Monday, August 4, 2008

Muscle Pain from Statins? Time for a genetic evalutaion!

In a wonderful demonstration of how successful genetic research on pharmacogenomics needs to be done. The SEARCH Collaborative Group successfully demonstrated the power of GWAS nested in other studies. This will be the hallmark of all great medication studies for the next 50 years. Do I mean GWAS? No!

I mean genetic analysis of the outliers in response to the medication being studied. I only ask myself, why in the hell hasn't pharma been doing this for the last 10 years????? I have my suspicions, but I'll never mention them in public.

So why am I so excited? know how the DTC companies test you for SNPs that give you odds ratios of 1.3 or less??? Imagine an Odds Ratio of 16 for developing statin induced myopathy!!! I was just talking to the founder of one of those sequencing companies worth billions of dollars about this the other day!!! Odds ratio of 16.9

Uhhhhh...Yeahhhh....that is a pretty good chance of getting the myopathies associated with this super common medication. How is that for immediately clinically relevant. Have the polymorphism?.....Don't take statins...instead take Fenofibrates.

This study was signed sealed and delivered to the door of the makers of Zocor(Merck). I sure hope they license that test!!!! Oh wait...this will likely be sold by deCode. Caraiso just told me that this SNP is on all of the big 3's chips! So who will contact their patients...sorry/customers? If the MD ordered the would be the MD's duty........This is why the government is trying to regulate this testing.....for this type of follow up. This is a great study it needs follow up!
How was it done? From NEJM...
Methods: We carried out a genomewide association study using approximately 300,000 markers (and additional fine-mapping) in 85 subjects with definite or incipient myopathy and 90 controls, all of whom were taking 80 mg of simvastatin daily as part of a trial involving 12,000 participants. Replication was tested in a trial of 40 mg of simvastatin daily involving 20,000 participants.

At first, I balked.....80 mg of Zocor. That's a whopping dose! Who the hell uses that much Zocor? So I tried to dismiss this study....but it was the replication in the 40mg Group that got me.....

So what did they find?

The noncoding rs4363657 SNP was in nearly complete linkage disequilibrium with the nonsynonymous rs4149056 SNP (r2=0.97), which has been linked to statin metabolism. The prevalence of the rs4149056 C allele in the population was 15%. The odds ratio for myopathy was 4.5 (95% confidence interval [CI], 2.6 to 7.7) per copy of the C allele, and 16.9 (95% CI, 4.7 to 61.1) in CC as compared with TT homozygotes. More than 60% of these myopathy cases could be attributed to the C variant. The association of rs4149056 with myopathy was replicated in the trial of 40 mg of simvastatin daily, which also showed an association between rs4149056 and the cholesterol-lowering effects of simvastatin. No SNPs in any other region were clearly associated with myopathy.

Why am I salivating over these results????

1. 16.9 is a huge Odds Ratio in a region where we know there has been lots of literature reporting the issues with Organic Anion Transporters and the bad effects of statins.

2. The replication in a clinically useful cohort.....40 mg of Zocor....there are tons of people on that dose.

3. Statin Myopathy can cause kidney failure from rhabdomyolysis(muscle breakdown). It is a dangerous side effect! Another clinically useful point!

4. There were NO OTHER SNPs associated with anything!!! The odds of that are preposterous! That IMHO means this is a very valid SNP. With everything, I suggest replication and then the marketing of this test to the physician. With the OTC Zocor in the UK....maybe they have an interest in DTC????? Or maybe not

The Sherpa Says:
Pharma are you listening? This is the way to save your medications and save you from lawsuits. The way to stay alive is by developing diagnostics that protect your patients and doctors.....If you don't follow this model, you will perish. When this study is replicated and the test available through a CLIA certified lab, Helix Health of Connecticut will test their patients for this very important genetic polymorphism.....As for the big 3 screens changing medications......Dr Agus should stick to Heme-Onc. The SNPs associated with heart attack in the SNP scan are separate from cholesterol....and statins have not been proven to lower risk via any mechanism other than lipid lowering....Either way, he would be better served testing the organic anion transporter.


cariaso said...

rs4149056 has another interesting property. It is present on the microarrays used by ALL 3 of the DTC testing platforms. If you've been tested by Navigenics, deCODEme, or 23andMe you already know your genotype at this position. And it isn't just statins, a large number of medications are affected by this snp. said...

Gene Sherpas: Personalized Medicine and You :
We found the information you post really useful, and we decided to make a review about it in our website.
You can check it out at :
Our score is :8.2

Steve Murphy MD said...

Thank I appreciate the link! I will check you out tonight!

Steve Murphy MD said...

Great point! Now we need to have these companies contact patients and make note of this for their benefits! I know the OATPs can have a heavy influence in many medications.....but the data isn't nearly as strong so far as this study. Once again....I am a huge supporter of genetic testing when we can use something with it....That being we have a few hundred people with this SNP out there. How do we contact them? What if they didn't pay for the subscription service? Don't we have a duty to follow up with them? I say yes. If the doctor ordered the test, the answer would be unequivocally yes. But in the case of the companies....they are not so sure....unless of course if you are paying them.


Anonymous said...


Should personalized genetic testing companies follow-up with their consumers when a new and important finding is published? Yes! However, there really isn't any difference between purchasing your own genetic test and purchasing a towel from Target. The company sells a product and the consumer purchases the product. The company selling the product does not need to contact each individual about the product they purchased. When consumer product recalls are done, they usually do it by website advertisement, mass mailings, and tv prodcasts (i.e., ABC Nightly News).

Anonymous said...

Major flaws presented by the three team panel!
1: “Personal genomics will explode.” No! Not if consumers have to flip the bill for the test.
2: “It doesn’t matter if the tests are regulated or not.”
3: “People can educate themselves about their genetic concerns.” Wrong. You need to understand how genetics works to educate oneself. Most people don’t even know how to pronounce DNA and you expect people to understand genetic risk?
4: “Family history is not accurate.” Ok, keep giving all patients the same amount of a medication for the same disease. Consistent medication dosage(s) for the same disease is no different (for accuracy) then a family history according to the panel.
5: “95% is preventable for the majority of people if taken preventive measures.” No! Only conditions that are operable can be treated (not prevented…prevention is the ‘prevent of something occurring’). Treatment is treating a condition and not preventing a condition.
6: “Get people to change their life style.” Yeah, that is a novel thought.
7: “Can change the way your genes are expressed in a few weeks to a few months just by how you eat.” Have you looked at the average diet of an American? Eating a lot of fruits and vegetables is not going to magically change how your genes are expressed in two or three weeks. My, oh my.

The Medical Quack said...

I also liked what you had to say as well! Great post and good information.

SciPhu said...

Reading the NEJM paper it looks as though testing for this SNP is only going to be useful for those on very high doses (80 mg daily). Isn't it reasonable to conclude that doctors should only use this test on patients considered for such a high dose ? Testing everyone doesn't look like a fruitful option (for Statins that is). Anyone agree/disagree with this conclusion ??

Steve Murphy MD said...

Anonymous Towel guy,
If you think buying a towel at Target is the same as getting a genetic are smokin something, or obviously don't know what the genetic test is being used for. Towels dry people off, hardly risky by any means....unless you have really bad psoriasis.

Genetic tests on the other hand......We the gov't just put several million in the kitty for researchers to discover the answer to that question....


Steve Murphy MD said...

The Medical Quack,
I love your site. It is an honor to be linked.

Steve Murphy MD said...

Read my post....this was replicated in a group which was on 40 of Zocor....a way more common dose....I was going to bounce the paper for the same reasons you put up there.....but they actually replicated it at 40mg.

SciPhu said...

I know it was replicated, and you may have misinterpreted my comment. So rephrasing then: Seeing that the occurrence of myopathy due to statin use is 1 in 10 000 on the 40 mg daily dose, what is the cost benefit of testing all those on 20-40 mg ? The cost benefit of such testing is more clear on the high dose, but there aren't many on that 80mg daily dose, or am I wrong ?

Jaxon said...

While doing some research on this topic I came across a very interesting article "Calf Injuries and Magnesium Deficiency." I would suggest reading it, as it offers some very interesting information on the correlation of poor health and magnesium deficiency. There's another really important article about magnesium as an alternatives to Lipitor called " Magnesium and Walking Will Always be Superior to Lipitor. Check them out if you have a chance.

Steve Murphy MD said...

You are correct. There are likely very few people on 80 mg. You may have to ask Merck for that data, but we rarely use 80 of Zocor or Crestor or Lipitor. There are way more people on 20-40. So if we look at the study, the number needed to harm at 80 is 67 patients on the drug. So the big question is what is the NNH for the 20-40........I will get back to you on that. It turns out that the snp is on all of the big 3's chips. Which means the cost to run is about 350 per sample. Or you could do targeted analysis for 150-250.......
I have to noodle it some more and get back to you... but I think you may be right.


Andreas said...

It is important to note that the odds ratios in the replication group were significantly weaker than those in the original group. This may be due to regression to the mean, or to the lower dose, or both. In any case, you will not get 16.9 in practice.

Even if you did, the phenotype they measured occurs only in 1.6% of cases in the 80 mg group, much less in the 40 mg group. What is worse, the placebo group in the replication had 12 cases, vs. 23 in the 40 mg group, meaning only 50% were attributable to statin.

Lastly, did you have a look at the definition of that phenotype? There are 49 cases of "definite" myopathy, defined by muscle symptoms plus 10 times upper limit normal CK. That is ok, although the literature is full of doubts about how relevant CK really is to statin muscular side effects.

Because 49 cases were not enough, they also included 49 cases of "incipient" myopathy. To be incipient, you did not need to have muscle symptoms. You needed to have 3 times elevated CK, not vs. ULN, but vs. baseline instead. Where it really gets strange is you also needed to have 1.7 fold elevated ALT baseline (a liver enzyme having little to do with muscle as far as I can tell). Why ALT? why 1.7? They do not explain. Lastly, the ALT only counts if it did not occur previously unaccompanied by CK elevation.

If I ever saw a contrived phenotype, this is it.

So, under the best of conditions, among those taking 80 mg, you catch only those 1.6% with the above collection of symptoms. In practice, rates of muscle commplaints range from 5-20%, so you are missing most of them. Furthermore, because the homozygote is 3%, only 50% of your positives can possibly be true. And this is not even considering the much weaker odds ratio in the replication group.

Clinically useful? I'd reconsider that.


Steve Murphy MD said...

@ Andreas
Where it really gets strange is you also needed to have 1.7 fold elevated ALT baseline (a liver enzyme having little to do with muscle as far as I can tell). Why ALT? why 1.7? They do not explain. Lastly, the ALT only counts if it did not occur previously unaccompanied by CK elevation.

Andreas, ALT is an enzyme ALSO in muscle tissue. FYI. That makes some sense me as to why it was included.

Lastly, the replication is key. As I said in my wrap up. If there is replication, it will be a useful test in those on this medication. Do you know how many people 2% is? A ton.

Replication will be key.


Andreas said...

Why include ALT in the definition of "incipient", but not "definite" myopathy? Why 1.7? Why is this not justified in the text? Is this some commonly known clinical definition that I am not aware of?

Replication is indeed key. However, in replication the marker was a LOT weaker than the original result. Use the replication odds ratio to judge the usefulness of the test, and it will come out even worse than I have so generously portrayed it.


Zocor Side Effects said...

My name is Wendy Rosko and i would like to show you my personal experience with Zocor.

I am 60 years old. Have been on Zocor for 8 days now. MD stopped med when I reported muscle pain and sent me for a blood test to measure muscle inflamation. Off med for 2 days and still having muscle cramps in arms and legs. No generalized aching.

I have experienced some of these side effects -
had leg cramps which I didn't realize were a side effect. On day 8 I woke up feeling as if I had the flu - bad muscle aches and cramps.

I hope this information will be useful to others,
Wendy Rosko