autoimmune and drug-induced causes of hemolysis. Delayed hemolytic transfusion reactions (DHTRs) present with milder symptoms including low-grade fever, jaundice, and a lower than expected posttransfusion hgb increment and are due to previously sensitized patients without detectable antibody at the time of crossmatch. DHTR occurs with IgG antibody-mediated complement fixation, manifesting as extravascular hemolysis.
Evaluation of a potential AHTR should include each blood unit transfused. Laboratory evaluation involves repeat crossmatch, performance of a direct antibody test (DAT; direct Coombs), as well as measurement of hgb, urinalysis, and plasma-free hgb or haptoglobin. With a more insidious DHTR leading to extravascular hemolysis, laboratory assessment should include hgb, reticulocyte count, DAT, indirect bilirubin and lactate dehydrogenase. Management of an AHTR is based on best practice and includes immediately stopping the transfusion, providing fluid support, and monitoring perfusion and urine output. Vasopressor support may be required as well as management of DIC with platelets and fresh frozen plasma (FFP). Transfusion of additional PRBCs should be avoided due to the potential for continued bystander hemolysis but can be given in the symptomatic patient or if with continued active bleeding.
2.5.2 Infection and Sepsis
Transfusion-transmitted viral infections have markedly decreased with improved donor screening and viral testing measures; current estimated rates of transmission include approximately <1 in 1.5 million for HIV, 1 in 300,000 for hepatitis B virus, and <1 in 2 million for hepatitis C virus (Busch et al. 2005; Dodd 2007; Dwyre et al. 2011). Advanced serologic screening measures have decreased the window period in which viral transmission has recently transpired but without positive testing measures, although, false-negative tests can still occur (Dwyre et al. 2011). Posttransfusion hepatitis is generally caused by viruses including hepatitis A, B, C and E in addition to CMV and EBV. CMV transmission can lead to primary infection in the previously CMV-seronegative recipient or reactivation in the previously infected recipient. CMV can lead to a mononucleosis-type syndrome and immunocompromised patients are at risk for more severe manifestations including nephritis, retinitis, interstitial pneumonitis, colitis and cytopenias (Rubin et al. 1985). Additional rare transmission of human T-lymphotropic retrovirus, human herpesvirus-8 and variant Creutzfeldt-Jakob disease has been noted (Dodd 2007). Parvovirus B19 is not routinely screened and can cause a prolonged reticulocytopenia and anemia in patients with underlying hematologic malignancies (Kaur and Basu 2005).
Bacterial infection must be considered in the patient with a new fever or fever increase ≥1 °C from the previous 24 h during transfusion. Blood contamination can occur at the time of collection or during processing and has decreased significantly with improved blood collection and screening procedures (Wagner 1997; Kuehnert et al. 2001; Stainsby et al. 2006; Dodd 2007). Bacterial infection is much more common with platelets compared to PRBCs since they are stored at room temperature and risk of contamination has been directly correlated to storage time (Morrow et al. 1991). Fatal infection is more likely due to Gram-negative endotoxin production as compared to the more commonly seen Gram-positive organisms (Arduino et al. 1989). Transfusion should be stopped in the event of fever or signs or symptoms of sepsis. The patient should be treated immediately with volume resuscitation, broad spectrum antibiotics and a transfusion workup commenced including blood cultures of the transfusion bag.
2.5.3 Allergic Reactions/Anaphylaxis
Allergic reactions are common, complicating 1–5 % of transfusions and more likely with platelet or plasma transfusion (Couban et al. 2002). Reactions are type I hypersensitivity mediated and usually mild with cutaneous manifestations although systemic symptoms related to anaphylaxis are possible and usually occur within minutes of transfusion commencement. Transfusion should be held once allergic symptoms manifest; diphenhydramine is usually sufficient to manage cutaneous symptoms. In the case of systemic symptoms consistent with anaphylaxis, the transfusion should not be restarted and the patient may require additional medications including epinephrine, steroids and fluid expansion. Any patient with a systemic reaction should be evaluated for IgA deficiency. Prophylaxis with diphenhydramine in the patient with a previous reaction has not been shown beneficial although it is often utilized (Wang et al. 2002; Sanders et al. 2005; Geiger and Howard 2007; Kennedy et al. 2008). It is reasonable to consider prophylactic corticosteroids prior to transfusion in the patient with multiple reactions; if symptoms continue, or if the patient is IgA deficient, washed blood products should be utilized.