Abstract A161: Bridging clinical trials and real-world practice: A meta-analysis of ponatinib safety in chronic myeloid leukemia

Introduction: Ponatinib [JC1] is a tyrosine kinase inhibitor with activity against all ABL kinase domain mutations including T315I, making it essential therapy for resistant chronic myeloid leukemia (CML). Its efficacy is tempered by arterial occlusive events (AOEs), the principal barrier to long-term use. Whether the safety profile established in the pivotal PACE trial reflects observations in routine clinical practice remains unclear. We performed a systematic review and meta-analysis comparing ponatinib safety for CML across clinical trial and real-world settings. Methods: We synthesized adverse event (AE) data from the PACE trial, its long-term follow-up, and five real-world cohorts from Europe, North America, and Asia, restricting all analyses to CML-specific patient subsets. Random-effects meta-analysis estimated pooled AE incidence; prespecified subgroup analyses examined heterogeneity by follow-up duration, AE ascertainment method, starting dose, and study design. Results: Six reports [JC2] comprising 995 CML patients were included; Philadelphia chromosome-positive ALL patients were excluded from this CML-specific analysis. There was substantial heterogeneity regarding rates of AOEs and other key toxicities (I2=63–93%), driven primarily by follow-up duration (median range12–56.8 months[JC3] across reports) and AE ascertainment method. Crude pooled AOE incidence was 2.8% (95% CI 0.3–7.8%; I2=82.7%); exposure-adjusted analysis showed AOE incidence fell from 15.8 to 4.9 events per 100 person-years as dose intensity declined from 32.1 to 20.4 mg/day, a pattern consistent with the dose-dependent AOE reduction reported in Japanese post-marketing data (6.8 per 100 person-years[JC4] ). AOE incidence was consistently lower across real-world cohorts with individualized, variable-dose protocols (0–7.3%; Italy, USA, Japan) than in the fixed 45-mg registration trial (31.1%). Hypertension (OR 2.14) and diabetes (OR 1.95) independently predicted AOE risk; prior nilotinib exposure was associated with subsequent AOE occurrence (11% vs. 0%, age-adjusted p=0.04). Conclusions: Real-world ponatinib safety data for CML diverge from clinical trial reporting due to differences in follow-up duration and AE ascertainment, limiting direct comparability. Dose optimization meaningfully reduced AOE incidence in both trial and real-world settings. Lower AOE incidence in cohorts with individualized dose-titration suggests that access to proactive monitoring, not drug biology alone, may modify toxicity burden across health systems. Standardized, harmonized safety reporting across future ponatinib studies is needed to translate trial-based safety data into actionable, equitable real-world clinical guidance.Generative AI assisted in editing this abstract's text; all scientific content and conclusions were reviewed and approved by the authors. Citation Format: Michael Phan, Michael Sketel, Marisol Miranda-Galvis, Josh Wilson, Jorge Cortes. Bridging clinical trials and real-world practice: A meta-analysis of ponatinib safety in chronic myeloid leukemia [abstract]. In: Proceedings of the 19th AACR Conference on the Science of Cancer Health Disparities; 2026 Oct 18-21; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2026;35(10 Suppl):Abstract nr A161.