1153 - SWITCHING FROM COVALENT BTKI TO BCL2i IS ASSOCIATED WITH IMPROVED CLINICAL OUTCOMES COMPARED TO SWITCHING TO A DIFFERENT COVALENT BTKI IN PATIENTS WITH CLL/SLL TREATED IN THE REAL-WORLD SETTING

Introduction:
Targeted agents, such as covalent BTKi- (cBTKi; i.e., acalabrutinib, ibrutinib, zanubrutinib) and BCL2i- based regimens (i.e., venetoclax), have demonstrated improved treatment efficacy and safety compared to CT/CIT for patients with CLL/SLL. However, some patients discontinue for reasons other than disease progression (PD) and switch to targeted agent with the same or different mechanism of action (MOA). This study assessed real-world treatment outcomes of patients with CLL/SLL who switched to a cBTKi- or BCL2i-based regimen following discontinuation of a cBTKi-based regimen.

Methods:
Data from the CLL Collaborative Study of Real-World Evidence (CORE), an international, retrospective observational study involving 23 centers, were included in this study. Adult BCL2i-naïve patients diagnosed with CLL/SLL were included if they discontinued their cBTKi-based regimen for reasons other than PD or completion of therapy and switched to a cBTKi or BCL2i-based regimen. Based on switching patterns, patients were classified into 2 mutually exclusive cohorts: cBTKi (i.e., cBTKi-cBTKi) or BCL2i (i.e., cBTKi-BCL2i). Treatment outcomes of the studied therapy included overall response (OR), progression-free survival (PFS), and time to next treatment or death (TTNT-D). OR was defined as complete or partial response. PFS was defined as the time from cBTKi or BCL2i initiation to PD/death (event), censoring at last follow-up date. TTNT-D was defined as the time from cBTKi or BCL2i initiation to initiation of next treatment/death (event), censoring at last follow-up date. Regression models, controlling for age at cBTKi or BCL2i initiation, sex, time from CLL/SLL diagnosis to cBTKi or BCL2i initiation, duration of discontinued cBTKi therapy, ECOG, and lactate dehydrogenase (LDH), assessed the odds of achieving an OR via a logistic regression and the time to PD/death and time to next treatment/death via a Cox proportional hazards (PH) regression.

Results:
A total of 121 BCL2i-naïve patients who discontinued their first cBTKi-based regimen in any line of therapy and switched to a subsequent treatment with a cBTKi- (cBTKi: 44 [36.4%]) or a BCL2i-based regimen (BCL2i: 77 [63.6%]) were included in the study. The most common subsequent regimen following cBTKi discontinuation was a second-generation cBTKi regimen (34/44 [77.3%]; acalabrutinib [30/34], zanubrutinib [4/34]) or BCL2i (venetoclax monotherapy: 43/77 [55.8%] and venetoclax+anti-CD20: 33/77 [42.9%]). Most common reason for prior cBTKi discontinuation was intolerance (cBTKi: 79.6%, BCL2i: 80.5%). The majority of patient characteristics were not significantly different between cohorts: median age at cBTKi or BCL2i initiation (cBTKi: 71.5 years, BCL2i: 69.0 years), gender (male: cBTKi: 63.6%, BCL2i: 68.8%), unmutated IGHV (cBTKi: 47.7%, BCL2i: 37.7%), and with del(17p)/TP53 mutation (cBTKi: 20.5%, BCL2i: 27.3%); however, there were some significant differences: median time from CLL/SLL diagnosis to cBTKi or BCL2i initiation (BTKi: 91.0 months, BCL2i: 62.6 months), median duration of discontinued-cBTKi (cBTKi: 19.6 months, BCL2i: 10.8 months), ECOG Performance Status 0 (cBTKi: 50.0%, BCL2i: 27.3%), and elevated LDH (cBTKi: 29.5%, BCL2i: 46.8%; Table). Median follow-up duration was 11.6 months for cBTKi and 16.4 months for BCL2i. Among 79 patients with recorded response, the OR was higher for BCL2i than cBTKi (cBTKi: 62.5%, BCL2i: 83.6%). The odds of achieving a response were significantly higher for BCL2i than cBTKi (unadjusted: 3.1 times [CI: 1.0; 9.1]; adjusted: 4.6 times [CI: 1.2; 18.0]). The PFS KM estimates were higher for BCL2i in relation to cBTKi at 12 months (cBTKi: 79.1%, BCL2i: 87.3%) and 24 months (cBTKi: 73.0%, BCL2i: 78.0%). Cox PH results revealed that treatment with BCL2i reduced the hazard of progression/death by 70% (unadjusted HR: 0.6 [CI: 0.3; 1.5]; adjusted HR: 0.3 [CI: 0.1; 0.8], p=.023) compared to cBTKi. The results were similar for TTNT-D (KM estimates at 12 months: cBTKi: 81.1%, BCL2i: 84.7%; at 24 months: cBTKi: 69.6%, BCL2i: 71.1%; unadjusted HR: 0.8 [CI: 0.4; 1.9]; adjusted HR: 0.4 [CI: 0.1; 1.0], p=.055).

Conclusion:
Patients who switched to a BCL2i-based regimen were more likely to respond to therapy and had lower hazard of progression/death; TTNT-D results also suggested a possible benefit of BCL2i over cBTKi. Switching to BCL2i-based regimens post-cBTKi demonstrates better treatment outcomes compared to cycling through consecutive cBTKis, further highlighting the effectiveness of BCL2i in real-world settings. Given the dynamic treatment landscape in CLL, the impact of switching to agent(s) with a different MOA or choosing to retreat with agent(s) in the same class are key considerations in optimizing clinical care.