Search for a command to run...
Abstract Background: Second-generation Bruton's tyrosine kinase inhibitors (BTKis), including acalabrutinib (acala) and zanubrutinib (zanu), improve upon the effectiveness and safety of first-generation ibrutinib in the treatment of Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (CLL/SLL). While both acala and zanu are approved for the treatment of CLL/SLL, emerging evidence suggests differences in their cardiovascular (CV) safety profiles. Utilizing a propensity score weighting approach, this study evaluated the real-world incidence of treatment-emergent (TE) CV events in patients (pts) receiving acala or zanu monotherapy in first-line (1L) and relapsed/refractory (RR) settings. Interim results are described here. Methods: The IntegraConnect PrecisionQ Database, which contains electronic health records from 3 million deidentified pts in the United States, was used to construct a retrospective cohort of CLL/SLL pts who initiated therapy on a second-generation BTKi therapy between 01JAN2022 and 31DEC2024. Patients were followed through 31MAR2025. ICD-10 codes were used to identify CV events including atrial fibrillation, atrial flutter, and arrhythmia (Afib) and hypertension. TE incidence was considered when a specified condition occurred during therapy. CV conditions occurring during therapy were not counted as incident if patients had the same condition in the two years prior to BTKi initiation. To adjust for differences in baseline demographics and clinical characteristics between pts receiving acala and zanu, standardized mortality ratio (SMR) weighting was applied using propensity scores derived from a logistic regression model. This model estimated the likelihood of acala treatment based on age, Eastern Cooperative Oncology Group (ECOG) performance status, comorbidities, prior B-cell lymphoma 2 (BCL-2) inhibitor use, prior chemoimmunotherapy, year of treatment initiation, and line of therapy (LOT). A standardized mean difference (SMD) <|0.1| was used to define good balance between groups. Exposure-adjusted incidence rates (EAIR), reported per 100 person-months; and weighted hazard ratio (HR) and 95% confidence intervals (CI) from Cox proportional hazard models were used to compare the TE incidence rates in the acala and zanu treatment groups. Results:The unweighted (weighted) sample included 1347 (1347) acala and 989 (1237) zanu pts, including 1157 (1157) acala and 803 (1064) zanu pts in the 1L and 190 (190) acala and 186 (173) zanu pts in second-line setting or later (2L+). After weighting, pts were balanced on all factors except year of BTKi initiation. The median age (in years) at initiation was 75 for acala and 74 for zanu. The percentage of acala and zanu pts with an ECOG score of 2+ was 7.6% and 8.4%, respectively. Baseline hypertension (9.8% for acala and 9.7% for zanu), bleeding and other cardiac risk factor conditions (2.5% for acala and 2.6% for zanu), prior BCL-2 inhibitor use (2.8% for acala and 2.4% for zanu), and prior chemoimmunotherapy use (13.7% for both) were similar between groups. The median (interquartile range [IQR]) follow-up was 16.5 (7.3, 26.8) months (mo) for acala and 16.2 (7.7, 22.0) mo for zanu. The overall incidence of TE CV events was higher for zanu vs. acala (12.0% vs. 9.4%, p<0.05). The EAIR was 1.5-fold higher for acala vs. zanu (0.85 vs 0.58 per 100 person-months) with a weighted HR of 0.73 (95% confidence interval [CI] 0.57, 0.92). The incidence of TE Afib was higher for zanu than acala (4.2% vs. 2.3%, p<0.01), with a 2-fold higher EAIR (0.28 vs. 0.14) and HR of 0.52 (95% CI 0.33, 0.81). When stratified by LOT, the incidence of TE Afib was higher for zanu than acala in 1L (3.7% vs. 2.1%, p<0.05) with a 2-fold higher EAIR and HR of 0.54 (95% CI 0.32, 0.90) and 2L+ (7.7% vs. 3.7%, p=0.09) with a 2.6-fold higher EAIR and HR of 0.45 (95% CI 0.18, 1.12). The overall incidence of TE CV events was higher for zanu than acala in 2L+ (17.9% vs. 10.5%, p<0.05), with a 2-fold higher EAIR and HR of 0.50 (95% CI 0.28, 0.88). Conclusions: These real-world analyses show TE CV events were statistically higher for zanu than acala - specifically for atrial fibrillation, atrial flutter, or arrhythmia, the EAIR was 2-fold higher in the zanu group compared to acala. CV events could lead to treatment interruptions and discontinuation, ultimately limiting clinical benefit and potentially increasing total cost of care. Further studies are warranted.