Now that everyone who wants to sound sophisticated about health care is talking about “personalized medicine”—even the president!—the open question is, what part of medicine is going to get personalized first? Which is to say, what can the intricacies of people’s genomes say about the best way to treat what ails them? The answers may well be in blood.

Blood, you see, doesn’t just come in types A, B, AB, and O. The “positive” or “negative?” Nope. In fact, let’s get all the way into the weeds: Scientists have since discovered over 300 proteins that contribute to blood type. The AB+ on your blood donor card? Yeah, that’s a massive oversimplification.

To be fair, that massive oversimplification for blood transfusions has worked pretty well for the past half century—except when it hasn’t. People who need multiple transfusions, as with sickle cell anemia, or people who have flat-out rare blood types need more precise blood typing. Even your average otherwise healthy patient could benefit in certain cases.

And the most precise blood typing you can get is with DNA. Blood genomics has come up in Europe, and this year in the US saw the creation of a new National Center for Blood Group Genomics. “It’s one of the first areas where you can really implement personalized medicine.” says Connie Westhoff, Director of Immunohematology, Genomics, and Rare Blood for New York Blood Center, one of the groups behind the national center.

Why is blood a good candidate for genetic personalization? It helps that how genes contribute to blood types is a bit more straightforward than, say, cardiovascular disease. A, B, and Rh (that’s the thing that gives you the positive/negative distinction) are all gene-encoded proteins that stick out like flags on the surface of red blood cells. When the immune system sees a foreign flag, it attacks—in which case the transfusion backfires. Now A, B, and Rh are the most prominent flags, so they are the most important. Behind them are about 35 other key proteins, which Westhoff says the National Center will test for in every single unit of blood they receive.

New Blood

Major blood centers sometimes do test for a battery of additional blood proteins. But it only makes economic sense for most centers to test blood for people who frequently get transfusions, like those who have sickle cell anemia. Getting a lot of transfusions can “sensitize” these patients to minor blood proteins, setting their immune systems on high alert for proteins that a first-time patient’s body might ignore. So matching an AB+ patient to an AB+ donor isn’t enough. “You probably only want to test group O donors,” says Jed Gorlin, medical director of Memorial Blood Centers, which is not affiliated with the National Center. People with blood type O blood, you may recall, are universal donors—though of course now doctors know that less common blood proteins can limit “universality.” Yup, it’s complicated.

It gets even more complicated with Rh-positive and Rh-negative. And here is where universal blood typing with DNA could help a larger pool of recipients—like any woman who ever plans on getting pregnant. Rh is actually not binary, positive or negative, but 50 different types of blood protein. (D, C, c, E, and e are the most common ones.) The positive or negative designation only refers to D, and it’s also possible to be neither positive or negative but “weak” or “partial” D. Confusing much? Well, these variations don’t register reliably on the typical lab test for Rh antigens either. Typing via DNA, on the other hand, cuts through the thicket and tells you exactly which Rh gene you have.

So, if a woman gets a blood transfusion with a mismatched Rh, she could become sensitized to that protein. If she later gets pregnant with a fetus who has a different Rh type—totally possible—her immune system, sensitized to search-and-destroy other, non-self Rh types, could attack the fetus. It’s a dangerous and sometimes fatal to the fetus.

Eventually, Westhoff wants to do even better than just looking for 35 markers. She recently coauthored a paper showing proof of concept for using next generation sequencing to type blood. By using next generation sequencing on all coding DNA, the group could look at all 300 blood proteins at once. The problem, though, is how you prioritize which proteins are most important. “We’re developing the algorithm to translate next generation sequencing for all these markers,” says Westhoff. One of the reasons for starting the National Center is to develop the computational knowledge and IT infrastructure that makes it possible.

The trouble with next generation sequencing for personalized medicine is that it’s expensive, and the payoff is uncertain. But you don’t need to get all the way to 300 blood proteins to get more personalized and precise blood typing. Thirty-five is already better than three.


Beyond Blood Type: Genomics Can Show What You’re Really Made Of