Scientists report on the discovery of a protein expression feedback mechanism that may explain why the BRAF inhibitor vemurafenib (PLX4032) is an effective therapy for melanomas carrying the BRAF V600E mutation but is much less effective against colon cancers carrying exactly the same mutation. A team led by scientists at The Netherlands Cancer Institute found that inhibiting BRAF in colon cancer actually triggers upregulation of the epidermal growth factor receptor (EGFR), which supports further tumor growth. Melanomas only express low levels of EGFR, so this feedback mechanism doesn’t occur.
René Bernards, Ph.D., and colleagues subsequently showed that treating mice carrying colorectal cancer xenografts using a combination of vemurafenib and an EGFR inhibitor was much more effective at holding back cancer growth than vemurafenib alone. Reporting their studies in Nature, the team suggests their data provides “a strong rationale for a clinical trial combining BRAF and EGFR inhibitors in BRAF V600E mutant CRCs.” Their paper is titled “Unresponsiveness of colon cancer to BRAF(V600E) inhibition through feedback activation of EGFR.”
Activating mutations in the BRAF oncogene occur in about 70% of primary melanomas and about 10% of colorectal cancers. However, while 80% of BRAFV600E melanomas will respond to vemurafenib, the response rate among relevant colon cancers is only 5%. To investigate the molecular mechanism behind this inherent resistance of BRAF-mutated colon cancers to vemurafenib, the researchers applied an RNAi genetic screening approach to see whether knocking out specific kinases might restore this drug sensitivity. To this end they screened an shRNA library representing all 518 human kinases and another 17 kinase-related genes.
The resulting data strongly indicated that suppression of EGFR synergized with BRAF inhibition. This notion was supported by studies in WiDr human colon cancer cells. Effecting EGFR silencing using an shRNA had no effect on cell proliferation, a result tallying with clinical observations that KRAS or BRAF mutant CRC cells don’t respond to EGFR-targeted mAbs, the authors note. In contrast, when EGFR-silenced WiDr cells were treated with vemurafenib, there was significant inhibition of proliferation.
Similar antiproliferative effects were observed in three different CRC cell lines treated with vemurafenib in combination with the anti-EGFR drugs cetuximab, gefitinib, and erlotinib. These findings are “consistent with the notion derived from the shRNA screen that EGFR inhibition is required to elicit a response to BRAF inhibition in CRC cells.”.
As to why EGFR inhibition renders CRC cell lines susceptible to BRAF inhibition, further antibody-based studies on lysates of drug-treated cells showed that vemurafenib treatment leads to strong activation of EGFR via Tyr 1068 phosphorylation. This feedback loop could be blocked by co-treatment of cells with vemurafenib and either cetuximab or gefitinib. Moreover, vemurafenib treatment inhibited MEK and ERK activation downstream of BRAF but activated AKT, which operates downstream of EGFR in a pathway parallel to BRAF, they note.
Significantly, treating CRC cell lines using the MEK inhibitor AZD6244 had the same EGFR-activating effects as vemurafenib therapy, in terms of EGFR phosphorylation, whereas administering both the MEK inhibitor and an anti-EGFR inhibitor blocked cell growth.
EGFR is expressed primarily in epithelial cancers, whereas melanomas are derived from the neural crest, providing an additional clue as to why melanoma responds better to vemurafenib than CRC. Indeed, when the investigators compared EGFR expression in a panel of BRAF(V600E) mutant melanoma, colon cancer and thyroid cancer cells, they found the melanoma cell lines expressed low levels of EGFR, whereas eight of 10 CRC cell lines examined expressed much higher levels of EGFR.
The two CRC lines that expressed only low levels of EGFR were subsequently found to be almost as sensitive to vemurafenib as the melanoma cell lines both in short-term and long-term assays. And, as expected, the thyroid cancer cell lines that also expressed high levels of EGFR were similarly sensitive to a combination of gefitinib and vemurafenib but not to vemurafenib monotherapy. In a final set of in vitro assays, the researchers demonstrated that ectopic expression of EGFR by normally vemurafenib-sensitive melanoma cells was enough to make them resistant to the drug.
The next step was to carry out in vivo evaluation. Immunodeficient mice carrying human WiDr and VACO432 CRC tumor xenografts were treated using either an EGFR inhibitor (cetuximab or erlotinib) or the BRAF-inhibitor PLX4720 (which is similar to vemurafenib but easier to formulate for use in mice) or a combination of EGFR inhibitor plus PLX4720. The effects in vivo mirrored those observed in the cell-based assays, in that animals given combination therapy demonstrated significant tumor growth inhibition, while those treated using either the EGFR inhibitor or PLX4720 derived very little benefit.
The Netherlands Cancer Institute researchers say human clinical trials evaluating combined EGFR and BRAF inhibition in CRC patients are warranted. They also point out EGFR expression levels may represent a clinically useful biomarker to help predict tumor response to vemurafenib monotherapy in BRAF mutant CRC tumors and aid in the selection of EGFR combination therapy in other cancers with the BRAF V600E mutation.
The frequent development of resistance to single targeted anticancer agents has led to a trend toward combining drugs in clinical trials, the team adds. “Our results highlight the power of ‘synthetic lethality’ genetic screens to identify which combinations of pathway inhibition are particularly effective. As such, these screens may help prioritize which combination therapies have the highest likelihood of being successful in the clinic.”