Lithium, the GOAT: DISCUSSION: Big Pharma's Influence on FDA Approval

Ethan: Dr. El-Mallakh, I really respect him because of how he teaches his students, pushes people to use clinical thinking skills. And evidence of that is in my 1^st year of residency, we were sitting in the ER and Dr. El-Mallakh, you asked me a question about why I wanted to use a certain medicine, and my answer was, ‘Well, it's FDA-approved for this condition,’ to which you said, ‘That's one of the worst reasons I've ever heard somebody say why they want to use a medicine.’

Rif: Unfortunately, it is. It's further unfortunate that FDA indications are becoming more potent. In medicine, it used to [00:01:00] be FDA indications were just the way into the marketplace. And then, clinical science determined the full range and scope of what a medication say, can do. And now, mainly because of changes in insurance and reimbursement for insurance, and some of the pushback that the FDA has done, FDA indications are becoming more important and clinicians get criticized for using things that don't have an indication. But, it's always important to understand that the only people that decide what drugs are gonna be indicated are NOT scientists, they're NOT clinicians, they're NOT physicians, they're business folks that are running a business company and they really make a business decision based on things like cost of development of a drug for an indication and. Market size and how much they're likely to make. So the [00:02:00] FDA indications are exclusively, and that's not an exaggeration, exclusively driven by market, not by medical need.

Ethan: Yeah, and that's turning into a big theme of the podcast. We have got to find a way to separate big business interests from medicine, similar to the (supposed) separation of church and state within government. It's really important that we have business interests and big pharmaceutical interests to help drive new drug development and discovery, but at the same time, when the primary outcome measure is profit, and the scientists are not able to do good science to determine what is efficacious and what is safe, we're in the situation we're in now where our healthcare outcomes are pretty piss poor, despite the fact that our healthcare spending almost doubles that of any other nation.

Ethan: It's [00:03:00] absolutely true. And I don't know what the solution is because America is the main innovator in healthcare and it is the main innovator in healthcare because of the size of the market and the potential for profit. That is something that's good. And, taking that away will reduce the likelihood of getting particularly groundbreaking treatments. For example, Nuplazid (Pimavanserin) is $56,000/year, because it's an orphan drug. You don't have to have as many patients to do your studies because, again, it's an uncommon disease. You don't get that many people, so the studies are limited. You only need 1 instead of 2. And then, you can charge more because the costs associated with development and the potential market are so… Disparate.

Ethan: I guess it's not really like [00:04:00] Aducanumab because the indication is for Alzheimer's.

Rif: It's expensive, I think, just because nobody will pay for it because it's not really approved…

Ethan: And, they also know that it's not going to be effective enough to gain popular use. My impression is that they drive that price up because they've already realized they're gonna lose a lot of money on it.

Rif: And they will! It's actually part of a new push by the FDA to approve things based on biomarkers rather than based on actual clinical outcomes, and it's not toooo new. For example, cholesterol-lowering drugs are exclusively about biomarkers. Cholesterol itself doesn't hurt you, but elevated cholesterol is associated with increased heart disease, and all sorts of vascular outcomes. And lowering cholesterol is lowering [00:05:00] the biomarker, and so it's approved exclusively as a biomarker. There were no studies showing, at least initially, that being on a cholesterol-lowering drug actually lowers your risk. All of them were approved based on biomarkers. Now, there have been studies showing that it does reduce your risk, but the reduction in risk is less than 20%.

Ethan: So, Walter Brown is a clinical Professor Emeritus of Psychiatry at Brown University (who) specializes in Bipolar 1 disorder. And, he had some comments on a podcast I was listening to recently stating that the major reason for low lithium use (in the US) is that after the 1980s, other drugs, particularly Depakote, came onto the market that could also prevent mania and depression. And drug companies marketed those heavily. Depakote, in the pharmaceutical industry's mind, took over [00:06:00] Lithium's gold standard role. He also emphasizes that the lithium chloride incident contributed to underutilization and that people didn't forget about that for several decades. But he does say the primary reason that it is underused is the aggressive marketing of other drugs. And Alec Coppin, the British psychiatrist who ran the long-term studies on Lithium's suicide prevention effect stated, because lithium is very cheap, there's not much money there commercially. The more that I look into it, the more I wonder, because of the pharmaceutical companies impact and hands in the medical education system, writing textbooks, and paying for a lot of medical schools’ needs, I kind of wonder. Because in medical school, all I heard about lithium was this intense side-effect profile. I didn't know anything about its’ anti-suicide benefit, that it's the most effective medication in [00:07:00] bipolar disorder. I kind of wonder if there is the drive to use the fear that already exists and push us further and further away from Lithium?

Rif: I'm sure companies have played a role in the reduction of lithium use. There's also been some problems with how we teach medicine, and, you know, in psychiatry, again, we don't know what causes our diseases. But we teach our students to look at clinical studies, right? So we have a scale that measures depression, and you know, you're depressed, it's 30, You give an antipsychotic, the scale goes from 30 to 15, you give lithium, it probably still goes from 30 to 15, but (sarcasm coming…) Gee, isn't an antipsychotic easier to use? It's so much safer. You don't have to worry about patient getting toxic. You don't have to worry about the same long-term problems. We don't [00:08:00] teach our students to think in terms of mechanism. If you think in terms of mechanism, the most reproduced abnormality in bipolar illness is an increase in intracellular sodium. There are so many studies, at least 80 or 90, that show that if you're manic or depressed and you have bipolar illness, your intracellular sodium is higher. Your transport of sodium is less. It is the most reproduced abnormality in bipolar illness. You have a drug that specifically normalizes intracellular sodium. It's been demonstrated to do that, and so if you think about mechanisms, the first drug you think of is lithium… But we don't teach mechanisms. We teach clinical medicine. And clinical medicine means you look at [00:09:00] the scale, if the depression goes from 30 to 15, it works, and lithium does from 30 to 15 as well. So there's no apparent benefit. Well, of course there's a ton of benefit. There's a benefit in that you actually are normalizing the physiology versus creating new abnormalities. In the physiology, which is of course what we do with antipsychotics, if we thought differently, we would do things differently. For us to be thinking differently, we have to have something to actually teach our students. This is the mechanism… and we don't actually have that clearly yet.

Ethan: To add to that, you have a reduction in that theoretical scale, you also have the same weight given to difficulty concentrating to suicidal thoughts. Those are the same reduction. So, you may have an antipsychotic that allows you to concentrate (bad example, sleep would have been much better) a little bit better, but you have a lithium that takes away your [00:10:00] suicidal thoughts. Yet, on that piece of paper, it doesn't look any different.

Rif: And it's even worse than that. These scales don't measure the full depressive syndrome in any one study. Specifics, weight changes, for example, generally don't reach statistical significance. You have to use the whole score. Otherwise, you need to do 1,000 patients. Having said that, the effect size is actually huge with lithium. You will see changes with small numbers. With a lot of the medicines that we use, you need larger numbers to see the same changes. And, that's something else that people don't appreciate, that smaller studies are actually in lots of ways better than larger studies. We teach our students that large studies are good because they're more generalizable, but smaller studies are actually quite impressive because if you're getting this big change with a small [00:11:00] study, that tells you that actually your patient, if they respond, they're gonna respond really well. And you don't see that with antipsychotics.

Ethan: And the primary researcher also is much more in tune with what's happening in front of them when they're looking at, say, 10 patients they start on lithium vs this big meta-analysis of 1000s of patients, which 95% of 'em, they've had zero input into their care. Beyond that, we could talk about how the industry-sponsored studies are positive 85% of the time versus the NIH studies are positive 50% of the time.

Rif: Be careful how you interpret those data? A company isn't gonna spend millions of dollars to do a study unless they know they're gonna make money. NIH is there to actually know the truth, and so NIH does fund a lot more truly experimental treatments. Companies do studies just to get drugs that they know work approved. The [00:12:00] FDA designs most industry-sponsored studies. NIH cares about how much a study is gonna cost. FDA studies are, to a large degree, better, because the FDA doesn't care how much it's gonna cost you. NIH studies are always smaller and that gives you a higher negative rate because they tend to be a single site. So, it's a more rarefied population. So, there are lots of other things that can explain. I don't share the same huge distrust that you verbalize regarding industry-sponsored studies, partly because they're not designed by the company, they're designed by the FDA.

Ethan: But, who's working at the FDA? You have way too much recycling of big pharmaceutical company executives (Rif: And now you’re really getting into…) who are working at the FDA and then going back to big pharmaceutical companies.

Rif: Yeah, so the [00:13:00] FDA utilizes a lot of folks that are specialists in the field. And the bigger problem isn't that these people are corrupt. The bigger problem is that they're taught in the same way that I talked about before, that it's a clinical study and that's the gold standard in clinical medicine. And we don't take mechanism into account, and when we don't take mechanism into account, I think that's a fatal flaw.

Ethan: I don't think that everybody in the industry is corrupt. There are definitely some corrupt people in the industry, and there's a lot of examples of that, like the Vioxx (Rofecoxib) example, Seroquel (Quetiapine) and Depakote (Valproic Acid) trying to get FDA approval in nursing homes for agitation and aggression, having ghostwriters and primary researchers that had 0 to do with the study. The pharmaceutical industry is really important for the development, but there have to be stronger regulations, and more [14:00] separation. You should not be able to go from the FDA 1 year to being a high-up executive at a pharmaceutical company the next year and vice versa. Those agencies need to be split and the primary motivation for us in medicine needs to be: find treatments that work most effectively instead of trying to force treatments that don't have good data. Those are the pieces that just bother me to the core.

Rif: I don't at all disagree with some of the things that you're saying, but I'm sort of imagining a listener going away believing that everything is fake, and that is absolutely not true. It is important, especially since we're talking to the public here, to actually measure what we say. I do agree with you that there is definitely a problem because I can design a study that will [00:15:00] be positive. Just by virtue of the design, I know what to do to make the study look positive. So there are ways of sort of, if you will, doctoring, what a word, the way that you do a study. But having said that, by and large, the studies that are designed by the FDA, that industry has to do are actually tremendously high quality. But you're absolutely right. There's lots of sloppiness that creates lots of problems. And generally the problems are exactly what you said. You have greater cost for the company because they rush, they select patients inappropriately. They include patients that shouldn't be done. You have to remember, these are generally new chemicals. They don't know enough about them, so they might design the study incorrectly. All of these things increase the number of negative studies.

So, [16:00] negative studies can happen for all sorts of reasons. And negative study isn't a bad thing. We actually need negative studies to help us learn what not to do. But, if they're not published, then we can't learn a whole lot further. And that has been the drive over the last 20 to 30 years is to publish negative studies. And the biggest reason that negative studies haven't been published, part of it is companies, but part of it is also the journals. If I am an editor of a journal, the measure of my journal is something called Impact Factor, how often the articles published in my journal get cited. If I accept a negative study to be published in my journal, nobody will cite it. And so it's not sexy. It's not gonna improve how my journal gets measured and, 5 years from now I'm gonna get fired as the editor because the Impact Factor of [17:00] my journal has dropped because I've published all of these negative studies, important studies to publish. That's one of the advantages perhaps of having the Internet journals. But you're absolutely right: Companies still don't want to publish their negative studies. But that's getting better. Companies are publishing negative studies. Usually they'll just combine 2 or 3 negative studies in one paper. And that's important because when you talk about meta-analysis and just to explain what a meta-analysis is, the prefix ‘meta’ means everything. A meta-analysis is an analysis of all the data. And because it's an analysis of all the data, you actually need to know about the negative studies. This is a problem ‘cuz a lot of times they don't get published. It's harder to get a negative meta-analysis because of that particular bias.

Ethan: It is really important, the point that you bring up about the listeners hearing what [00:18:00] I'm saying and thinking that all of it is fake. I don't want to insinuate that there is nothing that you can trust. But I do think it's important to highlight some of the historical examples of ways that this lack of separation between pharmaceutical companies and the agencies that regulate them.

Rif: I do agree that's a problem, but that also creates a certain efficiency. If I know how to get a drug through the FDA, then I'm more likely to get a drug that is good through the FDA rather than have it get stuck. Nothing is always all bad. Creating ways of reviewing the drugs rather than telling people how to live their own lives is probably where the solution is.