Childhood polycythemia vera and essential thrombocythemia: does their pathogenesis overlap with that of adult patients?

The incidence of polycythemia vera (PV) and essential thrombocythemia (ET) in children and adolescents is extremely low. The annual incidence of PV in patients aged less than 20 years is estimated to be about 2 new cases every 10 million people.1 Similarly, the annual incidence of ET ranges from 1 to 4 new cases every 10 million people.1–3 Therefore, on the whole it can be assumed that PV and ET are between 40 and 90 fold less frequent in children than in adults. The rarity of pediatric PV and ET has determined that, for many years, data on their clinical presentation and biologic features have been sparse and anecdotal.4–10 Furthermore, it has been generally accepted that specific diagnostic criteria developed for adult patients with PV and ET should also apply to pediatric cases.11, 12 Two recent reviews have extensively evaluated the pathogenesis of primary erythrocytosis and thrombocytosis (myeloproliferations independent of external influences) occurring in childhood.13, 14 The authors indicate that, as in the adult population, pediatric age PV and ET can also present as sporadic diseases.12,13 In addition, they underline that several cases of PV and ET reported in children are in fact familial diseases, caused by hereditary defects consisting of specific mutations of the erythropoietin receptor, thrombopoietin or MPL genes.15–20

To date, the discovery of specific genetic defects in adult Ph-negative myeloproliferative diseases (MPDs) has provided fresh insights into the understanding of the molecular pathogenesis of these disorders. Indeed, almost all patients with PV and about half of the patients with ET harbor a somatic point mutation in a highly conserved residue of the pseudokinase domain of the JAK2 tyrosine kinase, the JAK2 mutation.21–25 Furthermore, in the rare PV patients who proved JAK2 negative, other functionally similar JAK2 mutations involving the exon 12 have been described.26 In addition to the presence of specific molecular alterations, the vast majority of female patients with PV, and a significant proportion of those with ET, show a clonal expansion of hematopoiesis, indicating the neoplastic nature of the myeloproliferation.27 These specific MPD markers have been recently investigated in familial clusters of MPDs to explore whether JAK2 mutations could also be present in the germ line, as reported for mutations of the erythropoietin receptor, thrombopoietin or MPL genes.28–32 The results obtained have clarified that JAK2 is a somatic mutation secondarily acquired in this set of patients, occurring as frequently as in sporadic cases. In addition, like sporadic MPDs, these patients may also exhibit a clonal expansion of hematopoiesis.28–32 Although the primary pathogenetic alteration underlying this type of familial MPD still has to be defined, unlike the mutations of the erythropoietin receptor, thrombopoietin and MPL genes15–20 it does not appear to be hereditary.28–32 Interestingly, members belonging to the same family can develop MPDs with different phenotypes.28–32 This topic has been recently discussed in a well-presented review by Skoda and Prchal.33

Figure 1.Proposed diagnostic algorithm for childhood polycythemia (A) and thrombocythemia (B).

The increasing knowledge accumulated over recent years in adult MPDs has prompted hematologists to reconsider pediatric PV and ET.34–36 We have recently published a study carried out in 38 consecutive children with PV and ET,34 diagnosed in accordance with the criteria in use at the time of their first evaluation.11, 12 Among them, a patient with PV and 11 children with ET had a familial history of MPD. We have investigated the entire cohort of patients, including cases with a familial occurrence, for the presence of endogenous erythroid colony (EEC) growth, granulocyte PRV-1 RNA over-expression, JAK2 mutation and clonal hematopoiesis. Furthemore, in familial cases, mutations of the erythropoietin receptor, thrombopoietin and MPL genes were investigated. We found that 9 out of 11 patients with familial thrombocytosis showed an inherited activating mutation of MPL.20 Indeed, the first important observation of this study was the high frequency of hereditary forms in this unselected series of patients. Although the exact prevalence of hereditary disorders in our patient population could be overestimated, since 12 patients belonged to 5 different families, their overall occurrence in pediatric patients is undoubtedly noteworthy. We observed that children with familial disorders did not exhibit the JAK-2 mutation, always had a polyclonal hematopoiesis, and rarely showed EEC growth and PRV-1 RNA over-expression.34 These data indicate that in children investigated for PV or ET, a careful screening for familial thrombocytosis and erythrocytosis is mandatory, since familial MPDs observed at this age are most likely due to inherited, although often unknown, defects. Importantly, the identification of specific hereditary defects could help these young patients to avoid more invasive diagnostic approaches, such as a bone marrow biopsy. Among 26 children with non-familial MPD, the prevalence of the JAK2 mutation was similar in PV (37%) and ET (38%), while a significant proportion of female patients had a clonal hematopoiesis.34 The incidence of JAK-2 mutation in children with PV was significantly lower than in adult patients with PV investigated as control group (92% vs. 37%).34 By contrast, the difference in the proportion of JAK2 mutated patients between adults and children with ET appeared less pronounced (58% vs. 38%).34 In a previous study published in Blood by Randi et al.,35 of 20 children with sporadic ET diagnosed according to the PVSG criteria 11 were examined for the presence of the JAK2 mutation and for the clonality of hematopoiesis. The authors found that only 4 of them (25%) exhibited the JAK2 mutation and that the hematopoiesis was clonal in 4 out of the 15 female patients (28.5%). These findings differed significantly from the detection of 60% of JAK2 mutated patients and of 45% of clonal patients among the 47 adults with ET.35 More recently, El-Monheim et al. evaluated the presence of the JAK2 mutation in 9 children with primary thrombocytosis.36 While the bone marrow histology was suggestive of ET in all patients, the JAK2 mutation was detected only in 1 patient.36 On the whole, the data that have emerged from these studies indicate that children with PV and ET harbor the JAK2 mutation much less frequently than adults.34–36 Although it could be speculated that children showing a wild type JAK2 may become JAK2 positive at a later stage, the presence of EEC growth and of PRV-1 RNA over-expression in patients who proved negative for the JAK2 mutation34 suggests that, in pediatric MPDs, other molecular defects, functionally similar to the JAK2 mutation, could affect the JAK2 dependent signaling pathway.

At the time of writing, we have been investigating 47 consecutive children, of whom 19 had a familial MPD (5 PV and 14 with ET). We identified the MPL activating mutation in 3 further patients with familial thrombocytosis, while no hereditary genetic defects were found in the new cases of familial PV. In addition, all children with PV were investigated for the exon 12 JAK2 mutations,26 but no case carrying these mutations was found.

On the whole, the findings obtained by our group confirm that pediatric erythrocytosis and thrombocytosis are heterogeneous diseases, including both sporadic and hereditary disorders. In adult patients, the presence of JAK2 mutations plays a pivotal role in the classification of MPDs37 and offers the opportunity to modify the current diagnostic approach, based on the exclusion of a secondary myeloproliferation38. In contrast, pediatric cases often recognize different pathogenetic mechanisms. We have recently demonstrated that the proposed diagnostic guidelines of MPDs based on the detection of JAK2 mutations38 are not appropriate for pediatric patients39. Indeed, in the case of childhood MPDs, it should first be established if the disease is congenital or acquired. For this purpose, the first step of the diagnostic screening should investigate the possible presence of a familial occurrence suggestive of the congenital origin of MPD. In fact, this set of patients has a very low probability of carrying JAK2 mutations while they could exhibit hereditary mutations of the erythropoietin receptor, thrombopoietin or MPL genes. In addition, as recently suggested,34–36 JAK2 mutations are detectable only in a minority of children with non-familial PV and ET. For the diagnosis of all other cases, the extensive investigation of complementary MPD markers, such as clonality of hematopoiesis, EEC growth or PRV-1 RNA over-expression should be performed. In line with these observations, we propose a specific diagnostic approach for childhood MPDs (see Figure 1). Meanwhile, it is hoped that studies investigating alternative pathogenetic mechanisms in adult patients with wild type JAK2 MPDs will help to clarify the genetic alterations underlying childhood PV and ET.