September 2, 2011

Pharmacogenetics and Race in the Medical Record

There’s a discussion on a physician site I subscribe to about whether ‘Race’ should be included in the general demographics information of electronic medical records.  Here’s the reason it should.  Now that we are every so slowly receding from the political correctness of the last couple of decades of the 20th century, medical researchers are finding out that there are differences in both disease risk and care for certain groups of people.  We’ve know for years that your child’s chances of certain genetic conditions were different based on your and your reproductive-partner’s genetic background.  Everyone has known that children of Ashkenazi Jewish heritage (more or less of Eastern European background) were at higher risk for a terrible genetic disease called Tay-Sachs.  Dark-skinned people of African descent were at higher risk of sickle-cell disease; white kids of cystic fibrosis.  Finally, in the last decade or so, researchers have found the freedom to study some important medical problems, and they have found out important things.  Here’s a good summary of some findings in the growing field of pharmacogenetics, which looks for differences in the response to different medications.

Racial differences in response to cardiovascular medication.  [It’s pretty readable for non-doctors.]

It has been found, generally speaking and for example, that there’s a difference in response to a common blood anti-clotting medicine in Black, White, Hispanic, Asian people.  This could be really important to know.  Though not a cardiologist, I have the impression that treatment guidelines have incorporated some of these findings, and you might get a different dose or medication depending on your ethnicity.

Have I sold you on this? 

We have always known that different people respond differently to a given medication or dose.  Sometimes doctors have to try different formulations or doses to get the therapeutic response that will help the patient.  Clearly, there’s something about the way that patient is metabolizing the therapy that is special.  Presumably, it’s genetic.  Note—it might not be.  Grapefruit and grapefruit juice, for example, is a very potent inhibitor of (and this is common knowledge) the CYP3A4 form of the P450 enzyme.  A lot of drugs get digested by this enzyme.  Blood pressure drugs, heart rhythm drugs, cholesterol drugs, sleeping pills, anxiety drugs, antidepressants, antihistamines, some birth-control pills, some immune-suppressant drugs and some anti-HIV drugs.  Chances are really good that you or someone you love is taking a medication affected by grapefruit, especially by grapefruit juice.  Organic or not.  Did the people studying these drug-responses ask their subjects about grapefruit juice intake?  I don’t know. 

I’m bothered by the idea of having an ‘Ethnicity’ in your child’s medical chart.

So far, a lot of the research in this emerging field of Pharmacogenetics (also Pharmacogenomics) isn’t.  It’s really PharmacoRacialist research.  There’s painfully little genetics and lots of studies of ‘racial groups.’ 

I’m not trying to be politically correct here.  I wouldn’t doubt that epidemiologic research done in a remote Chinese village is probably representative of a relatively limited genetic pool.  True genetic research in the U.S. has depended for decades on relatively closed and technically inbred communities among the Amish.  One of the reasons behind the excellent and expert genetics professionals in Utah is precisely the multigenerational stability and consanguineous inbreeding available for research.  Genetics research looks for a link between subjects.  A lot of this current research has only an ‘ethnic’ label as the common link.

Partly, it’s the government’s fault.  Seriously.  The Census is probably the most important way our tax dollars get divided.  And, perhaps with good intentions, the government wants to know about the ‘racial’ makeup of this country. 

As far as I can tell, this is the latest official information from the Office of Management and Budget, which manages the Census.  

US Census Directive 15

Everything is going to be bases on the census, and the racial distinctions are sociopolitical, no longer genetic or medically-based in any way.  If you think certain groups need to be kept track of, write your congressperson. 

From a medical ethics point of view, there is clearly emerging data suggesting that different treatments or approaches might be optimal for those with different genetic ancestries.  These distinctions are unlikely to overlap much with meaningless Census distinctions, such as 'Hispanic' (which, after years of debate, has apparently been changed to “Hispanic or Latino.”  The 'Asian or Pacific Islander' category will be separated into two categories -- "Asian" and "Native Hawaiian or Other Pacific Islander."  Hey, at least the islanders live on islands for pete's sake, at one point in the 19th-century genetically isolated in the Darwin-in-the-Galapagos sense.  My Korean and Chinese families don't think they are Japanese.  What about Filipinos or Indonesians where different islands can mean different origins.

I have a family with 2 Chinese grandparents on one side and 2 grandparents from Portugal on the other.  They recently immigrated here from Brazil--what are they?  What are their 2 adorable kids?Are they Latino because Portuguese is a Romance Language?  Hispanic because they are from South America?  Asian because the kids have those sort-of-Asian eyes?

There’s an easy reality check.  Ask the Spanish-speaking people you know.  Believe me (and the US Census), they are all around you.  When I lived in Utah, which I considered to be the least ethnically-diverse place I had ever lived, we had a medical clinic that was pretty much 90% Spanish-speaking.  These families lived right there among, well, you know who, in Salt Lake City.  Here’s what I have noticed.  Ecuadorians hang out, mostly, with their fellow Ecuadorians.  They may have travelled through Mexico to get here, and they speak the same language (to me).  When a friend from Venezuela went out with me in Boston many years ago, he could tell immediately that people we met were speaking with accents from Peru or Guatemala or Cuba.  I’ll say he could tell if people came from Puerto Rico, but he said that he couldn’t understand much since they spoke so fast and dropped so many syllables.  He sometimes complained that he was too impatient to wait for a Colombian to finish the sentence.   He could tell when they started speaking—but not by how they looked.  It was a revelation to me when I attended the wedding of a distant relative in Paris.  Also there were relatives from Nice (on the Mediterranean coast of France).  My relatives said it was impossible to talk to them because “…they speak with such a heavy Southern accent.”  It took me a while to figure this out.  It was hard for me to reconcile Hercule Poirot and The Dukes of Hazzard.

poirot Dukes_of_Hazzard_Wallpaper_10_1024[Tasteful?  Professional?  Oh come on, what did you really expect?]

I think this is an ethics discussion because even with demonstrable medical distinctions, how does the working doctor categorize the folks sitting in the exam room, and their variegated offspring?  Do we wing it on skin pigment (I all shades of dark kids and light kids from curiously mixed parents). 

My personal opinion is that the idea of race is a sad vestigial construct without helpful meaning.  Here's what I might need to know:  have any of the child's relatives had sickle-cell disease?  Cystic Fibrosis? Hemophilia?  It wouldn't help me to have a place on the chart for 'ethnicity.'  But in the Social History section of my computerized charting program, I will note if a child is adopted and from where.  I will note there if there are medically-relevant familial risk factors. 

And what about those medical distinctions I noted above that ‘everyone has known for years?’  Even if this were true for one very long historical era, until World Wars and intercontinental migration, it isn’t very true here in California.  When I first moved here from Utah in 1999, I made a comment to one of the hematologists at the local children’s hospital about not seeing much sickle-cell disease in Utah.  She said she had a clinic full of white kids with sickle-cell, and I felt as ignorant as I was.  Here in California, every baby is screened for sickle-cell at birth (since 1990).   Everyone has known, since 1885, at least, that Mongolian Spots (benign bruise-like markings often appearing on the lower back of babies) were much more common in Asian children.  This was first noted, as a fact, by Dr. Erwin Bälz [I couldn’t make this stuff up!], the German-trained personal physician to the Meiji Emperor.  It was only around 100 years later that these were found to be correlated with skin pigment—the more pigment you had, the greater the chances of a mongolian spot.  But they are harder to spot in babies with very dark skin.  So they were noticed more in babies of Asian descent who had more pigment that Laplanders but less than equatorial Africans.  Ira Gershwin said it best:  It ain’t necessarily so.

It turns out that the whole ‘ethnic’ idea is largely confounded by genetics.  It’s certainly true that my ancestry is tough to follow prior to immigration from southern Russia.  Nobody kept track of peasants and who they mated with.  Still, there are lots on great kings who have kids with serving wenches.  And from there, it’s all down hill straight to me.  Which leads to this:  I’m descended from Charlemagne.  Yes, that Charlemagne.  Perhaps less likely an ancestor for you if your family recently came from Africa or Asia.  Your folks came from Africa in the 19th-century?  We can both attend the Charlemagne family reunion.  Just arrived?  Maybe not Charlemagne, but definitely Nerertiti.  Just like me.  Read this article from The Atlantic Magazine.  It will likely affect your view on race forever.  You can’t judge a book by its cover.

Because I don't think racial labels are generally helpful, I also believe that they are a potential problem. I wouldn't want my calling a patient 'white' when they are 1/16 Native American to interfere with their self-description. This may have legal implications, perhaps for Federal contracts or programs. Will it help them or hurt one of my patients in gaining a scholarship to define their race in a certain way, especially if whatever institution looks to their physician or medical record for verification? If I don’t think I, a fairly observant physician, can tell what race you are by looking at you, I certainly don’t have confidence that—even with training—college students working for the summer can categorize your ethnogenetic heritage when they come to your door to help with the census and try to give you a careful look in the hallway outside your apartment door by the dangling blinking buzzing insufficient fluorescent light that you’ve been bugging the cheapskate landlord to fix for months.