Cancer cannot be eliminated by evolution.
Most cancers occur after humans have finished breeding, so any genetic predisposition to cancer is not part of the natural selection process.
The other factor in the development of cancers is environmental exposures. It's a stark fact if one is a smoker, they are 20 times more likely to develop lung cancer than a non-smoker.
I don't know what the situation is now, but back in the seventies humans were creating about 12,000 new chemicals a year. It is inevitable some would be carcinogenic. However, association of a particular cancer with the causative carcinogen would take decades to surface.
Example: b-naphthylamine was first isolated from coal tar in the mid- 1800's. It was used in rubber as an anti-oxidant.
By the 1930's, medical researchers started noticing high levels of bladder cancer in rubber workers. It was not until 1974 that b-naphthylamine was officially designated as carcinogenic.
Humans have been consuming alcohol for millennia. It did not get designated as a Class 1 carcinogen until 1989. Back in the thirties and forties, mothers were encouraged to drink stout while pregnant.
Then there is synergism, the enhanced effect of a combination of chemicals. IMO it's going to take AI to sort that one out.