Friday, September 19, 2008

LRRK2, NYT, Trust and Sergey!

What I find interesting is how a SNP finding and a blogger can set the press a fire.....especially when the blogger is a junior member of the billionaire's club.


In this case I am speaking of course about Sergey Brin who recently published a blog post on his risk for Parkinson Disease. It helps understand why his wife launched 23andWe......

This research would have great potential if the Company 23andMe would obey the "rules and ethics" of common human research by allowing patients to remove their samples from 23andMe's database if tehy so chose.....and more importantly, consenting a patient when the sample (which 23andMe would now own) would be tested for other things...

23andMe's mission is to be the world's trusted source of personal genetic information

They have a long way to go........

That being said....I think we should talk about LRRK2 and what it means for Sergey and others who carry it. From Sergey

It is clear that I have a markedly higher chance of developing Parkinson's in my lifetime than the average person. In fact, it is somewhere between 20 percent to 80 percent, depending on the study and how you measure," Brin said.

That is 20 percent to 80 percent higher than the average risk......Not 80 percent likelihood to develop Parkinson Disease. That 80% risk is the case for BRCA carriers in breast cancer, but not this LRRK2! This was a statement on his post which some people may get confused by...so don't.

The New York Times reporter did

Mr. Brin, who made the announcement on a blog, says he does not have the disease and that the exact implications of the discovery are not clear. Studies show that his likelihood of contracting Parkinson’s disease in his lifetime may be 20 percent to 80 percent, Mr. Brin said.

Parkinson Disease affects approximately 1% of the population by age 65% and 4 to 5% by age 85 years. Therefore the lifetime risk is 2-5%. So a 1.2 to 2.1 Odds ratio would be 4% to 10% roughly. Not 80%!

There are several genes that have been linked to PD, including alpha synuclein, PARKN, PINK1, DJ1 which have all been linked to familial PD. LRRK2 is a realtive newcomer to the field with good publications starting in 2005. For reference the first BRCA publications were in 1994. It took us 10 years to really be able to fully understand and explain what the risks are. We still have problems with this though.

There are several misconceptions about BRCA as there are for LRRK2....worse yet, counseling for LRRK2 can be confusing. Which is why perhaps


Because there are only a small number of genes which are known to have a very substantial effect on health (e.g. 10 times the average risk)
I felt the possibility of discovering something very important to my health was just a hypothetical exercise. So, when my wife asked me to look up G2019S in my raw data (23andMe scientists had had the forethought to include it on their chip), I viewed it mostly as entertainment.....

LRRK2 is not one those genes that increases your risk by tenfold...

LRRK2 mutation accounts for 5 to 6% of familial PD and 1-2% of sporadic PD. Not exactly what I would call useful for a screening test. Mind you this is given for North Americans and Europeans.

These numbers however do change for a few ethnicities including North African Arabs, Ashkenzi Jews where it goes up to 29-37% of familial cases and 7-14% in sporadic....some studies say as high as 41% in sporadics, but the high sporadic data is not replicated.

What is familial PD? Well, at least one first degree relative must carry a diagnosis of PD.

I am not certain of Sergey's Ancestry.....perhaps he is African Arab or Ashkenazi Jewish? That might put his likelihood of having a mutation in LRRK2 higher.....

Interestingly in the largest study of Ashkenaim with PD to date the odds ratio to develop PD was 5.77 which is very high and fairly significant. Unfortunately, these data ONLY apply to Ashkenazi Jews from Israel as this was the population studied.

So don't go thinking it is that high for everyone. In this study, the Odds Ratio for developing PD in Ashkenazi patients with the LRRK2 G2019S mutation was 5.7 (CI 2.8 to 11.7). It is increases when you match age and sex to 8.6. Which has us thinking perhaps this is a sex effect? Not that sex, you perv!

What is the percent of unaffected carriers in the general population of Ashkenazi? 2.2%

That is a scary stat for my Jewish friends. That means this mutation is not nearly as common in the unaffected Ashkenazi Jewish population. This should be only interpreted for this Ethnicity! Not for Europeans or for Americans who do not have Ashkenazi ancestry!!

It was crazy that none of the other mutations in LRRK2 were detected in this Ashkenazi cohort. That certainly speaks for a founder effect!! Similar to the founder mutations in BRCA1/2

In dissecting out thet data a little more we see that the G2019S was found more commonly in Women and in Familial Cases for the Ashkenazi population.

So, I hope that helps. These data and perhaps the numbers Sergey quoted were likely those for Ashkenazi Jews. So if you are Tom Smith from Nebraska, chances are that the LRRK2 data you now could afford to get are not so dire. And even if you are Jewish, the penetrance of this mutation is 24 to 85%

Sergey's post was great and highlights a special issue. Ethnicity and its role in disease. We cannot apply one ethnic groups data to another. A study done on the Chinese does not apply to Northern Europeans. This is why SNP data can mislead and scare.

Here's the big question, when you found out you might get Parkinson Disease Sergey, were you alone? Would have liked to have someone there? Would you have liked to talk with a geneticist? Or was it like Joanna Mountain said to a friend of mine "Aww, Come on, this testing is just for fun"

The Sherpa Says:

At Helix Health of Connecticut, patients are seen that have family history of Parkinson Disease. I believe that all asymptomatic testing of this nature can be confusing and requires appropriately trained consultation....plain and simple. What if Sergey didn't have billions to through at this disease in hopes of a cure and to give him hope? He might be hopeless. In that case he might have even contemplated suicide......in the privacy of his own CPU......

7 comments:

Alberto said...

Thanks for clarifying these points so quickly! I was quite surprised when I read that "20-80%", it can be misleading indeed! I also think his wife's "23 and me" will get a lot of advertisement because of his statements (weird incipit for a blog in any case!).

Kevin said...

Sergey was born in Moscow to Jewish parents, so it is pretty likely that he has significant Ashkenazi Jewish ancestry

Steve Murphy MD said...

That misses the point Kevin. Even if Sergey is an Ashkenazi Jew, the highest risk is in African Berbers.....which I doubt he is.....

But the point was.....the press reprted his blog as if it were fact, with no data checking, no scientific opinion....

That is bad journalism, and it is precisely what is killing personalized medicine.

-Steve

Anonymous said...

"That is bad journalism, and it is precisely what is killing personalized medicine."

Agreed. I write articles about genetic health conditions for a hobby (full-time side job) and I never write a stat without using a primary source.

IQ said...

Definitily, Mr Brin is not a genius, but a lucky guy in the right moment at the right time.

Ashley said...

I would agree with IQ to an extent. I have no idea how intelligent Brin is, but it doesn't seem to me that the Google search engine is particularly amazing in programming terms. There are things like 3D video game engines which are far more sophisticated in terms of programming techniques than something like a search engine, which is, after all, only a glorified text parser. An international search engine is big and needs a lot of power to run it, but is the code that runs one as advanced as that which runs a game such as Halo 3 or a 3D Modelling Software such as 3DS Max? Brin and his colleague came up with a good idea in that they decided to rank search results based on the number of links to the site rather than the number of text matches. That's basically it. They weren't pioneers in some sort of technological breakthrough and they didn't invent search engines, despite the fact that everyone acts like they did. They didn't even invent the interface for Google. Altavista had virtually the same interface and still does, and God knows who else. I once had a class in Web programming in which the lecturer was crediting Google on their interface, which made me sick because they stole it.

Beth Corder, PhD said...

I am writing as a well-published genetic epidemiologist and as the scientific director of Matrix Genomics, Inc. Concerning LRRK2, close investigation of variation throughout the gene identified a distinct subset that carried a very specific very high risk signature for PD. http://www3.interscience.wiley.com/journal/122399048/abstract