Blood Journal
Leading the way in experimental and clinical research in hematology

Brief Report
The revised WHO diagnostic criteria for Ph-negative myeloproliferative diseases are not appropriate for the diagnostic screening of childhood polycythemia vera and essential thrombocythemia

  1. Luciana Teofili1,
  2. Fiorina Giona2,
  3. Maurizio Martini3,
  4. Tonia Cenci1,
  5. Francesco Guidi1,
  6. Lorenza Torti1,
  7. Giovanna Palumbo2,
  8. Angela Amendola2,
  9. Giuseppe Leone1,
  10. Robin Foà2, and
  11. Luigi M. Larocca3
  1. 1Department of Hematology, Catholic University, Rome, Italy;
  2. 2Division of Hematology, Department of Cellular Biotechnologies and Hematology, La Sapienza University, Rome, Italy;
  3. 3Department of Pathology, Catholic University, Rome, Italy


In the proposed revised World Health Organization (WHO) criteria for the diagnosis of BCR-ABL myeloproliferative diseases (MPDs), exclusion criteria have been replaced by the presence of JAK2 mutations. We applied these criteria to 45 children with MPDs: 13 with polycythemia vera (PV) and 32 with essential thrombocythemia (ET). Among these 45 patients, 12 with ET and 5 with PV had a familial history of MPD, and had been investigated for hereditary mutations of the erythropoietin receptor, thrombopoietin, or MPL genes. We found that the JAK2V617F mutation in children occurs less frequently than in adults, and that exon 12 JAK2 mutations are absent. On the basis of the revised WHO criteria, a significant proportion of childhood PVs were misdiagnosed. Furthermore, all familial ET, including patients carrying the hereditary MPLSer505Asn activating mutation, were erroneously diagnosed as MPDs. Our observations suggest that childhood MPDs require a set of specific diagnostic criteria.


The recent discovery of specific genetic defects in Philadelphia chromosome (Ph)–negative myeloproliferative diseases (MPDs)17 has given new insights in understanding the molecular pathogenesis of these disorders. They indicate the neoplastic nature of the disease and represent specific markers of MPDs that could modify the current diagnostic approach, based on the exclusion of a secondary myeloproliferation.8,9 For this purpose, a panel of clinical investigators and pathologists with expertise in myeloproliferative disorders has recently revised the World Health Organization (WHO) criteria for the diagnosis of BCR-ABL MPDs.10 The novelty of the proposed diagnostic guidelines is that JAK2 mutation analysis plays a pivotal role in the diagnosis of polycythemia vera (PV), while it complements bone marrow histology in essential thrombocythemia (ET) and idiopathic myelofibrosis (IM).10

MPDs in childhood are quite rare diseases.11,12 Although pediatric and adult MPDs exhibit similar hematologic findings,13,14 we and others have recently documented that the incidence of JAK2V617F mutations in childhood PV and ET is significantly lower than in adults.1517 Moreover, we could highlight that several pediatric cases of ET are in reality hereditary disorders, often resulting from the MPLSer505Asn activating mutation.15 In order to evaluate whether the revised WHO diagnostic criteria10 are appropriate for the diagnostic screening of pediatric PV and ET, in the present study we have applied them to a series of 45 children with PV and ET, including sporadic and familial forms.

Patients and methods

Patient samples were obtained following informed consent in accordance with the Declaration of Helsinki and with the approval from the Institutional Review Board of the Catholic University, Rome, Italy.

We evaluated 45 children with PV (n = 13) and ET (n = 32) consecutively observed between 1980 and April 2007 at the Division of Hematology of the University “La Sapienza” of Rome. All patients were diagnosed in accordance to the Polycythemia Vera Study Group (PVSG) and/or WHO criteria.8,9 A total of 17 patients (12 with thrombocythemia and 5 with polycythemia) had a familial history of MPD. In all these patients, the disease appeared to be inherited, with the disease present in one of the parents of the affected children.

Peripheral blood samples were collected after informed consent. The granulocyte cell fraction was isolated and subjected to molecular analysis for the JAK2V617F mutation.15,18,19 The exon 12 JAK2 mutations were investigated in patients with polycythemia who proved negative for the JAK2V617F mutation, according to the method described by Scott et al.6 In all patients with a familial history of polycythemia or thrombocythemia, a mutation analysis was carried out for the erythropoietin receptor gene (Epo-R; exon 8)20 and for the thrombopoietin (TPO; 5′ untranslated region)21 and MPL (exon 10) genes.22 All female patients were tested for clonality of hematopoiesis.15,18 In addition, all patients were evaluated for endogenous erythroid colony (EEC) growth.18 The methods used for JAK2, Epo-R, TPO, and MPL mutation analysis, hematopoiesis clonality, and EEC assays have been described in detail elsewhere.15,18,19

Results and discussion

Patients with polycythemia

The proposed revised WHO criteria for a diagnosis of PV require the fulfillment of both major criteria (increased red cell volume and presence of JAK2V617F mutation or other functionally similar JAK2 mutations) and of 1 minor criterion (panmyelosis at bone marrow histology, low serum Epo level, or EEC growth)10 In the absence of JAK2 mutations, the presence of 2 additional minor criteria is necessary for a diagnosis of PV.10 All evaluated patients had an increased red cell volume and showed a hypercellular bone marrow, as required.10 However, the JAK2V617F mutation occurred rarely in children with sporadic polycythemia (37%), and it was not detectable in patients with familial disorders (Table 1). Furthermore, in contrast to adult patients,6 no additional exon 12 JAK2 mutations were detected in JAK2V617F-negative cases. The serum Epo level was normal in most patients with sporadic polycythemia and in all patients with familial polycythemia (Table 1). As a consequence, the presence of EECs might play a fundamental role in confirming or excluding a diagnosis of PV. In line with the low incidence of JAK2 mutations, only 4 of the 13 investigated patients exhibited an EEC growth (3 with sporadic disorders and 1 with familial disorder). We failed to find Epo-R mutations among our patients with familial disorders (Table 1). On the whole, the proposed revised WHO criteria allowed a diagnosis of PV in 5 of 8 patients with sporadic polycythemia and in 1 girl with familial polycythemia (Table 1). Interestingly, this latter patient exhibited a polyclonal hematopoiesis in the human androgen receptor assay (HUMARA; Table 1).

View this table:
Table 1

Incidence of myeloproliferative markers and hereditary polycythemia markers in 13 children with PV

Patients with thrombocythemia

The proposed revised WHO criteria for a diagnosis of ET require the fulfillment of 4 criteria: (1) sustained platelet count of 450 × 109/L or more; (2) typical bone marrow histology; (3) exclusion of PV, IM, chronic myelogenous leukemia, myelodysplastic syndromes, or other myeloid neoplasms; and (4) presence of the JAK2V617F mutation or of another clonal marker, or, in the absence of a clonal marker, no evidence of reactive thrombocytosis.10 All pediatric patients with both sporadic and familial thrombocythemia fulfilled the first 3 criteria. With regard to the presence of a clonal marker, the JAK2V617F mutation was found in 40% of sporadic forms, while it was absent in all familial cases (Table 2). Furthermore, among the 18 female patients evaluated, hematopoiesis was clonal in 8 (all with sporadic thrombocythemia), while the other 10 patients appeared polyclonal (5 with sporadic and 5 with familial diseases; Table 2). Importantly, secondary causes of thrombocytosis listed in the proposed revised criteria10 were ruled out in all 32 patients. In 10 of 12 patients with familial thrombocythemia (4 of 5 investigated families), we found the MPLSer505Asn activating mutation, which is transmitted by dominant autosomal inheritance20 (Table 2). Nevertheless, all patients fulfilled the proposed criteria, independently of whether they had a sporadic or familial disorder (Table 2).

View this table:
Table 2

Incidence of myeloproliferative markers and hereditary thrombocythemia markers in 32 children with ET

The aim of this study was to assess the applicability of the proposed revised WHO criteria10 in childhood PV and ET. For this purpose, we applied the criteria to an unselected series of 45 children with PV and ET, diagnosed in accordance with the criteria in use at the time of their first evaluation.8,9 In more than one-third of cases, the disease appeared to be inherited. Although the exact prevalence of hereditary disorders in our patient population could be overestimated, since 17 patients with familial diseases belonged to 7 different families, their overall occurrence in pediatric patients is undoubtedly noteworthy. All children with familial disorders failed to exhibit JAK2 mutations, had a polyclonal hematopoiesis, and frequently showed the inherited MPLSer505Asn activating mutation. Nevertheless, 1 patient with familial PV and all patients with familial ET met the proposed revised WHO criteria.10 These observations suggest that the diagnostic screening of pediatric patients should first investigate whether other family members show similar hematologic abnormalities. In such patients, a careful screening for specific hereditary mutations should be performed: their detection would allow to exclude the presence of JAK2 mutations and to avoid more invasive diagnostic procedures such as a bone marrow biopsy. In particular, in order to prevent children with familial PV and ET from undergoing an inappropriate diagnostic process that may ultimately lead to an erroneous diagnosis of MPDs, guidelines for the diagnostic work-up of PV and ET in childhood should clearly state that in patients with a familial history of polycythemia and thrombocythemia hereditary mutations of Epo-R, TPO, or MPL must be excluded. Moreover, it should be recommended that all patients testing negative for these mutations be re-evaluated if new inheritable defects are described.

In our series of patients, the JAK2V617F mutation was found only in a proportion of children with sporadic MPDs. These findings confirm and expand similar results obtained in previous studies.1517 In addition, we have found that no exon 12 JAK2 mutations were detectable in JAK2V617F-negative patients with PV. We do not know if these patients will acquire JAK2 mutations later or if their disease originates from different pathogenetic mechanisms. Whatever may be the explanation, a significant proportion of childhood PV would nonetheless be misdiagnosed by using the proposed revised WHO criteria. In these cases, the presence of several alternative positive criteria, as listed in the currently used WHO criteria,9 should be considered. To this end, our findings indicate that the presence of a clonal hematopoiesis could represent a specific (albeit not sensitive) marker of MPD also in pediatric PV and ET.

Taken together, our observations suggest that childhood MPDs require a set of diagnostic criteria different from those proposed for adult MPDs. These appropriate diagnostic criteria should (1) exclude familial forms due to inherited molecular defects and (2) consider that pathogenetic alterations found in adult patients are detectable only in a minority of children. Otherwise, the application of diagnostic guidelines developed for adult patients to pediatric MPDs could, in the future, expose children to inappropriate molecularly tailored therapies.


Contribution: L.Teofili contributed to the study design, analyzed data, and wrote the manuscript. F. Giona, G.P., and A.A. enrolled patients and recorded clinical data. M.M., T.C., and F. Guidi. performed cell cultures and molecular analysis. L. Torti analyzed data. G.L. and R.F. critically reviewed the manuscript. L.M.L. designed the study and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Luigi M. Larocca, Istituto di Anatomia Patologica, Università Cattolica, Largo Gemelli 8, 00168 Roma, Italy; e-mail: llarocca{at}


This work was supported by Prin 2006, Ministero Università e Ricerca Scientifica (Rome, Italy) and by Fondi d'Ateneo, Progetti D1 2006-2007, Università Cattolica (Rome, Italy).


  • The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

  • Submitted June 6, 2007.
  • Accepted July 16, 2007.


View Abstract