Reporting of bone marrow biopsies to accurately classify myeloproliferative neoplasms significantly varies in routine practice, with many reports omitting relevant information and ancillary test results, an Australian study suggests.

Led by Australasian Leukaemia & Lymphoma Group (ALLG) members Dr Cecily Forsyth, Professor Andrew Grigg and Associate Professor David Ross, the MPN01 trial highlights the need for careful review of pathology to improve the reliability of a myeloproliferative neoplasms (MPN) diagnosis.

The researchers audited diagnostic information from 152 patients from eight participating sites enrolled in a national MPN registry of the ALLG between 2010 and 2016 and who had sufficient pathological data available for review.

Findings published in Pathology [link here] showed that bone marrow biopsies were performed in 74% of cases, however only two-thirds of these patients had pathological information that was deemed concordant with the stated diagnosis.

This was based on whether the doctors had documented and described accurately key diagnostic features present in the revised 2016 WHO Classification.

The authors said the main reasons for discordant results for the remaining third of biopsied patients were related to incomplete descriptions of megakaryocyte topography and morphology, inconsistent grading of reticulin fibrosis, and failure to integrate the available morphological and ancillary clinicopathological information.

This meant the specific MPN subtype might not have been reported correctly in 33% of cases, the authors noted, validating calls for a more standardised approach to bone marrow reporting and consistent reporting to cancer registries.

They identified that bone marrow biopsies were most frequent among patients with primary myelofibrosis (91%) and secondary myelofibrosis (100%), less so in unclassifiable MPN (76%) and essential thrombocythaemia (74%), and least frequently performed in polycythaemia vera (46%).

However, the researchers noted a trend for an increasing proportion of polycythaemia vera patients to undergo a biopsy if their diagnosis was made in the later years of the registry, which they said might have related to changes in diagnostic recommendations.

There was no significant change over the study period for other MPN subtypes.

Of the 40 patients to not undergo a biopsy, 26 would have been deemed necessary to fulfil the 2016 WHO criteria, the researchers added.

They were less likely to have a biopsy if a driver mutation had been identified.

“Our audit of the pathology information entered in a national MPN registry reveals deficiencies that should be addressed to meet international diagnostic standards and to improve clinical practice,” the authors wrote.

“Important elements of the bone marrow report were missing from a substantial proportion of MPN cases, and there was an inconsistent approach to semi-quantitative descriptors, such as cellularity and fibrosis.

“The growth in digital morphology has made it possible for bone marrow trephines to be included in external quality assurance programs and could provide a useful tool to improve consistency of bone marrow reporting in MPN.”

Novartis, Sanofi and Shire Australia Pty Limited provided funding for the studies, and Novartis and Sanofi reviewed the final manuscript with no editorial changes.