Cookies on this website
We use cookies to ensure that we give you the best experience on our website. If you click 'Continue' we'll assume that you are happy to receive all cookies and you won't see this message again. Click 'Find out more' for information on how to change your cookie settings.

Abstract Myeloproliferative neoplasms (MPNs) are clonal disorders characterized by excessive proliferation of myeloid lineages. Accurate classification and appropriate management of MPNs requires integration of clinical, morphological and genetic findings. Despite major advances in understanding the molecular and genetic basis, morphological assessment of the bone marrow trephine (BMT) remains paramount in differentiating between MPN subtypes and reactive conditions. However, morphological assessment is heavily constrained by a reliance on subjective, qualitative and poorly reproducible criteria. To address this, we have developed a machine-learning strategy for the automated identification and quantitative analysis of megakaryocyte morphology using clinical BMT samples. Using a sample cohort of recently diagnosed or established ET (n = 48) and reactive control cases (n = 42) we demonstrated a high predictive accuracy (AUC = 0.95) of automated tissue ET diagnosis based upon these specific megakaryocyte phenotypes. These separate morphological phenotypes showed evidence of specific genotype associations, which offers promise that an automated cell phenotyping approach may be of clinical diagnostic utility as an adjunct to standard genetic and molecular tests. This has great potential to assist in the routine assessment of newly diagnosed or suspected MPN patients and those undergoing treatment / clinical follow-up. The extraction of quantitative morphological data from BMT sections will also have value in the assessment of new therapeutic strategies directed towards the bone marrow microenvironment and can provide clinicians and researchers with objective, quantitative data without significant demands upon current routine specimen workflows.

Original publication

DOI

10.1101/762013

Type

Journal article

Publication Date

11/09/2019