NCI Scientists Test 5,000 Drug Combinations

Released: 22-Jan-2013 12:15 PM EST
Source Newsroom: National Cancer Institute (NCI) at NIH
Contact Information

Available for logged-in reporters only

Dr. Susan Holbeck, Ph.D., is a biologist in the Division of Cancer Treatment and Diagnosis at the National Cancer Institute. Her team spent two years testing all possible combinations of 100 cancer drugs currently approved by the FDA to see if any previously untested combinations are effective in certain cancers. 5,000 drug pairings were tested, totaling 300,000 experiments in all. Each pair of drugs was tested on the NCI-60, a panel of 60 diverse cell cultures, often referred to as cell lines. The NCI-60 allows researchers to analyze the anti-cancer properties of a compound on each of 60 unique cell samples from a wide range of cancers. As a result of this research, Holbeck’s team hopes to quickly bring several novel drug combinations to preclinical trials and then to the clinic.

Why did you decide to conduct this study?
Of the drugs that are out there, of all the possible pairs, less than a third of them have been tested together in people. That’s quite a large number that are already approved that have not been tested together. So, we hoped to find pairs of drugs that give a better effect than the single drugs alone. Can you get somewhere in treatment with a combination of drugs that you could not get with just the single agents? Once we find a drug pair that works better than we would expect for the single agents, we can take a look at the characterization of the cell line. We have a lot of molecular characterization data. We know the level of RNA expression in every gene in each of the cell lines. We have sequence data, protein data, and data on micro RNA expression. We can take all of that, do data mining and see why one cell line would respond well to combinations while another cell line would not. We can take a pair of observations and turn it into a rational, targeted drug that we can test in the clinic with a hypothesis.

What makes a study of this scale possible now?
The technologies for doing drug screenings have allowed smaller and smaller assays to be done and robotics have improved, so we decided it was a feasible effort to go ahead and do this. It is a one-time shot. It is a finite number of drugs and a finite number of cell lines and it has generated a large body of data that we hope will be useful to the research community.

About 0.1% of the tests produced positive results, showing a benefit in a number of cell lines. What does that mean moving forward?
When you are testing 300,000 combinations you can’t possibly take them all forward, even into animal testing, and certainly not into human testing. We take a look at the things that are most interesting. We want to see if we can discover, through data-mining, what might be things that we could develop a rationale for their use, and take those forward into animal testing. Then, we can see whether they actually have an affect on tumors, whether the combination is, in fact, better than a single agent on tumors in mice.

Your team tested the new combinations in human tumors grafted into mice. How does that work? Is that a common practice?
It is a very common practice. The mice are immunocompromised, so they don’t have an immune system. Their bodies are not going to fight off the graft. The cells can be implanted either under the skin of the mouse or implanted into a particular organ in the mouse and allowed to grow. You can measure how fast the tumor grows if you don’t treat it and how fast it grows if you give it drugs. You can use different doses of the drugs and different schedules. Whether this is actually predictive of what happens in the clinic, there is a lot of debate around that, but doing these experiments in mice allows you to say that you are able to deliver the drug to the tumor. You are able to have an effect on the tumor inside of something other than just a tissue culture dish. It is quite a big leap to go from cell lines to testing in humans. Mice are an intermediate step.

Why was it important for your team to test all possible drug combinations, stepping outside of what many scientists have long thought of as “rational combinations” (i.e. making scientific inferences about which pathways it might make sense to block simultaneously with different drugs)?
Let me make clear that the rational combinations are included in this study. Especially for a lot of the older drugs, their mechanisms are less well understood than the newer drugs. But even in the newer drugs, where someone has developed a drug that targets a specific kinase, for example, often, they will hit kinases in addition to the one that is being marketed. The drug will strongly affect one target and maybe less strongly affect another target but the secondary targets could provide a rational basis for combinations if we understood that.

Does the fact that all of these drugs are already approved by the FDA make the next stage of trials quicker or easier?
That is the hope. I am not an expert on the hurdles for getting a new use for an existing drug. Although, when you have a drug that is already approved, the hurdle is lower than it is for a drug that is not approved.


Comment/Share