Personalized medicine is the new Holy Grail of cancer therapy. The drugs we currently use are stupid. They don't actually target cancer cells, they target rapidly dividing cells. Cancer cells divide rapidly, which is why chemotherapy works, but plenty of other cells in our bodies divide rapidly, and that is why chemotherapy causes so many side effects.
Not only that, but not every tumor of the same type responds similarly to the same chemotherapy. For breast cancer, some tumors respond well to cyclophosphamide and doxorubicin, but others do not.
Cancer is, at its root, a genetic disease... meaning that changes in the genes within a cell cause it to transform from normal to cancer. A lot of recent work has gone into identifying the specific mutations that lead to a particular tumor type, and using this information to gauge risk and make treatment decisions.
A by-product of this work has been the development of so-called targeted therapies... drugs that interfere with the abnormal function of a mutated enzyme, for example. Because these drugs act only in cells that have that particular mutation, which presumably only happens in tumor cells, they are thought to be more specific and less prone to side effects. The ultimate goal, then, of personalize cancer medicine would be to identify mutations in an individual's tumor and prescribe a regimen of targeted therapies that are specific for that tumor. Not for the type of tumor... for the individual tumor.
An article in today's issue of the New England Journal of Medicine makes it clear that this approach is going to be more difficult than previously believed.
Most work aimed at identifying mutations in a specific tumor is based on a single biopsy of the tumor, the idea being that the important mutations will be present in every cell in the tumor. Today's article addressed this issue directly by comparing the mutations found in multiple different biopsies from the same tumor. What they found raises serious concerns. Only about 1/3 of the mutations this group identified were present in every biopsy specimen from the same tumor. The other 2/3 were found in only a subset of the biopsies.
At one level, this is not news. Cancer scientists have been aware for years that tumors are heterogeneous... that is, not all of the cells are the same. It stands to reason, then, that not every cell will have all of the mutations. That there will be some cells with fewer mutations, and some with more. And today's article does suggest that some mutations, probably the ones important for the original development of the tumor, are found throughout the mass. However, if only a single biopsy is performed and used as the basis for making treatment decisions, most of the identified mutations will NOT be common throughout the tumor.
It's not the end of the world, but it does mean that developing and testing this kind of treatment approach is going to be a lot messier than people have previously believed.
A Smarter War on Cancer
Is the Medical Community Complicit?
Genomic Medicine -- The Hope and The Hype