The science behind the discovery: Combination drug therapies for melanoma

To learn more about the science behind the discovery that combination drug therapies help overcome resistance problems in late-stage melanoma (see story), YaleNews posed some detailed questions to co-first authors David Stern and Marcus Bosenberg. Here are their answers. (Caution: heavy science ahead.)

Why might melanoma eventually become resistant to treatment, even when initial therapy with a single drug has been successful?

The general answer is that resistance is a common outcome for any single-agent therapy, and experience now shows that this is an especially prevalent problem for therapeutic agents targeted at components of signal transduction pathways (pathways that convert one chemical process into another*). For example, Zelboraf, for BRAF-mutated melanoma, and Tarceva, for EGFR-mutant lung adenocarcinoma, both lead to dramatic responses in patients with those mutations, but these responses range from months to a year or two on average.

(* Note: In normal cells, the decision to divide is regulated by hormones called growth factors. These hormones bind to cell surface receptors, and the receptors then activate "signal transduction" pathways that ultimately induce cell division. Most cancers are driven by excessive activity of either the receptors (e.g. EGFR in some lung cancer, HER2 in some breast cancer.)

Your study focused on mutated BRAF and RAS oncogenes. What is it that makes them so aggressive and difficult to treat?

We focused on BRAF and RAS oncogenes because these are the most common driver mutations in melanoma, with BRAF linked to 50% all of cases, and NRAS around 20%. Although metastatic melanoma is a very aggressive disease, there are a number of treatment options that work well for those with BRAF melanoma, including Zelboraf. The main issues here are the initial resistance of a minor subset of patients, and the treatment-selected resistance for most patients that occurs with a median of around 7 months.

RAS mutants are a different story. Mutations in one of the three RAS genes (NRAS, HRAS, and KRAS) are common in human cancer – accounting for as many as 30% overall. This has been known since the early 1980s, and there have been many sustained and aggressive efforts to directly attack RAS proteins. For single agents, there have been  no major successes in directly targeting the RAS proteins themselves. The problem lies mainly with the biochemical nature and structures of the RAS proteins, which lack major surface features that are susceptible to binding by drugs. Agents that prevent a required chemical modification of NRAS and HRAS proteins have shown promise in preclinical trials, but have not shown positive results in clinical trials.

What is the mechanism that enables the drug combinations you identified to be more more successful in fighting late-stage melanoma?

First a word of caution: What we have really done is identified combinations that are effective in inhibiting growth and promoting death of cell cultures that were derived from late-stage melanoma. Even the one animal study, for statin plus flavopiridol, is a fairly artificial approach, but a typical early-stage animal study.

Our study identified many combinations that were effective on BRAF melanoma, and a much smaller subset (mainly including statins) effective on NRAS melanoma. The general reason combinations may do better than single agents is that the two agents may inhibit two different processes that are important for the cancer, or that one agent inhibits the cancer process and the second pushes the cells over the edge into programmed cell death. The particular mechanisms of interaction will be specific to each drug combination pair, and we only began a mechanistic evaluation of two combinations in the paper.

For statins plus flavopiridol, each agent is doing something quite different. The main effect of statins for NRAS cells is to inhibit membrane localization of NRAS (by inhibiting the chemical modification that helps NRAS attach to cellular membranes). But, statins affect many important cancer pathways and we will be doing further work to see to what extent impact on metabolism, and other small proteins that are related to RAS and are known to be similarly affected by statins. Interestingly, these include RAC1, which a Yale melanoma sequencing study led by Ruth Halaban, showed is sometimes a melanoma driver. Flavopiridol inhibits cyclin-dependent protein kinases that are part of the cell division machinery, so that inhibition of these proteins directly interferes with the cell division cycle (and also transcription).

For BRAF, the one case we looked at in detail was a cell line that was resistant to Zelboraf/vemurafenib, despite having a BRAF mutation. We knew from earlier work that this line has very active EGF receptor, which is a hormone receptor commonly activated in cancers. We found that bringing in an EGF receptor inhibitor plus an AKT pathway inhibitor was effective in reducing cell growth, but that full suppression of the major cancer pathways beginning with EGFR, and including the BRAF to MAP kinase pathway and the parallel AKT signaling pathway required combination of three agents: Vemurafenib to knock out the effect of the BRAF mutation, Lapatinib to knock out the EGFR, and also the AKT inhibitor MK-2206, probably to counter a feedback circuit that often activates AKT signaling pathway when the RAF/MAP Kinase pathway is suppressed.

Many people were surprised to learn from your study that cholesterol-lowering statins were part of some of the successful drug combinations. What is it about statin drugs that helps fight melanoma?

As mentioned above, statins have many effects that work through inhibition of their target HMGCoA Reductase. Inhibiting this lipid synthesis pathway will affect production of many lipids, but also affect cellular metabolism, and indirectly but significantly affecting proteins in the RAS family, and cancer-related proteins RAC and RHO. In our cell culture studies with NRAS mutant cells, the impact on the NRAS driver itself is clearly a major component. But, the functional mix may be different in tumors, and this is an exciting area for further exploration.

How hard is it to get combination drug therapies approved by the Food and Drug Administration (FDA)?

The barrier is greatest where neither agent has been approved by the FDA, since the toxicity component is complicated; however, the FDA is now considering approval of selected drugs specifically as part of combination therapies.

When do you think your promising research might be ready for human clinical trials?

In theory, the results could set in motion the process of design, review, funding, and approval of clinical trials at any time. This is particulary true for the combination of a statin with flavopirodal (or another cyclin dependent kinase inhibitor), as the safety issues would be simpler to address for two drugs than for three. Much depends on the interest of the pharmaceutical companies and oncologists, which should become evident within a few months of publication. We are in active discussions with Paul Eder (leader of the Phase I group at YCC) and some drug companies.