Two U.S. groups have developed platforms that could improve the identification of cancer drug combinations that address drug resistance. A team from The University of North Carolina at Chapel Hill School of Medicine is using chemical proteomics to rationally design kinase inhibitor combinations that block signaling pathways mediating drug resistance,1 whereas Massachusetts Institute of Technology researchers are screening for targeted therapeutics that sensitize cancer cells to DNA-damaging agents when given sequentially rather than simultaneously.2

Each group has provided proof of principle for their platform by identifying a combination that showed efficacy in a mouse model of triple-negative breast cancer (TNBC).

Previous efforts to develop drug combinations that counter drug resistance in cancer have often focused on targeting genetic alterations in tumors that are acquired or become more prevalent in response to therapy. However, resistance can also be mediated by drug-induced changes in signaling pathways.

"Drug resistance is probably the biggest challenge that we face in cancer therapeutics today. I think the most important lesson from these two papers is that tumor cells can have an extraordinarily complex adaptive rewiring response to drug treatment, and we need to understand these responses at the protein network level, in addition to at the genetic level, in order to predict outcomes and select the best combination treatments," said Paul Workman, deputy CEO and director of the Cancer Research UK Cancer Therapeutics Unit at The Institute of Cancer Research.

Kinase to kinome

A team led by Gary Johnson set out to understand how kinase-signaling pathways in tumors respond to kinase inhibitors, with the hope that the information might guide the design of combination treatments. Johnson is chair of the Department of Pharmacology at the UNC School of Medicine.

The UNC researchers focused on TNBC, a subtype defined by the lack of three markers: estrogen receptor and progesterone receptor expression and HER2 (EGFR2; ERBB2; neu) amplification. There are no targeted therapies approved for TNBC, and chemotherapy is standard of care.

In tumor tissue from a patient with TNBC, more than 400 of the 518 human kinases were expressed. Of the 400 expressed kinases, about half could be profiled using chemical proteomics-a technique in which mass spectrometry is used to quantitatively measure the activation state of kinases.

The team found that tumor tissue had greater activation of the RAF-MEK-ERK pathway than adjacent normal mammary tissue. In human TNBC cell lines, the MEK inhibitor selumetinib initially blocked signaling downstream of MEK and blocked cell growth. However, after 24 hours of treatment, downstream signaling was reactivated and the tumor cells began to grow again.

Array BioPharma Inc. and AstraZeneca plc's selumetinib (AZD6244) is in Phase II testing in solid tumors.

Johnson's team hypothesized that kinase inhibitors against MEK could be activating other kinase-signaling pathways, thus blunting long-term efficacy. To find out if this is the case, the group used its platform to profile changes in kinase activation in tumors in response to selumetinib in a genetically engineered mouse model of TNBC.

In the tumors, giving selumetinib led to the upregulation of at least 30 kinases. The upregulated kinases included known targets of the multitargeted kinase inhibitor Nexavar sorafenib. Indeed, treating the TNBC mice with selumetinib plus Nexavar prevented the upregulation of about two-thirds of these kinases.

Onyx Pharmaceuticals Inc. and Bayer AG market Nexavar for liver and renal cancers. The drug is in Phase III testing in combination with chemotherapy in relapsed or refractory HER2-negative breast cancer.

Blocking the selumetinib-induced kinase upregulation resulted in a therapeutic benefit. In TNBC mice, selumetinib plus Nexavar led to tumor regression in 77% of the animals compared with 30% for selumetinib alone and 0% for Nexavar alone.

Results were published in Cell.

Building on the results in the mouse model, the UNC team is now using chemical proteomics to look at the effects of the GlaxoSmithKline plc MEK inhibitor trametinib (GSK1120212) on kinome activation in tumors from patients with TNBC. Trametinib inhibits MAP kinase kinase 1 (MAP2K1; MEK1) and MEK2 (MAP2K2).

In January, the researchers launched a trial to monitor the activity of kinases in tumor tissue before and after one week of treatment with trametinib. The trial is enrolling 10 patients and is expected to be completed by November. The trial is investigator sponsored, and the drug is not being dosed long enough to assess efficacy.

This month, GSK announced positive Phase III results with trametinib in metastatic melanoma patients with activating BRAF mutations. The molecule also is in Phase II testing to treat pancreatic cancer and relapsed or refractory solid tumors. Trametinib has not been tested in patients with TNBC.

Data linking the targeted inhibition of oncogenic pathways with the reprogramming of key kinases "suggests that the most effective approach to treating an individual's tumor may not only depend on the hardwired genetics but also on how each tumor changes at the molecular level after dosing with novel therapies," said Kiran Patel, development leader for GSK's MEK program.

"It is largely unknown how predictable or variable this reprogramming may be between different patients," said Patel. The UNC trial "will give us a more complete understanding of the potential diversity of kinase modulation by trametinib in triple-negative breast cancers, which could lead to more effective combination strategies for this tumor type in the future."

Johnson told SciBX the UNC team is planning to use the same proteomic approach to look at the kinome response to HER2-directed antibodies and small molecules in breast cancer and to targeted therapeutics in leukemia and pancreatic and renal cancers.

A patent application was filed on the technology and is available for licensing, Johnson said.

Drug combo two-step

An MIT group led by Michael Yaffe took a different tack to identify drug combinations that target signaling pathways: looking for drugs that were synergistic when dosed sequentially.

Yaffe is a professor in the Department of Biology.

Because some patients with TNBC respond to DNA-damaging chemotherapeutics, the researchers reasoned it might be possible to identify a targeted therapeutic that, if given in advance of chemotherapy, could rewire the signaling pathways of tumor cells to increase their sensitivity to DNA-damaging agents.

To look for such an effect, the MIT team screened combinations of targeted therapeutics and DNA-damaging agents in a TNBC cell line. In the screen, the combination of drugs-and the order in which they were dosed-was varied.

Consistent with previous results, the researchers found that treating TNBC cells with Tarceva erlotinib alone or coadministration of Tarceva and doxorubicin did not potently induce apoptosis. However, treating the cells with Tarceva at least four hours before doxorubicin induced activation of caspase-8 (CASP8; FLICE), a key mediator of apoptosis, and increased cell killing as much as fivefold over simultaneous treatment with the two drugs (p<0.0001).

Astellas Pharma Inc., Chugai Pharmaceutical Co. Ltd. and Roche market Tarceva, an epidermal growth factor receptor 1 (EGFR1; HER1; ERBB1) inhibitor, for pancreatic cancer and non-small cell lung cancer (NSCLC). The drug also is in Phase III testing for liver cancer.

The team also looked at the effect of the sequential combination in a panel of TNBC cell lines. Tarceva followed by doxorubicin induced synergistic cell killing in 4 of 10 TNBC cell lines tested. Cell line sensitivity correlated with basal EGFR activation-determined by the level of phosphorylated EGFR-not EGFR expression.

This result suggests the activity of the signaling pathway, rather than genetic alterations or gene expression, is important for determining drug response.

Finally, the team looked at the effects of the drug combination in vivo. In a xenograft mouse model of TNBC, doxorubicin alone, coadministration of Tarceva and doxorubicin, and sequential administration of Tarceva and doxorubicin all initially led to tumor regression. However, only the sequential combination prevented tumor regrowth.

Data were published in Cell.

"This work introduces a novel paradigm for combination treatments in oncology. What is particularly appealing is that the researchers could show that EGFR phosphorylation, not EGFR gene amplification, is predictive of the sensitizing effect of sequential erlotinib and doxorubicin treatment in the small sample set of cell lines tested," said Birgit Schoeberl, VP of discovery at cancer company Merrimack Pharmaceuticals Inc.

"I believe that this could be an interesting hypothesis to be tested clinically, especially since there are currently only limited treatment options for TNBC patients," she added.

However, Schoeberl noted that identifying TNBC patients with activated EGFR and optimizing a sequential dosing regimen could prove challenging.

"Since it is difficult to measure protein phosphorylation in tumor biopsies, the testing of EGFR phosphorylation as a potential predictive marker may be challenging clinically. Also, I think we would want to further understand how long a patient needs to be on anti-EGFR therapy prior to being treated with doxorubicin in order to obtain the maximal sensitizing effect to the subsequent chemotherapy," she said.

Yaffe, who is a member of Merrimack's scientific advisory board, told SciBX his team is working with a clinician at the Dana-Farber Cancer Institute to design a clinical trial of the sequential treatment regimen of Tarceva and doxorubicin.

A specific start date for the trial has not been determined.

The MIT team is now working on developing assays to look at EGFR activation in tumor biopsies and is looking for sequential treatments with potential efficacy in additional cancer indications, including pancreatic and lung cancers.

A patent application has been filed covering the technology, and the IP is available for licensing, Yaffe said.

Kotz, J. SciBX 5(24); doi:10.1038/scibx.2012.617
Published online June 14, 2012


1.   Duncan, J.S. et al. Cell; published online April 13, 2012; doi:10.1016/j.cell.2012.02.053
Contact: Gary L. Johnson, The University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, N.C.

2.   Lee, M.J. et al. Cell; published online May 11, 2012;
Contact: Michael B. Yaffe, Massachusetts Institute of Technology, Cambridge, Mass.


Array BioPharma Inc. (NASDAQ:ARRY), Boulder, Colo.

Astellas Pharma Inc. (Tokyo:4503), Tokyo, Japan

AstraZeneca plc (LSE:AZN; NYSE:AZN), London, U.K.

Bayer AG (Xetra:BAYN), Leverkusen, Germany

Cancer Research UK, London, U.K.

Chugai Pharmaceutical Co. Ltd. (Tokyo:4519), Tokyo, Japan

Dana-Farber Cancer Institute, Boston, Mass.

GlaxoSmithKline plc (LSE:GSK; NYSE:GSK), London, U.K.

The Institute of Cancer Research, Sutton, U.K.

Massachusetts Institute of Technology, Cambridge, Mass.

Merrimack Pharmaceuticals Inc. (NASDAQ:MACK), Cambridge, Mass.

Onyx Pharmaceuticals Inc. (NASDAQ:ONXX), South San Francisco, Calif.

Roche (SIX:ROG; OTCQX:RHHBY), Basel, Switzerland

The University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, N.C.