Wednesday, July 14, 2010

FDA Panel Recommends Keeping Avandia on the Market

As I mentioned in my post Avandia Smackdown!! , I suspected that Avandia would likely stay on the market. The story is all over the headlines, including the Wall Street Journal's FDA Panel Backs Keeping Avandia on the Market. The 12 panel members wanted to pull Avandia, but 20 said it should stay on the market. More interestingly, though the majority of the members felt there were signals of risk, the vote was 20 to 10 to proceed with the controversial TIDE trial, which Drs. Graham and Sen. Grassley said were unethical.

I have been following the hearing closely for the past few days via Twitter and live video feeds from CNN. Here are some initial thoughts.

1. Despite the panel's concerns about Avandia safety, the clear majority felt that there was simply not enough data. Many were concerned that studies like Nissen's meta-analysis were just not strong enough. Though they favored long term, randomized trials to definitely answer these questions, unfortunately, there were enough concerns about the RECORD study by some (low events, withdrawals, and some missing data) to be convincingly reassured.
2. The sentiment of many was because Actos didn't show as strong signals, it remained a better option. Many who chose to pull Avandia stated for the RECORD that Actos' availability was a deciding factor. This is concerning because the data for Actos safety is extremely weak (not that I think Actos is dangerous). One panelist stated what I have heard before, that "the absence of evidence does not equal the evidence of absence."
3. Many panelist stated that they were putting on their "public health" hats, meaning that even though scientifically they were not convinced of real harm, because there was a possibility of harm, they voted to remove or proceed with caution. This is VERY important, because as a clinician (which many on the panel were not), you have to balance risk and benefit every day. Is the side effects of a particular medicine worth the benefits of the medicine? Is the potential harm of radiation worth the need for a CT scan? I think if more practicing doctors were on the panel, fewer would have voted to remove the drug.

What now?
Ultimately, the FDA will decide what to do. Since the FDA doesn't have to agree with the panel and especially since the panel seemed split, the FDA could decide to remove Avandia anyway or keep on the market with certain restrictions. Given that 10 voted to have very strict warnings, it is likely that's what the FDA will do. These stricter warnings will likely include something like requiring only a diabetes specialist, i.e. an endocrinologist be able to write a prescription for Avandia. This language will be crucial, because even if Avandia stays on the market, if the restrictions are tough enough, no doctor will ever write for the prescriptions.

Despite all the holes that Avandia's opponents poked in the data, I remained convinved that the preponderance of the data points in Avandia's favor. It doesn't appear to cause increased heart attacks, it certainly doesn't cause increased death, seems to decrease stroke, and clearly decreases the use of insulin. I will continue to write for the product unless it is pulled from the market, newer restrictions make it virtually impossible to prescribe, or my patients request being placed on a different medicine.


Rich Meyer said...

While I agree that good science wins out over innuendos there are elements here that are very troubling. First, if there were in fact clinical trials with incomplete or false data how could the FDA rule that Avandia is safe and effective ? More importantly how can the FDA convince people that they are acting in their interests not the drug companies interest ? The real news story has been GSK's hiding clinical trial data and underreporting side effects. An investigation needs to be launched to ensure that data was not manipulated to meet end points. In my opinion the damage is done and I doubt that Avandia sales will maintain their current level.

Dr. Matthew Mintz said...

I agree completely with you that:
1. Damage is done, and Avandia is likely dead. I may be the only doc in the US that continues to prescribe it.
2. FDA does need to convince people it is acting in the interest of the people, which is why this has become so politicized
3. While I agree and investigation is probably needed and will certainly happen, this whole "GSK hid the data" is probably way overblown.
a. The fact that there are inconsistencies with some of the RECORD data does not mean that GSK knowingly hid any data. Several comments at today's panel were made that these inconsistencies are common and happen in all large clinical trials, with this trial being no different. In addition, the corrected data does not likely change the findings of RECORD. I have participated in clinical research. You are collecting a lot of data from all over. Record keeping (no pun intended) is very difficult.
b. The new NY times report of a "hidden" comparative study is also suspect. It was likely a small, head to head study looking at Actos vs. Avandia. I highly doubt that there was a statistically significant difference in heart attacks between the two drugs. Numbers would likely have been too small. More likely, the study showed a small lipid benefit with Actos, which is unlikely clinically relevant, given that all diabetics should be on a statin.
That said, having internal documents stating that the data "shouldn't see the light of day" doesn't make GSK look very transparent.
I am a big advocate of more transparency in pharma. I think any phase 3 or 4 trial should be pubically available and the companies should submit the raw data and let the FDA do their own analysis. The FDA should also have oversight/quality control over all industry sponsored phase 3 and 4 trials to ensure that they are run appropriately.

James Kildare said...

suffer from a chronic disease that is the back pain and I have already four years living with it, it's hard to say but the pains are intense and I have an 8 year old son asking if I can recover and get out of my bed to go for a walk with them park as a family ...

Anonymous said...

The FDA has been confusing the public since the Vioxx saga. The issue from Avandia has been totally different with Vioxx. The indication approved for Vioxx was to treat arthritis/joint pain. Increasing the blood coaulation resulting in a vessel block was not associated with arthritis, or the complication of arthritis, but considered as a drug toxicity/side effect.

Avandia was approved for the indication of lowering the blood glucose/diabetes. Myocardial infarction/heart attack is one of the complication in diabetes. However, the congestive heart failure caused from fluid retention is not a disease complication, but is the side effect from TZD's as a class effect, and we can consider this as the drug toxicity.

There is a distinction between drug toxicity and disease complication, and the drug safety surveilance only deals with the side effect/toxicity, not the disease complication. The disease complication is an issue that related to the efficacy of the drug. If a manufacture wants to claim their drug has a benefit of improving the lipids profile in diabetic population while it was approved for treating diabetes on lowering glucose, the manufacture will have to prove this from an efficacy trial, and files an IND under new indication, per se, improve the lipids profile. I don't think that any manufacture can claim a new indication through an outcome study which has been only used for safety observation.

Since Nissen scrued up this basic principal, the FDA got lost from this misleading that was manipulated by Nissen's financial sponsor, and the entire country got lost in the middle of nowhere after Nissen's meta-analysis published. This has been the best case to show how stupid the government is.