A few years back I lost a dear friend to a nasty little cancer called glioblastoma multiforme (GBM). It affects the glial cells of the brain and it is highly malignant. Back then, a diagnosis of GBM (also called a Grade IV astrocytoma) was a sentence to an early death.
From the outside, it hasn’t changed much today. A diagnosis of GBM means you will probably be dead within two years. But from the inside, changes in our cancer knowledge are coming so fast you can hardly see them for the dust.
But it didn’t start out very fast. Nor, for that matter, did it start with any of the fanfare we expect today for great scientific discoveries.
It started with a chicken.
Back in 1910 a farmer had a highly prized chicken with a tumor. He thought so highly of the chicken that he took it to the Rockefeller Institute in New York to try to get it cured. Cancers back then were mysterious things, so he didn’t expect much. What he got was a little below even his expectations.
The cancer specialist he was directed to, a man named Peyton Rous, not only couldn’t cure the chicken, he actually talked the farmer into letting him kill it! Rous must have had quite a gift for gab, since he not only talked the farmer into that, but he also persuaded him to let him have some other chickens from the same strain.
Using those chickens, Rous was able to conduct a remarkable experiment. He took some of the tumor from the original chicken, ground it up, and then put it through a filter so fine it filtered out the smallest cells, even microbes. Then he injected the product into some of the farmer’s chickens.
They immediately got cancer.
What Rous had done is present the first clear proof that some cancers, at least, came from viruses. Unfortunately, since everyone knew that was impossible, Rous had to wait until 1966 until he finally got his Nobel Prize. (I said it didn’t start out very fast.)
What Rous had discovered we now see as half of the basic cancer mechanism. The virus he had discovered had only four genes. Three of them were essential for the virus to reproduce. The fourth, a completely superfluous gene, was the first oncogene, a gene that stimulated unregulated multiplication of cells.
Over the last twenty years, building on Rous’ foundation, scientists have created a model of cancer that is elegantly simple. According to this model, all human cancers require the dysfunction of only one each of two types of genes: oncogenes and anti-oncogenes.
Oncogenes are (typically) dominant genes that stimulate abnormal proliferation of cells. They start out life as proto-oncogenes, usually having some function to do with cell growth and division. Then they are damaged. Which turns a nice, well behaved proto-oncogene into an oncogene that wants to produce wild, unregulated cell division and growth. Since the oncogene is a dominant gene, it is irrelevant if its paired gene is intact or not. The dominant gene controls behavior.
But nothing happens.
This kind of damage is common enough that the genome has produced a cancer prevention mechanism called a suppressor gene or an anti-oncogene. These are (typically) recessive genes whose job it is to inhibit or terminate cell proliferation. If either of the two suppressor genes is intact, the oncogene is inhibited. However, if both suppressor genes are inactivated or removed, then, and only then, can the oncogene run wild.
So it takes at least two separate “insults” to the genetic code to produce cancerous growth. The proto-oncogene must be converted into an oncogene and, even in families where one of the suppressor alleles is actually missing (i.e. high susceptibility “cancer families”), the other suppressor gene still has to be damaged.
Of course, it is typically a lot more complicated than that. A number of oncogenes have to be turned on and suppressors turned off to get that wild proliferation that is a malignancy. For instance, those new cells have to be supplied with blood, so a process called angiogenesis has to be turned on. Then too, other cells may have to be invaded, which involves shutting off some protective mechanisms. And, if the cancer is to metastasize, a complex process called migration has to be enabled.
Even a specific cancer, like glioblastoma multiforme, comes in various types, depending on what oncogenes are turned on and what suppressor genes are turned off. They may look the same and they may be just as fatal, but they are genetically different. This is important because knowing the mechanisms that produce malignancy opens the possibility of genetic therapies to combat them.
Imagine for a second that you want to go into a cell and do a little genetic repair. First thing you do is grab nature’s favorite gene splicer, the virus. A virus reproduces by inserting its own code into a cell’s DNA and turns the cell into a virus factory. So you remove the genetic material that allows the virus to reproduce its own genetic code. You replace that material with some other piece of code that you want to add to the DNA of a cell. You have created what is known as a vector virus. You inject your vector virus into the cell and it magically does its gene splicing. Only this time it inserts your new code into the existing DNA.
This sounds great. In theory, you could actually use this process to detect and repair the damaged genes that have turned the cells malignant. Or you could insert new, healthy suppressor genes in the chromosomes to shut down the cancer process. These and many other approaches have been tried, but the success has only been middling.
There are a number of complications. For one thing, it turns out viruses are not too specific about where they insert their new code, so it might be mislocated. Again, since each cancer is individual in its pattern of defective oncogenes and suppressor genes, there is no universal “magic bullet.” That is, your therapy apparently has to be customized for each individual. Then, too, how will you make sure that your therapy somehow reaches every single one of the cancerous cells? Any you miss will happily act as seeds to restart the entire uncontrolled growth process. And those are only some of the problems.
But if there are a lot of roadblocks, there are also a lot of great ideas being tried, too. For instance:
* Maybe, instead of trying to fix them, we will add some programming to make the cancer cells commit suicide.
* Maybe we will simply modify the cancerous cells to express some antigen on their surfaces to identify them to the immune system. Then we let the immune system clean them up.
* Maybe we will choose to target the immune system cells themselves and modify them to learn how to recognize and kill these specific cancer cells.
* And on and on.
Which ones will work? Who knows? But, looking through the forest of current failures and half successes, it is clear that we can now see the basic components of the problem. That, in itself, constitutes one of the major medical revolutions of our time. But more than that, it is clear that in a very few years we may see automated oncogene/suppressor mapping of cancers and targeted gene therapies to cure them
Of course, I like to remember it all started with some guy tucking his sick chicken under his arm and heading for Manhattan.