The Economist had an article on disease causing genes in a recent issue.
Common sense tells us that
… the older a gene is, the more likely it is to be part of the irreducible structure of being alive… Another reason for expecting that disease-related genes would be recently evolved is that the older a gene is, the more likely it is that errors and weaknesses that could lead to disease will have been eliminated by natural selection.
But the research the article describes finds that these genes are, in fact, very old.
…the researchers found that the majority of disease-causing genes were present in single-celled organisms and that most of the rest arose when multicellular creatures began to evolve. Genes specific to mammals, by contrast, barely ever carry diseases.
[The researchers] do not have an explanation for why genetic diseases seem to be caused so disproportionately by old genes, but their discovery does suggest that such diseases are an inescapable component of life which even evolution cannot get rid of.
Sounds like a legacy system issue. Very difficult to resolve. I would suggest to evolution that rebuilding from scratch may be the only way to proceed.