Breast Cancer and Genomics

We have known for years about BRCA mutations in breast cancer, fairly rare genes (1 in 1000) that carry a high risk for a woman developing breast cancer, and developing it early.  (I have one in my family.) Once a woman is known to have one of these her physician puts her in a different protocol, with more frequent mammograms, more education and training. But in the last few years full genome studies of tens and hundreds of thousands of people have revealed another category, ten times as large as the BRCA group that also has highly elevated risk.  This is because with that many samples and machine learning, identifying and adding together many SNP's of much smaller risk alone, but now in aggregate. Your doctor is likely not yet aware of this because it is still not long after the laboratory result, and it can take as much as a decade to integrate into regular practice.  But it's there and it's coming. More early detection is coming.

You may want to pass this along to anyone interested in breast cancer info.

Steve Hsu of Manifold (sidebar) had a Q&A about genomic prediction of disease about a year ago. It can be found as a podcast here, or on Youtube here. The key part about prediction begins after the 15 min mark on both and goes for about ten minutes more. They ran a standard medical cost benefit of what it would cost and save to take a full genome of every woman at this point, and found that it is already a bit beyond the break-even point.  And of course, you would now have the full genomes already in hand to apply to preventive medical procedures for any other heritable diseases. There isn't a good way to implement this in America that has a hodgepodge system, but it is likely that Britain's NHS or something similar will be moving to do this soon. 

Additionally, it is now starting to be used for improved embryo selection in in-vitro fertilization as well, and has an implantation rate of 73%, which is enough of an improvement over the current method to result in about 100,000 more births a year. (People give up because of money or discouragement after a few failures of implantation.) So you might want to pass along this podcast to anyone interested in IVF as well. Not a topic I have kept up with much, but I know some people care a great deal.

People get spooked about full genomes of everyone for worries about who will have the information, but the cat is long out of the bag on that.  You wont be made to by the government, but insurers may start to require it, and you can increasingly be identified via your relatives, even those distant enough that you don't know them. We ate that banana years ago, as Wally said in "Dilbert." People also get very nervous about embryo selection, because of the designer baby problem more than the disease-prevention problem. (And the embryos that don't make the cut are regarded as humans by some.) No company is doing selection for cognitive features at present, but an improvement of even five points in a generation will create a separate population pretty quickly when it's in effect for three generations. As I believe that not only IQ but such things as conscientiousness, determination, regularity of looks, and even charm might be heritable, it could be a big change.

But even if you don't opt in, someone will. China, probably.