Long-term hydroxyurea therapy alters bone marrow megakaryocyte morphology and associates with prognosis in myeloproliferative neoplasms

Hydroxyurea (HU) is a first-line treatment for myeloproliferative neoplasms (MPNs), but a quantitative, standardized framework for assessing its impact on megakaryocyte (MK) morphology and prognosis remains lacking. This retrospective, two-center study analyzed 22,097 MKs from 280 patients (118 polycythemia vera [PV], 162 essential thrombocythemia [ET]), classified into 12 subtypes using the 5th edition WHO classification in conjunction with objectively derived thresholds (minimum P-value approach). Our primary contribution is to establish a quantifiable scoring framework that translates descriptive WHO categories into reproducible prognostic metrics. Using this framework, HU significantly reduced normal MK proportion in JAK2-ET ( P = 0.0189), TN-ET ( P = 0.0233), and JAK2-PV ( P = 0.0427), while increasing abnormal subtypes (bulbous, separated nuclear MKs); conversely, IFN was associated with preserved normal MKs and reduced abnormal forms. Multivariate Cox regression, applied to our empirically derived cutoffs, identified protective thresholds (normal MK ≥ 58% in ET, ≥ 53% in PV; emperipolesis ≥ 2% in ET) and adverse cutoffs (Naked nuclei ≥ 13% in ET, ≥ 7% in PV; staghorn like ≥ 6% in ET; separated nuclear ≥ 3% in ET) for myelofibrosis progression (all P < 0.05). Our quantitative findings corroborate prior qualitative observations that long-term HU is associated with adverse MK remodeling, while IFN correlates with lower progression risk as captured by our threshold-based system. This standardized scoring approach—not the identification of new prognostic factors—provides a reproducible tool for monitoring treatment-related morphological changes and may inform risk-adapted therapy in MPN patients.