R EVIEW Neonatal and infant transfusions

D.A. SESOK-PIZZINIAND D. FRIEDMAN

Neonates are defined as infants younger than 28 dosing for platelet transfusions in these patients,and this days old, and in general many transfusion protocols are presents a clinical challenge owing to the limited data standardized to represent infants younger than 4 months from controlled clinical trials. of age. Often the distinction is made between pre- mature neonates and term neonates, and institutional Guidelines protocols may be established on the basis of age or birth The decision to transfuse is often considered in weight. When this is important in terms of transfusion terms of treatment of severe bleeding in an unstable protocols, premature neonates, neonates, and infants neonate or prophylaxis to prevent bleeding in a more will be described separately. stable neonate. The AABB Pediatric Hemotherapy Committee, by developing a practice consensus, was Indications for Neonatal and Infant Platelet one of the first groups to develop guidelines for Transfusions neonatal platelet transfusion for the purposes of Healthy term infants are born with the same platelet auditing.3 AABB members who were surveyed for the count as adults, whereas premature infants may have study noted that the platelet transfusion criteria for sick platelet counts that are on the lower end of the adult premature infants was too liberal at 100 × 109/L in normal range (150 × 109/L to 450 × 109/L). When the contrast to the medical practice at many institutions to platelet count is less than normal, this is an indication transfuse at less than 50 × 109/L. After that consensus for investigation and possibly treatment that may survey,other published guidelines emerged for infants. include platelet transfusions. Many disorders are asso- These guidelines recommended that stable neonates ciated with thrombocytopenia in neonates, and an with a platelet count of less than 50 × 109/L with active investigation for infections, drug exposures, auto- bleeding or an invasive procedure with production or alloimmunity,thromboses, neoplasms, and failure may require platelet transfusion. Neonates with genetic conditions is important for the evaluation of the platelet counts less than 30 × 109/L as a result of failure need for treatment or transfusion. of platelet production,but no identified bleeding or risk The degree of thrombocytopenia in newborns of bleeding,may also require platelet transfusion. In rare varies, although thrombocytopenia is the most instances, guidelines for neonates may more closely commonly reported hematologic abnormality in resemble guidelines for older children and adults, in neonatal intensive care units (NICU). According to whom transfusions are recommended for platelet Castle et al.,1 75 percent of sick neonates will have a counts of 5 to 10 × 109/L. A sick unstable neonate may transient thrombocytopenia by day 2 of life, which will be transfused if the platelet count drops below 100 × nadir by day 4 and return to normal counts by day 10 of 109/L with active bleeding or when an invasive life in 86 percent of neonates. These neonates may not procedure is anticipated in patients with sepsis, require a transfusion. However, approximately 50 disseminated intravascular coagulation, or other percent of hospitalized neonates will have a platelet mechanism of platelet consumption.4,5 count decrease to less than 100 × 109/L,and 20 percent The sick premature infant is at especially high risk will have a decrease to less than 50 × 109/L.1,2 These for an intracranial hemorrhage (ICH) from a low platelet neonates may require platelet transfusion. There is count or poor platelet function. These infants (younger ongoing controversy over the indications, trigger, and than 37 weeks) have decreased plasma coagulation

IMMUNOHEMATOLOGY, VOLUME 24, NUMBER 1, 2008 15 D.A. SESOK-PIZZINIAND D. FRIEDMAN

proteins and poor platelet function compared with thrombocytopenia are needed. These treatments may older children and adults. In addition, these premature include hemopoietic growth factors. Clinical trials are neonates have an underdeveloped subependymal needed to investigate the use of recombinant matrix that puts them at increased risk for an ICH, thrombopoietin to increase platelet production in particularly in the first 3 to 5 days after birth.6 A thrombocytopenic neonates. multicenter prospective, randomized controlled trial investigated whether early use of platelet concentrates Selection of Platelet Components reduced the incidence or extension of ICH in sick Platelet components may be platelet pheresis 2 neonates. This study concluded that neonates who (single-donor or platelets) or platelets were transfused when their platelet count decreased to (–derived platelets). Both platelet compo- 9 less than 150 × 10 /L did not show a significant nents are currently stored for 5 days,but they differ with difference in a new, or extension of an, ICH compared respect to their platelet counts and risk for adverse with the untransfused control group. Infants in the effects. Apheresis platelets are prepared using an control group received a transfusion when their platelet automated collection device from a single donor and 9 counts were less than 50 × 10 /L or the infant was collected in volumes of 200 to 400 mL, which includes bleeding. However,the study group did show an overall platelets, donor plasma, anticoagulants, leukocytes, and reduction in the use of plasma and RBCs compared with a few RBCs. Apheresis platelets that are leukocyte the controls. This study would suggest that prophylactic reduced should contain less than 5 × 106 WBCs and transfusion at a higher trigger value does not impact greater than 3 × 1011 platelets. In contrast,whole blood– clinical outcome for risk of ICH. derived platelets are prepared from centrifuged whole More recent guidelines for platelet transfusion blood to a final concentrated volume of about 50 mL. recommended a lower trigger value of less than 20 × Whole blood–derived platelets also contain donor 109/L or 30 × 109/L for platelet transfusions in term plasma, anticoagulant, leukocytes, and RBCs. Whole infants.7,8 The rationale for the lower trigger value is blood–derived platelets that are leukocyte reduced have that most serious bleeding caused by thrombocyto- less than 5 × 106 WBCs and greater than 5.5 × 1010 penia occurs in the first days of life. Therefore,patients platelets.11,12 severely thrombocytopenic from sepsis or necrotizing The standard dose for neonates is based on body enterocolitis beyond the first few days of life rarely have weight and is the same for either apheresis platelets or major hemorrhage. Other recommendations given are whole blood–derived platelets at 5 to 10 mL/kg for an a higher trigger level of less than 50 × 109/L for patients increase of 50,000/ L in the platelet count. Some who are at highest risk for hemorrhage caused by centers calculate doses based on random donor units or clinical instability or very low birth weight (less than their equivalents.13 In some institutions, dose is based 1000 g). Lastly,a trigger value of less than 50 × 109/L on the platelet count from the component, and the would be indicated for neonates with major bleeding aliquot given to the neonate is adjusted accordingly. If from a pulmonary, gastrointestinal, or renal source. In apheresis platelets are used for platelet transfusions,the reviewing studies of contemporary platelet transfusion platelets may be separated into smaller aliquots to practice in neonates, Murray and Roberts9 noted that decrease donor exposures and platelet wastage. The there were a variety of platelet triggers used in different aliquot or platelet syringe is considered an open system, neonatal intensive care units and that thrombo- and the platelets will expire within 4 hours. As long as cytopenic neonates were 10 times more likely to die the separation into a syringe is done with a sterile than neonates who did not receive transfusions. Del connecting device, the expiration time of the “mother Vecchio et al.10 also concluded that neonates who unit” should not change.11,12 However, the viability of receive more than four platelet transfusions had a risk the platelets may be affected by altering the volume of death 29.9 times that of neonates who did not within the storage bag. Transfusion services preparing receive a transfusion. This is most likely attributable aliquots from apheresis platelets should be aware of to the underlying disease and severity of the the manufacturer requirements for platelet storage. clinical condition resulting in the thrombocytopenia. This practice of providing platelet aliquots from The lack of definitive criteria for prophylactic apheresis platelets is controversial because of the platelet transfusions in this population caused the accuracy of the platelet component in volume only and authors to conclude that alternative treatments for not in final platelet concentration.

16 IMMUNOHEMATOLOGY, VOLUME 24, NUMBER 1, 2008 Neonatal and infant platelet transfusions

Both platelets and apheresis platelets require the Neonates may require additional special compo- same storage conditions, but the testing for bacterial nents as a result of the underlying disease causing the contamination remains disparate between the two thrombocytopenia. Neonates with neonatal alloimmune components. Platelet pheresis components routinely thrombocytopenia will require special platelets that lack undergo sampling for bacterial contamination via the platelet-specific implicated in the disease. culture methods. In the case of whole blood–derived These antigen-negative platelets may be available from a platelets,bacterial testing is often done by pH testing or maternal source or another allogeneic donor. It is also other methods less sensitive than those used for conceivable that maternal HLA passively apheresis platelets. In the future, changes in standard acquired in the neonate may impact response to platelet pooling practices for whole blood–derived platelets and transfusions. Transfusion-transmitted CMV is of concern improved testing methods may provide the most in the premature neonate because of the risk of serious sensitive and specific bacterial testing for both platelet CMV disease in premature infants. Components with components. reduced risk for CMV transmission are either negative when tested for CMV antibodies or leukocyte reduced. Risk of Platelet Transfusion in the Neonate Patients at risk for severe CMV infection include those and Infant with congenital disorders or AIDS, The infectious risk for each component may be hematopoietic progenitor cell transplant recipients, perceived as different for neonatal transfusions because organ allograft transplant recipients, premature infants of the ability to transfuse many more aliquots from during infancy, cancer patients undergoing intense apheresis platelets compared with whole blood–derived chemotherapy, and recipients of intrauterine trans- 5 platelets. This provides a mechanism for donor limi- fusion. Irradiation of platelet components to prevent tation with regard to platelet transfusion and therefore transfusion-associated GVHD is also indicated for exposes the neonatal patient to less infectious risk. premature infants. The requirement for CMV-negative Strict adherence to sterile technique for the aliquot and irradiated platelets in the term neonate is more preparation is necessary to avoid bacterial contami- controversial. nation during the aliquot manufacturing process. In addition to infectious risk, the neonate is Protocols for Neonatal and Infant Platelet exposed to other risks of transfusion with the use of Transfusion at The Children’s Hospital of either whole blood–derived platelets or apheresis Philadelphia platelets. Hemolytic risks from ABO-incompatible At The Children’s Hospital of Philadelphia (CHOP), plasma may be significant in the neonate owing to we support a very large infant intensive care unit with small blood volumes, although a precise risk estimate 60 NICU and 15 cardiac intensive care unit (CICU) beds. for neonates is not available. Accrediting agencies such Our NICU and CICU neonates and infants receive as the College of American Pathologists and AABB irradiated, “CMV-safe” leukocyte-reduced apheresis require protocols to ensure safety ofABO-incompatible platelets when a transfusion is required. ABO and D transfusions.11,14 If possible, infants should receive type-specific or compatible apheresis platelet aliquots only platelets that are ABO-group specific or are selected. If ABO group-specific or compatible is not compatible with the infant’s plasma. WhenABO-group available, saline replacement is performed to remove specific or compatible platelets are not available, the ABO-incompatible plasma. Saline suspension requires institution may elect to modify the component to centrifugation to remove plasma followed by the sterile remove plasma and reduce the risk for . addition of 0.9% normal saline, usually back to the These modifications include saline replacement, starting volume. These components are placed back on volume reduction, and washing.12 Depending on the the rotator for approximately 1 hour to ensure proper institution, all or some of these modifications may be mixing. Aliquots are manufactured in the available. With each modification, less platelets are with the use of a sterile connecting device. The syringe recovered and some will become activated,which can aliquots expire within 4 hours as an open system, and adversely affect the efficacy of the transfusion. Until the remaining platelets are stored in the original the platelet supply is sufficient for demand, these container with no change in expiration time. During modifications are necessary for platelet transfusion blood shortages,whole blood–derived platelets are used support in the neonate. for transfusion. Our present policy calls for irradiation

IMMUNOHEMATOLOGY, VOLUME 24, NUMBER 1, 2008 17 D.A. SESOK-PIZZINIAND D. FRIEDMAN

Fig. 2. Frequency distribution of patient ages at the time of platelet Fig. 1. Frequency distribution of 909 pretransfusion platelet counts in transfusion (N = 909) in infants who received platelet transfusions infants who received platelet transfusions while admitted to The while admitted to The Children’s Hospital of Philadelphia NICU Children’s Hospital of Philadelphia NICU between July 1,2006,and between July 1, 2006, and June 30, 2007. June 30, 2007. of all platelet products in inventory, but it excludes Neonates and infants on extracorporeal membrane irradiation of RBC components for term neonates in oxygenation (ECMO) are at considerable risk for surgery unless there is a specific medical indication for thrombocytopenia. Platelets are given at startup and at irradiation. circuit change to avoid the thrombocytopenia generally CHOP provides D type–specific or compatible plate- associated with the procedure as a result of platelets lets and would recommend RhIG administration in the binding to the tubing or activated within the ECMO event a female neonate with D– RBCs received D+ circuit. Additional dilutional effects from the massive components. Our policy reinforces concerns raised transfusion of RBCs and plasma also factor into the risk with the reported case from Brigham and Women’s for thrombocytopenia. Approximately 10 to 35 percent Hospital where an infant,at 17 weeks of age,developed of neonates receiving ECMO will have a hemorrhagic anti-D from a platelet transfusion. This case report event.16 Patient platelet counts need to be closely suggests that only a small amount of D+ RBCs may cause monitored during ECMO. Adjustments are required to formation in an infant just beyond 4 months of allow a higher transfusion of platelet dose based on age.15 Further data need to be collected to determine weight. At CHOP, ECMO infants and neonates are the sensitization risk of D+ platelet transfusion in D– transfused platelets to a maximum volume of 40 mL/kg. infants around 4 months of age. At CHOP,we have surveyed our institutional platelet Conclusions triggers in the NICU,and our data are similar to those published by Del Vecchio et al.,10 who noted a range of Neonatal and infant platelet transfusions are given platelet transfusion triggers. Figure 1 displays the platelet in a prophylactic or therapeutic setting to thrombo- count that immediately preceded the transfusion in 909 cytopenic neonates and infants,either to reduce the risk platelet transfusion events in our NICU. A trigger platelet of bleeding or for active bleeding. Because thrombo- count value of less than 100 × 109/L would account for cytopenia in a neonate can be caused by a variety of about 90 percent of the platelet transfusions ad- factors, it is important for the hospital to develop ministered in the CHOP NICU. However, according to indications, guidelines, and thresholds for platelet published guidelines, a more stringent trigger value of transfusions. Neonates and infants receiving trans- less than 50 × 109/L may have been considered in some fusions are at additional risk for other adverse reactions of these patients. Figure 2 shows that the majority of to blood products, including infection, immunologic platelet transfusions in neonates occurred in the first effects, transfusion-related acute lung injury, and fluid week of life. It appears that the majority of the platelet overload, among others. Institutional protocols should transfusions were given in accordance with published define the need for irradiation and CMV-negative guidelines. As we begin to look at our institutional data components in infants, neonates, and premature and outcomes, we may better understand the risks and neonates. Standardization of practice, when possible, outcomes of our NICU trigger values,and our guidelines allows for better process control and patient may change accordingly. management.

18 IMMUNOHEMATOLOGY, VOLUME 24, NUMBER 1, 2008 Neonatal and infant platelet transfusions

References 12.Brecher ME, ed. Technical manual. 15th ed. 1. Castle V,Andrew M, Kelton J, et al. Frequency and Bethesda, MD:AABB, 2005. mechanism of neonatal thrombocytopenia. J 13. Hillyer CD, Strauss RG, Luban NLC. Handbook of Pediatr 1986;108:749–55. pediatric . San Diego, CA: 2. Andrew M,Vegh P, Caco C, et al.A randomized, Elsevier,2004. controlled trial of platelet transfusions in 14. College of American Pathologists Commission on thrombocytopenic premature infants. J Pediatr Laboratory Accreditation Laboratory Accreditation 1993;123:285–91. Program Transfusion Medicine Checklist.Available 3. Strauss RG, Blanchette VS, Hume H, et al. National at: http://www.cap.org/apps/docs/laboratory_ acceptability of American Association of Blood accreditation/checklists/transfusion_medicine_oct Banks Pediatric Hemotherapy Committee guide- ober2006.pdf. Last accessed March 17, 2008. lines for auditing pediatric transfusion practices. 15. Haspel RL,Walsh L,Sloan SR.Platelet transfusion in Transfusion 1993;33:168–71. an infant leading to formation of anti-D: 4. Roseff SD, Luban NLC, Manno CS. Guidelines for implications for immunoprophylaxis.Transfusion assessing appropriateness of pediatric transfusion. 2004;44:747–9. Transfusion 2002:42:1398–413. 16. Festa CJ, Feng AK, Bigos D.Transfusion support in 5. Roseff SC.Pediatric transfusion:A physician’s hand- pediatric surgery, trauma, and the intensive care book. 2nd ed. Bethesda, MD:AABB Press, 2006. unit. In: Herman JH, Manno CS, eds. Pediatric 6. Strauss RG.Transfusion therapy for neonates.Am J transfusion therapy. Bethesda, MD: AABB Press, Dis Child 1991;145:904–11. 2002. 7. Murray NA, Howarth LJ, McCloy MP,et al. Platelet transfusion in the management of severe thrombo- Deborah A. Sesok-Pizzini, MD, MBA (corresponding cytopenia in neonatal intensive care unit patients. author), Medical Director, Blood Bank and Transfus Med 2002;12:35–41. Transfusion Medicine, The Children’s Hospital of 8. Gibson BE,Todd A, Roberts I, et al.Transfusion Philadelphia, Assistant Professor, Department of guidelines for neonates and older children. Br J Clinical Pathology and Laboratory Medicine, and Haematol 2004;124:433–53. David Friedman, MD, Associate Medical Director, 9. Murray NA, Roberts IA. Neonatal transfusion Blood Bank and Transfusion Medicine, The Children’s practice.Arch Dis Child Fetal Neonatal Ed 2004;89: Hospital of Philadelphia, Assistant Professor, F101–7. Department of Hematology, University of 10. DelVecchioA,Sola MC,Theriaque DW,et al.Platelet Pennsylvania School of Medicine, 34th Street and Civic transfusion in the neonatal intensive care unit: Center Boulevard, Philadelphia, PA 19104-4399. factors predicting which patients will require multiple transfusions.Transfusion 2001:41;803–8. 11. The AABB Standards Program Committee, ed. AABB standards for blood banks and transfusion services. 24th ed. Bethesda, MD:AABB, 2006.

IMMUNOHEMATOLOGY, VOLUME 24, NUMBER 1, 2008 19