ACR 2025 Daily Podcasts Day4a Save

This is an ACR twenty twenty five podcast coming to you from Chicago. Hope you enjoy it.

Hi,

everyone. My name is Brian Jaros, We are coming to you live here from ACR twenty twenty five.

Hi, my name is Doctor. Elaine Husney. I'm a rheumatologist from the Cleveland Clinic. Brian and I are going to kick off with, From Pain to Wellness in psoriatic arthritis.

Yeah, so this was a topic I was really excited to talk about. You'll hear definitely a lot during this conference about the new, therapeutics, biologics that we have in this disease state, which is definitely exciting, but I think, this topic explores other questions of how do we comprehensively or holistically treat these patients with psoriatic arthritis outside of just the medications that we're giving them. So to kick us off, Elaine's going to tell us about abstract number two three two seven. This is persistent pain in psoriatic arthritis, a distinct phenotype of depression, fatigue, and catastrophizing.

Yes. Thanks, Brian. This is a really, interesting, work done, at NYU. They have a great combined psoriasis psoriatic arthritis clinic. This is an abstract with Rebecca Haberman who's talking about pain and psoriatic diseases.

So they recruited about 91 patients and they split them up into three groups. So this is commonly the groups that we see, right? Patient with active PSA, a patient that has persistent pain but no swollen joints, and then a control group or those that are in remission. What they found, they looked at many, many different patient reported outcomes. The ones that I thought was interesting was when they were comparing the persistent pain, no swollen joint group versus remission, they found that they did have higher PROs related to catastrophizing pain, higher depression, not necessarily anxiety, higher fatigue, and overall just less quality less quality of life, unfortunately issues.

And so what they found is that that was equal I believe to the active group. The group that was having both tender and swollen joints and pain. So I think this really highlights that there are groups that we need to pay maybe closer attention to, right Brian? So perhaps the, even those with persistent pain, we need to really understand, their quality of life and their measurements and maybe we can intervene early and really improve their overall quality of life.

I completely agree, and that is a very perfect segue to talking about what interventions do we have, what ways can we help these patients recognize them early and exactly improve this quality of life metrics that, of course, to patients are very meaningful. So now I'll bring us to abstract three seventy one. This is lifestyle coaching in psoriatic arthritis, pilot findings from an online coaching program. Elaine will also tell us about this. This is a project she was intimately involved with.

So tell us how you came up with this and a little bit about the project.

Thanks so much for highlighting this. I think as, we know, our, keynote speaker was on wellness. Right? And so I think the issue of physician burnout and all these issues are all around us. But what can we do for our patients?

And so we have something called e coaching at the Cleveland Clinic, which just like it sounds, you know, we go to the patient, right? We don't have to have them come make an appointment for a practitioner. We do this all through email, and these e coaches are well trained, to connect with a patient and to see what, they can do with their chronic illness. So we have four pillars that we work with. The e coaches work on mindfulness, they work on exercise, they work on diet, as well as sleep.

So in this particular abstract, we enrolled, just a little over 86 patients to be exact. They all had psoriatic diseases and we looked at their patient reported outcomes before and after e coaching. And so e coaching is really just like it sounds. We connect with them by email at least twice a week is what has shown to be very helpful for the patients. If you do it less than that, sometimes you don't get as good of a result.

And we noticed that they had improvement in their self efficacy, improvement in their symptom management, and improvement in their medication management. So I think all of this is, really good. They can work on whichever pillar they want. So if they have sleep issues, they can focus on that. If they have mindfulness issues, they can work, work on meditation.

Or if they need to do all four, they can. But we go to them, in the e coaching. So I'm hoping that, you know, maybe in a larger study, maybe in other institutions, that they will pick up on, you know, maybe the use of an e coach, which is really a low touch, low cost alternative to really addressing some of those lifestyle pillars.

Definitely. And I imagine just thinking about the patients that I treat, a lot of them would be interested in this sort of platform. What was the reception that you heard from patients about this experience and participating in e coaching?

Yeah. I think, you know, interestingly, there was a lot of overwhelming, like, yes. I wanna do this. And then I think when they got into it and saw that it was a little bit of homework, we felt a little bit of a pause. So I think we do have to set some expectations.

This is not like you can sign up, and then all of these pillars will get miraculously better. Sure. So but we are able to slowly, step by step, you know, help them through each of the pillars. And I think once they understand that, we did get, you know, over 90%, you know, satisfaction. But I think you bring up a really good point.

I don't think it is something that is just a a one hit wonder once a week. Right? You really need to put in time, be consistent. Our project was over ten weeks.

I see.

Okay. And sounds like from your perspective, you think in terms of future directions, larger studies, larger cohorts of patients, is there anything else you see in kind of the wellness sphere of opportunities for improvement in psoriatic arthritis?

Great question. I think most people inherently want to get better, right? They inherently want to address these lifestyle, but they just don't know how.

Yep.

So I don't think it's a matter of trying to convince people. I don't think anyone's gonna say, I want less sleep. Right? So so I don't think we have to do that, but I think we have to do much better on how to get there.

Awesome. Awesome. Well, thank you so much for joining me today, Elaine. And that's all we have for today. We'll see you at the next video.

Thanks.

And I think Jack has to, encourage us to say who's the pain and who's the wellness of our group. Right? So so we'll let the audience decide.

Take care, everybody.

Hi. I'm at Janet Bourdeaux reporting at RheumNow. This is ACR twenty twenty five live in Chicago. I wanted to talk about the late breaking, poster number l b zero one. This was an interesting study.

So this was a randomized controlled trial on IL two therapy in SLE. IL two has been trialed in lupus many a time, and it's very tricky. Too much of a good thing, and it actually interferes and can make lupus worse. Low dose, might not bind well enough and might be as effective. So there's a sweet spot of whatever the molecule is and how it's, dosed.

So this was a large randomized controlled trial in patients with active lupus. And what they did was they had several doses of IL two or placebo added to standard of care. They had the usual primary outcomes, LLDAS, various responses, etcetera. So why am I reporting this? It's because it's very exciting.

Basically, there was clinical improvement that I think was not just statistically significant, but clinically relevant. And in the doses they chose, there was actually a positive dose response. So higher doses did better, and the lowest dose was closer to placebo, but still numerically above placebo. So not only did they have changes in lupus disease activity as we would expect, they also looked harder and found changes in c four with a dose response, and changes in some other, outcomes, some, ones that aren't traditionally used but might be relevant in IL two, treatment. So where will this fit in?

I think we have a plethora of a lot of, trials in lupus, but I think we want something safe and that is regulatory that can help our patients, without the toxicities. So if this continues in a positive, phase development, then I would think that this might be for our patients with moderate disease activity or for safety reasons to use it maybe over some other products. Don't know. More will come. Please follow us at RheumNow.

Thank you.

I'm Anthony Chan reporting here for RheumNow in in ACR twenty twenty five in Chicago. And today, I want to talk to you about a subject that we probably don't talk very much about in our clinics, which is the assessment of sexual function in axial spondyloarthritis. It appears that it's one of the important outcome measures that we should add to the many other outcome measures that we use to assess our patients with ankylosing spondylitis or excess spondyloarthritis. And there are two abstracts today presented that I wanted to talk to you. Firstly is two three four six.

This is a check study where they looked at the sexual function in patients who received biologic treatment. And in this study, they focused on men. And the patients who are about to start biologics, assessed their sexual function using the IIEF score, which is a validated score to check for sexual function at zero and then at six months. And the study objective was to see the impact of the biologics on these outcomes on top of the standard outcome, is the disease activity scores. There were fifty two male patients who were randomised into the study.

They had ankylosing spondylitis and the mean age was around 40 years. Around ninety two percent of them were HLEB27 positive. The majority of them, around ninety six percent, received anti TNF and around four percent had IL-seventeen inhibitors. They assessed sexual function but also they looked at traditional outcome scores, the CRP, S DES, health assessment questionnaires, SF 36, and also the EuroQual. The overall improvement at six months was significant in terms of the traditional outcomes, ESR, CRP, and STES, but also the improvement in quality of life, including the HAC and the Eurocall.

When they stratified these patients based on their baseline IIEF, erectile dysfunction scores, they found that there was a difference in the group that had received the biologics. There was an improvement in the score at the end of six months in those who were treated with biologics. There was also significant gains in many other aspects of their sexual function which was assessed. And this was maintained over the period of the study. In those patients who did not have any sexual dysfunction, then what the biologics did is they maintained score.

So what we're seeing is that in patients who had a low baseline score, we signify poorer sexual function, there was an improvement. But if they did not have sexual dysfunction, that was maintained throughout the whole six months. These were quite important studies that this shows that not only does our biologic DMARC treatments advance the improvement in terms of their health outcomes, their quality of life, their physical scores like the ESR and S test, but also sexual function. This highlights the importance of maybe addressing sexual health on top of our other outcomes that we assess. Similar to this in abstract two three one six, this is from a Brazilian study where they did a meta analysis in women this time with XBA.

And they also found that sexual defunct dysfunction and sexual health issues were were higher compared to age matched healthy controls. Again, highlighting that incorporating sexual health evaluation is is probably something we need to start to consider in, in assessing our patients with, axial spondyloarthritis. I'm Anthony Chan reporting here for RheumNow in the ACR 2025.

Hi, everybody. It's Mike Putman from the Medical College of Wisconsin. I'm here reporting live from, ACR 25 for RheumNow. Now today, I want to talk about one of the number one trends I have seen at the meeting this year. You may be expecting me to talk about CAR T or some new IL-seventeen inhibitor.

Nope. I am here to talk about TriNetX research. If you are like me and you've walked the halls or been to some abstract sessions, I am sure you have seen a lot of projects with the TriNetX logo and the TriNetX name at the top. You're probably asking yourself, where are all these coming from and why am I seeing so many projects based on this? Today, gonna try and give you a little rundown about how to read these projects and assess them because I think it's gonna be pretty important for interpreting the data that you saw this year.

So full disclosure, I do research with TriNetX. I actually like it quite a bit. And I think I've done some products that I really believe in using that dataset. But there are a couple problems. First, gonna talk about the good though.

So the TriNetX is a large federated database. It's a bunch of health care systems from around the country that have combined their data in a way that can be analyzed. So it's a lot of different inputs, but at the end of the day, all looks the same when you go to analyze it. They do have a lot of information. They have diagnostic codes, they have medications, they have just enough demographic data to get yourself in trouble.

And then the number one thing is that it's really easy to get access. If you're an institution that contributes data to TriNetX, you can get data back from TriNetX and get them about a 110,000,000 people from around The United States. If you have access to the world data, it's even 250,000,000. So quite a lot of people that you can study with real in a relatively frictionless manner. You just log on, throw some ICD codes into a query and bam, science.

There's a joke that I heard a while ago about amoxicillin. They said, the best thing about amoxicillin is that it is cheap, it is effective for treating ear infections, and it tastes like bubble gum. The problem with amoxicillin is that it's cheap, it's effective, and it tastes like bubble gum. So all the kids got it, and now there's a lot of immunity. So problem with TriNetX is that it's easy and so everyone's doing research with it, but that doesn't necessarily mean that the research is good.

And so a couple of things that are problematic with research from this area. You know, the number one problem that I see is that the sample sizes are just really big. You can really easily find a million patients and rare diseases, you'll get hundreds and hundreds of cases. So these are very amenable to submitting abstracts to the meeting because I haven't seen 300 cases of Behcet's disease but there's probably that many patients in the TriNetX database. And so these sample sizes encourage people to submit them and then makes it a lot easier to get that kind of work accepted.

The problem is that sample size isn't a magical fix. It's not like some sort of divine intervention where you you just get a big sample size and all of a sudden science happens. A large sample size often merely gives you a good precise estimate of the underlying bias of the research itself. So if you do a study where there is confounding by indication or channeling bias or selection bias or time related biases, you're just gonna get a really accurate measure of the problems of your study itself and not actually move science forward at all. So how do you tell fact from fiction in TriNetX projects?

The first is just ask yourself when you see a project, is this effect even plausible? You know, non rheumatic drug being applied in the rheumatic setting that has a two hundred percent reduction in mortality is just not plausible. Some drug that we have studied in a randomized controlled trial and we have a pretty good idea of its effect size and maybe it's kind of marginal and then we compare that head to head against one of our tried and true therapies that we know work really well, and it beats it out, that's just not plausible. So use your sniff test as a rheumatologist to ask yourself if whatever the purported association is could even be really plausible at all. The second thing to think about is just to ask yourself if the patients being studied even had the underlying disease itself.

Now, I mean, we all know about diagnostic coding. I'm guilty of this myself. There's a diagnostic code called non diagnostic immune findings or something. I put that on everybody. That's not the actual thing.

And so a lot of people put diagnostic codes around willy nilly and they don't necessarily mean the patient actually has the disease itself. So any project with one diagnostic code and no requirement for medications being received is likely to have a lot of people there who don't actually even have the disease itself. Third, ask if patients already had the outcome of question. A lot of the time you'll see studies where someone who probably already had a heart event, a major adverse cardiac event, or already developed renal disease is being put into a study and there's no effort to exclude such patients or make sure that they don't have the disease in question or the outcome in question. So an adequate look back period and an adequate exclusion of people who already had your outcome is pretty important.

Fourth, acid confounding or selection bias should be considered. There's a lot of the time when you read these studies and it's just kind of obvious that the people who are given this fancy DMARD compared to the people who are giving a off brand conventional synthetic DMARD, they're just different people. The decision to give that medication was different, the reasons for it were probably different. There's none of that in TriNetX. There's no notes, there's no explanation of why people were given things.

And so any study where it seems plausible that confounding or selection bias or something like that was a problem, it probably was. And then last but not least, and this is just a broad metric, ask yourself or ask the ask the person who's doing the the project if the in house statistics were used. So this database, you can just hop on there and throw in a query and click like t test, you know, it just does it for you. It's very unlikely that projects designed in that way have adequately assessed for the biases that I'm talking about. Most of the groups that are doing good research in this area, and there are a lot of groups that are doing good research in this area, are looking at the raw data and they're applying all of the modern fancy stats and statistical methodologies to it.

Just saying I logged on to TriNetX and used the propensity scores is is just not enough. So in conclusion, I I don't wanna be mean to all these studies doing TriNetX data. This is not a new thing. We have been doing observational studies with questionable conclusions forever. It's a thing that we've always been doing in medicine.

So TriNetX isn't necessarily the problem, but it's amplified a problem that already exists, which is that a lot of observational data is not true, or a lot of observational findings are not true, and it's given them these sample sizes that confer an error of plausibility about them that is probably undeserved. So really at the end of the day, the onus is on you. You have to be a physician who's capable of reading and interpreting the medical literature. And when you see something that's a purported finding, need to critically appraise it and ask yourself like, is this real and could this should this be affected by management? So I hope that was interesting.

I hope you saw a lot of great things at ACR twenty twenty five. Thank you for tuning in.

Hi, this is Doctor. Al Kim again from ACR Convergence twenty twenty five reporting from RheumNow. I'm from Washington University School of Medicine in St. Louis and the faculty lead for all the cellular therapy abstracts here at ACR. So this video, we're gonna talk about a single abstract here.

It's gonna be abstract 2,465 and it's testing actually an AstraZeneca CAR T cell that is targeting both BCMA and CD19. So the justification for doing this is that while most of the CAR T cells are targeting CD19 which gets rid of the majority of your B cells, everything from bone marrow B cells all the way through plasmablast, BCMA also allows you to peg down plasma cells. So these plasma cells usually are in your bone marrow, They don't express CD 19 and they're gonna be an important source of long lived antibodies. They could be an important source of autoantibodies in certain conditions. So that's a justification for looking at BCMA.

But there's a theoretical risk to this is that protective antibodies such as the ones that you generate through infection or vaccination could also be eliminated. So these are important early litmus tests despite the small studies to be able to determine whether or not these type of combinations are safe. So in this particular study, there's a dose escalating phase one open label study. So they basically dosed three people at a certain dose of CAR T cells. Once they deemed that safe and found no DLTs or dose limiting toxicities, they moved up to the next level, did the same thing with another set of patients and then did the same thing with a third level.

A total of 10 patients were enrolled in this study which was done in a single center in China. They all received lymphodepleting chemotherapy, your typical flus, derabine and cyclophosphamide in order to eliminate T cells so that there's more IL-seven and IL-fifteen for the CAR T cells to be able to grow into. So again, phase one study, so it's safety. We're going to start off with the routine CAR T cell safety issues including ICAST. So this is immune cell associated hemotoxicity syndrome.

We expect most people to get this because of the lymphodepleting chemotherapy and indeed nine out of ten participants had iCATS. The grades were pretty high. Six of them had grade three neutropenia or grade three or higher neutropenia and five had grade three or higher lymphopenia. So this was something most likely associated with the chemotherapy. Six out of ten had cytokine release syndrome.

All of them are grade one or two. This is pretty much to be expected and acceptable given the other CAR T cell trials in lupus and no one had ICANN. So this is the neurotoxicity syndrome. In terms of the PK or pharmacokinetic data, this looked good. Once the CAR T cells were infused, you saw a maximum expansion around day ten and then they eventually fell out of the bloodstream a couple weeks later.

That's associated with rapid peripheral B cell depletion. This is gonna be in the blood and then reemergence starting around four months afterwards in most of the participants. Efficacy was looked at week forty eight. So this is actually one of the few trials where we actually have almost year data in terms of efficacy. Interestingly enough, DORS was only hit in six out of the ten patients and one actually flared and had to withdraw from the study.

One other patient hit LLDAS, so low lupus disease activity, and interestingly all of these patients actually had lupus nephritis. So of the ones that hit week forty eight, which was going be nine because remember that one went through, six of them hit complete renal response. The main reason why they did not hit complete renal response or doris remission was because of proteinuria. So in those three patients they actually did repeat biopsies and found no activity. So the proteinuria was coming from something else.

The author was unable to tell me the specifics, but one could imagine it might be scarring. So the main differentiator of this product is that BCMA and indeed the protective antibody titers did drop. They checked mumps antibodies which were the most profound in terms of dropping Then there were also reductions in hepatitis B titers and tetanus titers although there was more variability. So this CAR T cell is definitely doing what it's supposed to do. It's getting rid of a lot of B cells.

It's also getting rid of your plasma cells most likely given the drops in antibodies, especially protective antibodies. As a reminder, ten out of ten of these people had hypogammaglobulinemia, which is something that is expected from something like this. But is this dual targeting really needed in something like lupus, given the data that we're seeing with just targeting CD19? And I think this is really dictated by the need to eliminate autoantibodies. These autoantibodies may not be as important to get rid of in lupus as opposed to the B cells whereas in other conditions such as inflammatory myopathies, Sjogren's, scleroderma, you may need to target those because they may be more pathogenic.

The story will eventually unravel itself and the answers hopefully will be clear soon.

Hi. I'm Sheila Reyes from The Philippines, and I'm here in Windu, Chicago reporting live for RheumNow at the ACR Convergence twenty twenty five. Hydroxychloroquine during these days of the conference has been, again, taking the spotlight. And, in this video, I will be talking about hydroxychloroquine and blood levels. Now just to give a brief background, we all know that hydroxychloroquine is a cornerstone in the treatment of SLE, and prior studies on dosing have shown that a dose of more than five milligram per kilogram conveys a twofold increased risk of retinal toxicity, whereas a dose of less than five milligram per kilogram increases the likelihood of lupus flares two to six times.

So in this video, I will be talking about abstract one seven two two entitled defining safe hydroxychloroquine levels blood levels rather by the group of doctor Shivani Garg who, that was presented during the, plenary session on Tuesday. Now the primary objective of this study was to validate the upper threshold of for therapeutic hydroxychloroquine blood levels, or blood level monitoring for routine clinical use. And the second objective of this study was to identify kidney function thresholds with potentially or that are potentially toxic hydroxychloroquine blood levels. Now pooled data from large SLE cohorts, including the SLIC, Wisconsin, and three French studies were analyzed, and hydroxychloroquine levels were, either taken from serum or whole blood, and it was taken at baseline. So what do the results show?

Among the one thousand eight hundred and forty two patients who were studied, four point four developed toxicity, majority, about ninety per sorry. Eighty five percent of which were retinopathies. And the odds of toxicity was about one point nine fold higher at blood levels more than one thousand one hundred fifty nanogram per ml, while levels that were less than 750 nanograms per ml was associated with a one point four fold higher odds of developing active SLE. Now they also saw that levels that went beyond 1,150 nanograms per mils were supratherapeutic, which means it had no added therapeutic benefit and was potentially toxic, therefore, defining the upper threshold of hydroxychloroquine blood levels at 1,150. Now in terms of, kidney function thresholds, they found that patients who were on CKD stage three stage three had twice the higher odds of supratherapeutic levels of hydroxychloroquine, and this is despite weight based dosing at less than or equal to five mgkg.

How can we apply this to clinical practice or what are the implications that these study results may have, well, it reinforces the importance of monitoring blood levels of hydroxychloroquine to guide its optimal use. Now although the study had strong points, they also had some limitations, and I think validation in larger clinical trials is still warranted. Follow me on x at RheumNow and tune in to RheumNow for more updates of the ACR twenty twenty five.

Hello. My name is doctor Rinalini Day. I'm a fellow in rheumatology and internal medicine working in London in The UK, and I'm delighted to be reporting for RheumNow from here in Chicago at ACR twenty twenty five. We had a really enlightening and exciting session yesterday morning here at Convergence, which was the great debate and the use of AI in rheumatology. AI is everywhere at this conference.

It's not been hard to miss. And so I'm delighted to be joined by the two speakers from that session. So I have with me today doctor Jeffrey Curtis, who is professor of medicine at University of Alabama in Birmingham, and also doctor Januz Yazdani, who is chief of rheumatology at San Francisco General Hospital and professor of medicine. Thank you very much for joining me for this. So I think, first of all, for people who weren't able to attend the session and also for people who are not at Convergence, I'd actually, first of all, like to get both of your thoughts on what you feel actually is is AI.

What is artificial intelligence? That's a big question. But if you can try and sum it up in a few words.

Certainly. Well, and there's a lot of definitions for AI or artificial intelligence. I think what's taken the world largely by storm is generative AI, where in general, AI could be thought of as computers doing tasks that humans historically have been used to doing. But with generative AI, AI is basically making up new content. That said, is it really reasoning?

No. In fact, it's pattern matching. It's sort of filling in the blank or if you will predicting the next word just like if you do a Google search and it tries to finish the beginning part of it. At its core, that's a lot of what generative AI is just on a massively sophisticated level. The other half of how I'd answer is that machine learning or predictive AI is essentially curve fitting.

We're gonna use AI or machine learning methods to essentially predict what's gonna happen. And the tie in to medicine is that's what doctors do all the time. You take a bunch of clinical data and figure out what's gonna go on, but that's what AI and machine learning is also well equipped to do.

Thank you. Yeah. And doctor Yazdani, did you want to add anything or refute that definition? No. No.

I think that's great.

I think maybe the key theme here is that computers are starting to do tasks that previously only humans can do. So maybe that's the simplest version of a definition for AI. Perfect. Okay.

So in yesterday's debate, doctor Curtis was arguing the pros of AI, and doctor Yazdani was arguing the cons. So it's impossible to cover this within just the space of ten minutes, but if we can maybe give, again, people who weren't there, the top level of what you what you feel AI is bringing to the table in rheumatology.

Certainly. There's a lot of things we could talk about, and obviously people can hopefully refer to the whole debate if they care to, but, I think there's some routine things where AI is saving us a lot of time. Office workflow, office automation Yeah. Education, those are big top level things, and there's tools like and open evidence and many others where AI is helping doctors learn things or manage patient portal messages or inboxes and you know get CME in the form of you know more sophisticated lit searches than just Google's gonna help you out. So some of it speaks to the routine daily tasks that we do just as part of learning as doctors.

Perfect. Okay. And the the drawbacks. I'm sure people have many drawbacks already that they're seeing of their own, but what would you feel at top level ones?

So one of the key things that many rheumatologists might not realize is that generative AI tools are often designated as not for clinical decision making. So things like AI scribes and chart summarization tools, the intention is that doctors will read every line of output and correct things that are wrong. But this leads to a situation where we have a lot of unregulated AI tools that are entering our clinics. Yeah. And actually, the onus for monitoring those is gonna fall on rheumatologists.

So people need to be really careful about checking the output of this technology.

And if I can just stay with you, doctor Yazdani. For people who are concerned about the use of AI in their clinics or that perhaps, you know, this is a global audience, perhaps working in a place where they're not quite up to speed with AI as yet, what do you have any tips for them? What they can do? What to look out for?

Yeah. I would say that I I would really recommend that people look at every single AI tool that's being offered to them in their clinic the same way that they would a drug or a device. Review the clinical data just like you would a clinical trial. What data was this trained on? Has this model been stress tested for things like bias?

We know these models actually tend to be very biased. Make sure that you are checking the output, and you've set up even, you know, the most basic system for doing some audit and making sure that, you know, periodically, you're checking to make sure the output is is fine.

Perfect. And doctor Curtis, similar sort of question for people who are maybe not maybe a bit scared of AI. Do you have somewhere where they can go to for resources or somewhere to start for thinking of applying AI in their practice?

I think a lot of what is lacking is just practical experience with the tools that are out there. I think it will be intuitive that physicians will gravitate towards whatever plug in that the EHR vendor either recommends or offers or endorses or that's compatible. But as Doctor. Yostani already just said, we need to figure out where did this data come from, how was it trained, how accurate is it. When it's reported, it's not just something that, oh, we just plug in and assume it works, it works well, and that the security and privacy should be automatically intrinsic because it's not.

So I think we need to rigorously pressure test not just the privacy, the safety, the security, the accuracy. I think that's one component. The second component is, is this really helping us? Are there unintended consequences? Are we saving time?

Or are we making more work for ourselves? Because I think it's a two sided hypothesis.

Okay. Yeah. No. Absolutely. And just one final question, actually.

I am as part of RheumNow, I've seen that there's a lot of AI related abstracts at this year's congress, which we've covered as part of the RheumNow coverage. Would either or both of you like to share one particular abstract, maybe from your own work or from someone else's work that you'd like to highlight and draw people to related to AI?

Well, it's easiest for me to talk

about our own work. So Yep. That's a fact. One

of our wonderful postdoctoral fellows, doctor Aglundes Garcia, is starting to use large language models to extract richer information from electronic rec health records. And I actually think there's a huge amount of promise here. Things that we previously couldn't get at, like the reasons that people are switching from one medication to another, I think this is just gonna open up a tremendous amount of opportunity for those of us who are doing clinical research, it'll ask allow us to just ask more interesting questions, more depth, more accuracy. So I think that's

really exciting. Perfect. And yourself, doctor Curtis?

To to pick one, and again drawing from my own work. So we have an AI focused abstract where patients are contributing PRO data collected on a smartphone once a week. And we were able to show that AI based algorithms over twelve to sixteen weeks after RA patients started either adalimumab or upadacitinib could very accurately classify whether patients were in low disease activity like a doctor would judge it. The paradigm is an app on a smartphone with patient data could tell you how your patients are doing and as an adjunct to a telehealth visit or for people who can't or won't come back or they live hundreds of miles away or there's a global pandemic, this sort of approach, I think, could be very helpful for between visit monitoring and patient triage.

Perfect. I know for a fact that those, abstracts have been covered on RheumNow, so do go and look at the coverage and the news articles to learn more about them. And thank you very much for joining me for RheumNow today. Please do go and watch the great debate on demand. I cannot recommend it enough.

If you'd like to know more about what's going on here at ACR twenty twenty five, do keep watching RheumNow, and thank you for joining us.

I'm Anthony Chen reporting here for RheumNow at ACR twenty twenty five in Chicago. And one of the interesting abstracts that was presented today was abstract number two three three one. And this talks about something that we probably don't talk very much about, which is coping strategies for our patients in axial spondyloarthritis. And this comes from the OASIS cohort where they have analyzed patients with axial spondyloarthritis. We know that our patients with axial spondyloarthritis have physical, emotional, and also social changes that can often affect how they feel and how they respond to the treatments that we are giving them.

So they have coping strategies that to help manage chronic disease, including pain. This study looked into how these coping strategies are being utilized by the patients, and secondly, what happens over a period of time whether these patients maintain their coping strategy or whether they change strategies as time went on. Also whether these coping strategies are somewhat related to disease activity. So the data came from the OASIS study, and they look at something called CORS, which is coping with rheumatic stressors. And these evaluation which looks at three main components.

Firstly, pain, secondly, limitation, and third, dependence. It is quite a comprehensive assessment with 61 items across eight strategies, and patients then were scored between one to four. And then they evaluated this according to the demographics of the patient, including their age, sex, and education, their health scores, including ESTI, BESFI, SF thirty six, and also to long detainee follow them over a period of time. So they had 116 patients who were included in this study, and they looked at the coping strategies. The most often used coping strategy was pain.

For pain, there was comforting cognitions, And for limitations, they use pacing. And for dependence, showing consideration. For associations, they were they looked at how these things then coped how these patients cope with these many challenges. Patients who use a lot of comforting cognitions, this was linked to those who had lower educational attainment and also higher BMI. For pacing, this was used more by older individuals.

And for showing consideration, this was used less with people with more comorbidities. What was surprising was these methods or coping strategies that were being used were not necessarily direct related to the outcome measures. So it looks like the disease scores, including STES, VESFI, SF thirty six, had very minor effect on which coping strategies that the patients use. And, also, the coping strategies did not really change over the two years of of the study. So this study really showed that these coping strategies are being utilized by our patients, and this is a very comprehensive way of analyzing it.

And patients use different ways to to cope with their chronic disease. The important thing is that this is not necessarily related to disease activity, and this is down to the individual person itself. So these are personal characteristics such as age, education, body mass index, and these are thought to be more important rather than the actual disease activity that determines how our patients use these coping strategies. So this again highlights to us that when we care for our patients with XBA in the personalization of this care, we should also try to tailor these coping strategies or coping styles in order for us to manage our patients with XBA better. I'm Anthony Chan reporting here for RheumNow at ACR twenty twenty five.

Hello, everyone. Today is day three of ACR twenty twenty five, and a lot of data being presented about patients, with axial spondyloarthritis. And, I found these few abstracts which are very clinically relevant for our patients, is those patients with axial spondyloarthritis that have inadequate response to first TNF, what to do in them next. There was a prospective study that was presented from 31 centers. The study looked at whether those patients with Axial SpA have inadequate response to TNF inhibitors should switch to a different TNF inhibitor or completely switch the class like to an IL-17A inhibitor.

The study showed that the primary endpoint at twenty four weeks was ASAS40 response, and the study showed there was no significant difference between the two groups, that ASAS40 response at twenty four, weeks was similar between the IL-17A, switch or a switch to a second TNF. So the take home message from this abstract was that, you know, this, switching is an individualized decision, shared decision making between the physician and the patient to decide where to switch next patient, where to switch to the next medication. Another abstract on the same line was from a Spanish registry which also looked at switching for patients to a different TNF inhibitor or to IL-17A inhibitor, and that registry also showed that the retention and survival was much longer on patients that switched from one TNF inhibitor to another TNF inhibitor suggesting that switching maybe within the same family may be better than switching to a different class. We did a real world study using TriNetX database where we looked at initial start, what should be their first start of biologic in these patients, and whether it should be a TNF inhibitor or an IL-17A inhibitor or a JAK inhibitor. We found that the odds of survival were much higher in TNF inhibitors than in patients with IL-seventeen and JAK inhibitors.

These data need to be looked into further, but maybe giving us some idea about what to start first in these patients with axial spondyloarthritis. Thank you.