In part two of this three-part Instagram Live discussion, 51˶ editor-in-chief Jeremy Faust, MD, Karl Nadolsky, DO, and Fatima Cody Stanford, MD, MPH, MPA, MBA, dive into the complexities of GLP-1 receptor agonists, unpacking their side effects, data controversies, and transformative potential in diabetes and obesity care.
Nadolsky is a clinical endocrinologist and obesity specialist at Holland Hospital in Michigan. Stanford is an associate professor of medicine and pediatrics at Massachusetts General Hospital and Harvard Medical School in Boston.
Following is a partial transcript of the video (note that errors are possible):
Faust: I want to move to some things about patients really are curious about, and one of them is side effects and what's real and what's not. Now, I'll frame this by saying I think that all medications we give have risks and benefits, and I think that there's a little bit, in my view, of obsessive attention on some of these things like, 'What is the discontinuation rate, what are the side effects?' All medications, especially injectables, really have a very high attrition rate. So this is not unusual, but what I'm noticing is something strange, which is, you'll see, I saw this that was observational data on which of the patients on, I think it was the semaglutide [Wegovy, Ozempic], do they have more pancreatitis and small bowel obstruction and gastroparesis were the ones they found. And what was so strange to me was in the actual randomized trials you didn't see [them], and the randomized trials were larger than the observational studies, which is a completely backward situation. I just have this sense that the magnifying glass is out, whether it's about the pancreatitis, the GI [gastrointestinal] stuff, whether it's about, someone asked about the optic nerve stuff, someone asked about mood disorders and all these things. Can we talk about the side effects and what you think is real, what you think is sort of maybe the magnifying glass being out?
Nadolsky: Yeah, I guess I'll go first. I think like you said, when you look at the meta-analyses of the huge trials, that's why we use randomized controlled trials, the adverse effects are pretty reasonable. I think the discontinuation rates of these are smaller than what you might think, oftentimes single digits due to side effects like the nausea and vomiting. And we know that the nausea is, it's pretty common relatively speaking, but people are usually able to weather that storm and we need to make sure we can deal with those for patients, help them tolerate it, help them get the response and personalized therapy. But there used to be a concern of the pancreatitis and really all the huge analyses of this in the trials don't show much of a concern. The one true statistical risk looks, in my opinion, based upon it, looks to be the gallbladder disease, something like 1.1 to 1.5% placebo versus the active arms, of a risk.
And so that's something I think that needs to be understood. Obviously those are small relative or small absolute numbers for sure, and it looks, even though you mentioned the mental health, I'd be interested to hear what Fatima says, but there were some of those reports that were raised and the Europeans and the FDA looked at it, ultimately it went away. And then there's actually some data, some huge retrospective records looking back, showing a reduced suicidal ideation that I saw, I can't remember where that was, maybe in a JAMA or something like that. Huge data sets, which to me makes more sense. If you're treating obesity, which has a high risk of mental health disorders, then maybe improving all those things will improve it. I mean, think about, look at, like you said, knee arthritis, the heart failure trials. I was most impressed as a guy who loves exercise and pushing fitness when we have people with heart failure, preserved ejection fraction, who can suddenly dramatically increase their 6-minute walk time and they're exercising more and they're feeling better and life is better. To me, it would make sense that maybe mental health in general would get better, but there were some signals. I'll get that. Fatima, what do you think of that?
Stanford: Well, I think first of all, we're looking at EHR [electronic health record] data. We have to recognize that this data is messy, and when you go looking for things, you're going to go find things. I can tell you as someone that's done EHR data analysis and when I go looking, I can find basically whatever I'm looking for if I twist and manipulate the data in any particular way. I think that when we're looking at the GLP-1 receptor agonist, people go in expecting to find things that are going to be negative, but we kind of go in setting up a hypothesis that we want to find something that is bad. A lot of people have still have this lens that obesity should only be fixed by looking at lifestyle modifications. This is really still the ongoing mantra within our society. We should not use anything outside of lifestyle modification, which is dietary modifications and or using exercise or exercise strategies.
When we start looking at pharmacotherapy or metabolic bariatric surgery, these are tools or strategies that we should not be investigating for this disease process. And so when we go looking for it, we will find it, particularly looking at EHR data, which is messy and actually highly flawed. I mean, I can tell you just from looking at this data, when we look at mental health and we look at the over 200 plus complications associated with obesity, we know that there is high degree of overlap. We pick a Venn diagram of any of these things, we're going to see high degree of overlap, unfortunately with a lot of the things that we're looking for. And so we find some unfortunate signals in these types of disease processes, whether it's depression, whether it's psychotic disorders like bipolar disorder, schizoaffective, schizophrenia, et cetera. And so when we go looking for it, like I said, we are going to often find it. And so I think this is what we're often seeing when we're looking at these retrospective analyses, and that's why we don't see it as often when we actually do a cleaner RCT [randomized controlled trial], which is a better form of looking at the data.
Faust: I'll give you an example of this. The pancreatitis signal in that JAMA paper was like 10x, meaning people had a 10 times higher rate of pancreatitis if they were on one of these agents compared to supposedly match controls on other agents. But then in the trials you don't see that signal. But I will see this patient, I've seen this patient in the ER [emergency room] several times. They just started their med and they have some nausea, vomiting, and abdominal discomfort, which are generally not terribly unexpected at first, mostly tolerable, but then a few can't tolerate. And that's understandable, but they go away. But because they're on the med, the doc says 'Uh-oh pancreatitis, go to the ER.' But if they're on any other medication with the same symptoms, they wouldn't be told to go to the ER. So it's almost like a self-fulfilling prophecy...
Nadolsky: And then they check the lipase, which is elevated in everyone -- a little bit -- taking them, and it signals the whole thing and they go down a rabbit hole.
For more, watch the video above.