Evaluation of HU-resistance in Adult Patients With Polycythemia Vera Who Meet PV-AIM Predictors
Key claim: This study aims to confirm that hemoglobin <15.5 g/dL and RDW ≥17% are real-world predictors of hydroxyurea failure in polycythemia vera patients, as originally identified by the PV-AIM machine learning model.
Abstract
The purpose of this study is to confirm the predictive factors for hydroxyurea (HU) failure (hemoglobin (HGB) <15.5 g/dL (9.62 mmol/L) and red blood cell distribution width (RDW) ≥17%) identified by machine learning in the polycythemia vera advanced integrated model (PV-AIM) project in the real-life setting.
Why this matters
Hydroxyurea resistance/intolerance is a key inflection point in PV natural history, triggering switches to ruxolitinib or interferon and affecting long-term thrombosis and transformation risk. Validating simple, routinely available CBC-derived predictors (HGB <15.5 g/dL, RDW ≥17%) from the PV-AIM model would give clinicians a pragmatic tool to anticipate HU failure before clinical resistance is declared, complementing emerging biomarker-guided approaches in this subtype.