2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Saturday, December 12, 2026
05:30 p.m. - 07:30 p.m.
MCCNO - Hall F
Outstanding mix of clinical abstracts, including clinical trials, retrospective analysis, and molecular observations in myelofibrosis, polycythemia vera, essential thrombocythemia and other myeloid neoplasms
Access, Representation, Population Health, Technology and Procedures, Molecular Testing, Omics Technologies, Pathology, Treatment Considerations, Combination Therapy, Drug Development, Therapy Sequence, Biological Therapies, Transplantation, Immunotherapy, Monoclonal Antibody Therapy, Non-Biological Therapies, Pharmacology, Women’s Health, Adverse Events, Artificial Intelligence (AI), Machine Learning, Biological Processes, Pathogenesis, Genomics, Hematopoiesis, Molecular Biology, Clinical Practice (Health Services and Quality), Patient Safety, Diseases, Bleeding and Clotting, Thromboembolism, Thrombotic Disorders, Myeloid Malignancies, Acute Myeloid Malignancies, Chronic Myeloid Malignancies, Research, AML, Translational Research, MPN, Clinical Research, Clinical Trials, Epidemiology, Health Outcomes Research, Patient-Reported Outcomes, Real-World Evidence, Registries, Survivorship, Study Population, Human, Adult, Elderly, Pediatric
PAPER

2048 Real-world hematocrit control in polycythemia vera: Longitudinal trajectories from 16,230 patients

Fahad Zamir, Fariha Zamir2, Ghaith Abu-Zeinah I, MD3
1 Weill Cornell Medicine - Qatar, Doha, Qatar
2 Weill Cornell Medicine - Qatar, Doha, Qatar
3 Weill Cornell Medicine, New York, New York, United States

Introduction: Polycythemia vera (PV) is a myeloproliferative neoplasm characterized by JAK2-mutated clonal erythrocytosis and increased thrombosis risk. Hematocrit (HCT) control <45% is a central modifiable determinant of thrombotic risk and mortality, yet nearly half of patients (pts) with PV have uncontrolled HCT at any time according to the REVEAL study. Longitudinal patterns of HCT control and the clinical characteristics associated with them remain poorly understood. In this study, we used unsupervised trajectory analysis to identify distinct HCT control patterns providing insight into the timing, durability or lack of HCT control and the factors associated with them.

Methods: Using the TriNetX electronic health record database, we conducted a retrospective cohort study of pts with PV. Pts were identified using ICD-10 code D45, with the earliest D45 date defining the baseline diagnosis. To confirm active PV at cohort entry and exclude pts likely diagnosed and treated before entering the data source, we required at least one HCT above the diagnostic threshold (>48% in women, >49% in men) within 1 year of baseline. All HCT values from baseline to last follow-up were included, except if measurements were after documented progression to myelofibrosis or acute myeloid leukemia. Demographics, clinical characteristics and cytoreductive drug (cytorx) use was collected. Pts were classified as high-risk if age is ≥60 years at baseline or with prior thrombosis, and as low-risk if age <60 years at baseline without prior thrombosis. We assessed HCT control (<45%) both at the individual patient level and at the population level using aggregated measurements. Distinct longitudinal HCT control patterns were identified using k-means clustering (k=4) of mean-normalized 3-month HCT trajectories over 5 years among pts contributing at least three time bins. Groups were compared using Fisher’s test for categorical variables and Welch’s t-test for means.

Results: 16,230 pts with PV met criteria, contributing 395,586 HCT measurements over a median follow-up of 2.0 years (range 0-25: only 6% had no follow-up). The majority were high-risk (61%), male (68%) and white (75%). Compared to low-risk pts, high-risk pts were older (mean age, 69 vs 46 years), with prior thrombosis (24% vs 0%), hypertension (56% vs 35%), and hyperlipidemia (48% vs 25%), all p<0.001. Mean baseline HCT values were similar at 52% (±5%) but longitudinal control was consistently worse among low-risk pts. HCT values ≥45% accounted for 56% of measurements in low-risk pts compared with 47% in high-risk pts (p<0.001). At the pt level, only 21% of low-risk pts versus 33% of high-risk pts had HCT control for ≥50% of measurements. Trajectory clustering showed the same 4 patterns in low and high risk but with different proportions (p<0.001): No control (NoC, 73% vs 61%), rapid sustained control (RSC, 13% vs 21%), early control with rebound (ECwR 8% vs 10%), and delayed control (DeC, 5% vs 8%). Most notable when comparing trajectories for high-risk pts: NoC were most likely male (70%) and RSC least likely (48%); NoC had similar rate of prior thrombosis (p=0.2, 24%), and higher hypertension, diabetes and hyperlipidemia; NoC were least likely to receive a cytorx (20% hydroxyurea (HU), <5% other) unlike RSC and ECwR (55-56% HU, 16-20% other). Similar trends for low-risk pts were seen with regards to sex and cytorx (10% HU, 6% other for NoC; 30-31% HU, 15-20% other for RSC, ECwR and DeC).

Conclusions: In this large real-world cohort of pts with PV, approximately half of all HCT measurements and the large majority of pts were not meeting the recommended target HCT <45%, and only a third or less maintained HCT control for majority of measurements. Worse control trajectories were associated with male pts, cardiovascular comorbidities, and less cytorx use. Surprisingly, NoC was the most common trajectory even in high-risk pts, and those with prior thrombosis, reflecting non-adherence to life-prolonging treatment standards. Low-risk pts had consistently worse HCT control than high-risk pts across measurement-level, pt-level, and trajectories. These findings highlight a large gap between evidence-based recommendations and real-world implementation of HCT control. Results support closer monitoring and treatment optimization of all pts with PV at a large scale, irrespective of risk category. Treatment-pattern analyses are ongoing.

Fahad Zamir: No relevant conflicts of interest to declare.