Abstract / Summary
The management of patients with essential thrombocythemia (ET) and extreme thrombocytosis (ExT) remains plagued by several unresolved issues. Therefore, we conducted a web-based survey to better understand hematologists’ clinical approaches to ExT and to compare them with responses generated by artificial intelligence (AI) programs. A total of 56 hematologists participated. Cytoreductive treatment was preferred by 41.1% of respondents at a platelet threshold of ≥ 1000 × 10⁹/L and by 83.9% at ≥ 1500 × 10⁹/L. Assessment of vWF: RCo at diagnosis was reported by 78.6% of hematologists. In the presence of cardiovascular risk factors and the absence of acquired von Willebrand disease (vWD), aspirin therapy was favored by 91.1% and 78.6% of respondents for platelet counts ≥ 1000 × 10⁹/L and ≥ 1500 × 10⁹/L, respectively. Prevention of thrombosis, prevention of bleeding, and symptom control were cited with similar frequencies as reasons for initiating cytoreductive treatment. Artificial intelligence (AI)-generated recommendations supported vWF testing, favored cytoreductive treatment, particularly at platelet counts ≥ 1500 × 10⁹/L, and generally recommended aspirin only for patients with platelet counts ≥ 1000 × 10⁹/L in the absence of acquired vWD. Overall, ExT appeared to be an important driver of treatment decisions, particularly at platelet counts ≥ 1500 × 10⁹/L. Comparisons with AI-generated recommendations revealed few differences in treatment approaches across some clinical scenarios, supporting the exploratory nature of the AI analysis.