Reproducibility of the WHO Histological Criteria for the Diagnosis of Philadelphia Chromosome-Negative Myeloproliferative Neoplasms

Umberto Gianelli; Anna Bossi; Ivan Cortinovis; Elena Sabattini; Claudio Tripodo; Emanuela Boveri; Alessia Moro; Riccardo Valli; Maurilio Ponzoni; Ada M Florena; Giulio F Orcioni; Stefano Ascani; Emanuela Bonoldi; Alessandra Iurlo; Luigi Gugliotta; Vito Franco


Mod Pathol. 2014;27(6):814-822. 

In This Article

Abstract and Introduction


This study, performed on behalf of the Italian Registry of Thrombocythaemias (Registro Italiano Trombocitemie), aimed to test the inter-observer reproducibility of the histological parameters proposed by the WHO classification for the diagnosis of the Philadelphia chromosome-negative myeloproliferative neoplasms. A series of 103 bone marrow biopsy samples of Philadelphia chromosome-negative myeloproliferative neoplasms consecutively collected in 2004 were classified according to the WHO criteria as follows: essential thrombocythaemia (n = 34), primary myelofibrosis (n = 44) and polycythaemia vera (n = 25). Two independent groups of pathologists reviewed the bone marrow biopsies. The first group was asked to reach a collegial ‘consensus’ diagnosis. The second group reviewed individually all the cases to recognize the main morphological parameters indicated by the WHO classification and report their results in a database. They were subsequently instructed to individually build a ‘personal’ diagnosis of myeloproliferative neoplasms subtype just assembling the parameters collected in the database. Our results indicate that high levels of agreement (≥70%) have been reached for about all of the morphological features. Moreover, among the 18 evaluated histological features, 11 resulted statistically more useful for the differential diagnosis among the different Philadelphia chromosome-negative myeloproliferative neoplasms. Finally, we found a high percentage of agreement (76%) between the ‘personal’ and ‘consensus’ diagnosis (Cohen’s kappa statistic >0.40). In conclusion, our results support the use of the histological criteria proposed by the WHO classification for the Philadelphia chromosome-negative myeloproliferative neoplasms to ensure a more precise and early diagnosis for these patients.


Since the introduction of the WHO classification criteria for the diagnosis of Philadelphia chromosome-negative chronic myeloproliferative disorders in 2001, much emphasis has been put on the bone marrow histological analysis as an irreplaceable tool for the diagnosis of these neoplasms.[1] Such criteria, stemming from the seminal work of leading experts,[2–5] established the boundaries among the different pathological entities classified as Philadelphia chromosome-negative chronic myeloproliferative disorders, providing indications on the morphological features to be associated with clinical and laboratory data in order to achieve a definite and reproducible diagnosis. The WHO classification criteria, revised in 2008, by means of incorporating recent molecular achievements, underscored the importance of an integrated approach in which the fine morphology has a central role.[6,7] Prototypical is the example of megakaryocytes, whose degree of atypia, proliferation and aggregation, have been considered the pillars of the differential diagnosis of these disorders. On the other hand, the frequent implementation of blood counts in routine clinical practice implies an increasingly early diagnosis and certainly contributes to the occurrence of Philadelphia chromosome-negative myeloproliferative neoplasms cases with markedly overlapping clinical and pathological features, which could hamper a clear-cut classification. These cases, regarded as ‘unclassifiable’ in the WHO classification, account for 10–15% of all myeloproliferative neoplasms and constitute a diagnostic ‘grey zone’ virtually linking all the Philadelphia chromosome- negative myeloproliferative neoplasms. Most of the diagnosis of myeloproliferative neoplasms ‘unclassifiable’ derive by the impossibility to reach a conclusive diagnosis in the presence of clinical or morphological features that, according to the WHO classification, are borderline with different myeloproliferative neoplasms. Recently, the reproducibility of histopathology of bone marrow according to WHO classification and its usefulness for identifying different myeloproliferative neoplasms has been questioned, [8–10] and subsequent controversies between supporters and opponents have been raised in this setting.

Provided that the correct classification of myeloproliferative disorders integrates clinical, molecular and pathological data, in this study, through the morphological analysis of bone marrow biopsy samples of a large series of Philadelphia chromosome-negative chronic myeloproliferative neoplasms at diagnosis, we aimed to test the inter-observer reproducibility of the histological parameters indicated in the WHO classification and propose a rational approach to the analysis of these variables in the differential diagnosis of myeloproliferative neoplasms.