Real-world comparison of treatment outcomes between BCL2i and 2nd generation BTKi therapy in first-line CLL patients

Abstract Background: First-line (1L) therapies for chronic lymphocytic leukemia (CLL) consist of B-cell lymphoma inhibitors (BCL2i) or Bruton tyrosine kinase inhibitors (BTKi). Current recommendations include second-generation covalent BTKi-based regimens, including acalabrutinib or zanubrutinib, or BCL2i-containing regimens, such as venetoclax with obinutuzumab (V+O). Historically, treatment effectiveness between classes is comparable, where treatment selection may be determined by certain risk factors, such as age or TP53 alterations. Real-world evidence on clinical outcomes in these data is limited, particularly among high-risk patients (pts). This work assesses the difference in time to next treatment (TTNT) among CLL pts receiving second-generation BTKi or V+O in the 1L setting, additionally stratified for risk factors such as age at treatment initiation, and genetic mutations (TP53 alterations and 17p deletion). Methods: The IntegraConnect PrecisionQ de-identified electronic health record (EHR) database, which includes data on over 2 million cancer pts across more than 500 care sites, was used to identify individuals with CLL who received second-generation BTKi or V+O as 1L therapy betweenJanuary 1, 2021 to February 28, 2025. The index date was the initiation of 1L therapy. Descriptive statistics were used to summarize demographic and clinical characteristics by treatment group. Wilcoxon rank sum tests and chi squared tests were performed to assess differences in continuous and categorical factors, respectively. A subset of patients whose data were manually abstracted from clinical documents from the E.H.R. were analyzed to assess 17p deletion or TP53 mutation status. Differences in TTNT by therapy were analyzed using Kaplan-Meier (KM) survival curves, with median estimates reported. Multivariable Cox proportional hazards regression was used to assess the impact of treatment on 1L TTNT, adjusted for age at treatment initiation, ECOG performance status, payer, and the region of the treatment facility, and hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) are reported. Among abstracted cases, KM curves and multivariable Cox models, adjusted for the same factors, were stratified by the TP53 or 17p deletion alteration status, to assess the potential influence on TTNT. Results: Overall, 3,844pts met study eligibility. The majority of pts received second-generation BTKi (n=2,903, 76%; V+O n=941, 24%). Among those who received BTKi, 2,338 (81%) were treated with BTKi monotherapy, and the remaining 565 (19%) received BTKi combination therapy, with the majority of those pts having received obinutuzumab (n = 302). Pts treated with BTKi were significantly older than those treated with V+O at index start (median age 74 vs 69, p<0.01). Additionally, there was a higher proportion of patients on Medicare/Medicaid and pts with an ECOG status of 2 or more among those treated with BTKi compared to V+O (Medicare/Medicaid 46% vs 39%, p<0.01; ECOG 2+ 11% vs 5%, p<0.01). There was a higher proportion of patients treated in the northeast region of the US among BTKi users compared to V+O (24% vs 17%, p<0.01). Median TTNT was not reached for either treatment group, but at 30 months, the event-free probability was 70% among pts that received second-generation BTKi, compared to 83% for V+O (KM logrank p<0.01). Pts that received V+O were over 40% less likely to initiate subsequent treatment than those that received BTKi (HR: 0.56, 95% CI: 0.46-0.69, p<0.01). There were 1,132 pts in the curated subset (BTKi n=858, 76%; V+O n=274, 24%). There was a significantly higher proportion of pts with a TP53 mutation or 17p deletion that received BTKi compared to V+O (18% vs 12%, p<0.05). Among those with TP53 mutation or 17p deletion, the difference in TTNT between BTKi and V+O was attenuated (HR: 1.33, 95% CI: 0.66-2.70, p=0.43). However, for those without TP53 mutations or 17p deletion, pts that received V+O were less likely to initiate next treatment at any given point in time (HR: 0.70, 95% CI: 0.50-1.00, p=0.05). Conclusions: In this real-world study analyzing CLL pts, those that received 1L V+O had significantly longer TTNT when compared to pts that received second-generation BTKi. However, this difference was not observed among a subset of pts with 17p deletions or TP53 mutations.