Ex vivo sensitivity to venetoclax is predictive of clinical activity

Abstract

A drug screening platform including 76 multiple myeloma (MM) relevant therapeutics was evaluated in 113 primary MM samples. The panel included venetoclax, a selective BCL-2 inhibitor that has shown exceptional antitumor activity in hematological malignancies. Here we explored the ex vivo venetoclax response and its association to clinical features and BCL-2 family RNA expression levels. Venetoclax was screened in a 7-point, 10-fold dilution starting at 10 μM, incubated for 24 hours, and response was evaluated through cellular viability. The majority of the 113 primary patient samples (93%) exhibited a dose response to the drug; broad mid-point EC50 ranges were observed (<0.1 to >10000 nM) and AUCs varied from 0.03584 to 0.9327. A mid-point EC50 lower than 100nM was measured in 49% of samples, demonstrating the overall potency of venetoclax in MM. As expected, samples from patients harboring t(11;14) had an increased ex vivo sensitivity to the drug (n=28; median AUC 0.1491) when compared to patients lacking the translocation (n=82; median AUC 0.3644) (p=0.0013). Additionally, increased sensitivity was demonstrated in samples from patients with newly diagnosed MM (n=35; median AUC 0.1977) when compared to relapsed MM (n=64; median AUC 0.4025) (p=0.0041); and standard risk (n=36; median AUC 0.2324) when compared to high risk (n=60; median AUC 0.4025) (p=0.0199). Other cytogenetic characteristics associated to venetoclax sensitivity was low plasma cell S-Phase (n=65; median AUC 0.2697) when compared to high S-Phase (n=31; median AUC 0.4353) (p=0.0035); samples lacking Gain(1q) (n=63; median AUC 0.2788) when compared to samples with Gain(1q) (n=47; median AUC 0.4047) (p=0.0383); and in samples lacking t(4;14) (n=99; median AUC 0.307) when compared to samples harboring the translocation (n=11; median AUC 0.5514) (p=0.01). Then, anti-apoptotic BCL-2 family member transcriptomic ratios were evaluated in the nine most (responders; median AUC 0.09409) and least (non-responders; median AUC 0.7195) sensitive primary patient samples. Overall BCL-2 RNA expression (p=0.0036), BCL-2/MCL-1 ratio (p=0.0019), and BCL-2/BCL-2L1 ratio (p=0.0106) were significantly increased in the responders when compared to the non-responders. Finally, two relapsed t(11;14) MM patients were treated with venetoclax, after the ex vivo drug screen results pointed to moderate sensitivity to the drug, with AUCs of 0.3380 and 0.3575. Both patients were treated with venetoclax in combination with dexamethasone or dexamethasone and carfilzomib and achieved partial responses. Collectively, our data obtained through a drug screening assay in ex vivo primary patient samples corroborates the strong preclinical and clinical associations of venetoclax response, including t(11;14) and favorable BCL-2 family profiles. The assessment of ex vivo sensitivity to venetoclax was shown to be an applicable tool to enrich for likely responders.