Abstract
Introduction.BCL-2 family members are crucial determinants for survival in normal and malignant B cells. Venetoclax, the BCL-2 targeting drug, was recently clinically approved for CLL. CLL cells cycle between the lymph node (LN) and peripheral blood, and between those compartments display large changes in BCL-XL and MCL-1, which are not targeted by Venetoclax. However, novel BH3 mimetics specific for BCL-XL and MCL-1 are under preclinical development. Ibrutinib is a clinically successful BTK inhibitor that forces Chronic Lymphocytic Leukemia (CLL) cells out of the lymph node. After prolonged application resistance develops in a fraction of patients, who then show fast disease progression. Thus, both Venetoclax and Ibrutinib have potential drawbacks when applied as single agents. For long-term successful application of new drug combinations, it is crucial to understand BCL-2 member functionality of malignant B cells in relation to their normal counterparts. In parallel, we studied clinical samples of CLL patients under Ibrutinib or Venetoclax treatment for changes in BCL-2 family members.
Experimental procedures.FACS sorting of tonsil B subsets, mRNA and protein profiling, co-culture of CLL cells, co-IP of BCL-2 members, cell death assays with BH3 mimetics specific for BCL-2, BCL-XL or MCL-1 (ABT-199, WEHI-539, A-1210477), intracellular FACS for BCL-2, BCL-XL and MCL-1 in CLL (CD19/CD5/CXCR4).
Results.We mapped the clearly distinct expression profiles for BCL-2 members in tonsillar naïve, germinal center (GC), plasma (PC) and memory B cells, which translate into different BH3 mimetic sensitivity profiles. In brief, naïve and memory cells rely on BCL-2, GC cells on MCL-1, and PC on BCL-XL. These approaches were extended to primary CLL cells. In a LN model where cells are fully resistant to single agent ABT-199 due to induction of MCL-1, BCL-XL and BFL-1, we find that dual or triple BH3 mimetic combinations restore killing. Moreover, there is differential sensitivity to (combinations of) BH3 mimetics between healthy and malignant B cells, suggesting there is a therapeutic window.
Furthermore, intracellular FACS staining for BCL-2, MCL-1, and BCL-XL was performed on patients treated with Ibrutinib or Venetoclax. In untreated patients, recent LN emigrants (CD5hi/CXCR4lo) have higher BCL-XL and MCL-1 expression than 'old' returning cells (CD5lo/CXCR4hi). This distinction collapses under Ibrutinib treatment, demonstrating that Ibrutinib affects pro-survival BCL-2 members in vivo. Interestingly, in patients that develop Ibrutinib resistance the MCL-1 or BCL-XL pre-treatment profile reappears. In contrast, under Venetoclax treatment BCL-2 member levels do not change and can even increase in some patients, underlining the different mode of action of the two drugs. The pattern differed among patients (n=5) and increases were observed for BCL-2, BCL-XL, or MCL-1.
Conclusions. BH3 mimetics have not only found their way to the clinic, but can also be used as tools for BCL-2 functionality profiling. Secondly, our ex vivo data underline that combination treatment between (paired) BH3 mimetics and/or Ibrutinib may afford better long-term efficacy in CLL than single agents. Thirdly, BCL-2 members may be useful as biomarkers for response and progression.
Eldering:Roche: Research Funding; Gilead: Research Funding. Brown:Pharmacyclics: Consultancy; Abbvie: Consultancy; Roche: Consultancy; Genetech: Consultancy. Kater:Celgene: Research Funding.
Author notes
Asterisk with author names denotes non-ASH members.