Joined: Jan. 2006
Again, on the off chance this is a serious inquiry, cancers are driven by some of the very factors that drive multicellular "lifestyles." It's very hard to be a multicellular animal--with lineages of generalized cell-types that divide and divide and specialize and specialize, requiring controls on what aspect of each cell's "total" genetic instructions to deploy, and when to deploy them--and then to turn around and turn off the division and say, "Okay, boys, that's enough now. Just sit there on idle, doing nothing but what you've been told--or, better yet, just die!--until you hear different!"
Cancers "take advantage" (by means of mutation and other errors) of the machinery of multicellularity that allows cells to live in concert, multiplying, specializing, mutually signalling and communicating, exporting and importing products of cellular metabolism, etc.
Because multicellularity is the result of an evolutionary process (that is, not designed with an eventual goal in mind, but a product of adaptation, exaptation, jury-rigging, trade-offs, and compromise among competing propulations of cells and cellular control mechanisms), it doesn't always work perfectly. Some of the balances between growth and termination of growth, supplying nutrition to "healthy" cells and denying nutrition to "unhealthy" cells, etc., can gang aft aglay. Usually it takes several different mutations or insults to derail the cellular maintenance "machinery," and then further mutations within the population of "unhealthy" cells (growing where and when they're not supposed to) to avoid the body's anti-cancer responses.
Just as the mother and embryo can be usefully looked at as in a state of evolutionary competition (at the same time as, and as well as, they should be looked at as cooperating) for scarce resources, healthy cells and cancer cells are also wrapped up in an evolutionary "arm's race." Those cancer cells that escape the body's control mechanisms--refusing to die, shut off, or express themselves in whatever limited fashion they are "supposed" to--no longer have the "interests" of the overall animal in mind. As the group of cancer cells grows and divides and further mutations within the cancerous lineage of cells generates still more variation--the cancer itself can evolve to evade the body's various defensive techniques.
If a population of unwanted rabbits (in Australia, let's say, where they are pests without adequate natural controls on their growth) is exposed to an effective poison, there will almost always be a few rabbits who happen to be somewhat or entirely resistant to the effects of the poison. Widespread application of the poison will initially reduce the rabbit population dramatically but--unless the poison is utterly effective--the population is all too likely to rebound. And the rabbits within the rebounding population, being descended from resistant or immune survivors, will be less affected by the next round of poison application.
So it goes for antibiotic resistance.
And so it goes for lines of cancerous cells. The body's controls work well, most of the time, to ward off cancers until the animal or human is mature enough to reproduce. Most--not all, but most--cancers are diseases of advanced age: because any given animal's ancestors have survived long enough to reproduce and support their offspring toward maturity (but have not necessarily survived much longer), cancers are fairly rare in the young. (Another way of phrasing this is, that the young who do contract cancer, and fail to fend it off, mostly do not survive to reproduce.) Because cancers that attack the body late in life do not face the same selection pressures (that is, the failure of an animal to survive a late-developing cancer will have little impact on the relative survival of that animal's offspring, which may already be grown and having its own offspring), many late-developing cancers have eluded the pruning shears of natural selection.
Remember, there is no "goal" of evolution where only the nice, disease-free, cute, warn'n'fluffy animals get to survive. Predators survive too, if they gobble enough prey. Bacteria and viruses survive too, if they are successful enough at colonizing and infecting their hosts.
And cancers survive too, because--if the "right" sequence of mutations and other propitious (for them) events occur, they can successfully out-compete the cells of their host body.
While cancerous cell lineages mostly do not survive the death of their "host," and so do not propogate in the usual evolutionary manner past that cut-off (though, up until that time, their growth and reproduction within the body can be analyzed from an evolutionary perspective), the body's anticancer mechanisms cannot be "perfect" (nothing in nature is perfect--everything comes with a metabolic "cost" or trade-off; absolute perfection of those mechanisms might well, for example, seriously inhibit the very processes that enable multicellularity; bodies tend to do a reasonably-good job defeating cancers of youth, but cannot, for the reasons previously suggested, easily "inherit" the ability to stave off cancers of later life), the same general types of mutations, break-downs, and failures tend to arise again and again in each new generation of animals, allowing a new cycle of cancers to get growing.
There are, of course, as one would expect from an evolutionary perspective, a few cancers which can "jump" or be transmitted from host to host--see a couple of interesting posts on Carl Zimmer's blog "The Loom," in this regard.
The last issue or so of "Scientific American" also has a good round-up article on the general topic of "Why we get cancer." That might be a good place to go next...
That should make for a start. Beyond that, others may be able to share some specifics, or you could take those suggestions and run with them (i.e., independent study--something that the internet enables with incredible facility).