Paper Details

ABSTRACT

Introduction: BCL2 G101V was the first on-target venetoclax resistance mutation described in patients with CLL (1). Other BCL2 mutations were later recognized including D103Y/E/V, R107_R110dup, A113G, R129L and V156D in patients treated with continuous venetoclax therapy (2).  Herein, we analyzed the spectrum of BCL2 mutations identified in samples collected over the last 5 years to update the BCL2 mutation landscape, with a focus on those arising after fixed-duration regimens.

Methods: We performed a retrospective analysis of consecutive CLL samples with detected BCL2 mutations referred to Molecular Haematology at Peter MacCallum Cancer Centre (PMCC, June 2021 to June 2026) for testing from patients with CLL being managed in routine practice. Targeted panel sequencing of all coding exons of BCL2 in addition to genes recurrently mutated in hematological malignancy was performed using library preparation including unique molecular indexes and duplex sequencing (sensitivity 1% variant allele frequency, VAF). Selected BCL2 variants were functionally evaluated with pooled prime editor screens, installing lesions within the endogenous BCL2 locus in venetoclax-sensitive cell lines. 

Results: BCL2 mutations were detected in samples from 43 patients (18 with >1 variant). The median aggregate VAF of detected BCL2 mutations was 5% (range, 1 to 47%). Fourteen patients had progressive disease on continuous venetoclax monotherapy (C-Ven), 3 patients progressed during or <3 months after fixed-duration venetoclax (grouped together with C-Ven), and 26 experienced relapse >= 3 months after fixed-duration venetoclax (FD-Ven). Among patients who received FD-Ven, 24 received rituximab-venetoclax (Ven-R) for relapsed/refractory CLL and 2 received BTKi-venetoclax as first-line treatment. Across all patients, the most common mutations detected were R107_R110dup (33%), A113G (30%), V156D (30%), G101V (28%) and D103 codon (23%). The dominant mutation (highest VAF) in individual patients was R107_R110dup (26%), A113G (23%), V156D (21%), G101V (21%), D103 (12%) and R129L (2%). Two patients had >1 co-dominant mutations of equal VAF. Other mutations detected in the cohort included TP53 (40%), NOTCH1 pathway (26%), XPO1 (14%) and RAS pathway (21%). Considering C-Ven vs FD-Ven, no difference in the frequency of individual BCL2 mutations was observed, however RAS pathway mutations were over-represented in disease relapsing after FD-Ven (31%, vs 6% for C-Ven). The prevalence of BCL2 mutations across the entire venetoclax-treated population was not estimable as samples were referred from across Australia; among patients from PMCC whose CLL progressed after Ven-R during the study period and tested, 5 (22%) had BCL2 mutations detected.

Conclusions: BCL2 mutations occur after both C-Ven and FD-Ven, including after first-line therapy, dispelling the expectation that fixed-duration regimens completely abrogate the emergence of resistance mutations. In the modern therapeutic and molecular diagnostic landscape, BCL2 mutations other than G101V occur at high frequencies and constitute important mechanisms of venetoclax resistance. This underscores the need for resistance testing beyond the G101V allele and for further research to examine potential differences in selection pressures for C-Ven and FD-Ven.