Pediatric Thrombocytosis

The physiologic reference range of platelet counts is 150-400 X 10⁹/L. A platelet count exceeding the upper limit is called thrombocytosis or thrombocythemia. Thrombocytosis is classified as either primary or secondary.

Primary thrombocytosis

Primary thrombocytosis (also called essential thrombocytosis, essential thrombocythemia, or primary thrombocythemia) consists of 2 types. The first is classical primary thrombocytosis and is caused by autonomous production of platelets unregulated by the physiologic feedback mechanism to keep the count within the reference range. It is a subset of myeloproliferative disorder (eg, essential thrombocythemia [ET]), myelofibrosis with myeloid metaplasia, polycythemia vera, chronic myelocytic leukemia (CML) or, in rare cases, acute myelocytic leukemia. ^[1]

Hematopoiesis in patients with ET is monoclonal and is caused by JAK2^V617Fmutation, thrombopoietin receptor gene (MPL) mutation, calreticulin gene (CALR) mutation, or other rare mutations. JAK2 mutation and CALR mutation are mutually exclusive, and there is a distinct phenotypic difference between these two. ^{[2, 3, 4]} . There have been other mutations reported, including ASXL1 and MPL^Y252H. ^[1] Though CML presenting with extreme thrombocytosis is rare, a boy aged 7 years was reported to have presented with moderate leukocytosis and a platelet count of 2.8 million/μL, with BCR-ABL mutations. ^[5]

The second type of primary thrombocytosis is, in most cases, familial and hereditary. It is caused by a mutation of either the thrombopoietin (TPO) gene or MPL. (Details of each gene mutation are described below.) Hematopoiesis in this type of mutation is polyclonal.

Secondary thrombocytosis

In contrast to primary thrombocytosis, secondary thrombocytosis is an exaggerated physiologic response to a primary event, such as an infection. In pediatrics, primary thrombocytosis is exceedingly rare, whereas secondary, or reactive, thrombocytosis is very common, particularly in infants.

Secondary thrombocytosis (the term reactive thrombocytosis is used in all subsequent discussions) is usually transient and subsides when the primary stimulus ceases. Despite the strikingly high platelet count (on occasions exceeding 1 million/μL), thrombotic and/or hemorrhagic complications are highly exceptional. This is in contrast to the thrombosis and bleeding that are reported complications of essential thrombocythemia (ET).

Reactive thrombocytosis is usually mediated by increased release of numerous cytokines in response to infections, inflammation, vasculitis, tissue trauma, and other factors. Thrombopoietin (TPO), the primary cytokine for platelet production and maturation, and interleukin (IL)-6 levels are usually initially elevated in response to the primary events mentioned earlier; they stimulate platelet production. However, serum or plasma levels of these cytokines do not seem to be correlated with degree of thrombocytosis.

Other cytokines may participate in the stimulation of platelet production. They include IL-3, IL-11, granulocyte-macrophage colony-stimulating factor (GM-CSF), and erythropoietin. These cytokines are directly or indirectly released during the primary events. When the original stimulation stops, the platelet count then returns to the reference range.

In severe infections, such as bacterial meningitis, one of the causes may be a rebound phenomenon after initial thrombocytopenia due to rapid consumption of platelets. This most commonly occurs in neonates and infants, indicating the labile nature of platelet count control in these subjects. Rebound thrombocytosis is also observed in the recovery phase of chemotherapy-induced thrombocytopenia and during the recovery phase of immune thrombocytopenic purpura (ITP).

The most common infection associated with thrombocytosis is pneumonia. Vlacha and Feketea described 102 children admitted with a diagnosis of lower respiratory tract infection; 49 of these children (median age 31 mo) developed platelet counts of over 500 X 10⁹/L. ^[6]

A retrospective study by Zheng et al found that out of 3156 children with respiratory tract infection, 817 (25.9%) had secondary thrombocytosis (500 X 10⁹ platelets/L or higher). ^[7]

A study from Taiwan on pediatric reactive thrombocytosis (platelet count >500,000/μL) showed a positive correlation between platelet count and WBC count and an inverse relation between platelet count and blood hemoglobin level. The same study reported that thrombocytosis is a significant independent risk factor for the length of hospital stay. A study done on an adult population in Israel showed that thrombocytosis is a risk factor for prolonged hospitalization in adults as well; the mortality rate of patients with thrombocytosis was significantly higher than that of patients without thrombocytosis. ^[8]

In some instances, such as chronic hemolytic anemia, the stimulus (hypoxia) to produce cytokines persists, causing long-term elevation of platelet counts. While thrombocytosis in association with iron-deficiency anemia is well documented, the mechanism remains unclear. Although elevated erythropoietin levels are observed in patients with iron-deficiency anemia and thrombocytosis, one study showed that these elevated levels had no correlation with platelet count or with levels of other cytokines potentially responsible for thrombocytosis, such as IL-6 and TPO. In some cases, an increased number of bone marrow megakaryocytes is observed. ^{[9, 10]}

In this context, a study on patients with chemotherapy-induced anemia shed some light on this subject. Patients randomly received intravenous (IV) iron, oral iron, or no iron in addition to erythroid-stimulating agents (ESA). The patients who received IV iron developed the least degree of thrombocytosis, and patients who received no iron developed the greatest degree of thrombocytosis, whereas the patients who received oral iron developed an intermediate degree of thrombocytosis. This observation suggested that although ESA causes thrombocytosis, iron deficiency itself was an additional factor to contribute to thrombocytosis. ^[11]

A rare disorder of unknown etiology, idiopathic cyclic thrombocytopenia is characterized by female predominance, fluctuation of platelet count with rebound thrombocytosis (with peak >1 million/μL), and median age of onset at age 35 years, although the youngest child described was aged 1 year. ^[9] If a patient with this diagnosis were to be evaluated during the rebound thrombocytosis, one might erroneously conclude that the patient developed acquired thrombocytosis.

Pituitary adenylate cyclase-activating polypeptide (PACAP) has been found to inhibit megakaryocytopoiesis and platelet function. PACAP deficiency observed in children with congenital nephrotic syndrome causes thrombocytosis in these patients, ^[12] whereas an extra dose of PACAP genes in partial trisomy 18p patients has been seen to result in prolongation of bleeding time and mild thrombocytopenia. ^[13]

Sporadic (nonfamilial) primary thrombocytosis is usually a clonal disorder, although nonclonal essential thrombocythemia has also been well documented. An MPL polymorphism gene, MPL^Baltimore, belongs to this polyclonal thrombocytosis (see below). The most common diagnosis in the pediatric age group is chronic myelogenous leukemia (CML). Polycythemia vera and myelofibrosis (MF) with myeloid metaplasia are other, rare diagnoses associated with primary thrombocytosis.

In primary thrombocytosis, primary and secondary hypercoagulable states frequently lead to thrombotic episodes and to a hemorrhagic tendency. In about 30% of pediatric cases, JAK2^V617Fmutation has been documented. More recently another mutation involving the CALR gene was independently documented by 2 groups of investigators in patients with myeloproliferative disorders. Though JAK2^V617mutation has been found in both polycythemia vera and essential thrombocytosis patients, CALR mutation was found only in patients with essential thrombocytosis. ^{[2, 3, 4]}

Familial or hereditary primary ET in children are heterogeneous disorders of different molecular abnormalities. Inheritance patterns vary; most familial thrombocythemia cases due to TPO gene mutations are transmitted in autosomal dominant manner. However, some are autosomal recessive. In one family, transmission appears to be X-linked recessive. ^[14]

At least two classes of molecular mutations that lead to familial thrombocytosis are known. One involves mutations of the TPO gene that result in increased TPO production by various mechanisms. The other involves mutations of the c-MPL gene that somehow constitutively maintain activated signal transduction, leading to continuous signaling for megakaryocytic proliferation. In some families, no specific molecular abnormalities have been found. Reported cases in which molecular abnormalities were investigated include those listed below. (Additional new mutations are likely to be reported in the future.)

Familial (hereditary) thrombocytosis reports are as follows:

• MPL mutation

  • A large Arab family with a p.Pro106Leu mutation and no thrombosis was reported by El-Harith et al. ^[11]

  • Abe et al reported an amino acid substitution of Trp(508) to Ser(508) in the intracellular domain of MPL. ^[15]

  • Ding et al reported 8 members of a Japanese family with a mutation in the transmembrane domain of MPL. ^[16]

  • An MPL gene polymorphism, designated as MPO^Baltimore(K39N substitution) causes little-to-moderate thrombocytosis (median of about 400,000) in heterozygous individuals and marked thrombocytosis (800,000-900,000) in homozygous persons. The frequency of MPL^Balitmore was found to be 7% in African American population. ^[17] Thus, some African Americans who were previously diagnosed to have essential thrombocytosis without detectable JAK2 or CALR mutation may have this polymorphism.

• TPO gene mutation or increased blood TPO level

  • Fujiwara et al reported on 3 members in a Japanese family with increased serum TPO levels and no mutation found in the TPO or MPL gene. ^[18]

  • Ghilardi et al and Kikuchi et al reported 4 members in 3 generations in a Japanese family who had a novel point mutation in the TPO gene. ^{[19, 20]}

  • Graziano et al reported on 3 members in a family who had a TPO mutation (G185T) and associated limb defects. ^[21]

  • Kondo et al reported on 5 members in 3 generations of a Japanese family who had a base deletion in the TPO gene (5'UTR). ^[17]

  • Liu et al reported on 11 members in a Polish family with a G→C transversion in the splice donor of intron 3 of the TPO gene. ^[22]

  • Robins and Niazi reported a mother and child with elevated TPO levels. The mutation was not studied. The child had a limb defect. ^[23]

  • Wiestner et al and Schlemper et al reported 11 members of a Dutch family with a G→C transversion in the splice donor of intron 3 of the TPO gene. ^{[24, 25]} Thrombosis and hemorrhage were noted.

  • Stockklausner et al in Germany reported 2 families due to TPO gene c. 13+1 G/C mutation in the splice donor of intron 3. Two members of 1 family had upper limb defects. ^[26] One of the family was previously described by Wiestner et al as above.

• Mutation of other genes

  • Homozygous mutations of interleukin 1 receptor antagonist (IL1RN) reported by Aksentijevich et al ^[27] and homozygous mutations of interleukin 36 receptor antagonist (IL36RN) reported by Rossi-Semerano L et al ^[28] caused significant thrombocytosis and leukocytosis in affected individuals. Treatment with anakinra normalized these counts.

  • In addition to blood count abnormalities, patients with IL1RN mutation showed skin pustulosis, skeletal abnormalities, hepatosplenomegaly, and pulmonary disease, whereas patients with latter mutations showed only dermatological manifestations (ie, systemic pustular psoriasis).

• Neither MPL gene nor TPO gene mutation found or studied

  • Stuhrmann et al reported on 4 Arab siblings with familial thrombocytosis. ^[14]

  • Tecuceanu et al reported on an Israeli-Jewish family with mild thrombocytosis (highest platelet count was 506 X 10⁹/L). ^[29]

  • Patients with microcephalic osteodystrophic primordial dwarfism (MOPD) type II have been described to have a moderate degree of thrombocytosis and leukocytosis. MOPD type 2 is caused by loss of function mutation of pericentrin (PCNT) gene. ^[30]

  • A girl aged 9 years with hypereosinophilic syndrome had, along with thrombocytosis and eosinophilia, dermatologic, neurologic, cardiac, pulmonary, and intestinal abnormalities. The patient, who had a FIP1L1-PDGFRA fusion gene, displayed a good response to imatinib therapy. ^[31]

Causes of secondary noninfectious thrombocytosis reported in the literature are listed below:

• Caffey disease ^[32]

• Granulocyte-colony stimulating factor treatment in neonates ^[33]

• Hepatocellular carcinoma ^[34]

• Low–molecular-weight heparin ^[35]

• Malignant ovarian tumors ^[36]

• Trauma ^[37]

• Toxocariasis - Two children who had extreme thrombocytosis along with hypereosinophilia were reported to mimic myeloproliferative disorder, due to toxocariasis; antihelmintic therapy caused the hematologic abnormalities to subside ^[38]

Acquired ET in children is similar to that found in adults, although JAK2^V617Fmutation (whose role in myeloproliferation is clear) and PRV-1 RNA positivity are seen less frequently in pediatric patients than in adults. CALR mutation has also been described, but the incidence of this somatic gene mutation in children has not been published. CALR mutation causes features of ET without causing other clinical features characteristic of myeloproliferative disorder.

The spleen is the major organ for the destruction of platelets; therefore, after splenectomy, a sharp rise in the platelet count is routinely observed, although the count subsequently undergoes a slow decrease to the reference range. Similarly, functional asplenia that may occur after splenic artery embolization results in thrombocytosis. Investigators in Israel reported a high frequency of thrombocytosis in asymptomatic hyposplenic or asplenic children. Assessments of the children for a splenic deficit, using technetium-99m sulphur colloid scintigraphy, were carried out mainly due to the occurrence of severe or recurrent infections and/or thrombocytosis and/or major immunodeficiency syndrome, with 50% of the imaging studies performed mainly because of persistent thrombocytosis. ^[39] Thus, it may be advisable to perform this imaging study in a selected group of patients who exhibit prolonged thrombocytosis without any obvious cause.

Frequency

United States

Dame and Sutor stated that the annual incidence of newly diagnosed primary thrombocytosis in childhood is 1 case per 10 million population. ^[40] According to these authors, about 75 children with primary thrombocytosis were reported from 1966-2000.

Dror et al published the results of an analysis of 36 children with essential thrombocytosis, but not the incidence of essential thrombocytosis. ^[30]

The frequency of reactive thrombocytosis is far more common than essential thrombocytosis and depends on age. Rates are highest during the first 3 months of life. Preterm infants have higher frequencies than those of term infants. According to Sutor's summary of several studies, 3-13% of hospitalized pediatric patients had a thrombocyte count of more than 500 X 10⁹/L. In one study, 0.5% of hospitalized children had a platelet count more than 800 X 10⁹/L. ^[32]

No evidence suggests that the incidences of either primary or reactive thrombocytosis vary significantly from one country to another or from one ethnic group to another. A Taiwanese study done at a general hospital indicated the incidence of reactive thrombocytosis to be 6.3% of all hospitalized children (birth to age 18 y). ^[33]

International

See above. The incidence of essential thrombocytosis is estimated to range from 1-4 cases per 10 million people younger than 20 years. ^[41]

Mortality/Morbidity

A study by Szuber et al of 361 patients aged 40 years or younger with myeloproliferative neoplasms found that those with essential thrombocythemia (ET) had a median survival period of 35 years, compared with 37 years for polycythemia vera and 20 years for primary myelofibrosis. ^[42]

Thrombotic or hemorrhagic complications caused by reactive or secondary thrombocytosis are described only anecdotally and must be regarded as extremely rare. However, in children with autoimmune disease or vasculitis, such as Kawasaki syndrome, thromboses do develop. In Kawasaki syndrome, this occurs particularly in the coronary arteries, and cardiac complications are the major causes of morbidity and mortality.

In patients with primary nonfamilial thrombocytosis, which is a myeloproliferative disorder, the frequency of thrombosis and/or hemorrhage widely varies among various reports in adults (20-84% for thrombotic complications and 4-41% for bleeding complications). Incidences of hemorrhagic and thrombotic complications in primary thrombocytosis of children are not known.

On the basis of experiences in young adults with primary thrombocytosis, these complications seem to occur less often in children than in adults. ^[43] Teofili et al reported a 0% rate of thrombosis in children with essential thrombocytosis, as opposed to 10 of 32 patients in a study of adults. ^[34] On the other hand, Dame and Sutor reported that about 30% of children with essential thrombocytosis had thromboembolic or hemorrhagic complications at the time of diagnosis or later, and that about 20% of initially asymptomatic children had these complications later. ^[40] These figures are similar to those of adults. Bleeding mainly involves the mucous membranes and skin (eg, GI hemorrhage, hemoptysis, postsurgical bleeding, bruises, epistaxis). Thrombosis involves the veins and arteries. The complication rates in familial thrombocythemia are not well described due to its rarity, but both thrombosis and hemorrhage occur. ^{[44, 43]}

Race

Essential thrombocytosis has no reported racial predisposition.

Sex

Although previously, no sex difference was reported in the frequency of essential or reactive thrombocytosis, the aforementioned study by Szuber et al found a female preponderance in ET in patients aged 40 years or younger. ^[42]

Age

Preterm infants and young infants do not maintain a platelet count in a range that is defined as normal for adults.

The frequency of reactive thrombocytosis is higher in infants and young children (see Frequency) than in older children. Preterm healthy infants have platelet counts higher than those of nonpreterm children. Lundstrom reported that the 95% limit for platelet counts in infants with a birth weight of less than 2000 g was 160-675 X 10⁹/L, with a median value of 375 X 10⁹/L. ^[36]

Matsubara et al reported an age-related shift in mean platelet counts. ^[37] According to the authors, 12.5% of infants younger than 1 month, 35.9% of infants aged 1 month, and 29.2% of those aged 2 months had platelet counts of 500 X 10⁹/L or more, whereas only 0.6% of children aged 11-15 years had such counts.

An age-related reference range of platelet counts in preterm infants (22-42 weeks' gestation) is available. ^[45] According to this article, the 95th percentile line exceeds 700,000 at 35-49 postnatal days in this cohort of patients.