DOI: 10.1542/neo.10-9-e435 2009;10;e435-e445 NeoReviews

Matthew A. Saxonhouse and Martha C. Sola-Visner Thrombocytopenia in the Neonatal Intensive Care Unit

http://neoreviews.aappublications.org/cgi/content/full/neoreviews;10/9/e435located on the World Wide Web at:

The online version of this article, along with updated information and services, is

Online ISSN: 1526-9906. Illinois, 60007. Copyright © 2009 by the American Academy of Pediatrics. All rights reserved. by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village,it has been published continuously since 2000. NeoReviews is owned, published, and trademarked NeoReviews is the official journal of the American Academy of Pediatrics. A monthly publication,

Thrombocytopenia in theNeonatal Intensive Care UnitMatthew A. Saxonhouse,

MD,* Martha C. Sola-

Visner, MD†

Author Disclosure

Drs Saxonhouse and

Sola-Visner have

disclosed no financial

relationships relevant

to this article. This

commentary does

contain a discussion

of an unapproved/

investigative use of a

commercial product/

device.

Objectives After completing this article, readers should be able to:

1. Discuss the incidence of neonatal thrombocytopenia.2. Describe the risks of a low platelet count in neonates.3. Develop a differential diagnosis for the neonate who has thrombocytopenia based on

time of presentation, risk factors, and signs and symptoms.4. Explain the potential mechanisms responsible for nonimmune thrombocytopenia.5. Develop a management plan for neonates who have immune or nonimmune

thrombocytopenia.6. Understand the risks and benefits of platelet transfusions in the neonatal intensive

care unit.

AbstractAs the survival of neonates cared for in the neonatal intensive care unit (NICU)has improved, hematologic issues have been recognized as clinically significantproblems in this population. Thrombocytopenia, in particular, is a common findingamong sick neonates, but there is considerable debate regarding the appropriateevaluation and management of affected infants. This article provides state-of-the artinformation on the pathophysiology, diagnosis, and treatment of neonatal thrombo-cytopenia. Specifically, the risks associated with low platelet counts in neonates arediscussed, and a practical approach to the differential diagnosis of neonates whodevelop thrombocytopenia is provided. Current recommendations for the manage-ment of immune and nonimmune varieties of thrombocytopenia also are reviewed,with an emphasis on the risks and benefits associated with platelet transfusions in thisage group.

Classification and Incidence of NeonatalThrombocytopeniaTraditionally, thrombocytopenia in neonates (as inadults) has been defined as a platelet count of less than150�103/mcL (150�109/L) (1) and has been classifiedfurther as mild (100 to 150�103/mcL [100 to150�109/L]), moderate (50 to 99�103/mcL [50 to99�109/L]), and severe (�50�103/mcL [50�109/L]). Recently, the definition of neonatal thrombocytope-nia was challenged by the results of a large populationstudy involving 47,291 neonates from a large multihos-pital system. In this study, the lower 5th percentiles forplatelet counts were 104.2�103/mcL (104.2�109/L)and 123.1�103/mcL (123.1�109/L) for infantsyounger than 32 weeks’ gestation and late preterm/termneonates, respectively. (2) Because this study did notexclude ill neonates, however, these values should be

*Assistant Professor, Division of Neonatology, Department of Pediatrics, University of Florida, College of Medicine, Gainesville,Fla.†Assistant Professor of Pediatrics, Division of Newborn Medicine, Children’s Hospital Boston and Harvard University School ofMedicine, Boston, Mass.

Abbreviations

CMV: cytomegalovirusELBW: extremely low birthweightGA: gestational ageGVHD: graft versus host diseaseHPA: human platelet antigenMPV: mean platelet volumeIPF: immature platelet fractionITP: immune thrombocytopenic purpuraIVIG: intravenous immune globulinNAIT: neonatal alloimmune thrombocytopeniaNICU: neonatal intensive care unitRP%: reticulated platelet percentagesTpo: thrombopoietinvWF: von Willebrand

Article hematology

NeoReviews Vol.10 No.9 September 2009 e435

interpreted as epidemiologic “reference values” forneonates admitted to the NICU rather than as “nor-mal values” for this population.

Applying the original definition of neonatal thrombo-cytopenia (ie, a platelet count �150�103/mcL[150�109/L]), several studies have shown that 18% to35% of all neonates admitted to the NICU developthrombocytopenia at some point during their NICUstays. (3)(4) The incidence of thrombocytopenia seemsto be inversely related to gestational age (GA), with thehighest numbers found among the most immature in-fants. In a recent cohort analysis, Christensen and asso-ciates (4) observed that 73% of extremely low-birthweight (ELBW) infants had one or more recordedplatelet counts below 150�103/mcL (150�109/L),most frequently during the first postnatal week. Further-more, within that group, thrombocytopenia occurred in85% of neonates who weighed less than 800 g at birthand in 60% whose birthweights were between 801 and900 g.

When interpreting a low platelet count in the NICU,it is important to recognize that improper collectiontechniques or unrecognized platelet clumping can pro-duce a falsely low value. This is particularly important ina well-appearing neonate who has no risk factors forthrombocytopenia and no signs of thrombocytopenia onclinical examination. In these cases, the low plateletcount should be confirmed by repeat sampling.

Neonatal Platelet Function and Risks ofThrombocytopeniaMultiple in vitro studies measuring platelet function incord blood samples (by platelet aggregometry and flowcytometry) have shown consistently that neonatal plate-lets (preterm � term) are hyporesponsive to most plate-let agonists compared with platelets from adults. (5)(6)Paradoxically, however, results of whole blood assays ofprimary hemostasis (such as the bleeding time or plateletfunction analysis) performed on term neonates or onterm cord blood samples suggested slightly enhancedhemostasis in neonates compared with adults. (7)(8)Similarly, whole blood coagulation assays such as throm-boelastography revealed that neonatal coagulation timeswere equal to or shorter than those from adults, againsuggesting adequate hemostasis. (9) Although the exactcauses for these findings are not completely understood,it has been theorized that the higher hematocrits, highermean corpuscular volumes, higher von Willebrand factor(vWF) values, and presence of ultra-large vWF polymersin neonates account for the disparities between the re-

sults of pure platelet function assays and tests of globalhemostasis. (10)(11)

Most of these studies focused exclusively on terminfants, but a number of recent reports have evaluatedthe effects of GA and postconceptional age (GA � post-natal weeks) on platelet function. Briefly, these studiesdemonstrated that: 1) platelet response to agonists andplatelet procoagulant activity were significantly de-creased in ELBW infants during the first days after birthcompared with adults, but improved by postnatal days10 to 14; (12) 2) bleeding times assessed on postnatalday 1 in infants born at 24 to 33 weeks’ gestation weretwice as high as those in infants born at 38 to 41 weeks’gestation, but also improved by postnatal day 10; (13)and 3) CT-ADP closure times (an in vitro measure ofprimary hemostasis) were inversely related to GA in bothcord blood and peripheral blood samples obtained onpostnatal days 1 to 2. (14) Although these observationsstrongly support the existence of developmental defi-ciencies in primary hemostasis in preterm neonates dur-ing the first 1 to 2 weeks after birth, the question ofwhether (and to what degree) such hemostatic abnor-malities contribute to the pathophysiology of intraven-tricular hemorrhage remains unanswered. Nevertheless,important factors must be considered when deciding thepoint at which the bleeding risk for a neonate who hasthrombocytopenia justifies intervention (ie, with platelettransfusions).

Approach to the Neonate Who HasThrombocytopeniaOne of the most helpful pieces of information whenevaluating a neonate exhibiting thrombocytopenia is thetime of presentation. Although there is some overlap, thepathologic processes that cause thrombocytopenia atbirth or during the first 72 hours after birth generallydiffer from those accounting for thrombocytopenia pre-senting after 72 hours. We, therefore, classify the causesof thrombocytopenia into early-onset and late-onset(Tables 1 and 2).

Early-onset ThrombocytopeniaThe initial evaluation for neonates who have early-onsetthrombocytopenia is based on the severity of thrombo-cytopenia, the clinical presentation (Table 1), and thematernal history. The latter is particularly important ifthe mother has autoimmune thrombocytopenia. If that isthe case, the most likely diagnosis in an otherwise well-appearing neonate is autoimmune thrombocytopenia,mediated by the transplacental passage of maternal auto-antibodies. Other important clues in the maternal history

hematology thrombocytopenia

e436 NeoReviews Vol.10 No.9 September 2009

are preeclampsia or chronic hypertension, which are fre-quent causes of placental insufficiency. This is the mostcommon cause of mild thrombocytopenia in healthy-appearing neonates (particularly those born preterm)and can be managed expectantly with close observation.(15) This type of thrombocytopenia is always mild-to-moderate, reaches a nadir on postnatal days 4 to 5, andinvariably resolves by 7 to 10 days after birth. Thus,either progression to severe thrombocytopenia or lack ofresolution within this timeframe should prompt evalua-tion for other causes. The clinical condition of affectedinfants should be monitored closely because thrombocy-topenia can be the first presenting sign of a seriouscondition (eg, sepsis). For this reason, many cliniciansorder blood cultures and consider administering antibi-

otics for well-appearing neonates inwhom the cause of the thrombocy-topenia is not yet clearly defined.

The presence of early-onsetthrombocytopenia (of any severity)in an ill-appearing term or pretermneonate should trigger suspicionfor bacterial or viral sepsis,TORCH (Toxoplasma, rubella, cy-tomegalovirus [CMV], herpes sim-plex virus) infections, or dissemi-nated intravascular coagulation (eg,related to perinatal asphyxia). Thephysical examination can provideimportant clues to the diagnosis,such as hepatosplenomegaly in con-genital viral infections. Appropriatetests should be ordered, and treat-

ment should be initiated based on index of suspicion.Early-onset severe thrombocytopenia in a well-

appearing neonate is primarily suggestive of immunethrombocytopenia, caused by the passage of antiplateletantibodies from the mother to the fetus. In the absenceof maternal thrombocytopenia, the most likely diagnosisis neonatal alloimmune thrombocytopenia (NAIT).However, if a careful physical examination reveals con-genital anomalies or dysmorphic features, a number ofgenetic disorders known to be associated with thrombo-cytopenia also should be considered and evaluated (Ta-ble 3). In these cases, the platelet size, represented bymean platelet volume (MPV), and a review of the periph-eral blood smear can assist in narrowing the differentialdiagnosis. For example, Jacobsen syndrome and Fecht-ner syndrome present with large platelets and otherassociated clinical findings, (16) and Wiskott-Aldrichsyndrome and X-linked thrombocytopenia present withsmall platelets (MPV typically �7 fL). Certain entities,such as congenital amegakaryocytic thrombocytopenia(mutations in the receptor for thrombopoietin withoutother congenital anomalies), can be almost indistin-guishable from NAIT in the neonatal period. Responseto treatments (or lack thereof), long survival of trans-fused platelets, and persistence of the thrombocytopeniabeyond the neonatal period are clues to the correctdiagnosis.

Late-onset ThrombocytopeniaThe evaluation of neonates who have late-onset throm-bocytopenia (Table 2) also is based on the severity of

Table 1. Differential Diagnosis of NeonatalEarly-onset* Thrombocytopenia

Clinical Presentation Degree of Thrombocytopenia Differential Diagnosis

Ill-appearing Variable Sepsis (bacterial, viral†)TORCH‡ infectionsBirth asphyxia

Well-appearing Mild-to-moderate Placental insufficiencyGenetic disorders (see Table 3)Autoimmune

Severe Neonatal alloimmunethrombocytopenia

Genetic disorders (see Table 3)Autoimmune

*�72 hours after birth†Causative viruses: cytomegalovirus, human immunodeficiency virus, herpes simplex virus, enteroviruses‡Toxoplasma, rubella, cytomegalovirus, herpes simplex virus

Table 2. Differential Diagnosis ofNeonatal Late-onset*Thrombocytopenia†

Clinical Presentation Differential Diagnosis

Ill-appearing Sepsis (bacterial, viral‡, fungal)Necrotizing enterocolitisInborn error of metabolism

Well-appearing Drug-induced thrombocytopeniaThrombosisFanconi anemia

*�72 hours after birth†Due to clinical variability in presentations, differential diagnosis ap-plies to all degrees of thrombocytopenia.‡Causative viruses: herpes simplex virus, cytomegalovirus, entero-viruses

hematology thrombocytopenia

NeoReviews Vol.10 No.9 September 2009 e437

thrombocytopenia and the clinical condition of the in-fant. In general, late-onset thrombocytopenia shouldprompt rapid evaluation and treatment for bacterial/fungal sepsis; viral infections such as herpes simplex virus,acquired cytomegalovirus, or enteroviruses; and necro-tizing enterocolitis (Table 2). If these most common

causes are ruled out, other potential causes include in-born errors of metabolism (eg, propionic acidemia,isovaleric acidemia, methylmalonic acidemia, and Gau-cher disease), thromboses (eg, renal vein thrombosis), ordrug-induced thrombocytopenias (eg, heparin, antibiot-ics). Fanconi anemia is a bone marrow failure syndrome

Table 3. Genetic Disorders Associated With ThrombocytopeniaCategory Subtype Other Clinical and Laboratory Findings

Chromosomal Trisomy 13 Aplasia cutis, congenital heart disease, cleft lip andpalate, polydactyly

Trisomy 18 Intrauterine growth restriction, congenital heartdisease, rocker-bottom feet, overlapping digits,hypertelorism, small mouth, clinodactyly

Trisomy 21 Congenital heart disease, transverse palmar crease,hypotonia, short neck with redundant posteriorfolds

Turner syndrome Congenital heart disease, cubitus valgus, webbedposterior neck, broad chest with wide-spacednipples, lower extremity edema

11 q terminal disorder (Jacobsensyndrome, Paris Trousseauthrombocytopenia)

Congenital heart disease, genitourinary anomalies,mild facial anomalies, abnormal brain imaging,limb anomalies

Familial May-Hegglin anomaly, Sebastiansyndrome

Giant platelets, neutrophilic inclusions

Fechtner syndrome Giant platelets, sensorineural hearing loss, cataracts,nephritis, neutrophilic inclusions

Bernard-Soulier syndrome Giant plateletsX-linked macrothrombocytopenia

due to GATA-1 mutationAnemia, genitourinary abnormalities (cryptorchidism)

Congenital amegakaryocyticthrombocytopenia

Abnormalities of head size and shape, developmentaldelay, congenital heart disease, cleft and high-arched palate, abnormal kidneys, optic atrophy,valgus and varus deformities, vertebral anomalies,coloboma, scoliosis, absent bone marrowmegakaryocytes

Wiscott-Aldrich syndrome Immunodeficiency, small platelets, eczemaAmegakaryocytic thrombocytopenia

and radioulnar synostosisRestricted forearm pronation, proximal radioulnar

synostosis in forearm, absent bone marrowmegakaryocytes

Fanconi anemia Hypopigmented and hyperpigmented skin lesions,urinary tract abnormalities, microcephaly, upperextremity radial-side abnormalities involving thethumb, pancytopenia (usually with onset inchildhood)

Thrombocytopenia and absent radii Shortened/absent radii bilaterally, normal thumbs,ulnar and hand abnormalities, abnormalities of thehumerus, congenital heart defects, eosinophilia,leukemoid reaction

Neonatal primary hemophagocyticlymphohistiocytosis

Fever, hepatosplenomegaly, hyperferritemia,hypertriglyceridemia, hypofibrinogenemia

Metabolic Propionic acidemia, methylmalonicacidemia

Failure to thrive, developmental delay, ketoacidosis,hyperglycinemia, hyperammonemia

Isovaleric acidemia Odor of sweaty feet, poor feeding, hypotonia,hyperammonemia, metabolic acidosis

Gaucher disease Hepatosplenomegaly, Gaucher cells in bone marrow

hematology thrombocytopenia

e438 NeoReviews Vol.10 No.9 September 2009

that rarely presents with cytopenia in the neonatal pe-riod, although cases have been described. (17)

Mechanisms of NonimmuneThrombocytopenia

Neonatal Platelet ProductionPlatelet production in neonates, as in adults, can berepresented schematically as consisting of four primarysteps: 1) production of thrombopoietin (Tpo), the mostpotent stimulator of platelet production; 2) proliferationof megakaryocyte progenitors (the cells that multiply andgive rise to megakaryocytes); 3) megakaryocyte matura-tion; and 4) generation and release of new platelets. (18)However, significant developmental differences exist be-tween neonates and adults in regard to platelet produc-tion. Specifically, plasma Tpo concentrations are higherin healthy neonates than in healthy adults, (19)(20)whereas neonates who have thrombocytopenia generallyhave lower Tpo concentrations than similarly affectedadults. (19)(21) Neonatal megakaryocyte progenitorshave a higher proliferative potential than adult progeni-tors, (22) but neonatal megakaryocytes are smaller and oflower ploidy than adult megakaryocytes (and, therefore,generate fewer platelets per megakaryocyte). Thus, it hasbeen postulated that under normal conditions, neonatesmaintain their platelet counts on the basis of the in-creased proliferative potential of their progenitors. (23)In response to thrombocytopenia, however, recent stud-ies have shown that neonates can increase the number,but not the size, of their megakaryocytes, (24) a devel-opmental limitation that might aid in explaining thepredisposition of ill neonates to develop thrombocytope-nia.

Measurements of Neonatal Platelet ProductionIn adults, bone marrow biopsy is the gold-standard testfor the mechanistic evaluation of thrombocytopenia. Inneonates, however, this procedure can be technicallydifficult and frequently is postponed until the infant isout of the neonatal period. With the hope of overcomingthis limitation, a number of potentially useful indirectmeasurements of platelet production were developedover the past decade, including plasma or serum Tpoconcentrations, (25) circulating megakaryocyte progen-itors, (26) reticulated platelet percentages (RP%), (27)and glycocalicin concentrations. (28) To date, most ofthese tests are only available in the research setting.However, a test similar to the RP% recently was devel-oped for clinical use. This RP% equivalent, termed the“immature platelet fraction (IPF),” can be measured by astandard hematology analyzer as part of a routine com-

plete blood count. Similarly to the RP% (and to thereticulocyte count in the evaluation of anemia), IPFvalues are elevated in thrombocytopenic conditions asso-ciated with increased platelet destruction (eg, immunethrombocytopenic purpura [ITP]) and decreased inthrombocytopenias due to decreased platelet production(eg, aplastic anemia, chemotherapy-induced thrombocy-topenia). (29)(30) Only one study has evaluated IPFvalues in neonates who do not have thrombocytopeniaand in neonates who have early-onset thrombocytope-nia. This single study reported that normal neonatal IPFvalues were higher than those previously reported inadults, likely reflecting an increased amount of plateletproduction in neonates (necessary to maintain a normalplatelet count during a period of rapid growth). (31)Most importantly, however, findings from this studysuggested that the IPF could be used (at least in certainvarieties of thrombocytopenia) to predict recovery of theplatelet count within the next 24 hours. Thus, the IPFhas the potential of becoming an important new tool inthe evaluation and management of neonatal thrombocy-topenia.

Mechanisms of Thrombocytopenia in SpecificNeonatal Illnesses

Although none of the indirect measurements has yetbeen validated adequately with concomitant bone mar-row or platelet kinetic studies in neonates, reports inadults and children suggest that the use of several tests incombination can help differentiate between disorders ofincreased platelet destruction and those of decreasedproduction. (28)(32)(33)(34) Indeed, the use of suchmarkers of thrombopoiesis already has provided signifi-cant insight into the mechanisms of some common vari-eties of neonatal thrombocytopenia, particularly thoseassociated with chronic intrauterine hypoxia/placentalinsufficiency and sepsis.

Several studies have implicated platelet underproduc-tion as the primary mechanism for the thrombocytopeniafrequently seen in fetuses and neonates exposed tochronic intrauterine hypoxia. (15)(19) Decreased con-centrations of circulating megakaryocyte progenitors anddecreased numbers of megakaryocytes in the bone mar-row of preterm neonates who have such thrombocytope-nia have been described and have been associated withlower-than-expected plasma Tpo concentrations. Invitro studies also have suggested that the hematopoieticmicroenvironment plays a significant role in mediatingchronic hypoxia-induced suppression of megakaryocyto-poiesis and that megakaryocyte progenitors from pre-

hematology thrombocytopenia

NeoReviews Vol.10 No.9 September 2009 e439

term neonates are more vulnerable than progenitorsfrom term neonates or adults. (35)

Several studies have focused on the pathogenesis ofthrombocytopenia in sepsis and have described elevatedTpo concentrations in septic neonates. A recent evalua-tion of 20 neonates who had sepsis or necrotizing en-terocolitis showed that Tpo and circulating megakaryo-cyte progenitors concentrations as well as the RP% wereelevated. (36) Taken together, these findings suggestedan upregulation of thrombopoiesis mediated by Tpo.Surprisingly, however, the degree of upregulation wasrather modest (two- to threefold), and the neonates whohad gram-negative sepsis did not exhibit the most signif-icant increases in thrombopoiesis, despite having moresevere thrombocytopenia and more severe illness. Thissuggested that the thrombopoietic response in neonatescan be downregulated during severe illness and can reacha state of “relative hypoproliferation.” (37)

Management of Neonatal ThrombocytopeniaNeonatal Alloimmune Thrombocytopenia(NAIT)

NAIT should be considered in any neonate who has aninitial platelet count of less than 50�103/mcL(50�109/L), especially in the absence of other riskfactors or clinical symptoms. A study of more than 150neonates who had thrombocytopenia found that when aplatelet count of less than 50�103/mcL (50�109/L)was used as a screen, 90% of the patients who had NAITwere identified. (38) In addition, the combination ofsevere neonatal thrombocytopenia with a parenchymal(rather than intraventricular) intracranial hemorrhage ishighly suggestive of NAIT. When NAIT is suspected,rapid blood testing is very important for timely andaccurate diagnosis. Blood should be collected from themother and father and submitted for confirmatory test-ing (if easily accessible). Antigen screening, based oncurrent recommendations, initially should include hu-man platelet antigen (HPA) 1, 3, and 5. (39) Thisevaluation should identify approximately 90% of cases ofNAIT. However, if the diagnosis is strongly suspectedand the initial evaluation results are negative, furthertesting should be undertaken for HPA 9 and 15 (andHPA 4 if the parents are of Asian descent). (39) Evalua-tion, if results are positive, reveals an antibody in themother’s plasma directed against the specific plateletantigen in the father. If blood cannot be collected fromthe parents in a timely fashion, the neonate’s serum canbe screened for platelet antibodies. However, a low an-tibody concentration in the neonate coupled with bind-

ing of the antibodies to the neonate’s platelets may resultin a false-negative result.

Due to the complexity of testing, evaluations shouldbe performed in a very experienced reference laboratorythat has a large number of typed controls available forantibody detection and the appropriate DNA-basedtechnology to type multiple antigens. (39) Imaging ofthe brain (ultrasonography, computed tomography scan,or magnetic resonance imaging) is mandatory andshould be performed as soon as possible when a diagnosisof NAIT is suspected. Findings from these studies dictatethe aggressiveness of the treatment regimen in the neo-nate as well as in the mother’s future pregnancies.

Based on recent data demonstrating that a large pro-portion of infants who have NAIT respond to randomdonor platelet transfusions just as well as neonates whohave nonimmune thrombocytopenia, random donorplatelet transfusions now are considered the first line oftherapy for infants in whom NAIT is suspected. (40) Ifthe patient is clinically stable and does not have evidenceof an intracranial hemorrhage, we recommend transfus-ing platelets if the platelet count is less than 30�103/mcL (30�109/L). If the patient has evidence of anintracranial hemorrhage, the goal is to maintain a plateletcount greater than 100�103/mcL (100�109/L). Inaddition to platelets, if a diagnosis of NAIT is confirmedor suspected, intravenous immune globulin (IVIG)(1 g/kg for up to 2 consecutive days) can be infused toincrease the patient’s own platelets and potentially toprotect the transfused platelets when they are adminis-tered. (41) If these treatment options do not efficientlyincrease the platelet count to a safe concentration within1 to 2 days, matched (antigen-negative) platelets must beprovided. Several versions of matched platelets exist,including: 1) maternal platelets, which need to be con-centrated prior to transfusion to limit the amount ofserum transfused to the baby; 2) donor platelets, whichare presumed to be matched because they are known tobe HPA1b1b and 5a5a (“PLA1 and PLA5 negative”) and,therefore, should be compatible in more than 90% ofcases of NAIT; (42) and 3) random donor-matchedplatelets, which are matched after obtaining the results oftyping. (39) In addition, methylprednisolone (1 mg/kgintravenously every 8 hours) might be used on the daysthat IVIG is being administered. (39) The clinical courseof NAIT is short in most cases, often resolving almostentirely within 2 weeks. However, to confirm the diag-nosis, it is important to follow the platelet count fre-quently until a normal count is achieved off treatment.

hematology thrombocytopenia

e440 NeoReviews Vol.10 No.9 September 2009

Autoimmune ThrombocytopeniaThe diagnosis of neonatal autoimmune thrombocytope-nia should be considered in any neonate who has early-onset moderate-to-severe thrombocytopenia and a ma-ternal history of either ITP or an autoimmune diseasewith or without thrombocytopenia. A recent retrospec-tive study of obstetric patients who had ITP (including ahigh number of mothers who had thrombocytopeniaduring their pregnancies) demonstrated a high incidenceof affected babies, with 25% of neonates exhibitingthrombocytopenia at birth, 9% having severe thrombo-cytopenia, and 15% receiving treatment for the throm-bocytopenia. (43) Based on this review, it is recom-mended that all neonates born to mothers who haveautoimmune diseases undergo a screening platelet countshortly after birth. If the platelet count is normal, nofurther evaluation is necessary. (44) If the neonate hasmild-to-moderate thrombocytopenia, the platelet countshould be repeated in 2 to 3 days. If the platelet count isless than 30�103/mcL (30�109/L), IVIG (1 g/kg forup to 2 consecutive days) is the first line of therapy. (45)Random donor platelets, in addition to IVIG, should beprovided if the infant has evidence of active bleeding.Cranial imaging should be used to evaluate for intracra-nial hemorrhage.

Nonimmune ThrombocytopeniaThe management of neonatal nonimmune thrombocy-topenia consists primarily of determining the cause andproviding diagnosis-specific therapy and supportive care.For patients who have moderate-to-severe thrombocyto-penia, the mainstay of therapy is platelet transfusion,which is indicated at different thresholds, depending onthe patient’s clinical status and whether there are anysigns of hemorrhage.

Platelet Transfusions in the NICURecent surveys, in concordance with prior studies, haverevealed substantial variability in platelet transfusionthresholds among neonatologists worldwide. (46)(47)Although the causes for such diversity are multifactorial,the lack of solid evidence to guide neonatal platelettransfusion decisions is likely a major contributing factor.To date, only one prospective, randomized trial hascompared different platelet transfusion thresholds in ne-onates. This study, which was limited to very low-birthweight infants during the first postnatal week, (48)found no differences in the incidence or severity ofintraventricular hemorrhages when platelet transfusionswere administered for platelet counts less than150�103/mcL (150�109/L) versus less than 50�103/

mcL (50�109/L). One additional retrospective studyaddressed the question of whether platelet counts of lessthan 50�103/mcL (50�109/L) could be toleratedsafely in the NICU. (49) The investigators concludedthat using a platelet count of 30�103/mcL (30�109/L)as a transfusion threshold was a safe practice for clinicallystable neonates, particularly after the first postnatal week.(49)

Based on this very limited evidence, we currentlyrecommend transfusing platelets to preterm neonates(�33 weeks gestational age) and to clinically unstableterm infants during the first postnatal week for plateletcounts below 50�103/mcL (50�109/L). After the firstpostnatal week, the threshold can be lowered to30�103/mcL (30�109/L) in clinically stable infants.Platelet transfusions should be given for platelet countsless than 100�103/mcL (100�109/L) only to neonateswho have signs of active bleeding. (44)

More consensus has been reached regarding whichblood products to use. Most experts agree that neonatesshould receive 10 to 15 mL/kg of a standard plateletsuspension, such as platelet concentrates (random-donorplatelets) or apheresis platelets. Whole blood-derivedplatelet concentrates (random-donor platelets) are ob-tained from a single donated unit of whole blood. Eachrandom-donor platelet unit has approximately 50 mL ofvolume and contains approximately 10�109 platelets per10 mL. (50) A single random-donor platelet unit usuallyis sufficient to provide a platelet transfusion to a neonateweighing 5 kg or less and should raise the platelet countto more than 100�103/mcL (100�109/L). (51) Thereis no need to pool more than one random-donor unit fora neonatal transfusion, a practice that (while still some-what prevalent) only increases donor exposures withoutany benefit. Apheresis platelets are collected via anapheresis machine that removes platelets from the donorwhile returning the remaining constituents of wholeblood. Each apheresis platelet unit contains five to eighttimes the number of platelets of a random-donor unit,suspended in approximately 300 mL of plasma. Thus,neonatal transfusions require only small aliquots of theselarge units. (52)

Two additional important considerations in neonatol-ogy are the prevention of transfusion-transmitted CMVinfections and graft versus host disease (GVHD). It iswidely accepted that neonates should receive only CMV-safe blood products. Which product is safest remainscontroversial, but most blood banks provide CMV-negative or leukoreduced products to neonates, whichreduces (but does not eliminate) the risk of transfusion-transmitted CMV infection. GVHD is effectively pre-

hematology thrombocytopenia

NeoReviews Vol.10 No.9 September 2009 e441

vented by irradiating cellular blood products prior totransfusing neonates. This practice is absolutely indicatedfor neonates who have suspected or confirmed immuno-deficiency as well as neonates receiving intrauterine orexchange transfusions, human leukocyte antigen-matched blood, or blood from a first- or second-degreerelative. (51)(53) However, because a primary immuno-deficiency might not yet have been recognized in theneonatal period, many blood banks choose to provideirradiated blood products to all neonates.

It is important for neonatologists to be aware of therisks associated with platelet transfusions for neonates.Specifically, platelet transfusions have been associatedwith transfusion-associated lung injury, a process charac-terized by the onset of hypoxemia and bilateral pulmo-nary infiltrates within 6 hours of a transfusion, which islikely underrecognized and underreported in the NICU.(54) From the infectious standpoint, the primary con-cern with platelet transfusions is the risk of bacterialcontamination, which is higher than the combined risk ofall viral infections for which platelets routinely are tested.(55) Several recent publications have shown a strongassociation between the number of platelet transfusionsand the mortality rate among NICU patients. (56)(57) Itis unclear from these studies whether this associationsimply reflects sicker patients receiving more platelets orplatelet transfusions adversely affecting outcome. Never-theless, platelet transfusion decisions in neonates shouldbe made thoughtfully, carefully balancing the risks andbenefits in each individual patient. It is also clear thatwell-designed randomized, controlled studies are neededto provide solid evidence on which to base transfusiondecisions.

Thrombopoietic Growth FactorsBecause of the risks associated with the use of bloodproducts, thrombopoietic growth factors have been in-vestigated as potential therapies for thrombocytopenia.Three thrombopoietic agents are currently available forclinical use. Recombinant interleukin-11 was the firstfactor approved by the United States Food and DrugAdministration for the prevention of severechemotherapy-induced thrombocytopenia. However,significant adverse effects, such as fluid retention andatrial arrhythmias, somewhat limited its use. (58)

In 2008, two novel thrombopoietic agents were ap-proved for the treatment of chronic refractory ITP: romi-plostim and eltrombopag. Both are thrombopoietin-mimetics, implying that they have no homology withthrombopoietin but act by binding to the thrombopoi-etin receptor on megakaryocytes, thus displaying very

similar effects to thrombopoietin. Both compounds havedemonstrated high efficacy and a favorable safety profilein adults who have thrombocytopenia. However, theirsafety and efficacy have not been evaluated in neonates.Based on recent data suggesting that thrombopoietin hasdifferent effects on neonatal compared with adultmegakaryocytes, (59) it seems judicious at this point topursue additional preclinical research on the effects ofthese compounds on neonatal organisms and to limittheir clinical use in neonates to well-designed clinicaltrials. Furthermore, 4 to 6 days are required for theplatelet count to start rising after therapy with thrombo-poietic agents. Because the median duration of throm-bocytopenic episodes in neonates is 7 days, the use ofthrombopoietic factors appears justified only in carefullyselected neonates who have varieties of thrombocytope-nia that are expected to be more prolonged.

ConclusionThrombocytopenia is a common problem in the NICU,particularly among preterm infants. Most cases are mildto moderate and do not warrant aggressive treatment.However, approximately 6% of all infants admitted to theNICU develop severe thrombocytopenia, defined as aplatelet count less than 50�103/mcL (50�109/L).A thorough and stepwise approach to the neonate whohas thrombocytopenia usually leads to the correct diag-nosis, allows appropriate treatment, and minimizes com-plications. However, there is little consensus on when toadminister platelet transfusions to avoid significant hem-orrhage, and current recommendations are based on verylimited scientific evidence. Large, well-designed, ran-domized, controlled trials are needed to address thisquestion.

References1. Chakravorty S, Murray N, Roberts I. Neonatal thrombocytope-nia. Early Hum Dev. 2005;81:35–412. Wiedmeier SE, Henry E, Sola-Visner MC, Christensen RD.Platelet reference ranges for neonates, defined using data from over

American Board of Pediatrics Neonatal-PerinatalMedicine Content Specifications• Know the causes and pathophysiology of

neonatal thrombocytopenia andthrombocytosis.

• Know the clinical and laboratorymanifestations and management ofneonatal thrombocytopenia and thrombocytosis.

hematology thrombocytopenia

e442 NeoReviews Vol.10 No.9 September 2009

47,000 patients in a multihospital healthcare system. J Perinatol.2009;29:130–1363. Castle V, Andrew M, Kelton J, Giron D, Johnston M, Carter C.Frequency and mechanism of neonatal thrombocytopenia. J Pedi-atr. 1986;108:749–7554. Christensen RD, Henry E, Wiedmeier SE, et al. Thrombocyto-penia among extremely low birth weight neonates: data from amultihospital healthcare system. J Perinatol. 2006;26:348–3535. Israels SJ, Odaibo FS, Robertson C, McMillan EM, McNicol A.Deficient thromboxane synthesis and response in platelets frompremature infants. Pediatr Res. 1997;41:218–2236. Rajasekhar D, Barnard MR, Bednarek FJ, Michelson AD. Plate-let hyporeactivity in very low birth weight neonates. Thromb Hae-most. 1997;77:1002–10077. Boudewijns M, Raes M, Peeters V, et al. Evaluation of plateletfunction on cord blood in 80 healthy term neonates using theplatelet function analyser (PFA-100); shorter in vitro bleedingtimes in neonates than adults. Eur J Pediatr. 2003;162:212–2138. Andrew M, Paes B, Bowker J, Vegh P. Evaluation of an auto-mated bleeding time device in the newborn. Am J Hematol. 1990;35:275–2779. Edwards RM, Naik-Mathuria BJ, Gay AN, Olutoye OO, TeruyaJ. Parameters of thromboelastography in healthy newborns. Am JClin Pathol. 2008;130:99–10210. Israels SJ, Cheang T, McMillan-Ward EM, Cheang M. Evalu-ation of primary hemostasis in neonates with a new in vitro plateletfunction analyzer. J Pediatr. 2001;138:116–11911. Roschitz B, Sudi K, Kostenberger M, Muntean W. ShorterPFA-100 closure times in neonates than in adults: role of red cells,white cells, platelets and von Willebrand factor. Acta Paediatr.2001;90:664–67012. Bednarek FJ, Bean S, Barnard MR, Frelinger AL, MichelsonAD. The platelet hyporeactivity of extremely low birth weightneonates is age-dependent. Thromb Res. 2008;124:42–4513. Del Vecchio A, Latini G, Henry E, Christensen RD. Templatebleeding times of 240 neonates born at 24 to 41 weeks gestation. JPerinatol. 2008;28:427–43114. Saxonhouse M, Garner R, Mammel L, et al. Platelet functionanalyzer (PFA-100) closure times are longer in neonatal bloodcompared to cord blood samples. Neonatology. In press15. Murray NA, Roberts IA. Circulating megakaryocytes and theirprogenitors in early thrombocytopenia in preterm neonates. Pedi-atr Res. 1996;40:112–11916. Drachman JG. Inherited thrombocytopenia: when a low plate-let count does not mean ITP. Blood. 2004;103:390–39817. Parikh TB, Udani RH, Nanavati RN, Rao B. Fanconi’s anemiain newborn. Indian Pediatr. 2005;42:285–28718. Kaushansky K. The molecular mechanisms that control throm-bopoiesis. J Clin Invest. 2005;115:3339–334719. Sola MC, Calhoun DA, Hutson AD, Christensen RD. Plasmathrombopoietin concentrations in thrombocytopenic and non-thrombocytopenic patients in a neonatal intensive care unit. Br JHaematol. 1999;104:90–9220. Dame C. Developmental biology of thrombopoietin in thehuman fetus and neonate. Acta Paediatr Suppl. 2002;91:54–6521. Murray NA, Watts TL, Roberts IA. Endogenous thrombopoi-etin levels and effect of recombinant human thrombopoietin onmegakaryocyte precursors in term and preterm babies. Pediatr Res.1998;43:148–15122. Sola MC, Du Y, Hutson AD, Christensen RD. Dose-responserelationship of megakaryocyte progenitors from the bone marrow

of thrombocytopenic and non-thrombocytopenic neonates to re-combinant thrombopoietin. Br J Haematol. 2000;110:449–45323. Mattia G, Vulcano F, Milazzo L, et al. Different ploidy levels ofmegakaryocytes generated from peripheral or cord blood CD34�cells are correlated with different levels of platelet release. Blood.2002;99:888–89724. Sola-Visner MC, Christensen RD, Hutson AD, Rimsza LM.Megakaryocyte size and concentration in the bone marrow ofthrombocytopenic and nonthrombocytopenic neonates. PediatrRes. 2007;61:479–48425. Emmons RV, Reid DM, Cohen RL, et al. Human thrombo-poietin levels are high when thrombocytopenia is due tomegakaryocyte deficiency and low when due to increased plateletdestruction. Blood. 1996;87:4068–407126. Saxonhouse MA, Christensen RD, Walker DM, Hutson AD,Sola MC. The concentration of circulating megakaryocyte progen-itors in preterm neonates is a function of post-conceptional age.Early Hum Dev. 2004;78:119–12427. Saxonhouse MA, Sola MC, Pastos KM, et al. Reticulatedplatelet percentages in term and preterm neonates. J Pediatr Hema-tol Oncol. 2004;26:797–80228. van den Oudenrijn S, Bruin M, Folman CC, Bussel J, de HaasM, von dem Borne AE. Three parameters, plasma thrombopoietinlevels, plasma glycocalicin levels and megakaryocyte culture, distin-guish between different causes of congenital thrombocytopenia.Br J Haematol. 2002;117:390–39829. Briggs C, Hart D, Kunka S, Oguni S, Machin SJ. Immatureplatelet fraction measurement: a future guide to platelet transfusionrequirement after haematopoietic stem cell transplantation. Trans-fus Med. 2006;16:101–10930. Briggs C, Kunka S, Hart D, Oguni S, Machin SJ. Assessment ofan immature platelet fraction (IPF) in peripheral thrombocytope-nia. Br J Haematol. 2004;126:93–9931. Cremer M, Paetzold J, Schmalisch G, et al. Immature plateletfraction as novel laboratory parameter predicting the course ofneonatal thrombocytopenia. Br J Haematol. 2009;144:619–62132. Koike Y, Yoneyama A, Shirai J, et al. Evaluation of thrombo-poiesis in thrombocytopenic disorders by simultaneous measure-ment of reticulated platelets of whole blood and serum thrombo-poietin concentrations. Thromb Haemost. 1998;79:1106–111033. Kurata Y, Hayashi S, Kiyoi T, et al. Diagnostic value of tests forreticulated platelets, plasma glycocalicin, and thrombopoietin levelsfor discriminating between hyperdestructive and hypoplasticthrombocytopenia. Am J Clin Pathol. 2001;115:656–66434. Robinson M, Machin S, Mackie I, Harrison P. Comparison ofglycocalicin, thrombopoietin and reticulated platelet measurementas markers of platelet turnover in HIV� samples. Platelets. 2001;12:108–11335. Saxonhouse MA, Rimsza LM, Stevens G, Jouei N, ChristensenRD, Sola MC. Effects of hypoxia on megakaryocyte progenitorsobtained from the umbilical cord blood of term and pretermneonates. Biol Neonate. 2006;89:104–10836. Brown RE, Rimsza LM, Pastos K, et al. Effects of sepsis onneonatal thrombopoiesis. Pediatr Res. 2008;64:399–40437. Harker LA, Finch CA. Thrombokinetics in man. J Clin Invest.1969;48:963–97438. Bussel JB, Zacharoulis S, Kramer K, McFarland JG, Pauliny J,Kaplan C. Clinical and diagnostic comparison of neonatal alloim-mune thrombocytopenia to non-immune cases of thrombocytope-nia. Pediatr Blood Cancer. 2005;45:176–18339. Bussel JB, Sola-Visner M. Current approaches to the evalua-

hematology thrombocytopenia

NeoReviews Vol.10 No.9 September 2009 e443

tion and management of the fetus and neonate with immunethrombocytopenia. Semin Perinatol. 2009;33:35–4240. Kiefel V, Bassler D, Kroll H, et al. Antigen-positive platelettransfusion in neonatal alloimmune thrombocytopenia (NAIT).Blood. 2006;107:3761–376341. Mueller-Eckhardt C, Kiefel V, Grubert A. High-dose IgGtreatment for neonatal alloimmune thrombocytopenia. Blut. 1989;59:145–14642. Murphy MF, Bussel JB. Advances in the management of alloim-mune thrombocytopenia. Br J Haematol. 2007;136:366–37843. Webert KE, Mittal R, Sigouin C, Heddle NM, Kelton JG.A retrospective 11-year analysis of obstetric patients with idiopathicthrombocytopenic purpura. Blood. 2003;102:4306–431144. Roberts I, Stanworth S, Murray NA. Thrombocytopenia in theneonate. Blood Rev. 2008;22:173–18645. Ballin A, Andrew M, Ling E, Perlman M, Blanchette V.High-dose intravenous gammaglobulin therapy for neonatal auto-immune thrombocytopenia. J Pediatr. 1988;112:789–79246. Baer VL, Lambert DK, Schmutz N, et al. Adherence to NICUtransfusion guidelines: data from a multihospital healthcare system.J Perinatol. 2008;28:492–49747. Josephson CD, Su LL, Christensen RD, et al. Platelet transfu-sion practices among neonatologists in the United States andCanada: results of a survey. Pediatrics. 2009;123:278–28548. Andrew M, Vegh P, Caco C, et al. A randomized, controlledtrial of platelet transfusions in thrombocytopenic premature infants.J Pediatr. 1993;123:285–29149. Murray NA, Howarth LJ, McCloy MP, Letsky EA, Roberts IA.Platelet transfusion in the management of severe thrombocytopenia inneonatal intensive care unit patients. Transfus Med. 2002;12:35–41

50. Saxonhouse M, Slayton W, Sola M. Platelet transfusions in theinfant and child. In: Hillyer C, Strauss R, Luban L, eds. Handbookof Pediatric Transfusion Medicine. Vol. 1. San Diego, Calif:Elsevier; 2004:253–26951. Strauss RG. Data-driven blood banking practices for neonatalRBC transfusions. Transfusion. 2000;40:1528–154052. Poterjoy BS, Josephson CD. Platelets, frozen plasma, andcryoprecipitate: what is the clinical evidence for their use in theneonatal intensive care unit? Semin Perinatol. 2009;33:66–7453. Gibson BE, Todd A, Roberts I, et al. Transfusion guidelines forneonates and older children. Br J Haematol. 2004;124:433–45354. Goldman M, Webert KE, Arnold DM, Freedman J, Hannon J,Blajchman MA. Proceedings of a consensus conference: towards anunderstanding of TRALI. Transfus Med Rev. 2005;19:2–3155. Fatal bacterial infections associated with platelet transfusions–United States, 2004. MMWR Morb Mortal Wkly Rep. 2005;54(7):168–17056. Baer VL, Lambert DK, Henry E, Snow GL, Sola-Visner MC,Christensen RD. Do platelet transfusions in the NICU adverselyaffect survival? Analysis of 1600 thrombocytopenic neonates in amultihospital healthcare system. J Perinatol. 2007;27:790–79657. Fernandes CJ, O’Donovan DJ. Platelet transfusions in infantswith necrotizing enterocolitis. Curr Hematol Rep. 2006;5:76–8158. Bussel JB, Mukherjee R, Stone AJ. A pilot study of rhuIL-11treatment of refractory ITP. Am J Hematol. 2001;66:172–17759. Pastos KM, Slayton WB, Rimsza LM, Young L, Sola-VisnerMC. Differential effects of recombinant thrombopoietin and bonemarrow stromal-conditioned media on neonatal versus adultmegakaryocytes. Blood. 2006;108:3360–3362

NeoReviews Quiz

1. Thrombocytopenia is defined as a platelet count less than 150�103/mcL (150�109/L) in neonates, as inadults. Using this definition, the incidence of thrombocytopenia among infants admitted to a neonatalintensive care unit has been shown to vary inversely with the gestational age of the infant. Of thefollowing, the estimated incidence of thrombocytopenia among infants whose birthweights are less than800 g, as reported by Christensen and associates, is closest to:

A. 25%.B. 40%.C. 55%.D. 70%.E. 85%.

2. Thrombocytopenia is classified based on the platelet count as mild (100 to 150�103/mcL [100 to150�109/L)]), moderate (50 to 99�103/mcL [50 to 99�109/L]), or severe (<50�103/mcL [50�109/L]). Itis also classified as early-onset (onset within 72 hours after birth) or late-onset (onset more than 72 hoursafter birth). This classification is helpful in identifying the cause of thrombocytopenia in neonates. Of thefollowing, the most common cause of mild, early-onset thrombocytopenia in a well-appearing neonate is:

A. Alloimmune disease.B. Birth asphyxia.C. Genetic disorder.D. Placental insufficiency.E. Viral/fungal sepsis.

hematology thrombocytopenia

e444 NeoReviews Vol.10 No.9 September 2009

3. Evaluation of a neonate who has late-onset thrombocytopenia is based on the severity of thrombocytopeniaand the clinical condition of the infant. Of the following, the most common cause of severe, late-onsetthrombocytopenia in an ill-appearing neonate is:

A. Bacterial/viral sepsis.B. Drug-induced thrombocytopenia.C. Fanconi anemia.D. Inborn error of metabolism.E. Venous thrombosis.

4. Platelet production in neonates, as in adults, occurs in four primary steps: production of thrombopoietin,proliferation of megakaryocyte progenitors, maturation of megakaryocytes, and generation and release ofnew platelets. Several tests have been designed to determine whether thrombocytopenia is caused bydecreased production of platelets or by increased destruction or consumption of platelets. Of the following,the test of platelet production that has the greatest potential for clinical application in the evaluation ofneonatal thrombocytopenia is:

A. Bone marrow biopsy.B. Circulating megakaryocyte progenitor count.C. Immature platelet fraction test.D. Serum glycocalicin concentration.E. Serum thrombopoietin concentration.

5. Neonatal alloimmune thrombocytopenia (NAIT) results from maternal immunization to a foreign fetalplatelet antigen derived from the father and subsequent transplacental passage of maternally derivedantibodies into the fetal circulation. Of the following, the first line of treatment for infants who havesuspected NAIT is the administration of:

A. Intravenous immune globulin.B. Methylprednisolone.C. Random donor platelets.D. Recombinant interleukin-11.E. Thrombopoietin mimetic.

6. The threshold for transfusion of platelets in neonates who have thrombocytopenia varies with the infant’sage and clinical stability as well as evidence of active bleeding in the infant. Of the following, thethreshold for platelet transfusion in a clinically stable neonate after the first week of age, as recommendedby Murray and associates, is a platelet count of less than:

A. 30�103/mcL (30�109/L).B. 50�103/mcL (50�109/L).C. 75�103/mcL (75�109/L).D. 100�103/mcL (100�109/L).E. 150�103/mcL (150�103/mcL).

hematology thrombocytopenia

NeoReviews Vol.10 No.9 September 2009 e445

DOI: 10.1542/neo.10-9-e435 2009;10;e435-e445 NeoReviews

Matthew A. Saxonhouse and Martha C. Sola-Visner Thrombocytopenia in the Neonatal Intensive Care Unit

 

& ServicesUpdated Information

s;10/9/e435http://neoreviews.aappublications.org/cgi/content/full/neoreviewincluding high-resolution figures, can be found at:

Permissions & Licensing

http://neoreviews.aappublications.org/misc/Permissions.shtmltables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://neoreviews.aappublications.org/misc/reprints.shtml

Information about ordering reprints can be found online: