Iris Otani, MD

 Gerry Lee, MD (Host): Hello, everyone, and welcome to another episode of Allergy Talk, a roundup of the latest in the field of allergy and immunology by the American College of Allergy, Asthma, and Immunology. For today's episode, we'll be reviewing more articles from Allergy Watch, a bi-monthly publication which provides research summaries to college members from the major journals of allergy and immunology, and by listening to this podcast, you can earn CME credit. For information about CME, head over to education.acaai.org/AllergyTalk. And we also have the ACAAI community on DocMatter, where we can continue the discussion about the articles today.

Well, hello again. My name is Gerry Lee. I'm an Associate Professor at Emory University and an Assistant Editor of AllergyWatch, and today, once again, we are joined by the Editor in Chief of AllergyWatch, Dr. Stan Fineman.

Stanley M. Fineman, MD, MBA: Hello, everybody, and glad to be here. I'm also an Adjunct Professor at Emory and past president of the American College.

Gerry Lee, MD (Host): And for the third chair, we are again welcoming Dr. Iris Otani, an Associate Professor of Medicine at UCSF and Assistant Editor of AllergyWatch. Iris, welcome back to AllergyTalk.

Iris Otani, MD: Hi, thanks so much for having me back. I'm an Associate Professor at UCSF in San Francisco and Medical Director for Allergy Immunology there as well.

Gerry Lee, MD (Host): Okay. Well, we have three more interesting articles to review. Iris, we'll start with you. I think you were learning more and more about use of auto injectors in different types of individuals. So what have we learned from this article?

Iris Otani, MD: Yes, so this article, is titled Frequencies and Predictors of Subcutaneous and Intraosseous Injection with 4 Epinephrine Autoinjector Devices, and I chose to review this paper because it looks at a really important question. Do patients really get an adequate dose of epinephrine when they use intermuscular epinephrine autoinjectors?

Are they actually getting it the right way? As a bit of background, I think we all know this, but prior pharmacokinetic studies have found that IM, mid anterolateral thigh injections, have faster peak epinephrine concentrations compared to the sub q injections, and that's why we recommend that patients get an IM injection of epinephrine for prompt treatment of anaphylaxis.

But this study suggests that there's a pretty good chance, especially in women, that auto injectors lead to subcutaneous administration, not IM. What the authors did is they compared needle length of four different epi autoinjectors with ultrasound measurements of skin to muscle depth and skin to bone depth of the mid thigh of 68 adults with allergic conditions to predict whether or not autoinjector use with and without maximal compression would lead to either sub Q or interosseous administration instead of IM administration. And they found that the predicted rates of sub Q injections were the highest with anapen that has the shortest needle length of 7.49mm, 65 to 66 percent regardless of compression. Kind of makes sense because the needle length is short. And then predicted rates of sub Q injections were the lowest with Emerade that has the longer needle length of 22.8 millimeters. So that one was a 21 percent of patients could get sub Q injections without compression. And then with maximal compression, 13%. On the other hand, maximal compression with this longer needle length introduced a small risk of intraosteus or IO injection. Now, the predicted rates of sub Q injections with EpiPen injects, so needle lengths are 15.5 and 15 millimeters, kind of medium length needles, were 37 to 38 percent without compression and then 29 percent with maximal compression. So not great, but not bad. And then the authors also looked at whether or not there were any differences in predicted rates of sub Q injections between gender and body habitus.

Female sex was associated with higher risk for having these sub Q injections instead of IM for all four auto injectors, regardless of whether or not there was compression used. Wide thigh circumference did predict sub Q injections, but only for the Anapen, the one with the shortest needle length. And then none of the other variables that they looked at, things like skin thickness, BMI, abdominal circumference, none of that really predicted whether or not there would be a higher likelihood for a sub q injection with any of the auto injectors.

So overall, I mean, I think this confirms that mid needle length auto injectors are the best in terms of maximizing the chance of IM injection without incurring this risk for having an IO injection. But I encourage people to take a look at figure four because that one, they show the projected frequencies of sub Q injections in men and women.

And the image is really striking because you can see that there's a 0 percent chance of men having a sub Q injection with the mid needle length auto injectors, but women have a 65 percent of having a sub Q injection, even if maximal compression is used when you're injecting, with the mid needle length auto injector.

So, it really begs the question, you know, when people are using EpiPen, especially in the heat of the moment when they're treating themselves for anaphylaxis, like, are they actually getting an IM injection? Or is there a chance that a lot of our patients are getting a sub Q injection?

And that's why sometimes, you know, people need another injection, or maybe the EpiPen doesn't work, the epinephrine auto injector doesn't work as expected. And I think it also kind of raises this question, you know, now there's this FDA approved epinephrine nasal spray. Is that potentially a better option for some people?

Because presumably there's not gonna be these differences in the administration for that. And I actually went back and looked at another article that was covered in AllergyWatch. The pharmacokinetic and pharmacodynamic comparison of epinephrine administered intranasally and intramuscularly by Tanimoto and all.

And in that one they show that the peak concentration with intranasal administration was lower or comparable to IM administration. The time to peak concentration was comparable and the improvements in blood pressure were comparable or better. So, I think this is all something to think about when we're prescribing epi for our patients and also really highlights, you know, for prescribing the IM epi it's good to potentially talk about this and make sure that we're going through um, correct admin technique with all of our patients.

Gerry Lee, MD (Host): So, Iris, I've heard that recommendation for patients with a higher body mass index, presumably with more subcutaneous fat in the thigh, that they administer it lower, like closer to the knee. Do you think we should change our recommendations for women versus men on how we administer epinephrine?

Iris Otani, MD: Yeah, I mean, I think I wouldn't necessarily change my recommendations between men and women, but I would say that I would kind of go over how it's important to make sure that after you press down on the EpiPen, like make sure to kind of push in and hold down. And if seems to not be working, it's possible that they might need another EpiPen. Make sure they have another EpiPen on hand. And I think it also makes me think a little bit more about this new intranasal spray.

Stanley M. Fineman, MD, MBA: You know it's interesting. Isaw this study and these are mostly things available in Europe. The only one we really have, I believe, is the EpiPen here. And so I remember when we first started seeing of the smaller children, like infants, who had problems with trauma from some of the administration, intramuscular administration of these devices.

But I think that's been taken care of with adjusted dosage and different needle length for those. But I do agree with you. I think the alternative types of administration, like intranasal or maybe even the sublingual might be you know, useful in the future. We've got to certainly have more experience because I don't think we've had that much experience with it yet.

Gerry Lee, MD (Host): Yeah, I'm excited for those products. I think they're smaller, they're more convenient, patients are more likely to use them. I think we have the general barriers of lack of track record, so we wanted to know comparative efficacy in real world and of course, cost. I think cost is probably the one we worry about the most, you know, in terms of making a decision what's best for our patients.

Okay, Stan, let's go to this next one. I'm so fascinated about how people are incorporating machine learning and AI into our field. So what are we learning about pollen?

Dr Stan Fineman: So this study, in fact, that's what kind of intrigued me about this study, that was presented with by Dr. Spreed in AllergyWatch. And the title of the article is Forecasting Daily Total Pollen Concentrations on a Global Scale. It was published in the journal Allergy in the May issue, May of 24. And it was an international group. There were one or two sites in the U. S. and they were looking at pollen data then they were trying to use different AI models to predict the concentrations of pollens. Right now we don't really have a good way of forecasting pollen count. I know there are some websites that, quote, forecast it, but quite frankly, we've been looking at pollen counts for years at our site, and we've collaborated with both the Emory Public Health Group and also the CDC, and we've been trying to figure out a way to forecast, but it's very, very difficult, and there are really no tried and true ways to do it.

So, automated intelligence, AI, everybody's a little bit concerned about it, I think, at some of my friends are concerned about what the influence in medicine is going to be, but I think this report basically shows that there could be some benefit to our understanding of how pollen is going to be affected by the different, weather, changes and so they use two types of AI learning.

One is something called CatBoost and quite frankly I really don't know anything about it except that it features these different categories and it's a way of, type of, I guess, ranking things it's using machine learning. It's a type of machine learning based on an algorithm, on decision trees.

Now that's the CatBoost. Now the deep learning that they used is another machine learning technique that uses these artificial neural networks to learn from the data. So it's not exactly a decision tree as far as I understand, but it's a different type of AI learning. And so they basically had the results of the pollen, in over 23 different cities, and they, around the world.

And they looked at the elevation, the various pollen counts and they, some of the variables that they were looking for besides the temperature, the ambient temperature, they also looked at the temperature in the soil at various depths and tried to predict it, and interestingly, they were able to predict some for certain cities I think in Australia, in Cape Town, in South Africa, and in Mexico City but there wasn't really a lot of consistency.

And, but I think that the reason that I brought this up is because I'm interested in pollen and the fact that it impacts our patients, it impacts other kinds of chronic diseases as well. And since we're seeing an increase in pollen because of our warming trends in the climate, that I think it's more and more important, but I think this is just an example.

I think the takeaway here is here's an example where AI can hopefully help us predict pollen counts and maybe advise patients, you know, when do they need to stay inside or when do they need to take extra precautions to avoid the pollen? Because the way we report it now, of course, is we just report the pollen count for the prior 24 hours because, we use a rotorod or then we just take that measurement for the prior 24 hours.

So this is, I think something that's going to come up and I just think it's an example of a pot a potentially positive use of automated intelligence for medicine.

Gerry Lee, MD (Host): What I always struggle with is I think absolutely for patients with pollenallergy, one strategy we can empower patients to control their symptoms is to try to minimize exposure on high pollen count days. But the flip side of it, and I kind of feel bad letting someone's disease control their life.

Do you know what I mean? Like, you know, a parent saying, well, I'm going to keep my kid in because there's a lot of pollen outside. I think on one hand that's helping, but the other hand there's like a cost to that, to a child's sort of social development. So I think pollen forecasting can help patients maybe know when they need to maybe take extra medicine or, try to minimize unnecessary outdoor exposure, I guess, but I always have always been more biased toward, yeah, let's just anticipate pollen and take the medicine you need to live a normal life than, oh, the pollen counts high, so you should not go outside. You know what I'm trying to say?

Stanley M. Fineman, MD, MBA: Well yeah, I mean, as you said, you could take precautionary measures, I mean, such as maybe increasing your maintenance meds if you need to during high pollen days, but I'm, all in with what you're telling the patients is, you know, you don't want to use pollen allergy to get out of PE outside. I would never write that.

Iris Otani, MD: Yeah, I think that's a great point, Gerry. I mean, often I think our role as allergists is really to help people live with their allergies and be able to live the life that they want. And I didn't think of that until you mentioned it, but I could definitely see how potentially this technology might lead to some people just being like, not going out today, but I can also see a lot of people really liking using like an app that's going to tell them, what the pollen count is going to be on any given day.

Stanley M. Fineman, MD, MBA: Yeah, and that's really the reason that I brought this up, was the AI part of it. We're also collaborating with the researchers at Emory and the Rollins Public Health School, who are using AI to count pollen. So we're using a pollen counter that can do an automatic intelligent pollen count.

It doesn't count as well as a hand count right now, but I think you're going to be seeing more data from these types of automated intelligence counters used around the world. And I just think it's something that we can keep our eyes out for and I just think it's interesting.

Gerry Lee, MD (Host): AI pollen counting is an amazing Stan. Boy, if we could get more information out into the people, because there's so few pollen counting stations in the country, if you look at NAB or so on. So that's exciting technology. I was just going to make the joke also that I think they should work on weather forecasting first before they work on pollen.

I think weather forecasting has a while, has a while to go too. They got to work on that. Anyways let's wrap this up. I got one last article to review and this is Eosinophil Depletion With Benralizumab for Eosinophilic Esophagitis. The first author was Mark Rothenberg from Cincinnati.

And, it's in the name. The name of the disease is eosinophilic esophagitis. So, I mean, the study is quite simple. Here's a drug that depletes eosinophils. The disease is eosinophilic esophagitis. Get rid of the eosinophils, call it a day, right? Like that, it just makes sense, right? But then when you look at this article, it's so fascinating.

So this is a double blind placebo controlled trial of benralizumab in eosinophilic esophagitis. This was reviewed in the last issue by Samantha Knox. And essentially they looked at, you know, histologic scores, which we should, that's how we track eosinophilic esophagitis. And then they had patient reported outcomes. Specifically, the primary outcome was the dysphagia symptom score, or DSQ. So their primary endpoint was not only an improvement of DSQ, but histological remission in this study was getting a peak eosinophil count under 6. So this is mostly an adult study, there's only 13 percent of patients were under 18, and when you give benralizumab compared to placebo, 87.4 percent reached that threshold of under 6. So that's awesome, right? You could just call it a day, right? Interesting that after 24 weeks of treatment, not only is there no change and a statistical improvement in dysphagia symptom questionnaire in the benralizumab group. If you look at the curves, there is this slight trend. It was not significant, but there's a slight trend that placebo got better symptom improvement! That the DSQ reduction during the study had a trend for improvement on placebo versus the drug. Again, we're gonna call that no difference, but I'm just, the curves are interesting, right?

They looked at other patient reported outcome measures. Other validated symptom scores, quality of life, and, no statistical difference. And then they did gene expression. And so when you look at gene expression in patients on benralizumab, they were more likely to have decreases in eosinophil specific genes like galactin or CCR4, but they had an increase of interleukin 5 and interleukin 13 expression, almost as if, like, if you get rid of EOs, the immune system is like motivated to make more T2 cytokine, right?

So what are we comparing this to? We're comparing to our current FDA therapy, which is dupilumab, right, which blocks IL 4 or 13. And clearly this study with benrolizumab has superior histologic remission, but they had a symptom improvement with dupilumab, blocking 4 and 13, and they're actually maybe seeing, I'm not going to call it worse because it was not statistically significant, but it actually was certainly not better than placebo. And so when we think about eosinophilic esophagitis, maybe it's actually a misleading diagnosis name because we characterize the inflammation as the disease, but you take away the EOs and it actually is driven something else. Like the EOs are an epiphenomenon. And I think that sort of closes the door on this, and helps us think about what's really driving this disease and how do we help our patients, and it's really going to be more upstream.

It's going to be T2 cytokines or so on. So I think that was just another study on a long line to examine this sort of difference between the IL 5 eosinophil side versus the T2 cytokine side, and I would say this would probably be the nail in the coffin that declares victory that addressing T2 cytokines like 4 and 13 is what's gonna get patients better with eosinophilic esophagitis.

Iris Otani, MD: Yeah, it's, really, another study showing that the anti IL 5, even though it decreases the eosinophils, like same as mepolizumab, this, the benrolizumab decreases the eos, but it doesn't really necessarily result in symptom improvement. And fortunately there's the anti IL 4, anti IL 13 therapies that have been beneficial.

Stanley M. Fineman, MD, MBA: Yeah. What you said, I think it's true, Jerry. Yeah, I think that would be our go to, would be IL 4 or 13 type treatment in these type of patients. think it's also, I just want to put a plug in for AllergyWatch. This is a study from the New England Journal, which, I don't know if all of our college members read on a regular basis, but obviously it's an important study and the findings are important.

It's the fact that we bring it up in AllergyWatch and talk about it in AllergyTalk, I think helps our members to sort of broaden their understanding of the current literature of some of our, you know, allergy developments.

Gerry Lee, MD (Host): Okay, well, those are our articles for today. If you like what you heard, please rate our podcast on Apple Podcasts. And we are always looking for more feedback on this podcast. If you have a correction, suggestion, or so on, please email us at allergytalk@acaai.org. Don't forget about CME credit.

That website is education.Acaai.org/allergytalk. And then to read archive issues of AllergyWatch, head over to college.acaai.org/publications/allergywatch. This is Gerry Lee. Thank you so much for listening again. We'll catch you next time. Enjoy the rest of your day.

ACAAI is presenting this podcast for educational purposes only. It is not medical advice or intended to replace the judgment of a licensed physician. The college is not responsible for any claims related to the procedures, professionals, products, or methods discussed in the podcast. And it does not approve or endorse any products, professionals, services, or methods that might be referenced.