Combination Of Kinase Inhibitors Overcomes B-Raf Inhibitor-Induced Paradoxical ERK Activation In CLL Cells In Vitro – Potential Implications For CLL Treatment

Chronic lymphocytic leukemia (CLL) remains incurable with current state of the art therapy creating the need for novel therapeutic concepts. Kinase inhibitors represent a promising strategy in the treatment of various malignancies including CLL. However, based on the recent experience with other targeted therapy compounds used as single agents, it appears important to identify additional targets and to evaluate therapeutic combinations targeting two or more critical signaling hubs in CLL cells. This strategy is likely to counteract the development of drug resistance more efficiently.

We and others recently showed that the Raf/MEK/ERK pathway plays a critical role for the in vitro survival of CLL cells and demonstrated that drugs such as sorafenib targeting all Raf-isoforms and other kinases, induce apoptosis. Here, we provide a detailed analysis of various B-Raf inhibitors, including sorafenib, PLX4720 (vemurafenib tool compound) and dabrafenib on apoptotic pathways in primary human CLL cells and their impact on CLL cell survival alone or in combination with MEK (U0126), dual phosphoinositide-3 kinase/mTOR inhibitor (BEZ235), and IGF1R inhibitors (AG1024, Picropodophyllin (PPP)).

10µM sorafenib, a concentration comparable to plasma levels of this drug in patients, strongly induced apoptosis in CLL cells, while the more specific B-Raf inhibitor PLX4720 did not affect viability. This discrepant finding may be attributed to PLX4720’s well-described paradoxical ERK activation reported in various solid tumor types lacking BRAF mutations and containing increased Ras-GTP levels. Indeed, only sorafenib reduced ERK phosphorylation in CLL cells, while PLX4720 treatment even enhanced ERK activation.

Interestingly, not only the B-Raf inhibitors PLX4720 or dabrafenib, but also lower doses (1-5µM) of sorafenib induced paradoxical ERK activation. Most likely, at these concentrations drug-bound Raf molecules act as potent allosteric activators on drug-free Raf monomers, thereby leading to an increase in ERK pathway activity. Importantly, paradoxical ERK activation coincided with enhanced viability. In line with findings in melanoma cells, BRAF (V600E) mutated CLL cells showed no paradoxical ERK activation and were effectively killed by the respective compounds.

To identify inhibitor combinations reducing the in vitro survival of BRAF wild-type CLL isolates, we combined the B-Raf inhibitors with the MEK inhibitor U0126 or the dual PI3K/mTOR inhibitor BEZ235 observing reduced ERK phosphorylation and significantly enhanced rates of cell death (p<0.0009 for U0126; p<0.01 for BEZ235). This indicates that the additional MEK or PI3K inhibition counteracts paradoxical ERK activation in vitro and may overcome the resistance to apoptosis induction mediated by paradoxical ERK activation. Likewise, we tested combinatory effects of IGF1R inhibitors (AG1024, PPP) with suboptimal sorafenib doses and observed a significantly enhanced cell death for the combinations AG1024 and sorafenib (p<0,0001) as well as PPP and sorafenib (p<0,0001).

In conclusion, this is the first description of a paradoxical ERK activation by Raf inhibitors in CLL cells with unmutated BRAF. Our observation of paradoxical ERK activation after treatment with suboptimal sorafenib doses may be of clinical importance since sorafenib is currently investigated in clinical trials in several malignancies including CLL. Our findings imply that Sorafenib plasma concentrations should be monitored when used for the treatment of CLL since lower plasma levels might promote paradoxical ERK activation and CLL progression. Furthermore, our results show the potential of therapies combining kinase inhibitors: additional MEK or PI3K inhibition or the concomitant inhibition of both pathways by IGF1R inhibitors may overcome paradoxical ERK activation. Furthermore, given our recent observation that the IGF1R is overexpressed in CLL cells, its inhibition by IGF1R kinase inhibitors might be advantageous over the single inhibition of MAPK or PI3K pathway components.