Why is it so hard to develop a drug that cures cancer?

By Carlo Maley, Guest Author

[SM: Carlo is an old friend of mine from MIT. We met at orientation in 1993. At the time, there was no formal field called Computational Biology, which was his area of interest.]

We all know that cancer is a big, scary problem. Approximately 40% of Americans will get some form of cancer in their lifetime. Many of us will die of the disease.

Nixon declared war on cancer in 1971 and despite huge gobs of cash being pumped into cancer research since that time, cancer has fought us to a draw. According to the American Cancer Society’s statistics, the rate of deaths due to cancer did not change between 1950 and 2004, even when you adjust for changes in lifespan.

That fact still astounds me, given how much medicine and technology have changed over the same period of time. For every advance in the treatment of one type of cancer, there has been a retreat in another type. According to a drug rehab in Miami FL, in the last few years, cancer rates and the absolute number of cancers deaths in the U.S. has decreased for the first time. But this is attributed to better screening (e.g., Pap smears, mammograms, colonoscopies, etc.) to detect cancer early when it can be removed surgically, and to cancer prevention (e.g., there are fewer smokers today). So where are the breakthrough drugs that can cure cancer?

The truth is that there are many drugs that can cure some cancers, some of the time. However, any oncologist will tell you that no single drug will work. Why is that? The short answer is: Evolution. The cells in a tumor evolve by natural selection. They generally mutate at a much higher rate than the rest of the cells in your body. This can happen for a variety of reasons including a mutation in the cancer cells that destroys the error correcting machinery that normally detects and fixes mistakes made when a cell copies its genome as it prepares to divide, or a mutation that mucks up the orderly segregation of the two copies of the genome into the two daughter cells. Whatever the reason, a tumor is a vast mosaic of mutant cells, with tens of thousands of mutations by some measures. Some of those mutations give a tumor cell a competitive advantage over its neighbors. We typically find mutations in cancer cells that make them less likely to die and help them divide faster than other cells. By the time a person notices a 1ccm lump and goes to the doctor, that tumor contains a billion cells. In many cases people show up in the clinic with a trillion cancer cells in their bodies.

Here is what generally happens next. If we can’t cut out the tumor, we give the patient chemo- or radiation therapy. Sometimes this has no effect, but often the tumor shrinks and may even disappear from MRI or X-ray scans. However, even a million cells are too few to be detected by MRI or X-rays. Within a few months or years, the tumor will reappear and now reapplication of that same therapy will have no effect.

So what is going on under the hood? Applying a chemotherapy is equivalent to spraying a field of crops with an insecticide. You will kill most of the insects but if there are any mutant insects that are resistant to the insecticide, they will start to proliferate no matter how many times you spray the field. In the case of cancer, the cancer drug may kill most of the cells in the tumor, but with billions of cells there, and tens of thousands of mutations, it is likely that there are some mutant cells that will be resistant to the therapy, no matter what therapy you choose. It is just a numbers game. Our therapies select for resistant cells, so that at relapse, when the tumor grows back, all the cells of the tumor are resistant to the therapy and reapplication of the therapy will no longer have any effect. There are other problems that make cancer therapies hard to develop. Resistance may also be caused by a failure of the drug to reach all the cancer cells. It can also be hard to find a drug that kills cancer cells and not too many normal cells. But the central problem remains: cancer drugs select for resistance.

We have learned that we have to use a cocktail of multiple drugs, for the same reason we do this with HIV – the chance that a cancer cell will be resistant to all the drugs in a cocktail is less than the chance it will be resistant to any one drug. But even cancer drug cocktails have only bought us an incremental improvement (and it is a nightmare to get multiple drug companies to agree on how to divide profits and liability when their drugs are combined in a cocktail). Rather than dealing with the problem of resistance, most cancer drug research is focused on finding a new drug that will kill cancer cells, and will inevitably select for resistance to that drug.

The fact that cells in tumors evolve is the central problem for cancer therapy and the reason cancer has been so hard to cure. This is the reason recent interest has turned to early detection and cancer prevention, where we might be able to head-off the problem before the disease becomes incurable.