Blood Adv. 2026 Jun 23:bloodadvances.2026020490. doi: 10.1182/bloodadvances.2026020490. Online ahead of print.
ABSTRACT
As the life expectancy of individuals with hemophilia increases, arterial thrombotic disease has emerged as an important clinical issue. Accurate risk stratification for primary prevention is increasingly critical, since managing arterial events is complex in hemophilia owing to bleeding risk. However, current prospective data on event incidence and risk prediction for hemophilia remain limited. We analyzed 5-year follow-up data from ADVANCE Japan-a nationwide prospective registry of Japanese adults aged ≥40 years with hemophilia. Participants with prior arterial thrombosis were excluded. Five-year arterial event risk was estimated using the Hisayama model, the Suita score, and QRISK3. Model performance was evaluated using sensitivity, specificity, and observed-to-expected event comparisons. Among the 599 participants (median age: 59 years; 83% hemophilia A, 63% severe disease), 28 first arterial events occurred over 5 years (4.67%). The sensitivity and specificity for identifying events were 70.5% and 68.0% for the Hisayama model, 73.3% and 61.2% for the Suita score, and 71.4% and 57.7% for QRISK3, respectively. The observed incidence was 2.93 events per 100 participants over 5 years, closely matching predictions by the Hisayama model and Suita score, whereas QRISK3 substantially overestimated the risk. Incorporating hemophilia-related clinical factors did not improve predictive performance beyond established scores. Arterial event incidence in Japanese adults with hemophilia was comparable to that expected in the Japanese general population. The Japanese population-based risk models-the Hisayama model-may support the development of primary prevention strategies for Japanese adults with hemophilia. This trial was registered with the Japan Registry of Clinical Trials (jRCT1030190046).
PMID:42335204 | DOI:10.1182/bloodadvances.2026020490